Docstoc

EPITHELIAL CELLS ATTACHMENT ON FIVE DIFFERENT DENTAL

Document Sample
EPITHELIAL CELLS ATTACHMENT ON FIVE DIFFERENT DENTAL Powered By Docstoc
					EPITHELIAL CELLS ATTACHMENT ON FIVE DIFFERENT DENTAL IMPLANT

               ABUTMENT SURFACE CANDIDATES




                                   by
                      Yves Alain Dietrich Sitbon




                A thesis submitted in partial fulfillment
                 of the requirements for the Master of
                Science degree in Operative Dentistry
                      in the Graduate College of
                        The University of Iowa


                               May 2009


             Thesis Supervisor: Professor Galen Schneider
                           UMI Number: 1464875




                          INFORMATION TO USERS


  The quality of this reproduction is dependent upon the quality of the copy
  submitted. Broken or indistinct print, colored or poor quality illustrations and
  photographs, print bleed-through, substandard margins, and improper
  alignment can adversely affect reproduction.
  In the unlikely event that the author did not send a complete manuscript
  and there are missing pages, these will be noted. Also, if unauthorized
  copyright material had to be removed, a note will indicate the deletion.




______________________________________________________________

                            UMI Microform 1464875
                        Copyright 2009 by ProQuest LLC
        All rights reserved. This microform edition is protected against
           unauthorized copying under Title 17, United States Code.
_______________________________________________________________

                               ProQuest LLC
                        789 East Eisenhower Parkway
                               P.O. Box 1346
                          Ann Arbor, MI 48106-1346
                                      Graduate College
                                    The University of Iowa
                                       Iowa City, Iowa




                              CERTIFICATE OF APPROVAL

                                 _______________________


                                     MASTER'S THESIS

                                      _______________


This is to certify that the Master's thesis of


                                 Yves Alain Dietrich Sitbon


has been approved by the Examining Committee
for the thesis requirement for the Master of Science
degree in Operative Dentistry at the May 2009 graduation.


Thesis Committee: ____________________________________________
                  Galen Schneider, Thesis Supervisor


                    ____________________________________________
                    Deborah Cobb


                    ____________________________________________
                    Deborah Dawson


                    ____________________________________________
                    Gerald Denehy


                    ____________________________________________
                    Georgia Johnson


                    ____________________________________________
                    Clark Stanford
To my wife Marie-Ève, for her permanent love and support.




                               ii
                              ACKNOWLEDGMENTS



        I would like to express all my gratitude for my research mentor and advisor Dr

Galen Schneider, who not only offered me to work on this subject and accepted me in his

laboratory, but also took time to guide and help me in a subject mostly unknown by me at

the time we started. Precision, concision, availability, and efficiency, are only a few of the

professional qualities he revealed during his mentoring, outnumbered only by his human

qualities.

        I would like to express all my gratitude to Dr Cobb who accepted me as a graduate

student in the very intense and valuable program she directs. The flexibility she offered

me allowed me to get the information and knowledge I needed the most, disregarding the

trouble that these special arrangements could create for her. Her support during these last

years has been a permanent source of energy and comfort.

        My gratitude also goes to Dr Denehy for his unconditional teaching generosity.

Not only did I learn a lot in clinical matters, but also in teaching skills and human

relationships. I wish I could become half-as-good a teacher for my own students. Thanks

for this very inspirational experience.

        Very special thanks for Dr Dawson who taught me every thing I know in

statistics, even though I realize that she just had time to barely scratch the surface of the

enormous amount of knowledge she could share. Because of my limitations in this

difficult but fascinating field, I was lucky enough to have her accept to perform and write

the whole statistical analysis of this manuscript.

        Finally, I would like to express my gratitude to Dr Johnson and Dr Stanford for

accepting to be part of the committee for this thesis, and for their guidance. I would also

like to thank Dr Aloysius J. Klingelhutz for providing the hTERT cells.




                                              iii
                                         TABLE OF CONTENTS




LIST OF TABLES............................................................................................................. vi

LIST OF FIGURES ......................................................................................................... viii

CHAPTER 1 INTRODUCTION ..................................................................................... 1
        1.1 Purpose ...................................................................................................... 4
        1.2 Hypothesis ................................................................................................. 4
        1.3 Aim 1 ......................................................................................................... 4
        1.4 Aim 2 ......................................................................................................... 4
        1.5 Aim 3 ......................................................................................................... 5
        1.6 Aim 4 ......................................................................................................... 5
        1.7 Objective.................................................................................................... 5

CHAPTER 2 BACKGROUND INFORMATION AND REVIEW OF THE
       LITERATURE................................................................................................. 6
       2.1 Brief history of implantology .................................................................... 6
       2.2 The implant-abutment complex: “anatomy” and terminology .................. 7
            2.2.1 The implant body............................................................................. 8
            2.2.2 The abutment ................................................................................. 11
       2.3 Cell and cytoskeleton............................................................................... 17
       2.4 Cellular junctions..................................................................................... 18
       2.5 Influence of the attachment phenomenon on cell biology....................... 22
       2.6 Role of the acquired exogenous pellicle .................................................. 23
       2.7 Cytokines ................................................................................................. 24
       2.8 Biological tissues in contact with the implant-abutment complex .......... 26
            2.8.1 Bone tissue..................................................................................... 27
            2.8.2 Oral mucosa................................................................................... 32
            2.8.3 Blood vessels and nerves............................................................... 37
       2.9 Anatomy of the soft tissue/tooth interface............................................... 37
       2.10 Anatomy of the soft tissue/implant interface......................................... 39
       2.11 Importance of the soft tissue cuff .......................................................... 43
       2.12 Influence of material and characteristics of the abutment surface ........ 44
            2.12.1 Materials for dental implant abutments ....................................... 45
            2.12.2 Impact of the nature of the substrate on cellular adhesion .......... 50
            2.12.3 Impact of the surface topography and chemistry on cellular
            adhesion .................................................................................................. 56
       2.13 Outcome of the review of the literature ................................................. 67

CHAPTER 3 MATERIALS AND METHODS ............................................................ 69
        3.1 Objective of the study.............................................................................. 69
            3.1.1 Aim I.............................................................................................. 69
            3.1.2 Aim II ............................................................................................ 70
            3.1.3 Aim III ........................................................................................... 70
            3.1.4 Aim IV........................................................................................... 71


                                                             iv
                 3.2 Background information on the investigation techniques ....................... 72
                      3.2.1 Cell culture .................................................................................... 72
                      3.2.2 Attachment assay........................................................................... 73
                      3.2.3 Scanning electron microscopy....................................................... 76
                      3.2.4 Immunofluorescence ..................................................................... 76
                      3.2.5 Polymerase chain reaction and electrophoresis ............................. 78
                      3.2.6 Cytokine analysis........................................................................... 82
                 3.3 Specimens preparation............................................................................. 86
                      3.3.1     Titanium and Zirconia Specimen Preparation.......................... 86
                      3.3.2     Cell culture ............................................................................... 88
                      3.3.3 Cells passaging .............................................................................. 88
                 3.4 Cell attachment assays............................................................................. 90
                 3.5 Cytokines-chemokines assay ................................................................... 91
                 3.6 Scanning Electronic Microscopy ............................................................. 92
                 3.7 Immunofluorescence................................................................................ 93
                 3.8 Polymerase Chain Reaction (PCR).......................................................... 93
                      3.8.1 RNA extraction.............................................................................. 93
                      3.8.2 Primers sequences ......................................................................... 94
                      3.8.3 Reverse transcription ..................................................................... 94
                      3.8.4 Electrophoresis .............................................................................. 95
                 3.9 Recapitulation of the repartition of the discs in the different groups
                 (aims I to III).................................................................................................. 95
                 3.10 Proliferation tests ................................................................................... 97
                 3.11 Statistical method................................................................................... 98

CHAPTER 4 RESULTS .............................................................................................. 100
        4.1 Attachment assays ................................................................................. 100
             4.1.1 Statistical Methods ...................................................................... 100
             4.1.2 Results ......................................................................................... 101
        4.2 Cytokines levels measurements............................................................. 102
             4.2.1 Statistical Methods ...................................................................... 102
             4.2.2 Results ......................................................................................... 103
        4.3 Proliferation assays................................................................................ 112
             4.3.1 Statistical methods....................................................................... 112
             4.3.2 Results ......................................................................................... 113
        4.4 Integrin 6 4 expression ....................................................................... 118
        4.5 Immunofluorescence assays .................................................................. 119
        4.6 Scanning Electrons Microscopy assays ................................................. 125

CHAPTER 5 DISCUSSION AND CONCLUSION ................................................... 130
        5.1 Discussion.............................................................................................. 130
        5.2 Conclusion ............................................................................................. 138

APPENDIX..................................................................................................................... 139

REFERENCES ............................................................................................................... 148




                                                                 v
                                                LIST OF TABLES



Table 1: Least squares means and their estimated standard errors for the attachment
         assay.................................................................................................................. 101

Table 2: Least squares means in pg/mL and their estimated standard errors for Il-6
         expression. ........................................................................................................ 105

Table 3: Least squares means in pg/mL and their estimated standard errors for IL-6
         expression following square root transformation. ............................................ 106

Table 4: Least squares means in pg/mL and their estimated standard errors for IL-6
         expression following square root transformation and outlier removal. ............ 106

Table 5: Least squares means in pg/mL and their estimated standard errors based on
         the full data set.................................................................................................. 108

Table 6: Least squares means in pg/mL and their estimated standard errors after
        removal of two putative outliers. ...................................................................... 108

Table 7: Least squares means in pg/mL and their estimated standard errors for IL-
         12 expression. ................................................................................................... 110

Table 8: Median, minimum and maximum of the TNF- levels within each surface
         treatment (pooling over all runs). ..................................................................... 111

Table 9: Comparisons of proliferation levels among 6 surfaces at day 3....................... 113

Table 10: Comparisons of proliferation levels among 6 surfaces at day 5..................... 114

Table 11: Comparisons of proliferation levels among 6 surfaces at day 7..................... 114

Table 12: Comparisons of proliferation levels among 3 days for TCP surface.............. 115

Table 13: Comparisons of proliferation levels among 3 days for AA surface. .............. 115

Table 14: Comparisons of proliferation levels among 3 days for BB surface................ 115

Table 15: Comparisons of proliferation levels among 3 days for CC surface................ 115

Table 16: Comparisons of proliferation levels among 3 days for DD surface. .............. 116

Table 17: Comparisons of proliferation levels among 3 days for EE surface. ............... 116

Table A 1: Batch YS08-2 data for attachment assay. ..................................................... 139

Table A 2: Batch YS08-3 data for attachment assay. ..................................................... 140

Table A 3: Batch YS08-4 data for attachment assay. ..................................................... 141



                                                                vi
Table A 4: Batch YS08-5 data for attachment assay. ..................................................... 142

Table A 5: Batch YS08-6 data for attachment assay. ..................................................... 143

Table A 6: Raw results for cytokines levels measurements ........................................... 144




                                                    vii
                                             LIST OF FIGURES



Figure 1: Root-form implant............................................................................................... 9

Figure 2: Single piece abutment-implant complex. ............................................................ 9

Figure 3: Implant body...................................................................................................... 10

Figure 4: Implant Straumann with smooth transmucosal crest module............................ 10

Figure 5: Different abutment types. .................................................................................. 11

Figure 6: Transmucosal neck of a Straumann implant. .................................................... 12

Figure 7: Implant- abutment-prosthesis complex. ............................................................ 12

Figure 8: Implant with external hexagon. ......................................................................... 14

Figure 9: Implant with internal octagon............................................................................ 15

Figure 10: Abutments for internal connection. ................................................................. 16

Figure 11: Solid abutment................................................................................................. 16

Figure 12: Schematic representation of a cell-cell adhesion. ........................................... 19

Figure 13: Schematic representation of a desmosomes. ................................................... 19

Figure 14: Schematic representation of focal adhesion. ................................................... 21

Figure 15: Schematic representation of a hemidesmosomes. ........................................... 21

Figure 16: Main tissue component of the oral mucosa. .................................................... 34

Figure 17: The dentogingival junction.............................................................................. 38

Figure 18: Difference between tooth- and implant-soft tissue interfaces. ........................ 40

Figure 19: TBR Z1             implant with zirconia transmucosal neck....................................... 49

Figure 20: Schematic representation of the particle counter. ........................................... 75

Figure 21: Schematic representation of a scanning electron microscope. ........................ 77

Figure 22: Polymerase chain reaction; cycle 1. ................................................................ 79

Figure 23: Polymerase chain reaction; cycle 2. ................................................................ 79

Figure 24: Polymerase chain reaction; cycle 3. ................................................................ 80



                                                             viii
Figure 25: Schematic drawing of the electrophoresis process.......................................... 82

Figure 26: Different combination of red dye for the beads. ............................................. 83

Figure 27: Different steps of a multiplex fluorescent beads-based immuno-assay. ......... 84

Figure 28: Laser measures in a multiplex fluorescent beads-based immuno-assay. ........ 85

Figure 29: Groups coding. ................................................................................................ 87

Figure 30: Constitution of the different batches (Aims 1 to 3)......................................... 89

Figure 31: Primers sequences for PCR. ............................................................................ 95

Figure 32: Samples repartition; aim I – cell attachment assay. ........................................ 96

Figure 33: Samples repartition; aim III – cytokines-chemokines assay. .......................... 96

Figure 34: Samples repartition; aim I - SEM and IF assays and aim II – PCR. ............... 97

Figure 35: Combined results for the five batches for each of the five surfaces for the
           attachment assay. ......................................................................................... 102

Figure 36: Average level of IL-1B for each of the five surfaces. ................................... 104

Figure 37: Average level of IL-6 for each of the five surfaces....................................... 107

Figure 38: Average level of IL-8 for each of the five surfaces...................................... 109

Figure 39: Average level of IL-12 for each of the five surfaces..................................... 110

Figure 40: Average level of TNF- for each of the five surfaces................................... 112

Figure 41: Comparisons of proliferation levels for six surfaces at days 3, 5, and 7.
           Means joined by a bar are not significantly different. ................................. 117

Figure 42: Comparisons of proliferation levels for days 3, 5, and 7 at each surface.
           Means joined by a bar are not significantly different. ................................. 117

Figure 43: Result of electrophoresis. .............................................................................. 118

Figure 44: Disc AA, x20................................................................................................. 120

Figure 45: Disc BB, x20. ................................................................................................ 120

Figure 46: Disc CC, x20. ................................................................................................ 120

Figure 47: Disc DD, x20................................................................................................. 121

Figure 48: Disc EE, x2o.................................................................................................. 121

Figure 49: Disc AA, x400............................................................................................... 122

Figure 50: Disc BB, x400. .............................................................................................. 122


                                                             ix
Figure 51: Disc CC, x400. .............................................................................................. 122

Figure 52: Disc DD, x400............................................................................................... 123

Figure 53: Disc EE, x400................................................................................................ 123

Figure 54: Disc AA, x600............................................................................................... 124

Figure 55: Disc BB, x600. .............................................................................................. 124

Figure 56: Disc CC, x60. ................................................................................................ 124

Figure 57: Disc DD, x600............................................................................................... 125

Figure 58: Disc EE, x600................................................................................................ 125

Figure 59: Disc AA, x1500............................................................................................. 126

Figure 60: Disc AA, x4500............................................................................................. 127

Figure 61: Disc AA, x20000........................................................................................... 127

Figure 62: Disc BB, x3500. ............................................................................................ 127

Figure 63: Disc CC, x3500. ............................................................................................ 128

Figure 64: Disc CC, x3500. ............................................................................................ 128

Figure 65: Disc DD, x3500............................................................................................. 128

Figure 66: Disc EE, x3500.............................................................................................. 129

Figure 67: Disc EE, x10000............................................................................................ 129




                                                             x
                                                                                           1


                                      CHAPTER 1
                                  INTRODUCTION



       Dental caries, as well as breakdown of the supportive tissues around the teeth

(periodontal disease), have been identified as two of the most common chronic illnesses.

They both can lead to loss of teeth. Classically, prosthetic appliances, whether fixed or

removable, are then used to replace these missing teeth. These treatment modalities are

not without shortcomings; fixed conventional prostheses require heavy sound tooth

structure removal, and removable prostheses are sometimes rather uncomfortable. Since

Brånemark’s first observation of the biocompatibility of titanium and its property to be

well accepted by bone, a phenomenon called osseointegration (Albrektsson, Branemark,

Hansson, & Lindstrom, 1981), a new treatment modality has emerged. It consists in

placing titanium screws, called implants, into the maxillary and/or mandibular bone to

replace the roots of the missing teeth. An intermediate piece (the abutment) is then place

onto the implant; it traverses the soft tissue barrier and provides support and retention for

the prosthesis, increasing dramatically the comfort of the patient while preserving his or

her remaining teeth, if any.

       Implant-supported prostheses have now become widely accepted as a safe and

predictable alternative to conventional prosthesis.     More over, due to the important

progress made in the understanding of the biological, biomechanical, and biochemical

principles that determine the clinical success of these implants, they have become the

standard of care in many clinical situations.

       If historically, osseointegration has been the major topic of interest in the implant

research area, the last decades have witnessed a parallel shift of the attention of

researchers toward the relationship between the abutment and the soft tissue around it.

By analogy with the osseointegration, this relationship is often referred to as “soft tissue
                                                                                            2


integration”. This concept is presently acknowledged as an important factor in implant

survival. Many authors have described both an epithelial attachment and a connective

tissue cuff around the abutment or the implant neck, relatively similar to the system

surrounding natural teeth (Bauman, Rapley, Hallmon, & Mills, 1993). This apparatus, as

for the natural teeth, seems to create a seal that prevents bacterial invasion, the latter

being known as a major disruptive factor of the implant – host interface, and therefore a

potential threat for the survival of the implant.

       The epithelium component of this attachment apparatus is the first barrier bacteria,

or any other antigen, would encounter and hence, is an important protection.               Its

connection to the abutment or implant neck also allows the soft tissues surrounding the

implant to keep a good spatial relationship with this implant, and for that reason, might

play an important role in the esthetic integration of the prosthetic work (Al-Sabbagh,

2006; Leblebicioglu, Rawal, & Mariotti, 2007; Lee & Jun, 2000; Wheeler, Vogel, &

Casellini, 2000). Finally, the epithelial cells, as well as the connective cells of this soft

tissue barrier, seem to play a role in the osseointegration process itself. Many growth

factors, such as morphogenic proteins and cytokines, are secreted by these cells and

probably have a direct influence on the bone histo-physiology.

       The first step a cell has to take to be able to differentiate and/or to express its
potential is often adhesion on a substrate, whatever this one might be. This adhesion, or

attachment of the cell occurs through specific structures, such as hemidesmosomes and

focal adhesions.     Specific protein structures (receptors) such as Integrins in focal

adhesions    (Chrzanowska-Wodnicka         &        Burridge,   1996),   and   Cadherins   in

hemidesmosomes (Kinch, Petch, Zhong, & Burridge, 1997), allow this attachment.

       The chemical nature and the topography of the substrate will considerably

influence the nature of the first layer adsorbed on its surface (biofilm), as well as the

following adhesion, proliferation, orientation, differentiation and expression of the cells,

whether epithelial or connective (Schaller, 2001; Schlaepfer, Hauck, & Sieg, 1999).
                                                                                            3


Different treatments for different biomaterials have been proposed to enhance the

properties of the surface of these biomaterials to promote cellular adhesion. Most often,

titanium is the material used for dental implants and abutments, and is therefore the most

widely studied. Coating with hydroxyapatite (HA), plasma-sprayed titanium oxides, or

acid etching are a few common procedures among many attempts made to try to improve

this material surface properties. Increasing the roughness is thought to be one of the

simplest factors to promote cellular adhesion, but this is cell-dependant. For example,

this has constantly been described for osteoblasts, but often contradicted for fibroblasts

(Grossner-Schreiber et al., 2006). As for the epithelial cells, many contradicting studies

were published, sometime presenting roughness as favorable to adhesion, sometimes as

inhibitory (Abrahamsson et al., 2002). More research is needed to clarify this.

       New materials have also raised a strong interest in this field of research, especially

zirconia. This ceramic material, well known for its biocompatibility and its mechanical

properties, has been extensively used in orthopedic surgery for total hip arthroplasty. Its

white color makes it particularly attractive when it comes to replacement of lost tooth

structures. Zirconia abutments have been proposed for many years to dental practitioners,

with the manufacturers advocating the esthetic, biomechanical and biological qualities of

the material.   The esthetic advantages are undeniable; the material avoids the gray
shadowing of the surrounding soft tissues, often seen with titanium abutments.

Moreover, the use of zirconia abutments combined with full ceramic crowns permits a

greater light transmission through the prosthetic element, and therefore enhances esthetic.

However, if zirconia has proven to be a highly biocompatible material, its effect in terms

of cellular adhesion, especially epithelial cells adhesion, is still not clear and needs to be

investigated.
                                                                                           4


                                        1.1 Purpose



       The purpose of this study is to determine if epithelial cells adhere and differentiate

differently to various surface micro- and nano-topographies, presented on titanium and

zirconium transmucosal abutment materials.              Specifically, the materials being

investigated are zirconia, machined titanium (control), titanium treated with hydrofluoric

acid, and two types of modified titanium.


                                       1.2 Hypothesis



       Our hypothesis is that these different implant abutment substrates will affect

epithelial cell attachment and differentiation differently.

       We have specific aims for this research, in order to be able to validate or infirm

our hypothesis:


                                          1.3 Aim 1



       Analyze differences in epithelial cell attachment and spreading on the different

substrates using cell attachment, immunofluorescent (IF) and scanning electron

microscopy (SEM) assays.


                                          1.4 Aim 2



       Assess 6 4 integrin expression in epithelial cells culture on different abutment

substrates.
                                                                                       5


                                           1.5 Aim 3



       Assess pro-inflammatory cytokines expression in epithelial cells culture on

different abutment substrates.


                                           1.6 Aim 4



       Analyze proliferation for epithelial cells culture on the five different abutment

substrates at 3 days, 5 days and 7 days.


                                       1.7 Objective



       The results of this study could lead to a better understanding on the influence of

the nature of the abutment material, and its surface topography, on epithelial cells

attachment, and their subsequent behavior. In turn, this knowledge could help designing

abutments with more desirable clinical features, such as surface favoring a strong

epithelial cell attachment. Finally, assessing the reaction of epithelial cells toward a

zirconium surface could provide evidence supporting the use of this more esthetic

material.
                                                                                        6


                                    CHAPTER 2
                   BACKGROUND INFORMATION AND
                       REVIEW OF THE LITERATURE



                           2.1 Brief history of implantology



       The idea of replacing teeth by implants is by no means recent. 4000 years ago,

ancient Chinese carved bamboo sticks the shape of pegs and drove them into the bone for

fixed tooth replacement. 2000 years ago, Egyptian used a similar method with precious

metal (Misch, 2005). Archeologists found evidence that already a thousand years ago,

occupants of what is now Honduras, developed a way to use tooth-shaped stones as dental

implants (Sullivan, 2001). In Europe, references to an implant in modern literature begins

in France, in 1809, and by the late 1800s, dentists on both sides of the Atlantic were

experimenting implants made of such things as extracted teeth and lead. An important

breakthrough was done in 1941 when Gustav Dahl, a Swedish doctor, placed a metal

structure below the periosteum, with vertical extensions protruding through the gingival.

Another advance came with the work of Leonard Linkow, who in 1946, introduced a self-

tapping titanium implant, and later created a blade implant that became the most widely

used implant design in the 1970s (Sullivan, 2001).

       In the 50s, Per-Ingvar Brånemark discovered accidentally that titanium appeared

uniquely compatible with bone in a rabbit. It took him approximately 10 years to get

struck by the clinical importance of his finding, when in the early 60s, he found that the

biocompatibility of titanium extended to the soft tissues and skin of human subjects

(Sullivan, 2001). He placed his first clinical implant in 1965. In the following 5 years,

his clinical results were unacceptably poor, with a failure rate of 50% (Albrektsson &

Wennerberg, 2005). It took until 1976 to have full recognition of this treatment modality
                                                                                        7


by the Swedish national health system, and another 5 years for this revolution to hit the

North American continent, mostly under the impulse of George Zarb, a dentist professor

at the University of Toronto, Canada (Sullivan, 2001). Along with 4 colleagues of the

same university, he went to Sweden in the late 1970s to get training in the Brånemark

method of implant placement. This training permitted a replica study in Canada, which

was to become a landmark since it was the first other location than Göteborg to confirm

the concept of osseointegration, developing at the same time one of the world’s most

important implant databank.      It is Zarb again that took the initiative to widen the

indications for osseointegrated oral implants, until then almost exclusively used on

completely edentulous patients.     He published clinical results in partly edentulous

patients, leading to acceptance of osseointegrated implants as a treatment modality for

these cases (Zarb, Zarb, & Schmitt, 1987; Zarb & Schmitt, 1993a; Zarb & Schmitt,

1993b).

        Oral implants have developed rapidly since then. After a focus on function,

improvements have shifted attention towards esthetic and simplicity of use. More or less

immediate loading has replaced delayed placement of the supra-structure; bone grafting is

now commonly use to provide the necessary bone volume, expanding the possibilities.

Tissue engineering is emerging and promises the opening of new horizons. However,
new implant designs and surfaces are launched at an increasing pace, sometimes without

any attempts at previous clinical testing, which start after the sales, not always for the

patient’s benefit.


                     2.2 The implant-abutment complex: “anatomy” and
                                       terminology



        Dental implants can be endosteal, subperiosteal or intra-mucosal.      The most

common form presently used is the root-form implant (figure 1), which is an endosteal
                                                                                             8


implant. The word endosteal is derived from endo, which means “within”, and from

osteal which means “bone”.        Most of the time, the implant is a complex structure

composed of 2 or more parts. Only a few numbers of implants, such as Nobel Biocare

Direct , have been designed as a single-piece system (figure 2). For two-piece (or more)

systems, one can schematically distinguished the implant body and the transmucosal part.


                                  2.2.1 The implant body



       The implant body is the part of the implant that will be placed in the bone and will

often slightly extend above the bone crest. It is divided in a crest module, the body itself,

and an apex (figure 3).

       The body can be a solid screw or bullet shaped. The solid screw is the most

frequent design, for it presents a better initial stability (a key for success) and offers more

bone-implant contact surface. Cylindrical or bullet-shaped implants are sometimes easier

to place, especially in soft bone, but necessitate a bioactive surface to compensate for a

reduced bone-implant contact area (Misch, 2005).           The crest module is the portion

designed to retain the prosthetic component and is so named because it is normally

located at the crest of the ridge (figure 3). In some implant designs by manufacturers
such as ITI, this module can also be the transmucosal part of the implant (figure 4). The

crest module is often smoother than the body, in order to impair plaque retention if crestal

tissue loss occurs (Misch, 2005). Even though a few authors could not demonstrate a

significant correlation between plaque accumulation and roughness of the transmucosal

part (Wennerberg, Sennerby, Kultje, & Lekholm, 2003), it has been reported quite

constantly that increased roughness leads to increased plaque accumulation (Quirynen et

al., 1993). On the other hand, it seems that polishing a surface under a threshold Ra of

0.2 m does not improve resistance to plaque adhesion and might actually decrease soft

tissue attachment on the abutment (Bollen et al., 1996).
                                                         9




Figure 1: Root-form implant.


Source: Rose L.F.; 2004, Periodontics, Elsevier Mosby.




Figure 2: Single piece abutment-implant complex.


Source: Nobel Biocare   Website.
                                                                        10




Figure 3: Implant body.


Source: Misch, C.E.; 2005, Dental Implant Prosthetic, Elsevier Mosby.




Figure 4: Implant Straumann with smooth transmucosal crest module.


Source: Straumann Website.
                                                                                     11


                                   2.2.2 The abutment



       Regardless of the type of dental implant, a transmucosal component is necessary

to provide support for the prosthetic work. For root implant, this component can present

a large variety of shapes, designs, and concepts (figure 5). We mentioned one-piece

systems, but most of the time, the implant system is a complex of two or more pieces.

The transmucosal part can simply be an extension of the crest module, as for certain

implants of the ITI system (figure 6), or can be a separated component, fixed on the

implant body by different means (screw, cement, “cold soldering”, “snapped”…), (figure

7). This transmucosal part can sometimes, in turn, provide support even to a third piece

before the prosthetic work, either fixed or removable, is finally placed on it.     For

simplification, we will call the transmucosal part “abutment”, even though the

terminology is not rigorously correct.




Figure 5: Different abutment types.


Source: Straumann Website.
                                                                        12




Figure 6: Transmucosal neck of a Straumann implant.


Source: Straumann Website.




Figure 7: Implant- abutment-prosthesis complex.


Source: Misch, C.E.; 2005, Dental Implant Prosthetic, Elsevier Mosby.
                                                                                         13


       The different designs allow two different protocols for the placement of the

implant:

        A traditional 2-stage surgical protocol, where the implant body is first submerged

under the soft tissues, (carefully closed by sutures) and exposed during a second surgery,

during which the transmucosal part is added.

       A 1-stage surgical protocol, where the implant body and the transmucosal part,

whether single-piece or two-components, are placed at the same time, and the

transmucosal part extend directly in the oral cavity, above the soft tissues. In this case,

the practitioner can decide to follow an immediate-loading protocol, or choose to differ

the loading, depending on the clinical situation.

       Some authors have reported a better stability of the soft tissue cuff around the

abutment when a two-stage technique is used. The reason could be that connective tissue

would have more time to attach to the structure before it has to compete for its surface

with the epithelial cells (Chehroudi, Gould, & Brunette, 1992). On the other hand, the

presence of a micro-gap between the implant and the abutment at the soft tissue level has

been shown to induce inflammation at that level (Abrahamsson, Berglundh, Glantz, &

Lindhe, 1998; Moon, Berglundh, Abrahamsson, Linder, & Lindhe, 1999). In fact, it is

now generally considered that there are no significant differences in soft tissue health and
structure between these 2 protocols. For example, Abrahamsson et al., in an experiment

on dogs, placed implant/abutment complexes following each protocol: on each dog, 3

implants were placed and submerged, receiving their abutment three months later (control

group), at which time another batch of three implants were placed on the contro-lateral

side, but receiving immediately their abutments (test group). Soft tissue health and bone

level were clinically and radiographically assess over another 6-month period, during

which strict oral hygiene was implemented. After these 6 months, dogs were sacrificed

and histometric and morphometric analysis were performed. No significant difference
                                                                                      14


could be observed between the two groups (Abrahamsson, Berglundh, Moon, & Lindhe,

1999).


2.2.2.1 Internal versus external connection



         The abutment, if not a simple extension of the implant body, can be fixed to the

latter by different means. A major distinction is internal versus external connection. In

systems with external connection, the crest module is generally a platform on which an

anti-rotational feature, most often a hexagon of limited height, protrudes (figure 8). A

corresponding female part at the base of the transmucosal piece is then placed over this

male part. The transmucosal piece is secured onto the platform by an independent screw.




Figure 8: Implant with external hexagon.


Source: Nobel Biocare     Website.
                                                                                         15


       Systems with internal connection can present a great variety of design. They all

have in common a hollow zone at the crestal end of the implant. This cavity can be

hexagonal or octagonal (figure 9), but also conical, cylindrical, triangular…A male part

of the proper design, emerging from the base of the abutment (figure 10), is inserted into

this cavity. The abutment can then be secured to the implant body by an independent

screw, as it is the case for an external connection, or can be cemented, or forced in place,

or rely on metal-to-metal friction and “cold soldering” for its retention. Sometimes, the

female part of the implant body is threaded and the male part of the abutment is a screw

emerging from the base of the abutment itself, which is then designed as a “solid”

abutment (figure 11).




Figure 9: Implant with internal octagon.


Source: TBR     Group Website.
                                                16




Figure 10: Abutments for internal connection.


Source: Straumann Website.




Figure 11: Solid abutment.


Source: Straumann Website.
                                                                                          17


                                2.3 Cell and cytoskeleton



       It is clearly beyond the scope of this report to describe in depth the very complex

biology of the cell in relation with its cytoskeleton. Nevertheless, a very brief overview

of some characteristics of this cytoskeleton will permit a better understanding of cellular

adhesion and parts of the experiments described subsequently.

       The cytoskeleton of a cell is a scaffold contained in the cytoplasm of all eukaryote

cells. This dynamic structure participates in different functions such as cell protection,

cell motion, maintaining the shape of the cell, intracellular transport as well as cellular

division. Three types of filaments, categorized by their size, constitute the cytoskeleton.

These three categories are: microfilaments, intermediate filaments and microtubules (Cate

& Nanci, 2008).

       The first category, the microfilaments, features actin monomers, grouped in

polymers, as its main constituent. The monomers are globular actin, called G-actin.

These monomers assemble in linear polymers called F-actin, themselves associated in

pairs to form a double helix protein creating the microfilament of approximately 7nm

diameter and a loop of the helix repeating every 37nm. Actin plays an important role in

muscle contraction as well as intra-cellular cargo transport, but these functions are not of

interest here. Actin microfilaments are concentrated just beneath the cell membrane and

are the most dynamic of the cytoskeleton filaments. They participate actively in cell

motility and mechanical support to the cell. A major function is also the network actin

filaments create with surrounding elements and their participation in some cell-cell and

cell-matrix junctions and hence their role in signal transduction (Cate & Nanci, 2008).
                                                                                           18


                                  2.4 Cellular junctions



       Specialized structures, called cellular junctions, are formed in cell-cell or cell-

extra-cellular matrix (ECM) contacts.      Three categories of cellular junctions can be

described: occluding, adhesive and communicating junctions. Only adhesive junctions

will be briefly described here.      One can distinguish cell-to-cell and cell-to-matrix

junctions. Cell-to-cell junctions comprise zonula adherens and macula adherens, also

known as desmosomes. The difference between these two junctions is their extension,

zonula encircling more or less completely the cell, and macula being more circumscribed.

Cell-to-matrix junctions comprise focal adhesions and hemidesmosomes.               Adhesive

junctions hold cells together or anchor them on the ECM (Cate & Nanci, 2008).

        At a molecular level, three components are implied: a transmembrane protein, a

cytoplasmic adapter and a cytoskeletal filament, these elements differing with the type of

junction.   For cell-to-cell junction, the transmembrane proteins are members of the

cadherin family, and are calcium-dependant proteins interacting with their analog in the

other cell. The cytoplasmic adapter is catenin, which interacts with the cytoplasmic

portion of the transmembrane cadherin. In the case of epithelial cells, the cadherin is E-

cadherin, the catenin are   and   catenin, and the cytoskeletal filament is actin (figure 12).

Catenins and actin filaments are concentrated on the cytoplasmic side of the cell

membrane and form a web structure. For desmosomes, the transmembrane proteins are

desmoglein and desmocollin, the catenins are desmoplakin and plakoglobin, and the

cytoskeletal elements are intermediate filaments (figure 13).             Desmoplakin and

plakoglobin form a dense plaque in the cytoplasm, and intermediate filaments attach to it

(Cate & Nanci, 2008).
                                                               19




Figure 12: Schematic representation of a cell-cell adhesion.


Source: Cate, T.; 2008, Oral histology, Elsevier Mosby.




Figure 13: Schematic representation of a desmosomes.


Source: Cate, T.; 2008, Oral histology, Elsevier Mosby.
                                                                                           20


         Cell-to-matrix junctions have a similar structure, but present different proteins and

anchor the cells to the ECM. In focal adhesion, the transmembrane protein is an integrin.

Integrins are heterodimers constituted by two subunits,          and    , with specificity to

varying ECM molecules. More than 20 combinations of different             and    subunits are

known. The cytoplasmic adapters are actinin, vinculin and talin, linking the integrin to

actin filaments (figure 14). Binding of the integrin to extracellular proteins permits

attachment of the cell and 2-way intracellular signaling (Cate & Nanci, 2008).             In

hemidesmosomes, the transmembrane protein is 6 4 integrin, which binds specifically

to laminin and collagen XVII. The complex 6 4 integrin/laminin-5 is believed to be the

conduit for hemidesmosomes-mediated cell signaling (Jones, Hopkinson, & Goldfinger,

1998).    The cytoplasmic adapters are BP230 and plectin, forming a dense plaque

providing anchorage for intermediate filaments (figure 15). Attachment of the cell with

these hemidesmosomes can be on the basal lamina or to molecules of the ECM through

other proteins (Cate & Nanci, 2008).

         The expression and organization of focal contacts and hemidesmosomes could

indicate the efficiency of cell adhesion (Lodish et al., 2000). In culture, most cells adhere

to their substrate through focal adhesion contact. Focal contacts are sites of mechanical

attachment to the extra-cellular matrix. They are points at which adhesion-related signal
transduction is initiated (Burridge, Fath, Kelly, Nuckolls, & Turner, 1988; Burridge &

Fath, 1989). It has been shown that the number of focal adhesion in attached cells can

vary with the roughness of the substrate. In their experiment, Größner et al. found that

this number was increased for smoother surfaces (Grossner-Schreiber et al., 2006). The

effect of the topography will however depend on the type of cells attached on the surface

(Hamilton, Chehroudi, & Brunette, 2007).
                                                                                  21




Figure 14: Schematic representation of focal adhesion.


Source: Schneider, G.; 2007, in bone pathophysiology class, University of Iowa.




Figure 15: Schematic representation of a hemidesmosomes.


Source: Cate, T.; 2008, Oral histology, Elsevier Mosby.
                                                                                         22


                   2.5 Influence of the attachment phenomenon on cell

                                          biology



       The cellular junctions described above participate directly on cell adhesion. They

are the very medium used by the cells to adhere to each another or to a substrate. For

example, perturbation in hemi-desmosomes integrity can lead to skin diseases such as

bullous pemphigoid or epidermolysis bullosa (Jones et al., 1998). There are an infinite

number of examples of the importance of cellular adhesion. Sometimes, this adhesion is

necessary, such as adhesion of the platelets on the endothelial lining of blood vessels, or

cell-to-cell adhesion of an epithelium.        In other cases, adhesion is a negative

phenomenon, such as adhesion of bacteria on artificial valves, or adhesion of dental

plaque on teeth. In the case of implant/abutment complexes, adhesion is fundamental for

osteoblasts on the implant body, fibroblasts and epithelial cells on the abutment. On the

contrary, adhesion of bacteria on those abutments can lead to pathological processes and

should be avoid.

       The functional activity of cells in contact with a surface is expected to be

influenced by that surface (Balasubramanian, Hall, Shivashankar, Slack, & Turitto, 1998;

Boyan, Hummert, Dean, & Schwartz, 1996). Cellular adhesion, by allowing inside-out

and outside-in signaling phenomenon, is a necessary step for many cellular activities,

such as differentiation, proliferation and cell expression (Jones et al., 1998; Miyamoto et

al., 1995). Adhesion of the cell on a substrate will determine its fate (Damsky & Ilic,

2002; Schaller, 2001). It will allow the cell to spread. This spreading will in turn, modify

the shape of the cell. There is abundant evidence that cell shape can regulate cell growth,

gene expression, secretion of proteinase, and ECM metabolism. Cell shape can in fact

govern whether individual cells grow or die, regardless of the type of protein that is used

to mediate adhesion (Chou, Firth, Uitto, & Brunette, 1995; Folkman & Moscona, 1978;
                                                                                         23


Hong & Brunette, 1987; Werb, Hembry, Murphy, & Aggeler, 1986). In 1999, Brunette

and Chehroudi, in a review article (Brunette & Chehroudi, 1999), stated: “it is abundantly

clear that surface topography affects cell shape and thus would be expected to modify

gene activity. Unfortunately, the effect on cell shape is complex, and little is known on

the effect of topography on gene expression. For example, culture of human fibroblasts

in vitro on machined grooved surfaces alters cell shape and regulates mRNA level,

stability, secretion, and assembly of the ECM protein fibronectin, but on the other hand, it

also alters the mRNA levels and stability for a matrix metalloproteinase” (Brunette &

Chehroudi, 1999).


                       2.6 Role of the acquired exogenous pellicle



       Adhesion of cells or bacteria to the implant surface does not occur directly, but

through a protein layer adsorbed on this surface.        This adsorption process begins

immediately after contact between the implant and host fluids, and results in a very thin

layer of proteins and proteoglycans attached to the substrate by electrostatic, Van der

Waals and hydrogen bonds. This layer, called acquired exogenous pellicle, will allow a

bridging between the surface and the cells or bacteria. Thus, the nature of this pellicle,

which can be considered as a part of the extra-cellular matrix (ECM), will influence the

type of bacteria or cells that will be able to adhere on the substrate, as well as the

behavior of the cells, once attached. Coating a surface with specific proteins can increase

or decrease cell attachment on that surface. For example coating titanium with collagen

IV has been shown to favor human epithelial cell adhesion on this surface, as a

vitronectin coating tended to decrease it (Park, Kim, & Ko, 1998). Adhesion of host cells

to the implant surface is mediated by integrin-receptors mechanisms between the ECM

proteins and receptors in the focal adhesion sites of the cells. These focal adhesion sites

contain integrins and recognize specific peptide sequences, for example RGD (Arginine-
                                                                                           24


Glycine-Aspartic acid), present in ECM proteins, such as fibronectin and vitronectin

(Baier, Meyer, Natiella, Natiella, & Carter, 1984; Ruoslahti & Pierschbacher, 1987).

ECM proteins also mediate bacterial adhesion to the implant surface through adhesins,

which are ECM-recognizing protein receptors on the bacteria cell wall.


                                       2.7 Cytokines



       Cytokines are small soluble proteins, synthesized by cells, generally immune

system cells, but by other cell types as well. Close in their functions to hormones or

neuromediators, cytokines are essential to communication between cells.              Through

specific receptors they can have a paracrine, endocrine, juxtacrine, or autocrine action. In

other words, they can act on close cells, distant cells, cells in contact, or on the very cell

that produced them.      A major difference between hormones and cytokines is that

hormones are secreted by a specific type of specialized cells. Also, targets for the action

of a hormone are generally more specific.

       The term cytokine was coined by Stanley Cohen in 1974, even though cytokines

have been described earlier; for example, Interferon was already identified in 1957.

       Their roles are multiple; intervening in processes such as inflammation, immune

response, hematopoiesis, reproduction, embryogenesis... They also contribute in

pathological processes such as autoimmune diseases, cancers, rheumatoid polyarthritis,

HIV infection and more.

       Cytokines are grouped in family, but not always following a rational order, as

their names and classification were often attributed as they were discovered, before their

real functions were known. Among cytokines can be found lymphokines and monokines,

interleukins, chemokines, and others.      Interleukins don't share common functions or

chemical features, but were only named as such as they were discovered. Chemokines,

on the contrary, have one common feature being their chemotactic property. As a matter
                                                                                      25


of fact, determining these functions is a difficult task because of redundancy (different

cytokines producing the same effect) and their pleiotropy (different cells producing the

same cytokines and cytokines acting on different cells). Moreover, an initial action is

often followed by a cascade of events, which makes investigations even more difficult.

Cytokines can also act synergistically or antagonistically (Cavaillon, 2005).     Often,

cytokines are described as pro-inflammatory (IL-1, IL-6, IL-12, IL-18, TNF- , INF- ...)

or anti-inflammatory (IL-4, IL-10, IL-13, INF- , TGF- ...), but even this distinction has

been challenged as being too simplistic (Cavaillon, 2001). Apparently, inflammatory

action depends on the cytokine amount, the nature of the target cell, of the activating

signal, the produced cytokine, the timing, the sequence of events, and even the

experimental design (Cavaillon, 1995).

       Keeping in mind the reserves above stated, it is still possible to define major

actions for certain specific cytokines. For example, IL-1, IL-6, and TNF- have regularly

been associated with bone resorption, and are considered pro-inflammatory cytokines, as

well as IL-12. On the other hand, IL-10 is generally considered an anti-inflammatory

cytokine.   IL-8 is a chemokine and has a chemotactic action on the neutrophiles

(Fitzgerald & Kreutzer, 1995).      The NCBI website provides useful information on

different cytokines:

                    IL-1beta gene encodes a protein that is a member of the
            interleukin 1 cytokine family. This cytokine is produced by
            activated macrophages as a proprotein, which is proteolytically
            processed to its active form by caspase 1 (CASP1/ICE). This
            cytokine is an important mediator of the inflammatory response,
            and is involved in a variety of cellular activities, including cell
            proliferation, differentiation, and apoptosis. The induction of
            cyclooxygenase-2 (PTGS2/COX2) by this cytokine in the central
            nervous system (CNS) is found to contribute to inflammatory pain
            hypersensitivity. IL-6 gene encodes a cytokine that functions in
            inflammation and the maturation of B cells. The protein is
            primarily produced at sites of acute and chronic inflammation,
            where it is secreted into the serum and induces a transcriptional
            inflammatory response through interleukin 6 receptor, . The
            functioning of this gene is implicated in a wide variety of
            inflammation-associated disease states, including susceptibility to
            diabetes mellitus and systemic juvenile rheumatoid arthritis. IL-8
                                                                                          26

           encodes for a protein that is a member of the CXC chemokine
           family. This chemokine is one of the major mediators of the
           inflammatory response. Several cell types secrete this chemokine.
           It functions as a chemo-attractant, and is also a potent angiogenic
           factor. This gene is believed to play a role in the pathogenesis of
           bronchiolitis, a common respiratory tract disease caused by viral
           infection. IL-10 encodes a protein that is a cytokine produced
           primarily by monocytes and to a lesser extent by lymphocytes.
           This cytokine has pleiotropic effects in immunoregulation and
           inflammation. It down-regulates the expression of Th1 cytokines,
           MHC class II Ags, and co-stimulatory molecules on macrophages.
           It also enhances B cell survival, proliferation, and antibody
           production. This cytokine can block NF-kappa B activity, and is
           involved in the regulation of the JAK-STAT signaling pathway.
           Knockout studies in mice suggested the function of this cytokine as
           an essential immunoregulator in the intestinal tract.

                                                                  NCBI website



       Even if their action is not totally understood, it is interesting to assess production

of such cytokines by specific cells, would it be only to be able to disclose a difference in

cell behavior under specific conditions, such as, for example, variable substrates the cells

would attach to, or be in contact with. Studies have shown that the type of material in

contact with cells can affect their production of cytokines (Perala et al., 1992; Schmalz,

Schuster, & Schweikl, 1998; Spyrou et al., 2002). Similarly, it has been reported that also

the topography of the substrate can influence the level of cytokines produced by the cells

in contact (Refai, Textor, Brunette, & Waterfield, 2004; Spyrou et al., 2002).

                   2.8 Biological tissues in contact with the implant-
                                   abutment complex



       The primary role of the implant / abutment complex is to provide support and

retention for various prosthetic elements, which will replace the missing tooth or teeth.

Typically, the implant body will replace the missing root, and the abutment will serve as a

connector between that implant and the prosthetic piece. The implant body does not have
                                                                                          27


the exact same relationship with the surrounding tissues as the natural root. The main

difference is the absence of the periodontal ligament (PDL), which in the case of a natural

tooth provides root anchorage to the bone socket through a net of fibers. This ligament

contains different cells and histological elements, which perform specific functions, such

as mechano-reception (providing feedback on the magnitude and direction of forces

applied on the tooth), or suspension, owing to the visco-elastic properties of this ligament.

The implant-prosthesis complex, because of the absence of this ligament, lacks some of

these sensorial and protection roles. The relationship between the implant body and the

surrounding tissue is achieved mainly by direct bone contact, or contact with connective

tissue that does not present the organization and structure of the PDL. The abutment, or

transmucosal component, also presents a major difference in its relationship with the

surrounding soft tissues, compare to the transgingival part of the natural tooth, and it is

once again related to the absence of direct fiber anchorage between these elements. After

a short description of the different histological tissue that contact the implant/abutment

complex, the differences between the natural tooth and the implant/abutment-complex

relationship with the surrounding soft tissues will be reviewed.


                                     2.8.1 Bone tissue



       Bone tissue is a mineralized connective tissue providing support, protection,

locomotion, and constitutes an important reservoir of minerals. It consists by weight in

about 28% type I collagen and 5% non-collagenous proteins. This organic matrix is

permeated by a substituted hydroxyapatite crystal (Ca10[PO4]6[OH]2), which makes up the
remaining 67% of bone.

       Bones present a dense outer layer constituted by cortical bone and a central cavity,

the medulla, filled by red or yellow bone marrow. This central cavity is sometimes

divided in small cells by bone trabeculae, creating a network called cancellous or spongy
                                                                                            28


bone. Cortical and cancellous bones have different behavior and metabolic responses, but

both share identical histological features, in that they consist of microscopic layers of

lamellae. These layers are organized either in concentric, circumferential, or interstitial

structures.

       Concentric lamellae constitute a cylinder of bone, called osteon, generally

oriented parallel to the long axis of bones. This osteon is the primary metabolic unit of

bone. A canal, the Haversian canal, is present in its center. It contains a capillary and is

lined by a single layer of bone cells. Adjacent haversian canals are linked by a network

of secondary canals, called Volkmann canals, also containing a blood vessel, thus

creating a rich vascular complex. Interstitial lamellae fill the spaces between the osteons.

The whole structure is embedded between an inner and an outer layer of bone, made of

circumferential lamellae.

       Bone tissue formation, turnover, and reparation are completed by specialized

cells, osteoblasts and osteoclasts, the former forming the bone, the latter resorbing it.


2.8.1.1 Osteoblasts



       Osteoblasts are mono-nucleated cells synthesizing collagenous and non-

collagenous bone matrix proteins. This matrix, called osteoid, will serve as a scaffold for

the later mineralization of the bone. Osteoblasts are issued form pluripotent cells of

mesenchymal origin (or ectomesenchymal origin in the head). Their differentiation is

controlled by complex mechanisms and feedback loops which will not be described here.

Pre-osteoblasts and osteoblasts are capable of mitosis, even though they are specialized

cells. Both exhibit high level of alkaline phosphatase on their plasmatic membrane outer

surface. This enzyme probably cleaves the organically bounded phosphate, thus likely

contributing to the initiation and propagation of the osteoid mineralization.
                                                                                           29


       Besides the matrix structural proteins (collagen I mainly), osteoblasts secrete non-

collagenous proteins such as decorin, thrombospondin, fibronectin osteopontin, etc. This

secretory activity depends on the developmental stage of the cell.

       In addition, osteoblasts and their precursors secrete a variety of growth factors and

cytokines, which participate in the regulation of the cellular function and bone

formation/resorption.    Examples of these proteins are Bone Morphogenic Proteins

(BMPs) such as BMP-2, BMP-7, Insulin-Like Growth Factor I and II (IGF-I and IGF-II),

Fibroblastic Growth Factor (FGF), Platelet-Derived Growth Factor (PDGF), etc.

       When bone formation ceases, a single-cell layer of osteoblasts extends along the

bone surface; the osteoblasts present then a flatter appearance and are called lining cells.


2.8.1.2 Osteocytes



       As osteoblasts form bone, a few of them will be entrapped within the mineralized

matrix and are then called osteocytes, occupying a cavity called osteocytic lacunae.

Small extensions interconnect the different lacunae, containing cell processes.          The

osteocytes are thus able to maintain contact with adjacent osteocytes or osteoblasts. This

communicative network allows maintenance of bone integrity and vitality, especially

through the repair of micro-cracks. If this network does not exist, hyper-mineralization

(sclerosis) and death of the bone occur.


2.8.1.3 Osteoclasts



       Osteoclasts are large multinucleated cells, found mainly at the bone surface. Their

main function is resorption, which, combined to the synthesizing action of the osteoblasts,

allows formation, remodeling and reparation of the bones. A typical feature of these cells

is the presence of a ruffle border, which consists in numerous evaginations of the
                                                                                          30


cytoplasmic membrane, oriented toward the bone surface. Around the ruffle border, a

peripheral adherent zone, called clear or sealing zone, created by the cytoplasmic

membrane of the osteoclast, attaches the cell onto the bone surface and seals the space

comprised between this ruffle border and the bone. Proton pumps associated with the

ruffle border will increase the hydrogen ion concentration in the sealed zone, thus

decreasing the pH. This drop in pH leads to the demineralization of the bone surface.

Vesicles are then released by the osteoclast in that space, liberating various enzymes,

such as cathepsins, which degrade the organic matrix exposed by the demineralization

process. Organic and inorganic products of the degradation processes are then evacuated

by endocytosis and transcytosis.      The depression left on the bone surface after the

resorptive action of the osteoclast is called a Howship’s lacunae.

       Osteoclasts have a hematopoietic origin, and, as the osteoblasts, have a multi-

stage differentiation process. Each stage is under the control of different cytokines,

growth factors and hormones. The mechanism is very complex, and will not be described

in details here. It is still worth mentioning that the differentiation of the osteoclast

precursors in mature osteoclasts depends on the presence of osteoblasts, through a

receptor-ligand coupling mechanism. On the osteoclast precursor surface, a receptor,

called RANK, has to bind to a ligand of the osteoblast surface called RANKL to be able
to differentiate in a mature active osteoclast. Therefore, resorption is more likely to occur

where osteoblasts are readily available to synthesize new bone. A soluble decoy receptor

secreted by the osteoblasts, osteoprotegerin (a member of the Tumor Necrosis Factors

super-family), can bind to the RANKL and block the interaction between the osteoblasts

and the osteoclasts precursors, thus blocking the differentiation of the latter, and playing

an important role in the regulation of osteoclasis.
                                                                                         31


2.8.1.4 Osseointegration



       Attempts have been made to obtain clinical success of implants without direct

bone contact, the interface being constituted either by fibrous tissue, trying thus to mimic

a periodontal ligament, or by polymethyl-methacrylate cements, then (and still) used in

orthopedic surgery, but providing poor clinical results (Zarb, Smith, Levant, Graham, &

Staatsexamen, 1979).

       In the early 60s, as Dr Brånemark was more and more conscious of the impact his

work could have on implantology, he believed that he had to coin a new term to refer to

the in-growth of bone in the threads and crevices of titanium implants. He settled upon

the term “osseo-integration”, derived from the Latin words os (bone) and integro (to

renew, to repair) (Branemark et al., 1969). This concept was not accepted in the first

placed; nor was it considered as a clinical achievement, neither was it regarded as a

biological possibility. It had to wait until 1976 to gain recognition when the Swedish

Board for Health and Welfare authorized 3 dental professors from universities other than

Göteborg to carry out a clinical review on a selected group of patients treated with

“osseo-integrated” implants. “Osseo-integration was confirmed a few years later by Zarb

and co-workers by the so-called “Toronto study” (Albrektsson & Wennerberg, 2005).

Following this study, in 1982 the “Toronto-conference” was held in front of an important

panel of representatives of most of all major schools of North America, implementing

definitively the concept and the term of osseointegration in all minds. Osseointegration

of titanium opened various medical treatments modalities such as bone-anchored hearing

devices or screws to support facial epitheses (Albrektsson & Wennerberg, 2005). But the

concept needed to be defined further: did osseointegration mean 100% bone-to-implant

contact around the entire circumference of the implant or was some interfacial soft tissue

acceptable? And at what level of resolution had the bone-implant contact to be direct? In
                                                                                        32


the early 1990s, a new definition was proposed:           “ a process whereby clinically

asymptomatic rigid fixation of alloplastic materials is achieved and maintained in bone

during functional loading” (Albrektsson & Wennerberg, 2005). One advantage of that

definition was its clinical nature. It is now widely accepted that osseointegration is

necessary for the clinical success of the implant, and therefore, this concept has given

pretexts for a formidable number of studies throughout the last five decades. But very

soon after function and stability of dental implants increased, in connection to improved

surgical protocols and better material characteristics for implants, esthetic and health of

the soft tissue surrounding the implant/abutment complex was sought after by clinicians

as well as patients, and interest in research branched toward that direction too.

        By analogy, the expression “soft tissue integration” is more and more frequently

used.   Before parameters affecting that soft tissue integration are reviewed, a brief

presentation of the anatomy of the soft tissue/implant interface will be presented, and

compared to the dento-gingival junction. A brief overview of the oral mucosa will be

given first.


                                    2.8.2 Oral mucosa



        The oral mucosa is the lining tissue separating the oral cavity from the

environment. It covers most of the surfaces of the oral cavity, except the surfaces of the

teeth. The junction between this mucosa and the teeth is a very specialized type of

attachment, which allows a semi-permeable seal around them, avoiding antigen (such as

bacteria) penetration. The oral mucosa consists of two separates tissue components: a

covering stratified squamous epithelium and an underlying connective tissue layer, called

lamina propria (figure 16). The two layers carry out common functions and should be

considered as an organ. These functions are protection, sensation, secretion and thermal

regulation. They may vary from one region to another one of the oral cavity. The
                                                                                         33


protective role is fundamental in the oral cavity, due to the direct relationship with the

external environment, the permanent presence of bacteria in the mouth, and the frequent

exposure to mechanical forces during chewing, biting, seizing food (Cate & Nanci, 2008).

An important detail is the fact that the cuff of soft tissue surrounding the implant neck or

abutment can be keratinized, attached mucosa, or mobile mucosa. Secretion of saliva is

also a fundamental function of the oral mucosa, but is not performed by the soft tissue

surrounding the implant complex.


2.8.2.1 Structure of the oral mucosa



        The interface between the epithelium and the connective tissue is usually

irregular.   Upward projections of connective tissue (connective tissue papillae)

interdigitate with epithelial ridges or pegs, called the rete ridges (Cate & Nanci, 2008).

At this interface, there is a basal lamina that appears as a structureless layer. Underneath

the connective layer is the submucosa of the oral cavity.          The boundary between

submucosa and mucosa is not clearly identified. This submucosa contains the major

blood vessels, nerves, and sometimes fatty or glandular tissue (figure 16). It is the

composition of that layer that will determine the flexibility of the attachment of the oral

mucosa to the underlying structures. In the region of gingival or parts of the hard palate,

the oral mucosa is attached directly to the periosteum of the underlying bone. This

configuration is then called a mucoperiosteum (Cate & Nanci, 2008).
                                                          34




Figure 16: Main tissue component of the oral mucosa.


Source: Cate, T.; 2008, Oral histology, Elsevier Mosby.
                                                                                       35




2.8.2.2 Connective tissue



       The connective tissue of the oral mucosa is also called lamina propria. It provides

support and nutriments to the oral epithelium. It is constituted by a ground substance in

which cells, fibers, blood vessels and nerves can be found. The ground substance is

composed of proteoglycans and glycoproteins, such as fibronectin, involved in

attachment, spreading and migration of cells.             It contains highly hydrated

glycosaminoglycans and serum-derived proteins. Two major types of fibers can be found

in this substance: the elastic fibers and the collagen (especially type I and III) fibers.

These fibers provide support and resistance to the mucosa, and are organized in bundles

following functional orientation (Cate & Nanci, 2008).

       The main cells of the lamina propria are the fibroblasts, responsible of the

elaboration of the fibers and the ground substance, playing thus a major role in

maintaining the integrity of the tissue. Moreover, their secretion of growth factors,

cytokines and inflammatory mediators, as well as their function of degradation of the

ECM allow them to actively participate in wound healing and tissue remodeling The

lamina propria contains also macrophages, histiocytes, mast cells, polymorphonuclear

leukocytes, lymphocytes, plasma cells and endothelial cells, all playing fundamental roles

in the inflammatory and immune defense systems (Cate & Nanci, 2008).


2.8.2.3 Epithelium



       The oral epithelium is the primary barrier between the external environment and

deeper tissues.   It is a stratified squamous epithelium.      Its structural integrity is

maintained by continuous cell renewal produced by mitotic division in the deepest layers.
                                                                                          36


The cells can therefore be considered as two functional populations: the progenitor

population and the maturing population (Cate & Nanci, 2008).


2.8.2.3.1 Epithelial proliferation



         The progenitor cells are situated in the parabasal two or three layers. Some of

them are stem cells, multiplying to produce new progenitor cells and maintain the

proliferative potential of the tissue. The others cells are amplifying cells which provide a

reservoir for the surface layers. The turnover time is the time taken for a cell to divide

and pass trough the entire epithelium. Its duration is about 41 to 57 days in the gingiva

and 25 days in the cheeks, to compare with 52 to 75 days in the skin and 4 to 14 days in

the gut. Keratinized epithelium turnover is slower than non-keratinized (Cate & Nanci,

2008).

         Proliferation and differentiation are influenced by different factors, such as drugs

(cancer chemotherapeutic drugs), time of the day, stress, and inflammation. The latter, if

mild, stimulates the turnover, or if severe, reduces it.       Various cytokines, such as

epidermal growth factor, interleukin-1, and transforming growth factor I and II, may also

influence the proliferative rate (Cate & Nanci, 2008).

2.8.2.3.2 Epithelial maturation



         The cells arising by division in the parabasal layers can remain in the progenitor

cell population or undergo maturation as they move toward the surface.             One can

distinguish two different patterns of maturation: keratinization and non-keratinization

(Cate & Nanci, 2008). Some areas of the oral mucosa, for example hard palate and

gingiva are inflexible, tough, resistant, and tightly bound to the lamina propria. This

results from the formation of a surface layer of keratin, a process known as keratinization.
                                                                                           37


Keratins constitute of a large family of proteins of different molecular weight. Epithelial

cells in certain areas produce them. These cells are then often called keratinocytes.

Keratin proteins will participate in the formation of intra or inter-cellular structures, such

as tonofibrils, desmosomes or hemi-desmosomes, thus contributing to the resistance

required to maintain cell shape and cohesion of the epithelial layer (Cate & Nanci, 2008).


                              2.8.3 Blood vessels and nerves



       Blood and nerve supply of the oral mucosa are rich and dense (Cate & Nanci,

2008). Except for that characteristic, blood vessels and nerves embedded in the soft and

hard peri-implant tissues have the same histological features and perform the same

function as elsewhere in the human body, and will therefore not be described here.


                      2.9 Anatomy of the soft tissue/tooth interface



       This interface has been heavily investigated and described for several decades

now. Two different zones can be described at this interface: an epithelial zone and a

connective tissue zone. The epithelial zone extends from the gingival crest to the most

apical end of the junctional epithelium, where the connective zone starts. The latter

extends in turn to the high of the alveolar bone, where the periodontal ligament starts.

       The epithelial component ranges usually between 0.5 and 3mm, dimensions above

that value being considered as pathologic (periodontal pocket). This epithelial zone can

be in turn divided in two distinct zones: the oral sulcular epithelium and the junctional

epithelium. The latter constitutes the actual tissue interfacing with the tooth surface, the

enamel in general (figure 17) (Cate & Nanci, 2008).
                                                                                       38




Figure 17: The dentogingival junction.


Legend: Oral epithelium (OE), oral sulcular epithelium, JE junctional epithelium (OSE),
   enamel (E), internal basal lamina (IBL), cementum (C), basal lamina (BL), connective
   tissue (CT), dentin (D), gingival crest (GC), gingival sulcus (GS).

Source: Cate, T.; 2008, Oral histology, Elsevier Mosby.




       The connective tissue part of that dento-gingival junction can also be divided in

two different compartments: the connective tissue supporting the epithelium above

described, and the connective tissue actually interfacing the tooth surface (generally the

cementum). The first one always presents an inflammatory infiltrate. The second is

essentially a mixed between collagen fibers and fibroblasts, supplemented by blood
                                                                                        39


vessels and nerves. Just above the crestal bone, the collagen fibers are found in the

lamina propria of the gingiva and form the gingival ligament (Cate & Nanci, 2008).

These fibers can be distinguished in different groups, depending on their origin,

termination, and orientation. The most numerous fibers are the dento-gingival fibers,

extending from the cementum to the lamina propria of the free and attached gingiva, these

fibers being oriented perpendicularly to the surface of the tooth.       Alveolo-gingival,

circular, dento-periosteal, and transseptal fibers form the other groups (Cate & Nanci,

2008).


                    2.10 Anatomy of the soft tissue/implant interface



         It is currently accepted to say that the interface between soft tissue and implant

has similarities with that of natural teeth (Myshin & Wiens, 2005). Specifically, the

epithelial tissue around the implant/abutment complex is remarkably similar to that of the

dento-gingival junction. Schupbach and Glauser could show that the structural properties

of the implant junctional epithelium correspond exactly to those around a tooth

(Schupbach & Glauser, 2007).

         If epithelium is similar, many authors have suggested a fundamental difference

between the peri-implant connective tissue and its counterpart around natural teeth:

around implants, which surface is not covered by cementum, collagen fibers run a course

more or less parallel to the abutment surface, and no fibers perpendicular to the

implant/abutment surface would be present (figure 18) (Berglundh et al., 1991; Buser et

al., 1992; Ruggeri, Franchi, Marini, Trisi, & Piatelli, 1992).
                                                                                        40




Figure 18: Difference between tooth- and implant-soft tissue interfaces.


Source: Rose L.F.; 2004, Periodontics, Elsevier Mosby.




       The connective tissue around the implant neck and/or abutment has long been

described as a scar tissue, with a predominance of collagen and few fibroblasts and

vascular structures. It was therefore argued that the tissue turnover of the peri-implant

mucosa was less rapid than that of gingival and that this attachment had poor regenerative

potential (Lindhe & Berglundh, 1998).

       In 1999, Moon et al. investigated the connective tissue in more depth, by

analyzing the repartition of the constituents of the connective tissue in relation to their

distance with the abutment.     They implanted 6 fixtures in each of 6 dogs, placed
                                                                                         41


abutments on these fixtures 3 months later, and finally sacrificed the dogs after another 6-

month period during which strict oral hygiene was implemented for the dogs.

Histological sections of the peri-implant tissues revealed that the distribution of the

constituents of the connective tissue within a 200 m band, varied from the surface of the

abutment toward a more peripheral zone. If previous observations of a scar-like tissue

were confirmed in the most external zone, closer to the surface of the abutment, more

fibroblasts and less collagen could be observed. The authors concluded that this part of

the connective tissue might have a higher turnover and a better healing potential after all,

and for this reason, could play a role even more important than formerly believed. They

confirmed the orientation of the collagen fibers to be parallel to that abutment surface,

and could not observed perpendicular fibers.         Because of those differences, they

recommended that the peri-implant soft tissue should not be called gingiva (Moon et al.,

1999).

         Schupbach and Glauser confirmed most of these findings in a report published in

2007.    They placed 12 experimental titanium mini-implants on five patients, and a

cylindrical abutment of 4mm was immediately attached onto the implant. All abutments

were made out of titanium, but with three different surface treatments. The surfaces were

either machined titanium, acid etched, or layered with oxidized microporous TiO2. After
8 weeks of transmucosal healing, the investigators retrieved the implants with a layer of

surrounding hard and soft tissue.     Light microscopy and transmission and scanning

electron microscopy were used to observe histological preparations made out of the

harvested samples. The authors reported here again the analogy between peri-implant

and peri-dental soft tissue organization. For all implants, the peri-implant mucosa was

composed of an epithelial and supra-crestal connective tissue cuff. The connective tissue

faced directly the implant surface. Collagen fibers and fibroblasts were loosely running

parallel to the surface of the implants and the abutments. They could observe for all three

surface types a densely packed fibers bundle with an orientation circumferential around
                                                                                         42


the implant in a horizontal plane, as seen around natural teeth. The most striking finding

of their experiment is probably the fact that they could demonstrate the existence of

functionally oriented fibers, extending more or less perpendicularly to the

implant/abutment surface, but this could be seen only for the oxidized-surface abutments

(Schupbach & Glauser, 2007).

       Other reports exist, where such perpendicular fibers have been described, in

particular when the surface of the implant or abutment contained porosities. Seventeen

years earlier, Steflik et al. could not only described the two-component attachment system

(epithelium-connective tissue), but also demonstrate the presence of such perpendicular

fibers (Steflik et al., 1990). In an in vitro experiment, Takata et al. could show that

bioactive materials such as bioglass or hydroxyapatite could lead to periodontal ligament

regeneration if periodontal ligament-derived cells could repopulate the surfaces, whereas

no such regeneration could be seen on titanium or partially stabilized zirconium surfaces

(Takata, Katauchi, Akagawa, & Nikai, 1994). In 1981 already, Schroeder et al. reported

on an experiment in which titanium implants, flame-sprayed with titanium powder,

providing a rough surface with pores 25 to 100 m in size, demonstrated non-inflammated

connective tissue with fibers inserted functionally at 90° into the plasma-sprayed neck.

Insertion was so strong that application of tensile strength dislodged particles of the
sprayed surface rather than detach the fibers from the surface (Schroeder, van der Zypen,

Stich, & Sutter, 1981). Such findings incited Deporter et al. to investigate on abutments

presenting a dual texture, with the most apical part being rough, to allow for the fibers of

the connective tissue to anchor on that surface, whilst a smoother coronal plaque would

limit bacterial attachment and, at the same time, allow epithelial cells to attach. These

experiments were sanctioned by a poor outcome, the porous zone being rapidly invaded

by bacteria leading to failure of the implant (Deporter et al., 1986; Deporter, Watson,

Pilliar, Howley, & Winslow, 1988).
                                                                                         43


       It seems nevertheless that such fiber anchorage would reinforce the attachment of

the connective tissue on the implant neck.       Many authors have suggested that this

connective portion would be the reason why epithelium down-growth is impeded, and

hence might contribute to a better stability of the soft tissue architecture and function

around implants. In 1983, Squier and Collins showed that porosities of 3 m would be of

ideal size to allow connective tissue penetration, and that this penetration would block

epithelium down-growth along the Millipore filters used for their experiment (Squier &

Collins, 1981).

       It is worth mentioning that in vivo studies of the soft-tissue/implant interface,

whether on animals or humans, are done in sites presenting keratinized epithelium. The

presence of such a keratinized mucosa is thought to be a positive factor in maintaining

soft tissue health (Myshin & Wiens, 2005). It would be interesting to investigate the

anatomy of the soft tissue/abutment interface when implants are placed in non-

keratinized, mobile mucosa.


                         2.11 Importance of the soft tissue cuff



       A firm attachment of the mucosa to a tooth or an implant is a prerequisite for the

perenity of the implant (Lindhe, Berglundh, Ericsson, Liljenberg, & Marinello, 1992). It

provides peripheral defense, and hemidesmosomes have a definite role in providing this

attachment. Recent data suggest that the hemidesmosomes could also act as specific sites

of signal transduction, and thus may participate in regulation of cell proliferation and

differentiation (Jones et al., 1998) and therefore in the osseointegration itself. The peri-

implant soft tissue cuff has to provide the same functions than the peri-dental gingiva.

Such functions are inflammatory and immunologic defenses, growth factors and

cytokines productions, filtering seal around the tooth or the implant, esthetic, and so on,

so forth. It has been stated that polymorphonuclear leukocytes migrating through the
                                                                                         44


junctional epithelium comprise the most important defense mechanism around implants

(Schupbach & Glauser, 2007). In addition, the junctional cells contain lysosomal bodies

as a further defense mechanism. Around natural teeth, the supragingival fibers apparatus

reinforces the attachment of the epithelium. This apparatus not only attaches the gingiva

to the tooth, but also creates a scaffold providing biomechanical resistance of the gingiva.

A similar apparatus is formed around implants, by fibers bundles running

circumferentially around the implant/abutment and fibers oriented longitudinally along

the implant/abutment long axis (Schupbach & Glauser, 2007).


                  2.12 Influence of material and characteristics of the
                                    abutment surface



       The different components of the soft tissue cuff around the implant neck and/or

the abutment, as well as the structure of this attachment apparatus were described above.

The importance of that cuff in terms of implant survival and esthetic outcome of the

restoration were also mentioned. For those reasons, it is obvious that the utmost attention

should be given to the properties of the implant abutment component to allow the

formation of that attachment, and all efforts should be made to improve it. Unfortunately,

the requirements for the abutment material and its properties are sometimes contradictory,

or difficult to achieve. For example, adhesion is desired for epithelial cells to provide a

seal around the transmucosal component, but at the same time, bacterial adhesion, which

can provoke a breakdown of the attachment, is not wanted. Furthermore, if adhesion of

the epithelial cells is necessary, down-growth of the same cells is to be avoided to prevent

formation of a periodontal pocket around the implant, or even worse, complete

marsupialization of that implant.
                                                                                            45


       Possible factors that could improve the relationship between the transmucosal

component and the soft tissue are the nature of the material, its topography, coatings, and

other possible chemicals treatments.


                     2.12.1 Materials for dental implant abutments



       Considering the elements presented above, it is clear that if the nature of the

substrate influences the type of acquired exogenous pellicle that will cover it, this will, in

turn, influence cellular adhesion and cell behavior. Titanium, gold base alloys, aluminum

oxide, and zirconium oxide are available as materials for abutments. These abutments

can be covered with dental porcelain extending more or less apically under the soft

tissues. The two materials of interest for the present experiment are titanium, either

machined, acid etched, or modified, and zirconium.


2.12.1.1 Titanium



       Titanium symbol is Ti and its atomic number of 22. Among its most remarkable

properties are its corrosion resistance, its strength-to-weight ratio, its poor heat and

electricity conductivity, and a very good biocompatibility. It is non-magnetic and is as

resistant as steel for about half its weight, but as steel, titanium structures are susceptible

to fatigue. A dioxide (and trioxide) layer spontaneously forms on titanium surfaces. This

layer is a few nanometers thick (3-5nm). It is the actual layer biological fluids and tissues

will contact when titanium dental (or else) implants are used. Use of this metal in

dentistry has been widely studied and Brånemark’s researches were probably the most

significant contribution to its acceptance in the medical field. Because of this acceptance,

the present review of the literature will not cover aspects such as biocompatibility or
                                                                                       46


physico-chemical properties.     These properties are incidentally close to those of

zirconium.


2.12.1.2 Zirconium



       The atomic number of zirconium is 40, and its symbol is Zr. As mentioned above,

it shares with titanium many physico-chemical features, such as resistance to corrosion,

strength, biocompatibility… A interesting property of zirconium is its polymorphism, this

metal presenting three different crystal configurations, namely monoclinic (M), cubic (C),

and tetragonal (T). At room temperature, the monoclinic form prevails and is stable up to

1170°C. Above, M T C transformation occurs. During cooling, a reverse T M

transformation takes place, accompanied by a volume expansion of 3-4%. C- and T-

phase can be stabilized at room temperature by adding oxides. The zirconium is then said

metastable, in so far that if stress is applied, the metastable C and T phases will then

transform back to M, with volume expansion. This for example, results in localized

compressive strengths around a crack that may develop under stress. The compressive

strengths will then “squeeze” the crack and stop it.       This is called transformation

toughening and participates directly in the good mechanical properties of zirconium.

Unfortunately, this transformation can occur even in more passive conditions, leading to

ageing of the material, roughening and particles loss. Zirconium oxide (ZrO2) is a

ceramic material, known as zirconia. The oxide mostly employed to reinforce zirconium

is yttrium oxide (Y2O3), leading to tetragonal zirconia polycrystal, and such names as

TZP or yttrium-reinforced zirconia have flourished in the medical and dental literature,

generally without being clearly explained. Zirconia has been largely employed in the

medical field, particularly in orthopedic surgery where it was proposed as hip prosthesis

material as early as 1969. Biocompatibility has been extensively investigated and is

excellent.   Notably, no carcinogenic or mutagenic effects could be associated with
                                                                                            47


zirconia (Covacci et al., 1999; Lohmann et al., 2002; Silva, Lameiras, & Lobato, 2002;

Torricelli et al., 2001). In the dental field, use of zirconia is more recent than titanium; it

has started to be employed in the early 1990, indications and commercial presentations

increasing since then.      In 1997, the first dental implant zirconia abutment was

commercialized (Zirabut , Wohlwena Innovative, Zurich, Switzerland). The white color

of this ceramic has promoted its use as an esthetic biomaterial, and reports have been

published on increased esthetic outcome when zirconia was used as an abutment material

instead of titanium (Watkin & Kerstein, 2008).          Even more, certain manufacturers

propose now different white shades to improve even more the esthetic, the initial white-

opaque color of this ceramic being sometimes too harsh.

       In spite of those interesting characteristics, zirconia is by far not a perfect material

when it comes to its medical use. In a recent review, Manicone et al. stated that even

though zirconia seems to be suitable for implant abutments, it still lacks enough research

to be fully accepted for this use (Manicone, Rossi Iommetti, & Raffaelli, 2007). In 2001,

a series of catastrophic failures of zirconia femoral implant heads compromised strongly

the future of that material as an implant material. Ageing has been designed as the source

of the problem. It is not yet clear when such ageing phenomenon will occur, and how to

prevent it (Clarke et al., 2003). Chevalier reviewed the current knowledge on this ageing
process, and emphasized the importance of a proper manufacturing and handling of the

product. He made a certain number of recommendations regarding manufacturing criteria

for zirconia as an implant material (Chevalier, 2006). In the dental field, only a few

clinical studies have been published on the use of zirconia as material for dental

abutments. Glauser et al. reported on 54 implants placed on 27 patients, restored with

zirconia abutments and full-ceramic crowns, and followed-up for four years.                No

abutment fractured over that period of time, nor did the crowns, but the authors reported

two screw loosening and three crown with minor incisal chipping. Favorable soft tissue

reactions, and usual level of bone loss were observed over the four years These results
                                                                                           48


must be carefully interpreted though, since there was a 33% dropout of the patients (and

restorations) (Glauser et al., 2004). It has also been reported that, for a similar roughness,

zirconia does not allow dental plaque to build-up as easily as on titanium (Scarano,

Piattelli, Caputi, Favero, & Piattelli, 2004). In 2006, Degidi et al. published on the

inflammatory activity around zirconia abutments, compared to titanium abutments. They

observed in their in vivo experimentation that after 6 months of healing, titanium

abutments seemed to induce a more severe inflammatory reaction than the zirconia

abutments, based on measures of vascular endothelial growth factor (VEGF) and nitric

oxide synthase (NOS) expression, as well as inflammatory infiltrate, proliferative activity

expression, and microvessel density (Degidi et al., 2006).

       Even though these in vivo studies presented a favorable outcome for zirconia, the

material still comes with disadvantages when used in the implant dental field. Fabrication

costs are high and the material is not available for every commercial implant system.

Concerns about detrimental effect on its properties when exposed to a wet environment

have also been raised for the dental use (Swab, 1991). Also, when zirconia is directly

veneered with dental porcelain, fracture of the veneering has been reported, and the core-

veneer interface is thought to be the weakest element of those restorations (Aboushelib,

de Jager, Kleverlaan, & Feilzer, 2007). Given the hardness of the material, possible
fretting wear of the fixture platform has been evoked (Brodbeck, 2003), and this could in

turn lead to an increased rotational freedom of the abutment around the anti-rotational

feature (hexagon generally), and induce more screw loosening (Binon & McHugh, 1996;

Binon, 1996). A new abutment with an original design has been proposed to avoid this

fretting wear (ZiReal    post, Biomet 3i); it features a titanium base fused to the zirconia

core. The purpose is to have a titanium-titanium interface between the abutment and the

fixture, and possibly avoid this wear phenomenon (Brodbeck, 2003). Longevity of the

fused interface between the titanium base and the zirconia core has yet to be established.

Another concept has been proposed for about 10 years in Europe and has recently been
                                                                                       49


made available in United States. It consists in a one-piece implant body/transmucosal

part, similar to Straumann ITI implants, but using both titanium and zirconium as core

material (TBR Z1 , Sudimplant, Toulouse, France). The osseous part of the implant is

made of sandblasted titanium, whereas the transmucosal part is made of zirconia (figure

19).   This design would use advantages of both materials where they are the most

required (Masini, 2005), and would avoid the inconvenient of titanium fretting wear by a

separate piece of zirconia. It also avoids the presence of a micro-gap at the soft tissues

level, authorizes a one-stage surgical procedure and simplifies the prosthetic restoration

of the tooth.

        In summary, it can be said that zirconium oxide, also known as zirconia, presents

attractive properties for the dental use. But concerns remains when it comes to ageing,

manufacturing, possible wear of opposing metal pieces, bond with dental porcelain, and

the lack of clinical studies. Moreover, investigation on epithelial cells attachment and

biology on that material is essentially lacking.




Figure 19: TBR Z1     implant with zirconia transmucosal neck.


Source: TBR     Group Website.
                                                                                          50


                2.12.2 Impact of the nature of the substrate on cellular

                                         adhesion



       Simion et al. could show that human gingival cells (fibroblasts and epithelial

cells) could develop on gold, gold ceramic, and passivated titanium and titanium alloy

(Ti6Al4V), but not as well on not passivated titanium and Ti6Al4V alloy. They could

also observe that if attachment and proliferation of the cells were equivalent on gold, gold

ceramic and passivated titanium and titanium alloys, the quality of this attachment was

not as good for gold and gold ceramic, the cell sheets offering less resistance to separation

from the substrate (Simion, Baldoni, & Rossi, 1991).

       Räisänen et al. investigated epithelial cells adhesion on five different materials

used for implant or implant-associated supra-structures. The surfaces were commercially

pure (c.p.) titanium, titanium alloy (Ti6Al4V), dental gold alloy, aluminum oxide, and

dental porcelain. The surfaces were polished to the smoothest surface possible, using

mechanical polishers up to grit 4000, and diamond pastes.                 Ti was polished

electrolytically. The rational of this extreme polishing was that fibroblasts and epithelial

cells are thought to prefer smooth surfaces. Cell coverage was 2.5 times higher for

metallic surfaces (Ti, Ti6A4V, and gold) than on ceramic materials (aluminum oxide and
dental porcelain). More cells seemed to attach the metallic surfaces. On porcelain, cells

were rounded and their edges appeared loosely attached to the surface (Raisanen et al.,

2000). The investigators further assessed attachment on the different surfaces, using

vinculin and 6 4 integrin immuno-staining.          On metallic surfaces, well developed,

round, or elongated focal adhesion contacts were disclosed by vinculin marking at the

periphery of the cells. On porcelain, focal contacts were seen only in cells growing in

islands.   However, most of the cells were solitary and rounded, and vinculin only

expressed diffusely in the cytoplasm.       Similarly, on metal surfaces, 6 4 integrin
                                                                                        51


immuno-fluorescent images were granular, whereas on ceramic surfaces, the images took

this aspect only on islands of cells. Otherwise, only small irregular patches were visible.

The authors concluded that in vitro, epithelial cells adhere more avidly on metallic

surfaces than on ceramic surfaces. In their discussion, they proposed the role of the oxide

layer on titanium surface, as a possible explanation of their observation. This oxide layer

is supposed to favor adsorption of physiological fluids and proteins and thus, to promote

cellular adhesion. But they did not discuss the fact that gold do not tend to be covered by

such an oxide layer, and appeared nevertheless to be a good substrate for the cells. The

other explanation proposed by the authors is the role of the surface roughness. They

noted that, in spite of their efforts to get a similar high polish on all surfaces, SEM

analysis showed that aluminum oxide and ceramic were the roughest surfaces, titanium

being the smoothest. The tendency of the cells to adhere on the substrates was correlated

to the roughness of the latter In that case, it might be possible that the nature of the

substrate is not, or not the only factor to influence cell adhesion.        If the second

explanation is accepted, soft tissue penetrating parts of dental implants should probably

be smooth (Raisanen et al., 2000).

        Welander et al. investigated on dogs the soft tissue barrier formed to implants

abutments of different materials. They placed two sets of four implants on each of six
dogs.   The two sets were placed at a three-month interval.            Two months after

implantation, fixtures were covered with two titanium abutments, one zirconium oxide

ceramic abutment (ZrO2), and one gold-platinum (Au-Pt) alloy, and this for each of the

two sets. Dogs were placed on a five-month plaque control program and euthanized two

months after the last set of four abutments was placed. The investigators evaluated the

distance between different landmarks of the peri-implant tissue (margin of the peri-

implant mucosa, apical termination of the barrier epithelium, abutment/fixture borderline,

and bone to implant contact). They also evaluated the amount of collagen, fibroblasts,

vessels and leukocytes within the connective tissue around the implant. Having placed
                                                                                       52


the two sets of implants/abutments at a different time, they could compare the results for

each material at two months and five months, and hence assess the stability of the soft

tissue/abutment complex over time.      Results demonstrate that ceramic and titanium

abutments provided a more stable soft tissue environment than Au-Pt abutments overtime

(from two to five months), the latter showing a tendency toward recession, usually

beyond the abutment/fixture borderline. Moreover, Au-Pt abutments were surrounded by

a connective tissue presenting more inflammatory cells and less collagen and fibroblasts

than ceramic and titanium abutments. In that regard, ceramic showed even better results

than titanium. The authors suggested that ZrO2 might provide better conditions for

epithelial cell attachment, but mentioned that others have proposed a difference in

bacteria colonization as a possible explanation for this difference between ZrO2 and

titanium (Welander, Abrahamsson, & Berglundh, 2008).

       About 10 years earlier, Abrahamsson et al. conducted a similar study. They

placed 3 implants on each side of the mandible of each of five dogs. After a 3-months

healing period, abutments were secured on the implants. For each dog, one control

abutment made of c.p. titanium and one abutment made of highly sintered ceramic

(Al2O3) were placed on each side, one gold abutment and one short titanium abutment

covered by a dental porcelain supra-structure were placed on the two remaining
abutments. After a 6-months period, with intensive plaque control measures, dogs were

sacrificed and histometric and morphometric analysis were done, using the same

histological landmarks than Welander et al. for their experiment. All gold abutments and

short titanium/dental porcelain abutments exhibited important recessions. The connective

tissue part was always in contact with the implant body itself, hence with titanium. Soft

tissue recessions were so important on 3 out of 5 gold abutments that the

implant/abutment interface was clinically exposed and even the epithelial attachment was

located on the titanium surface of the implant body. Bone loss was observable on all

these implants, probably occurring to provide the required space for the attachment
                                                                                        53


apparatus, which could apparently not consistently occur on these abutment materials.

On the contrary, both ceramic abutments and control abutments (c.p. Ti) presented two

distinct zones of tissue, always above the implant/abutment interface. These zones were a

junctional epithelium and a connective tissue zone, respectively of 2mm and 1.3mm

approximately. It should be mentioned that inflammatory infiltrate was systematically

observed at the abutment/implant interface for all abutment materials, provided some

connective tissue was present in regard of that zone. This infiltrate was slightly more

important for ceramic abutments than titanium abutments, probably because of a lesser

degree of adaptation between the ceramics abutments and the fixtures. The authors

concluded that material used for implant abutments was of decisive importance for the

quality of the attachment that occurs between the mucosa and the implant. Titanium and

Al2O3 abutments established similar conditions for mucosal healing to the abutment

surface and allowed the formation of an attachment that included one epithelial and one

connective portion. On the contrary, for gold alloy abutments or dental porcelain, no

proper attachment seemed to form at the abutment level but the soft tissue margin receded

and bone resorption occurred (Abrahamsson et al., 1998). Thomsen et al. suggested that

metallic surfaces do not induce adequate cell growth and adhesion; transition metals with

an oxide layer are more favorable (Thomsen et al., 1997).
       But these observations are not always accepted as relevant. In a recent review of

the literature, Linkevicius and Apse questioned the validity of conclusions drawn from

such animal experiments, referring to clinical trials on human were gold abutments

performed equally well, compared to titanium. They concluded their review by stating

that, as of the evidence available at this time, there is no significant difference between

abutments made of titanium, gold, zirconium or aluminum oxide in terms of stability of

the peri-implant tissues (Linkevicius & Apse, 2008). Even more, Abrahamsson himself

recently contradicted his previously published results in a new experiment, again on dogs.

Four different types of one-piece implant/abutment were designed for that experiment,
                                                                                           54


each implant/abutment being divided in three different zones (physically continuous, by

use of laser soldering), each made either of gold or titanium. The four designs consisted

in different combinations of gold and titanium along the three zones. Regardless of the

combination, and hence the material in contact with the soft tissues, the latter showed

similar anatomical features. The authors justified the dissimilar results for this new study

by differences in experimental design (Abrahamsson & Cardaropoli, 2007).                 It is

nevertheless obvious that the nature of the material used for the abutment influences the

biology and behavior of the soft tissues surrounding it. It is still not clear whether gold is

an acceptable alternative, but titanium and ceramic materials seem to be equally

appropriate choices.

       If titanium is still considered as the reference material, it is nevertheless possible

to improve the characteristics of that material. Shiraiwa et al. compared the attachment of

rat oral epithelial cells (OE), and their subsequent behavior on titanium, with plastic

(polystyrene) and glass surfaces. They assessed attachment, spreading, and proliferation

of the cells, and controlled the average roughness of the three different surfaces. Ra

values for the surfaces presented no significant difference, the values being comprised

between 0.03 m for glass and polystyrene, and 0.05 m for titanium. Therefore, the

authors assumed that results would not depend on the surface topography, but solely on
the nature of the substrate. They found that in culture between 3 to 72 hours, more OE

were attached on polystyrene than on Ti, while glass surfaces had an intermediate value.

The area of the cells increased gradually from 3 to 72 hours for the three substrates, but

the average area was greater for cells on polystyrene than on Ti. Cell colonies were

classified in three categories, depending on the number of cells forming the colonies. A

count of 2-cell, 3-to 5 cell, and 6-cell or more colonies was done. From 12 to 72 hours of

culture, polystyrene had the greatest number of 2-cell colonies, and Ti had the fewest.

The number of 3-to 5 cell colonies gradually increased on polystyrene and glass up to 72

hours, and was first observed on Ti after 72 hours. First appearance of 6-cell colonies
                                                                                        55


was after 96 hours for all substrates, but their number was greater on polystyrene than Ti

from 96 to 120 hours (Shiraiwa, Goto, Yoshinari, Koyano, & Tanaka, 2002). Similarly,

proliferation rate was at its maximum after 48 hours on cultures on polystyrene dishes and

on glass, and only after 72 hours on Ti. The authors also measured the amount of

Laminin-5 (LN-5), an extra-cellular protein secreted by epithelial cells, and considered as

crucial for adhesion of cells on the substrate. After 96 hours, there was a significant

difference in the LN-5 area between Ti and polystyrene. The investigators inferred from

this result that OE cells on Ti were poorly attached (Shiraiwa et al., 2002), in contrast

with the study from Baharloo described below (Baharloo, Textor, & Brunette, 2005). But

the latter study compares smooth Ti surfaces with rougher ones, and not with polystyrene.

This might be accountable for the different observations. Altogether, Shiraiwa et al.

conclude that initial attachment, spread, and proliferation of Ti was inferior to

polystyrene, and that properties of Ti substrate to which epithelial cells attached need to

be improved (Shiraiwa et al., 2002).

       In summary, abutment material certainly has an influence on the biology of the

surrounding soft tissues. Given the complexity of the different mechanisms that take

place during cell adhesion and growth phenomenon, it is impossible to create a research

model that would include all parameters. It is not yet clear whether other material than
titanium can be suggested as a valid alternative. Even titanium could be improved by

different treatments, as has been the case for osseointegration.            In particular,

modifications of the surface topography or chemistry of the abutment could lead to better

clinical results. More research could help clarify the situation.
                                                                                        56


                 2.12.3 Impact of the surface topography and chemistry

                                   on cellular adhesion



         After implantation, the fate of the implant surface has been described as a “race

for the surface” among protein adsorption, host cell adhesion, and bacterial adhesion

(Gristina, 2004). Cell spreading and locomotion on any substrate requires that sufficient

adhesion exist between the substratum and the cell to provide the necessary traction

(Brunette & Chehroudi, 1999). One of the surface properties that have been linked to cell

attachment is wettability, expressed in dynes/cm. A surface tension of 20 to 30 dynes/cm

has been found to be the minimum to exhibit some biologic adhesiveness. Material above

these values support bioadhesion to a higher degree (Baier, 1988). It should be noted that

a low surface energy can result from either the composition of the surface, or from

contaminants, such as grease, that have been imparted to the surface during fabrication,

cleaning, sterilization, or faulty manipulation during implantation (Brunette & Chehroudi,

1999).

         Attempts to enhance dental implant surface to promote osseointegration have been

numerous. Macro-retentive features such as screw threads with specific pitch, angle or

position, press-fit designs, or sintered bead technology have been used for that purpose.
Stability of the implant was then increased even more through micro- and nano-

modifications at the surface, either through changes in topography (roughness), or

chemico-biological means to manipulate cell biology on that surface (C. M. Stanford,

2006). Previously, most implants were either minimally rough, such as the turned screw

implant, or extremely rough such as plasma sprayed implants.            Today, the most

commonly used implant surface is slightly rough, with its surface roughness (Ra value)

around 1 or 2 m (Albrektsson & Wennerberg, 2005). Sandblasting is a very common

technique to provide the desire amount of roughness for the implant surface, resulting in a
                                                                                         57


roughness average (Ra) of 2,4 m, to compare with a Ra of 0,8 m for machined titanium

(Di Carmine et al., 2003). Another technique used to increase roughness is a gentle acid

etching of the surface. It modifies the roughness at a nanometer scale and has hence the

advantage of preserving macro- (micro-) retentive areas. Combination of specific macro-,

micro-, and nano- topographic features could be developed to optimize the surface.

Description of surface roughness alone does not always accurately characterize the

implant surface (C. M. Stanford, 2006), but is nevertheless an important factor for the

implant success, and it has been stated that soft tissue integration is even more influenced

by the surface texture than the process of osseointegration (Lauer et al., 2001). Finally,

roughness and chemical treatments have to be assessed simultaneously with geometrical

micro-features of the topography of the surface such as grooves, pits, their orientation,

spacing, depth, etc…(Brunette & Chehroudi, 1999).


2.12.3.1 Roughness



       In an in vivo study, Wennerberg et al. could not demonstrate any effect on surface

roughness on soft tissue health and structure. At the first appointment of their patients,

they placed abutments of different experimental roughness on functional, loaded implants

after removal of the prosthetic framework placed on these implants. For each patient, five

abutments were placed: one control, machined titanium abutment, and four rougher

abutments, obtain either by a different machining process, or by sandblasting with three

different particle sizes. At a subsequent appointment, amount of accumulated plaque and

marginal bleeding were assessed, and a histological sample was collected. The prosthetic

framework was then connected back to the implants. Even though profilometry showed a

significant difference of topography for the five different surfaces, no significant

difference could be observed in plaque accumulation or bleeding between the five groups.

But the time after which these assessments were made, following abutments placement,
                                                                                          58


was rather short (a month only), and the results, even though not significant, could still

show a tendency toward more inflammation and bleeding for the roughest surface

(Wennerberg et al., 2003).

       Similarly, Baumhammers et al. observed similar attachment of epithelial cells on

surfaces presenting different roughness (Baumhammers, Langkamp, Matta, & Kilbury,

1978), and Di Carmine et al. could not show any statistical difference in the surface

covered by epithelial cells on machined titanium (Ra = 0,8 m) or sandblasted titanium

(Ra = 2,4 m) after 24 hours culture. But the results of this latter experience tended to

indicate that these cells covered a larger surface on sandblasted surfaces. Moreover, cells

on these rougher surfaces exhibited numerous, long branched filopodia, closely adapted

to the surface, as cells on the smoother surface (machined Ti) did not present cytoplasmic

extensions. These extensions are suggestive of strong adhesion and spreading attitude

and hence, if the attachment was not statistically greater quantitatively, it might have been

better qualitatively for rough surfaces (Di Carmine et al., 2003).

In contrast Baharloo et al. found that smooth surfaces increased the growth, spreading,

membrane proximity, and adhesion of epithelial cells. They used five different substrates

for their study: Tissue Culture Plastic (TCP), TCP coated with titanium (TCP-Ti),

polished titanium (P), acid-etched titanium (AE), grit-blasted titanium ((B), and grit-
blasted and acid-etched titanium (SLA). The average roughness values for the rougher

surfaces were 0.58 m for AE, 5.09 m for B, 4.33 m for SLA, to compare with 0.06 m

for P. They cultured epithelial cells for five days. Starting at day 3 and up to day 5, the

smooth surfaces (TCP, TCP-Ti, and P) showed a significant increase of epithelial cell

growth. No significant difference was found between the different rough surfaces (AE,

B, SLA) over the 5-day culture (Baharloo et al., 2005). They used then a high population

density for cultures up to 28 days, to assess the effects on the rough surfaces on sheets of

epithelial cells that could be considered as analogous to epithelial tissue. Here again,

smooth surfaces allowed a better growth, the cells forming a confluent layer at day 1,
                                                                                            59


starting to layer at day 3, and even ongoing exfoliation cycles at day 28. On the other

hand, cells on rougher surfaces (AE, B, SLA) remained a confluent layer with no

stratification at day 3, and at day 7, the roughest surfaces (B, SLA) were presenting

patchy areas where the cells had detached. At day 14, on these surfaces, the cells were

forming distinct colonies separated by large gaps. At day 28, gaps were still present

(Baharloo et al., 2005). The authors also assessed cell areas, measuring it for three

different categories (singlet, doublets, and clusters) to take into account the contact-

induced spreading of the epithelial cells. They could show that, in agreement with

previous studies, epithelial cells in contact with other epithelial cells exhibited a threefold

increase in cell area, but this effect was less visible for rougher surfaces AE, B, and SLA.

Accordingly, focal adhesion, measured by immuno-gold staining and detection of

vinculin, where larger and more prominent on the polished surface compared to AE, B,

and SLA. On the rough surfaces, focal adhesions tended to be localized on the ridges

rather than the valleys of the surfaces. Epithelial cells were bridging above the valleys,

from one ridge to the next one. Hence, the distance between the cells’ membranes and

the substrate was higher in average (Baharloo et al., 2005). Because the area of focal

adhesion is thought to be related to the strength of cell adhesion, it seems likely that

epithelial cells attach more strongly on smooth surface. The authors concluded that,
overall, smooth titanium surfaces are better suited for devices interfacing with epithelium

than rougher surfaces, for the smooth surfaces promote better adhesion, growth, and

closer adaptation of epithelial cells to the Ti surface (Baharloo et al., 2005).

       Lauer et al., in an in vitro experiment, obtained comparable results, suggesting

again that glossy polished titanium allowed a better surface coverage by cultured human

gingival epithelial cells, compared to sandblast or plasma-sprayed titanium of superior

roughness. But although polished titanium favored cell attachment and proliferation

quantitatively, the quality of the attachment seemed to be of lesser quality (Lauer et al.,
                                                                                         60


2001), as was suggested in the above-mentioned study from Di Carmine et al (Di

Carmine et al., 2003).

       In summary, it is not clear yet whether a smooth or a rough surface is preferable

for epithelial cells attachment. Moreover, the effect of the roughness on fibroblasts and

bacteria needs to be assessed as well, since these cells and microorganisms do influence

directly the peri-implant tissues health. The fact that epithelial cells seem to act as a

group rather than individual cells complicates further the problem. Here again, more

research might help shed light on that aspect of soft tissue integration.


2.12.3.2 Acid-etch roughening



       Besides sandblasting and polishing, many other techniques have been proposed to

modify titanium surface roughness. Among them, acid etching has been rather frequently

studied. One advantage of this technique is that it does not modify the macro- and micro-

features of the surface. In 2002, Abrahamsson et al. used dogs to assess the difference in

the soft tissue surrounding titanium abutments of different characteristics. They placed 4

implants on five dogs, and after three months, connected two different types of abutments

on each dog: two abutments were regular machined titanium abutments, and two

abutments were dual thermal-acid-etched abutments (Osseotite®).             After a 6-months

period, during which a strict oral hygiene protocol was applied, they made histometric

and morphometric observations of the abutment/soft tissue interface, as well as an

assessment of the topography of the two different types of abutments. Even though

significant differences in the surface characteristics of these two types could be

determined, the authors failed to show any difference in the structure and composition of

the soft tissues surrounding these different abutments. They could observe the traditional

epithelium and connective parts of the abutment/soft tissue interface, and again could see
                                                                                           61


the presence of an inflammatory infiltrate in regard of the implant/abutment connection

(Abrahamsson et al., 2002).


2.12.3.3 Micro-topography



       There is large evidence that the topography of the implant surface influence

motion of the cells attached on that surface (Brunette & Chehroudi, 1999).              Both

epithelial cells and fibroblasts move on the substrate they have attached to. Fibroblasts

cells move as separate units, whereas epithelial cells move in sheets. A grooved surface

will direct the global direction of the displacement of the cells. More over, fibroblasts

will move faster if they are attached to a grooved surface. This has been attributed to the

fact that the cytoskeleton filaments and microtubules of the cell were aligned with the

groove and polarized the fibroblast (Damji, Weston, & Brunette, 1996). On a smooth

surface, the direction of the motion is more erratic and reorganization of this cytoskeleton

must occur before the cell change its direction, hence a lower average speed. Conversely,

the speed of displacement for epithelial cells is not affected by a grooved surface,

probably because these cells move in sheets and polarization of the cells does not readily

occur. A noticeable fact is that fibroblasts do present contact inhibition; they stop moving

when touching another fibroblast. They present then different behaviors: they can either

reverse their direction of motion, or they can slightly “shift to the side and continue in the

initial direction, or they can stop moving altogether. Epithelial cells, on the other hand,

will tend to cover the whole surface available. The sheets will adhere to each other upon

contact. When a fibroblast encounters a sheet of epithelial cells, it can either change its

direction, or invade the sheet. On grooved surface, fibroblasts will more often exhibit a

contact inhibition behavior and change their direction, compare to a smooth surface

where invasion of the epithelial sheet is more frequent (Damji et al., 1996).
                                                                                      62


         Cheroudi et al. have shown that micro-machined grooved or pitted surfaces can

produce connective tissue ingrowth, which, in turn, inhibits epithelial down growth. They

used an original design for their experiment, allowing them to have distinctive surfaces

for the connective tissue attachment (basic deep component, named BC) and the

epithelium attachment (skin penetrating component, named SPC).         Surfaces for the

epithelium attachment were smooth, whereas surfaces for connective cells attachment

were smooth, grooved (grooves of 30 m or 19 m depth), or tapered-pitted (120 m

deep).    Besides the different surfaces for the connective tissue attachment, they

investigated the effect on a one-stage surgical approach, and a two-stage surgical

approach. These experiments were done on rats. They reported the length of epithelial

attachment and connective tissue attachment along the implant at one, two, and three

weeks. They also indicated the numbers of failure for each type of surface or surgical

technique (one or two steps).

         For one-stage percutaneous implant, as a general rule, they observed a gradual

downward migration of the epithelium, which replaced the connective tissue, during the

two weeks following implantation. By three weeks, many implants with smooth surfaces

for BC (connective component) had either failed, because of epithelial marsupialization,

or had a very short connective tissue attachment.
         For two-stage implants, one week after the second surgery, connective-tissue

attachment contacted both the deep and the superficial component. Frequent penetration

of the cells in the grooves or pits could be seen on the deep component (BC) for grooved

or pitted surfaces. At two weeks, the down-growing epithelium on implants with smooth

deep components (BC) had replaced the connective tissue attachment on the SCP

(superficial component) and part of the BC (deep component). In contrast, epithelial

attachment had just reached the junction between SCP and BC on implants with 30 m-

deep grooves or 120 m-deep pits. At three weeks, epithelial attachment was generally

found on the BC, except for implants that had 30 m -deep grooves, in which case
                                                                                         63


epithelial attachment was located at the BC/SCP junction.          Connective tissue had

penetrated deeper in the grooves and totally filled the pits. These implants with 30 m-

deep grooves appeared to have the most stable connective attachment.

       Failures after three weeks were significantly associated with smooth BC (deep

component), for both one and two-step surgical protocol.

       In brief, it could be shown that surface topography significantly affected epithelial

attachment, connective tissue attachment, and recessions. More over, at either two or

three weeks post-implantation, there was significantly less recession on two-stage

implants than on one-stage implants, when comparing the same topography. Similarly,

two-stage implants presented a significantly greater connective tissue attachment. As a

general rule, recession increased and connective tissue attachment decreased with time.

The authors concluded their study by stating that in-growth of connective tissue could

inhibit epithelial down-growth, and that this epithelial down-growth was further reduced

by using a two-stage surgical protocol (Chehroudi et al., 1992). In a similar study, the

same authors could confirmed these results, showing again that connective tissue

ingrowth and mineralized tissue could be observed on micro-grooved surfaces, whereas a

thick capsule and epithelial downgrowth was seen around the smooth implants.

Moreover, grooved surface reduced significantly recession of the soft tissues around the
implant and time before failure of the implant (Chehroudi & Brunette, 2002).

       Influence of micro-features of the topography was further analyzed by the same

group of authors, who investigated on the effect of the orientation of the grooves, their

depth and their spacing (pitch). For this study, they prepared surfaces with smaller

groove depth, precisely 30, 22, 10, or 3 m. Pitch was varied, as well as the orientation of

the grooves, either parallel to the grand axis of the implant, or perpendicular to it.

Epithelial cells presented elongated nuclei on the smooth surfaces, and also on surfaces

with vertical grooves were cells lined up with the grooves.           In contrast, on the

horizontally grooved surfaces (10 and 3 m deep), cells with elongated nuclei were
                                                                                         64


observed only over the flat ridges between the grooves, as the cells within the grooves

had rounded nuclei. As for the 22 m deep horizontal grooves, no cells could be seen

inside them. Shape of the nuclei is a good predictor of the shape of the entire cell, and

hence, elongated nuclei could translate in flat, attached, cells (Grinnell, 1978). Also, the

authors could observe that epithelial attachment was the shortest for 22 m deep

horizontal grooves, followed by 10 m deep horizontal grooves and finally 3 m deep

horizontal grooves.    All horizontally grooved surfaces presented a shorter epithelial

attachment than smooth surfaces, which in turn were less than vertically grooved

surfaces. The pitch of the groove was not found as significant. Reciprocally, the longest

connective attachment was found on the 22 m deep horizontal groove surface and the

shortest was on vertically grooved surfaces. These results translated in lower recession of

soft tissue for the horizontal deep groove surfaces, compared to the increased recession of

the vertically grooved surfaces (Chehroudi, Gould, & Brunette, 1990).

       In summary, micro-topographical features added on the abutment surface odes

influence cells behavior in vitro. It seems that this influence also exists in vivo. An

appropriate combination of macro-, micro, and nano-features could allow an optimal

response of the soft tissues. Moreover, chemical treatments of the surface, especially by

coating, could improve this response even more.

2.12.3.4 Chemical modifications



Surface modifications have been employed to prevent non-specific protein adsorption and

to prevent adhesion of microorganisms such as bacteria, or fungi, or cells. For example,

coating the implant surface with PLL-g-PEG [poly(L-lysine)-grafted-poly(ethylene

glycol)] can render the surface resistant to full human serum and to cell adhesion and

proliferation (Tosatti et al., 2003). Modifying the surface coating by using a peptide-

functionalized PLL-g-PEG instead of a non-functionalized PLL-g-PEG have been shown
                                                                                          65


to re-establish cell adhesion whilst remaining resistant to non specific protein adsorption

(Tosatti et al., 2003; VandeVondele, Voros, & Hubbell, 2003).                 This peptide-

functionalized PPL-g-PEG presents a RGD motif, which is generally not recognized by

bacteria, but will allow cell attachment through integrin receptors. Maddikeri et al. were

able to show a highly significant reduction of different bacteria strains on a titanium

surface coated with the peptide-functionalized PLL-g-PEG, compared to an uncoated

control titanium surface. This kind of adhesion-selective surface treatment is promising,

but the possible problem of coating stability is still not elucidated (Maddikeri et al.,

2008). Physical hard coating of the implant surface with titanium nitride and zirconium

nitride has also been proposed as a mean to reduce bacterial colonization and possible

subsequent peri-implant pathology (Groessner-Schreiber, Hannig, Duck, Griepentrog, &

Wenderoth, 2004). Effect of such coating on epithelial cells and fibroblasts adhesion still

needs to be clarified.

       If the nature of the acquired exogenous pellicle affects cells’ attachment and

behavior on the implant surface, it is not the only factor to have such an influence. The

topography of the surface definitively shares this capability, as shown by a study of

Shuler et al. If Maddikeri focused on the effect of PLL-g-PEG coating on smooth

polished surfaces only, Schuler et al. used smooth and rough titanium surfaces (titanium
oxide sputtered-coated on epoxy blocks) to assess the effect of coating with PLL-g-PEG,

with or without peptide-function, respectively named RGD and PEG. The control used

for their study was commercially pure titanium oxide (named Ti), and also a scramble

peptide-function, with a RDG sequence instead of RGD, named RDG. Their results were
variable, depending on the type of cells (osteoblasts, fibroblasts, and epithelial cells). As

expected, all the different types of cells attached more (numbers of attached cells) on the

bioactive surfaces (Ti and RGD) compare to the inactive surfaces (PEG and RDG),

whether smooth or rough. But for the same chemical modification, osteoblasts attached

more on rough surfaces. The authors also found a direct relationship between the surface
                                                                                           66


density of RGD and osteoblasts’ attachment, confirming that the peptide-functionalized

PLL-g-PEG restores the capability of these cells to adhere on the substrate.              If a

concentration of more than 0.67 pmol/cm2 of RGD modified PLL-g-PEG was used,

osteoblast adhesion was restored to a value similar to the one found on TiO2 surfaces.

For fibroblasts, it could be observed that they attached more on smooth surfaces. As for

the epithelial cells, the pattern followed was more erratic; the number of attached cells

would sometimes be higher on rough surfaces, as when coated with PEG, or higher on

smooth surfaces, as it is the case for surfaces coated with RDG (Schuler et al., 2006).
       In summary, these chemical modifications could help prevent bacteria adhesion

on the abutment, and reduce the risk of dental plaque related pathology, i.e. peri-

implantitis. It remains unclear whether these coatings will degrade over time, and more

research is needed to clarify this point.


2.12.3.5 Effect of dental plaque and bacterial adhesion



As mentioned earlier, bacterial adhesion on the abutment surface can lead to peri-

implantitis, but it can also affect cell attachment, even before healing is achieved.

Kawahara et al. investigated on the effect of dental plaque extract on human fibroblasts

(HGF) and epithelial cells (HGE). They cultivated these cells on polyacrylic plates

coated with a titanium film. They exposed these cultures to three different concentrations

of dental plaque extracts: without filtration, named OPE, and filtered through 5 m and

0.22   m Millipore filters, named 5FE and 0.22FE respectively. Then, they assessed

growth rate, cell morphology, and cell adhesive strength for both the epithelial cells and

the fibroblasts. They further assessed the effect of plaque extract on a co-culture of HGF

and HGE. Dental plaque extracts were shown to have a significant effect on growth rate

for both types of cell, reducing this rate. Compared to that of HGE, the growth rate of

HGF was significantly more affected. After two days of exposure to 5FE, HGF exhibited
                                                                                         67


a reduction of 53.3% of their growth rate, to compare with a reduction of 30.1% for HE.

This result denotes that HGE have a higher resistance to dental plaque than HGF (31).

Both types of cells were affected by OPE, and morphological changes were observed:

HGF were shrunken and rounded after 2 days of exposure, and the cell-cell attachment of

the pavement –like monolayer cell sheet of HGE was partially detached and separated.

The authors indicate that these morphological changes were less pronounced for HGE

than HGF, here again suggesting a better resistance of HGE to OPE (31). Moreover,

adhesive strength (expressed as a ratio between the numbers of cell attached after shear

strength and before) was reduced for both types of cells after a 2-day exposure to at any

of the three different concentrations of dental plaque extracts. This reduction was greater

for HGF than HGE after exposure to OPE (95% and 87% respectively), to 5FE (77% and

60% respectively), and to 0.22FE (18% and 4% respectively) (31). Finally, the exposure

to 5FE of a co-culture of HGE and HGF lead to a higher spread on the substrate of HGE,

whilst HGF tended to “pile up” on the epithelial cells sheet. The authors attributed the

differences between fibroblasts and epithelial cells to the intrinsic behavior of the

epithelial cells, designed to protect the underlying tissues of external pathogens invasion.

The investigators concluded that dental plaque contamination might be one of the

possible mechanisms for apical epithelialization (Kawahara, Kawahara, Hashimoto,
Takashima, & Ong, 1998). Considering these results, it seems important to assess the

effect on a surface modification not only on the cells destined to attach on it, but also on

bacteria, since the latter can influence the adhesion of the former.


                      2.13 Outcome of the review of the literature



       Taking into account the research published so far, it is clear that cell attachment

and soft tissue integration mechanisms are essentially unknown. Many problems, such as

the number of different cell types involved, make experimental designs difficult to
                                                                                      68


establish. Little is known on the combined effect of macro-, micro-, and nano-features of

the surface topography on soft tissue. Even less is known about the response of the same

soft tissues when in relation with zirconia. For these reasons, a new experiment was

implemented, in order to improve knowledge of these complex interactions. Certain

aspects of epithelial cells behavior were examined in relation with different surfaces.

Machined titanium, being considered as the standard material for abutment, was used as a

control surface. Acid etching and two other modifications of titanium surfaces were used

as possible alterations to improve epithelial cells response. Finally, a zirconia surface

was added in the experimental group, to assess if any difference could be observable,

compared to traditional titanium abutment surfaces. The experiment will be described in

details in the following section.
                                                                                           69


                                     CHAPTER 3
                         MATERIALS AND METHODS



                                3.1 Objective of the study



       Assess behavior of human epithelial cells on different dental implant abutments

surfaces.


                                        3.1.1 Aim I



3.1.1.1 Specific Aim I



       Analyze differences in epithelial cell attachment and spreading on the different

substrates using cell attachment, immunofluorescent (IF) and scanning electron

microscopy (SEM) assays.


3.1.1.2 Hypothesis



       Epithelial cells will attach identically on treated titanium and zirconia substrates.


3.1.1.3 Rationale



       Analysis of epithelial cell attachment and cell morphology to different substrates

such as hydrofluoric treated titanium or zirconia substrates will lend insight into the

development of substrates that support better soft tissue responses during implant therapy.
                                                                                         70


                                       3.1.2 Aim II



3.1.2.1 Specific Aim 2



       Assess 6 4 integrin expression in epithelial cells culture on different abutment

substrates.


3.1.2.2 Hypothesis



         6 4 integrin expression between epithelial cells cultured on different abutment

substrates will be identical.


3.1.2.3 Rationale



       Analysis of 6 4 integrin expression during epithelial cell attachment to different

substrates such as hydrofluoric treated titanium or zirconia substrates will allow a better

understanding of the molecular mechanisms utilized by epithelial cells to attach to the

transmucosal abutment. This will allow the development of abutment biomaterials that

enhance the biological response of the epithelial tissues and thus increase the prognosis of

successful treatment during implant therapy.


                                      3.1.3 Aim III



3.1.3.1 Specific Aim 3



       Assess pro-inflammatory cytokines expression in epithelial cells culture on

different abutment substrates.
                                                                                         71


3.1.3.2 Hypothesis



       Pro-inflammatory cytokines expression between epithelial cells cultured on

different abutment substrates will be identical.


3.1.3.3 Rationale



       Analysis of cytokine expression during epithelial cell attachment to different

substrates such as hydrofluoric treated titanium or zirconia substrates will allow to assess

the biocompatibility of those materials and to assess if the cells present different gene

expression depending on the substrate they have attach to.


                                       3.1.4 Aim IV



3.1.4.1 Specific Aim 3



       Analyze proliferation differences for epithelial cells culture on the five different

abutment substrates at 3 days, 5 days and 7 days.


3.1.4.2 Hypothesis



       Proliferation of epithelial cells cultured on the five different abutment substrates

will be identical at 3 days, 5 days and 7 days.
                                                                                          72


3.1.4.3 Rationale



        Analysis of proliferation of the epithelial cells on different substrates such as

hydrofluoric treated titanium or zirconia substrates will help developing materials for

dental implant abutments with an optimal proliferation rate of the epithelial cells onto

their surface.


                    3.2 Background information on the investigation
                                        techniques



        In this section the basic principles of the different techniques used for the

investigation will be briefly reviewed.      The actual materials and methods will be

described subsequently in the according section.


                                     3.2.1 Cell culture



        As will be described later, this experiment was done using human gingival

epithelial cells. These cells are not easily obtained, and therefore rat cells are often used

in similar studies. One advantage of using human cells is an easier correlation between

the observations and clinical extrapolation. The human cells used for the experiment

were immortalized by activation of human telomerase (hTERT), a ribonucleoprotein that

maintains telomere length during cell division (Blackburn; Greider, 1991). A telomere is

a region of repetitive DNA at the end of chromosomes, which protects the end of the

chromosome from destruction. Its name is derived from the Greek nouns telos (               )

"end" and mer s (μ       ) "part". During cell division, the enzymes that duplicate the

chromosome and its DNA cannot continue their duplication all the way to the end of the

chromosome.      If cells divided without telomeres, they would lose the end of their
                                                                                          73


chromosomes, and the necessary information it contains. The telomeres are disposable

buffers blocking the ends of the chromosomes and are consumed during cell division and

replenished by an enzyme, the telomerase reverse transcriptase. Telomere shortening in

humans induces replicative senescence that blocks cell division. This mechanism appears

to prevent genomic instability and development of cancer in human aged cells by limiting

the number of cell divisions.       Malignant cells that bypassed this arrest become

immortalized by telomere extension mostly due to the activation of telomerase, the

reverse transcriptase enzyme responsible for synthesis of telomeres. Telomerases are part

of a protein subgroup of specialized reverse transcriptase enzymes known as TERT

(TElomerase Reverse Transcriptases) that are involved in synthesis of telomeres in

humans.

       These immortalized cells were culture in order to obtain a sufficient number.

Classical cell culture and passaging techniques were used and will not be described here.


                                 3.2.2 Attachment assay



       As mentioned earlier, attachment of a cell is very often the first step the latter has

to take before it can multiply or differentiate, or express its potential. Sometimes, this
adhesion is not desired, as for bacteria on a catheter, or dental plaque on teeth or dental

implants. On the other hand, adhesion might be of vital importance as, for example, for

platelets on an injured blood vessel wall, in order to enable coagulation. Attachments

assays can therefore shed light on the interaction between a specific type of cells and a

specific type of substrate, in specific conditions. This potentially allows improvement of

this interaction (whether adhesion is wanted or not) by modifying the nature of the

substrate, its topography, or, if possible, the conditions at the time of interaction (pH,

temperature, etc.).
                                                                                          74


       Attachment assays can be of many forms, such as description of the shape or the

volume of the cell, quantification of the cell movements, numbers of cells attached on the

substrate, or proliferation of the cells on that substrate. For this study, a combination of

methods was used: number of attached cells, proliferation of the cells after attachment,

qualitative appearance of the cells on the substrate, and organization of the actin

cytoskeleton.   These two last methods were used combined with other techniques

(Scanning Electron Microscopy and Immuno-Fluorescence Microscopy respectively),

which will be described in the corresponding section.

       Determining number of attached cells can be done in different ways.               For

example, a specimen is seeded with a known number of cells. After cells have been left

on the surface for a sufficient time to allow attachment, the seeding media is removed

from the surface of the sample, and one can either count the cells on the specimen

surface, or count cells still in suspension in the media. The latter method allows a faster

and more precise count. The seeding media is aspirated from the specimen surface after

the required amount of time, and added to a volume of conductive liquid (Isoton). Cells

in suspension are then counted with an automatic particle counter (figure 20). This

counter detects and measures the change in electrical resistance produced by a particle or

cell suspended in the conductive liquid traversing a small aperture. The particle, or cell,
functions as a discrete insulator, and its passage through the aperture modulates the

impedance of the electrical path between two submerged electrodes located on each side

of the aperture. At the time the particle passes in the aperture, it creates a small current,

suitable for counting and sizing the particle. Once the particles in suspension are counted,

this number is retracted from the number of cells used to seed the sample, and this gives

the number of cells left attached on the disc. The time within which cells are left on the

specimen should be short enough to avoid cell multiplication as a possible bias.
                                                                                          75




Figure 20: Schematic representation of the particle counter.


Source: Instruction manual, Beckman Coulter Z1 Coulter Particle Counter.




       On the contrary, for proliferation assays, cells are left on the specimen for a longer

period of time (to allow cell multiplication), and counted at predetermined time points.
The technique differs also by the fact that for a proliferation test, the seeding media is

also aspirated, but discarded. The specimen is then flooded with a media allowing cell

detachment, generally a mix of trypsine and EDTA. This last media is collected with the

now detached cells, added to a volume of conductive liquid, and here again, run through a

particle counter. The number obtained is a direct reading of the actual number of cells

that were attached onto the specimen at a specific time point.
                                                                                        76


                          3.2.3 Scanning electron microscopy



       Scanning Electron Microscopy (SEM) is a well-known method of investigation

and will be only very briefly described here.

       Whereas an optical microscope uses photon to provide an image of the specimen,

SEM, as its name indicates, uses electrons. In the casing of the microscope, an electron

gun generates these electrons, and a first coil, or condensing lens, focuses the electron

beam. A second coil deviates the beam, which is swept over the specimen surface,

providing the “scanning part” of the process. Finally, a third coil, or focusing lens,

permits the focusing of the image. As the beam plays over the surface of the specimen,

usually coated with very conductive metal (gold or platinum), secondary electrons are

released, gathered by a detector, amplified, and the signal is transmitted to a monitor for

visualization of the specimen (figure 21).


                               3.2.4 Immunofluorescence



       This method of investigation finds numerous applications in different scientific

research fields, such as medicine, biology, histology, etc. A complete description of these
techniques is beyond the scope of this thesis, but a brief summary of the principles

follows.

       Generally, an indirect technique with two different antibodies is used. The first

one is used against the antigen of interest. Then, a secondary dye-coupled antibody that

recognizes the primary antibody is used. The fluorescent dye is observed with a specific

microscope and allows visualization of the targeted antigen.

       Direct techniques are also available, as the one used in this experiment, which

does not require an antibody, but uses phalloidin instead. Phalloidin is a toxin found in

certain mushrooms (Amanita Phalloides), and binds to actin filament.            It is used
                                                                                     77


conjugated to a fluorescent molecule, such as rhodamine isothyocyanate (RITC) or

fluorescein isothyocyanate (FITC). Here again, fluorescence is observed with a specific

microscope, and permits a qualitative assessment of the cytoskeleton of the cells.




Figure 21: Schematic representation of a scanning electron microscope.
                                                                                      78


                   3.2.5 Polymerase chain reaction and electrophoresis



       Polymerase Chain Reaction, or PCR, is a method to replicate a DNA sequence in

high quantities in a short period of time. Applications are numerous and diverse, such as

research, DNA sequencing, synthetic production of organic molecules, forensic

techniques, etc.

       First, a sample of DNA is collected. Primers are specifically designed for the

target segment of the DNA double helix. These primers are complementary to this

segment at its 5’ and 3’ ends. Therefore, two primers are needed: the forward primer

reading one strand, the reverse primer reading the other one. The DNA and nucleotides

are placed in a media, and the two strands of the DNA double helix are separated at a

specific temperature (denaturation), usually around 94-96°C.       Temperature is then

lowered, and primers will bind to their matching strand of DNA. This is the annealing

step. Under specific conditions of temperature, and the presence of the appropriate

enzyme (usually Thermus aquaticus (Taq) polymerase, a heat-resistant enzyme, or

another DNA polymerase), each strand will be paired to a new sequence of corresponding

nucleotides (extension). At the end of the cycle, the doubled-strand of DNA is now

replicated (figure 22). A new cycle can start, after which there will be twice as much
double strands of DNA (figure 23). At each new cycle, the number of strands is doubled,

leading to a very fast increase in the quantity of available DNA (figure 24). This DNA is

then present in amounts large enough to allow for research, production of specific

proteins, or other purposes.
                                                 79




Figure 22: Polymerase chain reaction; cycle 1.


Source: Cornell University website.




Figure 23: Polymerase chain reaction; cycle 2.


Source: Cornell University website.
                                                                                        80




Figure 24: Polymerase chain reaction; cycle 3.

Source: Cornell University website.




       An interesting variant of this technique is to use mRNA instead of DNA. The

mRNA contained in a cell at a specific time point is representative of the activity of the

cell at that time point; at least as far as gene expression is concerned. If the cell

membrane is broken and the mRNA collected, the latter can be reverse transcribed into

cDNA. This cDNA can, in turn, by amplified as described above. Again, specific

primers are designed, but this time for a determined type of protein, of interest for the

investigator. Therefore, these primers will allow the replication of the cDNA strand only

if it contains the appropriate gene. In other words, the cDNA will be replicated only if it

was issued from a mRNA coding for that specific protein of interest. The more mRNA

the cell had expressed for that gene at the time of membrane lysis, the more cDNA will be

present at the time of amplification, and the more corresponding replicated cDNA will be

present at the end of the cycles. This technique is sometimes referred as “expression

profiling”.
                                                                                         81


       The cDNA is then submitted to electrophoresis, and the presence of the sequence

coding for the gene of interest can be assessed in a semi-quantitative way.

Electrophoresis is a relatively simple technique where an agarose gel is placed between

two electrodes and bathed in an electrolytic solution. A current runs then from one

electrode to the other. The cDNA is dyed (usually with ethidium bromide) and placed in

the solution, and because of its negatively charged sugar-phosphate backbone, is attracted

by the cathode. The gel creates a resistance to this migration and this resistance increases

with the size of the cDNA. Hence, a cDNA with a specific size (number of base pairs)

will stop at a specific point in the gel, depending on the intensity of the current and the

concentration of agarose in the gel. This amount of cDNA creates a band at that location.

The ethidium bromide fluoresces under specific lighting conditions (UV light) and

permits visualization of the band. The intensity of the band allows for a semi-subjective

way of assessing the amount of cDNA blocked at this place. If there are several types of

proteins in the same batch, multiple bands can be obtained within the same lane.

       This technique is specifically interesting if different batches of identical cDNA are

compared. Each batch is run on a different lane. The intensity of the band obtained for

each batch allows for a direct comparison of the amount of cDNA present, and hence, for

the amount of mRNA initially present in each sample. A base-pair scale on the side of
the gel provides a mean of controlling whether this band can correspond to the cDNA, of

know number of base-pairs, coding for the sequence of interest.
                                                                                       82




Figure 25: Schematic drawing of the electrophoresis process.


Legend: 1) Agarose gel with three slots/well. 2) Injection of DNA ladder into first slot.
   3) DNA ladder injected -Injection of samples into second and third slots. 4) A current
   is applied. The DNA moves toward the positive anode due to negative charges on its
   phosphate backbone. 5) Small DNA strands move fast, large DNA strands move
   slowly through the gel. The DNA is not normally visible during the process, so a
   marker is added to the DNA to avoid it to run off the gel. 6) Color marker is added to
   the DNA ladder.

Source: Wikipedia Website, Agarose Gel Electrophoresis.




                                 3.2.6 Cytokine analysis



       Measuring the amount of cytokines or chemokines released by a batch of cells

requires high-end equipment. First, the investigated cells are covered by a culture media,

and left on their support for a certain period of time. The media (supernatant) is then

collected and cold-stored until use.
                                                                                      83


       5.6 m polystyrene beads are dyed with a specific ratio of two different red dyes.

The ratio allows a specific recognition of the type of bead. For each dye, there are 10

gradations of saturation available. Hence, by combination, up to 100 different bead types

can be created (figure 26).




Figure 26: Different combination of red dye for the beads.


Source: Luminex     Corporation Website.



       Anti-bodies to different human cytokines are attached on the beads. Each type of

bead (as determined by dye-ratio) wears antibodies to one specific cytokine. The beads

are then mixed in the supernatant and left for a certain period of time to allow for

cytokine attachment on the antibodies. Unbound material is then removed by filtration.
                                                                                       84


Anti-human multi-cytokine biotin reporter is then added and binds to the cytokines

themselves bound to the beads. Streptavidin–phycoerythrin, a fluorescent-signal emitting

molecule, is placed in the mix and in turns, binds to the anti-human multi-cytokine biotin

reporter (figure 27).




Figure 27: Different steps of a multiplex fluorescent beads-based immuno-assay.
                                                                                        85


       The solution is then run through a specific machine. As it passes through a

narrow aperture, allowing passage of just one bead at a time, the solution is illuminated

by two different lasers. The first one, a red laser, detects the beads and recognizes the

specific dye signature and therefore, the type of cytokines attached to that specific bead.

A second laser (green) reads the amount of fluorescence carried by that bead and permits

to determine the number of cytokines actually attached on the bead (figure 28).




Figure 28: Laser measures in a multiplex fluorescent beads-based immuno-assay.


Source: Luminex     Corporation Website.
                                                                                        86


       Readings are then compared to a standard curve, and concentration of each type of

cytokine in the solution is extrapolated by a software. Because of the 100 possible

different combination of dye in the beads (and hence 100 different types of beads), levels

of up to 100 different cytokines can be evaluated within the same sample.


                                3.3 Specimens preparation



                 3.3.1 Titanium and Zirconia Specimen Preparation



       This experiment was done by culture of human epithelial cells on specimens

(discs) made out of the five different materials to be investigated.

       All specimens were prepared and provided by Astra Tech for use in the proposed

study. Discs were prepared under the following parameters: they were 6.25 mm in

diameter and 2 mm thick. All were sent sterilized, in a sealed package, and ready for

tissue culture. Additionally, discs were sterilized again under UV light for 10 minutes for

each side. Five different types of surfaces were investigated: one Zirconia surface (Zir)

and four different types of titanium surfaces. The titanium surfaces were: 1) Machined

Titanium (MT); 2) Hydrofluoric acid treated titanium (HFAT); 3) Modified Titanium 1

(MODTi1); and 4) Modified Titanium 2 (MODTi2).

       A double letter labeled each type of surface: AA, BB, CC, DD, and EE (figure

29). The investigators did not know which surface corresponded to which label, and

therefore, the study was masked. Masking could not be totally achieved, the zirconium

surface being readily recognizable due to its specific white color.         Also, machined

titanium presents with a specific aspect. Nevertheless, all efforts were made to assess

differences between the discs in an objective way.
                                                                                        87




Figure 29: Groups coding.




       Sample size was variable for the different stages of the experiment and will be

specified for each of those stages. The total number of discs for the first two aims of the

experiment was 15 for each type of surface. Five batches were created, labeled YS08-2

to YS08-6, each containing 3 discs of each surface (see figure 3.11).

       Sample size was variable for the different stages of the experiment and will be

specified for each of those stages. The total number of discs for the first two aims of the

experiment was 15 for each type of surface. Five batches were created, labeled YS08-2

to YS08-6, each containing 3 discs of each surface (figure 30).
                                                                                      88


                                   3.3.2 Cell culture



       Established oral hTERT-immortalized human adenoid keratinocytes were used for

this experiment and were obtained from Dr Aloysius J. Klingelhutz (Department of

Microbiology, University of Iowa, Iowa City, IA). The characteristics of these cells have

been previously described (Farwell et al., 2000).

       Oral hTERT keratinocytes were grown in a flask containing KSFM (keratynocytes

serum free medium) supplemented by EGF (Endothelial Growth Factor); 0.16ng/mL and

BPE (Bovine Pituitary Extract); 25 ug/mL. Media also contained 10% FBS (fetal bovine

serum) and Penicillin-Streptomycin antibiotic (Penicillin; 10000 units/mL and

Streptomycin; 10000 g/mL).


                                  3.3.3 Cells passaging



       Passages numbers to obtain the necessary concentration of cells were:

Batch YS08-7 : passage 9

Batch YS08-2 : passage 4

Batch YS08-3 : passage 4
Batch YS08-4 : passage 5

Batch YS08-5 : passage 7

Batch YS08-6 : passage 8
                                                                                   89




Figure 30: Constitution of the different batches (Aims 1 to 3).




       After passaging, the media (supplemented KFSM) was removed and the flask

rinsed with PBS (phosphate buffered saline). Then 2mL of 0.05% trypsine / 0.02%

EDTA was added to detach the cells, and the flask was incubated for 5 minutes at 37°C.
                                                                                       90


The cells and the 2mL of trypsine-EDTA were then pipetted , and added into 6mL of

supplemented KSFM. 100 L of this mix, called Single Cell Suspension (SCS) was then

taken and placed it into 9.9mL of Isoton. Cells were then counted with a Beckman

Coulter Z1 Coulter Particle Counter.         Two measures were taken and averaged, to

determine the number of cells per milliliter. The SCS was then diluted adequately to

obtain a concentration of 50,000 cells/mL.


                               3.4 Cell attachment assays



       For each surface, one batch of three discs was prepared by placing the three discs

in plastic tubes of matching diameter. The Tygon tubes were treated to prevent cell

attachment on them.     For that purpose, they were cleaned with RBS 35 detergent

concentrate (Pierce Chemicals), diluted at 20mL/ 1L H2O, for 5 minutes, followed by a

15-minute rinse with running tap water. After that, Aquasil Siliconizing Fluid (Pierce),

diluted to 1:100, was added for 1 minute. This product is an organosilane compound,

which coats the surface of the tubing and prevents cellular attachment. It was then rinsed

with running tap water for 20 minutes and again with tap distilled water for 10 minutes.

After the last rinse, tubes were air-dried on paper towels overnight. Finally, tubes were

gas sterilized, discs forced into them and the gap between the both sealed with melted

parafilm. The upper surface of the discs was then flooded with 0.5 ml of SCS (50,000

cells/mL). Two 0.5mL samples of the SCS were saved in the middle of the process, and

mixed in 9.5mL of Isoton, to check the actual concentration of the cells in the diluted

media. These two samples were called Post dilution Reference (PDR) 1 and 2. The SCS

was left on the discs, and the latter placed in an incubator (37°C, 5%CO2) for 2 hours to

allow for the cells to attach. The SCS was then removed from each wells and added to

8.5mL d’Isoton in a culture cup. Discs were immediately rinsed twice with 0.5mL PBS.

These PBS increments were pipetted each time and added in the culture cup. Final
                                                                                       91


volume of the sample for each disc was therefore 10mL. After this procedure, discs were

immediately covered with 0.5mL of supplemented KSFM.                 The unattached cells

contained in the collected media were then counted with the Beckman Coulter Z1 Coulter

Particle Counter. Taking into account the real concentration of the cells in the media

(determined by averaging number of cells in PDR1 and 2), number of attached cells was

deducted from number of cells still in suspension. This experiment was repeated for the

five batches (YS08-2 to YS08-6), each time with all three discs per surface. Hence the

sample size was 15 discs for each of the five surfaces.


                              3.5 Cytokines-chemokines assay



       After the attachment assays, the same 15 discs for each surface (75 discs total)

were immediately flooded with 0.5mL supplemented cell-free KSFM and left for 24

hours into the incubator at 37°C. After this incubation period, the KSFM was collected

from each well and 0.5mL PBS dispensed in them. KSFM samples were placed in a

freezer at -20°C until use.

       Cytokines and chemokines concentrations (pg/mL) in cell culture supernatants

were determined using commercial multiplexed fluorescent bead-based immunoassays

(Millipore, Billerica, MA) in the Luminex 100 IS Instrument (Luminex, Austin, TX) as

previously described (Pingel, LC, et al., 2008). Kit 48-002 was used which detects IL-

12[p70] (Th1 cytokine), IL-10 (Th2 cytokine), IL-1 , IL-6, TNF           (pro-inflammatory

cytokines), and IL-8 (chemokine).       Briefly, 50       L cell culture supernatants were

incubated with anti-human multi-cytokine beads at 4°C for 18 hours. Unbound material

was removed by filtration. 25       L of anti-human multi-cytokine biotin reporter was

added, and reactions were incubated at room temperature for 1.5 hours in the dark.

Twenty-five    L of streptavidin–phycoerythrin was then added, and the plates were

incubated at room temperature for an additional 30 minutes. Twenty-five L of stop
                                                                                       92


solution was added, and the plates were read in a plate reader (Model 100 IS, Luminex,

Austin, TX).     Concentrations of cytokines in each sample were extrapolated from

standards (2.3-to-5000 pg/mL) using Beadview software (Millipore, Billerica, MA).

       After this cytokine/chemokine-expression assay, the same discs were used for the

other stages of the experiment: SEM, immunofluorescence and PCR assays. Repartition

of the discs for these three assays will be described in the corresponding sections.


                           3.6 Scanning Electronic Microscopy



       For the SEM, four discs for each surface were used, from batches YS08-2 and

YS08-3 (two discs for each surface per batch). At 24 hours, media used for the previous

assay (cytokine/chemokine expression) was removed, and discs were rinsed twice with

PBS. Cells were fixed on the discs using a mix of 3% formaldehyde/3% glutaraldehyde

for 30 minutes. Then, discs were rinsed with sodium cacodylate buffer (0.2M) for 5

minutes, covered with fresh cacodylate buffer and stored at 4°C until further use. They

were then dehydrated in graded ethanol: 30%, 50%, 75%, 90%, 95%, and then 2 times in

100%, for 10 minutes at each concentration. Discs were further treated twice with 100%

hexamethydisilazane for 5 minutes, and air-dried in a vacuum dessicator. An Au/Pd

coating was applied to the specimen surface with an Emitech K550 sputter coater. The

specimens were examined in a Hitachi S-4800 scanning electron microscope at an

accelerating voltage of 3.0 KV and a working distance of 3.1 mm, at a magnification
ranging from 1000X to 50,000X. Spreading and morphology of the cells were assessed

qualitatively for each type of surface.
                                                                                     93


                               3.7 Immunofluorescence



       For this immunofluorescence assay, two discs for each surface were used from

batches YS08-2 and YS08-3 (one disc for each surface per batch). Actin filaments were

stained with phalloidin conjugated to a fluorophore (FITC, Molecular Probes) in order to

analyze phenotypes of cells cultured on the various types of discs, using previously

described methods (Schneider and Burridge, 1994). Briefly, discs were flooded with PBS

just after the cytokine/chemokine assay. Then, the cells on the discs were fixed in 3.7%

formaldehyde (10mL 37% formaldehyde, 90mL PBS) for 10 minutes. Discs were then

flooded with Universal Buffer for 5 minutes (UB; 30mL 5M NaCl, 50mL 1M Tris

(pH7.6), 10mL 10% Na Azide, 910mL dH20), and placed in 0.5% Triton-x (0.5mL

Triton-X 100, 99.5mL UB), for permeabilization of the cellular membranes, for 7

minutes. Afterwards, Phalloidin-RITC (1:500) was applied for 60 minutes. Finally, discs

were covered with UB and stored at 4°C until use. At that time, discs were rinsed with

deionized water, air-dried and then mounted on microscope slides. Images were captured

with an Olympus BX40 microscope equipped with epifluorescence optics at

magnifications ranging from 20X to 600X.


                        3.8 Polymerase Chain Reaction (PCR)



                                 3.8.1 RNA extraction



       All discs of batches YS08-4 to YS08-6 were used for that experiment. At 24

hours, after the cytokine/chemokine assay, media was removed and the discs were

covered with 0.5mL PBS. RNA was first extracted and isolated from the cells attached
                                                                                      94


on the discs by using the RNeasy-mini Kit from Qiagen. RNA was then eluted from the

filter using 50 L of 0.1mM EDTA and stored at -20°C until further use.


                                3.8.2 Primers sequences



       Primers sequences were obtained from NCBI website (figure 31). Accession

number was NM_001005619 for Homo Sapiens Integrin beta4 ( 4) and NM_001079818

for Homo Sapiens Integrin alpha6 (ITG 6). GAPDH (control) primers were designed by

Applied Biosystem. The primers were obtained from Integrated DNA Technologies, Inc.


                               3.8.3 Reverse transcription



       The RNA was then reverse transcribed into cDNA using an Applied Biosystems

Taqman Reverse Transcription Reagents kit (using Random Hexamers). Incubation was

done at 25°C for 10 minutes, followed by reverse transcription at 48oC for 30 minutes and
inactivation at 95°C for 5 minutes.

       Then, both amplicon targets ( 6 and       4) and the endogenous rRNA control

(GAPDH) were amplified in separate tubes by PCR with AmpliTaq Gold DNA

Polymerase.   The thermal cycling parameters were 95oC for 10 minutes to activate

AmpliTaq Gold DNA Polymerase, followed by 40 cycles of 95oC for 15 seconds

(denaturation of DNA) and 60oC for 1 minute (annealation and extension). PCR reactions
were performed in a PTC-200 MJ Research cycler.
                                                                                        95




Figure 31: Primers sequences for PCR.




                                  3.8.4 Electrophoresis



       The PCR products were separated on an agarose gel (Ultra Pure DNA Grade

agarose; Bio-Rad) containing ethidium bromide (0.125μg/ml), visualized on a

Transilluminator (Fisher, Pittsburgh PA) and photographed with a Photo-Documentation

Camera and Hood (Fisher).


                3.9 Recapitulation of the repartition of the discs in the
                            different groups (aims I to III)



       Figures 32 to 34 summarize the repartition of the discs in the different groups, and

the sample size for each group.
                                                                        96




Figure 32: Samples repartition; aim I – cell attachment assay.




Figure 33: Samples repartition; aim III – cytokines-chemokines assay.
                                                                                        97




Figure 34: Samples repartition; aim I - SEM and IF assays and aim II – PCR.


                                 3.10 Proliferation tests



       Again, the same five types of surfaces were investigated, and Tissue Culture

Plastic was added this time. Discs were not mounted in tubes, but placed directly into 96

well-culture plates. Proliferation tests were done at three days, five days and seven days.

For each time point, three discs per surface type and 3 wells (TCP) were used.

       Discs and empty wells were each seeded with 30 L of KSFM containing 25,000

cells. This volume was spread on the whole disc surface. After waiting an hour to allow

for cell attachment, 300 L of KSFM was added to flood the well.

       At required time points, KFSM was aspirated from each well and the wells rinsed

with 300 L PBS. The latter was also aspirated and the wells flooded with 300 L 0.25%

trypsin-0.1% EDTA for 15 minutes at 37°C in the incubator to detach the cells. An

optical microscope was used to verify that cells were detached, using the TCP wells as

controls. The trypsin-EDTA mix was then pipetted 12 times to insure separation of the
                                                                                          98


cells, and mixed into 9.1mL of Isoton. Wells were rinsed twice with 300 L of Isoton that

was pippeted 4 times to ensure final detachment of the cells. Both increments of Isoton

were added to the previous trypsine-EDTA/Isoton mix. This solution was then run

through the Beckman Coulter Z1 Coulter Particle Counter to obtain the number of cells in

each well, and cell density on the disc surface was calculated.


                                  3.11 Statistical method



       An overall 0.05 level of Type I error was set for the statistical analysis. For the

attachment assay, a mixed model approach was used for statistical analysis to evaluate

surface effects: surface type was treated as a fixed effect, and run (batch) was treated as a

random factor.    Factor effects are considered to be fixed if the levels in the study

represent all possible levels of the factor about which inferences are to be made. In this

instance, we were interested in whether or not the five surface types differed in terms of

percentage cell attachment. Factors are considered to be random if the levels used in the

study represent only a random sample of a larger set of potential levels. In this instance,

we regarded the runs as representing a random sample from the universe of such runs that

could have been completed and evaluated. In such instances, there was no particular

interest in estimating the specific outcome for a given run, but rather in taking into

consideration the variability among runs in the context of the analysis. Residual analyses

were conducted to evaluate conformance to model assumptions. The MIXED procedure

of SAS® Version 9.1 (Littell, Milliken, Stroup, Wolfinger, & Schabenberger, 2006) was
used to fit these models and test these hypotheses.

       For cytokines levels measurements again, where appropriate, a mixed model

approach was used for statistical analysis to evaluate surface effects on the level of

production of a particular cytokine: surface type was treated as a fixed effect, and run

(batch) was treated as a random factor. Residual analyses were conducted to evaluate
                                                                                      99


conformance to model assumptions. Data transformations were considered as needed.

The MIXED procedure of SAS® Version 9.1 (Littell et al., 2006) was used to fit these

models and test these hypotheses. This approach was taken for the evaluation of IL-12,

IL-6 and IL-8.

       However, for two of the cytokines evaluated (IL-1B and TNF ), there were

multiple instances where the cytokine was present at levels too low to be measured

(coded as 0), necessitating a different approach.        Two statistical procedures were

considered: The first was the approach of Conover and Iman (Conover & Iman, 1976;

Conover & Iman, 1981) which utilizes rank-transformations in the context of standard

linear modeling approaches specifying surface and run effects. This rank approach is

considered very robust, and represents a bridge between parametric and nonparametric

methods. The second was the application of Friedman’s test to the median cytokine level

based on the triplicate measurements in each surface-run combination. This provides a

nonparametric test of surface effect after adjustment for run effect.

       Except in two instances when a putative outlier (BB at day 5 and CC at day 7) was

removed from the data, each mean count was based upon three replicate measurements.

       Adjustments had to be made to account for the multiple comparisons:

       Within each day, all pairwise mean comparisons among six surfaces were made,

i.e. 3x15=45.

       Within each surface, all pairwise mean comparisons among 3 days were made, i.e.

6x3=18.

       These 73 multiple comparisons were taken into account by adjustment using the

Holm modification of the Bonferroni method, in conjunction with an overall Type I error

of 0.05.
                                                                                         100


                                      CHAPTER 4
                                       RESULTS



       In this section, results for each of the assays will be presented, starting with the

assays leading to quantitative data allowing for a formal statistical analysis. The other

assays, with a more qualitative and semi-subjective approach, will be presented next.

Therefore, the outline of this section does not follow the chronology of the assays

themselves, nor does it the order of the Materials and Methods section.


                                  4.1 Attachment assays



       For this assay, the blinded cell surface designations used were AA, BB, CC, DD,

and EE. Three replicate disks were evaluated for each of the five surfaces within each of

five runs (batches), for a total of 15 evaluations per surface and 75 assessments total. Cell

attachment for a given disk was measured as percentage of cells attached.


                                 4.1.1 Statistical Methods



       A mixed model approach was used for statistical analysis to evaluate surface

effects: surface type was treated as a fixed effect, and run (batch) was treated as a

random factor. Residual analyses were conducted to evaluate conformance to model

assumptions. The MIXED procedure of SAS® Version 9.1 (Littell et al., 2006) was used
to fit these models and test these hypotheses.
                                                                                       101


                                        4.1.2 Results



          While significant run effects were identified (p<0.001), there was no evidence of

differences in levels of cell attachment among the five surface treatments (p=0.11).

Results for individual batches are shown in the appendices section (6.1).
          The least squares means (i.e. means adjusted for run effects) and their estimated

standard errors are given in table 1.



                                 Standard

Surface                          Estimate                       Error

AA                               56.6362                        3.7504

BB                               64.4412                        3.7504

CC                               62.5390                        3.7504

DD                               62.5772                        3.7504

EE                               59.2875                        3.7504




Table 1: Least squares means and their estimated standard errors for the attachment assay.




          Figure 35 shows the combined results for the five batches for each of the five

surfaces.
                                                                                     102




Figure 35: Combined results for the five batches for each of the five surfaces for the
          attachment assay.




                          4.2 Cytokines levels measurements



       The object of the present analyses was to compare the effect of the five different

surfaces on cytokine production of the epithelial cells attached on the discs. For this

purpose, after the attachment assay, cells were left on the discs, in the appropriate

media, for 24 hours. Levels of cytokines released in the media were then measured.

Cytokines of interest were IL-1, 6, 8, 10 and 12, TNF- .       All cytokine levels were

measured in pg/mL.


                               4.2.1 Statistical Methods



       Where appropriate, a mixed model approach was used for statistical analysis to

evaluate surface effects on the level of production of a particular cytokine. Residual

analyses were conducted to evaluate conformance to model assumptions.               Data

transformations were considered as needed. The MIXED procedure of SAS® Version
                                                                                       103


9.1 (Littell et al., 2006) was used to fit these models and test these hypotheses.   This

approach was taken for the evaluation of IL-12, IL-6 and IL-8.

       However, for two of the cytokines evaluated (IL-1B and TNF ), there were

multiple instances where the cytokine was present at levels too low to be measured

(coded as 0), necessitating a different approach.       Two statistical procedures were

considered: The first was the approach of Conover and Iman (Conover & Iman, 1976;

Conover & Iman, 1981) which utilizes rank-transformations in the context of standard

linear modeling approaches specifying surface and run effects. The second was the

application of Friedman’s test to the median cytokine level based on the triplicate

measurements in each surface-run combination. This provides a nonparametric test of

surface effect after adjustment for run effect.


                                        4.2.2 Results



       Raw results for cytokines levels measurements are provided in the appendices.


4.2.2.1 IL-1B



       There were multiple instances where the cytokine was present at levels too low to

evaluate; in fact, 60 of the 75 measurements (80%) fell into this category.     For this

reason, the rank-transform approach of Conover and Iman (1976, 1981) using all data

points, and the nonparametric Friedman test utilizing the median response within a

surface-run combination were used, as outlined previously.       Neither the rank-based

modeling (p=0.09) nor Friedman’s test (p=0.24) provided evidence of a difference in

responses among surfaces after adjustment for run effects. The existence of run effects

was supported by the Conover-Iman approach (p=0.002). The median cytokine level for
                                                                                     104


each surface (pooling over all runs) was zero. (Zero values were used to denote the “too

low to measure” responses).

       Figure 36 shows the average level of IL-1B for each of the five surfaces. Vertical

bars represent standard deviation. It is worth reminding that only 15 samples could be

taken into account for this cytokine, and that these 25 discs were unevenly distributed

among the surfaces (5 discs for AA and CC, 3 discs for DD, and 1 disc for Bb and EE).




Figure 36: Average level of IL-1B for each of the five surfaces.




4.2.2.2 IL-6



       There appeared to be some slight deviation from model assumptions using the

untransformed data, so possible normalizing transformations were explored. The square

root transformation appeared to address potential non-normality, although there did
                                                                                       105


appear to be a single outlier; analyses were repeated after exclusion of that single

observation.

          Conclusions were entirely consistent whether or not transformation and/or outlier

removal were employed. In all instances, significant run effects were identified (p< 0.002

in all instances).

          Based upon the analysis of the untransformed data, there was no evidence of

differences in levels of IL-6 among the five surface treatments (p=0.56). These results

were consistent with those obtained after square root transformation (p=0.62) and

following square root transformation with outlier removal (p=0.45).

          The least squares means in pg/mL (i.e. mean values on the untransformed scale,

adjusted for run effects) and their estimated standard errors are given in table 2.

Following square root transformation, the least squares means in pg/mL (i.e. mean

values on the transformed scale, adjusted for run effects) and their estimated standard

errors are given in table 3.

          Following square root transformation and outlier removal, the least squares

means in pg/mL (i.e. mean values on the transformed scale, adjusted for run effects) and

their estimated standard errors are given in table 4.



                                Standard
Surface                         Estimate                       Error
AA                              43.8133                        8.1911
BB                              46.6867                        8.1911
CC                              47.2860                        8.1911
DD                              43.7867                        8.1911
EE                              35.5427                        8.1911



Table 2: Least squares means in pg/mL and their estimated standard errors for Il-6
          expression.
                                                                                     106


                                 Standard

Surface                          Estimate                        Error

AA                               6.3831                          0.6363

BB                               6.7212                          0.6363

CC                               6.6307                          0.6363

DD                               6.4897                          0.6363

EE                               5.8193                          0.6363




Table 3: Least squares means in pg/mL and their estimated standard errors for IL-6
          expression following square root transformation.




                                 Standard

Surface                          Estimate                        Error

AA                               6.3831                          0.6206

BB                               6.7212                          0.6206

CC                               6.9027                          0.6242

DD                               6.4897                          0.6206

EE                               5.8193                          0.6206




Table 4: Least squares means in pg/mL and their estimated standard errors for IL-6
          expression following square root transformation and outlier removal.




          Figure 37 shows the average level of IL-6 for each of the five surfaces.
                                                                                          107




Figure 37: Average level of IL-6 for each of the five surfaces.




4.2.2.3 IL-8



       There appeared to be a deviation from model assumptions of normality, as well

as some obvious outliers, using the untransformed data. However, after removal of two

outliers, diagnostics indicated that assumptions of normality were not unreasonable.

Results changed very little when the outliers were removed.

       Conclusions were entirely consistent whether or not outliers were removed. In

both sets of analyses, significant run effects were identified (p< 0.0001 in all instances).

       Based upon the analysis of the full data set, there was no evidence of differences

in levels of IL-8 among the five surface treatments (p=0.46). The significance probability

for the analogous results after outlier removal was p=0.44.

       Based on the full data set, the least squares means in pg/mL (i.e. mean values

adjusted for run effects) and their estimated standard errors are given in table 5. After
                                                                                      108


removal of two putative outliers, the least squares means in pg/mL (i.e. mean values

adjusted for run effects) and their estimated standard errors are given in table 6.



                               Standard

Surface                        Estimate                        Error

AA                             496.22                          180.88

BB                             581.40                          180.88

CC                             663.57                          180.88

DD                             610.07                          180.88

EE                             483.73                          180.88




Table 5: Least squares means in pg/mL and their estimated standard errors based on the
          full data set.




                                        Standard

          Surface                       Estimate                        Error

          AA                            496.22                          180.77

          BB                            506.04                          181.42

          CC                            663.57                          180.88

          DD                            610.07                          180.88

          EE                            483.73                          180.88




Table 6: Least squares means in pg/mL and their estimated standard errors after removal
           of two putative outliers.
                                                                                    109


       Figure 38 shows the average level of IL-8 for each of the five surfaces.




Figure 38: Average level of IL-8 for each of the five surfaces.




4.2.2.4 IL-10



       All IL-10 measurements were designated as “too low to be measured” and so no

formal analysis was possible.


4.2.2.5 IL-12



       While significant run effects were identified (p=0.002), there was no evidence of

differences in levels of IL-12 among the five surface treatments (p=0.92).
                                                                                      110


          The least squares means in pg/mL (i.e. means adjusted for run effects) and their

estimated standard errors are given in table 7.



                                 Standard
Surface                          Estimate                       Error
AA                               1.4447                         0.06789
BB                               1.4560                         0.06789
CC                               1.4900                         0.06789
DD                               1.4387                         0.06789
EE                               1.4780                         0.06789



Table 7: Least squares means in pg/mL and their estimated standard errors for IL-12
          expression.




          Figure 39 shows the average level of IL-12 for each of the five surfaces.




Figure 39: Average level of IL-12 for each of the five surfaces.
                                                                                       111


4.2.2.6 TNF-



          There were multiple instances where the cytokine was present at levels too low to

evaluate; this occurred in 7 of the 75 measurements (9.3%). For this reason, the rank-

transform approach of Conover and Iman (1976, 1981) using all data points, and the

nonparametric Friedman test utilizing the median response within a surface-run

combination were used, as outlined previously.            Neither the rank-based modeling

(p=0.46) nor Friedman’s test (p=0.76) provided evidence of a difference in responses

among surfaces after adjustment for run effects.          The existence of run effects was

supported by the Conover-Iman approach (p<0.0001).

          The median, minimum and maximum of the cytokine levels within each surface

treatment (pooling over all runs) are given in table 8.



Surface                         Median                          Minimum -- Maximum

AA                              1.31                            0 – 2.06

BB                              1.59                            0 – 2.53

CC                              1.78                            0 – 2.43

DD                              1.59                            0 – 2.25

EE                              1.41                            0 – 2.34




Table 8: Median, minimum and maximum of the TNF-                 levels within each surface
          treatment (pooling over all runs).




          Figure 40 shows the average level of TNF- for each of the five surfaces.
                                                                                         112




Figure 40: Average level of TNF- for each of the five surfaces.




                                 4.3 Proliferation assays



       The object of this analysis was to compare the effect of the five different surfaces

on proliferation after 3, 5, and 7 days, of the epithelial cells attached on the discs. Also,

Tissue Culture Plastic wells were included as a sixth type of surface, serving as a control.

At each time point, cells were suspended in the culture wells, the media removed and the

cells counted.


                                 4.3.1 Statistical methods



       Except in two instances when a putative outlier (BB at day 5 and CC at day 7) was

removed from the data, each mean count was based upon three replicate measurements.

       Adjustments had to be made to account for the multiple comparisons:

       Within each day, all pairwise mean comparisons among six surfaces were made,

i.e. 3x15=45.
                                                                                          113


          Within each surface, all pairwise mean comparisons among 3 days were made, i.e.

6x3=18.

          These 73 multiple comparisons were taken into account by adjustment using the

Holm modification of the Bonferroni method, in conjunction with an overall Type I error

of 0.05.


                                         4.3.2 Results



          The data shown below provide strong evidence (p<0.0001) of interaction between

the two factors, DAY and SURFACE. This indicates that both factors were important

and affected proliferation in a non-additive manner.          The proportion of variability

explained by this interaction model (R2) was 96.8%.
          Tables 9 to 11 present the results at 3, 5, and 7 days respectively. Within a table,

means sharing a same letter (between parentheses) are not significantly different after

adjustment for multiple comparisons using the Holm modification of the Bonferroni

method, in conjunction with an overall 0.05 level of Type I error.



Surface                                          Mean count

EE                                               1441.00        (a)

AA                                               1294.67        (a); (b)

TCP                                              1251.33        (a); (b)

BB                                               911.67         (a); (b)

DD                                               678.00         (a); (b)

CC                                               368.33               (b)




Table 9: Comparisons of proliferation levels among 6 surfaces at day 3.
                                                                                     114


Surface                                       Mean count

AA                                            2796.67       (a)

TCP                                           2759.67       (a)

EE                                            2191.33       (a); (b)

BB                                            1576.50       (b); (c)

DD                                            1382.00       (b); (c)

CC                                            807.00        (c)




Table 10: Comparisons of proliferation levels among 6 surfaces at day 5.




Surface                                       Mean count

TCP                                           3979.33        (a)

AA                                            3585.33        (a)

EE                                            3207.33        (a)

CC                                            2109.50        (b)

BB                                            2105.33        (b)

DD                                            1531.67        (b)




Table 11: Comparisons of proliferation levels among 6 surfaces at day 7.


           Tables 12 to 17 represent the comparison between each time point for the six

different surfaces. Within a table, means sharing a same letter (between parentheses) are

not significantly different after adjustment for multiple comparisons using the Holm

modification of the Bonferroni method, in conjunction with an overall 0.05 level of Type

I error.
                                                                              115


Day                                         Mean count
7                                           3207.33        (a)
5                                           2191.33        (b)
3                                           1441.00        (c)



Table 12: Comparisons of proliferation levels among 3 days for TCP surface.




Day                                         Mean count
7                                           3585.33         (a)
5                                           2796.67         (a)
3                                           1294.67         (b)



Table 13: Comparisons of proliferation levels among 3 days for AA surface.




Day                                         Mean count
7                                           2105.33         (a)
5                                           1576.50         (a); (b)
3                                           911.67          (b)



Table 14: Comparisons of proliferation levels among 3 days for BB surface.


Day                                         Mean count
7                                           2109.50         (a)
5                                           807.00          (b)
3                                           368.33          (b)



Table 15: Comparisons of proliferation levels among 3 days for CC surface.
                                                                                      116


Day                                          Mean count
7                                            1715.67         (a)
5                                            807             (a)
3                                            368.33          (a)



Table 16: Comparisons of proliferation levels among 3 days for DD surface.




Day                                          Mean count
7                                            3207.33         (a)
5                                            2191.33         (b)
3                                            1441.00         (b)



Table 17: Comparisons of proliferation levels among 3 days for EE surface.




       Figure 41 presents the results of the mean counts for the six surfaces at each time

point. Figure 42 presents the results of the mean count for the three days for each of the

six surfaces. For both graphs, the two putative outliers were removed from the results,

and the adjustment above described was made.

       From the tables and graphs obtained from this assay, it can be observed that at day

3, the only significant difference in proliferation rate was between CC and EE, the latter

presenting a higher rate. At day 5, TCP showed a significantly higher rate than any other

surface, except for EE. EE had also again a significantly higher rate than CC. At day 7,

two distinct pools could be observed: TCP, AA, and EE showed higher proliferation rates

than BB, CC and DD.
                                                                                      117




Figure 41: Comparisons of proliferation levels for six surfaces at days 3, 5, and 7. Means
           joined by a bar are not significantly different.




Figure 42: Comparisons of proliferation levels for days 3, 5, and 7 at each surface.
          Means joined by a bar are not significantly different.
                                                                                     118


       From a time point perspective, TCP was the only surface showing a significant

difference at each of the time points.        DD, on the contrary, showed no significant

difference between any of these points. BB had a slightly less marked pattern, where

there was no difference between day 3 and 5, and no difference between day 5 and 7, but

nevertheless a significant difference between day 3 and 7. CC and EE had no difference

between day 3 and 5, but a significant difference between these two time points and day

7. As for AA, there was a significant difference between day 3 and the two others, but

none among day 5 and 7.


                               4.4 Integrin 6 4 expression



       After isolation of the mRNA of the cells attached on the discs, cDNA was

obtained by reversed transcription and duplicated with PCR techniques using primers

targeting 6, 4, and GAPDH genes. The cDNA was then run through an agarose gel by

electrophoresis and pictures were taken to assess the presence of a band for each surface

type, as well as the intensity of the band.




Figure 43: Result of electrophoresis.
                                                                                        119


       It can be seen from figure 43 that both subunits ( 6 and 4) of the integrin

receptor, as well as the control gene (GAPDH) were expressed for each surface type, and

that there seem to be no significant difference in the intensity of the band between each

surface for each of these three genes. Therefore, it is possible to consider that expression

of integrin 6 4 by the hTERT cells attached on the discs is identical regardless of the

surface type.


                             4.5 Immunofluorescence assays



       Fluorescent-labeled Phalloidin was used to mark the actin molecules of the

epithelial cells’ cytoskeleton. No quantitative analysis was done for this part of the

experiment, which aim was more of an observational kind. Therefore, statements such as

average, low, or high density, just reflect the overall impression of the examiner, and

should be considered with circumspection. Moreover, no formal randomization of the

areas observed on each disc was done, and again, this assay was merely aimed at gaining

some insight on the actin distribution within the cells.

       The lowest magnification (20X) allowed observing almost the entire surface of the

disc and the pattern of cell repartition on its surface (figures 44 to 48). AA presented a

rather regular pattern of distribution over its surface, with a density that appeared to be

average. BB distribution was more difficult to assess because of the scarcity of the cells

on this surface. CC showed an annealed pattern, with most of the cells laid at the

periphery of the disc, close to its margin. DD showed a somewhat similar pattern, but

with a second concentration of cells at the center of the disc. Both CC and DD surfaces

presented with an average density of cells. Finally, EE presented a rather regular pattern,

similar to AA, with a cell density seemingly higher.
                           120




Figure 44: Disc AA, x20.




Figure 45: Disc BB, x20.




Figure 46: Disc CC, x20.
                                                                                      121




Figure 47: Disc DD, x20.




Figure 48: Disc EE, x2o.




       Higher magnification (400X, figures 49 to 53) allowed observation of the general

contour of the cells, as well as observation of the cell-cell relations.      No marked

differences could be observed between the different surfaces. Generally, cells were of

similar size and had a rather round contour, except for DD cells which presented a

slightly more irregular shape. Also, CC cells, even though of rather round contour,

presented a slightly more jagged outline than AA, BB and EE. It could also be observed

than when cells were clustered, they tended to take a more polygonal form (picture 4.6).
                            122




Figure 49: Disc AA, x400.




Figure 50: Disc BB, x400.




Figure 51: Disc CC, x400.
                                                                                        123




Figure 52: Disc DD, x400.




Figure 53: Disc EE, x400.




       At the highest magnification (600X), the previous observations were confirmed.

DD and CC cells presented a more irregular contour than BB, AA and EE cells. For these

two last groups (AA and EE), cells were decidedly of rounder shape. Actin distribution

in all the cells was rather consistent between surfaces, with a well-defined peripheral ring

of actin, except maybe for AA cells, where a slightly more diffuse distribution of the actin

could be seen (figures 54 to 58).
                            124




Figure 54: Disc AA, x600.




Figure 55: Disc BB, x600.




Figure 56: Disc CC, x60.
                                                                                  125




Figure 57: Disc DD, x600.




Figure 58: Disc EE, x600.




                     4.6 Scanning Electrons Microscopy assays



       The very same remarks than for immunofluorescence assays can be made here.

No quantitative analysis was done for this part of the experiment, which aim was again

more of an observational kind.    Here again, no formal randomization of the areas
                                                                                         126


observed on each disc was done, and this assay was merely aimed at gaining some insight

on the appearance of the cells attached on the discs.

       No cells could be observed on any of the EE discs (figure 66) prepared for this

SEM assay. Hence, no assessment on the morphology of the cells could be done for this

type of surface. But in some instances, cell remnants could still be seen (figure 67).

       The four remaining type of surface presented cells with similar aspect (figures 59

to 62). The observed size of these epithelial cells was in the 25-30 m-range. Numerous

filopodias could be seen, radiating from the cell bodies. Cells were well spread on their

substrates, maybe a little bit less for AA surfaces, where again, the cells appeared to have

a slightly rounder shape (figure 59). AA cells nevertheless seemed to be “strongly”

anchored on the surface when considering the number and the repartition of the filopodias

(figure 60 and 61). Bridging between the cells could be observed in some occasions

(figure 64).




Figure 59: Disc AA, x1500.
                              127




Figure 60: Disc AA, x4500.




Figure 61: Disc AA, x20000.




Figure 62: Disc BB, x3500.
                             128




Figure 63: Disc CC, x3500.




Figure 64: Disc CC, x3500.




Figure 65: Disc DD, x3500.
                              129




Figure 66: Disc EE, x3500.




Figure 67: Disc EE, x10000.
                                                                                130


                                  CHAPTER 5
                     DISCUSSION AND CONCLUSION



                                   5.1 Discussion



      In this in vitro experiment, attachment and biology of human gingival

epithelial cell on five different surface candidates for dental implants abutments

were investigated.     Investigations techniques such as attachment assays,

proliferation assays, scanning electron microscopy, immuno-fluorescence,

polymerase chain reaction and electrophoresis, and cytokines expression

measurements, were used.

      The surface candidates were: a polished titanium surface serving as a

control, one acid-etched titanium surface, two modified titanium surfaces, and

one zirconia surface. The surfaces were received and investigated in a masked

approach. At the time this manuscript was written, the manufacturer (Astra) did

not release any further information about the different surfaces. Therefore, a

double letter, i.e. AA, BB, C, DD, and EE, will still design them. Information about

roughness or chemical treatments were not provided, but one can assume that

the machined titanium surface was the smoothest of the four titanium surfaces,

and that the zirconia surface was a yttrium-reinforced zirconium oxide. Because

of its white color, zirconia could be readily recognizable (EE), and machined

titanium surface could possibly be identified too (probably AA).          A formal

investigation on micro-topography of these surfaces was not part of the

experiment, but the SEM observations demonstrated subjectively a clear

difference between each surface in terms of topography.
                                                                                    131


       The first assay was an attachment assay, aiming at assessing in a

quantitative way, how human gingival immortalized epithelial cells (hTERT cells)

would attach on the different surfaces. Roughness, as well as nature of the

material and possible chemical treatment of the surface, are known to influence

cell adhesion and behavior. The effect of such characteristic varies with the type

of the cell; a fibroblast will behave differently than an osteoblast or an epithelial

cell (Hamilton et al., 2007). Osteoblasts have rather constantly been shown as

having a great affinity for rougher surface. It seems that fibroblasts share the

same affinity, and most of the studies propose a slightly rough surface as a better

substrate for fibroblasts adhesion.      As for epithelial cells, it is not yet clear

whether a smooth or a rough surface is preferred. Published results are often

contradictory. Baharloo et al. suggested a smooth surface as more favorable for

epithelial cells adhesion (Baharloo et al., 2005), as well as Hormia et al (Hormia,

Kononen, Kivilahti, & Virtanen, 1991), whereas Di Carmine et al. (Di Carmine et

al., 2003) suggested a slightly rough surface as a better substrate and

Baumhammer et al. could not established a difference between smooth and

rough surfaces for epithelial cells adhesion (Baumhammers et al., 1978).              A

possible explanation for these contradictory results is the great variability of
experimental designs among those studies. Roughness is just but one factor.

Nature of the surface, type of topography, anisotropy or not of the surface,

chemical treatments of the surface, are all confounders that make interpretation

difficult. Moreover, epithelial cells used for the studies are often of different kinds,

sometime issued from transplants, sometime form immortalized cells, sometime

from human origin, sometime from animal origin. Even the provenance of the

cells in a same species, such as skin epithelial cells, or gingival epithelial cells,

may influence the results (Marmary, Brunette, & Heersche, 1976). The fact that

epithelial cells tend to create a confluent layer acting as a single unit, might also
                                                                                    132


explain divergent results (Hamilton et al., 2007). These results also depend on

the concentration of the cells in the media, the time allowed for attachment, so

on, and so forth. It has been said that such a confluent layer is necessarily less

influenced by the substrate nature and topography than a single cell.

       In this experiment, no significant difference could be found in cell

attachment after two hours, reinforcing the idea that the substrate might not have

the highest influence on this type of cells. Machined titanium has been used for

more than five decades in the medical and dental field and has since then proven

its great biocompatibility, even though there are reports of possible amelioration

of cell affinity for that material by different treatments.       The fact that no

differences could be shown for the initial attachment (two hours) on human

gingival epithelial cells on the four test surfaces, compared to the machined

titanium surface (control), reflects a biocompatibility similar to that of machined

titanium when it comes to initial cell attachment, which has been proven to be a

critical time-period for cell adhesion (C. M. Stanford, Solursh, & Keller, 1999).

       Attachment of the epithelial cells was then further assessed by PCR and

electrophoresis assays, in order to determine expression of 6 4 integrin, a cell

surface receptor involved in cell-matrix and hemidesmosomes adhesion. Again,
no difference could be shown between the five surfaces, confirming that human

gingival epithelial cells showed a similar tendency to attach onto the different

substrates, in spite of their probable topographic differences.

       Expression of other proteins, that is six different cytokines, was also

performed using commercial multiplexed fluorescent bead-based immunoassays

to determine cytokines and chemokines concentrations in cell culture

supernatants. Levels did not differ significantly between surfaces. This result

suggests that the five materials influence on production of tested cytokines by the

epithelial cells attached on their surface was comparable, and again that
                                                                                    133


behaviour of these cells was similar for each of the abutment material candidates.

Epithelial cells are know to produce IL-1 , but not IL-1 ; also, they are not

producing high levels of TNF- .         On the other hands, reports have been

published that mentioned the influence of the roughness of the substrate on

cytokines expression (Refai et al., 2004; Spyrou et al., 2002). The different result

observed in the present study might be explained by two factors: first, the

roughness of the five surfaces used here were not formally investigated in terms

of roughness, and the difference between them might be less than for Refai et al.’

or Spyrou et al.’ experiments. Second, Refai et al. made their observation for

macrophages, and Spyrou et al. for osteoblasts, not for epithelial cells. This, by

itself, may explain the difference with the results of the present investigation.

       Morphology of the attached cells, as well as distribution of the actin

filaments in their cytoplasm were evaluated through SEM and immuno-

fluorescence assays.      Contradicting the results of the attachment assays,

densities of the cells observed on the different surfaces using immuno-

fluorescence, were not similar.      One titanium surface (BB) showed a lower

density, as EE (zirconia surface), showed a denser coverage by the cells. A first

possible reason for these discrepancies is the fact that for the SEM and IF
assays, cells were left on the discs for 24 hours, and not two hours as for the

attachment assay. The proliferation tests at three, five, and seven days have

shown that the growth of the cells on the surfaces were not identical over time.

More over, these observations need to be considered very carefully though, since

only two discs for each type of material were examined, and no formal

randomization of the observations was made. The investigators rather aimed at

taking images of “representative” areas of the discs. Even more, within a same

material, differences could be seen between the two discs. The second BB disc

was quite more covered than the one whose picture is reproduced in the present
                                                                                  134


manuscript.    In the same way, the second EE disc presented far less cells

attached on its surface. Moreover, analysis of cell density on the discs after

preparation for immuno-assay was by no means an intended objective, but more

an incidental observation, of little scientific value.       It seems that taking into

account only the formal quantitative analysis given by cell attachment and

proliferation assay is a safer approach to analyse the differences between the

surfaces. In the same way, the statement that DD, and CC surfaces presented

cells with a more irregular contour suggesting a better attachment, compared to

the rounder cells on the other surfaces, even though consistent to the

subsequent proliferation assays, should be taken with circumspection, for the

reasons above mentioned.        One should rather observe that, over all, cells

presented with a relative similitude on all five surfaces.

       The exact same observations can be made for the SEM assay where,

again, no randomization and not quantitative analysis were performed. If certain

minor differences could be observed between the discs, they should not allow to

draw any conclusion, except that all cells observed presented a potential to

attachment. It is nevertheless striking to observe that on zirconia discs, not a

single cell could be found at the time of the observation, even though attachment
assays, and even the immuno-fluorescence pictures, clearly showed that cells did

attached on those surfaces. This absence of cells in the SEM pictures could be

explained by the preparation of the discs for the observation, the different steps

applied to the samples having probably leaded to a loosening of the cells. In

such case, given the constant with which zirconia discs appeared bare, and that

all other surfaces systematically showed cells on their surface, it should be

assumed that attachment of the cells on zirconia, even though as frequent as on

other discs, as shown by the attachment assay, might not be as strong
                                                                                    135


qualitatively. Lauer already exposed the fact that a high level of attachment does

not necessarily mean a good quality of that attachment (Lauer et al., 2001).

       Proliferation assays concluded this set of initial assessment of the

behaviour of human gingival epithelial cells on the five surfaces under

investigation. A tissue culture plastic (TCP) surface was added as a control for

this assay. BB, CC and DD surfaces showed a significant lower proliferation rate

over time, compared to AA, EE, and TCP surfaces. The difference with the

attachment assay, where no differences could be seen, might be explained by

two factors: first, the initial count for the proliferation assay was at three days,

compared at two hours for the attachment assay. Second, the concentration of

cells seeded in the wells for the proliferation assay was higher than the one for

the attachment assays. Investigators have shown that proliferation and adhesion

are dependant on the initial concentration of cells (Hamilton et al., 2007; Lauer et

al., 2001; Liu & Karasek, 1989). This might be of special importance when trying

to extrapolate in vitro results to in vivo situations where cell concentration is

higher (Brunette & Chehroudi, 1999).

       AA surfaces were most probably the machined titanium surfaces, and EE

the zirconium surfaces. Along with TCP, these were apparently the smoothest
surfaces of this experiment. Hence, it seems that under the conditions of the

experiment, smooth surfaces do not promote initial adhesion of the epithelial cells

(two hours), but favour their subsequent growth.           If capacity of growth for

epithelial cells is undeniably sought after for dental implant abutment material, it

is not necessarily true that the highest rate of proliferation is desirable. Epithelial

cells will compete with the fibroblasts for the abutment surface, and they have a

heavy inherent advantage for that race. Capacity of the material to favour even

more epithelial cells could be detrimental to connective tissue attachment. More

important, proliferative cells (blastic cells) tend to be less differentiated than cells
                                                                                   136


with a lesser proliferative potential. Similarly to this experiment, Lauer et al. could

show that smoother surfaces presented a higher level of proliferation at six days,

and that plasma-sprayed and sandblasted titanium surfaces presented a lower

level of growth. They have also suggested that this lower level of adhesion could

reflect a lower level of differentiation (Lauer et al., 2001). The lower proliferation

rates observed for BB, CC, and DD, could reflect a more “advanced” stage of the

cell life cycle, and maybe a greater expression of keratin proteins, for example.

Such a hypothesis could be tested in a subsequent experiment by evaluating

expression of such keratin proteins. It has been proven that specific keratin types

are produced at specific stages of differentiation of the keratinocytes (Gasparoni,

Squier, & Fonzi, 2005).

       This experiment does present a certain number of limitations.           One of

them is inherent to in vitro experiments, which results cannot readily be

transposed to the clinical floor (Brunette & Chehroudi, 1999; Brunette, 1999).

Moreover, if human gingival cells were used, eliminating a possible bias that

animal cells could carry, these were immortalized cells, and it has been shown

that such cells do not perform the exact same biological functions than regular

ones. They can be more compared to cancerous cells, i.e. pathological cells.
Therefore, expression of certain genes might be influenced by this fact (Farwell et

al., 2000). If certain properties of a surface can promote a certain type of cells, or

a certain behavior, it remains to be elucidate whether this promotion is desirable

or not.   For example, if rough surfaces were to actually promote a better

differentiation of epithelial cells, it has also been demonstrated that it would

promote a better bacterial adhesion and that this could elicit peri-implantitis.

Similarly, if epithelial adhesion and growth is favored, fibroblasts would have

more difficulty to attach on the same surface, and connective stabilization of the

epithelial attachment would be compromised.         Production of pro-inflammatory
                                                                                 137


cytokines might be the sign of suffering cells, but on the other hand, could lead to

a better line of defense in the peri-implant area and a better healing capacity, due

to the chemotactic properties of those cytokines. It is therefore not clear whether

one wants a lower or a higher level of cytokines production during the initial

healing phase of peri-implant tissues.      In the present experiment, the only

conclusion that can be made from the cytokines assays is that no difference in

their levels could be shown at 24 hours, but not whether this is favorable to the

test surfaces compared to the control one (machined titanium). Maybe a higher

or a lower level would have been profitable.

       Beyond the aims explored in this study, many other factors have to be

considered before accepting one or the other of those surfaces as good surface

material for a dental implant abutment.        For example, zirconia, even though

proved to be similar to the other surfaces for the characteristics investigated here

(exception maybe for the quality of the attachment of cells), is still a material

raising a lot of concerns. Ageing has almost brought this material out of the

orthopedic field (Clarke et al., 2003), and is still a Damocles sword hanging over

it. Even if the consequences of a failure are not as catastrophic in the dental

field, this ageing phenomenon might be a problem too, for materials are
submitted to extremely difficult conditions (temperature changes, cyclic loads,

sometimes of heavy magnitude, acidity, humidity…). Wear of opposing metal

surfaces is a real problem, and new designs elaborated to avoid this are still to be

tested clinically.

       Overall, research for the best abutment material is very complex due to the

antagonist properties that are desirable. Ongoing in vitro experiments need to be

done, and reinforced later by long-term randomized clinical trials. Until then, it is

unreasonable to assume that such or such material is a better choice than
                                                                               138


another one, and smooth titanium could probably still be considered as the

present standard, to which other surfaces have to be compared.


                                  5.2 Conclusion



      In this experiment, no difference could be shown between the tested

surfaces for human gingival epithelial cell attachment at two hours, expression of

adhesion protein 6 4 integrin, six pro-inflammatory cytokines production, and

morphologic appearance of these same cells after 24 hours.

      Proliferation tests showed significant differences in proliferation rates for

these cells at three, five and seven days between the different surfaces.

      More investigations are needed before any clinical recommendation can

be made.
                                                                                       139


                                         APPENDIX




Attachment assay - Batch YS08-2


Post-dilution reference: PDR1 = 873 ; PDR2 = 729     Average =   801




                            Unattached    Attached   Attached              Standard
Surface       Disc                                               Average
                            cells         cells      cells (%)             deviation
              1             347           454        56,68
AA            2             298           503        62,80       60,09     3,12
              3             314           487        60,80
              4             242           559        69,79
BB            5             258           543        67,79       69,37     1,42
              6             236           565        70,54
              7             306           495        61,80
CC            8             314           487        60,80       63,67     4,14
              9             253           548        68,41
              10            336           465        58,05
DD            11            300           501        62,55       59,80     2,41
              12            330           471        58,80
              13            444           357        44,57
EE            14            356           445        55,56       49,48     5,59
              15            414           387        48,31




Table A 1: Batch YS08-2 data for attachment assay.
                                                                                       140




Attachment assay - Batch YS08-3


Post-dilution reference: PDR1 = 669 ; PDR2 = 821     Average =   745




                            Unattached    Attached   Attached              Standard
Surface       Disc                                               Average
                            cells         cells      cells (%)             deviation
              1             330           415        55,70
AA            2             268           477        64,03       61,16     4,73
              3             270           475        63,76
              4             219           526        70,60
BB            5             158           587        78,79       73,15     4,89
              6             223           522        70,07
              7             267           478        64,16
CC            8             233           512        68,72       66,31     2,29
              9             253           492        66,04
              10            156           589        79,06
DD            11            168           577        77,45       76,82     2,61
              12            194           551        73,96
              13            262           483        64,83
EE            14            249           496        66,58       65,01     1,48
              15            271           474        63,62




Table A 2: Batch YS08-3 data for attachment assay.
                                                                                      141




Attachment assay - Batch YS08-4


Post-dilution reference: PDR1= 838 ; PDR2 =840      Average =   840




                           Unattached    Attached   Attached              Standard
Surface      Disc                                               Average
                           cells         cells      cells (%)             deviation
             1             399           441        52,50
AA           2             457           383        45,60       49,96     3,80
             3             405           435        51,79
             4             360           480        57,14
BB           5             318           522        62,14       60,08     2,61
             6             328           512        60,95
             7             344           496        59,05
CC           8             374           466        55,48       56,55     2,17
             9             377           463        55,12
             10            310           530        63,10
DD           11            324           516        61,43       61,67     1,33
             12            332           508        60,48
             13            345           495        58,93
EE           14            351           489        58,21       58,77     0,50
             15            343           497        59,17




Table A 3: Batch YS08-4 data for attachment assay.
                                                                                      142




Attachment assay - Batch YS08-5


Post-dilution reference: PDR1= 652 ; PDR2 =744      Average =   698




                            Unattached   Attached   Attached              Standard
Surface      Disc                                               Average
                            cells        cells      cells (%)             deviation
             1              387          311        44,56
AA           2              346          352        50,43       45,99     3,93
             3              398          300        42,98
             4              285          413        59,17
BB           5              293          405        58,02       53,25     9,28
             6              401          297        42,55
             7              305          393        56,30
CC           8              306          392        56,16       54,49     3,02
             9              342          356        51,00
             10             384          314        44,99
DD           11             351          347        49,71       48,81     3,46
             12             337          361        51,72
             13             336          362        51,86
EE           14             318          380        54,44       53,53     1,45
             15             319          379        54,30




Table A 4: Batch YS08-5 data for attachment assay.
                                                                                      143




Attachment assay - Batch YS08-6


Post-dilution reference: PDR1= 1030 ; PDR2= 1098    Average =   1064




                           Unattached    Attached   Attached              Standard
Surface      Disc                                               Average
                           cells         cells      cells (%)             deviation
             1             434           630        59,21
AA           2             352           712        66,92       65,98     6,35
             3             300           764        71,80
             4             401           663        62,31
BB           5             296           768        72,18       66,35     5,17
             6             377           687        64,57
             7             260           804        75,56
CC           8             290           774        72,74       71,68     4,51
             9             354           710        66,73
             10            403           661        62,12
DD           11            345           719        67,58       65,79     3,17
             12            344           720        67,67
             13            300           764        71,80
EE           14            371           693        65,13       69,64     3,91
             15            298           766        71,99




Table A 5: Batch YS08-6 data for attachment assay.
                                                                                           144




Luminex 100 IS       Kit Number 48-002       lot #1392783         01/08/2008


                                    IL-10      IL-12    IL-1B    IL-6    IL-8     TNFa
Spot   Content      Conc     Tube   pg/mL      pg/mL    pg/mL    pg/mL   pg/mL    pg/mL
1      Std A - S8   5000,0          3520,0     2510,0   2730,0   3970,0 4060,0    2710,0
2      Std A - S8   5000,0          6170,0     7180,0   7910,0   5520,0 4340,0    <HIGH>
3      Std B - S7   1667,0          1980,0     2350,0   1690,0   1970,0 1750,0    1750,0
4      Std B - S7   1667,0          1680,0     1900,0   2290,0   1830,0 1940,0    1680,0
5      Std C - S6   556,0           550,0      590,0    568,0    529,0   606,0    559,0
6      Std C - S6   556,0           528,0      539,0    518,0    496,0   558,0    531,0
7      Std D - S5   185,0           194,0      173,0    177,0    193,0   185,0    188,0
8      Std D - S5   185,0           185,0      182,0    186,0    197,0   209,0    191,0
9      Std E - S4   61,7            57,8       55,6     64,4     57,6    59,4     59,9
10     Std E - S4   61,7            64,2       65,9     61,5     60,9    59,2     63,7
11     Std F - S3   20,3            20,8       20,9     20,0     20,2    19,9     20,0
12     Std F - S3   20,3            19,7       20,0     21,6     20,9    20,0     19,3
13     Std G - S2   6,9             6,9        7,4      7,0      7,2     6,9      7,1
14     Std G - S2   6,9             7,1        6,6      6,5      6,8     7,1      7,3
15     Std H - S1   2,3             2,3        2,5      2,4      1,9     2,3      2,2
16     Std H - S1   2,3             2,3        2,1      2,2      2,8     2,3      2,4
17     SB           0,0
18     SB           0,0
19     AA-1                  YS-08-2 <LOW> 1,5          1,0      68,9    809,0    2,0
20     AA-2                  YS-08-2 <LOW> 1,7          1,1      64,2    970,0    2,1
21     AA-3                  YS-08-2 <LOW> 1,6          <LOW> 52,2       754,0    1,7
22     BB-1                  YS-08-2 <LOW> 1,2          <LOW> 44,6       526,0    1,6
23     BB-2                  YS-08-2 <LOW> 1,5          1,1      58,2    628,0    1,8
24     BB-3                  YS-08-2 <LOW> 1,7          <LOW> 90,5       1150,0   2,4
25     CC-1                  YS-08-2 <LOW> 1,7          1,2      79,2    1090,0   2,3
26     CC-2                  YS-08-2 <LOW> 1,7          1,0      76,0    935,0    2,1




Table A 6: Raw results for cytokines levels measurements
                                                                         145


27    CC-3            YS-08-2 <LOW> 1,6   1,4    88,8   1130,0   2,3
28    DD-1            YS-08-2 <LOW> 1,6   <LOW> 52,8    760,0    1,9
29    DD-2            YS-08-2 <LOW> 1,8   1,1    62,4   880,0    1,8
30    DD-3            YS-08-2 <LOW> 1,6   <LOW> 63,6    907,0    1,8
31    EE-1            YS-08-2 <LOW> 1,4   <LOW> 27,1    321,0    1,4
32    EE-2            YS-08-2 <LOW> 1,5   <LOW> 32,2    549,0    1,6
33    EE-3            YS-08-2 <LOW> 1,4   <LOW> 9,0     21,0     1,1
34    AA-1            YS-08-3 <LOW> 1,1   <LOW> 10,5    120,0    <LOW>
35    AA-2            YS-08-3 <LOW> 1,4   1,0    32,5   679,0    1,9
36    AA-3            YS-08-3 <LOW> 1,4   1,1    26,7   370,0    1,2
37    BB-1            YS-08-3 <LOW> 1,5   <LOW> 40,6    669,0    1,9
38    BB-2            YS-08-3 <LOW> 1,6   <LOW> 50,4    1150,0   2,5
39    BB-3            YS-08-3 <LOW> 1,5   <LOW> 29,9    448,0    1,6
40    CC-1            YS-08-3 <LOW> 1,4   <LOW> 33,4    618,0    1,8
41    CC-2            YS-08-3 <LOW> 1,6   1,1    50,7   743,0    1,8
42    CC-3            YS-08-3 <LOW> 1,5   <LOW> 36,9    574,0    1,8
43    DD-1            YS-08-3 <LOW> 1,6   <LOW> 44,6    691,0    2,0
44    DD-2            YS-08-3 <LOW> 1,7   2,2    79,8   982,0    2,3
45    DD-3            YS-08-3 <LOW> 1,3   <LOW> 31,9    597,0    1,6
46    EE-1            YS-08-3 <LOW> 1,6   <LOW> 38,7    910,0    2,2
47    EE-2            YS-08-3 <LOW> 1,5   <LOW> 30,6    682,0    1,9
48    EE-3            YS-08-3 <LOW> 1,7   <LOW> 40,6    1010,0   2,3
49    AA-1            YS-08-4 <LOW> 1,5   1,1    77,1   1120,0   1,6
50    AA-2            YS-08-4 <LOW> 1,6   <LOW> 52,8    842,0    1,8
51    AA-3            YS-08-4 <LOW> 1,6   <LOW> 68,3    751,0    1,6
52    BB-1            YS-08-4 <LOW> 1,5   <LOW> 58,2    929,0    2,1
53    BB-2            YS-08-4 <LOW> 1,8   <LOW> 60,6    981,0    2,0
54    BB-3            YS-08-4 <LOW> 1,8   <LOW> 60,3    1110,0   2,0
55    CC-1            YS-08-4 <LOW> 1,7   1,1    60,3   1250,0   2,2
56    CC-2            YS-08-4 <LOW> 1,6   <LOW> 60,9    1330,0   2,3
57    CC-3            YS-08-4 <LOW> 1,5   <LOW> 57,0    1380,0   2,0
58    DD-1            YS-08-4 <LOW> 1,5   <LOW> 58,2    1160,0   2,2




Table A 6 continued
                                                                         146


59    DD-2            YS-08-4 <LOW> 1,6   1,0    57,3   1280,0   2,0
60    DD-3            YS-08-4 <LOW> 1,5   <LOW> 39,7    803,0    1,6
61    EE-1            YS-08-4 <LOW> 1,7   <LOW> 87,1    918,0    1,8
62    EE-2            YS-08-4 <LOW> 1,3   1,0    44,3   828,0    1,7
63    EE-3            YS-08-4 <LOW> 1,6   <LOW> 47,3    679,0    1,9
64    AA-1            YS-08-5 <LOW> 1,3   <LOW> 11,9    52,3     1,0
65    AA-2            YS-08-5 <LOW> 1,4   <LOW> 28,0    208,0    1,1
66    AA-3            YS-08-5 <LOW> 1,3   <LOW> 16,1    106,0    <LOW>
67    BB-1            YS-08-5 <LOW> 1,4   <LOW> 31,9    226,0    1,2
68    BB-2            YS-08-5 <LOW> 1,3   <LOW> 31,5    223,0    1,5
69    BB-3            YS-08-5 <LOW> 1,2   <LOW> 19,5    104,0    <LOW>
70    CC-1            YS-08-5 <LOW> 1,3   <LOW> 22,2    128,0    1,1
71    CC-2            YS-08-5 <LOW> 1,4   <LOW> 28,0    174,0    1,0
72    CC-3            YS-08-5 <LOW> 1,3   <LOW> 18,8    97,6     <LOW>
73    DD-1            YS-08-5 <LOW> 1,3   <LOW> 29,9    215,0    1,1
74    DD-2            YS-08-5 <LOW> 1,4   <LOW> 30,6    160,0    1,2
75    DD-3            YS-08-5 <LOW> 1,1   <LOW> 20,2    128,0    1,1
76    EE-1            YS-08-5 <LOW> 1,4   <LOW> 23,5    182,0    1,0
77    EE-2            YS-08-5 <LOW> 1,4   <LOW> 35,0    209,0    1,3
78    EE-3            YS-08-5 <LOW> 1,4   <LOW> 24,1    182,0    <LOW>
79    AA-1            YS-08-6 <LOW> 1,4   <LOW> 57,6    202,0    1,3
80    AA-2            YS-08-6 <LOW> 1,3   <LOW> 44,9    294,0    1,2
81    AA-3            YS-08-6 <LOW> 1,5   <LOW> 45,5    166,0    1,3
82    BB-1            YS-08-6 <LOW> 1,2   <LOW> 32,5    132,0    1,2
83    BB-2            YS-08-6 <LOW> 1,2   <LOW> 39,4    185,0    1,2
84    BB-3            YS-08-6 <LOW> 1,4   <LOW> 52,2    260,0    1,3
85    CC-1            YS-08-6 <LOW> 1,5   <LOW> 53,4    308,0    1,7
86    CC-2            YS-08-6 <LOW> 1,2   <LOW> 6,8     <LOW> <LOW>
87    CC-3            YS-08-6 <LOW> 1,3   <LOW> 36,9    196,0    1,2
88    DD-1            YS-08-6 <LOW> 1,2   <LOW> 24,8    180,0    <LOW>
89    DD-2            YS-08-6 <LOW> 1,3   <LOW> 37,5    237,0    1,5
90    DD-3            YS-08-6 <LOW> 1,2   <LOW> 23,5    171,0    1,3




Table A 6 continued
                                                                     147


91    EE-1            YS-08-6 <LOW> 1,3   <LOW> 35,0   280,0   1,3
92    EE-2            YS-08-6 <LOW> 1,4   <LOW> 30,6   238,0   1,3
93    EE-3            YS-08-6 <LOW> 1,5   <LOW> 28,0   247,0   1,4
94    Media           Media   <LOW> 1,2   <LOW> 3,6    <LOW> <LOW>
95    media           media   <LOW> 1,1   <LOW> 6,0    <LOW> <LOW>
96    media           media   <LOW> 1,4   <LOW> 5,2    <LOW> 1,0
96    media           media   <LOW> 1,4   <LOW> 5,2    <LOW> 1,0




Table A 6 continued
                                                                                       148


                                   REFERENCES




Aboushelib, M. N., de Jager, N., Kleverlaan, C. J., & Feilzer, A. J. (2007). Effect of
   loading method on the fracture mechanics of two layered all-ceramic restorative
   systems. Dental Materials : Official Publication of the Academy of Dental Materials,
   23(8), 952-959.

Abrahamsson, I., Berglundh, T., Glantz, P. O., & Lindhe, J. (1998). The mucosal
    attachment at different abutments. an experimental study in dogs. Journal of Clinical
    Periodontology, 25(9), 721-727.

Abrahamsson, I., Berglundh, T., Moon, I. S., & Lindhe, J. (1999). Peri-implant tissues at
    submerged and non-submerged titanium implants. Journal of Clinical
    Periodontology, 26(9), 600-607.

Abrahamsson, I., & Cardaropoli, G. (2007). Peri-implant hard and soft tissue integration
    to dental implants made of titanium and gold. Clinical Oral Implants Research,
    18(3), 269-274.

Abrahamsson, I., Zitzmann, N. U., Berglundh, T., Linder, E., Wennerberg, A., & Lindhe,
    J. (2002). The mucosal attachment to titanium implants with different surface
    characteristics: An experimental study in dogs. Journal of Clinical Periodontology,
    29(5), 448-455.

Albrektsson, T., Branemark, P. I., Hansson, H. A., & Lindstrom, J. (1981).
    Osseointegrated titanium implants. requirements for ensuring a long-lasting, direct
    bone-to-implant anchorage in man. Acta Orthopaedica Scandinavica, 52(2), 155-
    170.

Albrektsson, T., & Wennerberg, A. (2005). The impact of oral implants - past and future,
    1966-2042. Journal (Canadian Dental Association), 71(5), 327.

Al-Sabbagh, M. (2006). Implants in the esthetic zone. Dental Clinics of North America,
    50(3), 391-407, vi.

Baharloo, B., Textor, M., & Brunette, D. M. (2005). Substratum roughness alters the
   growth, area, and focal adhesions of epithelial cells, and their proximity to titanium
   surfaces. Journal of Biomedical Materials Research.Part A, 74(1), 12-22.

Baier, R. E. (1988). Surface preparation. The International Journal of Oral Implantology
    : Implantologist, 5(2), 15-19.
                                                                                          149


Baier, R. E., Meyer, A. E., Natiella, J. R., Natiella, R. R., & Carter, J. M. (1984). Surface
    properties determine bioadhesive outcomes: Methods and results. Journal of
    Biomedical Materials Research, 18(4), 327-355.

Balasubramanian, V., Hall, C. L., Shivashankar, S., Slack, S. M., & Turitto, V. T. (1998).
    Vascular cell attachment and procoagulant activity on metal alloys. Journal of
    Biomaterials Science.Polymer Edition, 9(12), 1349-1359.

Bauman, G. R., Rapley, J. W., Hallmon, W. W., & Mills, M. (1993). The peri-implant
   sulcus. The International Journal of Oral & Maxillofacial Implants, 8(3), 273-280.

Baumhammers, A., Langkamp, H. H., Matta, R. K., & Kilbury, K. (1978). Scanning
   electron microscopy of epithelial cells grown on enamel, glass and implant materials.
   Journal of Periodontology, 49(11), 592-597.

Berglundh, T., Lindhe, J., Ericsson, I., Marinello, C. P., Liljenberg, B., & Thomsen, P.
    (1991). The soft tissue barrier at implants and teeth. Clinical Oral Implants Research,
    2(2), 81-90.

Binon, P. P. (1996). The effect of implant/abutment hexagonal misfit on screw joint
    stability. The International Journal of Prosthodontics, 9(2), 149-160.

Binon, P. P., & McHugh, M. J. (1996). The effect of eliminating implant/abutment
    rotational misfit on screw joint stability. The International Journal of Prosthodontics,
    9(6), 511-519.

Blackburn, E. H. (1991). Telomeres. Trends in Biochemical Sciences, 16(10), 378-381.

Bollen, C. M. L., Papaioanno, W., VanEldere, J., Schepers, E., Quirynen, M., & Van
    Steenberghe, D. (1996). The influence of abutment surface roughness on plaque
    accumulation and peri-implant mucositis. Clinical Oral Implants Research, 7(3),
    201.

Boyan, B. D., Hummert, T. W., Dean, D. D., & Schwartz, Z. (1996). Role of material
   surfaces in regulating bone and cartilage cell response. Biomaterials, 17(2), 137-146.

Branemark, P. I., Adell, R., Breine, U., Hansson, B. O., Lindstrom, J., & Ohlsson, A.
    (1969). Intra-osseous anchorage of dental prostheses. I. experimental studies.
    Scandinavian Journal of Plastic and Reconstructive Surgery, 3(2), 81-100.

Brodbeck, U. (2003). The ZiReal post: A new ceramic implant abutment. Journal of
    Esthetic and Restorative Dentistry : Official Publication of the American Academy of
    Esthetic Dentistry ...[Et Al.], 15(1), 10-23; discussion 24.

Brunette, D. M. (1999). In vitro models of biological responses to implants. Advances in
    Dental Research, 13, 35-37.
                                                                                         150


Brunette, D. M., & Chehroudi, B. (1999). The effects of the surface topography of
    micromachined titanium substrata on cell behavior in vitro and in vivo. Journal of
    Biomechanical Engineering, 121(1), 49-57.

Burridge, K., & Fath, K. (1989). Focal contacts: Transmembrane links between the
    extracellular matrix and the cytoskeleton. BioEssays : News and Reviews in
    Molecular, Cellular and Developmental Biology, 10(4), 104-108.

Burridge, K., Fath, K., Kelly, T., Nuckolls, G., & Turner, C. (1988). Focal adhesions:
    Transmembrane junctions between the extracellular matrix and the cytoskeleton.
    Annual Review of Cell Biology, 4, 487-525.

Buser, D., Weber, H. P., Donath, K., Fiorellini, J. P., Paquette, D. W., & Williams, R. C.
    (1992). Soft tissue reactions to non-submerged unloaded titanium implants in beagle
    dogs. Journal of Periodontology, 63(3), 225-235.

Cate, T., & Nanci, A. (2008). Oral histology (7th edition ed.). St Louis, Missouri 63146:
    ELSEVIER MOSBY.

Cavaillon, J. M. (1995). Cytokines in inflammation. [Les cytokines de l'inflammation]
   Comptes Rendus Des Seances De La Societe De Biologie Et De Ses Filiales, 189(4),
   531-544.

Cavaillon, J. M. (2001). Pro- versus anti-inflammatory cytokines: Myth or reality.
   Cellular and Molecular Biology (Noisy-Le-Grand, France), 47(4), 695-702.

Cavaillon, J. M. (2005). Molecular mediators: Cytokines. Encyclopedia of molecular cell
   biology and molecular medicine (2nd Edition, vol.8 ed., pp. 431-431-460).
   Weinheim, Germany: Wiley-VCH Verlag.

Chehroudi, B., & Brunette, D. M. (2002). Subcutaneous microfabricated surfaces inhibit
   epithelial recession and promote long-term survival of percutaneous implants.
   Biomaterials, 23(1), 229-237.

Chehroudi, B., Gould, T. R., & Brunette, D. M. (1990). Titanium-coated micromachined
   grooves of different dimensions affect epithelial and connective-tissue cells
   differently in vivo. Journal of Biomedical Materials Research, 24(9), 1203-1219.

Chehroudi, B., Gould, T. R., & Brunette, D. M. (1992). The role of connective tissue in
   inhibiting epithelial downgrowth on titanium-coated percutaneous implants. Journal
   of Biomedical Materials Research, 26(4), 493-515.

Chevalier, J. (2006). What future for zirconia as a biomaterial? Biomaterials, 27(4), 535-
   543.
                                                                                          151


Chou, L., Firth, J. D., Uitto, V. J., & Brunette, D. M. (1995). Substratum surface
   topography alters cell shape and regulates fibronectin mRNA level, mRNA stability,
   secretion and assembly in human fibroblasts. Journal of Cell Science, 108 ( Pt 4)(Pt
   4), 1563-1573.

Chrzanowska-Wodnicka, M., & Burridge, K. (1996). Rho-stimulated contractility drives
    the formation of stress fibers and focal adhesions. The Journal of Cell Biology,
    133(6), 1403-1415.

Clarke, I. C., Manaka, M., Green, D. D., Williams, P., Pezzotti, G., Kim, Y. H., et al.
    (2003). Current status of zirconia used in total hip implants. The Journal of Bone and
    Joint Surgery.American Volume, 85-A Suppl 4, 73-84.

Conover, W. J., & Iman, R. L. (1976). Some alternative procedures using ranks for
   analysis of experimental-designs. Communications in Statistics, 5(14), 1349.

Conover, W. J., & Iman, R. L. (1981). Rank transformation as a bridge between
   parametric and nonparametric statistics. American Statistician, the, 35(3), 124.

Covacci, V., Bruzzese, N., Maccauro, G., Andreassi, C., Ricci, G. A., Piconi, C., et al.
   (1999). In vitro evaluation of the mutagenic and carcinogenic power of high purity
   zirconia ceramic. Biomaterials, 20(4), 371-376.

Damji, A., Weston, L., & Brunette, D. M. (1996). Directed confrontations between
   fibroblasts and epithelial cells on micromachined grooved substrata. Experimental
   Cell Research, 228(1), 114-124.

Damsky, C. H., & Ilic, D. (2002). Integrin signaling: It's where the action is. Current
   Opinion in Cell Biology, 14(5), 594-602.

Degidi, M., Artese, L., Scarano, A., Perrotti, V., Gehrke, P., & Piattelli, A. (2006).
   Inflammatory infiltrate, microvessel density, nitric oxide synthase expression,
   vascular endothelial growth factor expression, and proliferative activity in peri-
   implant soft tissues around titanium and zirconium oxide healing caps. Journal of
   Periodontology, 77(1), 73-80.

Deporter, D. A., Friedland, B., Watson, P. A., Pilliar, R. M., Howley, T. P., Abdulla, D.,
   et al. (1986). A clinical and radiographic assessment of a porous-surfaced, titanium
   alloy dental implant system in dogs. Journal of Dental Research, 65(8), 1071-1077.

Deporter, D. A., Watson, P. A., Pilliar, R. M., Howley, T. P., & Winslow, J. (1988). A
   histological evaluation of a functional endosseous, porous-surfaced, titanium alloy
   dental implant system in the dog. Journal of Dental Research, 67(9), 1190-1195.

Di Carmine, M., Toto, P., Feliciani, C., Scarano, A., Tulli, A., Strocchi, R., et al. (2003).
    Spreading of epithelial cells on machined and sandblasted titanium surfaces: An in
    vitro study. Journal of Periodontology, 74(3), 289-295.
                                                                                         152


Farwell, D. G., Shera, K. A., Koop, J. I., Bonnet, G. A., Matthews, C. P., Reuther, G. W.,
    et al. (2000). Genetic and epigenetic changes in human epithelial cells immortalized
    by telomerase. The American Journal of Pathology, 156(5), 1537-1547.

Fitzgerald, J. E., & Kreutzer, D. L. (1995). Localization of interleukin-8 in human
    gingival tissues. Oral Microbiology and Immunology, 10(5), 297-303.

Folkman, J., & Moscona, A. (1978). Role of cell shape in growth control. Nature,
    273(5661), 345-349.

Gasparoni, A., Squier, C. A., & Fonzi, L. (2005). Cytokeratin changes in cell culture
   systems of epithelial cells isolated from oral mucosa: A short review. Italian Journal
   of Anatomy and Embryology = Archivio Italiano Di Anatomia Ed Embriologia,
   110(2), 75-82.

Glauser, R., Sailer, I., Wohlwend, A., Studer, S., Schibli, M., & Scharer, P. (2004).
    Experimental zirconia abutments for implant-supported single-tooth restorations in
    esthetically demanding regions: 4-year results of a prospective clinical study. The
    International Journal of Prosthodontics, 17(3), 285-290.

Grinnell, F. (1978). Cellular adhesiveness and extracellular substrata. International
    Review of Cytology, 53, 65-144.

Greider, C. W. (1991). Telomeres. Current Opinion in Cell Biology, 3(3), 444-451.

Gristina, A. (2004). Biomaterial-centered infection: Microbial adhesion versus tissue
    integration. 1987. Clinical Orthopaedics and Related Research, (427)(427), 4-12.

Groessner-Schreiber, B., Hannig, M., Duck, A., Griepentrog, M., & Wenderoth, D. F.
    (2004). Do different implant surfaces exposed in the oral cavity of humans show
    different biofilm compositions and activities? European Journal of Oral Sciences,
    112(6), 516-522.

Grossner-Schreiber, B., Herzog, M., Hedderich, J., Duck, A., Hannig, M., & Griepentrog,
    M. (2006). Focal adhesion contact formation by fibroblasts cultured on surface-
    modified dental implants: An in vitro study. Clinical Oral Implants Research, 17(6),
    736-745.

Hamilton, D. W., Chehroudi, B., & Brunette, D. M. (2007). Comparative response of
   epithelial cells and osteoblasts to microfabricated tapered pit topographies in vitro
   and in vivo. Biomaterials, 28(14), 2281-2293.

Hong, H. L., & Brunette, D. M. (1987). Effect of cell shape on proteinase secretion by
   epithelial cells. Journal of Cell Science, 87 ( Pt 2)(Pt 2), 259-267.
                                                                                        153



Hormia, M., Kononen, M., Kivilahti, J., & Virtanen, I. (1991). Immunolocalization of
   proteins specific for adhaerens junctions in human gingival epithelial cells grown on
   differently processed titanium surfaces. Journal of Periodontal Research, 26(6), 491-
   497.

Jones, J. C., Hopkinson, S. B., & Goldfinger, L. E. (1998). Structure and assembly of
    hemidesmosomes. BioEssays : News and Reviews in Molecular, Cellular and
    Developmental Biology, 20(6), 488-494.

Kawahara, H., Kawahara, D., Hashimoto, K., Takashima, Y., & Ong, J. L. (1998).
   Morphologic studies on the biologic seal of titanium dental implants. report I. in vitro
   study on the epithelialization mechanism around the dental implant. The
   International Journal of Oral & Maxillofacial Implants, 13(4), 457-464.

Kinch, M. S., Petch, L., Zhong, C., & Burridge, K. (1997). E-cadherin engagement
    stimulates tyrosine phosphorylation. Cell Adhesion and Communication, 4(6), 425-
    437.

Lauer, G., Wiedmann-Al-Ahmad, M., Otten, J. E., Hubner, U., Schmelzeisen, R., &
    Schilli, W. (2001). The titanium surface texture effects adherence and growth of
    human gingival keratinocytes and human maxillar osteoblast-like cells in vitro.
    Biomaterials, 22(20), 2799-2809.

Leblebicioglu, B., Rawal, S., & Mariotti, A. (2007). A review of the functional and
    esthetic requirements for dental implants. Journal of the American Dental
    Association (1939), 138(3), 321-329.

Lee, E. A., & Jun, S. K. (2000). Achieving aesthetic excellence through an outcome-
    based restorative treatment rationale. Practical Periodontics and Aesthetic Dentistry
    : PPAD, 12(7), 641-8; quiz 650.

Lindhe, J., & Berglundh, T. (1998). The interface between the mucosa and the implant.
    Periodontology 2000, 17, 47-54.

Lindhe, J., Berglundh, T., Ericsson, I., Liljenberg, B., & Marinello, C. (1992).
    Experimental breakdown of peri-implant and periodontal tissues. A study in the
    beagle dog. Clinical Oral Implants Research, 3(1), 9-16.

Linkevicius, T., & Apse, P. (2008). Influence of abutment material on stability of peri-
    implant tissues: A systematic review. The International Journal of Oral &
    Maxillofacial Implants, 23(3), 449-456.

Littell, R. C., Milliken, G. A., Stroup, W. W., Wolfinger, R. D., & Schabenberger, O.
     (2006). SAS® for mixed models (2nd ed. ed.). Cary, NC: SAS Institute Inc.
                                                                                          154


Liu, S. C., & Karasek, M. (1989). Isolation and growth of adult human epidermal
     keratinocytes in cell culture. 1978. The Journal of Investigative Dermatology, 92(4
     Suppl), 164S; discussion 165S.

Lodish, H., Berk, A., Zipursky, L. S., Matsudaira, P., Baltimore, D., & Darnell, J. (2000).
    Molecular cell biology (4th edition ed.). 41 Madisohn Avenue, New York, New York
    10010: W. H. Freeman and Company.

Lohmann, C. H., Dean, D. D., Koster, G., Casasola, D., Buchhorn, G. H., Fink, U., et al.
   (2002). Ceramic and PMMA particles differentially affect osteoblast phenotype.
   Biomaterials, 23(8), 1855-1863.

Maddikeri, R. R., Tosatti, S., Schuler, M., Chessari, S., Textor, M., Richards, R. G., et al.
   (2008). Reduced medical infection related bacterial strains adhesion on bioactive
   RGD modified titanium surfaces: A first step toward cell selective surfaces. Journal
   of Biomedical Materials Research.Part A, 84(2), 425-435.

Manicone, P. F., Rossi Iommetti, P., & Raffaelli, L. (2007). An overview of zirconia
   ceramics: Basic properties and clinical applications. Journal of Dentistry, 35(11),
   819-826.

Marmary, Y., Brunette, D. M., & Heersche, J. N. (1976). Differences in vitro between
   cells derived from periodontal ligament and skin of macaca irus. Archives of Oral
   Biology, 21(12), 709-716.

Masini, V. (2005). Versatility of transmucosal implants with zirconium ring. Minerva
   stomatologica, Vol. 54(1-2), 69-69-77.

Misch, C. E. (2005). Dental implant prosthetics (1st edition ed.). St louis, Missouri
    63146: Elsevier MOSBY.

Miyamoto, S., Teramoto, H., Coso, O. A., Gutkind, J. S., Burbelo, P. D., Akiyama, S. K.,
   et al. (1995). Integrin function: Molecular hierarchies of cytoskeletal and signaling
   molecules. The Journal of Cell Biology, 131(3), 791-805.

Moon, I. S., Berglundh, T., Abrahamsson, I., Linder, E., & Lindhe, J. (1999). The barrier
   between the keratinized mucosa and the dental implant. an experimental study in the
   dog. Journal of Clinical Periodontology, 26(10), 658-663.

Myshin, H. L., & Wiens, J. P. (2005). Factors affecting soft tissue around dental implants:
   A review of the literature. The Journal of Prosthetic Dentistry, 94(5), 440-444.

Park, J. C., Kim, H. M., & Ko, J. (1998). Effects of extracellular matrix constituents on
    the attachment of human oral epithelial cells at the titanium surface. The
    International Journal of Oral & Maxillofacial Implants, 13(6), 826-836.
                                                                                           155


Perala, D. G., Chapman, R. J., Gelfand, J. A., Callahan, M. V., Adams, D. F., & Lie, T.
    (1992). Relative production of IL-1 beta and TNF alpha by mononuclear cells after
    exposure to dental implants. Journal of Periodontology, 63(5), 426-430.

Pingel, L.C., Kohlgraf K.G., Hansen C.J., Eastman C.G., Dietrich D.E., Burnell K.K., et
    al. (2008). Human beta-defensin 3 binds to hemagglutinin B (rHagB), a non-fimbrial
    adhesin from Porphyromonas gingivalis, and attenuates a pro-inflammatory cytokine
    response. Immunology and Cell Biolgy 86,643-649.

Quirynen, M., van der Mei, H. C., Bollen, C. M., Schotte, A., Marechal, M., Doornbusch,
    G. I., et al. (1993). An in vivo study of the influence of the surface roughness of
    implants on the microbiology of supra- and subgingival plaque. Journal of Dental
    Research, 72(9), 1304-1309.

Raisanen, L., Kononen, M., Juhanoja, J., Varpavaara, P., Hautaniemi, J., Kivilahti, J., et
    al. (2000). Expression of cell adhesion complexes in epithelial cells seeded on
    biomaterial surfaces. Journal of Biomedical Materials Research, 49(1), 79-87.

Refai, A. K., Textor, M., Brunette, D. M., & Waterfield, J. D. (2004). Effect of titanium
    surface topography on macrophage activation and secretion of proinflammatory
    cytokines and chemokines. Journal of Biomedical Materials Research.Part A, 70(2),
    194-205.

Ruggeri, A., Franchi, M., Marini, N., Trisi, P., & Piatelli, A. (1992). Supracrestal circular
   collagen fiber network around osseointegrated nonsubmerged titanium implants.
   Clinical Oral Implants Research, 3(4), 169-175.

Ruoslahti, E., & Pierschbacher, M. D. (1987). New perspectives in cell adhesion: RGD
   and integrins. Science (New York, N.Y.), 238(4826), 491-497.

Scarano, A., Piattelli, M., Caputi, S., Favero, G. A., & Piattelli, A. (2004). Bacterial
    adhesion on commercially pure titanium and zirconium oxide disks: An in vivo
    human study. Journal of Periodontology, 75(2), 292-296.

Schaller, M. D. (2001). Biochemical signals and biological responses elicited by the focal
    adhesion kinase. Biochimica Et Biophysica Acta, 1540(1), 1-21.

Schlaepfer, D. D., Hauck, C. R., & Sieg, D. J. (1999). Signaling through focal adhesion
    kinase. Progress in Biophysics and Molecular Biology, 71(3-4), 435-478.

Schmalz, G., Schuster, U., & Schweikl, H. (1998). Influence of metals on IL-6 release in
   vitro. Biomaterials, 19(18), 1689-1694.

Schroeder, A., van der Zypen, E., Stich, H., & Sutter, F. (1981). The reactions of bone,
    connective tissue, and epithelium to endosteal implants with titanium-sprayed
    surfaces. Journal of Maxillofacial Surgery, 9(1), 15-25.
                                                                                         156


Schuler, M., Owen, G. R., Hamilton, D. W., de Wild, M., Textor, M., Brunette, D. M., et
    al. (2006). Biomimetic modification of titanium dental implant model surfaces using
    the RGDSP-peptide sequence: A cell morphology study. Biomaterials, 27(21), 4003-
    4015.

Schupbach, P., & Glauser, R. (2007). The defense architecture of the human periimplant
    mucosa: A histological study. The Journal of Prosthetic Dentistry, 97(6 Suppl), S15-
    25.

Shiraiwa, M., Goto, T., Yoshinari, M., Koyano, K., & Tanaka, T. (2002). A study of the
    initial attachment and subsequent behavior of rat oral epithelial cells cultured on
    titanium. Journal of Periodontology, 73(8), 852-860.

Silva, V. V., Lameiras, F. S., & Lobato, Z. I. (2002). Biological reactivity of zirconia-
    hydroxyapatite composites. Journal of Biomedical Materials Research, 63(5), 583-
    590.

Simion, M., Baldoni, M., & Rossi, P. (1991). A study on the attachment of human
    gingival cell structures to oral implant materials. The International Journal of
    Prosthodontics, 4(6), 543-547.

Spyrou, P., Papaioannou, S., Hampson, G., Brady, K., Palmer, R. M., & McDonald, F.
    (2002). Cytokine release by osteoblast-like cells cultured on implant discs of varying
    alloy compositions. Clinical Oral Implants Research, 13(6), 623-630.

Squier, C. A., & Collins, P. (1981). The relationship between soft tissue attachment,
    epithelial downgrowth and surface porosity. Journal of Periodontal Research, 16(4),
    434-440.

Stanford, C. M. (2006). Advancements in implant surface technology for predicable long-
    term results

Stanford, C. M., Solursh, M., & Keller, J. C. (1999). Significant role of adhesion
    properties of primary osteoblast-like cells in early adhesion events for chondroitin
    sulfate and dermatan sulfate surface molecules. Journal of Biomedical Materials
    Research, 47(3), 345-352.

Steflik, D. E., McKinney, R. V., Sisk, A. L., Koth, D. L., Singh, B. B., & Parr, G. R.
     (1990). Ultrastructural investigations of the bone and fibrous connective tissue
     interface with endosteal dental implants. Scanning Microscopy, 4(4), 1039-47;
     discussion on 1047-8.

Sullivan, R. M. (2001). Implant dentistry and the concept of osseointegration: A historical
     perspective. Journal of the California Dental Association, 29(11), 737-745.

Swab, J. J. (1991). Low-temperature degradation of Y-TZP materials. Journal of
   Materials Science, 26(24), 6706.
                                                                                        157


Takata, T., Katauchi, K., Akagawa, Y., & Nikai, H. (1994). New connective tissue
    attachment formation on various biomaterials implanted in roots. The International
    Journal of Oral & Maxillofacial Implants, 9(1), 77-84.

Thomsen, P., Larsson, C., Ericson, L. E., Sennerby, L., Lausmaa, J., & Kasemo, B.
   (1997). Structure of the interface between rabbit cortical bone and implants of gold,
   zirconium and titanium. Journal of Materials Science.Materials in Medicine, 8(11),
   653-665.

Torricelli, P., Verne, E., Brovarone, C. V., Appendino, P., Rustichelli, F., Krajewski, A.,
    et al. (2001). Biological glass coating on ceramic materials: In vitro evaluation using
    primary osteoblast cultures from healthy and osteopenic rat bone. Biomaterials,
    22(18), 2535-2543.

Tosatti, S., De Paul, S. M., Askendal, A., VandeVondele, S., Hubbell, J. A., Tengvall, P.,
    et al. (2003). Peptide functionalized poly(L-lysine)-g-poly(ethylene glycol) on
    titanium: Resistance to protein adsorption in full heparinized human blood plasma.
    Biomaterials, 24(27), 4949-4958.

VandeVondele, S., Voros, J., & Hubbell, J. A. (2003). RGD-grafted poly-L-lysine-graft-
   (polyethylene glycol) copolymers block non-specific protein adsorption while
   promoting cell adhesion. Biotechnology and Bioengineering, 82(7), 784-790.

Watkin, A., & Kerstein, R. B. (2008). Improving darkened anterior peri-implant tissue
   color with zirconia custom implant abutments. Compendium of Continuing
   Education in Dentistry (Jamesburg, N.J.: 1995), 29(4), 238-40, 242.

Welander, M., Abrahamsson, I., & Berglundh, T. (2008). The mucosal barrier at implant
   abutments of different materials. Clinical Oral Implants Research, 19(7), 635-641.
   doi:10.1111/j.1600-0501.2008.01543.x

Wennerberg, A., Sennerby, L., Kultje, C., & Lekholm, U. (2003). Some soft tissue
   characteristics at implant abutments with different surface topography. A study in
   humans. Journal of Clinical Periodontology, 30(1), 88-94.

Werb, Z., Hembry, R. M., Murphy, G., & Aggeler, J. (1986). Commitment to expression
   of the metalloendopeptidases, collagenase and stromelysin: Relationship of inducing
   events to changes in cytoskeletal architecture. The Journal of Cell Biology, 102(3),
   697-702.

Wheeler, S. L., Vogel, R. E., & Casellini, R. (2000). Tissue preservation and maintenance
   of optimum esthetics: A clinical report. The International Journal of Oral &
   Maxillofacial Implants, 15(2), 265-271.

Zarb, G. A., & Schmitt, A. (1993a). The longitudinal clinical effectiveness of
    osseointegrated dental implants in anterior partially edentulous patients. The
    International Journal of Prosthodontics, 6(2), 180-188.
                                                                                      158


Zarb, G. A., & Schmitt, A. (1993b). The longitudinal clinical effectiveness of
    osseointegrated dental implants in posterior partially edentulous patients. The
    International Journal of Prosthodontics, 6(2), 189-196.

Zarb, G. A., Smith, D. C., Levant, H. C., Graham, B. S., & Staatsexamen, W. Z. (1979).
    The effects of cemented and uncemented endosseous implants. The Journal of
    Prosthetic Dentistry, 42(2), 202-210.

Zarb, G. A., Zarb, F. L., & Schmitt, A. (1987). Osseointegrated implants for partially
    edentulous patients. interim considerations. Dental Clinics of North America, 31(3),
    457-472.

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:5
posted:5/16/2012
language:English
pages:170
fanzhongqing fanzhongqing http://
About