Fall 2012 • Volume 11 / Issue 3 • ISSN 1538-8786

BioProcessing
J O U R N A L
Trends & Developments in BioProcess Technology

A Production of BioProcess
Winter 2010/2011 BioProcessing Journal • 3 Technology Network
• www.bioprocessingjournal.com
 TRENDS & DEVELOPMENTS IN BIOPROCESS TECHNOLOGY

Cell-Based Potency Assays:

Expectations and Realities
By ROBERT D. SCHROCK

activity based on the attribute of the product that is

A
mong the battery of assays
needed to release viral-based
gene therapy biologics products,
the potency assay is typically the
most customized and unique
of all the tests found on a Certificate of Analysis
(COA). Where many safety, purity, and to some
extent, identity assays can be applied with little
to no custom modifications, the potency assay
should be directly related to the biological activity
of the construct. This aspect of the potency assay
requires that development, characterization, and
validation efforts be started early in the lifecycle
of the product, and adequate resources should
be allocated to this end. This article will cover a
brief overview of available regulatory guidances,
discuss hallmarks of a good cell-based potency
assay, and provide a case study of an assay that was
developed for a Phase 3 gene therapy product.
Introduction
Potency is typically thought of as “the capacity of a
product to produce an expected biological activity.” In
more formal language, the International Conference
on Harmonization (ICH) (section Q6B, Guidance for
Industry: Specifications: Test Procedures and Acceptance
Criteria for Biotechnological/Biological Products)[1] defines
potency as “…the quantitative measure of biological
linked to the relevant biological properties...”
A potency assay should be designed to:
• Measure a relevant biological activity of the product
• Measure a property specific to the product
• Be quantitative
• Include use of a reference standard (RS) for
comparison
Potency assays should be intended for use in lot release, to
demonstrate lot-to-lot consistency, and indicate stability.
Finally, the assay should be robust enough that it can be
validated.
The demonstration of potency can range from animal
models to cell-based assay systems that incorporate direct
measurement of a biological activity. In vivo animal
models that measure, for example, weight gain/loss or
antibody titer, can be a direct measure of biological
activity in a dose-dependent response. However, this
type of assay is very difficult to validate. Ex vivo assays
utilize cultures of primary cells from a specific tissue
upon which the product has activity. Examples of this
approach are signaling pathways and activities of growth
factors or hormones. Due to a lack of a consistent and
well-characterized cell bank, these assays can have
considerable variability. In vitro cell-based assays include
well-characterized cell banks and often incorporate
biochemical methods such as protein kinase activation,
transgene expression, or reporter gene expression. Based
on considerations of variability, reproducibility, and cost,
the goal should be to develop a cell-based potency assay
with a measurable, quantitative endpoint.

ABOUT THE AUTHOR

Robert D. Schrock, PhD (robert.schrock@lonza.com), Head of Quality Control
Lonza Viral-Based Therapeutics, 8066 El Rio St., Houston, Texas USA | Phone: 713-568-6190
Website: www.lonza.com/custom-manufacturing/biological-manufacturing/viral-based-therapeutics.aspx
This article is based on Dr. Schrock’s presentation at the ISBioTech 2nd Annual Meeting held in Rosslyn, Virginia, April 2-4, 2012.

Fall 2012 BioProcessing Journal • 4 • www.bioprocessingjournal.com

 The primary regulatory guidances are listed in
Table 1. The United States Pharmacopeial Convention
(USP) <111>[2] is applicable primarily toward the use
of in vivo potency assays — typically animal models or
microbiological assays. Recent publications, USP <1032>[3],
<1033>[4], and <1034>[5] have much more information
applicable to cell-based potency assays. These guidances
expand on expected attributes of potency assays and
describe various types of assays and parameters that are
to be included during validation.
Lifecycle of Potency Assay Development
Development of the potency assay should be initiated
early in the product lifecycle, during pre-clinical stages.
Potency assays are typically the most challenging to
implement as the molecular mechanisms of therapy need
to be understood to some extent. The basic requirements
of the assay should be demonstrated: (a) quantitative;
(b) specific to the product; and (c) stability-indicating.
By Phase 1 stage, important parameters (i.e., temperatures,
range, incubation limits, curve fit reliability) should be
defined (as part of the general qualification assay). It
may be acceptable for the assay specification to be “report
results” at this stage until enough data points are collected
to determine an acceptable range. Once these parameters
are known, pre-validation (planning of validation
strategy) activities should be underway by the time the
product is in Phase 2 trials. Full validation work should
be underway by late Phase 2 and early Phase 3 stages. At
TABLE 1. Regulatory publications.
TABLE 2. Lifecycle of potency assay development and validation.
Fall 2012 BioProcessing Journal •
this point, the assay method should be pretty well locked
down, although refinements that improve efficiency,
repeatability, and robustness can be implemented. It
is important that no fundamental changes to the assay
are made at this point unless the intent is to perform
comparison studies to demonstrate clear advantages in
the method. By the time the product enters Phase 2 trials,
specifications should be set. The specification range can
be wide, perhaps stating a minimum value, with the
intention of narrowing this range based on data from
multiple production lots (Table 2).
In Vitro Cell-Based Potency Assay Approaches
As mentioned earlier, the overarching premise behind
cell-based potency assays is to demonstrate that a product
has the expected biological activity. This is typically
customized based on the product. For example, infection
of a tumor cell line with an adenoviral product that
delivers an important tumor suppressor gene would
cause growth inhibition and cell death. When compared
to an “ad-empty” vector that does not contain the tumor
suppressor, the therapeutic vector concentration that
inhibits cell growth (IC50) for this cell line is much lower
than for the control vector.[10] Another example is an
enzyme delivered by a recombinant virus that activates
an inert pro-drug given in trans. The activated drug then
inhibits growth in a specific cell line.[11] In all cases, the
activity measured is calculated and statistical analysis can
be applied to the data. It is advantageous to identify a
5 • www.bioprocessingjournal.com
RS (in these examples, a well-characterized virus lot) to
be used to show assay comparability, determine system
suitability parameters, and to follow assay trending.
Matrix Approach
A suite of assays can be used to demonstrate a cascade
of activities that are essential for biological activity.
This approach is illustrated by using the case study of
a gene therapy vector that provides a functional tumor
suppressor gene that, when expressed in susceptible cells,
induces apoptosis. Four separate assays measure the steps
to true biological activity: (1) the virus particle (VP)
concentration is measured by absorbance at 260 nm
although this is more a measure of strength; (2) the
infectious titer measures the ability of the virus to infect
cells; (3) the Western blot or (preferably) ELISA assay
measures expression of transgene-encoding protein; and
(4) the IC50 assay most directly measures the biological
activity on susceptible cells. Taken together, this series of
assays is a powerful measure of the potency of the lot.[8]
The matrix approach is shown schematically in Figure 1.
Cell Line Attributes
Any cell line used in a quality control (QC) assay
needs to be well-characterized and tested.[8] Cell line
origin and history should be well-documented. Sterility,
Mycoplasma, percent viability, growth characteristics,
and some performance information from the potency
assay should be documented. A tiered cell bank system
(master and working cell bank) should be used to ensure
consistency in assays over time.[3] In addition, knowledge
FIGURE 1. Matrix approach to demonstrating potency.
Fall 2012 BioProcessing Journal •
of performance over time after thaw will help reduce
variability, so a maximum number of passages after thaw
should be established, typically as part of the validation.
Assay Design: Mitigating Variability
Variability is one of the biggest challenges one faces in
establishing a potency assay and in cell-based assays in
general. Variability is inherent in cell-based assays. A high
number of replicates per data point help mitigate variability;
therefore 96-well plate formats are often used. Well-to-
well variability should be explored during development,
including positional variability, such as edge effects. Other
parameters that affect variability are cell health, confluency,
passage number, and incubator conditions (including
temperature gradients).[12, 13] In our experience, cell health
is the largest single contributor to variability and risk of
invalid assays. Therefore, analysts should be well-trained
to assess cell health by appearance, and be enabled to abort
the assay if obvious cell stress is observed at early stages.
It is advisable, that standard operating procedures (SOPs)
include specific instructions on how to carry and maintain
cell lines. In addition, the assay protocol should have a
good balance of data points, replicates, and range without
increasing the number of plates used past the point where
the assay is too onerous to perform.
Case Study: Potency Assay
for a Gene Therapy Adenovirus
The remainder of this article will describe our experience
in developing and validating a bioactivity assay to assess
the potency of a gene therapy recombinant adenovirus.
6 • www.bioprocessingjournal.com
This project was undertaken well before any detailed
guidances were available for cell-based potency assays,
though it is a good general example of the development
and validation process for assay development. This
replication-deficient adenovirus was engineered to
express a tumor suppressor gene that is either deficient
or mutated in many cancer cells. Upon infection, the
expressed tumor suppressor induced apoptosis in
susceptible cells. By adding increasing amounts of VP/cell
(multiplicity of infection [MOI]), growth inhibition could
be quantified and an IC50 calculated (expressed in VP/
cell). A schematic of the assay setup is in Figure 2.
The bioactivity measured in this assay has been
linked to the ability of the tumor suppressor-containing
adenovirus to shrink tumors in animal models. Apoptosis
of tumor cells in vivo and in vitro has been demonstrated
through in situ staining for apoptotic markers. Tumor
shrinkage and in vitro growth inhibition of cells were
shown to be dependent on the presence of the tumor-
suppressor transgene, as opposed to ad-empty vectors that
did not contain this transgene.
Early versions of the assay were not quantifiable
and assay results were expressed as pass/fail. Cells were
infected at MOIs of 100, 500, and 1000 VP/cell. The test
article passed if the growth of the cells were reduced
≥ 50% (compared to no infection) at 1000 VP/cell after
four days of incubation. MOI data at 100 and 500 VP/cell
FIGURE 2. Assay design for increasing amounts of adenovirus-expressed tumor suppressor induces cell death.
TABLE 3. Developing a process to identify parameters that affect variability.
Fall 2012 BioProcessing Journal •
were included to accumulate data for future revisions of
the assay. Further assay development was carried out to
ensure the quantifiable nature of the assay. This included
adding additional data points to generate a more accurate
and precise curve fit and to verify that the four-parameter
sigmoidal curve fit model was appropriate. During the
pre-validation phase in developing this assay, various
parameters were examined as sources of variability.
Results are shown in Table 3.
Overall, the two major factors contributing to
variability were cell health (minimizing cell stress) and
analyst training. The method was revised to minimize
cell stress by not removing media from the cells before
infection. Concentrated virus dilutions were simply
added to the existing media in the 96-well plates. There
was little variance (amongst experienced analysts)
whether the virus dilution series was prepared in tubes or
plates. However, using a 96-well serial dilution plate and
transferring the fluid from this plate to the cell-containing
plate accelerated this step considerably, reducing time
spent during the infection step and making the assay
less sensitive to analyst experience. Finally, results from
analysts not experienced in this type of cell-based assay
varied considerably. Further training usually rectified this.
In addition, results were most consistent at four days post-
infection. Once these improvements were implemented,
the assay was consistent, with a good signal that detected
7 • www.bioprocessingjournal.com
 to cell-based potency
assay validation. Detailed
statistical analyses that can
be applied to this type of
assay will not be covered
in this article. Discussions
of statistical analysis as it
applies to biological assays
can be found in USP 35-
NF 30 chapters <1032>[3],
<1033>[4], <1034>[5] and
in EP chapter 5.3.[9]
According to USP
<1033>[4], the validation
parameters that should be
addressed in a cell-based
potency assay are: linearity,
relative accuracy, and
FIGURE 3. Determination of bioactivity for adenoviral gene therapy product. intermediate precision (IP)
which includes repeatability,
growth inhibition and four-parameter sigmoidal curve fit range, and specificity. Linearity is used to verify the
R2 values that were typically ≥ 0.9. An example of the data relative accuracy and effective range of the dilutional
output from the quantitative assay is shown in Figure 3. linearity method. Each potency assay typically measures
In this experiment, the test article and a reference control a unique activity. Therefore, the term “relative accuracy”
were prepared on separate plates in parallel. applies as there is often no way to confirm absolute
After improving the assay, the curves from 35 assays accuracy using an accepted alternative method. The
were analyzed (Figure 4). The minimum asymptote range is defined by the user and needs to be optimized
is small and represents the part of the curve where to cover values that are to be expected from the product.
there is 100% cell death. This parameter was the most IP is performed by evaluating intra-laboratory (different
variable. The primary indicator of potency is the curve fit analysts/days within a laboratory) and inter-laboratory
parameter, C, the IC50. Only one test of
35 was outside two standard deviations
of the mean value of 251 VP/cell. Based
on this assessment, the preliminary
specification for this viral construct
was set at an IC50 of 100 -500 VP/cell.
The specification would later evolve to
express the result as a ratio of sample
IC50 to reference IC50 that was run in
the same assay. The example in Figure 3
shows that this ratio would be 0.9.

Validation of Potency Assay:

Expectations
USP 35-NF 30: <1033>[4] (Biological
Assay Validation) describes general goals
in validation of relative potency assays.
This chapter is a guidance although
it does outline current expectations
for validation of cell-based potency
assays. The case study presented is
intended to illustrate a basic approach FIGURE 4. Curve fit statistics.

Fall 2012 BioProcessing Journal • 8 • www.bioprocessingjournal.com

 TABLE 4. Dilutional linearity of potency assay for adenoviral gene therapy vector.

(variation between laboratories) results. Specificity of an
assay is used to demonstrate a lack of interference from
components that are likely to be present in the sample
matrix. While not specifically required per USP <1033>[4],
robustness may be added to assess whether such things as
plate manufacturers, cell passage, and slight temperature
variations affect the outcome of the assay.
Validation of Potency Assay for
Gene Therapy Adenoviral Product
Data from the aforementioned adenoviral gene therapy
product is shown below. Dilutional linearity was shown
by the analysis of three different lots, run in duplicate,
with the assays performed on three different days. Virus
concentrations from 12-3000 VP/cell were added to
wells containing 10,000 cells/well. After addition of
alamarBlue® (Life Technologies Corp.), which indicates
the extent of cell metabolism based on decreasing
absorbance at 570 nm, the plate is read for absorbance in a
plate reader. The results are shown in Table 4.
The expected correlation of determination (R2) was
≥ 0.9 for a four-parameter sigmoidal curve. For these
assays, the mean R2 value was 0.999, demonstrating
excellent curve fit. Based on the % relative standard
deviation (% RSD) values of the absorbance values at the
various VP/cell concentrations, the limit of quantitation
was shown to be 480 VP/cell as higher VP/cell values
resulted in ~100% cell death. Based on the linear portion
of the curve, the range of the method was shown to be
77- 480 VP/cell.
To determine relative accuracy, the approach was to
prepare a standard curve in the manner of the assays
shown in Figure 3. In addition, two samples within the
linear range of the curve (200 and 400 VP/cell) were tested
as unknowns within the same experiment. The values
from these samples were calculated from the standard
curve and then compared to the theoretical values of
200 and 400 VP/cell. A % recovery was determined
([calculated value ÷ theoretical value] × 100). This was
repeated using three different lots, in duplicate, for each
lot. The acceptance criterion for % recovery was 50 -150%.
The results from these experiments are contained in
Table 5. The relative accuracy acceptance criterion for all
experiments was met. The mean % recovery for the 400 VP/
cell sample was 87% and 80% for the 200 VP/cell sample.
Repeatability of the assay was determined by testing ten
replicates from one lot at three different concentrations
(77, 192, and 480 VP/cell). As for relative accuracy, the %
recovery was calculated by comparing the calculated value
to the theoretical values. The mean and % RSD of the ten

TABLE 5. Relative accuracy of potency

assay for adenoviral gene therapy vector.

Fall 2012 BioProcessing Journal • 9 • www.bioprocessingjournal.com

replicates were calculated. The acceptance criterion for
% recovery was 50-150%, and the criterion for % RSD
was ≤ 40%. The results from these experiments are in
Table 6. Both acceptance criteria for this experiment were
met, demonstrating that the assay gives consistent values
within the usable part of the standard curve.
To assess intra-laboratory precision, the assay was
performed on three different days using two analysts, and
two concentrations (200 and 400 VP/cell) were tested
within each experiment. Six replicates were assayed each
day, along with a standard curve run in the same plate
each day. For each day, the calculated VP/cell value was
averaged and the % RSD and % recovery calculated for the
six replicates. The % recovery acceptance criterion over the
three days of the test was 50-150% which was met (91%
for 400 VP/cell and 89% for 200 VP/cell concentrations).
The % RSD values were acceptable at < 10% for each day.
These results are summarized in Table 7.
Inter-laboratory precision measures how well an assay
transfers from lab-to-lab, including different analysts,
on different days and with different instruments. It
demonstrates how resistant an assay is to intangibles
that are not defined in the protocol. To demonstrate
inter-assay precision, the IC50 of a single lot of product
was determined three times each by two laboratories
(using different analysts, equipment, and on different
days). The mean IC50 values and % RSD of the three
runs from each laboratory were compared. Results from
the internal laboratory were an IC50 of 227 VP/cell and a
% RSD of 16.7%, and results from the outside laboratory
were an IC50 of 185 VP/cell and a % RSD of 6.8%. The
variability of the results was found to be within the
expected range of the assay (± 50%). These results are
shown in Table 8.
For cell-based potency assays, specificity involves
demonstrating certain components that may be expected
to be present in the sample do not interfere with the assay.
In this example, it was reasoned the most likely source of

TABLE 6. Repeatability of potency assay

for adenoviral gene therapy vector.

TABLE 7. Intra-laboratory precision of potency

assay for adenoviral gene therapy vector.

TABLE 8. Inter-laboratory precision of potency

assay for adenoviral gene therapy vector.

Fall 2012 BioProcessing Journal • 10 • www.bioprocessingjournal.com

interference would be inactive virus. Empty adenoviral
capsids could also contribute to specificity loss, but the
presence of unstable inactive virus was thought to be
the most probable and abundant source of interference,
especially during forced degradation stability studies. The
inactivated virus could then be used to spike active virus
and the IC50 values compared ± the spike. To produce
inactive virus, purified final product was heat-inactivated
(56 °C for one hour) and used as the challenge compound.
As shown in Table 9, inactivation was complete. Heat
inactivation resulted in an IC50 value of > 3000 VP/cell
for the inactivated virus alone. In a preparation of equal
parts active and inactive virus, the measured activity of
the active virus was essentially the same as when measured
alone (IC50 values of 205 and 180 VP/cell, respectively).
These results demonstrated that the assay could effectively
measure potency in the presence of inactive product, thus
the assay had acceptable specificity.
In addition to the assay specificity experiment, the
specificity of the system was explored by comparing
IC50 values obtained from the cells typically used in the
assay compared to non-susceptible cells. The IC50 values
indicated that the non-susceptible cells required over
three times more virus to achieve an equivalent reduction
in cell metabolism than when compared to the cells used
in the assay (data not shown).
It is advantageous to demonstrate the robustness of
the potency assay. In this case, robustness was tested by
assessing different cell culture plate suppliers, media
source, cell passage number, as well as the effects of a
lower incubator temperature (it is known that higher
incubator temperatures severely reduce infectivity). The
TABLE 9. Specificity of potency assay
for adenoviral gene therapy vector.
TABLE 10. Robustness of potency assay for adenoviral gene therapy vector.
Fall 2012 BioProcessing Journal •
IC50 values were variable between assays for some of these
variances. However, the relative accuracy within each
assay was more consistent. When two concentrations of
sample (200 VP/cell and 400 VP/cell) were compared to
a standard curve within the same assay (as in Table 5),
% recovery ranged from 73-117%, well within the
acceptance criterion of 50-150% for relative accuracy.
These results (Table 10) indicated that, although absolute
IC50 values might change in some cases, relative accuracy
was not meaningfully compromised nor was the R2
correlation coefficient for the four-parameter sigmoidal
curve fit. In other words, a product tested relative to a RS
would be expected to yield the same IC50 ratio as long as
the test article and reference were tested under the same
conditions.
The use of a RS is important throughout development,
validation, and trending of assay performance. During
development, a standard may be a process development
lot or Phase 1 material and can be used for early system
suitability testing. As long as it is (approximately)
representative of the manufacturing process for future
good manufacturing practices (GMP) lots, this approach
can help elucidate the true variability of the assay. Once
the product is in late-phase clinical development, an
official RS should be well-characterized (e.g., from part of
a GMP clinical trial lot). As part of this characterization,
an external RS (if available) should be tested side-by-
side as part of internal RS release. This provides a good
reference point between internal RS lots. For example,
even if the adenoviral reference material cannot be
tested for transgene bioactivity, the comparisons from
other tests such as infectious titer and virus particle
11 • www.bioprocessingjournal.com
enumeration give more confidence that the internal potency assay development early in a product’s lifecycle.
RS is representative of the product that will be tested. It is essential to understand some of the molecular
Internal RS should be stored in stable conditions. mechanisms of the product’s therapeutic effect and
Trend analysis of RS is the ideal way to assess assay narrow down the assay to a single, measurable biological
drift, QC cell bank comparability, analyst to analyst activity. The capabilities and limits of the assay should be
comparability, and changes in assay variability. understood before validation begins, with plans to validate
the assay once the product is in Phase 3 clinical trials.
Conclusion
The cell-based potency assay is usually the most
customized assay on a biological product’s COA, and ACKNOWLEDGEMENT
it typically requires a higher commitment of time The author would like to thank Ricardo Jimenez, Dr. Jessica Carmen,
and resources to develop. It is, therefore, wise to start and Kim Yang for assistance and critical review of this manuscript.

REFERENCES

[1] International Conference on Harmonization. ICH Topics Q6B-
Guidance for Industry: Specifications: Test Procedures and Acceptance
Criteria for Biotechnological/Biological Products. 18 August 1999,
vol 64, p 44928, http://www.ich.org/products/guidelines/quality/
article/quality-guidelines.html.
[2] USP <111> Design and Analysis of Biological Assays. USP 35–
NF 30:106, http://www.usp.org/meetings-courses/workshops/past-usp-
workshops/usp-bioassay-guidance-chapters.
[3] USP <1032> Development and Design of Biological Assays.
USP 35–NF 30: 162, http://www.usp.org/sites/default/files/usp_pdf/
EN/2010-03-25_1032_PF36(4)_w_line_numbers.pdf.
[4] USP <1033> Biological Assay Validation. USP 35–NF 30: 5174,
http://www.usp.org/sites/default/files/usp_pdf/EN/2010-03-25_1033_
PF36(4)_w_line_numbers.pdf.
[5] USP <1034> Analysis of Biological Assays. USP 35–NF 30: 5186,
http://www.usp.org/sites/default/files/usp_pdf/EN/2010-03-25_1034_
PF36(4)_w_line_numbers.pdf.
[6] USP First Supplement. Design and Analysis of Biological
Assays. USP 35–NF 30, http://www.uspnf.com/uspnf/
display?cmd=jsp&page=chooser.
[7] USP Second Supplement (not yet official). USP Bioassay Guidance
Chapters, 2012, http://www.usp.org/es/node/595.
[8] Center for Biologics Evaluation and Research (CBER). Guidance
for Industry: Potency Tests for Cellular and Gene Therapy Products.
Rockville (MD): CBER. Jan 2011, http://www.fda.gov/downloads/
BiologicsBloodVaccines/GuidanceComplianceRegulatoryInformation/
Guidances/CellularandGeneTherapy/UCM243392.pdf.
[9] Statistical Analysis of Results of Biological Assays and Tests:
European Pharmacopoeia 7th Edition. July 2012, http://online6.edqm.
eu/ep705/.
[10] Peng Z. Current Status of Gendicine in China: Recombinant
Human Ad-p53 Agent for Treatment of Cancers. Human Gene Therapy,
2005; 16 (9): 1016-1027.
[11] Trask TW, Trask RP, Aguilar-Cordova E, Shine HD, Wyde PR,
Goodman JC, Hamilton WJ, Rojas-Martinez A, Chen SH, Woo SLC,
Grossman RG. Phase I Study of Adenoviral Delivery of the HSV-tk Gene
and Ganciclovir Administration in Patients with Recurrent Malignant
Brain Tumors. Molecular Therapy, 2000; 1 (2): 196-203.
[12] Rieder N, Gazzano-Santoro H, Schenerman M, Strause R, Fuchs C,
Mire-Sluis A, McLeod LD. The Roles of Bioactivity Assays in Lot Release
and Stability Testing. BioProcess International, 2010; 8 (6): 33-42.
[13] Joelsson D. A Practical Guide to Design of Experiments (DOE) for
Assay Developers, 2010, http://www.potencyassay.com/wp-content/
uploads/2010/06/DOE_for_assay_developers_Chp1_Rev-1.0.pdf.

BioProce Winter
200
9/2
ISSN 1538 010

ssing
-8786

J O U
R N
Trends
A L

editor@
an
d Deve

A Call For
lopme
nts in
BioPro
cess Te
chnolog
y

bioprocessing
ARTICLES journal.com
Vol. 8/N
o. 4

ww w.
bio pr
oc es sin
gj ou rn
al. co m

call-for-articles-J83.indd 1 10/16/2012 11:01:32 AM

Fall 2012 BioProcessing Journal • 12 • www.bioprocessingjournal.com