Process Proposal for MAb Production in
Dunaliela tertiolecta
Bi Motum
Moving Tomorrow’s Technology
Katie Thompson
Lisa Feenstra
Dieu Pham
Rodolfo Ham-Zhu
Report Submitted: 6/6/2006
For: Dr. Haluk Hamamci
MAb Background and Process Proposal
Therapeutic Protein History
The introduction of monoclonal antibodies has enriched the tapestry of the biological and
biomedical field given that it has extended our understanding of immunology and provided new
strategies for the combating and treatment of immune deficient diseases. Produced through the
use of hybridoma cells for therapeutic purposes, recombinant protein antibodies like Avastin,
which is currently being marketed as a treatment for colon and rectal cancer, have pioneered a
multi-billion dollar biotechnology industry.
Vascular Endothelial Growth Factor, VEGF, a growth signal sent out by the tumor,
enables tumors to receive nutrients. Avastin functions by attaching directly to VEGF therefore
preventing the tumor blood vessels from growing and spreading. Targeting the tumor’s blood
vessels reduces the size of the tumor because in order for tumors to grow they need a constant
supply of oxygen and other nutrients. Thus by cutting off the tumor’s supply line, they can no
longer grow and spread.
Market Considerations________________________________
Antibody drugs are expected to exceed $68 billion in worldwide sales by the year 2010
(Liszewski). The recent prospects of expanding Avastin to treat both lung and breast cancer
patients may play a large part in the therapeutic protein market. Considering that the yearly
number of diagnosed colon cancer patients is about one third of both breast and lung cancer
combined (see Figure 4), and that the dosage for the potential patients is twice that of patients
currently using Avastin, it is no wonder that the demand for this product is expected to clearly
surpass the supply (see Tables 1-3). If we do some rough calculations with current Genentech
capacity in mind, we can approximately quantify the increased demand of this particular product.
As illustrated in Table 2, 1.6 million grams of additional Avastin is needed yearly to meet
demands for all three cancers. Considering that Genentech plans on utilizing a 90,000 liter
facility to meet this demand and current mammalian cell lines produce roughly only 0.3grams of
MAb per liter of production (Hu, Wei-Shou), Table 3 illustrates that the exceeded yearly demand
for this protein alone could be approximately 19,000 grams.
1
Figure 1 – Number of cancer cases annually
Source: Biotechnology Industry Organization
Table 1: Current Avastin Use: Colon Cancer Patients
Avastin sales for 20051 $1,133,000,000
Cost per gram of Avastin1 $5,500
Grams sold in 2005 206,000grams
2
Mg Avastin per kg body weight for colon cancer patients 5mg/Kg
Mg used per dose (using 70kg as average body weight) 350mg
Grams used per year, per patient (assuming one dose every two weeks) 8.4grams
Number of patients in 2005 using Avastin (g sold/g per patient) 24,524
3
Number of colon cancer patients in 2005 154,000
Percentage of colon cancer patients using Avastin 16%
Table 2: Projected Avastin Use Increase: Lung and Breast Cancer
Number of current lung cancer patients3 167975
3
Number of current breast cancer patients 433935
Potential increase in Avastin patients due to lung cancer (using 16% of total) 26876
Potential increase in Avastin patients due to breast cancer (using 16% of 69430
total)
Mg Avastin per Kg body weight for both lung and breast cancer2 10mg/Kg
Mg used per dose (using 70kg as average body weight) 700mg
Grams used per year, per patient (assuming one dose every two weeks) 16.8grams
Yearly increase of Avastin sold due to lung and breast cancer 1.6 million
grams
Yearly Avastin sales increase due to lung and breast cancer $8.9 billion
2
Table 3: Increased Production Demand
Genentech’s expected capacity to meet this demand2 90,000Liters
Yield MAb per liter using current mammalian cell line 0.3grams/Liter
Gram MAb per gram Avastin (approximate)4 .025g/g
Total gram of MAb produced per fermentation 27000 grams
Total gram of Avastin produced per fermentation 1,080,000 grams
Total gram of Avastin needed (current plus potential increase) 1,823,941grams
Total grams of Avastin needed that will exceed Genentech 743,941grams
capacity
Total grams of MAb needed that will exceed Genentech 19,000 grams
capacity
Information sources for Tables 1-3:
1
-Genentech Financial Report 2005
2
-Hu, Jim, Blogs for Industry
3
-American Cancer Society
4
-Food and Drug Administration
In addition to the demand increase due to the existing drugs on the market ( 22% MAb),
the pipeline of 200+ pending therapeutic protein pharmaceuticals (57% MAb) also pushes for
increased production levels (Steiner). Because of the expected rise in required production, most
of the pharmaceutical and biotechnology companies already outsourcing some of their
biopharmaceutical manufacturing are planning to increase their spending in the next two years.
Median contract spending in 2004 was reported as roughly $3 to $4 million per contract
organization per year and this number has been predicted to double by the end of this year (Fox).
The size of the worldwide contract manufacturing market is expected to reach $2.5 billion this
year (Fox). Considering the nearly two-fold increase from the $1.7 billion earned in 2004,
contract manufacturing is looking at a very bright future.
Process Cost Considerations
As competition and production volumes increase in response to the accelerated demands,
incentives for improving efficiencies are expected to take a big leap forward. In other words, low
production costs will be crucial for the success of our company. Although lower production
costs are desired, lower yields are not. When considering current expression systems utilized to
fabricate monoclonal antibodies, there are both advantages and disadvantages in terms of the
bottom line.
There are currently two strategies for producing monoclonal antibodies like Avastin:
CHO cells lines and plant cultures. There are many differences between plant, and CHO based
expression systems. These differences become very important when considering how to make
our organization as competitive as possible and are summarized below in Table 4.
3
Table 4: Comparison of CHO and Plant Cell Expression Systems
Expression CHO Plant Cells (Traditional and algae)
System
Shear Resistance Lowest Medium
Respiration Rate Lowest Medium
Growth Rate Medium Slow (traditional plants such as tobacco)
Fast (algae)
Doubling Time 1 day One week (traditional plant)
3.5 hours (algae)
Excreted Protein Yes Yes
Post-translational Yes. Post-translational Yes, with some restrictions.
Modifications modifications are almost Plant proteins may be immunogenic.
identical to humans.
Problems faced in Oxygen transfer is often a Getting enough light to cells is often a
production problem because cells cannot problem. Plant cultures can grow densely
withstand high agitation or and become viscous.
aeration.
Media Complex. Contains serum, Simple.
which is expensive and easily
contaminated.
Safety Can transmit disease-causing Cannot transmit animal diseases.
agents such as viruses.
Other Suitable for large and Easy to scale up.
complicated proteins.
Currently monoclonal antibodies like Avastin (Bevacizumab) are being mass produced in
a Chinese Hamster Ovary (CHO) mammalian cell expression system in a nutrient medium
containing the antibiotic gentamicin. Through post-translational modifications and
Bevacizumab’s covalent structure, the molecule was modified to posses similar characteristics of
a human IgG1 (European Medicines Agency).1 The current expression vector used to express
this antibody is the plasmid pSVID5.ID.LLnspeV.xvegf36HC.LC. The plasmid is encoded with
Bevacizumab and then inserted into CHO DP-12 (European Medicines Agency)2. To begin the
inoculation train 20L of culture cells are inoculated, and the cells sub-cultivated every three to
four days in a subsequent series of increasing volumes (80L 400L 2000 L). An aliquot of
the cells are then placed in a 12000L CHO cell fed-batch suspension culture for harvesting
(European Medicines Agency).3 Post fermentation, the cells are run through a recovery train
which isolates and purifies the therapeutic protein. Due to the animal protein requirements of this
1
Scientific Discussion. European Medicines Agency. 4 Apr. 2006. pg 7.
2
IBIDEM. pg 4.
3
IBID. pg 5.
4
cell line, viral contamination is an issue and involves rather expensive filtering and deactivation
procedures. Overall, the current technology is expensive and time consuming.
Alternatively, two approaches exist to produce these proteins from plants: plant cell and
algae suspension culture. Due to damage by twisting motion (in turbulent mix regime) to plant
cells, gas-to-liquid mass transfer is limited. Because of plant’s low cell growth rate, aseptic
conditions in plant suspension cultures have to be maintained for two to four weeks. Although
shear stress affects the growth of algaein a concentric tube airlift photobioreactor (Contreras, et.
al.), the effects are less drastic than in plant cell cultures (Wu, et. al). In addition, algae grows
much faster than plant cells. For these reasons, algae suspension cultured in photobioreactors are
the optimum choice. Photon flux density (PFD) is one of the most important limiting and to-be-
controlled variables (Merchuck, et. al.). Mathematical models that relate biomass productivity to
average photon flux density –which can be controlled by varying agitation speed, have been
developed by Merchuck. According to Merchuck’s model, the optimum algae concentration is
108 cell/ml – higher than in plant cells. Assuming that productivity is proportionally related to
cell concentration, algae cultures should yield much higher production rates.
Algae is a good candidate because of substantially lowered manufacturing and
maintenance costs as well as higher yields. In order to produce Bevacizumab, an expression
system with post translational modification (PTM) is usually required. PTM’s such as
phosphorylation, glycosylphosphatidylinositol and ubiquitination have been confirmed to exist in
algae strains such as Dunaliela tertiolecta (Kwon et al., 2005). Like CHO cells, algae cell
cultures need to be first grown in a laboratory. But as pointed out in Table 5 below, the algae
cells grow much more quickly than CHO cells. The algae strain Dunaliela tertiolecta will first be
inoculated in disposable wave bags with magnetic stirrer bars for light and CO2 mixing. Wave
bags are superior for inoculation usage because they are biocompatible and have temperatures
tolerance from 0 to 50ºC. In addition, given that wave bags are disposable, a separate CIP unit is
not required for this step of the process. This allows for easier validation and a reduced factory
start-up time. Once the suspension culture is made, it can be used to inoculate a
photobioreactor. A fed-batch system can then be used to operate the photobioreactor to grow the
algae cells and produce the desired product. After the initial inoculation, a sample of the
previous batch culture can be saved to inoculate the next batch. The media for algae cells does
not have to have precursors and is therefore considered simple and free of viral contaminants.
The algae cells need to have light, carbon dioxide, and nutrient rich water. The media needs to
have a source for carbon, nitrogen, sulfur, and phosphate. Once the growth is complete, the cells
are sent to a recovery train which involves centrifugation for both cell isolation and cell debris
removal, cell lysing to extract the desired product, and further purification steps involving
Protein A columns, Gel Chromatography, and Anion Exchange Chromatography. The general
schematic of the proposed process is represented below in Fig. 2.
5
Inoculation Train Cell Concentration
Upstream Processing Build-up
Growth Photobioreactor Cell Growth
Cell Isolation, Broth
Centrifugation Removal
Cell Lysis
Centrifugation Cell Debris Removal
Downstream Processing
Protein A Column Protein Isolation
Gel Column Improper Protein
Chromatography Removal
Cation Exchange Protein Purification
Column
Anion Exchange Protein Purification
Column
Final Protein Product
Figure 2: Process flowsheet of MAb Production via Dunaliela tertiolecta algal cells
6
The following is an estimated timeline for the completion of the project.
Timeline
Program Phase; 1 Month
1. Facility Program
a. Furnishes all data for plant and decide plant capacity
b. List Operational Description- list of all products to be produced, a general description
of a the production process, expectation of process control, and a statement on the
future plans of the plant
c. List of Organization Chart- anticipated facility staff and their main duties
d. Master List of Required Spaces and why they are needed
e. Functional Resolution Analysis- include warehousing needed and ingredient
preparations
f. Comparative Analysis of Spatial Sizes- compares and contrast areas and volumes
using modular increments
g. Estimation of the volume prediction and area needed
2. Site Analysis
a. Decide on location to ensure that it can accommodate the large facility.
b. Location and property survey
c. Conduct a topographic map of the site and explore the subsurface of the desire area
3. Design Criteria
a. Rough design estimate should be done for each piece of material
4. Utility
a. List all utilities needed and their approximate cost; include also cost of peak demands
for each utility item
b. Report all capacity of all utilities used, ie 50 psig of steam available for heat
exchanger
5. Target data for plant completion and execution
6. Prepare a rough project budget
Conceptual Design Phase; 1.5 Months
1. Process Flow Diagrams
2. Process is scaled up and defined and utilities systems are outline
3. List of equipment and their specifications
4. Facility Layout is developed
5. Decide if project is economically feasible by performing detail economic analysis
6. Develop alternative schemes and estimates
7. Generate a general project schedule
Design Development Phase; 1.5 Months
1. P&ID, instrumentations and flow diagrams
2. Prepare list of all equipment needed
3. Prepare equipment specifications
4. Prepare lists of where equipments can be purchase
5. Prepare Piping diagrams
6. Estimate a general cost by taking vendors quotations
7. Also note equipment that needs to be shipped or transported to plant and their costs
7
8. Identify emergency power requirement
9. Control system Configuration
10. Heating, ventilation, and air conditioning systems are defined if needed
11. Define electrical systems
12. Design layout of plant structure
13. Notify FDA concerning plans for project completion and start-up.
Detail Design Phase; 3-4 Months
1. Prepare detail structural drawing of building
2. Show how construction is performed
3. Prepare detail mechanical drawing of building and operating equipment
4. Prepare detail electrical drawings of all operating equipment
5. Purchase equipment
a. Have manufacturer of all equipment bought pre-tested before shipping
6. Deliverable equipment
a. Backup/Extra Equipment Parts
b. Documentation (include number of copies)-including operating manuals, test
results, guarantees, help manuals
c. Drawings-including final P&IDs and wiring diagrams of all equipment purchased
d. Backup copies of software and important manuals
e. List of company contacts and technical support policy/numbers
7. Construction Bid Documents
8. Construct design of upstream and downstream units
Construction Phase; 4 -5 Months
1. Site Preparations
2. Build plant
3. Include all needed wiring and mechanical parts
4. Install all the needed equipment and check to make sure everything is running properly
Validation Phase; 6 Months
1. Document Preparation- prepare checklist and procedures to be used during filed validation
2. Installation Qualification
3. Operational Qualification
Start-Up
8
Upstream Processing
Culture Requirements
Light Energy
The light will be provided using monochromatic lights at various sizes specified in the
following specification tables. Light transfer is an issue, and becomes increasingly so as density
increases, therefore, the size of the light source increases rather exponentially as the size of the
reactor increases. Chlorophyl absorbs at 650-700 nm; this is equivalent to 4.6x1014-4.3 x1014Hz.
At this energy, Planck’s law indicates 3.3-3.5x1018 photons Joule-1. A proton flux density of
58μmol m-2s-1 has given a maximum biomass yield of 1.2 g mol (Janssen et al., 2000). Assuming
the PFD optimum value to be 58μmol m-2s-1, and taking 3.4 x1018 photons Joule-1
(monochromatic light), the required power per area will be 10.3 Wm-2. Yet a more conservative
PFD estimate of 200-300 μmol m-2s-1(Quigg & García-González) is considered, which would
correspond to 34.46-51.7 W m-2. The light concentrations can be measured using a
photosynthetic photon flux density device (PPFD) using a quantum sensor connected to a
quantum photometer.
Nutrients
To determine the type of media that will be used and the amount of nutrients required,
elements present in the algae strain used for our production line were determined and are shown
below in Table 5. In addition, 3.5% protein/dry weight (w/w) ( Dr. Baez ) is assumed for
protein yield per biomass. This results in a production of 17,150 kg of dry weight per year of
biomass. A summary of the amount of substrate required for the production of such amount of
biomass is shown in Table 6.
9
Table 5: Elements Found in Dunaliela tertiolecta
Element MW Proportion Total Proportion by Normalized
[g/mol] by moles weight weight [%] to 4 g/day
[moles]
C 12 222 2664 64.21 2.57
N 14 38 532 12.82 0.51
P 30.97 1 31 0.75 0.03
S 32 0.28 9 0.22 0.01
K 39 0.36 14 0.34 0.01
Mg 24.3 0.37 9 0.22 0.01
Ca 40 0.019 1 0.02 0.00
Sr 87.62 0.0081 1 0.02 0.00
Fe 55.84 11.3 631 15.21 0.61
Mn 54.9 1.9 104 2.51 0.10
Zn 65.4 1.49 97 2.35 0.09
Cu 63.5 0.67 43 1.03 0.04
Co 58.9 0.01 1 0.01 0.00
Cd 112.4 0.1 11 0.27 0.01
Mo 95.94 0.011 1 0.03 0.00
Total 4148.60806 100 4
Source: Phytoplankton Dynamics Laboratory. University of Texas A&M at Galveston.
Table 6: Substrates Required for 1715kg of Dry Biomass
Compound Substrate Proportion Kg MW Percent Substrate
by weight element element required
[%] needed in [Kg/year]
[Kg/year] substrate
C CO2 64.21 1101.20 44 0.27 4037.7
N KNO3 12.82 219.86 101.1 0.14 1587.7
P K2HPO4 0.75 12.86 136.1 0.23 56.5
S CuSO4·5H2O 0.22 3.77 249.7 0.13 29.3
Ca CaCl2·2H2O 0.02 0.34 147 0.27 1.3
Fe FeCl3, Na2EDTA 15.21 260.85 162.2 0.34 758.2
Mn MnCl2·4H2O 2.51 43.05 147.9 0.37 116.0
Co CoCl2·6H2O 0.01 0.17 237.9 0.25 0.7
Mo NH4)6Mo7O24·4H2O 0.03 0.51 1235.9 0.54 0.9
Trace Various 4.22 72.37 - - -
100 1715
Source: Aizawa and Miyachi, Brown et al., Mil'ko, Grant; Borowitzka and Borowitzka,
McLachlan, Massyuk.
Given that the amount of substrate used in CHO cells does not vary with the production
10
of different proteins (Dr. Stark), it is reasonable to assume that the majority of nutrients are used
for biomass production. From literature, the highest protein yield in plants ranges from 35-50%
protein/dry weight basis. Despite the relatively low substrate use for protein production, it is not
reasonable to ignore the substrates used given that 500 kg is required per year.
Table 7 shows the percent concentration of elements in Bevacizumab and N-
glycosylation. Oxygen is generated from carbon dioxide according to the stoichiometric formula
n CO2 + 2n H2O + ATP + NADPH → (CH2O)n + n O2 + n H2O; for this reason, carbon dioxide
is the limiting reagent. Also, nitrogen and hydrogen percent concentration is higher in the protein
than in N-glycosylation. This is important because mass balances calculations are based on the
protein percent concentration4.
Table 7: Percent Concentration of Elements in Belvacizumab and N-glycosylation
Element Proportion in Proportion in Substrate Kg element Kg substrate
protein N-glycosylation needed needed
(% weight) (% weight) [Kg/yr] [Kg/yr]
C 69.5% 40.8% CO2 347.5 1275
O 12.5% 51.9% CO2 0 0
N 9.25% 1.1% KNO3 46.2 334
H 8.75% 6.2% Various 43.8 Various
Source: US Patent & Trademark Office (for protein amin oacid sequence), Garret & Grisham
(Glycosylation), Julio Baez (for substrate used).
Upstream Design Specifications_________________________
Medium Preparation
The primary method of sterilizing the media is through heat. The gaseous CO2 supply
does not need to be heat sterilized but must be filter sterilized before it enters the reactor. Before
heat sterilizing the media, it should be microfiltered in order to remove any large particles.
Then, the media should be heat sterilized using steam to the point at which the probability of
contamination is very low. Heat sterilizing the media will also degrade some of the media
components. Thus the initial concentration of nutrients in the media must account for this
reduction to the final concentration. After the media is heat sterilized, it must be quarantined
before it is used in order to verify that it is not contaminated.
The liquid media is sterilized in a continuous sterilization system. Continuous
sterilization systems have advantages over batch sterilization systems such as lower energy costs,
4
If percent composition of protein is used in mass balance, nitrogen and hydrogen will be in excess. Thus allowing
protein production at is maximum rate.
11
more efficient use of energy, and the fact that the time to sterilize is lower because the
temperature is higher. The system (detailed specifications listed below in Table 8 ) consists of
two heat exchangers, a retention coil, and a holding tank. Two heat exchangers are used in order
to recover the energy used to heat up the media. The primary heat exchanger has a hot incoming
feed from the sterilized media leaving the retention coil. Energy is recovered when the hot
sterilized media heats up the incoming cold media feed that needs to be sterilized.
The media first enters the primary heat exchanger and is heated from 30˚C to 120˚C. The
cold media is heated up by the hot sterilized media, which is cooled from 130˚C to 50˚C. The
media then enters the secondary heat exchanger where it is further heated by steam to 130˚C.
The media next enters the 33 meters long retention coil at a velocity of 0.11 m/s. It takes the
130˚C media 5 minutes to travel through the retention coil, where the majority of any
contaminants are killed. The sterilized media then returns to the primary heat exchanger, heating
up the fresh, incoming media. The sterilized media is then held in a 10,000 L collection tank. A
detailed schematic of this process is illustrated below in Fig. 3.
Table 8: Media Preparation Specifications
Equipment Material Dimensions
Collection Tank 316L SS 10,000 L. Diameter of 1.62m and
height of 4.86m.
Regen. Heat Exchanger 316L SS 16 plates. Each plate is 40 cm x
30 cm x 11.17 mm.
Steam Heat Exchanger 316L SS 24 plates. Each plate is 40 cm x
30 cm x 11.17 mm.
Retention Coil 316L SS 3 inch inner diameter. 3.5 inch
outer diameter. 91.4 meters long.
Centrifugal Pump 316L SS Magnatex Pumps, ISO 9000
Temperature probe WQ101 Temperature Sensor
pH Probe PHR-212
Flow Meter Celcon PS601C
Liquid Level Meter LV800 Series Level Meter
Acid Supply Tank 316L SS 1,000 L. Diameter of 0.75m and
height of 2.25 m.
Base Supply Tank 316L SS 1,000 L. Diameter of 0.75m and
height of 2.25 m.
Diaphragm Valve High-Purity Two-Way
Diaphragm Valve EW-98613-56
Sampling Probe PFA 100 Probe
12
Temperature
Probe
Incoming CIP
Steam
Incoming Medium Feed Temperature
Probe
Temperature
Start-up Recycle Heat Exchanger Probe
Retention Coil
P-7
Pump Flow Meter
Regenerating Heat Exchanger Steam
Acid Supply Base Supply
Heat Exchanger
Condensate
PLC
PLC Collection
Level Meter
Tank
Sampling Probe
pH Meter
Pump
PLC
Dispense for inoculation medium To bioreactor
Figure 3: Medium Preparation P&ID
13
Inoculation Train
From Dunaliella tertiolecta’s annual biomass production and the saturation
concentration, the volume to process annually should be 343,000 L (sample calculations shown
in Appendix A). This means that 34300 L of inoculate stream is required. Detailed specifications
of the inoculation train are listed below in Table 9.
Growth rate for green algae at light saturation differs from strain to strain. For example,
Scenedesmus obliquus doubles 2.2 times per day, whereas Chrorella pyrenoidosa doubles 9.2
times per day. Since growth rate data is not available for Dunaliela tertiolecta, an estimate has to
be made. Given that most algal strains double at 3.5 hours5, a conservative estimate is 4 hours.
10% inoculants in the inoculation train would, therefore, take 16hours to grow (see calculation in
Appendix A).
Due to absolute sterility as well as cost effectiveness, disposable wave bags were chosen
as the reactor for the inoculation train. Since the largest wave bags produced are 2000L and the
working volume is 50%, the inoculation scheme will be as follows (Fig. 4):
10mL 16100mL 161L 1610L 16100L 161000L 16 Photobioreactor
hrs hrs
hrs
hrs
hrs
hrs
Figure 4: Inoculation Flow Sheet
The total time to inoculate the photobioreactor will be 96 hours, or 4 days. Due to the
working capacity of 50%, the actual size of the bags (and test-tube – 10mL) used are two times
that of the loaded volume listed above. The first inoculation will be from a standard Petrie dish
and into a 20mL disposable test tube filled with 9mL of fresh medium using a disposable
inoculation needle. The second inoculation will involve a 200mL wavebag. The bag is first filled
with sterilized air until rigid and then filled with 90mL of fresh medium. The bag is then placed
on the mechanical system and then rocked at 15 rocks per minute at an angle of 6 degrees. The
temperature is allowed to equilibrate and then the inoculant (10mL) is transferred via peristaltic
pump. The same procedure is repeated for each of the remaining steps in the train: 10% inoculant
from the previous step and 90% fresh medium (of half of the bag volume).
In addition to the light energy mentioned above, each step of the train requires adequate
CO2. The CO2 will be provided using simple plastic tubing (much like those used at Genentech)
of 1” diameter. For the larger bags, 200L and above, multiple tubes will be placed to ensure
adequate CO2 transfer. Conveniently, the bags in question come equipped with filters to ensure
absolute sterility of any incoming substance. The CO2 requirement, as further discussed in the
previous section, “Medium Preparation”, is approximately 60% of the cell biomass. The
dissolved CO2 concentrations will be measured with a sensor. The pH of the trains should also be
measured and will be monitored and maintained at pH 6.8 with a standard probe. A sampling
probe will also be utilized to extract samples and monitor the cell density of each step of the
train. The bags have custom made probes to ensure absolute sterility and each probe
measurement will take place every 6 hours during each step of the train
5
Julio Baez.
14
Table 9: Innoculation Specifications
Innoculation Step 1 2 3 4 5 6
Wave Bag 20mL 200mL 2L 20L 200L 2000L
Dimensions
Innoculant Volume 1mL 10mL .1L 1L 10L 100L
Media Requirements 9mL 90mL .9L 9L 90L 900L
Light Source 34.46-51.7 W m-2 “ “ “ “ “
CO2 Supply
Time Required 16hrs “ “ “ “ “
Total Time for 96hrs
Innoculation
CO2 Supply Lines 1-1” silicon tube “ “ “ “ “
CO2 Valves membrane “ “ “ “ “
CO2 Pressure At pump-Positive displacement – “ “ “ “ “
Element release at 1000psi
CO2 Flow Element Magnetic meter : 0-62.3 L/min “ “ “ “ “
CO2 Probe Electrode: 10mbar minima “ “ “ “ “
pH Probe Electrode: 7.0+ 5 “ “ “ “ “
Light Probe Quantum:2-300 µmol m-2s-1 “ “ “ “ “
Sampling Probe Optical “ “ “ “ “
Transfer Pump Diaphragm: 0-62.3 L/min “ “ “ “ “
Transfer Tubes 1.5” silicon tubing “ “ “ “ “
Photobioreactor
In order to successfully produce 600kg of MAb per year, as the calculation highlighted in
the appendix illustrates, 343000 L of saturated cultivation solution will be required. This will
require 10- 10,000L tubular photobioreactors that are detailed below in Table 10. Every batch
will require two days: 1 to fill the tanks, and one to run the reaction. To meet the MAb demand,
we will perform 36 batches (fed batch) per year.
The tubular reactors currently used are in the order of 1cm-2.5cm (García-González et al.,
Mullikin and Rorrer). This is due to the light attenuation, modeled by Beer-Lambert Law,
exponentially decaying:
Iz = Io exp(-kXz)
Where: Iz PFD at z depth [μmol m-2s-]
Io PFD at surface [μmol m-2s-]
K Culture attenuation coefficient [L/mg cm]
X Concentration of biomass [mg dry weight/L]
Z Depth [cm]
Source: Tyler and Smith.
15
Because of this exponential reduction of light, small tubes are necessary. Calculations in
the appendix show that a 1 ½ inch tube is needed if light attenuation effects and pressure drops
are considered. While a 1.22 inch is best to reduce light attenuation, the pressure drop for a 2
hour filling cycle is 146645 Pa/m as compared to 46751.7 Pa/m in 1.5 inch pipes (see appendix for
all calculations).
The small diameter in the tube will require each reactor to have 15,860 m in length,
which will be a challenge to pump. If the photobioreactor was to be filled in 2 hours, the pressure
drop from pipe will be 8 million Pa –high pressure. For this reason, it is better if the
photobioreactor was filled over a 24 hour period to reduce the total pressure drop down to 0.5
million Pa. Using these sizing and flow rates, the total pressure drop will be 88 psi, which is
below the 1000 psi that diaphragm valves can generate.
In addition to satisfying pumping requirements, photobioreactor tubes should also be
resistant to the temperatures and pressures of the photobioreactor. Out of the many possibilities,
silicon flexible tubing was chosen because it is transparent, smooth, and it can support 88 psi.
Even though standard manufacturing practices using mammalian or bacterial cells require
the employment of impellers for mixing and jacket characteristics for adequate heat removal, our
design does not require the use of impellers or jackets. Algae in general prefer stagnant
environments and vigorous agitation hinders growth. Because light absorption is the only
limiting reagent in our reaction, and agitation does not increase light absorption, agitation is not
required for optimal growth. Dunaliella tertiolecta is also an alga that is capable to grow and
produce proteins optimally at ranges from 25-35ºC. Generally, algae thrive at higher
temperatures and therefore heat removal is an unnecessary expenditure. This allows our
organization to utilize a room thermostat from a standard HVAC to control temperature.
As mentioned previously, CO2 transfer will be accomplished using simple silicon tubing
of 1” diameter. Unlike CHO and E. coli expression systems, gas transfer is not a limiting reagent
in our algae process. The CO2 will be filtered upon entry into the photobioreactor to ensure
absolute sterility. In addition, our exit gas stream will also be filtered to eliminate outside
contamination.
The medium, detailed previously, will be fed into the photobioreactor in a fed batch
manner and filtered upon entry. The detailed process diagram is illustrated below in Figure 5.
16
Table 10: Tubular Photobioreactor Specifications
Tube Volume 10,000
Loading Volume 8,000
Wall Thickness 0.75 inches
Tube Material Flexible vitron tubing
Feed Type Fed Batch
Light Requirements 34.46-51.7 W m-2
CO2 Flow Rate 200 g/h
Media Flow Rate 6.9L/min
Time in tube 1 day (2 day if you count filling time)
Number of Cycles 36 per year
Productivity 35% of biomass
Innoculation Transfer Pipes 1.5” silicon tubing
Innoculation Transfer Pump (to Tube Diaphragm: 0-62.3 L/min
Photobioreactor)
CO2 Supply Lines 1” silicon tubing
CO2 Filters Hydrophobic, membrane-pleated: 0.01µm
rating
CO2 Control Valves Membrane valves, autoclavable
CO2 Pressure Element At pump-Positive displacement – release at
1000psi
CO2 Flow Element Magnetic meter : 0-62.3 L/min
CO2 Probe Electrode: 10mbar minima
Medium Supply Pipes Stainless steel, electropolished, 2.5”, tilted
Medium Filters Hydrophobic, membrane pleated: 0.02 µm
Medium Control Valves Membrane valves, autoclavable
Medium Flow Element Magnetic meter : 0-62.3 L/min
Sampling Probe Optical
pH Probe Electrode: 7.0+ 5
Light Probe Quantum:2-300 µmol m-2s-1
Tube Pressure Gauge At pump-Positive displacement – release at
1000psi
17
Figure 5: Photobioreactor P&ID
Deoxygenation
Oxygen removal is crucial for the photosynthetic activity of plant cells. Not only does
oxygen inhibit photosynthesis, but also high irradiance levels inside the photobioreactor with
high dissolved oxygen leads to photo-oxidation. A photobioreactor lacking dissolved oxygen
(DO) has an increased photosynthetic activity of 14%. Photosynthetic activity has been known to
decrease 35% when the DO content is 1.38 mol/m3 at 20ºC (Rubio et al. 1999). For these reasons
deoxygenating is required for photobioreactors. Currently, there are three ways in which this can
be achieved: Catatalytic oxygen removal systems (CORS), Activated Carbon Deoxygenation
(DEOX), and Gas Transfer Membranes (GTM).
CORS removes oxygen by treating media with hydrogen, and passing it trough a
palladium catalyst to produce water that is less than 1ppb in dissolved oxygen concentration.
DEOX, on the other hand, feeds a chemical oxygen scavenger known as hydrazine; the stream is
then directed to a ion-exchange resin vessel to removed trace impurities leached by the activated
18
carbon. This process yields a purity of 1ppb of dissolved oxygen. Finally GTM is the passing of
the stream through an impermeable membrane, but with gas transfer. This method also yields to
less than 1ppb in dissolved oxygen. Alternatively, hybrid methods have been suggested to reduce
oxygen to 7,(Kemp, 2004) and the contaminants are acidic using a cation
exchange column is appropriate. Here the desire proteins will bind to the column while the
contaminants and unwanted proteins will not be retain and leave as waste. Before every run the
column is sterilized with NaOH and regenerated with NaCl. Afterwards media from the gel
filtraion column is loaded on the column and allowed to bind. The proteins are then eluded with
a solution of .03M Na-Phosphate and .1M NaCl at pH 6.5. It is estimated that 95% of the
proteins will be recovered. In addition, samples of the out coming effluent will be tested every 6
hours with a sampling probe. The cation exchange column utilized is fully described in Table ,
and is illustrated in Figure 8. Additional equipment involved in this recovery step are described
in Table .
Anion Exchange Chromatography
Two anion exchange chromatography columns operated identically in parallel will
immediately follow cation exchange separation and are used for capturing the remaining trace
contaminants such as host cell protein and DNA residues from the protein media. Its main
function is to remove the negatively charged contaminants. It also separates all protein A leakage
products and removes small amount of oligomers from the antibody. Thus a large amount of host
cell proteins (hcp) will be bound while the neutral antibodies are allowed to flow through. The
typical contaminant clearance value for hcp removal is 10 for every ng/ml while for DNA it is
100 for every pg/ml (Kemp, 2004). At this step, the column is operated in the flow through mode
(Kemp pg. 93). Given that most antibodies are basic proteins, pI >7, (Kemp, 2004) and the
contaminants are acidic, using an anion exchange column is appropriate. In order to maintain
high volumetric flow rates and to prevent bottlenecking, the column is required to have a
minimum depth of approximately 10cm (Kemp, 2004, pg. 93). Determined by Necina et al.
(1993) the main ligand for the anion exchange column is the amino group. The anion exchange
column utilized is fully described in Table 7, and is illustrated in Figure 8. Additional equipment
involved in this recovery step are described in Table 8.
The anion exchange column is operated by first loading the incoming protein solution
from the protein A column onto the column. The contaminants will bind to the resin while the
protein will flow through. Before every run the column is regenerated with NaCl followed by a
NaOH wash. The mAb is then eluted using Na-Phosphate and NaCl solutions, pH 8.5. The
recovery rate is estimated to be 95%. A sampling probe will also be utilized to extract samples
every 6 hours during the anion exchange recovery step.
47
Table 7: Ion Exchange Operating Modules
CM- Hyper D Cation Exchange Mono Q Anion Exchange
Chromatography Chromatography (2)
Mass (kg) in per cycle 18.5 kg 8.77 kg
Mass (kg) out per cycle 17.5kg 8.3 kg
Recovery % .95 .95
Bed volume 2769.95 L 4154.92 L
Bed height 0.95m 1.09 m
Bed diameter 1.92 m 2.19m
Concentration of 6.67mg/mL 2.11mg/mL
protein in
Concentration of 2.11mg/mL 0.33 mg/mL
protein out
Conductivity Ranges 4-10 4-6
(AU)
Typical Buffer 8309 L of 0.03M Na-Phosphate, 12.5L of 50mM Na-Phosphate,
Media** 8309 L of 0.03M NaCl, pH 6.5 12.5L of 1M NaCl pH 6.3*
Buffer time 280 min 280 min
Elution Buffer 8309 L of 0.03 M Na-Phosphate 12.5L of 0.03 M Na-Phosphate
containing 8309 L of 1M NaCl, pH containing 12.5L of 1M NaCl,
6.5 pH 8.5
Elution flow rate 9.89 L/min 14.84L/min
Elution Time 280 min 280 min
Volumetric flow rate 0.0099 m^3 /min 0.014m^3/min
Resin type Fractogel EMD SE Hicap(M) beads Q-Sepharose, Preswollen in
20% ethanol
Resin pore size 800 A ~4,000,000 Da exclusion limit
Resin particle size 90 um 45 μm (wet)
Resin binding capacity 36mg of mAb mg/mL 30mg of mAb/mL
Resin storage NA 2-8°C
temperature
Particle size NA Not less than 95% than within
distribution 45 and 165 microns
Regeneration buffer 1-2 CV( or 2769L) of 1-2 M of 1-2 CV ( or 8309 L) of 1-
NaCL 2MNaCl followed by 1 CV ( or
8309 L ) of .1M NaOH in
8309L .5 MNaCL
Sanitization 2729L of .1-.5 M NaOH 8309L of.1-.5 M NaOH
Sanitization Flow rate 9.89L/min 14.84 L/min
Resin usage Needs to replace every 56 cycles Needs to be replaced every 20
cycles
Regenerate time 280 min 280 min
48
Resin Regeneration 1832.07mL/min 1044.23 mL/min
flow rate
Resin Volume 512978.4 mL 292397.7 mL
Linear gradient flow 9.892659 L/min 14.83899 L/min
rate
*Column outer 2.49m 2.85 m
diameter
*Column inner 1.91m 2.19m
diameter
*Column height 1.24 m 1.42 m
*Column Volume 3600.93 L 5401.40 L
Absorbance 280 nm 280nm
Number of cycles 36 cycles per year 36 cycles per year
Cycle time 14 hr 14 hr
pH condition range 1-13 2-12
Column Pressure drop 8 bar 1000 psi
limit
Actual estimated 5.47 bar 455.32 psi
Pressure drop of
column
Source: Sigma-Aldrich website, Subramanian,
http://www.merck.de/servlet/PB/show/1225330/w214154_Fracotgel_WF.pdf
*
taken to be 30% bigger than bed dimensions
Table 8. Additional Equipment for the Cation and Anion Exchange Column and their
Dimensions
Material Dimension Purpose
Pump 316L The pump removes solvents from the
elution, buffer, or regenerating tanks and
delivers it to the anion exchange column at
the appropriate flow rate
Na- 316L Diameter: 1.289m; Solvent used for the buffer and elution
Phosphate Height: 3.86m media for the cation and anion exchange
tank -5000L columns
NaOH tank- 316L Diameter: 1.289m; Solvent used for regenerating the anion
5000L Height: 3.86m column. Also use as a media for sanitizing
the cation exchange columns
NaCl tank- 316L Diameter: 1.289m; Solvent used for regenerating both the
5000L Height: 3.86m cation and anion column. Also used in the
buffer and elution media for the anion and
cation exchange columns
49
Elution 316L Diameter: 1.62m, It is also used for eluding the bound proteins
buffer tank Height: 3.25m from the cation exchange columns.
for cation-
10,000L
Elution 316L Diameter: 1.62m, Used for eluding the bound contaminants
buffer tank Height: 3.25m from the anion column. The contaminants
for anion- will leave the column and go to a waste
10,000L tank.
Buffer tank 316L Diameter: 1.62m, Use to maintain the appropriate pH
for anion Height: 3.25m
exchange
columns-
10,000L
Buffer tank 316L Diameter: 1.62m, Use to maintain the appropriate pH
for cation Height: 3.25m
exchange
columns-
10,000L
Regeneration 316L Diameter: 1.62m, Use to regenerate the column so that the
tank- Height: 3.25m contaminant from the incoming influent
10,000L stream will bind to the column
Air vent 316L To remove air from the highest point of a
coil or piping assembly
Valves 316L Use to monitor and control amount of fluid
leaving and entering
Sample Use to check the media in and leaving the
Probes anion and cation exchange column
UV and Monitor the absorbance (at 280nm) and
conductivity conductance of the product leaving the
monitor effluent stream.
Pressure Use to monitor the pressure of the influent
gauge
Air Vent For air bubbles or bypass
Piping Inner Diameter: 1.5 Transport media from one location to
1.5 inch inch another
Outer Diameter: 1.37
inch
50
Buffer Media .05M Na-Phosphate
(.03M Na-Phosphate,
Air .1M NaOH
.03 NaCL) .1M
Media from Vent
P-37
P-41 E-3
V-15 E-1
gel filtration P-42
P-29 P-35 NaCL
E-14 V-18
V-11 Elution Buffer P-20
P-43(.03M Na-Phosphate, P-30
E-6 Buffer Media
P-32
Pump V-23 1M NaCL,pH 6.5) P-46 PLC Elution Buffer ( 50mM of Na-Phosphate,
P-22
Pressure S-4 S-3
E-10
(.03M Na-Phosphate, .1MNaCL)
Gauge P-45
P-19
I-1
1MNaCL, pH 8.5) pH probe
P V-6 S-1 PLC
P-12
I-8 pH probe S-6
E-15 S-2
V-14
S-5 V-24
pH probe E-13
P-5
I-2
I-16 V-3
Influent PLC
P-23 E-11
V-13 V-2
E-9 P-3 pH probe
pH probe P-8
P-18
pH probe V-1 V-10
P-28
Regenerant NaCL
P-47 P-55
pH probe P-1
P-34 P-31
Computer LabView Software
P-43 P-2
Anion Exchange P-15 P-11
P-40 pH probe P-17
Column
I-6
V-25 V-7
Cation exchange P-44
Column
Anion Exchange
Column
E-7
E-5
P-38
Sampling Probe UV Monitor
Target
P-36
Protein
P-57 I-4
V-22 Conductivity Monitor
P-27
P-54
V-5
V-20
E-12 E-2
waste T V-9
E-16 V-16
Target I-3 P-33
Conductivity Cell V-4
Protein
Waste P-27 Target P-6 P-21 P-7 P-14
P-49
I-10
Protein Conductivity Cell
P-24
Sampling Probe V-19 T
V-17 I-13
Waste UVI-7Cell
V-21 I-5
Conductivity Monitor Sampling Probe Target Protein
P-27 P-10
T
P-50 T
I-15
UV Cell
I-11
P-52
Computer LabView Software
I-12
UV Monitor
I-14
P-51
E-4
Figure 8. P&ID of Cation and Anion Exchange Columns
51
Downstream CIP/SIP Systems__________________________
For the downstream process, the same multi-use cleaning system that was used to clean
the upstream units will be utilized. The figure of the multi-use cleaning system is available in
part 2 of the project. Again cleaning reagents are dispersed through the circuits for different
intervals of time at varying temperatures. The specifications for the CIP/SIP systems are
described in Table 29. A typical cleaning schedule for our system is illustrated in Table 30.
The water for injection is used to supply the water necessary in the CIP and SIP systems.
Combined, these processes required 54,520 L of WFI. Before using the water, it first needs to be
sterilized. The water is carbon filtered and then undergoes reverse osmosis. The reverse osmosis
membrane captures ions, microbes, endotoxins, and other small contaminants (Lyderson, pg.
548).
52
Table 29: CIP/SIP Specifications for Downstream Processing
Material Description Dimension
Scavenge Return Pump The return pump pumps the fluids out of the
clean process equipment, and either returns
it to the CIP unit (recycling) or pumps it
directly to the drainage system.
Double Pipe Heat exchanger The built-in heating elements heat the fluid
to a specified temperature.
Piping 316L SS Use to connect equipment to one another 1.5 inch
Sodium hydrochlorite 10000L 316L SS Detergent for CIP. Use to id media and Diameter:
tank algae residues 1.62m, Height:
4.86m
Water tank 10000L 316L SS For multiple cleaning and solution Diameter:
preparation purposes. 1.62m, Height:
4.86m
Sodium hydroxide 10000L 316L SS Chemical for cell disruptor. Diameter:
1.62m, Height:
4.86m
Sodium hydrochloride 10000L 316L SS Chemical for cell disruptor. Diameter:
1.62m, Height:
4.86m
Acetone 100L 316L SS Chemical for centrifuge. Diameter: 0.34,
Height: 1.05
H3PO4 10000L 316L SS Protein A Column Diameter:
1.62m, Height:
4.86m
Guanidine hydrochloride10000L 316L SS Protein A Column Diameter:
1.62m, Height:
4.86m
C2H6O10000L 316L SS Ion Exchange Column Diameter:
1.62m, Height:
4.86m
0.05 M CH2O2 in 40% C2H6O and 316L SS Ion Exchange Column Diameter:
60% H2010000L 1.62m, Height:
4.86m
53
0.2 NaH2PO4 + 0.3 M 316L SS Ion Exchange Column Diameter:
CH3COONa10000L 1.62m, Height:
4.86m
Water 10000L tank 316L SS Water comes from WFI . Use for cleaning with Diameter:
and for buffer media 1.62m, Height:
4.86m
Sodium hydroxide 10,000L tank 316L SS Second solvent used for cleaning purposes with Diameter:
1.62m, Height:
4.86m
Control valves Use to monitor and control amount of fluid
leaving and entering
Cold Water Supply tank (3000L) 316L SS Supply the system with cool to cold water with Diameter:
1.08m, Height:
3.25m
Steam Supply tank (3000L) 316L SS Supply heat exchanger with steam with Diameter:
1.08m, Height:
3.25m
Condensate Return box (200L) 316L SS Return the collected condensate amount to with Diameter:
heat exchanger .4m, Height:
1.3m
Pressure relief valve Use to protect system’s pressure vessels
Duplex Filter Filter steam from heat exchanger
PLC (include software and
printer) Built-in PLC unit with an operating panel.
Placed in the control room.
The PLC is programmed individually for
each cleaning task. Each task is fully
documented by a laser printer.
Sample probe Use to check the detergent and buffer media
Positive Displacement Pump Use for injecting chemicals through an the
pipeline system at precise rates.
CIP pump The frequency controlled supply pump
builds up the pressure to the spray nozzle
and washing spear, that is fitted in the
process equipment to be cleaned.
Supply Vessels 316L SS Holds the cleaning reagent and or water for
the unit to be clean
Compressed Air Filtered compressed air purge system
ejects residual water from the lines.
WFI Carbon Filter Water scrub unit, Model: WSU1000P 60gpm capacity
WFI Reverse Osmosis Membrane Osmonics, Model: M-CB4040AD 4gpm capacity
1.0µm rating
Sources: ACCompacting, MorkUSA, LHS Company
(Note: Assume 3:1 Height to Diameter ratio for Tank dimensions)
54
Table 30: Typical Cleaning Schedule
(Note: An air blow step occurs in between each of the cleaning step)
Equipment Operation Flow Amount Time Temperature Pressure
rate
Cell Water Rinse 38 L/min 760 L 15-20 Ambient 2-8 bar
disruptor min (water
feeding)
Sodium 38 L/min 15.048 of 15-20 Ambient to
hydroxide wash NaOH + min 75 ºC
(0.1, ) 135.4 gal of
water
Water Rinse 38 L/min 570L or 5-15 Ambient 2-8 bar
150.48 gal min (water
feeding
pressure)
Sodium 38 L/min 15.048 gal of 15-20 Ambient to
hydrochloride Sodium min 75 ºC
wash Hydrochlorite
+ 135.43 gal
of water
Water Rinse 38 L/min 380 L or 5-10 Ambient 2-8 bar
100.32 gal min (water
feeding
pressure)
Purged/Drainage 38 L/min ----- 10-15 ------
min
Steam sterilized 130 lb/hr 43 lb 5-20 Ambient to 3-5 bar
min 100C
Centrifuge Water rinse 10 L/min 10L 1 min 100 ºC 2-8 bar
(first step) (water
feeding)
Water flood 10 L/min 10L 1 min 100 ºC 2-8 bar
(water
feeding)
Water rinse 10 L/min 6L 0.6 min Ambient 2-8 bar
(water
feeding)
Acetone flood 10 L/min 5L 0.5 min Ambient 2-8 bar
(water
feeding)
Water rinse 10 L/min 40L 4 min 100 ºC 2-8 bar
(water
feeding)
55
Centrifuge Water rinse 10 L/min 40L 4 min 100 ºC 2-8 bar
(2nd step) (water
feeding)
Water flood 10 L/min 40L 4 min 100 ºC 2-8 bar
(water
feeding)
Water rinse 10 L/min 30L 3 min Ambient 2-8 bar
(water
feeding)
Acetone flood 10 L/min 25L 2.5 min Ambient 2-8 bar
(water
feeding)
Water rinse 10 L/min 40L 4 min 100 ºC 2-8 bar
(water
feeding)
All 1000L Water Rinse 38 L/min 760 L 15-20 Ambient 2-8 bar
tanks-ie min (water
Holding feeding)
Tank,
Elution tank,
Regeneration
tank, Buffer
tank, Citrate
tank, Tris-
HCL tank
Sodium 38 L/min 15.048 of 15-20 Ambient to
hydroxide wash NaOH + min 75 ºC
(0.1, ) 135.4 gal of
water
Water Rinse 38 L/min 570L or 5-15 Ambient 2-8 bar
150.48 gal min (water
feeding
pressure)
Sodium 38 L/min 15.048 gal of 15-20 Ambient to
hydrochloride Sodium min 75 ºC
wash Hydrochlorite
+ 135.43 gal
of water
56
Water Rinse 38 L/min 380 L or 5-10 Ambient 2-8 bar
100.32 gal min (water
feeding
pressure)
Purged/Drainage 38 L/min ----- 10-15 ------
min
All 5000L Water Rinse 38 L/min 760 L 15-20 Ambient 2-8 bar
tank- NaCl min (water
tank, NaOH feeding)
tank, Na-P
tank
Sodium 38 L/min 15.048 of 15-20 Ambient to
hydroxide wash NaOH + min 75 ºC
(0.1, ) 135.4 gal of
water
Water Rinse 38 L/min 570L or 5-15 Ambient 2-8 bar
150.48 gal min (water
feeding
pressure)
Sodium 38 L/min 15.048 gal of 15-20 Ambient to
hydrochloride Sodium min 75 ºC
wash Hydrochlorite
+ 135.43 of
water
Water Rinse 38 L/min 380 L or 5-10 Ambient 2-8 bar
100.32 gal min (water
feeding
pressure)
Purged/Drainage 38 L/min ----- 10-15 ------
min
2000L Water Rinse 38 L/min 760 L 15-20 Ambient 2-8 bar
Affinity Prep min (water
Protein A feeding)
container
57
Sodium 38 L/min 15.048 of 15-20 Ambient to
hydroxide wash NaOH + min 75 ºC
(0.1, ) 135.4 gal of
water
Water Rinse 38 L/min 570L or 5-15 Ambient 2-8 bar
150.48 gal min (water
feeding
pressure)
Sodium 38 L/min 15.048 gal of 15-20 Ambient to
hydrochloride Sodium min 75 ºC
wash Hydrochlorite
+ 135.43 of
water
Water Rinse 38 L/min 380 L or 5-10 Ambient 2-8 bar
100.32 gal min (water
feeding
pressure)
Purged/Drainage 38 L/min ----- 10-15 ------
min
Protein A Acid wash- 64 L/min 7700L 15-20 Ambient 2-8 bar
Column H3PO4, pH 1.5 min (water
feeding)
Water flood-4-6 22 L/min 7700L 15-20 Ambient 2-8 bar
M Guanidine min (water
hydrochloride feeding)
Storage - C2H6O 22 L/min 3080L 15-20 Ambient 2-8 bar
min (water
feeding)
Ion Cleaning-0.05 22 L/min 21,000L 15-20 Ambient 2-8 bar
Exchange M CH2O2 in min (water
Column 40% C2H6O and feeding)
60% H20
Water wash- 22 L/min 21,000L 5-10 Ambient 2-8 bar
H20 min (water
feeding)
Conditioning- 22 L/min 21,000L 15-20 Ambient 2-8 bar
0.2 NaH2PO4 + min (water
0.3 M feeding)
CH3COONa
58
Salt removal- 22 L/min 21,000L 5-10 Ambient 2-8 bar
H20 min (water
feeding)
Non-specific 22 L/min 1400L 15-20 Ambient 2-8 bar
saturation- min (water
Gelatin feeding)
Cost Analysis of Downstream Process Design
The five sections of the downstream processing include: centrifugation, cell disruption,
another centrifugation step, two protein A affinity columns operated in parallel, a gel filtration
column, a cation exchange column, and two anion exchange chromatography columns operated
in parallel. The cost associated with both installation, maintenance, and operation of each section
have been determined and listed in the following tables. The final yearly operational cost of the
downstream side of the process sums up to be $ 14.1 Million, while the one time start up cost is
estimated to be $10.3 Million. This cost does not take into account employees compensation or
building costs. This cost also does not yet include the cost of the gel filtration column.
Cell Harvest : Centrifugation
Table 31: Fixed Capitol Costs for Cell Harvest Centrifugation
Item Quantity Unit Total Cost
Cost
BRPX617SFV, 20L, 8000rpm, 60cm Diameter 2 $500,500 $1,001,031
Diaphragm Valve 2 $449 $898
Butterfly Valves (CIP) 6 $300 $1,800
Centrifugal Pump 8 $3,000 $24,000
Process Piping (20% of major equipment cost, Atkinson __ __ $200,206
1991)
Installation (25% of major equipment, Atkinson, 1991) __ __ $250,258
Validation (20% of major equipment, Atkinson, 1991) __ __ $200,206
Shipping (40% of major equipment cost, Atkinson, 1991) __ __ $400,000
Total Capitol Investment for Cell Harvest Centrifugation $2.1 million
59
Table 32:Yearly Operational Cost for Cell Harvest Centrifugation
Item Quantity per Year Unit Cost Total
Cost
Maintenance 2 3% of Capitol Investment $45,000
(Atkinson, 1991)
Cooling Water 100,000L, 36 times per year: $0.50 per Liter $1,800
3,600,000L
Biological Waste 66% of 100,000L, 36 times per $0.20/100L $4783
Treatment year: 2.4million Litres
Steam Sterilization 500tonne $15/tonne $7500
Energy Consumption 35,840 kWh $0.09/kWh $3226
Total Yearly $62,310
Operational Costs
Cell disruption
Table 33: Capitol Costs for Bead Mill Operation
Item Unit Quantity Total Cost
Cost
316L s.s. Dyno-Mill MULTI-PILOT $229,000 5 $1.15
305 Liter , 5000rpm, 70L/min million
316L Buffer Solution Holding Tank, 381 Liter $900 1 $900
316L Mixing Tank, 1906 Liter $4000 1 $4000
Installation Costs (25% major equipment cost, __ __ $286,000
Atkinson, 1991))
Process Piping (20% of major equipment cost, __ __ $230,000
Atkinson, 1991)
Splitters $100 7 $700
Diaphragm Valve $449 4 $1,800
Centrifugal Pump $3,000 7 $31,000
Shipping (40% of major equipment, Atkinson, 1991) __ __ $460,000
Validation (up to 20% of major equipment, __ __ $230,000
Atkinson, 1991)
Total Fixed Capitol Costs for Bead Mill $2.4 million
60
Table 34: Yearly Operational Costs for Bead Mill Operation
Item Unit Cost Quantity per Year Total
Cost
Maintenance 3% of Capitol 2 $72,000
Investment, Atkinson,
1991
Cooling Water $0.50 per Liter $34,000L, 36 times per $612,000
year: 1.2 million liters
Buffer Solution (plus $3.00 per liter $8,400 L. 36 times per $907,200
reducing agents and year: 302,400 liters
inhibitors)
Energy Consumption $0.09/kWh 40,000 kWh $3,600
Total Yearly Operational $1.6 million
Costs for Bead Mill
If the lack of cell wall on our algae strain results in easy cell rupture, an alternative , more
cost effective Microfluidizer system can be used instead. Homogenizers were priced around
$278,000 in 1986. This equipment, adjusted for 3% annual inflation, will cost approximately
$500,000 in 2006. Because the microfluidizer system includes the control elements,
programmable logic controllers, and heat exchanger, the cost analysis for cell disruption will not
require the addition of these elements. The only additional cost associated with this system is the
cost of electricity and the cost for running the heat exchanger, which was estimated to annually
cost $19000 and $4000, respectively. For calculating total capitol investment as well as selling
price required, we will use the more costly bead mill system to ensure proper budgeting.
Cell Debris Removal: Centrifugation
Table 35: One Time Costs for Cell Debris Removal Centrifugation
Item Quantity Unit Cost Total Cost
BRPX617SFV, 57L, 9,000rpm, 100cm Diameter 2 $1,430,044 $2,860,088
Diaphragm Valve 2 $449 $898
Butterfly Valve (CIP) 6 $300 $1,800
Centrifugal Pump 8 $3,000 $24,000
Process Piping (20% of major equipment, Atkinson, 1991) __ __ $572,018
Installation (25% of major equipment , Atkinson, 1991) __ __ $715,022
Shipping (40% of major equipment, Atkinson, 1991) __ __ $1.2
million
Valladation (20% of major equipment, Atkinson, 1991) __ __ $580,000
Total Fixed Capitol Costs for Cell Debris Removal $6.0 million
Centrifugation
61
Table 36 :Yearly Operational Cost for Cell Debris Removal Centrifugation
Item Quantity per Year Unit Cost Total
Cost
Maintenance 2 3% of Capitol Investment $126,000
(Atkinson, 1991)
Cooling Water 100,000L, 36 times per year: $0.50 per Liter $1,800
3,600,000L
Biological Waste 5% of 100,000L, 36 times per $0.20/100L $360
Treatment year: 180,000 Litres
Steam Sterilization 500tonne $15/tonne $7500
Energy Consumption 35,840 kWh $0.09/kWh $3226
Total Yearly $138,886
Operational Costs
Affinity Column
Table 37: Cost of Equipment and Operation of Affinity Column.
Equipment Units Needed Cost per unit
Affinity Column Shell 2 Columns $8,032.55*
Holding Tank 1 $20,855.36*
Resin 1540 L per year $6,000 per liter of affi-prep protein A**
Tris-HCl 180,000 L per $1.47 per liter***
year
Sodium Citrate 180,000 L per $0.08 per liter****
year
Adsorption spectrophotometer 2 $5,000
equipped with LabView to
monitor adsorption
Centrifugal Pump 1 $707.28
Motor for Pump 1 $1850
Pressure Meter 2 $500 per meter
Temperature probe 2 $260 per probe
pH Probe 2 $550 per probe
Flow Meter 2 $125 per meter
Tris-HCl Buffer Container 1 $20,855.36*
Sodium Citrate Elutant 1 $20,855.36*
Container
Affi-prep Protein A Container 1 $9,177.89*
Diaphragm Valve 12 $449.30 per valve
Diaphragm Membranes 48 per year $100 per membrane
Sampling Probe 2 $1095.00 per probe
62
Piping (20% of major $32,000
equipment)
Maintenance (3% of capital $5,000
costs)
Installation (25% of major $40,000
equipment)
Total Fixed Capital Costs $162,320.00
Total Annual Operational $9.5 Million
Cost
*
(Seider, 2004)
**
(Harrison, pg. 367)
***
(Bio-Rad website)
****
(Sigma-Aldrich website)
Cation Exchange Columns
Table 14. Cost of Cation Exchange Column
Equipment Type Quantity Cost (for each) Total cost
Resin (replace Fractogel EMD 512978.4 mL $457053.5
every 56 cycles) SE Hicap(M)
beads $445.49 for
500 mL
Column Cation 1 $374,000 $374,000
Membranes for 28 $100 each $2800
valves
Elution Buffer 10,000L; 316L 1 $20,000 $20,000
tank
Buffer tank 10,000L; 316L 1 $20,000 $20,000
Computer 1 $4095.0 $8190
Program
Software (ie Lab
view)
UV and PC 1 $2105 $4210
Conductivity
monitor
Sample probes PFA 100 1 $1095 $6579
Pumps Diaphragm 1 $5000 each $5000
Valves High Purity Two 7 $449.30 $3143
Way diaphragm
Valve- EV-98613-
56
63
Pressure gauge High-accuracy 1 $800.0 $800.0
digital test gauge,
0 to 30" Hg range
EW-68330-00
Electricity 856 kW* .8 cent per kW
$68.48
per cycle
316L Piping (1.5 Estimated to be $ 18000
in) 20% of major
equipment
NaOH cost 5538 L $63.78 per 18L $ 2211.04 per
cycle
NaCl cost 19387 L $81.70 per 300mL $45316.67 per
cycle
Na-Phosphate 16618 L $51.0 for 2.5kg $ 16972.8 per
cycle
Motor for Pump 32" Commerical $1849.95 $1859.95
Mower with 10.5
B&S Engine 1
Pressure Release At pump 6- come with $0 $0
pump
Conductivity Cell 1 $5000 $5000
UV Cell 1 $5000 $5000
Direct Soft PLC 1 $395.0 each $790
software
PLC controller 2 $1095.0 $2190
STV Sanders 1 $1573.0 $3146
Pressure relief
valve
Maintenance Estimated to be $2700 per year
3% of major
equipment
Installation Estimated to be $22500
25% of major
equipment
Shipping Estimated to 40% $36000
of major
equipment
Validation Estimated to be $18000
20% of major
equipment
Total Fixed .54 Million
Capital Costs
64
Total 7.5 Million
operational cost
Anion Exchange Columns
Table 14. Cost of Anion Exchange Columns
Equipment Type Quantity Cost (for each) Total cost
Resin (replace Q-Sepharose, 555666.67 mL $852392.66
every 20 cycles) Preswollen in per column
20% ethanol $383.50 for
500mL
Column Anion 2 $254,000 $508,000
Membranes for 72 $100 each $7200
valves
Regeneration 10,000L; 316L 1 $20,000 $20,000
Buffer tank
Elution Buffer 10,000L; 316L 1 $20,000 $20,000
tank
Buffer tank 10,000L; 316L 1 $20,000 $20,000
Computer 2 $4095.0 $8190
Program
Software (ie Lab
view)
UV and PC 2 $2105 $4210
Conductivity
monitor
Sample probes PFA 100 6 $1095 $6579
Pumps Diaphragm 5 $5000 each $25,000
Valves High Purity Two 18 $449.30 $8087.4
Way diaphragm
Valve- EV-
98613-56
Na-Phosphate 5,000L; 316L 1 $10,000 $10,000
tank
NaCl tank 5,000L; 316L 1 $10,000 $10,000
NaOH tank 5,000L; 316L 1 $10,000 $10,000
Air Vent 1 $1200 $1200
Pressure gauge High-accuracy 1 $800.0 $800.0
digital test gauge,
0 to 30" Hg range
EW-68330-00
65
Electricity 856 kW* .8 cent per kW
$68.48
per cycle
316L Piping (1.5 300ft Estimated to be $ 124600
in) 20% of major
equipment
NaOH cost 16618 L per $63.78 per 18L $ 58883 per
cycle cycle
NaCl cost 16643 L per $81.70 per $4532443 per
cycle 300mL cycle
Na-Phosphate 832L per cycle $51.0 for 2.5 L $ 16972.8 per
cycle
Motor for Pump 32" Commerical $1849.95 $9249.75
Mower with 10.5
B&S Engine 5
Pressure Release At pump 6- come with $0 $0
pump
Conductivity Cell 1 $5000 $5000
UV Cell 1 $5000 $5000
Direct Soft PLC 2 $395.0 each $790
software
PLC controller 2 $1095.0 $2190
STV Sanders 2 $1573.0 $3146
Pressure relief
valve
Maintenance Estimated to be $18690 per
3% of major year
equipment
Installation Estimated to be $155750
25% of major
equipment
Shipping Estimated to be $249200
40% of major
equipment
Validation Estimated to be $ 124600
20% of major
equipment
Final Fixed $2.9 Million
Capital Costs
Total $5.3 Million
Operational
Cost
Source: Sigma Aldrich website, Harrison
*
The electricity cost is estimated as 20% of the process equipment cost
66
**
The maintenance and installation cost is estimated as 3% of the process equipment cost
CIP/SIP for Downstream Process
Table 39: Cost of CIP/SIP system for downstream process
CIP materials Quantity Cost per unit Total Cost
316L piping 1500ft $14.07 ft $21,105
Electricity 515kW $ 203/kW $104545
NaOH 58.75 L per cycle $5.16 per kg $30315
H20 54,519.37 L per cycle $.50 gallon $2710
H3PO4, pH 1.5 10000 L $466 per gallon $1211600
CH3COCH3 30L per cycle $87.4 per gallon $2644
CH6ClN3 3840 kg per cycle $354.3 per kg $1360512
C2H6O $16000 L per cycle $40.9 per gallon $170144
CH2O2 69kg per cycle $151 per gallon $10419
NaH2PO4 983 kg per cycle $172 per kg $169144.8
CH3COONa 738 kg per cycle $125 per kg $92250
Gelatin 3kg per cycle $138 per kg $615
Total Fixed Capital $21,105
Costs
Total Operational Cost $ 3.1 Million
67
Economic Feasibility of MAb Production
in
Dunaliela tertiolecta
Fixed and Operational Costs of Entire Process____________
The total process for producing monoclonal antibodies includes both an upstream and a
downstream. The upstream includes four steps: medium preparation, inoculation train,
photobioreactor, and CIP/SIP. The downstream includes seven steps: centrifugation, cell
disruption, another centrifugation step, two protein A affinity columns operated in parallel, a gel
filtration chromatograph column, a cation exchange column, and two anion exchange
chromatography columns operated in parallel. The total cost of the process is summarized in
Tables 40 and 41 below. The total one time cost for the entire process is $13.8 million and the
total annual operational cost is $18.9 million.
Table 40: Total One Time Costs for Entire Process
One Time Costs Cost ($)
Upstream Process $3.5 Million
Downstream Process $10.3 Million
Total Fixed Capital Costs for Entire Process $13.8 Million
Table 41: Total Annual Operational Costs for Entire Process
Annual Operational Costs Cost ($)
Upstream Process $6.1 Million
Downstream Process $14.1 Million
Total Annual Operational Costs for Entire $20.2 Million
Process
68
Selling Price to Receive 20% Return on Investment________
The following tables show potential sell prices to receive the minimum 20% return on
investment (ROI) for the first year of production (Table 42) as well as years after (Table 43). We
have included 50% contingency and 10% interest rates on money borrowed to account for the
high risk business venture we are pursuing. We have also accounted for employee salaries,
pensions, and healthcare. Considering that current contract manufacturing procedures result in
MAb selling prices between $2000 and $5000, our proposed algal process is a very competitive
business venture.
Table 42: Sell price per gram of MAb required to receive 20% ROI for first year of
production
Total Annual Operating Costs (includes maintenance and utilities)
Total Capitol Investment (includes shipping and validation)
Annual Labor and Payroll (300 employess with avg. salary of $100,000 per year to $30
include pensions and healthcare) million
Start-up Expenses (10% of capitol investment, Atkinnson, 1991)
Contingency (50% of capitol investment for unventured process, Atkinson, 1991)
Other Plant Overhead (telephone, sewer, freight, travel, legal, rentals) $400,000
Taxes and Insurance (6% of capitol investment, Atkinson 1991)
Annual Expense due to Interest Rate for Capitol Investment (10%)
Sales and Marketing
Total costs for first year of operation
Price per gram MAb to receive minimum 20% ROI
Table 43: Sell price per gram of MAb required to receive 20% ROI every other year of
production
Total Annual Operating Costs (includes utilities) $20.2
million
Annual Labor and Payroll (300 employees with avg. salary $100,000 per year to $30 million
include pensions and healthcare)
Contingency (50% of capitol investment for un-ventured process, Atkinson, 1991) $6.9 million
Other Plant Overhead (telephone, sewer, freight, travel, legal, rentals) $400,000
Taxes and Insurance (6% of capitol investment, Atkinson 1991) $828,000
Annual Expense due to Interest Rate for Capitol Investment (10%)
Sales and Marketing
69
Total annual costs after first year $58.3
million
Price per gram MAb to receive minimum 20% ROI $117
Nomenclature
Symbol Meaning
F Force
Q Charge
E Dielectric constant
r Radius
Vg Sedimentation
velocity
n Plate number
w Angular velocity
g Gravity constant
Σ Sigma Factor
a Area
ρ Density
μ Viscosity
70
Works Cited
ACCompacting. Master 280 Washing System. Copyright 2005.
http://www.accompacting.com/html/ico/washer_mastercl280.html. April 1,2005.
Alfa Laval. http://www.alfalaval.com . May 16th, 2006.
American Cancer Society. Cancer Statistics 2006: A Presentation by the American Cancer
Society. http://www.cancer.org/docroot/stt/stt_0.asp April 10th, 2006.
Atkinson, Bernard. 1991. Biochemical Engineering and Biotechnology Handbook.
Stockton Press. New York, NY.
Baez, J. Talks about different kinds of bioreactors. Fibrogen. 2006.
Barbosa, Maria J., Marcel Janssen, Nienke Ham, Johannes Tramper, and Rene´ H. Wijffels.
Microalgae Cultivation in Air-Lift Reactors: Modeling Biomass Yield and Growth Rate
as a Function of Mixing Frequency. Wiley Interscience Journal, 2002.
Bio-Rad Website. http://www.bio-
rad.com/B2B/BioRad/jsp/default.do?BV_SessionID=@@@@0629350043.1147759348
@@@@&BV_EngineID=cccgaddhkjmhjhkcfngcfkmdhkkdflm.0&catalogNumber=161-
0798&contentDescrip=1.5+M+Tris-
HCl%2C+pH+8.8%2C+1+L&nextPage=BRCatalogPage&oid=16362
Bio-Rad Website. http://www.bio-
rad.com/B2B/BioRad/product/br_category.jsp?BV_SessionID=@@@@1869198434.114
7637990@@@@&BV_EngineID=ccccaddhkifekmkcfngcfkmdhkkdflm.0&categoryPath
=%2fCatalogs%2fLife+Science+Research%2fChromatography+%7c+Protein+Purificatio
n%2fChromatography+Media%2fAffinity&catLevel=5&divName=Process+Separations
&loggedIn=true&lang=English&country=US&catOID=-
31129&isPA=false&serviceLevel=Lit+Request
Biotech Industry Organization. Advantages of Plants to Produce Therapeutic Proteins. 2006.
http://www.bio.org/healthcare/pmp/factsheet3.asp April 9th, 2006.
Contreras A, Garcia F, Molina EC, Merchuck JC. 1998. Interaction between CO2 mass transfer,
availability and hydrodynamic stress in the growth of Phaeodactylun tricornutum in a
concentric tube airlift photobioreactor. Biotechnol Bioeng 60:317-325.
CNNMoney.com. “Genentech: An expensive darling” by Aaron Smith. January 10, 2006.
http://money.cnn.com/2006/01/09/news/companies/genentech/> April 8, 2006
Crabtree, Penny. Algae to Antibodies. Union Tribune. May 20th, 2005.
http://www.signonsandiego.com/news/business/biotech/20050520-9999-
lz1b20algae.html April 8th, 2006.
Crisenten L, Iversen JL. A Bubble Column Photobioreactor with Online Optimization of Light
Intensity and Determination of Photosynthetic Quotient. Institute of Biochemistry and
Molecular Biology. University of Southern Denmark. Odense, Denmark.
Datar, Cartwright, and Rosen. Process Economics of Animal Cell and Bacterial
Fermentations. Biotechnology. Volume 11 (1993). 349-57.
71
Davies, Julian E. and Arnold L. Demain, editors in chief. Manual of Industrial Microbiology
and Biotechnology, 2nd Edition. Washington, D.C.: ASM Press, 1999.
Deindoerfer, F. H. and A. E. Humphrey. Microbiological Process Discussion: Analytical
Method for Calculating Heat Sterilization Times. Applied Microbiology. 1958.
D’Elia, Nancy A., Bjorn K. Lydersen, and Kim L. Nelson. Bioprocess Engineering: Systems,
Equipment, and Facilities. New York: Wiley, 1994.
Demain, Arnold L. and Nadine A. Solomon, editors. Manual of Industrial Microbiology and
Biotechnology, 1st Edition. Washington, D.C.: ASM Press.1986.
Food and Drug Administration. Avastin Consumer Reports. July 21st, 2004.
http://www.fda.gov/cder/consumerinfo/druginfo/Avastin.HTM April 10th, 2006.
Fox, Sandra. Contract Manufacturing Fills Industry Niche. Genetic Engineering News. Volume
25, No. 17. October 1st, 2005.
http://www.genengnews.com/current/article.aspx?id=777 April 8th, 2006.
García González M, Moreno J, Manzano JC, Florencio FJ, Guerrero. Production of Dunaliella
salina biomass rich in 9-cis-B-carotene and lutein in a closed tubular photobioreactor. J.
Biotech. 115:81-90. 2005.
Genentech .“What is Avastin?” 2005. April 8,
2006
Genentech. “Lucentis” March 2006
http://www.gene.com/gene/pipeline/status/vascular/amd/index.jsp?> . April 8, 2006
Genentech. 2005 Annual Reports. March 14th, 2006.
http://www.gene.com/gene/ir/financials/annual-reports/2005/index.jsp April 9th, 2006.
Gorry SB, Haas WE, Mahaffee JW. Effective Deoxygenation by a Hybrid Process Combining
High Transfer Membranes with Catalytic Oxygen Reduction. Technical Paper. Ionics
Pure Water Solutions.
Greenovation. 2006. http://www.greenovation.com/Projects_Production.htm April 10th, 2006.
Grisham RH, Grisham CM. Biochemistry. 2nd Edition. Forth Worth: Saunders. 1999.
Harrison, Roger G., Paul Todd, Scott R. Rudge, and Demetri P. Petrides. Bioseparations Science
and Engineering. New York: Oxford University Press, 2003.
Hetherington, P.J., Follows, M., Dunnill, P., and Lilly, M.D. (1971) Release of protein from
bakers’ yeat (Saccharomyces cerevisiae) by disruption in an industrial homogenizer.
Biotech Bioeng. 13(4):549-560
Hu, Jim. The Price of Life. Blogs for Industry. February 20th, 2006.
http://dimer.tamu.edu/simplog/archive.php?blogid=3&pid=3265 April 9th, 2006.
Hu, Wei-Shu. Overview of Cell Culture Technology. University of Minnesota. 2004.
HEADING 1 April 9th, 2006.
Ion Exchange. Copyright 2003. May 12, 2006
Janssen M, Janssen M, Winter M, Tramper J, Mur LR, Snel J, Wijffels RH. Efficiency of light
utilization of Chlamydomonas reinhardtii under medium duration light/dark cycles. J
Biotech. 78: 123-137.
Kargi, Fiket and Michael L. Shuler. Bioprocess Engineering: Basic Concepts, 2nd Edition. New
Jersey: Prentice-Hall, Inc., 2002.
Kwon SJ, Choi EY, Choi YJ, Ahn JH, Park OK. Proteomic studies of post-translational
modifications in plants. J. Exp. Bio. 57:7:1547-1551. 2002.
Lewin, Daniel R., J. D. Seader, and Warren D. Seider. Product and Process Design Principles:
Synthesis, Analysis, and Evaluation, 2nd Edition. New York: Wiley, 2004.
72
LHS Company. CIP Cleaning Systems. Copyright 2005 http://www.l-h-s.co.uk/cip.htm. April 1,
2005
Liszewski, Kathy. Exploiting Post Translational Modifications. Genetic Engineering News.
Volume 25, No. 1. January 1st, 2005.
http://www.genengnews.com/current/article.aspx?id=862 April 8th, 2006.
Longo presentation notes. Powerpoint slides given in lecture for ECH 161B. Winter 2006.
Maffei, K. KMPT Process Technology. Free of contamination: CIP cleaning in a GMP
centrifuge. http://www.kmpt.com/index.htm
Merchuk JC, Ronen M, Giris S Arad (Malis). Light/dark cycles in the growth of the red
microalga Porphyridium sp. J. Chem Technol Biotechnol. 59: 705-713.
Microfluidizer Corporation. http://www.microfluidicscorp.com/equipment/610.html
Mitsubishi Kakoki Kaisha, LTD. GMP-Compatible Horizontal Peeler Centrifuge.
http://www.kakoki.co.jp/english/products/m/m-004.html
Morgan, Andrew. Barnum Mechanical Inc. presentation. Thursday, May 4, 2006.
MorkUSA. Mobile CIP Units. Copyright 2003
http://www.morkusa.com/Default.asp?PageId=159. April 1, 2005.
Mullikin RK, Rorrer GL. A tubular recycle photobioreactor for macroalgal suspension cultures:
biohydrogen. New York: Plenum Press, 1998.
Nest Group, The. PolyLC Ion Exchange Column Care and Cleaning.
http://www.nestgrp.com/protocols/polylc/iex_clean.shtml
Pearson, Sue. Genetic Engineering News vol. 26, no. 6. Bioprocessing: Multiproduct
Manufacturing Facilities. 2006.
Quig A, and Beardall J. Protein turnover in relation to maintenance metabolism at low photon
flux in two marine microalgae. Plant cell and environment. 26: 693-703.
Rubio FC, Acien FG, Sanchez JA, Camacho FG, Grima EM. Prediction of disolved oxygen and
carbon dioxide concentration profiles in tubular photobioreactors for microalgal culture.
Biotech & Bioeng: 62:1:71-86, John Wiley & Sons, Inc.
Ryu, Dewey, Kyung Kang, William N. Drohan, and Carolyn L. Orthner. “Effect of Matrices on
Affinity Purification of Protein C.” Biotechnology and Bioengineering 39: 1086-1096.
New York: Wiley, 1992.
Sepharose Flow Rate.Copyright 2006.
. May 11, 2006
Sepharose Base Ion exchange Media Information Sheet. Copyright 2005.
.
May 12, 2006.
Sigma-Aldrich Website.
https://www.sigmaaldrich.com/catalog/search/ProductDetail/SIAL/S1804
Stark, D. Talks at UC Davis. Genentech. 2006.
Steiner, Dr. U. Presentation on MAb Market. Bayer . 2005.
Production of recombinant proteins by plants April 10th, 2006.
Subramanian, G. Antibodies, vol. 1: Production and Purification. New York: Kluwer
Academic/ Plenum Publishers, 2004.
Subramanian, G. Antibodies, vol. 2: Novel Technologies and Therapeutic Use. New York:
Kluwer Academic/ Plenum Publishers, 2004
Wu XiaoXi, Merchuck JC. 2002. Simulation of algae growth in a bench-scale bubble column
73
reactor. Wiley Periodicals, Inc. 2002.
Appendix
Calculations for Process Capacity, Bioreactor Design, Media Requirements, and Flow
Rates will be in the Final Report.
Upstream Media Sterilization Heat Exchanger Calculations
assume a volume of 10,000 Liters of media needs to be sterilized
sterilization of 10 m3
Time for sterilization 2 hr
Flow rate 1.389 L/s 0.001389 m3/s
KJ/(Kg
specific heat of water 4.186 *K)
KJ/(Kg
specific heat of sea water 3.93 *K)
Regen. Heat Exchanger
T hot in 130 C
T hot out 50 C
T cold in 30 C
T cold out 120 C
choose plate data
width, w 40 cm 0.4 m
height, h 30 cm 0.3 m
spacing 0.4 cm 0.004 m
number of
plates 16 plates
port size 30 cm
Thickness of the plates, a 0.00117 m
delta T ln 358.0519
delta T 373.15 K
Q 2169.453 KW
Q delta T ln 2081.674
74
A 1.68 m2
U 3460.648 W/(m2*K)
thermal conductivity
k (hot stream) 0.016 W/(m*K)
k(cold stream) 0.58 W/(m*K)
k (SS) 17 W/(m*K)
Pr=Cp*u/k
Pr (hot stream) 196.5
Pr (cold
stream) 6.775862
Re 2320.719
Nu=0.023*Re^(.8)*Pr^(1/3)
Nu(hot stream) 718.4363
Nu(cold
stream) 233.8419
hydraulic diameter
Dh 0.002333
h=Nu*k/Dh
h(hot) 4926.754 W/(m2*K)
h(cold) 58130.34 W/(m2*K)
1/U 0.000289 (m2*K)/W
U 3460.212 W/(m2*K)
assume diameter of pipe is 3 inches 0.0762 m
cross sectional area of
pipe 0.00456 m2
velocity = 0.304556 m/s
Steam Heat Exchanger
T hot in 140 C
T hot out 120 C
T cold in 120 C
T cold out 130 C
choose plate data
width, w 40 cm 0.4 m
height, h 30 cm 0.3 m
75
spacing 0.4 cm 0.004 m
number of
plates 24 plates
port size 30 cm
Thickness of the plates, a 0.00117 m
delta T ln 287.577
delta T 293.15 K
Q 1704.342 KW
Q delta T ln
A 2.64 m2
U 2202.231 W/(m2*K)
thermal conductivity
k (hot stream) 0.016 W/(m*K)
k(cold stream) 0.016 W/(m*K)
k (SS) 17 W/(m*K)
Pr (hot stream) 209.3
Pr (cold
stream) 245.625
Re 2320.719
Nu(hot stream) 733.709
Nu(cold
stream) 773.9121
hydraulic diameter
Dh 0.002333
h(hot) 5031.488 W/(m2*K)
h(cold) 5307.185 W/(m2*K)
1/U 0.000456 (m2*K)/W
U 2193.003 W/(m2*K)
how much steam is needed
76
length of holding tube = v* 5 mins Q= 1704.342 kW
0.0762
diameter of holding tube 3 inches m m dot = 5.813889 L/s
cross sectional area of
tube 0.00456 m2 for .5 hrs,
Liters of
steam per
heat
velocity in sterilized
tube 0.304556 m/s m= 10465 batch
length of tube 91.36691 m
Upstream CIP Calculations
CIP Appendix
Velocity= ,408 where Q is the flow rate and d is the
Reynolds number = (d*v*p)/mu Q/d^2 diameter
d inner diameter 2 inch
v fluid velocity 6.51 ft/s or 122.9609 m/min
p fluid density 60.6 lb/ft^3
flow
mu viscosity 0.35 cP Line size d inside Re rate
30 L/min or
1.5 1.37 243013 38gpm
For
steam
30 gpm or 14434.62 lb/hr Re 192000 for inner diameter 1.37
Cost Analysis water(L) sodium hydroxide
1liter= .264 gal 760 49.6 gal
684 Sodium hydrochlorite
570 49.6 gal
steam
380 58400 lb
684
380
380
380
380
total water 4598 L or 1213.872 gal of water
flow rate of sodium hydrochlorite and sodium hydroxide
Re= 6.31 *W/ d * mu where W is lb/hr
77
Re 4000
W 303.962
flow rate 0.625481 gpm
Centrifugation Calculations: Cell Removal
First Centrifuge: Remove Cells from Broth
Cell Density Calculations
Vcell
cell D (um) dry cell (g) water/cell mass (g) (um3) Vcell(cm3) cell density (g/cm3)=p
1.10E+0
10 1.15E-10 0.8 5.75E-10 523.3333 5.23E-10 0
Disc-Type Centrifugation Equation: Maximum Feed Rate
Q
2a 2 o g 2n 2 Ro 3 R13 cot
v g
9
3g
bowl radius
theta (cm)
42 30
u Ro Ri
a (cm) p(g/cm3) po(g/cm3) g (cm/s2) g/(cm*s) n w(rad/s) (cm) (cm)
0.0005 1.10E+00 1 980 0.798 100 628 25 8
Q (l/min) Q (m3/min) Feed Rate From Fermentation
70 0.07
Q (cm3/s) vg sigma
3742.539 6.7357E-06 5.56E+08
0.224552 Maximum Feed Rate Allowed to Achieve Seperation (m3/min)
Therefore, our process flow rate is acceptable to achieve desired separation in the chosen
centrifuge.
78
Time Required to Achieve Seperation
2 ro
Gt t Generalized Gt for Algae Cells
g 3.00E+05
w(rad/s) Ro g t(seconds)
29.81865
628 25 980 39
Centrifuge Capacity Centrifuge Height
Volume(L) Q(L/min) time to separate (s) Volume Cone = 1/3*pi*r2*H (m3) r(m) h(m) h(cm)
34.78843 70 29.81865 0.034788 0.25 0.53 53.18
Centrifuge Time For Total Process Batch Volume of 100,000L
Batch Volume per
Q(L/min) Reactor Total Batch Volume Total time (min) Time (hours)
23.80
70 10,000 100000 1428.571 952
Second Centrifuge: Remove Cell Debris from Protein Broth
Cell Density Calculations
Vcell
cell D (um) dry cell (g) water/cell mass (g) (um3) Vcell(cm3) cell density (g/cm3)=p
10 1.15E-10 0.8 5.75E-10 523.3333 5.23333E-10 1.10E+00
debris mass debris V cell debris density
7.48E-11 6.54167E-11 1.14E+00
Disc-Type Centrifugation Equation: Maximum Feed Rate
Q
2a 2 o g 2n 2 Ro 3 R13 cot
v g
9
3g
theta bowl radius (cm)
42 40
79
u
a (cm) p(g/cm3) po(g/cm3) g (cm/s2) g/(cm*s) n w(rad/s) Ro (cm) Ri (cm)
0.00025 1.14E+00 1.07 980 0.798 100 837.3333 35 8
Q (l/min) Q (m3/min) Feed Rate From Fermentation
70 0.07
Q (cm3/s) vg sigma
3432.21 1.24E-06 2.769E+09
0.205933 Maximum Feed Rate Allowed to Achieve Separation (m3/min)
Therefore, our process flow rate is acceptable to achieve desired separation in the chosen
centrifuge.
The maximum throughput of the centrifuge will allow for fluctuations of up to 75% more of the expected process flow rate.
Time Required to Achieve Seperation
2 ro
Gt t Generalized Gt for Algae Cells
g 2.00E+06
w(rad/s) Ro g t(seconds)
837.3333 35 980 79.87139
Centrifuge Capacity Centrifuge Height
Volume(L) Q(L/min) time to separate (s) Volume Cone = 1/3*pi*r2*H (m3) r(m) h(m) h(cm)
93.18329 70 79.871394 0.093183 0.35 0.726764 72.67643
Centrifuge Time For Total Process Batch Volume of 100,000L
Batch Volume per
Q(L/min) Reactor Total Batch Volume Total time (min) Time (hours)
70 10,000 100000 1428.571429 23.80952
Lysis by microfluidizer
ln(1-R) = -kNbPa
b = 0.3
k = 0.3 MPa-a
a = 0.6
p = 20,000 psi = 137 MPa
R = Extent of disruption
N = Number of pass
Ln(1-R) = -5.74
1-R = exp (-5.74)
80
R = 0.9968
99.68%
Heating
~0.33 Kg ice/L required to cool microfluidizer
(0.33 Kg ice /L )(334 KJ / KJ ice) ( 35 L/min) = heat generated
= 3857.7 KJ/min
=923 Kcal/min or 6.86 Kcal/L
Cooling
923 Kcal/min means 923 Kg change 1ºC every minute
Assuming heat transfer coefficient to be limited by water (and not by resistance of metal)
= 250 BTU / ft2 hºF = 1420 W / m2K
Assume 1 m2 surface area.
6.86 Kcal/L corresponds to 6.86 º C increase since 1L ~ 1 Kg
∆Tln = (15- 6.85) / (15 / 6.86)
= 10.4 ºK
q = UA∆Tln
= (1420 W / m2K)(1 m2)(10.4 ºK)
= 886 KJ/mim
(3857.7 KJ/min) / (886 KJ/min) = 4.35 min
Calculations for Protein A Affinity Column
Determining mass flow rates and volumetric flow rates for Affinity Columns.
conc of mAb coming out of 2nd centrifuge:
0.006mg/mL
want flow rate of 22
mass in per cycle 15829.04
number of cycles 36
mass out of end ion exchange column
16.66667kg/cycle
mass in of end ion exchange column
0.9595% separation
17.54386kg/cycle
mass out of affinity column
17.54386kg/cycle
0.9595% separation
mass in to affinity column
18.46722kg/cycle
81
mass in 18467.22mg/cycle
vol in 3077870mL/cycle
vol in 3077.87L/cycle
time 140min
flow rate 21.98479L/min
assume time is 140min
Now find area using flow rate assume 3:1 diameter to height ratio
But the height should not be much larger than 2 feet.
So we will have 2 protein A affinity columns in series
D= 1.792257m 5.880037feet
H= 0.61m 2.001288feet
V= 1.538935m^3
A(cs) = 2.522844m^2
Determining amount of resin needed, container size, and flow rate
resin needed resin container
12 mg protein/mL resin V= 2000L
mass in 9.71959kg/cycle V= 2m^3
mass in 9719590mg/cycle H = 3D
resin capacity 12mg protein/mL D= 0.921318m
resin vol 809965.8mL H= 2.763953m
resin vol 809.9658L per column
resin load time 60min
resin flow rate 13.49943L/min
Determining amount of elutant needed and flow rate
mass out 17.54386kg/cycle
mass out 17543860mg/cycle
elute volume 2500L
elute volume 2500000mL
conc out: 7.017544mg/mL
Determining pressure drop down column. From Bio-Rad website, the maximum pressure drop
the beads can withstand is 1,000 psi. http://www.bio-
rad.com/B2B/BioRad/product/br_category.jsp?BV_SessionID=@@@@1869198434.114763799
0@@@@&BV_EngineID=ccccaddhkifekmkcfngcfkmdhkkdflm.0&categoryPath=%2fCatalogs
%2fLife+Science+Research%2fChromatography+%7c+Protein+Purification%2fChromatograph
y+Media%2fAffinity&catLevel=5&divName=Process+Separations&loggedIn=true&lang=Engli
sh&country=US&catOID=-31129&isPA=false&serviceLevel=Lit+Request
p v
L k
82
maximum L allowable
max. del p 6896552N/m^2
3
2
dp
k L= 1.983478m
150 (1 ) 2 The height of the column cannot exceed 2 meters
dp= 45um
dp= 0.000045m
assume void fraction is .1
e= 0.1
k= 1.66667E-14m^2
100feet
L= 0.61m 30.48037m
assume u is that of water
u= 0.000798kg/m*s
v=Q/A
Q= 0.010992393m^3/min
Q= 0.000183207m^3/s
A= 2.522844356m^2
v= 7.2619E-05m/s
delta p = 2120970kg/(m*(s^2))
2120970N/(m^2)
307.54065psi
Cost Calculations
Tris-HCl buffer needed
180000L per year
from bio-rad
1 L = $22 for 1.5M
so dilute 1.5M to .1M by 15 dilutions to get 1L of .1M = $1.47
cost of 180000 L $ 264,000.00
1 L = $1.47
Sodium Citrate elutant needed
180000L per year
5 kg = 214british pounds
Assume density of 1 kg/m^3
5 m^3 = 403.9464US Dollars
5 m^3= 5000L
price 0.0807893$/L
cost of 180000 L $14,542.07
83
Cost of 2,000 L protein A container
V (in gallons)
Cost =375*V^.51
from page 556 in Seider
V= 2000L
V= 528.40159gallons
cost $9,177.89
cost of affinity column shell
V= 1540L
V= 406.86922gallons
cost $8,032.55 per column
Anion Exchange Column Calculations
Anion Exchange Column
Total Mab 600kg Mab
total cycle 36cycles
mab/cycle 16.66666667mass mout
% recovery 0.95
mass in 17.54385965kg Mab
Find flow rates and bed volumn dimensions
For two column
time
assume mAb conc in 7.018mg/mL
mass in 8.771929825kg/cycle
mass in 8771929.825mg/cycle
Volume in 1249918.755mL/cycle
Volume in 1249.918755L/cycle
time 55min
flow rate 22.72579555L/min
For two column
Volume per column 624.9593776L
Assume 3:1 diameter to height 0.624959378m^3
Diameter 1.336475661m OR 4.276722ft
Height 0.445491887m OR 1.425574ft
Flow rate per column 11.36289778L/min
Pressure drop - Use Darcy Law k 44.03
pressure drop 687.15psi
mass out 8.333333333kg
elution vol 3CV 1874.878133L
1874878.133mL
conc. Out 4.444733333mg/mL
84
Find resin volumne
Use binding capacity 30mgAb/ml
know need 16.67kg of Ab
so calculated ml of resin needed 555666.67mL
85