Arginine functionalisation of hydrogels for heparin binding –a

							Arginine functionalisation of hydrogels for heparin binding –a novel
approach to developing a pro-angiogenic biomaterial.




Louisa Gilmore , Steve Rimmer , Sally L Mc Arthur , Shweta Mittar and
Sheila MacNeil




16th May
Abstract
The process of new blood vessel formation from previously established blood
vessels is vital for wound healing and is regulated by a number of stimulators
and inhibitors of which vascular Endothelial Growth Factor (VEGF) and
heparin play major roles. VEGF stimulates proliferation and migration of
endothelial cells and is influenced by interactions with heparin-like
glycosaminoglycans, which act to bind and stabilise this labile protein in vivo.

The aim of this study was to synthesise a biomaterial which exploits the ability
of heparin to bind and release bioactive VEGF to stimulate angiogenesis in a
wound or for tissue engineering applications. The approach explored was to
amine modify a biocompatible hydrogel to bind heparin based on electrostatic
interactions which in turn would bind VEGF.

Hydrogels were produced from poly (NVP-co-DEGBAC-co-AA) and
functionalised with either synthetic trilysine or triarginine peptides. These
peptide sequences interacted with the highly sulphated heparin which in turn
was able to interact with the heparin-binding domain of VEGF165. Both
hydrogels were able to bind and release bioactive VEGF165 with the arginine-
modified     hydrogel    outperforming      the   lysine-modified    hydrogel.
Cytocompatibility of the base hydrogels was confirmed in vitro with primary
human dermal fibroblasts and human dermal microvascular endothelial cells
(HDMECs). The triarginine-functionalised hydrogels , and to a lesser extent
the lysine functionalised hydrogels were found to significantly stimulate
HDMEC proliferation once loaded with heparin and VEGF suggesting that
these hydrogels are a promising approach for providing a pro-angiogenic
stimulus for neovascularisation.


Introduction

There is a major need for biomaterials to promote angiogenesis both for
patients with chronic non-healing wounds and also for the tissue engineering
community where tissue engineered materials often fail because of slow
neovascularisation.

In the normal healthy adult, wound healing occurs in a smooth and well
regulated manner (unless infection intervenes ) but in the elderly or others
with vascular problems ,often diabetes related, the surrounding vasculature is
compromised and wounds can fail to heal. This failure to heal of chronic
wounds (encompassing venous ulcers, pressure ulcers or neuropathic ulcers)
is a major drain on western healthcare systems worldwide and a severe
burden for these patients .
Additionally in tissue engineering, tissues such as laboratory expanded skin
can be produced which are capable of providing a good barrier function but
these can fail to engraft on the patient because of delays in angiogenesis.
This is because normally a conventional split thickness skin graft contains
intrinsic vasculature architecture and this graft will initially survive via
inosculation between some of these vessels and the underlying blood vessels
of the wound bed. In tissue engineered skin there are no blood vessels to
connect with the underlying wound bed and these tissue engineered grafts
are entirely dependent on neo-angiogenesis. This is acknowledged as a
major problem for engraftment of all tissue engineered constructs with the
exception of the avascular tissues of cartiliage and cornea.

The process of new blood vessel growth into tissues, neovascularisation , is a
fundamental requirement for the developing embryo and for normal wound
healing in adults and throughout the female reproductive cycle (Klagsbrun and
D'Amore 1996). Healing in normal wounds occurs as a sequential cascade of
overlapping processes that involves the release of proangiogenic growth
factors from the injured tissue. These diffuse into the nearby tissue and bind
to specific endothelial cell receptors on the pre-existing blood vessels,
initiating a cascade of events. Very early in wounding vasodilatation occurs
increasing vascular permeability in response to a major proangiogenic
mitogen, vascular endothelial growth factor (VEGF). This allows release of
plasma proteins which produces a provisional scaffold for endothelial cell
migration (Dvorak, Brown et al. 1995; Dvorak 2000). Vasodilatation is
followed by proteolytic degradation of the basement membrane and
extracellular matrix (19). The endothelial cells then proliferate and migrate
towards the injured tissue regulated by locally released chemical stimuli with
further ECM degradation to accommodate the sprouting vessel (Lamalice, Le
Boeuf et al. 2007). The sprouting blood vessels, initially present as a solid
cord, then form a lumen and increase in vessel diameter and length by the
initial thinning of the endothelial cells and fusion with other blood vessels.
Finally maturation of the blood vessel occurs, including the introduction of
mural cells (pericytes or smooth muscle cells), production of basement
membrane and, in many cases, vascular regression (Darland and D'Amore
1999).

This complex sequence of activities is regulated by a number of angiogenic
stimulators and inhibitors. Key stimulators include vascular endothelial growth
factor (VEGF), fibroblast growth factor-1 (FGF1) and 2 (FGF2), transforming
growth factor alpha (TGFα), transforming growth factor beta (TGFβ) and
platelet derived growth factor (PDGF). There are also a range of inhibitors
and angiogenesis is controlled by the balance between stimulators and
inhibitors.

As the challenge of promoting angiogenesis is a major one there is an
extensive research in this area..      Initial     approaches focussed on
administering single growth factors such as VEGF or bFGF . While these
were shown to stimulate angiogenesis in animal and small scale clinical
studies (Takeshita, Zheng et al. 1994; Hopkins, Bulgrin et al. 1998;
Schumacher, Pecher et al. 1998; Laham, Sellke et al. 1999; Kipshidze,
Chekanov et al. 2000; Sun, Chen et al. 2005) effects were minor and some
studies also identified (as reviewed by Epstein et al and Fuchs et al. 2001)
some undesirable results. These were the development of new blood vessels
in non-target tissues, an increase of vascular permeability in non-target
tissues and the growth of tumours. This led to the view that low and localised
doses of angiogenic growth factors are preferable to one off large doses of
these expensive peptides. Accordingly considerable research effort was then
directed towards the localised delivery of proteins using biomaterials as
delivery/release vehicles for angiogenic factors.

In vivo heparin and heparan sulphate GAGs play a significant role in
regulating angiogenesis and accordingly several natural and synthetic routes
to a heparin-mimetic have been described. These are capable of releasing
bioactive growth factors but they also generally fail to offer spatial localisation
as the heparin itself can be rapidly lost from the wound site. Accordingly
current studies seek to immobilise heparin. For example Steffens et al
covalently incorporated heparin into collagen matrices using 1-ethyl-3-(3-
dimethylaminopropyl)-carbodiimide (EDC) and N-hydroxysuccinimide (NHS)
and found an increase in endothelial cell proliferation compared to the use of
collagen matrices alone. A further increase in proliferation was observed as
VEGF was loaded into these matrices. Another two-component delivery
system was reported by Jeon et al (Jeon, Kang et al. 2006). Heparin was
incorporated covalently into poly(L-lactide-co-glycolide) nanospheres in the
presence of EDC to enable long-term zero-order delivery of bFGF. These
nanospheres were then further incorporated into the fibrin gels to get a
release period of 4 weeks and the bFGF was shown to be bioactive in vitro
leading to microvessel formation in a mouse ischemic limb model.

In summary the evolving literature provides a strong consensus that while
pro-angiogenic growth factors can stimulate angiogenesis these are better
delivered released from a biomaterial which initially binds them then slowly
releases these proteins. This prevents uncontrolled diffusion of growth factors
and improves the stability of these labile peptides which are not degraded
while bound . There are also now several promising studies incorporating
heparin into biomaterials, suggesting the presence of heparin is able to further
improve the ability of growth factor delivery to blood vessels to stimulate
angiogenesis.

Accordingly in the current study we have sought to take things one step
further to make a cost effective, simple and reproducible biomaterial which
can bind heparin on the basis of hydrostatic rather than covalent interactions
and then rely on the high affinity binding of this native heparin for VEGF to
produce a hydrogel capable of promoting angiogenesis. To this end
hydrogels were functionalised with basic peptide sequences to explore the
hypothesis that a sequence of basic peptide sequences would bind heparin
which in turn would bind and then release VEGF in a biologically active form.

RESULTS

Peptide functionalisation of hydrogels

Poly(NVP-co-DEGBAC) (PND) and poly(NVP-co-DEGBAC-co-AA) (PNDA)
hydrogels were synthesised by thermal and UV polymerisation and analysed
for equilibrium water content and for residual monomer concentration by
NMR.
The optimisation of peptide coupling was then undertaken . Activation of the
hydrogels was by DCC and NHS in DMF at room temperature for 24 hours,
This was followed by peptide coupling in PBS (pH 8.5) at 0°C for 72 hours,
resulting in the synthesis of trilysine- and triarginine-functionalised poly(NVP-
co-DEGBAC-co-AA) hydrogels. Figure 1 summarises this synthesis and
peptide modification.

The presence of peptides was confirmed by ToF-SIMS analysis as shown in
Figure 2 where fragments characteristic for lysine and arginine are indicated.
Colorimetric analysis of hydrogels using TNBS gave average peptide loadings
of 0.14g/g hydrogel and 0.20g/g hydrogel for lysine and arginine respectively
as summarised in Table 1 for 5 separate batches of hydrogels.




Table 1: Summary of properties of hydrogels
                                    Hydrogel
 Hydrogel Properties                PND PNDA           PNDKKK         PNDRR
                                                                      R
 % composition(a)
 NVP                                  98   93      93           93
 AA                                   0    5       5            5
 DEGBAC                               2    2       2            2
 EWC (%)(b)                           86.5 92.9    92.9         92.9
 Peptide       loading      (mg/mg ~       ~       0.14      ± 0.20       ±
 hydrogel)(c)                                      0.01         0.01
(a) The composition of the hydrogels were calculated from the original
monomer composition utilised in synthesis (b) EWC results were an average
of 6 measurements (c) Peptide loading as determined by indirect colorimetric
TNBS analysis. Results are an average of 5 separate batches of polymers
used in cell culture experimentation.
NMR?? showed the peptide-functionalised hydrogels contained no residual
monomer and to have suitable mechanical properties and so were analysed
further for heparin and VEGF binding as well as in cell culture
experimentation.


Ability of peptide functionalised hydrogels to bind heparin

The heparin used was the sodium salt from porcine mucosa , a natural
unfractionated heparin with molecular weight ranging between 17,000 to
19,000 Da. The binding of heparin to the functionalised hydrogels was
analysed using fluorescently labelled heparin and X-ray photoelectron
spectroscopy (XPS) ( Figure 3) and fluorescently labelled heparin ( Figure 4).

XPS was used to quantitatively assess heparin binding to acid-and peptide-
functionalised hydrogels by analysis of the sulphur (S2p) spectrum (sulphate
content of heparin) and the effect of heparin on the carbon peak (C 1s) of the
hydrogel spectra. Analysis of the hydrogels was performed before and after
heparin binding. S2p signal depicted the qualitative approach to assessing
heparin binding thereby providing information on the relative sulphur content
of the different hydrogels pre and post heparin binding. This data is shown in
Figure 3a and where it was clear that there was no increase in the sulphur
content post exposure to heparin with acid-functionalised hydrogel. This infers
that little heparin remain on this material after the extensive washing protocol.
With respect to the trilysine hydrogel a slight increase was seen in the sulphur
signal showing that heparin has attached to the surface and remained
attached throughout the washing process (Figure 3b). The greatest increase
in intensity of the sulphur spectrum was seen with the triarginine hydrogel,
confirming that more heparin was retained by this hydrogel (Figure 3.c).

Analysis of the C1s profiles (Figure 3) revealed that there was no significant
change in the ratio of carbon environments for acid –functionalised hydrogels
pre and post exposure to heparin suggesting that no or little heparin was
bound in this case. In contrast in the case of peptide (KKK- and RRR-)
functionalised hydrogels before and after exposure to heparin showed a clear
decrease in the contribution from C-O/C-N bonds (286eV) and amide bonds
(CNO, 288eV) and an increase in acid bonds (COOH 289eV) inferring that
the bonds previously observed for the hydrogel may now be partially masked
by the presence of heparin. These results are summarised in Table 2.

Figure 4 shows a quantitative analysis of the binding of fluorescently labelled
heparin to the hydrogels. After extensive (x 7) washes, no fluorescent tagged
heparin was found in the wash solutions of the hydrogels exposed to heparin.
This was a good indication of the strength of the binding of the heparin to the
hydrogels and this washing protocol was then used for all future experiments.
This approach showed that hydrogels functionalised with arginine peptides
bound 20.7% of the heparin in these experiments compared to trilysine
functionalised hydrogels which bound 13.5%.            In contrast the acid
functionalised hydrogels bound only 5.2% of the heparin. These values
equate to 0.03mg, 0.08mg and 0.12mg of heparin bound per mg of dry
hydrogel for acid, KKK and RRR functionalised hydrogels respectively.

Ability of hydrogels to bind and release VEGF

The acid and peptide functionalised hydrogels were then loaded with heparin
and then with VEGF and the binding of VEGF to the hydrogels and its
subsequent release were then examined, assaying the VEGF in the media
and washes by ELISA. Results are shown in Figure 5.

Analysis of the VEGF detected in the wash solutions compared to the amount
of VEGF initially exposed to the hydrogels showed that the peptide-
functionalised hydrogels were able to absorb significant amounts of VEGF.
PND and PNDA polymers were found to absorb 24.6ng and 27.5ng per
hydrogel disc respectively, in comparison to 28.9ng and 31.5ng for the
PNDKKK and PNDRRR materials. These values equate to binding capacities
of 21.5ng, 24.0ng, 25.2ng and 27.5ng of VEGF per milligram of dry hydrogel
for PND, PNDA, PNDKKK and PNDRRR respectively (Figure 5). After three
washes at 4°C to remove unbound material it was calculated that the
nonfunctionalised PND retained 21.7ng (19.0ng/mg dry hydrogel) of the
VEGF it was exposed to, which was slightly less than the 23.5ng (20.5ng/mg
dry hydrogel) retained by the PNDA polymer. In contrast statistical analysis (
ANOVA analysis with post-hoc tukey correction) showed significantly more
VEGF( 30.8ng (26.9ng/mg dry hydrogel)) remaining attached to the PNDRRR
hydrogel after washes.

Release of VEGF at 37ºC ( Figure 5b) showed that all polymers released
VEGF throughout the 72hrs of the study period . The VEGF release pattern
was very similar for the PND, PNDA and PNDKKK hydrogels and over this
period, a total of 19.8ng (17.3 ng/mg dry hydrogel), 19.4ng (16.9 ng/mg dry
hydrogel), and 20.1ng (17.6 ng/mg dry hydrogel) of VEGF was released per
hydrogel disc for the 3 polymer types respectively, a release of 91%, 83% and
79% of the protein remaining after the wash steps for PND, PNDA and
PNDKKK. At all time points analysed, the PNDRRR hydrogel was found to
release more VEGF than the other polymers, with 27.0ng (23.6 ng/mg dry
hydrogel) of VEGF released per hydrogel disc in the 72 hour period, 88% of
the protein that remained after the wash step. ANOVA with post-hoc Tukey
analysis of the results revealed that this difference in release was significant
after 6 hours. This result is as expected when considering the higher heparin
loading of the RRR-functionalised material.


Effect of peptide-functionalised         hydrogels     on cell     viability and
attachment

All hydrogels were then examined for both biocompatibility ,looking at the
metabolic ability of fibroblasts and endothelial cells in indirect contact with the
hydrogels , and also for the ability of these modified hydrogels to support cell
adhesion and metabolic activity.

Results are shown in Figures 6-8. Hydrogels were placed in culture wells and
cells added and cell morphology was visualised by photographing cells using
both light and fluorescence microscopy and metabolic activity was measured
quantitatively using the MTT assay for both the cells growing on the TCP and
also for those cells which attached to the hydrogels.

The fibroblasts used for testing hydrogels were from different passages and
donors and the hydrogel samples were also from multiple batches.
Fibroblasts were also cultured in fibroblast growth media both with and
without 10% FCS (fetal calf serum) for 96 hours to assess if presence of
serum influenced the reaction of the cells to the presence of the hydrogel.

Cells on the TCP surrounding the hydrogels were seen to grow up to the
edges of all of these hydrogels - not shown as photography of non-labelled
cells on these hydrogels was difficult. Accordingly as seen in Figures 6 and 8
cells were then fluorescently labelled with Cell TrackerTM Red for
photography. Cells on the TCP had normal morphology suggesting that these
hydrogels are cell compatible. In contrast, very few cells were observed to
attach to the PNDA hydrogel , a few attached to the PNDKKK hydrogel and
significantly more attached to the PNDRRR hydrogels as shown in Figure 6
and assessed quantitatively in Figure 7. Fibroblasts on the PNDA and
PNDKKK hydrogels also appeared rounded in contrast to those on the
PNDRRR hydrogel which had a more normal morphology.

Examination of the tissue culture plastic wells after removal of the hydrogel
post-MTT assay revealed that no fibroblasts were attached to the tissue
culture plastic under the hydrogel, therefore for quantitative analysis the MTT
response of the control was adjusted to give an equal tissue culture plastic
surface area to that available in the hydrogel-containing wells. Figure 7 shows
that cell viability of the fibroblasts grown on TCP adjacent to the hydrogels
was unaffected by the presence of the hydrogels. For cells grown on the
hydrogels there was a significant increase in viability only in case of the
PNDRRR hydrogel.


Figures 6 and 7 show results for cells grown with 10% foetal calf serum in the
media. The same experiments were also conducted in the absence of serum (
not shown) . Fibroblasts appeared to have normal morphology adjacent to the
hydrogels and MTT analysis based on 3 separate experiments showed no
significant effect of culturing fibroblasts in serum-free media for PND, PNDA
or PNDKKK hydrogels. In case of PNDRRR hydrogels however, all the cells
on TCP and surrounding the hydrogel had a rounded morphology and
metabolic activity assessed by MTT activity was reduced by approximately
50% (p<0.001). In serum free media no fibroblasts attached to any of the
hydrogels ( not shown).

Similar experiments exposing hydrogels to human dermal microvascular
endothelial cells for 96hrs were conducted in endothelial cell media containing
5 % serum. The quantitative MTT data (not shown) confirmed that none of the
four hydrogels tested had any cytotoxic effect on the HUDMECs.

Assessment of endothelial cell responses to peptide-functionalised
hydrogel loaded with heparin and VEGF.
The hydrogels were then incubated for 96 hrs with heparin and with VEGF
and their effect on endothelial cells examined. Endothelial cells were
examined both in the presence of 5 and 2% FCS. (Endothelial cell
experiments with hydrogels were conducted in media with 2% rather than 0%
FCS as these cells were not sustainable in serum-free media for the length of
these experiments.)

Qualitative results for the ability of endothelial cells to attach to hydrogels are
shown in Figure 8 for fluorescently labelled cells. There was no attachment of
cells to the PND and PNDA hydrogels, a few cells attached to PNDKKK but
clearly more cells attached to the PNDRRR . Endothelial cells attached to the
hydrogels had a rounded appearance and appeared to be in clumps or
aggregates or cells.
In contrast endothelial cells which grew on the TCP had a cobblestone
appearance (a hallmark of endothelial cells) and proliferated right up the
edge of the hydrogels ( not shown). On removal of the hydrogels, as with the
fibroblasts, the endothelial cells had not attached to the underlying tissue
culture plastic.
.
The hydrogels loaded with heparin and VEGF had a slight stimulatory effect
on the metabolic activity of the adjacent endothelial cells. While this was
significant for 2 out of 3 experiments when cells were cultured in 5% FCS (
not shown) it was seem nuch more clearly when the FCS was reduced to 2%
as shown in Figure 9 . Here results are significant in 3 out of 3 experiments
when endothelial cells are cultured next to PNDRRR hydrogels preloaded with
heparin and VEGF. A lesser stimulation was seen in 2 out of 3 experiments
with PNDKKK hydrogels loaded with heparin and VEGF.


DISCUSSION

To be done



METHODS

To be done




References ( INCOMPLETE)
Darland, D. C. and P. A. D'Amore (1999). "Blood vessel maturation: vascular
development comes of age." J Clin Invest 103(2): 157-8.

Dvorak, H. F. (2000). "VPF/VEGF and the angiogenic response." Semin
Perinatol 24(1): 75-8.

Dvorak, H. F., L. F. Brown, et al. (1995). "Vascular permeability
factor/vascular endothelial growth factor, microvascular hyperpermeability,
and angiogenesis." Am J Pathol 146(5): 1029-39.

Epstein, S. E., S. Fuchs, et al. (2001). "Therapeutic interventions for
enhancing collateral development by administration of growth factors: basic
principles, early results and potential hazards." Cardiovasc Res 49(3): 532-42.

Hopkins, S. P., J. P. Bulgrin, et al. (1998). "Controlled delivery of vascular
endothelial growth factor promotes neovascularization and maintains limb
function in a rabbit model of ischemia." J Vasc Surg 27(5): 886-94; discussion
895.

Jeon, O., S. W. Kang, et al. (2006). "Long-term and zero-order release of
basic fibroblast growth factor from heparin-conjugated poly(L-lactide-co-
glycolide) nanospheres and fibrin gel." Biomaterials 27(8): 1598-607.
Kipshidze, N., V. Chekanov, et al. (2000). "Angiogenesis in a patient with
ischemic limb induced by intramuscular injection of vascular endothelial
growth factor and fibrin platform." Tex Heart Inst J 27(2): 196-200.

Klagsbrun, M. and P. A. D'Amore (1996). "Vascular endothelial growth factor
and its receptors." Cytokine Growth Factor Rev 7(3): 259-70.

Laham, R. J., F. W. Sellke, et al. (1999). "Local perivascular delivery of basic
fibroblast growth factor in patients undergoing coronary bypass surgery:
results of a phase I randomized, double-blind, placebo-controlled trial."
Circulation 100(18): 1865-71.

Lamalice, L., F. Le Boeuf, et al. (2007). "Endothelial cell migration during
angiogenesis." Circ Res 100(6): 782-94.


Schumacher, B., P. Pecher, et al. (1998). "Induction of neoangiogenesis in
ischemic myocardium by human growth factors: first clinical results of a new
treatment of coronary heart disease." Circulation 97(7): 645-50.

Sun, Q., R. R. Chen, et al. (2005). "Sustained vascular endothelial growth
factor delivery enhances angiogenesis and perfusion in ischemic hind limb."
Pharm Res 22(7): 1110-6.

Takeshita, S., L. P. Zheng, et al. (1994). "Therapeutic angiogenesis. A single
intraarterial bolus of vascular endothelial growth factor augments
revascularization in a rabbit ischemic hind limb model." J Clin Invest 93(2):
662-70.