It!"

 

X

93W
. ."6,
l ‘1‘

1'

,.

.‘r'.
age
...|

hag-.99

9'91;
VT?

”3’9

a
‘v
.9

,.

 

f.
9

ﬁai
. 399:5.
9

 

LIBRARIES
a MICHIGAN STATE UNIVERSITY
F f , , EAST LANSING, MICH 48824-1048
r: ‘ j l , .

v

“2930179

This is to certify that the
thesis entitled

Optimization and Control of Recombinant Protein Expression

presented by
Casey Preston

has been accepted towards fulfillment
of the requirements for the

MS. degree in Chemical Engineering and
Materials Science

 

Major Professor’sﬁgnature

/2/m/OH

 

Date

MSU is an Afﬁrmative Action/Equal Opportunity Institution

PLACE IN RETURN Box to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

l

 

999.99

i
{I TU

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/CIRCIDateDuo.p65-p.15

OPTIMIZATION AND CONTROL OF RECOMBINANT PROTEIN EXPRESSION
By

Casey Preston

A THESIS
Submitted to
Michigan State University

In partial fulﬁllment of the requirements
For the degree of

MASTER OF SCIENCE
Department of Chemical Engineering and Materials Science

2004

ABSTRACT
OPTIMIZATION AND CONTROL OF RECOMBINANT PROTEIN EXPRESSION
By
Casey Preston

The skills required for the Optimization and control of protein expression in
recombinantly engineered microorganisms is of increasing value to chemical engineers
and makes them attractive candidates for positions in the biotechnology industry.
Therefore, protocols have been developed for use in the Biochemical Engineering
Teaching Laboratory. Procedures for batch fermentations in shake ﬂasks, fed-batch
fermentations in the New Brunswick BiOFlo IIC, SDS-PAGE, absorbance analysis, and
media selection are included.

Novel and robust software has been developed for use in the control of fed-batch
bioreactors. The software allows the operator of the equipment to easily develop and
implement ﬂexible, knowledge-based control schemes based on the environmental
variables monitored by the control unit. A section of the Operator’s manual and
descriptions of the controller components and concepts are presented.

A 185 residue segment of the HA2 domain of hemagglutinin has been expressed
in the E. coli host Rosetta pRARE in both batch and fed-batch fermentations.
Isotopically labeled amino acids have been selectively incorporated into the expressed
protein for the purposes of analysis by rotational-echo double resonance (REDOR). The
expression optimization procedure that has been followed utilizing the developed

protocols is outlined.

ACKNOWLEDGMENTS

I would ﬁrst like to thank past members of Dr. Worden’s research group who
supported me and taught me how to do research: Tyler Ames, Mark Widman, David
Knop, and Prashant Srivistava. I also want to thank the new member’s: Neeraj Kohli,
Kai Wang, and Brian Hassler. It has been a pleasure working with you and I wish you all
success. I also appreciate all of the help that I have been given from the faculty and
ofﬁce staff of the chemical engineering department throughout my time at Michigan State
University. Speciﬁcally, thank you JOAnn Peterson for keeping me on track to graduate
and Dr. Bruce Dale for agreeing to be on my committee.

I want to acknowledge the people who made this thesis possible. Jun Sun, I thank
you for teaching me much of what I know about recombinant engineering and guiding me
through my work. I also need to thank Jason Criscione for collaborating with me on the
project. Your help made the work much easier, quicker, and more enjoyable. I could not
have asked for a better research partner. Similarly, thank you Dr. David Weliky for being
on my committee, developing the project, and giving me so much support and
information. You all helped to make my return to graduate school worthwhile.

This work also would not have been possible without the support of my family.
Dad, Mom, Carrie, and Ian- thank you for keeping me going. I would not have ﬁnished
otherwise.

Finally, thank you Dr. Mark Worden for everything that you have done for me.
You were the major force behind my education in engineering and I would not have been

able to succeed with you. I cannot thank you enough.

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................. vi
LIST OF FIGURES ........................................................................................................... vii
1. Introduction ................................................................................................................. 1
1.1 Recombinant Engineering ................................................................................... l
1.2 Teaching and Research Opportunities ................................................................. 2
1.3 Available Resources ............................................................................................ 4
1.4 Thesis Contents ................................................................................................... 5
2. Isotopic Labeling of Hemagglutinin ........................................................................... 7
2. 1 Hemagglutinin ..................................................................................................... 7
2.2 Expression ........................................................................................................... 9
2.3 Methods ............................................................................................................. 10
3. Effect of Glucose Concentration in M9 Media ......................................................... 12
3.1 Background ....................................................................................................... 12
3.2 Results ............................................................................................................... 13
3.3 Discussion ......................................................................................................... 16
4. Catabolite Inhibition of Induction ............................................................................. 19
4. 1 Background ....................................................................................................... 19
4.2 Results ............................................................................................................... 20
4.3 Discussion ......................................................................................................... 2 1
5. Effect of Temperature and IPTG Concentration on Induction .................................. 24
5. 1 Background ....................................................................................................... 24
5.2 Results ............................................................................................................... 25
5.3 Discussion ......................................................................................................... 28
6. Expression of HA2_185 in a Fed-Batch Bioreactor .................................................. 32
6. 1 Background ....................................................................................................... 32
6.2 Results ............................................................................................................... 35
6.3 Discussion ......................................................................................................... 36
7. Presto Control ............................................................................................................ 4O
7. 1 Background ....................................................................................................... 40
7.2 Results ............................................................................................................... 41
7.3 Discussion ......................................................................................................... 42
APPENDIX A: PROTOCOLS FOR PROTEIN EXPRESSION ...................................... 46

iv

APPENDIX B: PRESTO CONTROL MANUAL .......................................................... 124

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

LIST OF TABLES

Table 1: Overall Effect Of Glucose on Growth ................................................................. 15
Table 2: Separating Gel Solutions (15 mL) .................................................................... 108
Table 3: Stacking Gel Solution (10 mL) ......................................................................... 108
Table 4: Standard Fed Batch Control .............................................................................. 136
Table 5: Reduced Fed Batch Control .............................................................................. 142
Table 6: Maximum DO Control ...................................................................................... 142
Table 7: Exponential Corrected Feed Forward Control .................................................. 143
Table 8: Minimum Feed Rate Control ............................................................................ 143
Table 9: Maximum Feed Rate Control ............................................................................ 144
Table 10: Saturation Pulse Test Control ......................................................................... 145
Table 11: Feed PID 1 & 2 ............................................................................................... 146

vi

LIST OF FIGURES

Figure 1: Effect of Glucose on Growth ............................................................................. 14
Figure 2: Effect of Glucose on Speciﬁc Growth Rate ...................................................... 15
Figure 3: Effect of Glucose on Growth Before and After Induction ................................ 20
Figure 4: Effect of Glucose on HA2_185 Expression ....................................................... 21
Figure 5: Effect of Temperature and IPTG Concentration on Growth ............................. 26
Figure 6: Effect of Temperature and IPTG on HA2_185 Expression after 3 hrs ............. 27
Figure 7: Effect of IPTG on Inclusion Body Formation ................................................... 28
Figure 8: Growth in the BiOFlo IIc .................................................................................... 35
Figure 9: Expression of HA2_185 in the BioFlo IIc ......................................................... 36
Figure 10: Presto Control Response to Substrate Depletion ............................................. 42
Figure 11: Glucose Limited Growth ............................................................................... 129
Figure 12: kLa for the BioFlo IIc .................................................................................... 131
Figure 13: Mass Transfer and Consumption in the BiOFlo IIc ....................................... 133

vii

1. Introduction

1.1 Recombinant Engineering

Proteins represent the most abundant class of organic molecules in the human
body. They function in regulating cellular responses, maintaining cell shape, transporting
molecules, initiating a body’s defensive mechanisms, and catalyzing reactions. The
human body is believed to contain 10,000 to 20,000 different proteins and over 2000 of
these are believed to be enzymes. Proteins have incredible potential as marketable
products in the biotechnology and pharmaceutical industries, yet the majority of these
proteins have yet to be studied. Techniques such as polymerase chain reaction have been
developed that allow the genes that encode these proteins to be identiﬁed, but the proteins
themselves are very difﬁcult to create. Smaller proteins can be synthesized in a step-wise
fashion, but proteins over 100 amino acids are difficult to produce synthetically.
Therefore, larger proteins are more easily created using a biological host; the anabolic
mechanisms of a cell are more proﬁcient at producing complex macromolecules than an
organic chemist. Even so, the natural abundance of a protein in a cell is generally very
small making the puriﬁcation of the protein both difﬁcult and uneconomical.

Recombinant engineering (expression of foreign DNA in host cells) is commonly
used to overcome the low yields found in native expression. In recombinant engineering,
genes that produce the protein of interest are isolated from the native host and cloned into
an expression vector. A host cell for the vector is selected that is easily grown and
controlled. The host is then transformed by the vector and utilized in a fermentation

process to produce the protein at the quantities required for study or commercial sale. In

this way, intracellular proteins can be overexpressed to such an extent that they may
make up as much 50% of the ﬁnal cell culture protein mass (Grabski et al., 2003; Laarge
and Langosh, 2001). With the proper host and bioreactor equipment, extracellular
proteins can be produced to an even greater overall concentration (Bailey and Ollis,
1986).

Recombinant engineering has become the prevalent method of producing proteins
in the biotechnology and pharmaceutical industries. Sales of protein based biotechnology
products are expected to be $32 billion by 2006 (Shuler and Michael, 2002).
Recombinantly produced pharmaceuticals such as antibodies, human growth hormone,
and insulin are used in the treatment of such diseases as cancer, arthritis, diabetes, and
viral illnesses. In addition, proteins are beginning to have applications in sensors, food
processing, and enzymatic catalysis. The biotechnology industry presents increasing

employment opportunities for chemical engineers with the proper training.

1.2 Teaching and Research Opportunities

Many of the skills and techniques that characterize the discipline of chemical
engineering are applicable to the ﬁeld of biotechnology. In particular, chemical
engineers are required to fulﬁll bioprocess puriﬁcation roles such as the development Of
chromatography operations and crystallization and tableting procedures. Another ﬁeld in
which chemical engineers can be beneﬁcial is bioreactor development. In essence,
fermentation is the same as the traditional forms of reaction engineering for which
chemical engineers are trained. Reactants (air and feed sources) and a catalyst (cells) are
added to a reactor and a value added product (protein) is created. Also, the analytical

methods of chemical engineering are essential for reliable reaction control of the 100 L to

10,000 L bioreactors found in industry. Classical reaction issues such as heat production,
heat transfer, mixing, reaction stoichiometry, and reaction rate all apply. Therefore,
chemical engineering students with a background in the basic laboratory techniques of
recombinant engineering and fermentation are attractive candidates for bioprocessing
positions.

Skills related to fermentation can also be of beneﬁt to those pursuing academic
research in other ﬁelds involving bioscience, such as proteomics and biochemistry. A
primary objective of much research in biochemistry is to understand the structure and
function of proteins in the body (Julka and Regnier, 2004; Zhu et al., 2002). When the
proteins become too large for direct synthesis, many biochemists utilize recombinant
engineering. Unfortunately, due to a lack of equipment and expertise, some researchers
attempt to produce the proteins in poorly regulated and modeled environments. Batch
fermentations in shake ﬂasks are often used even though they do not allow active control
of growth or protein production. The inability to monitor the environmental conditions
and physiological state of their expression hosts in these batch processes can lead to
many problems (Liden, 2002). Impediments to recombinant protein expression include
the stringent response of the cell due to a lack of substrate (Notley and Ferenci, 1996),
toxicity of the heterologous protein or its inducer (Glick, 1995), toxicity of fermentation
byproducts such as acetate (Konstantinov et al., 1990; Akesson et al., 1999), catabolite
repression of expression (Grossman et al., 1998), oxidative stress and protein degradation
(Konz et al., 1998), oxygen starvation (Chen et al., 1999), and inhibition of growth due to
high metabolite concentrations (Lee, 1996). Chemical engineering fundamentals such as

process control, reaction rate calculations, and modeling of the mass transport

characteristics of the system can be used to solve many of the issues facing growth and

protein production in cell cultures.

1.3 Available Resources

The Chemical Engineering and Materials Science department is especially well
suited to providing learning and research opportunities for students interested in
recombinant engineering and fermentation processes. The Biochemical Engineering
Teaching Laboratory in rooms 3269 and 3262 of the Engineering Building houses much
of the equipment required for basic experiments related to the Optimization of protein
expression in batch and fed-batch fermentations. This equipment includes 1 L and 10 L
stirred tank bioreactors, a laminar ﬂow hood, electrophoresis equipment, incubators,
glassware, and large and small centrifuges. More importantly, expertise and research
opportunities can be supplied via the Protein Expression Laboratory (PEL) and the
multiple faculty members in the chemical engineering department with projects related to
biochemical engineering. Students can potentially gain experience in bioreactor process
development through independent research projects or assigned laboratory experiments

Some of the fermentation infrastructure in the Biochemical Engineering Teaching
Laboratory is a result of the work done by David Knop (Knop, 2002). As part of his
thesis project, the control system for the 1 L bioreactors (BioFlo He) in the teaching
laboratories was substantially improved. Much of the internal controller software for the
bioreactors was bypassed and a stand alone control unit was implemented. Graphical
programming software and additional control modules were added to increase the control
capabilities. The 1 L bioreactors now have the ability to be driven by very robust control

schemes as compared to the limited control capabilities of many commercial units.

Operators can employ previously impossible control schemes for fed-batch fermentations

of recombinant microorganisms.

1.4 Thesis Contents

The work done in for this thesis is intended to help students and researchers use
the equipment in the Biochemical Engineering Teaching Laboratory. Appendix A of this
thesis contains protocols related to the optimization of protein expression in recombinant
cell cultures. These procedures were developed speciﬁcally for use with the equipment
available in the teaching laboratory. Protocols are written for batch fermentations in
shake ﬂasks and test-tubes, as well as fed-batch fermentations in the BioFlo IIC.
Protocols are also included for use Of the autoclave, the spectrophotometer, the various
centrifuges, the SDS-PAGE equipment, and for choosing and preparing media. Each
protocol contains an introduction to the procedure, a list of the required equipment, and
hints for students who have never before used the equipment or worked with recombinant
cell cultures. It is hoped that these protocols will be the basis fOr laboratory experiments
developed for use in the Biochemical Engineering Laboratory courses.

Appendix B consists of a portion of the manual that describes the use of Presto
Control. Presto Control is the control system developed for operating the BioFlO IIc
during fed-batch fermentations. It was developed in response to complaints that the
previous control system was inﬂexible and caused controller responses that were not well
understood. The previous control system was created by David Knop speciﬁcally for
producing extracellular secondary metabolites in high cell density cultures (Knop, 2002).
However, the controller’s ability to deal with the issues encountered in cultures

expressing recombinant proteins via induction was limited. In addition, the computer

interface for the controller did not make obvious to the operator the controller actions that
were being implemented. Presto Control was designed speciﬁcally to address those
issues.

The primary text of this thesis is an analysis of example experiments to
demonstrate the use of the protocols and the control software described in the
Appendices. The experiments were an attempt to Optimize the expression of a segment
of the inﬂuenza fusion peptide hemagglutinin that had been cloned onto a pET vector in
the Escherichia coli host strain Rosetta pRARE. The experiments follow the protocols in
Appendix A. Experiments were also utilized to test the ability of Presto Control to
maintain Optimal growth and expression conditions in the BioFlo He. The discussion Of

the results helps to illustrate the features of the software.

2. Isotopic Labeling of Hemagglutinin

2.1 Hemagglutinin

Cellular infection by viral DNA results from the fusion of the protective
membranes of the virus and the cell. The union of the lipid bilayers of the virus and the
cell is facilitated by a class of proteins generally identiﬁed as “fusion glycoproteins”
(Hernandez et al., 1996). A hydrophobic domain of the fusion glycoprotein called the
fusion peptide embeds itself in the hydrophobic lipid layer of the cellular membrane. A
conformational change in the fusion glycoprotein then causes the viral membrane to fuse
with the cell membrane and viral DNA to be released into the cytoplasmic space in the
cell. The structure and conformation changes of the fusion glycoprotein are of interest to
researchers because of their relationship to infection and disease. A full understanding
virus-cell fusion may one day lead to treatments for viral illnesses.

Hemagglutinin (HA) is the fusion protein characteristic of inﬂuenza. It is a
homotn'meric glycoprotein that induces fusion with an endosomal membrane in the cell
when the local pH drops to 5.0 (Kim et al., 1998). It is the most widely studied of the
fusion glycoprotein (Wilson et al., 1981; Wiley and Skelhel, 1987). Hemagglutinin is
divided into two subunits, HA1 and HA2, by proteolytic cleavage after expression in the
infected cell. HA2 is of particular interest because it houses the fusion peptide and
exhibits massive, pH triggered, conformational changes (Kim et al., 1996). It is the
subunit that is primarily responsible for the actual fusion event (Leikina et al., 2001;

LeDuc et al., 2000; Epand et a1. 1999).

Recently, speciﬁc sections of HA2 have been recombinantly expressed in E. coli
and studied by spin-labeling electron paramagnetic resonance (EPS) techniques
(Macosko at al., 1997; Kim et al., 1998; Kim et al., 1996) and atomic force microscopy
(AFM) (Epand et al., 2001). Dr. David Weliky of the Department Of Chemistry at
Michigan State University believed that rotational-echo double resonance (REDOR)
analysis of HA2 would provide further insight into the conformational changes of the
protein (Yang et al., 2002). REDOR is a nuclear magnetic resonance (NMR) technique
that utilizes the dipolar coupling between adjacent, isotopically labeled elements to ﬁlter
out the magic angle spinning (MAS) signal of the labeled elements from the natural
abundance signals. This allows the local conformation of the two isotopically labeled
elements to be determined. REDOR has been successfully used to determine the
conformation of lipid bound proteins in solution (Murphy et al., 2001; Wang et al., 1997).
Its use in studying HA2 would allow a better understanding of the structure of fused HA2
before and after its pH induced conformation change.

One of the most efﬁcient means Of producing hemagglutinin is through the use of
recombinant engineering techniques. A DNA construct of a segment from the N terminal
end of HA2 was obtained from Yeon-Kyun Shin, a professor in the Department of
Chemistry at the University of California, Berkeley. The construct encodes for the ﬁrst
185 residues (HA2_185) of the peptide and it contains the 127 amino acid section studied
by his research group (Kim et al., 1996). HA2_185 is characterized by a 25 amino acid
fusion peptide at the N terminus and a longer chain of more hydrophilic residues at the C-
terminus. The hydrophobic character of the fusion peptide makes HA2_185 both

difﬁcult to express and utilize in the soluble, active form required for REDOR analysis.

The protein can only remain in its native conformation during expression by interacting
with the internal lipid membranes of the host cell (Kim et al., 1996). Therefore, the
fusion peptide has a tendency to aggregate and form inclusion bodies at high expression
levels and limited membrane surface areas. Inclusion bodies are dense, intracellular
aggregates of denatured protein that often characterize heterologous protein expression in
prokaryotes. In order to overcome these issues and produce the protein in the quantities

required for REDOR analysis, Optimization of the expression system was attempted.

2.2 Expression

For expression, the HA2_185 gene was cloned into the pET-24b(+) vector and
used in the transformation of the E. coli host Rosetta pRARE (Novagen). The utilization
of this expression system immediately placed certain constraints on media selection.
First, the pET vector and Rosetta act in concert to maintain stringent control of the
induced protein by way of a T7 lac promoter (Novagen, 2002). Therefore induction of
protein expression occurs upon the introduction of isopropyl B—D,thiogalactopyranoside
(IPTG) to the media. Kanamycin at a recommended concentration of 15 ug/mL is
required in the fermentation media to prevent plasmid loss. Rosetta also contains the
plasmid pRARE which expresses eukaryotic speciﬁc codons not normally produced in E.
coli host strains. This plasmid is maintained by chloramphenicol at a suggested level of
34 ug/mL.

Carbon and nitrogen sources for the media were chosen based on the needs Of the
REDOR experiments. REDOR analysis requires that the peptide bond between speciﬁc
pairs of amino acids be labeled with '3 C and 15N atoms. The procedure for isotopically

labeling amino acids in recombinantly expressed proteins is well documented (Muchmore

et al., 1989; Lian and Middleton, 2001; Goto and Kay, 2000). Generally, the labeled
amino acids are added directly before induction of the protein of interest. The host strain
preferentially incorporates the labeled residues into the protein rather than building the
amino acid from the available nitrogen and carbon sources in the media. Therefore,
complex media containing protein cannot be used because it contains unspeciﬁed
amounts of unlabeled amino acids. Instead, a deﬁned media must be employed. M9 was
chosen for the experiments described in this work due to its simplicity in preparation and
cost effectiveness. Glucose was utilized as the substrate because it is the preferred
carbon source for the host strain. As long as glucose was present during induction, the
culture was unlikely to break down the added isotopically labeled amino acids for
anabolic and catabolic purposes. Metabolism of the labeled amino acids could lead to
incorporation of the labeled atoms into other, non-speciﬁc amino acids. The resulting

isotopic dilution might add noise or diminish the REDOR signal.

2.3 Methods

All fermentation experiments followed the protocols described in Appendix A.
Times and quantities suggested in Appendix A were the values used unless otherwise
stated. M9 media was used for growth and it was prepared as described in the MEDIA
section of Appendix A. Chloramphenicol and kanamycin were added to each volume of
fresh media to obtain 20 mg/L and 50 mg/L respectively. Initial colonies of the
expression host were obtained from a single glycerol freeze source that had been
inoculated onto M9 plates containing 4 g/L glucose. When the OD600 reached a value
between 1.0 and 2.0, suspension cultures were transferred to new media at a dilution ratio

ranging from 5% to 10%. Fed—batch fermentations were carried out in the BioFlo 11c

10

(New Brunswick) following the set-up and procedure outlined in the FED-BATCH
FERMENTATIONS section of Appendix A. During the fermentation, glucose was fed at
a concentration of 50% w/v utilizing Presto Control and the Standard Fed Batch control
scheme described in Appendix B.

Analytical techniques also followed the protocols in Appendix A. The OD600 of
the cultures was measured and analyzed for the speciﬁc growth rate (SGR) as described
in the CELL DENSITY- OD600 section of Appendix A. SDS-PAGE analysis of cell
samples was undertaken as described in the SDS-PAGE section using the FB-VElO-l 10
mL Fischer vertical electrophoresis stand (Fischer Scientiﬁc). All gels were poured with
a 12% acrylimide concentration in the separating gel. When phosphate buffer was used
in the cell fractionation procedure, 0.5% n-lauroyl sarkosine was added to aid in
solubilization of the membrane and HA2_185. Detergent had to be added to any

preparation of the puriﬁed protein in order for it to remain in solution

11

3. Effect of Glucose Concentration in M9 Media

3.1 Background

Experiment were conducted to determine the culture’s speciﬁc growth rate (SGR)
and growth limiting nutrient in M9 media. It had been assumed that the maximum cell
density obtainable in batch growth in M9 would be dependent on either the glucose or the
ammonium chloride concentration. Glucose is the only source of carbon in M9 and its
depletion leads to a massive reduction in catabolic and anabolic activities of the cell
culture. The presence of glucose in the media, however, can also cause growth inhibition
as a result of the formation of byproducts related to anaerobic metabolism of the substrate
(Lee, 1996). For example, acetic acid, a fermentation byproduct produced at high
glucose concentrations, can cause inhibitory effects in E. coli at concentrations as low as
2 g/L (Konz et al., 1998). In bioreactors with pH control, acetic acid can also indirectly
cause growth inhibition as a result of the salt formed via the base that is added to
maintain the pH.

The nitrogen source in M9 media is ammonium chloride and its depletion also
leads to growth inhibition. The molar ratio of carbon to nitrogen in stande M9 media is
6.2. It is on the same order of magnitude as the ratio of carbon to nitrogen in the cell, 4.0.
These calculations assume a concentration of 1.0 g/L ammonium chloride, 4.0 g/L
glucose, and a general cellular composition of CH1,gOo,5No_2 (Bailey and Ollis, 1988).
Therefore, either component could be the growth limiting nutrient depending on the

yields of the substrate.

12

Batch growth experiments were conducted with three 50 mL Rosetta cultures in
250 Erlenmeyer ﬂasks in order to determine the effect of increased glucose concentration
on cell growth. The initial concentration of glucose in the media was either 5 g/L, 10
g/L, or 20 g/L. The ammonium chloride concentration for each ﬂask was increased to 5.0
g/L over the 1.0 g/L suggested for standard M9. This was done to reduce the chance of
nitrogen becoming the growth limiting nutrient at the relatively high glucose
concentrations utilized in this experiment. An ammonium chloride concentration of 5
g/L is just below the concentration seen to inhibit growth in E. coli (Lee, 1996). The
optical density (OD600) was monitored and the pH of the cultures were checked when
they had reached stationary phase. Then, additional M9 buffer and salts were added to

the media to discover whether it would spur growth.

3.2 Results

Figure 1 shows the OD600 of the culture at various glucose concentrations. Figure
2 exempliﬁes the method used to determine the speciﬁc growth rate (SGR) of the culture
as described in the CELL DENSITY - OD600 section of Appendix A. Lines were ﬁtted
using least squares analysis with the y-axis deﬁned as the natural log of the ratio of A and
A0, where A is the OD600 at the time speciﬁed on the x-axis and A0 is the OD600 at time
equals zero. The slope of the line is the SGR and it is listed next to the legend entry for
Figure 2. Table 1 describes the overall growth of the culture at various initial glucose

concentrations. It includes the pH and OD600 of the culture 20 hrs after inoculation.

13

Figure 1: Effect of Glucose on Growth

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

3
/l-* -
2.5 1* :1
2
o
3
n 1 5
O
1 + 5 g/l. glucose
-I- 10 g/L glucose
+ 20 g/L glucose
0.5
O . i .
O 10 15

Time (hr)

14

20

Figure 2: Effect of Glucose on Speciﬁc Growth Rate

 

3.5

3

 

 

 

 

 

 

   
 

 

 

‘ O 5 g/L initial glucose slope (SGR)-‘- 0-40
I 10 g/L initial glucose slope (SGR): 0.40
A 20 g/L initial glucose slope (SGR): 0,38

 

 

 

 

 

 

4 6 8 10
Time (hr)

Table 1: Overall Effect of Glucose on Growth

 

 

 

 

 

Initial Glucose Concentration SGR Final OD600 Final pH
5 g/L 0.40 m" 2.472 6.21
10 g/L 0.40 hr'r 2.685 4.21
20 g/L 0.38 hr" 2.325 4.28

 

 

 

 

 

After the culture had reached stationary phase at 20 hours, 10 mL of M9 buffer
and salts were added to the 34 mL of broth still left in each shake ﬂask. The addition of
the buffer caused the pH of the cultures to rise. The additional M9 buffer had little effect
on the growth of the 5 g/L culture. The cell density of the 10 g/L and 20 g/L glucose
cultures, however, increased slightly over the next 2 hours until the pH again dropped to

a level below 5.0 (results not shown).

15

3.3 Discussion

The results shown in Figure 1 and Figure 2 suggest that the growth limiting factor
in these batch cultures was the buffering capacity of the media. The low pH that resulted
in the high glucose concentration cultures was far outside the optimum range of 6.5 to 7.5
for cell growth (Bailey and Ollis, 1986), and may have triggered the onset of stationary
phase. Increasing the pH by adding buffer allowed additional cell growth to occur. This
new growth stopped when the pH once again dropped below the threshold value of 5.0.
Therefore, nutrient limitations did not appear to be a growth limiting factor in these
cultures.

The 5 g/L glucose culture, however, did not respond to the addition of M9 buffer.
Since the primary nutrient in the buffer is ammonium chloride, it can be assumed that
nitrogen had not become depleted in this culture. Also, the pH was never low enough in
this culture either before or after the addition of the buffer to induce a stationary phase.
Instead, the culture had apparently consumed its entire carbon source by the time it
reached its ﬁnal OD600 of 2.5.

The speciﬁc growth rate and ﬁnal cell concentration of the culture were not
noticeably affected by the concentration of glucose in the media. The glucose
concentration itself was not expected to lead to SGR limitations since substrate inhibition
is normally only found at glucose concentrations greater than 50 g/L (Lee, 1996). Acetic
acid formation at high glucose concentrations, however, is a common problem in
culturing E. coli. This byproduct is produced when the glucose concentration is high
enough to cause its uptake rate to exceed the oxygen uptake rate required for full

catabolic oxidation. A characteristic of this process is a relatively rapid growth of the

culture. In E. coli, acetate is generally only produced at an SGR above 0.35 hr'l (van der
Walle and Shiloach, 1998; Lee, 1996). This SGR is similar to the growth rate found here.
Rosetta, however, is a derivative of the B strain and this host tends not to produce acetic
acid at toxic levels irrespective of the glucose concentration (Shiloach et al., 1996).

At the low cell densities allowed by shake ﬂasks, the growth of Rosetta is not
noticeably inhibited by nutrient depletion. Instead, the growth limiting factor in batch
fermentations is believed to be the buffering capacity of the media. In M9 media, the
utilization of ammonium chloride as a nitrogen source causes the release of hydrogen
ions into the media (Liu et al., 2001). Over time, these hydrogen ions accumulate and
overcome the buffering capacity of the media, thereby causing a drop in the pH.
Contributions to the hydrogen ion concentration by acidic fermentation byproducts such
as acetic acid are also a possibility (Jaradat and Bhunia, 2002; Lee, 1996). These
experiments show that the initial glucose charge speciﬁed in the recipe for M9 becomes
completely consumed at about the same time that the buffering capacity of the media is
lost. To combat the pH drop, base could be added that would moderate the pH and thus
allow growth to higher potential cell densities. This type of fed-batch fermentation is
much easier in a bioreactor with automated pH control, though. Furthermore, oxygen
transfer limitations in a shake ﬂask could begin to become an issue at high cell densities.
Therefore, high cell density culture growth requires the use of specialized equipment
such as the BioFlo IIc.

When studying protein expression in shake ﬂask experiments, a steady state
glucose concentration does not necessarily need to be maintained because the

concentration of the substrate does not have a noticeable effect on the SGR. If it did have

an effect, then any testing of expression levels would need to take glucose concentration
into account. The glucose levels would have to remain consistent in order to prove that a
change in the SGR during induction would be due to protein expression and not to a
change in the glucose concentration. This is not an issue in these optimization
experiments. Instead, an E. coli culture in M9 should be induced at a cell density near
1.0 in order to maintain physiological pH.

This experiment was unable to test whether the accumulation of byproducts such
as acetic acid could affect cell growth in high cell density fed batch fermentations. The
relatively short amount of time allowed for culture growth in shake ﬂasks does not allow
inhibitory concentrations of byproducts to be produced. It is unclear whether the drop in
pH in the high glucose concentration cultures resulted from either ammonium uptake,
acetic acid production, or both. The creation of acetic acid is a possibility, however,
since the speciﬁc growth rate calculated in these experiments is relatively high. A shake
ﬂask experiment was attempted in which small quantities of glucose were added at
regular time intervals in order to maintain a low average glucose concentration. Even in
these conditions, the pH dropped to inhibitory levels by the time an OD600 of 2.5 was
reached (results not shown). Therefore, experiments related to glucose concentration and

acid formation are best left to fed-batch experiments in the bioreactor.

4. Catabolite Inhibition of Induction

4.1 Background

The lacUV5 promoter employed to express recombinant DNA in the Rosetta/pET
system is based on a lac promoter. In the presence of lactose or the non-metabolizable
analog IPTG, the wild type lac promoter initiates the production of galactosidase.
Galactosidase is an enzyme required by the cell for metabolism of lactose and galactose.
The wild type lac promoter also requires the presence of CAMP before it allows bonding
of polymerase. The culture only produces CAMP when glucose is below 1%w/v as a
signal that substrates other than the preferred carbon source of glucose should be
catabolized (Hirschel et al., 1980). The lacUV5 promoter includes three point mutations
to the wild type lac promoter that are intended to reduce its dependence on CAMP for
induction. Previous researchers have found that lacUV5 does not experience catabolite
inhibition in the presence of glucose (Ochocka et al., 2003; Hirschel et al., 1980), but
glucose has been found to have an effect on basal level expression (Grossman et al.,
1998; Novy and Morris, 2001). Therefore, the effect of glucose on expression of
HA2_185 in Rosetta was tested.

The host strain was grown in 50 mL cultures in M9 media with initial
concentrations of either 2 g/L or 20 g/L glucose. Seven hours after inoculation, 5 mL of
each culture were collected for use as the uninduced, control samples and transferred to
10 mL culture tubes. IPTG was added to the remaining culture in each shake ﬂask to
yield a ﬁnal inducer concentration of 1 mM. Multiple samples of the induced and

uninduced cultures were taken at 2 hrs after induction. A portion of the samples were

19

lysed and separated into soluble and insoluble fractions and analyzed by SDS-PAGE
according to the protocol in Appendix A. The rest of the samples were retained as total
cell protein (TCP) fractions. The T CP fraction is a solubilization of the entire culture

pellet without any puriﬁcation

4.2 Results

The growth curves for both the induced and uninduced cell cultures are shown in
Figure 3. The SGR for the uninduced cultures were calculated to be 0.36 hr'l for the
culture with 2 g/L initial glucose and 0.34 hr'l for the culture with 20 g/L initial glucose
concentration. The induced cultures showed signiﬁcant growth inhibition during protein
expression.

Figure 3: Effect of Glucose on Growth Before and After Induction

 

 

 

 

 

 

 

 

 

 

 

 

 

2.5
+ 2 g/L hitial Glucose
. . Uninduced 9
+ 20 g/L hitial Glucose Cultures {-
2 ~ "-0-3-2 g/L uninduced
- - -I - -- 2O yL uninduced
1 5 I
a ' "
co . .-
o .'
O 1 ,' 'ﬂu-ced
Cultures
Time of
hduction
O 5
. J
O 9 i i .
0 2 4 6 8 10

Time (hr)

20

Figure 4 exhibits the results of the SDS-PAGE analysis of the culture samples.
Lane 9 contained the TCP fraction of a sample taken at the time of induction. Lanes 1
through 8 contained the protein from the culture samples taken 2 hrs after the time of
induction. Speciﬁcally, lanes 5 through 8 contained the TCP fractions, lanes 1 through 2
contained the soluble fractions, and lanes 3 through 4 contained the insoluble fractions of

the induced culture samples.

Figure 4: Effect of Glucose on HA2_185 Expression
1 2 .3. 4 5
. ’ MW ‘ ‘~

l0)
loo
Ito

 

4.3 Discussion

The SGR was again found to be independent of the glucose concentration. Also,

the average SGR calculated in these experiments, 0.35 hr", was similar to the SGR of

21

0.39 hr'l found in the experiment previously discussed. Induction, however, had a very
signiﬁcant effect on the growth rate. Two hours after induction, the OD600 of the
uninduced cultures had increased by approximately three times the magnitude of the
increase in the OD600 of the induced cultures. This could have been a result of product
inhibition, toxicity of the IPTG, or the diversion of cellular resources to the production of
HA2_185.

The SDS-PAGE analysis shown in Figure 4 indicated that the Rosetta/pET
expression system had the ability to express HA2_185 at high enough concentrations to
satisfy the mass and puriﬁcation requirements of the project. Lanes 1 through 6 of Figure
4 had 23 kDa bands that were missing in the uninduced fraction. HA2_185 has a weight
close to 23 kDa. The similarity of the intensity of the HA2_185 bands in lanes 5 and 6
indicated that the difference in glucose concentration between the two cultures had not
resulted in a variation in expression levels.

Lanes 1 through 4 shown in Figure 4 compared the insoluble and soluble
fractions of the TCP samples ran in lanes 5 and 6. The insoluble fractions contained the
dense inclusion bodies formed by the HA2_185, while the soluble fractions contained the
HA2_185 that had not aggregated into inclusion bodies during expression and lysis. This
HA2_185 fraction was assumed to primarily consist of membrane bound and detergent
solubilized HA2_185 in its native conformation. For the purposes of this thesis, the
fraction of expressed protein that did not form inclusion bodies was designated “soluble
HA2_185”. The heavy bands in lanes 3 and 4 suggested that the induction conditions of
37 °C and 1 mM IPTG utilized in this experiment resulted in the formation of inclusion

bodies of HA2_185. In addition, the bands were more intense than those displayed in

22

lanes 1 and 2. This suggested that less protein remained in the soluble than in the
insoluble form after two hours of induction at these conditions. The visual appearance of
the gel suggested that a higher portion of the HA2_185 was found in the insoluble
fraction of the culture sample at 20 g/L glucose than in the 2 g/L sample. The total
amount of all the proteins in the 20 g/L insoluble fraction was also greater, however.
Correspondingly, the 20 g/L soluble fraction had a smaller total amount of protein. It
was likely that the separation of the soluble fraction from the insoluble fraction during
preparation of the 20 g/L sample was incomplete. This would account for the appearance
of the higher mass of inclusion bodies in the 20 g/L culture sample.

An inducer concentration of 1 mM IPTG is recommended for maximum
expression levels in the pET vector (Novagen, 2002). The T7 promoter used in the pET
expression system is very strong. Upon full induction, the pET vector is designed to set
all of the anabolic activity of the cell towards production of the recombinant protein. As
shOwn in Figure 3, this causes a signiﬁcant reduction in cell growth. In addition, the high
induction activity causes large levels of inclusion body formation. This suggests that the
induction conditions used in this experiment should be modiﬁed if increases in the

overall yield of the soluble fraction versus the insoluble fraction are desired.

23

5. Effect of Temperature and IPTG Concentration on Induction

5.1 Background

As discussed in section 4, the Rosetta/pET expression system tends to form
inclusion bodies at high induction strength. In recombinant protein expression, it is often
desirable to have control of the yield of soluble protein. A reduction in inclusion body
formation can be accomplished by decreasing the induction levels during expression
(Laage and Langosch, 2001; Novagen, 2002). Lower induction levels act to reduce the
rate of protein formation and lower the concentration of unbound protein in the
cytoplasm. This reduced protein concentration decreases the interaction between the
hydrophobic regions of the induced proteins and lowers the probability of aggregation.

Induction levels in Rosetta can be reduced by lowering the IPTG concentration.
Wild type E. coli utilize lac permease for active uptake of lactose analogs (Garrett and
Grisharn, 1999). Rosetta, however, has a mutation that inactivates the lach gene that
encodes for this transport protein (Novagen, 2002). Therefore, IPTG enters the cell via
passive diffusion that is dependent on the concentration of IPTG in the media. Since the
activity of the lac promoter is dependent on the local IPT G concentration, the induction
level can be controlled by varying the amount of IPTG added to the fermentation broth.

Induction levels are also dependent on temperature because protein is formed by
enzymatically catalyzed reactions. Enzymatic activity decreases with decreasing
temperature. Therefore, if the temperature in the bioreactor is reduced, the rate of protein

formation is also reduced.

24

An experiment was conducted to determine the effect of temperature and IPTG
concentrations on cell growth, protein expression, and inclusion body formation. A
single test tube culture of Rosetta was used to inoculate 300 mL of M9 media containing
4 g/L of glucose and the required antibiotics as described in Appendix A. The culture
was incubated in a 1 L Erlenmeyer ﬂask until an OD600 of 1.0 was reached at 10 hours. It
was then split into six 50 mL fractions in 250 mL shake ﬂasks. IPTG was added to the
ﬂasks to produce duplicate cultures with inducer concentrations of 0.5, 0.05, or 0 mM.
Half of the duplicate shake ﬂask cultures were then transferred back to the 37 °C
incubator. The other three ﬂasks were kept at room temperature (~25 °C). Samples were

collected over ﬁve hours and analyzed by SDS-PAGE.

5.2 Results

Figure 5 shows the effect of the IPTG concentration and the temperature on the

cell densities of the cultures after induction.

25

Figure 5: Effect of Temperature and IPTG Concentration on Growth
3.5

 

+Room T, 0.5 mM IPTG
+Room T, 0.05 mM IPTG
+Room T, 0 mM IPTG
--->1(-- 37 c, 0.5 mM IPTG ........... + ----------- 49
3 1 --O--37 0.0.05 liPTG .T
--+--37C,0mM|PTG "

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2.5
+' /
a ’ .
8 2 , f0 ----- .._. ...... " '-
o ......
0".
1.5
1
0.5 I I I 7
1o 11 12 13 14 15

Time (hr)
Figure 6 shows the results from the SDS-PAGE analysis of the soluble and

insoluble fractions of the induced cell culture 3 hours after induction. Figure 7 shows the

insoluble fractions from the induced cultures grown at room temperature.

26

Figure 6: Effect of Temperature and IPT G on HA2_185 Expression after 3 hrs

 

 

 

 

 

 

 

 

 

L09! 1 2 a 4 5 5 z a a

50+ .

40—»

V ,- at; h’. J» 3.5"

30—»

20* HA2_185

15—»

Lane IPTG Fraction Temp. Lane IPTG Fraction Temp.

1 0 mM soluble room 6 0.05 mM insoluble room
2 0.05 mM soluble room 7 0.5 mM insoluble room
3 0.5 mM soluble room 8 0.05 mM insoluble 37 °C
4 0.05 mM soluble 37 °C 9 0.5 mM insoluble 37 °C
5 0.5 mM soluble 37 °C

 

 

 

 

 

 

27

 

Figure 7: Effect of IPTG on Inclusion Body Formation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

to: 1 2 a. .4 5 a z a a a; 11
50+]
40-»
a 29.4.999- ., m, 3,0 2.. : 1 ,an _
- ‘ H 2_185
15—» M '
Lane Time IPT G Lane Time IPTG Lane Time IPTG
1 0hr 0.5mM 5 3hr 0.5mM 9 2hr 0.05mM
2 0.5 hr 0.5 mM 6 5 hr 0.5 mM 10 1 hr 0.05 mM
3 1 hr 0.5 mM 7 5 hr 0.05 mM 11 0.5 hr 0.05 mM
4 2 hr 0.5 mM 8 3 hr 0.05 mM
5.3 Discussion

As was expected, the uninduced culture grew faster at 37 °C than at room

temperature (~25 °C) (Shuler and Kargi, 2002). Induction also had a very signiﬁcant

effect on the growth rate and the extent of the growth inhibition during induction was

dependent on the temperature. The OD600 of the cultures induced with an IPTG

concentration of 0.5 mM IPTG increased very little after induction. The culture at 37 °C

28

 

and 0.5 mM IPTG entered a stationary phase within 1 hour of induction while the cell
density of the culture at room temperature continued to increase slightly. At 37 °C, the
growth rate of the culture with 0.05 mM IPTG was substantially inhibited in comparison
to the uninduced culture. Up until 3 hrs after induction, both of the cultures at 37 °C
increased cell mass at a greater rate than the cultures at room temperature. After 3 hrs of
induction, the cultures at 37 °C entered into a stationary phase that probably resulted from
nutrient depletion. The cultures at room temperature continued to grow until the end of
the 5 hour induction period. Also, induction at 0.05 mM IPTG did not appear to have
inhibitory affects on the cellular growth rate at room temperature.

The intensity of the HA2_185 band was not noticeably different in lanes 2
through 5 as shown in Figure 6. All of the cultures had produced about the same amount
of soluble HA2_185 after 3 hrs of induction. In fact, a comparison of the soluble
fractions from all of the time points showed that the quantity of soluble HA2_185
displayed in Figure 6 was the maximum attainable quantity in, the culture in these
experiments (results not shown). Within 3 hours of induction, all of the cultures had
reached a maximum limit on the amount of soluble HA2_185 that they could produce and
this maximum quantity was the same irrespective of induction conditions. The cultures
with higher induction strengths reached their soluble HA2_185 limits within an hour,
while the cultures with lower inducer concentrations took longer. A difference, though,
could be found in the total mass of inclusion bodies formed. As can be seen in lane 11 of
Figure 6, the culture with the highest induction strength, 37 °C and 0.5 mM IPT G,
produced signiﬁcantly more insoluble HA2_185 than the others. The samples ran in

lanes 8 though 10 and all exhibited similar HA2_185 band intensities. The culture

29

induced at room temperature and 0.5 mM IPTG exhibited slightly less insoluble protein,
but this was attributed to its lower cell density relative to the other two samples.

The gel in Figure 7 was run in order to determine the time rate of change of
inclusion body formation. Lanes 1 through 6 showed that inclusion bodies continued to
increase with time in the culture at 0.5 mM IPT G and room temperature. A signiﬁcant
mass of insoluble protein was present within half an hour of induction. In comparison,
inclusion bodies did not appear in the culture subjected to a lower concentration of
inducer until 2 hours after induction. Its overall production of HA2_185 after ﬁve hours
of induction, however, had far exceeded that created by the culture at 0.5 mM IPT G.
This is a result of the higher cell density obtained in the culture at 0.05 mM IPTG.

The induction process has a toxic effect on cell growth that is relevant to the
optimization of overall protein yield. If recombinant protein expression causes a large
enough decrease in cell growth or cellular metabolism, then the overall protein yield may
be very low. This is exempliﬁed by a comparison of the ﬁnal HA2_185 mass obtained in
the cultures displayed in Figure 7. While the culture at 0.5 mM IPT G initially had a
greater rate of HA2_185 production, the culture at 0.05 mM IPTG ultimately produced a
higher mass of HA2_185 because it did not enter into a stationary phase as did the other.
Optimally, inducer activity should be high enough to direct the majority of the cellular
anabolic activity towards recombinant protein production. The induction level, however,
should not be so high that it causes a stress response in the cells and forces the culture
into a stationary phase. In some cases, the stress response resulting from recombinant
protein expression has been shown to become more pronounced at higher temperatures

(Hunke and Betton, 2003). This effect was seen in these experiments and suggests that

30

both the temperature and inducer strength should be lowered to in order to maintain cell
growth.

The total amount of soluble HA2_185 that is obtained per mass of culture is
independent of the induction strategy. This suggests that the amount of HA2_185 that
can remain in a soluble state in a Rosetta cell is limited by the membrane surface area
with which the protein can interact. During expression, the hydrophobic fusion peptide
of HA2_185 must become incorporated into the hydrophobic portion of a lipid bilayer in
order for it to remain soluble. Otherwise, interactions between the peptides causes them
to aggregate and create inclusion bodies. A saturation limit is reached after a speciﬁc
quantity of the peptide has become embedded in the membrane. When this solubility
constraint is attained, additional expressed fusion protein must necessarily form inclusion
bodies.

Under the above hypothesis, the yield of soluble protein can be maximized by
developing an induction strategy that results in the formation of the maximum amount of
allowable soluble protein while minimizing inclusion body formation. For the cultures
induced at lower temperatures in this experiment, inclusion bodies only began to form
after the saturation limit for the soluble protein was reached. Kim et a1. (1996) found the
maximum incorporation level of the fusion protein to be 50% of the total protein in the
membrane. In the experiment, the culture subjected to the lower inducer concentration
did not form signiﬁcant quantities of inclusion bodies until much later in the induction
phase. It also did not experience the growth inhibition displayed by the cultures at higher
induction levels. Therefore, recommended conditions for production of soluble

HA2_185 in E. coli include lower temperatures and reduced IPT G concentrations.

31

6. Expression of HA2_185 in a Fed-Batch Bioreactor

6.1 Background

Based on the results of the batch growth and expression experiments, an induction
scheme was devised to produce isotopically labeled HA2_185 in a fed-batch bioreactor.
Optimal results were based on four criteria:

1) High overall yield of active HA2_185 per mass of cell protein

2) Maximum incorporation of the labeled amino acid

3) Minimal isotopic dilution

4) Minimal cost

HA2_185 needs to be produced in high overall concentrations due to the requirements
of the puriﬁcation process. The immobilized metal afﬁnity chromatography (IMAC)
operation used to purify HA2_185 preferentially binds the heterologous protein relative
to the native constituents of the cell. If the majority of the afﬁnity sites are occupied by
HA2_185 after the cell lysate and column resin are equilibrated, then the other proteins
can be easily washed from the chromatography column and the HA2_185 can be eluted at
a high purity. If, however, the HA2_185 is at a low overall concentration compared to
the contaminant proteins, equilibrium conditions favor that at least some contaminant
protein binds to the column. This reduces the effectiveness of the chromatography
puriﬁcation.

The puriﬁed protein must also be in its native, solubilized state or the NMR
results are not indicative of the active structure of the protein. An attempt was made to

solubilize the HA2_185 inclusion bodies using the denaturant urea. The unfolded protein

32

was then allowed to renature by removing the urea from the buffer solution in the
presence of detergent. Unfortunately, the solubilized protein was only partially active in
producing fusion events at low pH (results not shown). This suggested that a portion of
the protein remained in a denatured state after solubilization. Due to time constraints,
more complex renaturation procedures were not attempted.

Therefore, only the soluble fractions of HA2_185 are suitable for use in NMR
experiments. The shake ﬂask experiments suggested that the maximum overall
concentration of soluble HA2_185 is dependent on the saturation limitations of the lipid
bilayer of the cell. After the maximum amount of soluble protein is produced, the rest
enters into inclusion bodies. Since inclusion bodies cannot be used for NMR, there
production is a waste of cellular capacity and should be minimized. This is especime
true considering the cost of the labeled amino acid used to form them.

The issue of isotopic dilution can be handled in multiple ways (Muchmore et a1,
1989). The most effective solution is to utilize an auxotrophic host strain that can neither
break down nor produce speciﬁc amino acids. The expression system for this project had
already been chosen, however. Another common method of reducing isotopic dilution is
to use a deﬁned media that includes a full complement of all of the amino acids required
by the cell. If the culture does not need to produce amino acids, then the transamination
reactions required for amino acid production are not active and isotopic dilution is
unlikely. This type of deﬁned media is very expensive, however, so this approach
conﬂicted with the desire to minimize cost. Therefore, M9 medium was used, and the
labeled amino acid was added at induction. This labeling procedure has worked

successfully for other researchers (Verdemato et al., 2000)

33

Practical steps can be taken to decrease the likelihood that the labeled amino acid
are metabolized by the cell. First, as glucose is the preferred substrate, it can be kept at
concentrations above 1 g/L for the duration of the fermentation, which would exceed the
uptake saturation level for the sugar (Levisauskas, 2003). Secondly, very high induction
levels can cause a stress response in the cell that leads to the production of proteolytic
enzymes (Medina et al., 2002; Schweder et al., 2002). This may cause unnecessary
degradation of HA2_185 and increase the possibility of isotopic dilution. Therefore,
lower levels of expression are preferred.

Based on the rationalization given above, a strategy for HA2_185 expression in
fed-batch fermentations was developed. The set-up and operation of the BioFlo IIC were
as described in the FED-BATCH section of Appendix A. M9 medium with an initial
concentration of 4 g/L glucose, 20 mg/L chlorimphenicol and 50 mg/L kanamycin was
employed as the growth media. The bioreactor was inoculated with 45 mL of Rosetta at
an OD600 of 1.8. Glucose was fed manually throughout the fermentation at a rate
required to keep the substrate from becoming depleted. A total of 18.5 g glucose was
added by the end of the fermentation. The BioFlo He was set to control temperature at 37
°C, and air ﬂow was maintained at 0.1 Umin throughout the exponential growth phase of
the culture. Agitation was controlled to maintain a dissolved oxygen concentration of
30%. Ammonium hydroxide was added as needed to maintain the pH at 7.0. When the
culture reached an approximate OD600 of 7 at 12.75 hrs after inoculation, the temperature
in the bioreactor was lowered to 26 °C. The culture and BioFlo IIc controller were given
15 minutes to adjust to the change in temperature. Then, 200 mg of leucine and 0.2

mmols of IPTG were added to the l L of fermentation broth. The induction phase was

34

allowed to proceed for two hours during which culture samples were taken every half
hour. The samples were fractionated into soluble and insoluble fractions and analyzed by
SDS-PAGE. At the end of the induction period, the culture was harvested so that the

HA2_185 could be puriﬁed and analyzed by NMR.

6.2 Results

Figure 8 shows the OD600 of the fermentation culture in the BioFlo IIc prior to induction.
The calculated SGR for the culture was 0.40 hr". The OD600 of the culture at the
conclusion of the induction phase was 8.6. Figure 9 is the SDS-PAGE analysis of the
soluble and insoluble fractions over the two hour induction period.

Figure 8: Growth in the BioFlo IIc

 

 

 

 

 

 

 

 

 

 

ODGOO
o-sroootsmoaxroo

I I I I I I I I I I I I

01234567891011121314
Time(hr)

 

 

35

Figure 9: Expression of HA2_185 in the BioFlo Ilc

 

no; 1 2 a 4. 5 5. z a a

64+

38+

26+

19+,

1
15+
HA2_18

6 +
Lane Time Fraction Lan Time Fraction
1 0hr soluble 6 0.5 hr insoluble
2 0.5 hr soluble 7 1 hr insoluble
3 1hr soluble 8 1.5 hr insoluble
4 1.5 hr soluble 9 2 hr insoluble
5 2 hr soluble

 

 

 

 

 

 

 

 

 

 

 

 

 

6.3 Discussion

Growth and expression in the bioreactor closely mimicked the results found in the
shake ﬂask experiments. The SGR was the same in both the fed-batch fermentations and
the batch fermentations. Nutrient depletion or excessive fermentation byproduct creation
also did not appear to be an issue at the cell densities reached in this fermentation; the

growth rate was nearly constant prior to the time of induction.

36

The IPTG concentration of 0.2 mM and the expression temperature of 26 °C were
speciﬁcally chosen to reduce the likelihood of a stress response in the culture. The
lowered temperature decreased the overall growth rate, but the cells continued to grow
and reached a ﬁnal OD600 over 1 unit higher than at the time of induction. In addition,
the oxygen uptake rate of the culture continued to increase as indicated by the continuing
ramp of the agitation rate (results not shown). In order to maintain a constant dissolved
oxygen concentration, the agitation rate had to increase in order to balance the rate of
oxygen transfer with the increased rate of oxygen uptake required for cell growth and
protein expression.

As shown in Figure 9, very little inclusion body mass was found in the insoluble
fractions of the culture samples. The overall quantity of insoluble protein found in the
SDS-PAGE analysis was also very low, though, as indicated by the faint bands in lanes 6
through 9. Therefore, it was difﬁcult to draw conclusions about the extent of inclusion
body formation in relation to the overall cell mass. Because the insoluble HA2_185 band
never exceeded the intensity of the band at ~38 kDa, it was assumed that inclusion body
formation was not excessive. HA2_185 levels in the soluble fraction noticeably
increased over the course of the induction phase. Comparing the relative intensities of
the HA2_185 bands in Figure 4 and Figure 6 implied that more soluble protein could
have been produced before the solubility limit was reached. Even so, the mass of
HA2_185 in the harvested cell culture was high enough relative to contaminating proteins
that satisfactory purity was obtained through the puriﬁcation procedure. When Rosetta

was allowed to express for six hours at the same conditions in a separate experiment, the

37

HA2_185 concentration far exceeded the previously observed soluble fraction limit
within three hours (results not shown).

Incorporation of the isotopically labeled leucine into the protein for purposes of
NMR study was also successful. Based on the NMR signal, the ratio of labeled leucine to
unlabeled was estimated at 50%, a high enough incorporation rate for analysis by
REDOR (results not shown). Isotopic dilution was not signiﬁcant enough to affect the
results of the NMR studies.

These experiments illustrated the use of the protocols described in Appendix A to
optimize protein production in recombinant expression systems. Results from batch
experiments were used to develop a successful expression strategy for fed-batch
fermentations. In addition, simple tests such as OD600 and SDS-PAGE were used to
determine the SGR of a host strain, inhibitory effects of the induction system, the rate of
inclusion body formation in recombinant E. coli, and media selection.

Signiﬁcant limitations to this work and the supplied protocols exist with regards
to fed-batch fermentations. Relatively small quantities of protein were required for the
REDOR analysis, so the growth of high cell density cultures in the BioFlo BC was not
mandatory for the success of this project. The maximum dry weight of the culture never
exceeded 5 g/L. Attempts were made to grow this host to high cell densities in
conditions similar to those used in this experiment, but growth inhibition was found to
varying degrees when the culture reached dry weights above 5 g/L (results not shown).
Others have obtained dry weights of E. coli greater than 150 g/L (Lee, 1996). At these
cell concentrations, inhibition by toxic byproducts and oxygen transfer limitations

become increasingly problematic. The ability to optimize the control scheme to

38

overcome these limitations was addressed in the development of Presto Control,

however.

39

7. Presto Control

7.1 Background

Presto Control is the name given to the controller software designed for
conducting fed-batch fermentations in the BioFlo He. The reasons for its creation have
already been discussed in the introduction in section 1. A description of Presto Control’s
features and an introduction to ferrnentor control can be found in Appendix B.

Riesenberg (1999) and Rani (1999) review the vast number of control schemes
and concepts developed to produce high cell density cultures. Other than the stress
response caused by anoxic conditions in a bioreactor, poor control of substrate
concentration is the principal concern in high cell density culture growth and recombinant
protein expression (Konz et al., 1998; Notley and Ferenci, 1996). For example,
starvation for more than a brief period of time can irreparably reduce both growth rate
and maximum cell density if it is allowed. Generally, though, feed depletion is not
harmful if it can be corrected within a few minutes (Suzuki et al., 2000).

The Standard Fed Batch control scheme available in Presto Control utilizes a
speciﬁc method for detecting and correcting substrate starvation. This scheme is
described in detail in Appendix B. In brief, the control unit detects a spike in the
dissolved oxygen (DO) caused by a starvation induced drop on the cellular respiration
rate of the cells. It then reacts by adding a speciﬁed amount of feed. This is a commonly
utilized feed strategy (Nor et al., 2001; Knop, 2002; Suzuki et al., 2000; Oh et al., 1998).

To test this controller response, the BioFlo He was prepared and inoculated with

Rosetta following the protocols in Appendix A. The Standard Fed Batch control recipe

40

was utilized as described in the protocols, and an initial 10 g charge of feed was added in
the ﬁrst stage. Eight hours after inoculation, substrate depletion was detected by the
controller, and the results showing controller compensation are presented below. The
response of the controller was governed by stage 2 and stage 10 of the Standard Fed

Batch control scheme.

7.2 Results

Figure 10 shows the data points taken by the fermentation logging function of
Presto Control at the time of the feed depletion event. The time points of the DO and the
total mass of glucose added were recorded every thirty seconds. The stage logging
feature was referenced to give the exact times of the stage switches. The DO set-point
and trigger values are indicated on the graph. Also, the period of each stage is designated
with the dotted lines. At the time of feed depletion, the control scheme was in stage 2.

At a recorded time of 16.1843 hrs into the fermentation, the controller detected that the
D0 in the BioFlo IIc had exceeded 15%. Since Trigger A of stage 2 is set to call stage 10
when the DO exceeds 15 %, stage 10 was initiated. Immediately, the feed pump began
adding feed as speciﬁed in stage 10. At 16.1912 hrs (25 seconds later), the controller
detected that the DO had dropped below 15% again. This triggered stage 2 and the ﬁnal

10 grams of feed speciﬁed in stage 10 were added.

41

Figure 10: Presto Control Response to Substrate Depletion

 

 

 

 

 

 

 

 

 

 

 

20 28
Stage 2 5 5 Stage 2
I I I -
s s - -~ 27
E 1'1—Stage 10 .'
15 :‘L L ‘” 25
l 1 I
oo Trigger ’: : I -_ 25
Value 5 E I A
o : :' " 3’
A O : : I di— 24 n
2\110 .10. 4' i 9 I 8
O I i i .9 3:90 23 E
a DO Set-Point E E °: 0 0 0° ’0 ’ ° 8
l i I u.
: : I “ 22
l i I
1 H
5 : :I 4 21
III-Illri .00
E E I Feed Mass " 20
0 I . . I I m I 19
16.1 16.15 16.2 16.25 16.3 16.35 16.4

Time (hr)

7.3 Discussion

The Standard Fed Batch control scheme responded as expected to the feed
depletion event. Unfortunately, due to the 30 second scan rate for variable logging, only
a broad understanding of the probe and controller responses can be obtained from the
data shown in Figure 10. The following description should provide greater insight into
the events.

Before feed depletion, the DO began to slowly rise over a two minute period as
the glucose concentration dropped to inhibitory levels. When the glucose was completely

consumed, the DO rose from 12% to 15% within 5 seconds. In response, the feed pump

42

immediately began to add glucose at a rate which brought the glucose to a concentration
above 0.05 g/L within 5 seconds. Uptake of the glucose and reestablishment of the
original oxygen uptake rate was nearly instantaneous. The DO probe, however, exhibited
an approximate 5 second time lag, which delayed the drop in measured oxygen. This
response delay allowed the D0 to spike to over 20% before the DO began to decrease.
Initially, the DO decreased very quickly but. as it got closer to its steady state value, the
rate of decrease declined. At 25 seconds, the DO ﬁnally dropped below 14% and the rest
of the 10 grams was added.

If the DO had not dropped below 14% within 30 seconds, feed would have been
cutoff by the triggering of stage 1 1. This action is designed to prevent excessive
amounts of feed from being added due to rises in the DO that are not caused by feed
depletion. It is possible that the DO response to feed might be slow enough that a control
scheme might discontinue feeding even if the detected DO spike was due to glucose
depletion. At the tubing size and glucose concentration recommended in Appendix A,
0.35 g of glucose would have been added to the fermentation media before the thirty
second cut-off point. This is well above the saturation limit of glucose for E. coli
(Levisauskas et al., 2003). Therefore, even if the control scheme did halt feeding early
by triggering stage 11, the DO would quickly drop back down to steady state and the
controller would be prepared to detect the next DO spike.

The response of Presto Control to a feed depletion event is a very basic example
of the capabilities of the program. Appendix B explains its features and its use in

controlling a fed-batch fermentation. More importantly, Presto Control is ﬂexible

43

enough that an operator can tune the control parameters and design control schemes

according to the needs of each protein expression system.

44

APPENDICES

45

APPENDIX A

PROTOCOLS FOR PROTEIN EXPRESSION

46

APPENDIX A: TABLE OF CONTENTS

INTRODUCTION ............................................................................................................. 49
BACKGROUND ........................................................................................................... 49
GENERAL SAFETY GUIDELINES ........................................................................... 50

BATCH FERMENTATIONS ........................................................................................... 51
TEST TUBE CULTURES ............................................................................................ 52
SHAKE FLASK CULTURES ...................................................................................... 53

FED BATCH FERMENTATIONS .................................................................................. 56
BIOFLO IIC INTRODUCTION ................................................................................... 56
BIOREACT OR PREPARATION CHECK LIST ......................................................... 57
BIOREACTOR PREPARATION ................................................................................. 58

pH Probe Calibration ................................................................................................. 59
Bioreactor Assembly ................................................................................................. 59
Bioreactor Sterilization ............................................................................................. 61
DO Probe Ionization .................................................................................................. 63
FEED VESSELS ........................................................................................................... 64
LOADING PRESTO CONTROL ................................................................................. 66
CONTROL PREPARATION ....................................................................................... 67
Mixing ....................................................................................................................... 67
Temperature Control ................................................................................................. 67
pH Control ................................................................................................................. 68
Feed Pump Calibration .............................................................................................. 70
Feed Control .............................................................................................................. 72
Air Control ................................................................................................................ 73
DO Probe Calibration ................................................................................................ 74
Agitation Control ....................................................................................................... 75
MEDIA ADDITIONS ................................................................................................... 77
INOCULATION ........................................................................................................... 79
BIOREACT OR GROWTH ........................................................................................... 80
ANT [FOAM ADDITION ............................................................................................. 82
BIOREACT OR SHUTDOWN ..................................................................................... 83

SUPPLEMENTARY MATERIAL ................................................................................... 85

MEDIA .......................................................................................................................... 85
M9 ............................................................................................................................. 85
Modiﬁed M9 ............................................................................................................. 86
LB .............................................................................................................................. 86
TB .............................................................................................................................. 86
Deﬁned Media ........................................................................................................... 87
Stock Solutions .......................................................................................................... 89
Antibiotics ................................................................................................................. 90

AUTOCLAVING .......................................................................................................... 92

AGAR PLATES ............................................................................................................ 94

INOCULUM CULTURES ............................................................................................ 95
Agar Plates ................................................................................................................ 95

47

Test Tube Cultures .................................................................................................... 96

Shake Flask Cultures ................................................................................................. 97
CELL DENSITY — OD600 ........................................................................................... 99
CULTURE SAMPLES ............................................................................................... 101
INDUCTION ............................................................................................................... 103
CULTURE HARVESTING ........................................................................................ 105
SDS-PAGE .................................................................................................................. 107

Gel Casting .............................................................................................................. 107

Cell Fractionation .................................................................................................... 1 l 1

Sample Preparation ................................................................................................. 114

Electrophoresis Procedure ....................................................................................... 116

Gel Staining and Analysis ....................................................................................... 119

Stock Solutions ........................................................................................................ 121

48

INTRODUCTION
BACKGROUND

This is intended as a general protocol for producing protein in microbial cultures using
the equipment available in room 3262 of the Engineering Building at Michigan State
University. The included material is particularly directed towards optimizing
intracellular protein production in recombinant E. coli. It is split into three sections. The
ﬁrst section describes batch growth and production experiments in shake ﬂasks and test
tubes. The second, much longer section, contains protocols for completing fed batch
protein production experiments using the BioFlo He. The third section contains
descriptions of tools and techniques that are relevant to both batch and fed-batch projects.
It includes a protocol for SDS-PAGE, cell density measurements, autoclaving,
centrifugation of the culture, and various media recipes.

This manual assumes that the operator has already created or obtained the host strain and
induction procedure. It does not include information on the creation or transformation of
host cells and vectors for recombinant engineering. Molecular Coning: A Laboratory
Manual (Sambrook and Russell, 2001) can be referenced for these procedures. In
addition, the “pET System Manual” (Novagen, 1992) has a good overview of common
recombinant E. coli strains and induction systems.

These protocols assume that the operator has some knowledge of aseptic culturing
techniques and cellular biochemistry and growth kinetics. Every operator has their own
personalized style and beliefs on cell culturing and protein production, so view this
manual as suggested practice and not as scientiﬁc law.

A choice needs to be made regarding the growth media. The best media for use in
fermentations varies depending on the cell type and the ﬁnal product desired. Different
media compositions and their applications can be found in MEDIA on pg. 85.

The goal in recombinant engineering is to overexpress a heterologous protein in a cell
culture. Two common methods for achieving this goal exist. The ﬁrst is to express the
protein as a batch culture in a shake ﬂask or test tube. The second is to grow the culture
in a stirred tank bioreactor using probes and media feeds to control the environmental
conditions. Other methods such as continuous growth reactors and immobilized cell
biocatalysis exist, but they require special equipment and are not covered in this protocol.
Batch suspension cultures and fed batch fermentations in bioreactors each have speciﬁc
advantages over the other.

The major advantage of batch fermentations in shake ﬂasks or test tubes is that they do
not require specialized bioreactor equipment. In addition, the preparation time for the
experiment is much lower. All that is required is an enclosed vessel and a shaker to keep
the culture suspended. The disadvantage is that the environmental conditions of a
suspension cultures are difﬁcult to control. Feed concentration, pH levels, and oxygen

49

concentration in the media can be impossible to maintain. Therefore, the cell density in a
shake ﬂask culture is limited by the buffering capacity of the media and the minimal
oxygen transport allowed by the shaker and the ﬂask. Generally, the maximum dry
weight of cells per liter of media that can be obtained in a batch culture is 2.5.

In contrast, fed batch fermentations in a bioreactor can produce cell densities as high as
50 times those in a shake ﬂask. Substrate can be fed on demand to keep the culture from
starving without toxic overfeeding. Acid or base can be added to maintain physiological
pH. Agitation and aeration can be ramped up to maintain the desired dissolved oxygen
concentration for growth. All of these processes, however, require specialized equipment
and probes. In addition, the culture volume is limited by the size of the bioreactor. The
bioreactor described in this protocol, the BioFlo He, has a working volume of 1 L. Each
experiment requires a relatively large quantity of media and, therefore, a large
expenditure of time and money. Furthermore, if only one bioreactor is available,
fermentations cannot be run simultaneously and the results compared.

For these reasons, the choice to use a suspension culture or a stirred tank bioreactor is
dependant on the goals of the fermentation. Batch fermentations work well as
preliminary steps in optimizing cell growth and protein production. Small batch
experiments with differing media or induction conditions can be run simultaneously and
the results used to determine the optimum expression strategy. The expression strategy
can then be employed in high cell density, fed batch fermentations in order to produce
large amounts of the desired protein. In some cases, however, the expression strategy is
dependant on the cell density, so the BioFlo He is required for the ﬁnal optimization
steps. Conversely, the protein being produced might only be needed in small quantities
and expression in a shake ﬂask might be ideal.

GENERAL SAFETY GUIDELINES

o If any problem occurs, immediately contact the safety representative for the lab. The
name and number of the contact person is posted on the door to the lab.

The pressure in the air lines and the bioreactor needs to be kept under 15 psig.
Filters and drains should be monitored for ﬂow resistance or plugging.

Eye, arm, leg, and feet protection should be worn while working in the lab.

Gloves should be worn when working with materials that have been exposed to the
culture.

Rubber gloves and splash goggles must be worn when working with acid and base
tubing and containers.

Acid and base containers should only be opened in the fume hood.

All vessels, including the bioreactor, must be vented during autoclaving.

Thick autoclave gloves must be worn when handling recently autoclaved vessels.
All centrifuges need to be properly balanced and the tops to centrifuge tubes and
containers must be tightly sealed.

All surfaces and materials that come into contact with the culture must be sterilized.

Pipette tips and syringes used to handle the culture should be placed in a biohazard
bag after use.

50

BATCH FERMENTATIONS

These protocols assume that the purpose of the project is to test multiple conditions in
simultaneous experiments in order to determine the optimum expression strategy. They
are broken into two categories: test tubes and shake ﬂasks. Ultimately, the optimized
expression strategy can be used to produce large amounts of protein in large volume
shake ﬂasks or in the BioFlo Hc.

When working with a new host or induction system, certain environmental parameters
should be checked for their inﬂuence on protein expression. They are:

0 Temperature

Bulk media composition

Initial substrate concentration

Inducer concentration

Induction time

Knowledge of the expression host or recombinant protein can supply information on
which variables might be important. Other environmental conditions such as cell density,
pH, and dissolved oxygen concentration can only ‘be analyzed in the BioFlo 110.

As with all fermentations, the most reliable method of beginning a project is to use an
inoculum culture grown up in a series of increasing volume transfers. A full discussion
of inoculum cultures can be found in INOCULUM CULTURES on page 95. In addition,
thought needs to be given to the way in which simultaneous batch culture experiments are
to be analyzed. If batch cultures are to be compared, they must all come from the same
inoculum culture. Furthermore, the inoculum culture should only be separated at the time
that the variance in the tested environmental condition is introduced. The growth and
expression characteristics of separated cultures can drift apart and cause non-speciﬁc
variability in the compared results.

The following is an example culture transfer scheme for simultaneous comparison of
IPTG inducer concentrations and induction time. The listed steps include inoculum
culture transfers.

Streak glycerol freeze sample onto agar plate.

Grow to until visible colonies are formed.

Transfer single colony from plate to 5 mL of media in a 25 mL test tube.

Grow to OD500 of 1.5.

Transfer 5 mL test tube culture to 200 mL media in l L shake ﬂask.

Grow to OD600 of 1.0.

Separate 200 mL culture into four 50 mL cultures in 250 mL shake ﬂasks.

Add either 0.1 mM, 0.2 mM, 0.4 mM. or 0.8 mM IPT G to each ﬂask.

Grow and take samples at 0, 1, 2, 4, and 6 hour time points.

TEST TUBE CULTURES

Due to the small volumes available in test tubes, it is difﬁcult to take more than one
culture sample without signiﬁcantly changing the culture conditions. Therefore, test
tubes are primarily used to determine the ﬁnal results of an expression strategy. They are
also useful for testing a large range of varying media compositions. Each test tube
contains a speciﬁc media type or carbon source and the ﬁnal expression levels are tested
to determine if a speciﬁc media type acts to inhibit protein production.

W

Media

Inoculum

Sterilized ﬂask

1 or more sterilized test tubes with tops
Sterilized pipette tips

Sterilized wooden colony applicator or metal loop
Sterilized graduated cylinder (optional)
Sterilized pipette tips

Pipetter

Inducer

Marking tape

Permanent marker

Time: 9 hrs to 18 hrs

Procedure

1) Prepare an inoculum culture on an agar plate or in a test tube according to
INOCULUM CULTURES on pg. 95.

2) Prepare the media to be tested according to MEDIA on pg. 85.

3) In the laminar ﬂow hood, transfer the sterilized media to the desired number of
sterilized test tubes with caps.

0 Suggested: 5 mL of media in a 25 mL test tube.

4) Move the inoculum plate or test tube culture from the incubator to the laminar ﬂow
hood.

0 The culture should not have been refrigerated before transfer. It should be in mid-
exponential growth phase at an OD600 z 1.0.

5) Inoculate the test tube with a single colony from the plate or a small volume from the
test tube.

0 Remove the cap to the test tube and sterilize the lip of the glass by passing it
through a ﬂame from the Bunsen burner. For an agar plate, remove the cover and
touch a sterilized wooden applicator to a visible colony then carefully swirl it in
the media in the test tube. For inoculum from a test tube, use the pipette to add
1% to 10% culture volume to fresh media.

6) Label the tubes with your name, date, and the components of the culture.
7) Incubate the tubes in the desired growth conditions.
0 Suggested: 37 °C, 250 RPM

52

8) When the culture reaches the desired cell density for induction, transfer the test tube
back to the larrrinar ﬂow hood.

0 Suggested: When the broth becomes opaque (OD600 of approximately 1.0).

9) Add the required amount of inducer to bring the inducer concentration to the desired
induction strength.

0 Suggested: Dilute 0.1 mL of 1 M IPTG to 1 mL with sterilized RO water. Add 50
uL of 0.1 M IPTG to the 5 mL test tube culture for an induction strength of lmM.
10) Return the culture to an incubator which has been set it to the desired temperature.
0 Maximum total expression in E. coli usually occurs at 37 °C.
0 Do not allow the culture to remain unagitated for extended periods of time at high
cell concentrations or the culture can become oxygen starved.

11) At the end of the induction period, take a 1 mL culture sample according to
CULTURE SAMPLES on pg. 101 and a cell density measurement according to
CELL DENSITY — OD600 on pg. 99.

0 Suggested: 4 - 6 hr induction period.

12) Sterilize the remaining culture in the test tubes and inoculum vessel.

0 Method 1: Place in a test tube rack and autoclave the culture according to
AUTOCLAVING on pg. 92.

0 Method 2: Move the test tubes to the walk-in hood in 3269 or a fume hood. Fill
each test tube with 10% bleach solution and let sit for 10 minutes.

13) Clean the test tubes.

0 Dump the sterilized culture down the drain and wash the test tubes with soap and
water. Then rinse them with R0 water and allow them to dry.

14) Analyze the culture samples according to SDS-PAGE on pg. 107.

SHAKE FLASK CULTURES

Shake ﬂasks have a larger volume then test tubes and can supply multiple culture samples
from a single vessel. Therefore, the operator can obtain a time course analysis of the
growth rate or expression levels of the culture. Any experiment that can be done in a test
tube can be expanded to a shake ﬂask experiment. Test tube cultures, however, require
less time and material.

Large volume shake ﬂask cultures can also be used to produce protein for
experimentation purposes. Many biochemistry labs that do not have stirred tank
bioreactors use shake ﬂasks for all of their recombinant protein work. If a shake ﬂask is
to be used to produce protein, the operator must remember that the maximum cell density
in a batch experiment is inhibited by the buffering capacity of the media. In general, the
maximum OD600 that can be reached before the physiological pH is lost is between 3 and
4. M9 media results in acidic conditions while LB media tends towards basic conditions.
Also, for E. coli, 1 g of glucose provides enough substrate for a little over a half unit
increase in the OD600. Therefore, 4 grams of glucose can potentially give a ﬁnal cell
culture density indicated by an OD600 of approximately 2.5. The quantity of substrate
available in the normal recipe for LB media that does not include glucose provides
similar results. For these reasons, it is common practice to induce protein expression in a

53

batch culture when the OD600 has reached 1.0. This allows the culture to continue to
grow and produce protein before the limitations of the media are reached.

Supplies
Media

Inoculum

Multiple sterilized Erlenmeyer ﬂasks
Sterilized pipette tips

Pipetter

Inducer

Sterilized graduated cylinder (optional)
Marking tape

Permanent marker

Time: 9 hrs to 18 hrs

Procedure

1)
2)
3)

4)

5)

6)
7)

8)

9)

Prepare an inoculum culture in a test tube or shake ﬂask according to INOCULUM

CULTURES on pg. 95.

Prepare the media according to MEDIA on pg. 85.

Transfer the sterilized media to one or more sterilized shake ﬂasks.

- For maximum oxygen transport, the media should not be greater than 20% of the
total ﬂask volume.

Transfer the inoculum culture from the incubator to the laminar ﬂow hood.

0 Suggested: OD600 of 1.0 for inoculum.

0 Do not allow the culture to sit too long without agitation or the cells can suffer
from the stringent response caused by oxygen starvation.

Inoculate the shake ﬂasks with the inoculum culture.

0 Remove the covering from both culture vessels and ﬂame the opening to sterilize.
Use the pipette to add 1% to 10% culture volume to the fresh media in the shake
ﬂasks. Otherwise, just the ﬂame the lip to the inoculum vessel and pour the
culture directly into the ﬂask.

Label the ﬂasks with your name, date, and the components of the culture.

Incubate the shake ﬂasks.

0 Suggested: 37 °C and 250 RPM in the enclosed incubator in the teaching lab.

When the culture the reaches the desired cell density for induction, transfer the shake

ﬂasks back to the laminar ﬂow hood.

0 Suggested: OD600 of 1.0.

Remove the top to the shake ﬂask and ﬂame the opening.

10) Take a 1 mL culture sample according to CULTURE SAMPLES on pg. 101 and a

cell density measurement according to CELL DENSITY — OD600 on pg. 99. This
will be the non-induced 0 hr time point for the expression.

11) If multiple environmental conditions or media component concentrations are to be

tested for their impact on induction, split the large shake ﬂask culture among smaller
shake ﬂasks.

54

0 Remove the coverings to the sterilized shake ﬂasks and ﬂame the openings. Use
either large volume sterilized pipette tips or a sterilized graduated cylinder to split
the culture equally among the given containers. Maintain a culture volume of
approximately 20 % of the total shake ﬂask capacity.

12) Add any additional media components to the separated cultures.

0 A change in carbon source for the culture source cause a lag phase in growth or
change in the metabolism of the culture. This makes determination of the effect
of substrate on the rate of protein expression difﬁcult.

13) Add the required amount of inducer to bring the inducer concentration to the desired
induction strength.

0 Suggested: 1 mM IPTG for full induction of Lac promoters.

14) Incubate the cultures in the desired environmental conditions.

0 Do not allow the culture to remain unagitated for extended periods of time when
at high cell concentrations or the culture can become oxygen starved.

15) Take additional culture samples and OD600 readings during the course of induction.

0 Suggested: 0.5, 1, 2, 3, 4, 6 hrs.

16) At the end of the induction period, determine the cell density and take a ﬁnal 1 mL
culture sample.

0 Suggested: 4 — 6 hr induction period.

17) If the culture is to be processed and the protein puriﬁed, harvest the cells according to

CULTURE HARVESTING on pg. 105.

18) Sterilize the remaining culture in the shake ﬂasks and inoculum vessels.

0 Method 1: Partially ﬁll the ﬂasks with water and autoclave the culture according
to AUTOCLAVING on pg. 92.

0 Method 2: Move the shake ﬂasks to the walk-in hood in 3269 or a fume hood.
Fill each test tube with bleach to produce a 10% ﬁnal concentration Md let sit for
10 minutes.

19) Clean the glassware.

0 Dump the sterilized culture down the drain and wash the ﬂasks with soap and

water. Then rinse them with R0 water and allow them to dry.
20) Analyze the culture samples according to SDS—PAGE on pg. 107.

55

FED BATCH FERMENTATIONS
BIOFLO IIC INTRODUCTION

The BioFlo He is a 1 L working volume, bioreactor vessel and control console designed
for growing microbiological cultures to high cell densities. This section is intended as an
introduction and step by step procedure to operating the BioFlo Hc bioreactors in room
3262 of the biochemical engineering teaching lab. The following protocols include tips
and suggested operating parameters for ﬁrst time users of the equipment. Many of the
suggestions primarily target projects involving induced, intracellular protein production
by recombinant E. coli, so other microbiological systems require modiﬁcations to this
protocol.

Opto 22 is a set of software programs and processing units designed for industrial
automation projects. Due to the limited control capabilities built into the BioFlo 11c,
Opto 22 has been used to add functionality to the bioreactor system. “Presto Control” is
the name given to the controller and display programs designed using Opto 22 that allow
staged controller recipes to regulate the process variables of the BioFlo Hc. While an
understanding of the software driving the controller of the BioFlo HC is not required to
run a project, it is helpful. It is possible for the culture or the equipment to function
outside of the parameters expected by the controller and quick changes might need to be
made to the controller recipe to keep the project from failing. The Presto Control
Manual, describes the features of the program and how individualized recipes can be
created.

For background on fed batch bioreactors, New Brunswick has published some
introductions to fermentation and equipment in “Fundamentals of Fermentation:
Techniques For Benchtop Fermentors” (R & D Lab, date unknown) and “An Introduction
to Fermentation: Fermentation Basics” (Pumphrey and Julien, 1996). Useful information
on bioreactors can be also be found in reviews by Lee (1996), Liden (2002), and
Riesenberg (1999).

56

BIOREAC TOR PREPARATION CHECK LIST

Prepare and sterilize media components

Prepare and autoclave feed vessels, shake ﬂasks, and test tubes
Prepare and pour plates

Inoculate plates

Inoculate test tube cultures

Inoculate shake ﬂask cultures

Calibrate pH probe

Clean and assemble bioreactor

Autoclave bioreactor

Add media components to bioreactor and feed vessels

Hook up bioreactor to BioFlo IIc Console and feed vessels
Ionize DO probe

Calibrate DO probe

Setup Presto Control and BioFlo IIc Console for bioreactor control

Inoculate bioreactor

57

BI UREA C TOR PREPARA T I ON

The bioreactor needs to be sealed and sterilized before introduction of the culture. The
following procedure is a list of all of the basic tasks that must be accomplished before the
bioreactor is ready to be used. Due to the requirements of the DO probe calibration
procedure, the equipment must be autoclaved at least six hours before the culture is
introduced. Do not wait until the culture is nearly ready for transfer before starting this
part of the protocol.

All of the BioFlo 110 control parameters are set using the “Selector” and “Mode” knobs
on the front of the BioFlo IIc console. The “Increase/Decrease” switch changes the
values in each mode.

Mode Options:

CONTROL: Shows the actual value of the selected monitored variable. This does
not have to be selected for the BioFlo Do to control the variable; the control is a
continuous process that cannot be turned off.

SET POINT: Shows the set point value for the selected environmental variable.
When in D0 control mode, the displayed set point value for agitation is the
maximum RPM value allowed.

ZERO: Shows the current value of the selected variable and used for calibrating
the pH and DO probes. When this mode is selected and the Increase/Decrease
button is pressed, the current value of the monitored variable is set to the new
displayed value.

SPAN: Shows the current value of the selected variable and used for calibrating
the pH and DO probes. When this mode is selected and the Increase/Decrease
button is pressed, the current value of the monitored variable is set to the new
displayed value. In conjunction with the value set in the “ZERO” mode, the span
value creates a calibration curve for the entire range of values.

Selector Options:

RPM: Controls the motor that spins the impeller. In DO control mode, this
option is used to set the minimum and maximum allowed RPM values.
NUTRIENT: Disabled. This function has been bypassed so that it can be
controlled by Presto Control.

pH: Monitors the pH of the media and controls both the acid and the base pumps.
DO: When the DO control mode is selected, the set point for this mode will
control the agitation between the maximum and minimum set points for RPM.
Otherwise, this mode just monitors the DO and allows calibration of the DO
probe.

ANT IFOAM: The probes are not available for the use of this controller.
TEMPERATURE: Controls the pump, heater, and valves that are used to supply
cooling or heating water to the water jacket.

58

pH Probe Calibration

The pH probe calibration should be performed prior to autoclaving the bioreactor.
Otherwise, the probe and the bioreactor can become contaminated when the probe is
removed and reinstalled during the calibration process. Autoclaving the probe in a liquid
environment does not affect its functionality and the BioFlo He maintains calibration set
points even when it is turned off. The probe should be calibrated in buffers heated to the
operating temperature of the bioreactor but this can be difﬁcult to accomplish.

Supplies

Bioreactor pH probe
BioFlo IIc pH probe cable
RO water bottle

Rinse container

pH 7.0 buffer

pH 4.0 buffer

Time: 10 min

Procedure

1) Remove the probe from its storage solution and rinse it with R0 water.

2) Attach the pH cable to the probe and to the BioFlo He.

3) On the BioFlo IIc control unit, turn the Selector knob to “pH” and the Mode knob to
“Zero”.

4) Insert the pH probe into pH 7.0 buffer and wait for the reading to stabilize.
0 The nonreﬁllable pH probes respond slowly and may take a couple minutes to

fully stabilize.

5) Use the Increase/Decrease button to adjust it to 7 .0.

6) Remove the probe from the pH 7.0 buffer and rinse it with R0 water.

7) Place the probe in pH 4.0 buffer and switch the Mode knob to “Span”.

8) After the reading has stabilized, use the Increase/Decrease button to adjust it to 4.0.

9) Detach the pH probe and rinse it with R0 water. The probe is now calibrated and
ready to be autoclaved with the bioreactor.

10) Return the probe to its storage buffer until the bioreactor is ready for the head plate to
be installed.

Bioreactor Assembly

The exact requirements for assembling the bioreactor depend on the parts available and
the type of ports on the head plate. Some engineering ingenuity might be required to get
everything installed and functioning properly.

Su lies

Media

Bioreactor vessel

Head plate

Feed port
Culture/media/antifoam port

59

3 head plate screws
Inoculation port cap
Stoppers or caps for unused ports
Therrnowell port

Sample port and assembly
Sparging port and assembly
Impeller shaft and assembly
2 impellers

Condenser port

Two 0.2 um air ﬁlters
Bafﬂes

Scissors

Multiple II) tubing

pH probe

DO probe

Hex head wrenches

Time: 20 minutes

Procedure

1)

2)

3)

4)

5)
6)
7)

Check that the bioreactor vessel, bafﬂes, head plate and attachments are all clean.

0 It is difﬁcult to remove all of the detergent from the vessel when it is cleaned.
Rinse the vessel thoroughly before use or excessive foaming can occur during
agitation.

0 If feed needs to be controlled precisely, a piece of tubing attached to a pipette or
syringe tip should be attached to the outlet of the feed port on the underside of the
head plate. The outlets of the standard metal ports can cause erratic feeding due
to large drop sizes or feed solution clinging to the underside of the head plate.

Check that the impeller drive shaft does not easily slide up and down and that it turns

with only moderate effort and without the bearings squeaking or grinding.

o If the shaft slides up and down, a new underside seal assembly will need to be
ordered and installed.

0 If the shaft is grinding, a new ball bearing assembly will need to be ordered and
installed.

Install the impeller blades in the desired positions on the impeller shaft.

0 Suggested: Use two of the six spoke, ﬂat bladed impellers with a diameter of 2
1/8”. The lower impeller should be positioned at the end of the impeller shaft just
above the sparger. The second impeller should be just below the surface of the
liquid (about 2 V2” above the lower impeller).

Place the bafﬂes in the bioreactor

o In order to allow all of the port connections to be made, the gap in the bafﬂes
needs to face the cooling water ports on the bioreactor base.

If they are present, seal the addition ports in the sides of the glass vessel.

Add media to the vessel according to MEDIA on pg. 85.

Place the head plate on the glass vessel.

0 Make sure that the black rubber O-ring is in place in the groove on the underside
of the head plate and that it forms a seal with the lip of the glass vessel.

0 Line up the pH and DO probe ports with the gap in the bafﬂes and look through
the ports to ensure there is adequate space for safe probe insertion.

8) Remove the pro-calibrated pH probe from its storage buffer, rinse it with R0 water
and slide it into the head plate.

0 Most likely, the pH probe being used does not screw into the head plate and the
seal is maintained with just an O-ring.

0 If adjustments need to be made to the bafﬂes or the head plate, do not use the pH
probe as a lever to force movement because it may break.

9) Check the membrane of the DO probe for damage. Insert the probe into the head
plate and screw in the adapter.

0 If the probe appears damaged, the probe was responding slowly or erratically in
previous fermentations, or it has been over six months since changing the
electrolyte, check the probe’s instruction manual for the protocol on how to
service it.

0 The DO probe shaft may be too long for the vessel. A black O-ring can be slid
partially up the body of the DO probe and used to form a seal between the
adapter, the DO probe, and the head plate.

10) Raise the stainless steel ring encircling the glass vessel and secure the head plate with
the thumbscrews.

0 Do not fully tighten the thumb screws before autoclaving or the glass lip may
crack.

11) Recheck that the impeller can spin freely and that all of the probes and ports line up
correctly
12) Attach an autoclavable 0.2 pm air ﬁlter to the air inlet hose on the head plate. Attach

a second ﬁlter to the air outlet hose on top of the exhaust gas condenser.

0 Twist the ﬁlter onto the end of the tubing so that the tubing reaches at least to the
second ridge of the ﬁlter. Filters and the ﬁlter tubing can be secured with zip ties.

o The ﬁlters can be reused from previous experiments if they have not become
plugged and signiﬁcantly impede airﬂow. If the removal of either the inlet or
outlet ﬁlter from the airﬂow path causes a large airﬂow spike by the controller,
the ﬁlter should be replaced.

0 If plugging is a concern, T-adapters can be utilized to allow two ﬁlters to be used
in parallel on the inlet or outlet line. If one ﬁlter becomes plugged, the other will
still function.

Bioreactor Sterilization
Certain principles should be kept in mind when preparing the bioreactor for autoclaving.
0 Do not over tighten any of the screws or caps. They can become overly tight or
cause cracking of the glass vessel during autoclaving.
0 Any tubing or ports that stick down into the media within the bioreactor must be
closed off before autoclaving to prevent media loss. These include the sample
port and the air inlet.

61

0 Air must be allowed to move in and out of the vessel during autoclaving in order
to allow equilibration of the pressure. Media can be forced out of the vessel
during depressurization of the autoclave if equilibration is not maintained.

0 Moving metal parts and ﬁlter openings should be covered with aluminum foil to
prevent damage.

0 Certain components of the media should not be autoclaved and must be ﬁlter
sterilized. These solutions should be added after the bulk media has cooled.

Supplies
Aluminum foil
Clamps

Time: 1 hr

Procedure
1) Clamp off the inlet air hose.

0 Do not clamp the exhaust gas line. The exhaust gas line allows airﬂow for
pressure equilibration during autoclaving.

2) Seal off the acid, base, and glucose addition ports with small sections of clamped
tubing.

0 If antibiotics and mineral salts are going to be added after the bioreactor has been
autoclaved and hooked up to the BioFlo IIc console, a small 0.2 pm syringe ﬁlter
can be attached to the tubing of one of the inlet ports before autoclaving and used
to aseptically transfer the media components.

3) Clamp the tubing attached to the sampling port and close the sampling port valve.
4) Cover all inlets, outlets, probe connections, and moving metal pieces with aluminum
foil.

0 These parts include the impeller shaft, the outlets of the air ﬁlters, all clamps, the
tops of the pH and DO probes, and any ports that may have been left open.

0 The aluminum foil increases heat transfer and reduces the chances of corrosion
due to condensation on the metal parts.

5) Unscrew the cap to the inoculation port so that it is just resting on the port.
0 This keeps pressure from building up within the vessel and keeps the threaded
seal from becoming too tight during heating and cooling in the autoclave.
6) Recheck to be sure that all of the ports with openings into the media are sealed.
7) Place the vessel upright in the autoclave.
8) Flip the Power and Control switches to the “On” position on the autoclave.
9) Open the cooling water and steam valves located behind the autoclave.
10) Close the autoclave door and lock it in place by turning the handle all the way to the
right.
11) Set the Steam Sterilize Time to 35 minutes and the Steam Dry Time to 00 minutes.
12) Press the Reset button.
13) Press the “Liquid Cycle” button to start the autoclave process.
14) Wait for the autoclave to ﬁnish.

62

o The autoclave will run for 35 minutes once it reaches 132 °C. It will then spend
several minutes exhausting the steam. The buzzer will sound when the exhaust
cycle is complete. The buzzer can be stopped by pressing the reset button.

15) Close the steam valve behind the autoclave.
16) Open the door, remove the vessel and carry it to the BioFlo IIc stand.

0 If possible, allow the autoclave to continue to cool for at least 15 rrrinutes before
opening the door. This prevents excessive volume reduction of the media due to
boil off when the door is opened and the remaining pressure is released.

0 Do not leave liquid in a closed autoclave with the steam valve open for more than
a couple of hours or a large portion of the media may evaporate.

17) Close the cooling water valve to the autoclave.

DO Probe Ionization

The DO probe functions by measurement of the current produced by a continuous voltage
between an anode and a cathode. Oxygen diffuses across the probe membrane that
separates the potassium chloride electrolyte solution from the bioreactor media and is
reduced at the cathode. The reduction is balanced by the oxidation of silver at the anode.
The current produced is directly proportional to the oxygen concentration if the voltage is
constant. Therefore, the electrodes of the DO probe must be fully polarized before
calibration. Ionization of the DO probe is accomplished by having it hooked to the
BioFlo IIc console with the consoles power turned on for a minimum of 6 hours.
Deionization is not instantaneous, so the probe can be unhooked for a short period of time
during the ionization process if need be. The ﬁrst steps of the procedure are concerned
with connecting the BioFlo IIc to its console.

Time: 6 hrs

Procedure

1) Make sure that the main water supply to the BioFlo He is turned on.

2) Check that the water line into the back of the BioFlo He is open.

3) Turn on the power to the BioFlo He.

4) Remove the foil from the probe connections to allow any trapped water to evaporate
quickly.

5) Remove the clamp on the air inlet.

0 During autoclaving, the air hose may become permanently compressed. Open the
clamp slightly and rotate it 90 degrees so that it forces the compressed portion of
the tubing to remain open.

6) Lightly tighten any of the loosened screws or port caps.
7) After the vessel is cool enough to touch, attach the DO probe cable to the probe and
the BioFlo He.

0 The bioreactor can be quickly cooled down by using the cooling jacket as
described in Temperature Control on pg. 67.

63

FEED VESSELS

The optimum feed source for culture growth is dependant on the host strain and the
expression system. In all cases, the feed should consist of a highly concentrated carbon
source. In order to reduce the volume increase caused by the addition of feed, the carbon
source should be near to its solubility limit while not being too viscous to pump. A
nitrogen source is also be needed if the base being used to maintain the pH is not
ammonium hydroxide or ammonia.

0 Glucose: E. coli grow preferentially on glucose but certain protein expressions
systems are catabolically inhibited by glucose. In some cases this is a desired effect
because it inhibits basal level (leaky) expression that can inhibit the culture during the
growth phase. Glucose can also lead to excessive levels of acetic acid formation by
certain strains of bacteria.

0 Glycerol: Glycerol does not lead to catabolite inhibition, but its uptake is slower than
that of glucose. Glycerol has been shown to stop acetate production by some strains
of E. coli (Lee, 1996).

0 Lactose/Galactose: Galactose induces expression in systems utilizing a Lac
promoter. Some host strains are incapable of active uptake of the catabolite.

0 LB: LB supplies complex forms of all the needed nutrients including nitrogen, but
continuous feeding of this source can lead to the buildup of slowly metabolized, toxic
components.

0 Other sources: Fructose, maltose, lactate, acetate, ethanol, complex oils.

The following procedure explains how to set-up a feed vessel. Besides the feed source,
these vessels are also used to pump in acid, base, non-autoclavable components and the
culture.

Supplies
1 ft of medium diameter tubing

1 ft of small diameter tubing

500 mL screw top glass jar

Circular metal plate with inlets and outlets

Cylindrical screw top for sealing the metal plate to the glass jar
Tubing clamp

Small air ﬁlter
Tin foil
Plastic tubing adapter

Procedure
1) Thoroughly wash and rinse the glass jar and its top.
2) Rinse water through the tubing to make sure that it is not plugged or dirty.
3) Cut a length of medium diameter tubing that is slightly longer than the length of the
glass jar. Attach it to the outlet on the underside of the metal plate.
4) Attach a foot long piece of medium diameter tubing to the outlet on the top of the
plate that is in-line with the tubing that was already connected.

5) Cut a short length of medium diameter tubing. Attach an air ﬁlter to one end. Attach
the other end to the opening on the top of the plate that is not in line with the piece of
tubing that was already connected.

6) Add the feed solution if it is to be autoclaved in the vessel rather than being supplied
from a stock solution after autoclaving.

7) Thread the tubing and ﬁlter through the screw top and lightly screw the assembly to
the top of the bottle.

8) Use a tubing adapter to attach the foot long section of smaller diameter tubing to the
medium diameter tubing attached to the head plate.

0 For best control of the rate of feed addition, this tubing should have a bore size of
1.0 mm. This smaller tubing allows more precise control of the feed. If precise
control is not needed or the vessel will be used to add culture media, larger
diameter tubing can be used. The mixed assemblage of larger and smaller
diameter tubing is required because it can be very difﬁcult to get the smaller
diameter tubing attached to the metal ports of the feed vessel.

9) Clamp the outlet to the tubing.

10) Wrap the clamp, the outlet of the tubing, and the ﬁlter in aluminum foil.

11) Place the vessel(s) in an autoclave tray.

0 Wrap the end of the tubing around the top of the jar so that outlet is not lying in
condensed steam at the end of the autoclave cycle.

12) Autoclave for at least 20 minutes as described in AUTOCLAVING on pg. 92.

13) Remove from the autoclave and tighten the screw top.

65

LOADING PRESTO CONTROL

The following protocol demonstrates a quick procedure for starting Presto Control. The

controller is needed for attachment of the feed tubing and calibration of the DO probe.

Three components of the system have to be operating correctly for everything to be

working:

1) The snap control unit and I/O modules must be operating and being sent signals from
the BioFlo 11c and its probes.

2) The control unit must have the correct OptoControl strategy downloaded and in
“Run” mode and it must be communicating with the computer.

3) The computer must have the correct OptoDisplay Project loaded for communicating
with the OptoControl program downloaded to the control unit.

The Quick Start procedure starts Presto Control if the correct strategy and display

program have already been loaded. If the controller or display programs are incorrect or

they are not communicating properly, use the Full Load procedure.

Time: 2 min

uick Start Procedure
1) Open the “Shortcut to OptoDisR” icon on the desktop. The Presto Control display
screen should open.
2) Event Log Viewer should list:
0 Com Port: snap=>Open Primary
0 Controller: snap=>Attaching to Scanner
3) Close the Event Log Viewer.
0 If the program opens to a screen displaying a green stop button at the top, a recipe
is already started. Click on “Stop” to stop the recipe and exit to the set-up screen.

Full Load Procedure
1) Click on the “OptoControl” icon on the desktop to start OptoControl.
2) Click on “File”, “Open Strategy”, and select the strategy located at
D:\PrestoControl\ControllerProgram\PrestoController.
3) Select “Download Strategy” from the “Controller” menu. Wait for the control
program to download.
4) Select “Debug” from the “Mode” menu.
5) Select “Play” from the “Debug” Menu
6) Exit OptoControl.
7) Click on the “Shortcut to OptoDisC” to start OptoConﬁgurator.
8) Click on “File”, “Open Strategy”, and select the open the project located at
D:\PrestoControl\DisplayProgram\PrestoDisplay.
9) Click on “File”, and “Save Project and Load Runtime”.
10) Event Log Viewer should list:
0 Com Port: snap=>Open Primary
0 Controller: snap=>Attaching to Scanner
11) Close the Event Log Viewer.

66

CONTROL PREPARA TI ON

The following sections describe how to set up the BioFlo 11c for control operations. The
procedures can be done at any time before inoculation. In general however, there is no
need to have agitation, pH, or temperature control of the bioreactor while ionization of
the DO probe is occurring.

Mixing

Agitation is controlled directly by the BioFlo He. The Presto Control program can only
monitor the RPM. Two modes of control exist: manual RPM set-point, and PID control
by DO set point. When the temperature or pH control is active, the agitation should be
set to at least a minimum value in order to maintain mixing in the bioreactor. This
section describes setting the system for manual set-point control of agitation for mixing
purposes. The agitation can be set between 25 and 1000 RPM.

Supplies
Motor

Hex head wrenches
Time: 1 min

Procedure
1) Place the servo motor on top of the drive shaft and make sure that the multi-jaw
coupling is linking the drive shaft and the motor.

0 Over time, the socket on the top of the drive shaft may drop too low. If the
coupling is no longer connecting with the drive shaft, use a hex head wrench to
raise the socket.

2) Connect the motor’s control cable to the BioFlo He.

3) Turn the Selector knob to “RPM” and the Mode knob to “Set Point”.

4) Using the Increase/Decrease switch to adjust the set point to 50 rpm.

5) Turn on the motor using the “Motor” switch on the front of the BioFlo IIc console.

Temperature Control

The temperature of the bioreactor is monitored by a resistance temperature detector
inserted into the therrnowell in the head plate and controlled by a heating/cooling jacket
cradling the bottom of the bioreactor. The default operation of the temperature controller
is to recirculate the water through the jacket of the bioreactor. When the temperature is
too low, a heater is turned on to warm the water circulating through the jacket. When the
temperature is too high, the BioFlo IIc stops recirculating the jacket water and allows
fresh cooling water to enter from the water lines. The process is controlled by a PID
controller with a set point that can range between 20 °C and 60 °C.

Su lies

Resistance temperature detector
Glycerin or water

67

Time: 1 nrin

Procedure '
1) Insert the temperature probe into the thermowell on the head plate and connect the
cable to the BioFlo HC.

0 Heat transfer to the temperature probe can be improved by partially ﬁlling the
thermowell with water or glycerin.

2) Turn the Selector knob to “Temperature” and turn the Mode knob to “Set Point”.
3) Use the Increase/Decrease switch to adjust the set point.
0 Suggested: 37 °C is the standard growth temperature for E. coli.
4) Turn the Mode knob to “Control”.
5) Connect the upper water lines to the exhaust gas condenser.

0 The top line goes to the top port for each set of connections. The top lines are
always connected Inst.

6) Attach the lower water lines to the metal bioreactor heating jacket.

0 Do not attach the water lines unless the temperature probe is connected and
operating properly. The heating can become uncontrolled and the water jacket
can become very hot.

7) Prime the recirculation loop for 30 seconds by holding down the Prime switch on the
right side of the BioFlo IIc.

pH Control

The pH is controlled directly by the BioFlo IIc. When the pH is outside of the dead band
established by the set point, either the base or acid pump is set to operate until the pH is
returned to the set point. With large diameter tubing and/or highly concentrated acids or
bases, the control parameters of the BioFlo He can lead to overshoot and a cycle of
alternating acid and base addition. This leads to a buildup of the salt concentration in the
media and can cause culture death. To prevent this, only put one pump in control of
maintaining the pH. A minimal media with ammonium as the source of nitrogen leads to
the production of hydrogen ions due to cell metabolism. The pH of this type of media
can be maintained with just the base pump placed on “AUTO”. The acid pump can be
turned off. In addition, if the base being used is concentrated ammonium hydroxide or
ammonia, the culture has a continuously replenishing nitrogen source. Complex media
tends toward basic conditions and can usually be controlled with just the acid pump. 6 N
hydrochloric acid is commonly used. The following procedure should be done while
wearing rubber gloves, full coverings for the arms and legs, and chemical splash goggles.

Supplies
Protective clothing

Acid vessel

Base vessel

Acid

Base

Labeling tape
Permanent marker

68

Time: 10 rrrin

Procedure
1) Add the acid and base to the sterilized feed vessels described in FEED VESSELS on
pg. 64.

2)

3)

4)

5)

6)

Do this in a fume hood. Do not do it in the laminar ﬂow hood. Wear rubber
gloves with gauntlets, a long sleeved lab coat, and chemical splash goggles.
Under no circumstances open an ammonium hydroxide or sulfuric acid container
outside of a fume hood.

Acid and base vessels can be reused in subsequent bioreactor projects. The
solutions do not need to be replaced between each use.

Always keep the tubing outlets above the liquid level of the acid and base
containers and ﬁrmly clamped.

Inspect any plastic tubing adapters to make sure that they are not damaged by
previous acid or base contact.

Label the vessels with contents, name, and date.
Connect the acid and base lines to their ports on the head plate.

The acid and base ports are the small, bent, ridged tubes that are permanently
attached to the head plate. Make sure that the ports and tubing outlets are above
the liquid level of the acid and base containers. Quickly remove the tubing used
to seal the ports during autoclaving and push the acid and base tubing into place.
If possible, attach the tubing before removing the clamps.

Airﬂow through the bioreactor keeps positive pressure in the vessel and limits the
chance of contamination.

If the tubing is being reused from a previous experiment or the clamp is close to
the outlet, attach a second clamp slightly farther up the tubing. Then, scissors can
be used to cutoff the outlet clamp and a sterile portion of the tubing exposed for
attachment to the head plate. Be very careful to protect against any splashing that
may occur due to acid or base caught between the two clamps.

Thread the acid, base and feed lines into their respective pumps.

Determine the rotation direction of the pump head and the amount of tubing
needed to reach the bioreactor from the pump. The tubing with the smallest inner
diameter (1 mm bore size) should be wound through the pump head. Switch the
pumps to “ON” and carefully thread the tubing into the moving pump. If the
tubing gets caught on one of the metal guide pegs, it can be torn. Switch the
pumps to “OFF” when completed.

Prime the tubing.

0 Prime the tubing by switching the pumps to “ON” until drops can be seen coming
from the outlet of the ports.

0 If the pumps do not seem to be able to start the ﬂuid ﬂowing, priming can be
accomplished by either lifting the outlet of the media container above the inlet to
the pump or by pulling the ﬂuid through the tubing with a sterilized syringe
before attaching it to the port.

Attach the pH probe cable to the probe.

69

0 If the pH probe does not screw into the head plate, then the sterile seal is
maintained with O—rings only. Because of this, excessive movement of the probe
can introduce contamination into the vessel. To avoid this, detach the cable from
the BioFlo Ho and screw the cable onto the probe by turning the cable, not the
probe. The other end of the cable can then be connected to the BioFlo Ho.

7) Turn the Selector knob to “pH” and the Mode knob to “Set Point”.
8) Use the Increase/Decrease switch to adjust the set point.
0 A pH of 7.0 is standard physiological pH.
9) Return the Mode knob to Control.
10) Flip the Acid and Base switches on the side of the BioFlo IIc to “AUTO”.

0 When the pH controller has added the necessary amounts of acid and base to
reach its set point, switch “OFF” the pump that is not needed for maintaining the
pH during culture growth.

Feed Pump Calibration

This procedure can be used to determine the exact operating parameters of the nutrient
pump for the BioFlo DC. The pump speeds do not drift so, if the speed is already known
or can be read from the table in Feed Control on pg. 72, this section can be skipped. Feed
control on the BioFlo He is complicated by the fact that the pumps only have one speed
when they are turned. Therefore, the digital feed pump must use an on/off cycle to mimic
an analog rate. Presto Control has been designed to allow the user to determine the
“Pulse Length” and “Pump Calibration” needed for tight control of feed by the digital
mimic. The following procedure is not required if the Presto Control recipe being used is
a batch fermentation or does not utilize open loop or PID control of the feed.

This protocol directly determines two parameters.

1) The maximum feed rate of the pump when it is turned on

2) The minimum length of time needed for the pump to form one full feed droplet.

This 2nd parameter is important because effective feed control requires that at least one
feed droplet is added to the bioreactor every time the feed pump is turned on. Otherwise,
the analog rate might specify the current feed rate to be a certain value even though the
actual feed rate during the current on/off cycle is zero. This can be especially
problematic at very low feed rates when the on/off cycle length is very long. This same
consideration needs to be taken into account when deciding on a length of time for the
“PID On/Off Cycle Length”. If a full droplet takes 2 seconds to form and the “PID
On/Off Cycle Length” is set to 4 seconds, an output of the P11) controller of less than
50% of maximum might allow the pump to cycle on and off without any feed actually
being added. If the “PID Scan Rate” is also set to 4 seconds, the controller would update
its output based on its previous speciﬁed feed rate even though no feed had actually been
added. This would contribute to poor controller performance. Therefore, the feed PID
scan rate and cycle length should be scaled according to the maximum feed rate of the
tubing and the feed requirements of the culture. Alternatively, a better ﬁx for this control
limitation is to attach a small pipette or syringe tip to the outlet of the feed line. This
reduces the amount of feed required to form a full droplet. If the added liquid volume is
not a consideration, the feed concentration can also be reduced.

70

M

10 mL graduated cylinder
1 ft feed tubing

Outlet of feed line

Small Erlenmeyer ﬂask
Water or feed solution

Time: 15 min

Procedure

Pump Calibration
1) Thread 3 1 ft length of feed tubing into the nutrient pump head.

2)

3)

4)

5)

6)
7)

0 This tubing needs to be the same as the tubing used to pump the feed into the
bioreactor. The outlet should also to be the same as the outlet of the feed port into
the bioreactor. Depending on the preparation of the bioreactor, this may be a
metal tube, another piece of plastic tubing, or a pipette or syringe tip.

0 Switch the pump to “AUTO”. In the “Feed” section of Presto Control, set the
“Pump Time” to 30 seconds or more. Set the feed to “Full On” and carefully
thread the tubing into the moving pump. If the tubing gets caught on one of the
metal guide pegs, it can be torn. Leave an equal amount of tubing exiting and
entering the pump head.

Fill the Erlenmeyer ﬂask with at least 10 mL of feed solution or water.

0 If the feed solution is exceptionally viscous, the substitution of water may not be
appropriate.

Place the inlet of the tubing into the solution in the Erlenmeyer ﬂask and the outlet of

the tubing into the mouth of a 10 mL graduated cylinder.

Prime the tubing.

0 Set the “Pump Time” to 2 or 3 rrrinutes. Prime the tubing by setting the feed to
“Full On” until drops can be seen coming from the outlet of the tubing. Set the
feed to “Full Off”.

0 If the pump does not seem to be able to start the ﬂuid ﬂowing, priming can be
accomplished by either lifting the liquid level of the Erlenmeyer ﬂask above the
inlet to the pump or by pulling the ﬂuid through the tubing with a syringe.

Determine the maximum feed rate of the pump.

0 Set the “Pump Time” to 120 or more seconds. Set the feed to “Full On”. When
the feed timer has stopped, calculate the maximum feed rate of the Nutrient pump
based on the “Pump Time” and the volume of feed added to the graduated
cylinder.

Click on the orange “Control” button.

Click on “Pump Calibration” and enter the calculated maximum feed rate of the

nutrient pump.

Pulse Length

1)
2)

Maintain the same preparation as used for calculating the “Pump Calibration”.
Click on “Pulse Length” in the FEED CONTROL section of the main window and
enter 200 ms.

71

3) In the same window, click on “Fraction of Time On” and enter 0.5 or an alternate
desired value.

4) Click on feed time and enter 60 seconds or more.

5) Click on “Feed Pulse Control”.

0 The nutrient pump will turn the pump on for the time speciﬁed in “Pulse Length”
and turn it off for a length of time speciﬁed by “Fraction of Time On”. For
example, the pump would be on for 0.2 s and off for 0.6 s when the “Pulse
Length” equals 200 ms and “Fraction of Time On” equals 0.25.

6) Increase the “Pulse Length” in increments until at least one full liquid droplet is
produced from the end of the tubing outlet.

0 For a very small pipette syringe tip, this may be as little as 200 ms. Small bore
size tubing may produce one droplet every 1 second. Large metal tubing may
require 2 seconds.

7) Click on the orange “Control” button.
8) Enter the new “Pulse Length” in the Feed Control section of the pop—up window.
9) If desired, enter a corresponding “PID On/Off Cycle Length” and “PID DO Scan

Rate”.

0 Assuming that the PID feed control output will demand a feed rate of at least 20%
of the maximum rate, the “PID On/Off Cycle Length” should have a value of 5
times the “Pulse Length”. This will allow at least one drop to be formed each
time the pump switches on during PID feed control.

Feed Control

The feed should be consistent with the carbon source chosen from MEDIA on pg. 85. Do
not add more feed solution to the feed vessel than is needed for culture growth or can be
added to the bioreactor without fear of overﬂowing. A controller problem can lead to
uncontrolled overfeeding.

Supplies
Feed

Feed vessel
Labeling tape
Permanent marker

Time: 10 min

Procedure
1) On the main screen of the Presto Control program, click on the orange “Control”
button. Make sure that the current parameters are set to the desired values.

0 If tight control of feed is desired, and the minimum “Pulse Length” and exact
“Pump Calibration” have not been determined, check Feed Pump Calibration on
pg. 70 for the procedure.

0 Suggested “Feed” control parameters:

0 Pulse Length: 1000 ms
0 Feed Concentration: 0.50 g/mL
0 Pump Calibration: 1.41 mIJmin

72

PID On/Off Cycle Length: 5000 ms
PID DO Scan Rate: 5.0 5
Maximum Volume: 200 mL
Proportional: -1.00
Integral: 1.00
0 Derivative: 0.00
2) If applicable, transfer the feed from the stock feed bottle to the feed vessel.

0 Take the bottle of sterilized stock feed solution to the laminar ﬂow hood and open
it. Use the Bunsen burner to ﬂame the top of the bottle. Unscrew the cap of the
feed vessel and ﬂame its top. Pour the desired amount of feed from the feed stock
bottle to the feed vessel while attempting to keep the feed from dripping down the
sides of the bottle. Reﬂame the lips of the containers and replace the tops.

3) Label the vessel with contents, name, and date.
4) Connect the feed line to its port on the head plate.

0 Use the same technique deseribed for attaching the pH tubing.
5) Thread the feed line into the nutrient pump head.

0 Determine the direction of rotation of the pump head and the amount of tubing
needed to reach the bioreactor from the pump. The tubing with the smallest inner
diameter should be wound through the pump head. Switch the pump to “AUTO”.
In the “Feed” section of Presto Control, set the “Pump Time” to 30 seconds or
more. Set the feed to “Full On” and carefully thread the tubing into the moving
pump. If the tubing gets caught on one of the metal guide pegs, it can be tom.

6) Prime the tubing.

0 Set the “Pump Time” to 2 or 3 minutes. Prime the tubing by setting the feed to
“Full On” until drops can be seen coming from the outlet of the port. Set the feed
to “Full Off”.

0 If the pumps do not seem to be able to start the ﬂuid ﬂowing, priming can be
accomplished by either lifting the outlet of the media container above the inlet to
the pump or by pulling the ﬂuid through the tubing with a sterilized syringe
before attaching it to the port.

00000

Air Control

Two gas sources exist for Sparging into the bioreactor and the desired source is selected
by a three way valve attached to the top of the BioFlo He. The ﬁrst source comes from a
nitrogen cylinder that is controlled manually. The second source is pressurized
environmental air controlled by an airﬂow regulator in conjunction with Presto Control.
Both have to pass through a manual regulator on the front of the BioFlo 110. This valve
must always be Opened far enough so that the added ﬂow resistance of the valve does not
prevent the controller from reaching the maximum desired airﬂow output.

Time: 2 min

Procedure
1) Click on the orange “Control” button. Make sure that the current parameters are set
to the desired values.

0 Suggested “Air” control parameters:

73

PIT) DO Scan Rate: 5.0 5

Minimum Air Flow Rate: 0.05 Umin

Maximum Air Flow Rate: 1.05 Umin

Proportional: 1.00

Integral: 5.00

0 Derivative: 0.00

2) Uncover the air inlet and exhaust ﬁlters and attach their respective lines.
0 Make sure that the tubing is not kinked on either of the air lines.

3) Twist the air ﬂow source valve attached to the t0p of the BioFlo IIc to the right so that
it selects for “air”.

4) Make sure that the air ﬂow regulator on the front of the BioFlo He is open.

5) Check to make sure that the airﬂow regulator attached to the air line on the south wall
has pressure and that the valves are open.

6) Test various “Air Flow Rates” in the feed section of Presto Control to make sure that
the controller can reach and maintain the chosen air ﬂow.

<><><><><>

DO Probe Calibration

The electrolyte solution in the DO probe must be ionized before it can be calibrated.
Therefore, the probe must be connected to the BioFlo BC with its power turned on for at
least 6 hours before inoculation. The output of the DO probe depends heavily on the
temperature of the media, so the probe should be calibrated at the operating temperature.
The zero point oxygen concentration is set using nitrogen saturated media. The 100% D0
concentration value is set using air saturated media at the maximum mixing conditions.

Time: 15 min

Procedure

1) Turn the Selector knob to “Temperature” and the Mode knob to “Set Point”.

2) Use the Increase/Decrease switch to change the set point to the operating temperature.

3) Switch the Selector knob to “RPM” and the Mode knob to “Set Point”.

4) Increase the agitation speed to the maximum process value.
0 The maximum suggested agitation rate is 750 RPM.

5) Select nitrogen as the inlet gas by turning the 3-way valve on top of the BioFlo 110 so
that it points left.

6) Open the main valve on the nitrogen cylinder.

7) Open the smaller, pressure control valve by twisting it clockwise until the line has
approximately 15 lb*ft of pressure

8) Set the ﬂow rate of nitrogen entering the vessel to 1 Umin using the manual ﬂow
regulator on the front of the BioFlo Dc.

9) Wait for the DO reading on the BioFlo 110 to stabilize - it should be near to 0%.
0 Stabilization takes a few minutes.
0 If nitrogen is not available or the DO probe needs to be quickly recalibrated

without nitrogen sparging, the zero point can be established by unplugging the
DO probe from the BioFlo HC.
10) Turn the Selector knob to “DO” and the Mode knob to “Zero”.
11) Push the Increase/Decrease switch once to adjust the display reading to 0%.

74

12) Turn off the main valve on the N2 cylinder.

13) When the gauges on the regulator drop to 0 psi, shut the secondary valve.

14) Turn the 3—way valve to point to the right so that it selects for environmental air.

15) Click on “Airﬂow Set Point” in Presto Control to set the maximum air sparge rate
that will be used in the control recipe.

0 The maximum suggested air ﬂow rate is 1 Umin.

0 Open up the manual air ﬂow regulator on the front of the BioFlo He by giving it
an extra counter-clock wise turn. This reduces the restriction on the airﬂow and
allows the controller to reach its maximum air ﬂow.

16) Wait for the DO reading on the BioFlo He to stabilize - it should be near 100%.

17) Turn the mode knob to “Span”.

18) Use the Increase/Decrease switch to adjust it to 100%.

19) Click on “Airﬂow Set Point” in Presto Control and enter “0” into the popup window
to stop the air ﬂow.

20) Decrease the agitation to 50 rpm by moving the selector knobs to “RPM” and “Set

Point” and press the decrease button until desired value is reached.

Agitation Control

When in D0 control mode, the BioFlo IIc sets the agitation rate based on the DO set
point entered into the BioFlo IIc console. Presto Control cannot control or monitor the
DO set point used for control of agitation. Therefore, the independent control processes
of the BioFlo He can interfere with some of the staged controller recipes of Presto
Control. Unless the DO set point control is manually switched off on the BioFlo He, the
BioFlo He continues to correct for errors in its DO set point by changing the agitation.
This can conﬂict with PID and open loop control of the airﬂow or feed by Presto Control.
The rate of change of agitation through the BioFlo BC is so slow, however, that it rarely
interferes with the normally faster response of the PID controller used in Presto Control
or conﬂicts with the detection of a DO spike monitored in some controller schemes. A
large problem can result, though, if the DO set point of the BioFlo He is accidentally set
to a different value than the DO set point in Presto Control.

A few methods can be used to make sure that the lack of interaction between Presto
Control and the agitation controller does not become a problem. The ﬁrst method is
already a part of the control program. The RPM value reported to Presto Control is
greater than the actual reading taken by the BioFlo IIc. Therefore, if a control recipe calls
for a maximum agitation rate of 750 RPM before switching to the next stage, and the
maximum agitation rate set on the BioFlo He is 750 RPM, it is still possible for a stage
switch to occur. If this signal offset had not been implemented, signal drift could cause
the RPM value reported to Presto Control to become smaller than the actual reading taken
by the BioFlo Hc. Therefore, 750 RPM would never be detected and the stage switch
could never occur. A second method of reducing controller conﬂict is implemented by
just switching off the DO control mode on the BioFlo IIc console when the RPM has
reached its maximum value. The third method is to set the DO set point on Presto
Control slightly lower than the set point on the BioFlo IIc. Therefore, when the agitation
has reached its maximum value, the switch to PID control of airﬂow or feed maintains

75

the DO below the DO set point for agitation control and forces the agitation to remain at
its maximum value.

If the Presto Control recipe being used does not rely on ramping of the agitation rate, this
next part can be skipped. Instead, just set the desired, constant agitation rate as described
in Mixing on pg. 67.

Time: 1 min

Procedure

1)
2)

3)
4)

5)

6)

Turn the Selector knob to “DO” and the Mode knob to “Set Point”.

Use the Increase/Decrease switch to adjust the set point to the desired value.

0 Use the same or slightly higher value used in Presto Control. The minimum set
point to ensure aerobic growth is 10%.

Turn the Selector knob to “Agitation” and the Mode knob to “Set Point”.

Use the Increase/Decrease switch to adjust RPM to the desired minimum value.

0 Suggested: 50 rpm.

Press the DO Control Active button. The DO Control Active light will come on.

0 Once this control has been set, do not press the button again while agitation is
ramping up or it will force the RPM to immediately go to its maximum value.

Use the Increase/Decrease switch to adjust the set point to its maximum value.

0 750 RPM is the maximum suggested agitation rate.

0 The agitation will ramp up slightly while the maximum value is being set, but it
will slowly settle back down to its minimum value if the D0 is above its set point.

76

MEDIA ADDITIONS

The salts, antibiotics, and trace minerals that can not be autoclaved with the buffered
media in the bioreactor need to be added individually. The technique for these additions
varies depending on the circumstances. Three methods are listed here.

Method I

If the tubing and probes have not been connected to the BioFlo IIC and none of the salts,
amino acids, or antibiotics degrade quickly, the media components can be added in the
hood after the bioreactor has cooled down and before the DO probe has been booked up
for ionization.

Time: 5 min

Su lies

Media components

Sterilized pipette tips

Pipetter

Sterilized graduated cylinders (optional)

Procedure

1) Move the bioreactor into the laminar ﬂow hood.

2) Unscrew the inoculation part, one of the smaller ports, or remove the tubing covering
the feed ports.

3) Use sterilized pipettes or sterilized graduated cylinders to add each of the media
components.

4) Replace the port covers or tubing and move the bioreactor back to the BioFlo IIc
stand.

Method 2
If the media components are in small volumes, they can be added through a syringe ﬁlter
into one of the ports.

Time: 5 min

Su lies

Media components

Rinse water

Sterilized pipette tips
Pipetter

Sterile 0.2 um syringe ﬁlter

Syringe

Procedure
1) Unclamp the piece of tubing covering one of the open ports.

77

2) Quickly attach a small 0.2 pm syringe ﬁlter onto the end of the sterile tubing attached
to the port.
0 The syringe ﬁlter can be attached before autoclaving in order to reduce the chance
of contamination.
3) Screw the end of a 3 or 5 mL syringe onto the ﬁlter.
4) Remove the plunger.
5) Pipette the desired quantity of one of the components into the syringe body.
6) Replace the plunger and force the media component into the bioreactor.
o If the volume being added is very small, water can be used to ﬂush the solute
through the ﬁlter.
0 Antifoam cannot be passed through a ﬁlter.
7) Repeat steps 4 through 6 for each component.

Method 3

If a large volume needs to be added and none of the components might come out of
solution due to interactions at higher concentrations, they can all be pumped into the
bioreactor together.

Time: 10 min

Supplies

Media components

Sterilized pipette tips

Pipetter

Sterilized graduated cylinders (optional)
Sterilized feed vessel

Procedure

1) Move all of the media components and a sterilized feed vessel into the laminar ﬂow
hood.

2) Unscrew the top of the feed vessel and ﬂame the glass lip.

3) Add each of the media components to the feed vessel.
0 Either aseptically pipette each component from a stock feed solution or dissolve

the solutes and sterile ﬁlter them into the vessel using a syringe and ﬁlter.

4) Flame the lip of the feed vessel and replace its top.

5) Move feed vessel to the BioFlo IIc stand.

6) Connect the feed tubing to an open port on the head plate.
0 Use the same technique described for attaching the pH tubing.

7) Thread the feed tubing into the antifoam pump head.
0 Use the same technique described for the pH tubing.

8) Set the antifoam pump to “ON” and wait for all of the media to be added to the
bioreactor.

9) Switch the antifoam pump to “OFF”.

78

INOC UIAT ION

The inoculation process can cause a stress reaction in the culture or lead to contamination
of the bioreactor. The following method describes the best way to keep the bioreactor
from becoming contaminated. Unfortunately, it also allows the culture to become starved
of oxygen as it is being pumped. The best way to prevent oxygen starvation is to take the
inoculum culture directly from the incubator, unscrew the large inoculum port of the
bioreactor, and dump the culture directly into the bioreactor. It is also the best way to
cause contamination.

It is important to keep the cell density of the culture within the optimum bounds of the
exponential growth phase. Allowing the culture to stagnate too long at too high of a cell
density can lead to a long lag phase, or poor growth and production at the beginning of
the run. In contrast, a low inoculating cell density can lead to a lengthy lag phase.

If the initial feed source for the culture is added by the ﬁrst stage of the Presto Control
control scheme, the control scheme should be started while the inoculum is being pumped
into the bioreactor.

Supplies
Inoculum

Sterilized feed vessel
Time: 10 min

Procedure

1) In the laminar ﬂow hood, withdraw up to 1 mL of culture and pipette it into a cuvette.

10) Check the OD600 of the sample according to CELL DENSITY - OD600 on pg. 99.
0 The culture should have an OD600 between 0.8 and 2.0.

2) If the OD500 is within the proper range, transfer the culture to a sterile feed bottle
described in FEED VESSELS on pg. 64.

3) Connect the inoculum tubing to an open port on the head plate.
0 Use the same technique described for attaching the pH tubing.

4) Thread the feed tubing into the antifoam pump.
0 Use the same technique described for the pH tubing.

5) Set the antifoam pump to “ON” and wait for all of the inoculum to be added to the
bioreactor.

79

BI OREA C T OR GROWTH

The best control scheme for a project depends on the speciﬁc requirements of the culture
for growth and production. The control scheme recipes available for download are
described in the Presto Control Manual along with a full description of program
Operations. Individualized recipes can also be designed and saved using Presto Control.

Ideally, once the control scheme is started, the operator should no longer need to make
changes to the system except during induction or antifoam additions. Culture conditions,
however, should be monitored at least every 6 hours. Check to make sure that the D0 is
being maintained at the desired set-point. Also, verify that the agitation, airﬂow, and
feed proﬁles represent what would be expected from the control recipe and exponential
growth. Recognize that the OD600 of an E. coli culture can double in less than an hour.
Therefore, if the oxygen transfer capability of the bioreactor is at 50 % of maximum, it
can become maxed out within an hour. If one of the goals of the project is to keep the
culture in an exponential growth phase, the culture might need to be harvested or induced
close to this time point. The bioreactor should be monitored closely near to the end of the
run.

The length of the growth phase in the bioreactor can vary drastically depending on the
culture and the media and the exact timing cannot be standardized. The cell density
should be checked regularly as described in CELL DENSITY — OD600 on pg. 99. If the
culture needs to be tested for product formation, samples can be taken as described in
CULTURE SAMPLES on pg. 101. Periodically, antifoam needs to be added as
described in ANTIFOAM ADDITION on pg. 82.

Time: 12 hrs -— 48 hrs

Procedure
1) Verify that all of the control parameters are set correctly in the “Control” window.
2) Check to make sure that the feed pump is set to “AUTO”.
3) Check that all of the set points and control values are at their desired values.
4) If applicable, check to make sure that the BioFlo 110 D0 control light indicates that
the RPM ramp is active.
5) Click on the desired control strategy recipe to download (both parts if applicable).
0 For induced protein expression in recombinant E. coli, use Standard Fed Batch
Control.
0 The screen will freeze for about 1 minute while downloading.
6) Press the “Start” button to begin the recipe.
0 A dialog box will appear listing the name of the logging ﬁle. Record this name.
7) Make any online changes that are needed for the recipe. '
8) Monitor culture growth.
0 Check the OD600 and the environmental conditions at least every 6 hours.
Monitor the culture more regularly as the cell density increases.
0 Take culture samples if necessary.

80

9) When the culture has achieved the desired cell density, induce according to
INDUCTION on pg. 103.

81

AN T IF 0AM ADDITION

As the culture grows, foam is sometimes produced on top of the liquid. If this foam
enters the exhaust ﬁlter it can block the air ﬂow and lead to a pressure build up in the
vessel. This changes the oxygen transfer characteristics of the system and it can
negatively impact culture growth. Also, if the foam leaks past the seals of the bioreactor,
it can allow culture to escape or lead to contamination of the bioreactor. The addition of
antifoam dissipates this buildup and prevents these problems. Antifoam is an oil that
reduces the surface tension of the media. It is standard practice to add a few drops of
antifoam every six hours or whenever the cell density is measured. If the culture
produces high levels of foam it might need treatment more often.

Supplies
Sterilized antifoam agent

Feed vessel
Labeling tape
Permanent marker

Time: 5 min

Procedure
1) If needed, transfer sterilized antifoam to a sterile feeding bottle with the appropriate
tubing and connections.

0 The antifoam vessel can be reused in subsequent bioreactor projects. The
antifoam does not need to be replaced between each use.

2) Label the antifoam bottle with name, date, and contents.
3) Connect the antifoam tubing to an open port on the head plate.

0 Use the same technique described for attaching the pH tubing.
4) Thread the tubing into the antifoam pump.

0 Use the same technique described for the pH tubing.

5) Prime the tubing by setting the antifoam pump to “ON” and wait for drops to form at
the port outlet.
6) Let a few drops fall into the media.

0 A single drop should noticeably reduce foaming.

0 Addition of antifoam affects the surface tension of the media and will
immediately cause a large decrease in dissolved oxygen. An increase of 100 rpm
or 0.1 Umin of airﬂow may be required to maintain a constant DO concentration
after antifoam addition. If the current stage in the controller recipe is set to detect
a sudden change in dissolved oxygen or any of the process control parameters, the
stage may need to be put on manual hold and the expected response monitored
closely until the PID controllers have adjusted to the change.

7) Flip the antifoam switch back to the “OFF” position.
8) If the foam level does not drop within a few seconds, repeat the procedure.
9) Repeat whenever a time point is taken or foam forms.

82

BIOREA C TOR SH U TDO WN

If the culture needs to be harvested, shutdown of the bioreactor and harvesting of the cells
should be performed as quickly as possible. Without feed, airﬂow, or agitation, the
culture might exhibit a stringent response that can negatively affect the product. This
entire procedure should be done wearing hand and eye protection. Furthermore,
disconnection of the acid and base vessels should be done with splash goggles, rubber
gloves, and covering for the arms and legs.

Everything that has come into contact with the culture must be sterilized. This can be
accomplished by either soaking it in 10 % bleach or autoclaving it. Preferentially, large
liquid volumes should be autoclaved, but the bioreactor and unclean glass vessels should
not be autoclaved without liquid in them.

Supplies
Large and small ﬂasks and beakers

10 % bleach solution
RO water bottle
Clamps

Soap and brushes

Time: 1 hr

Procedure

1) Click the “Stop” button in Presto Control. This will stop the control recipe and stop
data recording. _

o The log of bioreactor data can be accessed in the folder D:\ Presto
Control\Controller Program\Fermentation Logs. It will be named according to the
date it was recorded.

2) Turn off the Power switch to the BioFlo HC.

3) Shut off the water supply to the unit.

4) Close the air/nitrogen valve by turning it so that it points away from the front of the
BioFlo lie.

5) Use clamps to close off the tubing to the acid, base, antifoam, and glucose feeds.

6) Disconnect the tubing from the bioreactor.

0 Use the proper protection for handling the acid and base lines.

0 Keep the outlets of all of the tubes above the liquid levels of the vessels.

7) Remove the motor and place it on its stand on top of the BioFlo IIc console.

8) Disconnect the pH and DO cables and remove the temperature probe from the
thermowell.

9) Remove the pH and DO probes. Rinse them with R0 water. It may be necessary to
lightly scrub them to remove culture and wipe them with 70% ethanol solution to
remove culture. Return them to their stands.

0 The pH probe should be stored with it tip covered in a KCl solution or a buffer
solution with a pH between 4.0 and 7.0.

83

10) Disconnect the water lines from the heat exchanger and the exhaust condenser.

0 Remove the bottom connector ﬁrst, followed by the top.

11) Disconnect the air inlet line.

12) Carry the bioreactor to a nearby sink or to the wash basin. Set it in the basin and
unscrew the head plate.

13) Remove the head plate and set it in the sink. Rinse it with a 10% bleach solution to
kill the culture on it.

14) Remove the air inlet and exhaust ﬁlters. Dispose of them in a biohazard bag or save
them for future use.

15) Carefully pour the culture into a large beaker.

16) If the cell culture is to be processed for product, see the HARVESTING CELLS
protocol on pg. 105.

17) If the culture is not going to be harvested, it needs to be sterilized. This can be done
using one of two methods. The ﬁrst method is the best option.

0 Method 1: Cover the beaker containing the culture with aluminum foil.
Autoclave the culture according to AUTOCLAVING on pg. 92. Any extra liquid
cultures, ﬂasks, or test tubes can also be sterilized in this manner.

0 Method 2: Pour the culture into a bucket and place it in the walk-in hood in 3269
or in a fume hood. Make sure the hood fan is on. Add ~500 mL of bleach and let
the culture stand for at least 30 rrrinutes. Disposing of the culture in this way
produces poisonous chlorine gas, so it must be done in the hood.

18) Return the bioreactor to the basin. Rinse it with a 10% bleach solution to kill any
remaining cells. Let the bleach contact it for 10 minutes.

19) Wash the head plate and bioreactor and any feed vessels thoroughly with detergent
and tap water. Be sure to scrub the impeller shaft, impellers and bafﬂes. Rinse all
ports (i.e. air inlet, sparger, acid, base, antifoam, glucose, condenser and sampling
port) and tubing by running water through them.

20) Rinse the head plate and bioreactor several times to remove any remaining detergent.

21) Give all items two ﬁnal rinses with R0 water.

22) Reinstall the head plate and return the bioreactor to the BioFlo IIc console or to the
designated storage area.

23) If applicable, remove any autoclaved culture containers from the autoclave after the
autoclave has cooled down. Pour the liquids into the drain and wash and rinse the
containers.

SUPPLEMENTARY MATERIAL

MEDIA

The following is a list of commonly used culture media types and stock solutions. See
the review by Zhang and Greasham ( 1999) for a discussion of other types of media and
the nutrient needs of microorganisms. The preparation procedures group the media
components that can be mixed together before autoclaving without fear of degradation or
reaction. The components listed after the autoclaving step should be added after the bulk
media has cooled. The given amounts are the quantities that would be required to
prepare the 1 liter of media used in the BioFlo IIC. The recipes should be scaled down as
appropriate for producing agar plates and inoculum media, but do not mix and autoclave
incompatible components.

M9

E. coli are facultative microorganisms and can produce nearly every component of their
cellular structure from a nitrogen, oxygen, and carbon source. M9 (minimal media)
supplies these components in their simplest form and is the cheapest of the growth media.
The carbon source is glucose. Nitrogen is supplied via ammonium chloride and its
uptake by the cells causes hydrogen ion formation. Therefore, potassium phosphate and
sodium phosphate are used as buffering agents. Sodium chloride increases the ionic
strength of the media and magnesium sulfate provides the trace mineral magnesium. Not
all strains of E. coli are able to grow on basic minimal media. In such cases, other trace
minerals and nutrients can be added.

Composition

1 g/L NH4C1

6 g/L NazHPO4
3 g/L KHzPO4
0.5 g/L NaCl

4 g/L glucose
2 mM MgSO4

Preparation (1 L)

200 ml 5X M9 salts

790 Hi] deionized H20

20 g agar if creating plates
Autoclave in bioreactor or ﬂask

Cool to 60°C

Sterilized 8 ml 50% glucose
Sterilized 2 ml 1 M MgSO4
Antibiotics

Immediately pour plates if applicable

85

Reference: pET System Manual (2002)

Modiﬁed M9
This is an example of a modiﬁed minimal media. The following additions to M9 are
recommended when culturing JM 109.

Composition

M9 media

0.1 mM CaClz

1 mM thiamine-HCI

Preparation (1 L)

Standard M9 preparation

Sterilized 0.1 mL 1M CaClz
Sterilized 1 mL 1 M Thiamine-HCI
Immediately pour plates if applicable

Reference: Yanisch-Perron et al. (1985)

LB

LB is the most common complex media used for cell culturing. It relies on processed cell
material and protein supplements to supply nutrients, and extra buffering components and
salts are not needed. Cultures grow better and faster on LB than on minimal media.
Unfortunately, the media is more expensive and does not allow tight control of amino
acid, protein, or sugar concentration for purposes of labeling and protein expression.

Composition

10 g/L Bacto tryptone

5 g/L Yeast extract

10 g/L N aCl

Preparation (1L)

0 200 mL 5X LB

0 800 mL deionized H20

0 20 g agar if creating plates

0 Autoclave in bioreactor or ﬂask
0 Cool to 60°C

0 Antibiotics

0 Immediately pour plates if applicable

Reference: pET System Manual (2002)
TB

Terriﬁc broth is a variation of LB that is designed to improve growth in certain cultures
by adding more carbon and nitrogen sources.

86

Composition

12 g/L Bacto tryptone
24 g/L Yeast extract
4 ml glycerol

2.31 g/L KH2P04
12.54 g/L KH2P04

Preparation (1 L)

900 ml deionized water

12 g Bacto tryptone

24 g Yeast extract

20 g agar if creating plates
Autoclave in bioreacotor or ﬂask
Cool to 60°C

Sterilized 5 ml 80% glycerol
Sterilized 100 ml sterile K phosphate
Antibiotics

Immediately pour plates if applicable

Reference: Sambrook and Russel (2001)

Deﬁned Media

This media was designed to allow isotopic labeling of speciﬁc amino acids in proteins
expressed by E. coli. In order to suppress amino transferase and proteinase activity of the
culture, the media supplies excess concentrations of most of the amino acids and
metabolites needed by E. coli for growth and protein expression. For strains exhibiting
poor growth on M9, this media is also a good reference for possible additions to minimal
media that might improve metabolism.

Composition

0.50 g/L alanine

0.40 g/L arginine

0.40 g/L aspartic acid
0.05 g/L cystine

0.40 g/L glutamine
0.65 g/L glutarnic acid,
0.55 g/L glycine

0.10 g/L histidine

0.23 g/L isoleucine
0.23 g/L leucine

0.42 g/L lysine hydrochloride
0.25 g/L methionine
0.13 g/L phenylalanine
0.10 g/L praline

2.10 g/L serine

87

0.23 g/L thrionine
0.17 g/L tyrosine
0.23 g/L valine

0.50 g/L adenine

0.65 g/L guanosine
0.20 g/L thymine
0.50 g/L uracil

0.20 g/L cytosine
1.50 g/L sodium acetate
1.50 g/L succinic acid
0.50 g/L NH4Cl

0.85 g/L K2PO4

2 g/L glucose

4 mM MgSO4

0.01 mM FeCl3

2 mg/L CaC12*2HzO
2 mg/L ZnSO4*7H20
2 mg/L MnSO4*HzO
50 mg/L L-tryptophan
50 mg/L thiamine

50 mg/L niacin

1 mg/L biotin

Preparation

960 mL of water
0.50 g alanine

0.40 g arginine

0.40 g aspartic acid
0.05 g cystine

0.40 g glutarnine
0.65 g glutanric acid,
0.55 g glycine

0.10 g histidine

0.23 g isoleucine
0.23 g leucine

0.42 g lysine hydrochloride
0.25 g methionine
0.13 g phenylalanine
0.10 g praline

2.10 g serine

0.23 g thrionine
0.17 g tyrosine

0.23 g valine

0.50 g adenine

0.65 g guanosine

88

0.20 g thymine

0.50 g uracil

0.20 g cytosine

1.50 g sodium acetate

1.50 g succinic acid

0.50 g NII4CI

0.85 g KzPO4

Autoclave in bioreactor or ﬂask
Cool to 60°C

Sterilized 40 mL 50% glucose
Sterilized 4 mL M gSO4
Sterilized 1 mL 0.01 M FeCl3
10 mL trace minerals
Antibiotics

Reference: Muchmore et al. (1989)
Stock Solutions

80% Glycerol (100 mL)
0 20 mL of RO water in 100 mL graduated cylinder
0 Add glycerol to 100 mL by volume displacement

50% glucose (1 L)

0 Heat and stir 600 mL of RO water

0 Slowly add 500 g of glucose (dextrose)
0 Add water to 1L

0 Autoclave

K phosphate (100 mL)

0 Stir 80 mL of RO water
0 Add 2.31 g KH2PO4

0 Add 12.54 g KH2PO4
0 Add water to 100 mL

0 Autoclave

5X M9 salts (l L)

Stir 900 mL of RO water

Add 5 g of NH4C1

Add 30 g Na2HPO4

Add 15 g KH2P04

Add 2.5 g NaCl

Adjust pH to 7 with 1N NaOH
Add water to 1 L

89

0 Autoclave

5X LB (1 L)

Stir 800 mL of RO water

Add 50 g Bacto tryptone

Add 25 g Yeast extract

Add 50 g NaCl

Adjust pH to 7 with 1N NaOH
Add water to 1L

Autoclave

1M MgSO4 (100 mL)

Stir 90 mL of RO water
Add 12 g MgSO4

Add water to 100 mL
Autoclave

l M IPTG (isopropyl B-D,thiogalactopyranoside) (10 mL)
Open 1 g IPTG container.
Dissolve contents in total 4.2 mL of water

Filter sterilize
Store at —20°C.

0.01 M Fer (100 mL)
Stir 90 mL of RO water
Add 0.16 g FeCl3

Add water to 100 mL
Autoclave

Trace Minerals and Vitamins (100 mL)
Stir 90 mL of RO water
Add 20 mg CaC12*2HzO
Add 20 mg ZnSO4*7HzO
Add 20 mg MnSO4*HzO
Add 500 mg L-tryptophan
Add 500 mg thiamine
Add 500 mg niacin

Add 10 mg biotin

Add water to 100 mL
Filter sterilize

Antibiotics

The use of antibiotics in the media serves two purposes. One is to prevent contamination
by undesired microorganisms. The second purpose is to select for host strains that
contain the plasmids that express the protein. Each plasrrrid encodes for both a

90

component of the protein expression system, and a protein that makes the cell immune to
a speciﬁc antibiotic. Therefore, only the cells containing the expression plasmid(s) can
survive in the growth media. The following is the preparation procedures for certain
antibiotics used in the pET expression system (Novagen, 2002). The needed antibiotics
should be added to the desired concentration at each phase of the culture transfer.
Antibiotics should never be autoclaved; use a 0.2 um syringe ﬁlter and syringe for
sterilization. They should be stored at -20°C once they are in solution.

Carbonicillin (disodium salt) (5 mL)
250 mg in 5 mL deionized water.
Filter sterilize

Store at —20°C

Use at 50 ug/ml (1:1000)

Ampicillin (sodium salt) (5 mL)

125 mg in 5 mL deionized water.
Filter sterilize

Store at -20°C.

Use at 50 ug/ml (1:1000)

Chloramphenicol (5 mL)
170 mg in 5 mL ethanol.
Filter sterilize

Store at —20°C

Use at 34 ug/ml (1:1000)

Kanamycin (sulfate) (5 mL)

0 150 mg in 5 mL deionized water.

Filter sterilize

Store at —20°C.

Use at 30 ug/ml for cells containing kanR plasmids (1:1000)
Use 15 ug/ml for cells with a chromosomal kanR gene (1:500)

Tetracycline (5 mL)

0 62.5 mg in 5 ml ethanol.

0 Filter sterilize

0 Store at —20°C.

0 Use at 12.5 ug/ml (1:1000)

Rifampicin (5 mL)

5 mg in 5 ml 67% methanol, 0.17 N NaOH.
Filter sterilize

Use at 200 ug/ml within 5 days (1 :50)
Protect from light.

91

AUTOCIA VING

All liquids added to the bioreactor should be sterilized. The autoclave can be used for
sterilization if the media components are not temperature sensitive and do not react at
high temperatures. Media that degrades in the autoclave or small volumes of solutions
can be sterilized by passing them through a 0.2 pm ﬁlter. The autoclave is also used to
kill cell cultures so that they can be poured down the sink or thrown out with the trash.
Some points:

0 Glucose has a tendency to turn media brown when autoclaved with other media

components.

0 Certain salts (i.e. MgSO4) react with other salts at high temperatures or high
concentrations.

Antibiotics are degraded by heat.

If a trace mineral is to be added to the media but its reactivity is unknown, it is

safer to autoclave it separately from the bulk media and add it after the liquids

have cooled.

DO and pH probes should only be autoclaved if their tips are submersed in media.

Always use autoclave gloves when removing objects from the autoclave.

Do not overﬁll containers or they might boil over.

The “Liquid Cycle” slowly releases the pressure as the autoclave cools. This

prevents liquid from violently boiling out of its container. The “Dry Cycle”

immediately vents the steam at the end of the autoclave cycle and can be used to
sterilize empty vessels.

0 If agar plates or plastic biohazards are being destroyed, they should be placed in
an orange biohazard bag and set in an autoclave tray with water covering the
bottom. The bag should be autoclaved for 45 minutes on the “Liquid Cycle”. It
can be thrown out with the rest of the trash if it is sealed up in a secondary black

garbage bag.
Supplies

Bottles, ﬂasks, or containers
Aluminum foil

Labeling tape

Permanent marker

Procedure
1) Prepare the media or desired stock solutions.

0 If the autoclave is not immediately available, store in the refrigerator to prevent
contamination.
2) Pour the solution into a bottle or Erlenmeyer ﬂask.

0 To prevent boiling over, do not ﬁll the container more than 2/3rds full.
3) Cover the top of the container.

0 If the bottle has a screw top, loosely screw on the cover.

92

0 If the container is a ﬂask, seal the opening by crushing a piece of aluminum foil
around the outside edges of the mouth.

0 The containers cannot be sealed airtight or pressure equilibration cannot occur.

4) Use labeling tape and a permanent marker to label the ﬂask with name, contents, and
date.

5) Set the container(s) in an autoclave tray and place the tray in the autoclave.

6) Flip the Power and Control switches to the On position on the autoclave.

7) Open the cooling water and steam valves, located behind the autoclave.

8) Close the autoclave door and lock it in place by turning the handle all the way to the
right.

9) Set the Steam Sterilize Time to at least 20 minutes and the Steam Dry Time to 00
minutes. Larger media volumes will require more time to fully sterilize.

10) Press the Reset button.

11) Press the “Liquid Cycle” button to start the autoclave process.

12) Wait for the autoclave to ﬁnish.

0 The autoclave will run for the set length of time once it reaches 132 °C. It will
then spend several minutes exhausting the steam. The buzzer will sound when the
exhaust cycle is complete. The buzzer can be stopped by pressing the reset
button.

13) Close the steam valve behind the autoclave.
14) Shut off the valve to the steam supply.

0 If possible, allow the autoclave to continue to cool until the internal temperature is
below 100 °C before opening the door. This prevents excessive volume reduction
of the media due to boil off when the remaining pressure is released by the door
being opened.

0 Do not leave liquid in a closed autoclave with the steam open for more than a
couple of hours or a large portion of the media will evaporate.

15) Open the autoclave door.

0 Some steam will be released when the door is opened. Do not place your head

directly over the top edge of the door while the pressure is being released.
16) Remove the autoclave tray using a pair of heavy autoclave gloves.

0 Check to make sure that the aluminum foil has not been blown off the top of the
containers. If it has, it can be quickly replaced without a strong risk of
contamination.

17) Shut off the cooling water supply.

93

AGAR PLATES

Preparing agar plates is tricky when media components need to be added to the bulk
media after it has cooled. A subjective measurement of the correct time to ﬁnish mixing
the media and pour the plates is when the ﬂask containing the solution is no longer
uncomfortably hot to hold. Condensation might form on the top of the plate during
cooling and storage, so they should be stored upside down. Try not to allow this to drip
down onto the agar and cause mixing of the culture colonies during inoculation and
incubation. The following protocol is taken from the Experiment Manual for the

Biochemical Engineering Laboratory (Worden et al., 2002).

Supplies

Autoclaved media with 2 % agar
Extra components and antibiotics
Sterile Petri dishes

Paraﬁlm or bag

Permanent Marker

Procedure

1) Prepare and autoclave the media in a ﬂask with 2% agar added to the solution.

0 After autoclaving, the solutions may be cooled by partially submerging the ﬂasks
in cold water. The agar solution must not be cooled below 50°C, however, and it
must also be swirled continuously during cooling. Otherwise, the agar will
solidify on the inner wall of the ﬂask. Its freezing point is 42°. Once the agar has
solidiﬁed, it must be reheated to 100°C to be liqueﬁed again.

2) Remove the aluminum from the mouth of the ﬂask and ﬂame the lip until it is hot.

3) If applicable, add the other media components, such as antibiotics, by sterile pipette.

o In order to make sure that the addition of the other media components does not
cause solidiﬁcation of the media, swirl the media while adding the solutions.

4) Lift the lid of a plate, and pour enough agar solution into the plate to just cover the
bottom (about 15-20 mL). Replace the lid.

0 Any bubbles that may have formed during pouring may be broken by gently
swirling the plate. They may also be broken after the gel has solidiﬁed by passing
a Bunsen burner ﬂame across the top of the gel.

0 Stack the next plate to be poured on top of the one just poured to insulate the plate
and minimize condensation inside the lid.

5) Repeat the previous step until all of the plates are poured.

6) Immediately after pouring the plates, rinse the remaining agar down the drain with
plenty of hot water. Once the agar has solidiﬁed, it can be difﬁcult to remove from
the ﬂask.

7) Wait for the plates to cool completely.

8) If not being used immediately, wrap Paraﬁlm around the outside edge of each plate or
place all of the plates into a sealed bag.

9) Label the plates with name, contents, and date.

10) Store the plates upside down at 4°C.

94

IN OC ULUM CULTURES

In order to easily monitor cell growth and reduce the lag in growth caused by a massive
dilution in cell concentration, inoculum cultures are ﬁrst grown up in a series of
increasing volumes of broth. The dilution ratio for each volume transfer can be as high
as 1:50 without a signiﬁcant risk of a long lag phase. For 1 L fermentations, a common
series of culture transfers is: glycerol freeze —» agar plate —9 5 mL media in a test tube —»
50 mL media in a shake ﬂask —> 1 L media in a bioreactor. A culture should be
transferred to a larger volume of media during mid exponential growth. Therefore, it is
best to transfer the culture when it has reached an OD600 of between 0.8 and 2.0. Most
shake ﬂasks and test tubes cannot support exponential growth at a cell concentration
indicated by an OD600 above 2.0 due to oxygen transfer or buffer limitations. On the
opposite extreme, an OD600 below 0.8 might lead to a lag phase in growth at any dilution
ratio greater than 1:10. The procedure for the OD600 measurement can be found in CELL
DENSITY - OD600 on pg. 99.

A single cell line should be used as the basis for an inoculum culture. Individual colonies
of a single cell line can be obtained by streaking a culture sample onto an agar plate.
Then, after the cells have multiplied to form visible, separated colonies, a colony is
removed from the plate and transferred to growth media in order to start a suspension
culture from a single cell line. Ideally, the cell culture sample for an inoculum plate
should come from the original glycerol freeze created directly after the culture was
transformed. Cultures that have been grown to high cell densities and transferred to
fresh media multiple times might grow and express proteins poorly due to a loss of
plasmid(s) or natural selection for slower growing or more stringent cell lines. If a
glycerol freeze is not available, however, the culture used to start the inoculum plate can
come from a suspension culture or a different plate. If it comes from a plate, the plate
should not have been incubated at growth temperatures (37 °C) for more than 24 hours
and should not have been stored in the refrigerator (8 °C) for more than a week before
transfer. Similarly, the cell density in a shake ﬂask should not have grown above an
OD600 of 2.0 and should not have been allowed to sit without agitation for extended
periods of time.

If possible, the culture should be grown on the same carbon source from one media
transfer to the next. Switching its primary substrate causes a lag in growth as the culture
adapts its metabolism to the new substrate. In general, E. coli grows quicker on LB broth
than on minimal media and glucose. Also, if the product is expressed by genes on
plasmids, all of the growth media should contain the antibiotics that select for those
plasmids.

Agar Plates

Supplies
1 or more agar plates
Glycerol freeze or other culture source

Wand or wire loop

95

Paraﬁlm
Permanent marker

Time: LB -— 12 hrs; M9 - 24 hours

Procedure
1) Warm the plates to incubation temperature and place in the laminar ﬂow hood.
0 If the culture to be used was stored on a plate in the refrigerator, warm the culture
back up to growth temperature for an hour in the incubator.
2) Inoculate the plate with cells from the culture sample
0 Flame the wand or wire loop and allow it to cool. Push the tip of the wire loop
into the glycerol freeze or the culture source and remove. Starting in one comer
of the plate brush the loop back and forth across one side of the plate. Reﬂame
the loop and allow it to cool. Then, starting on the edge of the section of the plate
that was previously streaked, brush the loop across a new comer of the plate while
make sure to cross over into the previous section to pick up cells. Repeat this
dilution procedure until every section of the plate has been streaked.
0 Do not allow the glycerol freeze to warm up to room temperature. Return it to the
-80 °C freezer immediately after use.
3) Place the cover on top of the plate and wrap the outer edge with paraﬁlm.
4) Label the plate with your name, date, and the components of the culture.
5) Incubate the plate upside down in the incubator.
0 Suggested: 37 °C
0 Rich media plates (i.e. LB) are usually incubated overnight for 8 to 12 hr.
Minimal media plates (i.e. M9) may take 24 hr before visible colonies are formed.
6) If the cultures are not going to be used immediately after incubation, they should be
stored upside down at 4 °C.

Test Tube Cultures

841291148

Media

Inoculum

Sterilized ﬂask

l or more sterilized test tubes with tops
Sterilized pipette tips

Sterilized wooden colony applicator or metal loop
Sterilized graduated cylinder (Optional)
Marking tape

Permanent marker

Time: LB — 6 hrs; M9 - 12 hrs

Procedure
1) Prepare the media according to MEDIA on pg. 85.

0 Prepare enough media for at least two test tube cultures and two shake ﬂask
cultures (100 mL total).

96

2) In the laminar ﬂow hood, transfer the sterilized media to one or more sterilized test
tubes with caps.

0 Suggested: 5 mL of media in a 50 mL test tube.

3) Move the inoculum plate from the incubator to the laminar ﬂow hood.

0 If the culture to be used was stored on a plate in the refrigerator, warm the culture

back up to growth temperature for an hour in the incubator.
4) Inoculate the test tube with a single colony from the plate.

0 Using a sterilized wooden applicator or 200 uL pipette tip, touch it to the visible
colony on the plate and then carefully drop it or swirl it in the media in the test
tube.

0 Prepare a second culture at the same time in case the ﬁrst culture has too long or
too short of a lag phase for the time schedule.

5) Label the tubes with your name, date, and the components of the culture.
6) Incubate the tubes.

0 Suggested: 37 °C and 250 RPM in the enclosed incubator in the teaching lab.

0 Transfer the culture when it begins to become opaque (OD600 of approximately
1.0).

0 LB cultures usually need to incubate for 6 hours. Minimal media cultures may
require 12 hours or more. Allow more time if the culture is being transferred to a
media with a different culture source.

Shake Flask Cultures

m

Media

Inoculum

Sterilized 250 mL Erlenmeyer ﬂask
Sterilized pipette tips

Pipetter

Sterilized graduated cylinder (optional)
Marking tape

Permanent marker

Time: LB — 3 hrs; M9 - 6 hrs

Procedure
1) Prepare the media according to MEDIA on pg. 85.
2) Transfer the sterilized media to one or more sterilized shake ﬂasks.
0 Suggested: 5 mL culture into 45 mL of fresh media in a 250 mL Erlenmeyer ﬂask.
o For maximum oxygen transport, the media should not be greater than 20% of the
total ﬂask volume.
3) Transfer the inoculum test tube culture from the incubator to the laminar ﬂow hood.
0 Do not allow the culture to sit too long without agitation or the cells can suffer
from the stringent response caused by oxygen starvation.
4) Inoculate the shake ﬂasks with test tube culture.

0 Flame the edges of the test tube and pour the culture directly into the media in the
ﬂask.

97

5) Label the ﬂask with your name, date, and the components of the culture.
6) Incubate the shake ﬂasks.

0 Suggested: 37 C and 250 RPM in the enclosed incubator in the teaching lab.

0 Transfer the culture when it begins to become opaque (OD600 of approximately
1.0).

98

CELL DENSITY - OD600

Below an optical density (OD) of 1.0, a cultures absorbance at a wavelength of 600 nm is
directly proportion to its cell density. Therefore,

A = CX (1)
where A is the OD600, C is the proportionality constant, and X is the cell mass
concentration. The OD600 can be used to determine the speciﬁc growth rate, it, of the
culture. The speciﬁc growth rate of the culture is deﬁned as:

_1dX

.11 = X? (2)

where t is the time. Assuming that the speciﬁc growth rate is constant, Equation 2 can be
integrated to obtain an expression for X.

X = X02”t (3)
where X0 is the cell mass concentration at time equals 0. As can be seen, the cell mass is

an exponential function of the speciﬁc growth rate. Rearranging equation 3 and
substituting the OD600, A, for the cell concentration, Equation 4 is obtained.

A
Ln —— = [It 4
A0 ( )
Where A0 is the OD600 at time equals zero. A semi log plot of the ratio of A to A0 versus
the time produces a line with a slope equal to the speciﬁc growth rate. Note that these
results are only applicable when the culture is in exponential growth phase and the
measured 01350015 below 1.0.

This procedure relates to the use of the Turner SP-890 UVN IS Spectrophotometer
located in room 3262 of the Engineering building. The instrument has multiple modes
depending on whether it is to be run as a stand-alone unit or connected to a computer.
See the manual for general instructions on its use and how to change modes. The
protocol listed here for determining cell density applies to most host types used in
recombinant engineering. The optimal wavelength used for the measurement is different
for some species, and optical measurements are not suitable for determining the density
of some cultures such as fungi. If the accuracy of these methods is in doubt, serial
dilution and weighing experiments should be used to check the assumption of linear
dependence.

Supplies
1 or 2 cuvette tubes

Dilution media

Test tubes for dilutions
Pipette tips

10% bleach and ﬂask

Procedure
1) Turn on the power switch in the back.

99

2)

3)
4)

5)

6)
7)
8)

9)

Wait for the system to initialize and the System Main Menu screen to be displayed.

0 Do not select a mode.

0 If the system is already in a speciﬁc mode, press “ESC” to return to the main
menu.

Press the “A” button. LAMBDA should be displayed on the top line of the screen.

Enter “600.0” in the “it” section of the screen. Then press “ENTER” and wait for the

spectrophotometer to adjust to the new wavelength. The new wavelength will be

displayed in the upper right comer, along with the absorbance.

Insert a cuvette containing your blank solution into the back sample holder.

0 Minimal media and water do not signiﬁcantly absorb light at 600 nM. Water can
be used as the blanking medium with little loss in accuracy.

Close the lid and press “AUTO ZERO”. Remove the cuvette and discard the blank.

Pipette your culture sample into the cuvette and return it to the back sample holder.

Close the lid and record the absorbance reading displayed in the “DATA” ﬁeld of the

screen.

Use serial dilutions to dilute the culture until the OD600 is below 1.0.

10) Calculate and record the absorbance based on the dilutions made.
11) Rinse the cuvettes out into a labeled ﬂask containing 10% bleach to kill the culture

cells.

Calculations

Cells/m1; OD600 x 8x108 Cells / dilution ratio
Cell Dry Weight = OD600 x 0.43 g/L ldilution ratio
Cell Wet Weight 2 3 x Cell Dry Weight

100

CULTURE SAMPLES

In order to monitor protein expression or cellular responses to environmental variables,
culture samples are collected and stored for later analysis by SDS—PAGE or enzyme
assay. Culture can be collected in 1.5 mL microcentrifuge tubes and spun down to obtain
cell samples for analysis. At cell densities indicated by an OD600 below 5, 1 mL samples
generally produce the required protein concentrations for analysis by SDS-PAGE. At
high cell densities, a 1 mL sample volume requires large volumes of Laemrrrli buffer for
solubilizing the cell mass. Therefore, the sample volume should be reduced in relation to
the increase in cell density as the OD600 rises above 5. When checking protein expression
after induction, a sample should be taken before induction, at 0.5 hours after induction, at
1 hour after induction, and every hour after that until the culture is harvested.

Supplies

Syringe

1.5 mL microcentrifuge tubes
Test tubes 1
Pipette tips

Pipetter

Permanent marker

Small ﬂask

10% bleach solution

Time: 5 nrin

Procedure

1) Connect a 5 mL syringe to the end of the tubing attached to the sampling port.
2) Open the sampling valve.

3) Flush the sampling port.

0 Withdraw ~5 mL of culture with the syringe. Close the sampling valve to prevent
drainage of the culture out of the bioreactor and remove the syringe. Discard the
ﬁrst 5 mL into a ﬂask containing 10% bleach solution to kill the cells.

4) Take a culture sample.
0 Reattach the syringe and open valve as before. Withdraw a second 5 mL of
culture. Close the clamp and valve and remove the syringe.
5) Transfer the culture sample to a test tube.
6) Use a portion of the culture sample to determine the OD600. Record this for reference
to the 1 mL pellet sample.
7) Pipette 1 mL of the culture sample into one or more 1.5 mL microcentrifuge tubes.

0 At high cell concentrations, smaller volumes of culture sample may be easier to
prepare for analysis by SDS-PAGE. For reference, a 1 mL culture sample
harvested at an OD600 of 10 produces the desired cell mass for viewing by SDS-
PAGE using 1 mL of Laemmli buffer. Higher cell concentrations will require
more work to digest with Laemmli buffer due to the limited volume of the
rrricrocentrifuge tube.

101

0 If only one sample is taken, ﬁll a second microcentrifuge tube with 1 mL of
water. This tube will serve to balance the microcentrifuge.

8) Place the tubes across from on another in whatever microcentrifuge is available. The
hinge to the lid of each tube should face the outside of the rotor.

9) Centrifuge at full speed for 30 seconds.

10) Remove the tubes.

0 If available, save the counter-balance tube for later use.

11) If the product being checked is intracellular, discard the culture supernatant by
pouring it into a ﬂask containing 10% bleach to kill any remaining cells. If the
product is extracellular, pour the supernatant into a new microcentrifuge tube.

0 A twisted piece of KimiWipe can be used to mop up any liquid left in the pellet
sample tube. Make sure not to disturb the pellet.

12) Label the sample tube or a box containing all of the sample tubes with your name,
contents, and date.

13) Store the sample in whatever freezer is available.
0 A -40 °C or -80 °C freezer is ideal.

102

INDUCTION

In recombinant engineering, the desired protein product is usually encoded on a plasmid
that has been introduced into the host cell. Transcription of the DNA on this plasmid is
designed to only occur when induced by a change in the media or environmental
conditions by the operator. This allows the host culture to grow to a high cell density
without the heterologous protein interfering with the metabolism of the cells. Induction
strategies include changes in temperature, changes in pH, and the introduction of a new
chemical. The following protocol describes a protocol for induction by the addition of
isopropyl B—D,thiogalactopyranoside (IPTG). IPTG is a non-metabolizable analog of
lactose that can induce transcription of genes controlled by Lac promoters. This form of
induction is utilized by the pET expression system and it is the most common inducer
used in this lab. In fed batch fermentations in the BioFlo He, the temperature, pH, or DO
can also be changed to decrease degradation of the protein being expressed.

Supplies

Sterilized l M IPTG or other inducer
Sterilized pipette tips

Pipetter

Time: 1hr - 6 hrs

Procedure for Fed Batch Fermentations in the BioFlo [Tc

1) Make any changes to the control recipe needed for induction.

2) If applicable, change the set point of any of the controlled environmental conditions
such as pH or temperature that need to be changed. Allow the bioreactor to reach its
new set point. ‘

0 A drop in temperature or change in pH will slow down the metabolism of the
culture and cause an increase in the DO. If the current stage in the controller
recipe is set to detect a sudden change in dissolved oxygen or any of the process
control parameters, the stage may need to be put on manual hold and the expected
response monitored closely until the PID controllers have adjusted to the change.

3) Unscrew and remove the cap to the inoculation port.

0 This exposes the bioreactor to the environment and a chance of contamination.
This late in the fermentation and at this high of a cell density, contamination is no
longer a concern as long as the length of the induction phase is less than several
hours.

4) Add the IPTG or other inducer to the broth.

0 1 mL of 1 M IPTG will give the standard induction concentration of 1 mM IPTG.

o IPTG is toxic to the cells and may cause a rise in the DO as the oxygen uptake
rate decreases. As with a change in the pH or temperature described above, the
stage may need to be put on manual hold and until the PID controllers have
adjusted to the change. If the recipe is currently employing PID control of the DO
by feed rate, the drop in the metabolism rate will likely cause overfeeding. The
controller increases the feed rate in an attempt to reduce the rising DO

103

concentration that is a result of the maximum oxygen uptake rate decreasing
below oxygen transfer rate of the bioreactor. In other words, the DO
concentration is forced above the set-point and the controller tries to respond by
dumping feed into the bioreactor even though the culture’s rate of oxygen
consumption is no longer being limited by a lack of substrate. The correct
controller response is to return the recipe to DO control by airﬂow with an open
loop feed scheme to keep the culture from becoming substrate limited.

5) Return the cap to the inoculation port.

6) Periodically monitor and take samples from the bioreactor during the induction phase.

7) When the induction phase is complete, stop the recipe and harvest the culture.

0 4 to 6 hours is the standard length of most IPT G induction phases.

Procedure for Batch Fermentations
8) If the temperature is to be reduced to improve expression, set-up a water bath shaker
for use.

0 Water bath shakers are located on the back wall of near to the door that connects
the two rooms of the Biochemical Engineering Teaching Lab. Use a bucket to
add water to the basin until the ﬂask clamps are covered. Turn of the power and
the heater and set the temperature knob to the desired set-point. Monitor the
temperature to make sure that the heating coils of the shaker are producing the
desired bath temperature.

9) Remove the culture from the incubator and place it in the laminar ﬂow hood.

0 Do not allow the culture to sit without agitation for extended periods of time. The
D0 in the ﬂask can become low enough to trigger a stress response in the cells
that will affect protein expression.

10) Remove the foil or cap from the batch culture container and ﬂame the opening.

11) If desired, take a cell density measurement and culture sample for analysis as the zero
time point of the induction phase.

12) Add the IPT G or other inducer to the broth.

0 Suggested: l M IPTG diluted 1:1000 by the batch fermentation broth will give the

standard induction concentration of 1 mM IPTG.
13) Reapply the tin foil or cap to the opening of the batch culture container.
14) Return the culture to the incubator or to the water bath shaker.
15) If desired, periodically take cell density measurements and culture samples from the
shake ﬂask in order to determine the time course of the induction phase.
16) When the induction phase is complete, stop the recipe and harvest the culture.
0 4 to 6 hours is the standard length of most IPTG induction phases.

104

CULTURE HARVESTING

At the end of the project, the entire culture needs to be harvested from the bioreactor. If
the product is an intracellular protein created by induction, the cell mass must be
separated from the broth and the product puriﬁed. If the product is extracellular, only the
liquid needs to be saved. Do not allow the culture to sit at room temperature for extended
lengths of time or product degradation can occur. This is why the centrifuge is cooled to
4 °C. Also, it is imperative that the centrifuge is balanced properly.

Supplies

Several 500 mL centrifuge bottles with lids
Spatula

Plastic storage container with lid

10% bleach

Soap and brushes

Labeling tape

Permanent marker

Time: 20 min

Procedure

1) Turn on the Sorvall RC-SB centrifuge in Room 3269.

2) Press the “Door” button while lifting the lid to access the rotor.

3) Install the correct rotor for the by gently placing it on the spindle in the center of the
centrifuge.

o The KA-lO composite rotor should be used for 500 mL bottles.

0 The Sorvall SA-600 rotor should be used for 30 mL samples

4) Close the door and adjust the centrifuge temperature to 4 °C

0 The controls are located below the temperature gauge on the left side of the unit.
The blue line represents the low temperature limit and the red line represents the
upper limit. The red line should be placed at 5 °C and the blue line just above 0
°C. Allow some time for the centrifuge to cool down.

5) Carefully pour the culture into the centrifuge bottles.

0 The bottles should be ﬁlled no higher than 80% full. Overﬁlling them can cause
them to leak while being centrifuged.

6) Take the samples to the double pan balance in 3269 and zero the balance using the
sliding weights.
7) Balance the bottles.

0 Inspect the rotor to determine the number of sample slots that must be used to
provide counter balancing. If only one sample bottle is being used, a water-ﬁlled
counter balance will be required. Two bottles containing cells can be placed on
the balance and culture transferred from one to the other until they are balanced.
While all the bottles placed in the rotor should contain approximately the same
volume, only those placed across from each other need to be carefully balanced.

105

0 Make sure that the lids of the 500 mL bottles have o—ring seals and are fully
sealed to prevent leakage.

1) Check that the centrifuge has reached 4 °C and open the door and unscrew the rotor
lid.

0 Even if the centrifuge has not completely reached 4 °C, it is better to centrifuge
and freeze the culture immediately rather than allow degradation of the product
due to the stress response of the cells.

2) Place the sample bottles in the centrifuge and make sure that the placement acts to
counterbalance the rotor.

3) Replace the rotor lid and screw it counter clockwise into place. Tighten the second
screw to lock the lid in place.

4) Close the door.

5) Set the Angular Velocity and the Time.

0 Suggested: 5,000 rpm; 10 minutes

6) Press the Start button.

0 The centrifuge must be balanced or the moorings of the spindle may snap. If a
high-pitched whine or a vibration begins to occur, immediately press the “Stop”
button and allow the rotor to slow down. When the “Door” button is lit, you can
press it to open the door. Remove the centrifuge tubes and double check that they
are balanced. Clean and disinfect any culture sample that may have spilled into
the centrifuge. If leakage or poor counterbalancing did not cause the problem,
contact the lab safety representative because there may be some other problem.

7) When the rotor has stopped and the “Door” light is lit, open the lid and unscrew the
rotor lid clockwise. Remove the rotor.

8) Remove the tubes from the rotor.

0 Check for spills inside the rotor or centrifuge. If culture has leaked into the
centrifuge, clean and sterilize the area.

9) If the supernatant is to be discarded, autoclave it or add bleach to it as described for
culture sterilization in BIOREACT OR SHUTDOWN on pg. 83.

10) If the product being checked is intracellular, discard the culture supernatant by
pouring it into a ﬂask and either add 10% bleach or autoclave it as described for
culture sterilization in BIOREACT OR SHUTDOWN on pg. 83. If the product is
extracellular, pour the supernatant into a new sample bottle.

11) Determine the weight of a plastic container that will be used to store the cells.

12) Use a spatula to spoon the cells from the centrifuge tubes into to the sample container.

13) Measure the weight of the plastic container with the added cells.

14) Calculate the weight of the cell pellets and label the culture container with your name,
contents, date, and cell mass. Store it in a -40 °C f or -80 °C freezer.

15) Rinse the 500 mL centrifuge bottles and any other containers with 10% bleach to kill
the remaining cell culture. Wash and rinse the centrifuge bottles and other containers.

16) Store it in an available freezer.

0 A -40 °C or -80 °C freezer is ideal.

106

SDS-PA GE

Sodium Dodecyl Sulfate Poly-Acrylimide Gel Electrophoresis (SDS-PAGE) is a
ubiquitous and non-speciﬁc method of determining protein composition in biological
samples. SDS-PAGE separates proteins along a linear sample lane on a gel sheet
according to MW. If the size of a protein is distinct from the rest of the proteins in the
sample, it is viewed as a sharp band in the sample lane. Multiple samples can be run side
by side and the intensity of the bands compared to each other to determine the relative
concentrations of each protein in the sample.

SDS-PAGE functions by unfolding the protein and attaching negative charges along its
length. The protein is then loaded onto a gel matrix and subjected to a voltage potential.
The voltage draws the protein through the gel and the gel acts as a selective inhibitor to
the protein’s migration. The larger the size of the protein, the more resistance it
experiences and the slower it migrates through the gel. The smaller proteins encounter
less resistance and therefore travel through the gel faster than the larger proteins. When
the proteins have been fully separated across the length of the gel, they are dyed. The
intensity of the bands that are created by the dyed protein is directly proportional to the
mass of protein in that band.

SDS-PAGE generally serves two purposes as a tool for optimizing recombinant protein
expression. First, it allows the researcher to get a general idea of the purity of a protein
sample that has been puriﬁed. Be separating the proteins by weight, any bands that are
visible besides the one for the desired protein indicate that the product sample has not
been fully puriﬁed. The second use of SDS-PAGE is to determine the expression level of
an induced protein in a host organism. The strength of the product band relative to the
other bands of the sample indicates the extent to which the protein was overexpressed.

The following protocol is intended for use speciﬁcally with the FB-VElO—l 10 mL
Fischer vertical electrophoresis stand. All SDS-PAGE equipment function similarly,
though. Wear gloves and safety goggles at all times during these procedures. Many of
the components such as acrylimide, and 2-mercaptoethanol are toxic. Others chemicals
can stain the skin and clothing. In addition, proteins from exposed ﬂesh on the bands can
contaminate the samples.

Gel Casting

The electrophoresis stand allows two gels to be run at once. This can save a great deal of
time and it allows more samples to be compared to each other. As long as the gel and
sample compositions are similar, the procedure and results for running two gels as
compared to one gel should be the same. In order to make sure that the two gels do not
vary, the gels should be poured at the same time using a single set of solutions. The gel
caster has the capability to hold two sets of plates at one time. The metal plates must face
each other on the inside so that the clear glass plates allow the operator to view the gels
that are being poured.

107

The gels are poured in two layers having differing acrylimide concentrations and
differing pH’s. The top layer, in which the protein is loaded, is called the stacking gel.
The low polymer concentration in stacking gel allows all of the proteins to migrate at the
same velocity and forces them to stack up at the interface between the stacking gel and
separating gel. The separating gel is the lower layer that is used to separate the proteins.
The desired composition of the separating gel is dependant on the size of the protein that
is to be separated and viewed. Gels are deﬁned by their concentration of polymerizable
component. The stock monomer solution used in forming the gel is a mix of 30%
acrylimide and 0.8% bisacrylimide for cross linking. A 12 % gel contains 12% by weight
of acrylimede. Free radical polymerization of the gel is initiated by ammonium
persulfate. N ,N,N ’,N ’ tetramethylethylenediamine (TEMED) is added to protect the free
radicals from quenching.

Each acrylimide concentration has only a certain range of protein weights that it can
separate. Gels with a very high polymer concentration resist the movement of the larger
proteins to a greater extent than the smaller proteins. Therefore, the smaller proteins
reach the bottom of the gel before the larger proteins begin to visibly separate near the
top of the gel. Lower polymer concentrations can be used to separate larger proteins.
The larger proteins are allowed to travel through the gel, but the smaller proteins all
crowd together at the bottom of the gel.

The following tables contain the recommended concentrations of each component for the
stacking and separating gel solutions. The solution that is prepared can be used to create
two gels. The ammonium persulfate and TEMED should only be added right before the
gel is poured.

Table 2: Separatirg Gel Solutions (15 mL)

 

 

 

 

 

 

 

 

 

Polymer Concentration 6 % 8 % 10 % 12 %

Target Protein Weights 110 kDa 70 kDa 40 kDa 25 kDa
Water 8.1 mL 7.1 mL 6.1 mL 5.1 mL
Separating Buffer 3.8 mL 3.8 mL 3.8 mL 3.8 mL
30:0.8 Acrylimide/Bis 3.0 mL 4.0 mL 5.0 mL 6.0 mL
10 % Ammonium Persulfate 150 11L 150 1.1L 150 ILL 150 uL
TEMED 12 11L 9 p.L 6 11L 6 11L

 

 

 

 

Table 3: Stacking Gel Solution (10 mL)

 

 

 

 

 

 

 

 

 

 

Polymer Concentration 3 %
Water 6.5 mL
StackingBuffer 2.5 mL
30:0.8 Acrylimide/Bis 1 mL
10 % Ammonium Persulfate 60 pL
TEMED 10 p.L
Su lies

DI water

Separating buffer

 

30:0/8 acrylimide/bis solution
10% ammonium persulfate
Stacking buffer

Pipette tips

Gel caster stand

Glass gel plate

Notched metal plate

White Teﬂon plate spacers
White Teﬂon sample comb
Gel plate plastic pouch

Gel caster pressure plate
Butanol

Two small beakers or 25 mL tubes

Time: 45 rrrinutes

Procedure

1)

2)

3)

4)

5)
6)

7)

8)

Clean the gel caster stand and the plastic front pressure plate with soap and water.

Wipe dry.

Clean the glass plate and the white metal plate with the tabs at the upper comers.

Rinse with ethanol or acetone and allow it to dry.

0 The gel that is formed will not stick to wet surfaces, so the plates must be
completely dry before they are placed in the pouch. This is why the plates are
rinsed with a volatile liquid such as acetone or ethanol. The acetone or ethanol
can be quickly removed from the plates by holding the plate with tongs and
passing it over the ﬂame from a Bunsen burner. They can also be wiped dry with
a KimiWipe.

Clean the plastic pouch and allow it to dry.

0 Check that the pouch does not leak.

- The pouch can be dried quickly by rinsing it with ethanol or acetone forcing air to
circulate inside it.

Place the glass plate, the notched metal plate and two white teﬂon spacers into the

pouch.

0 The white spacers are placed between the edges of the two plates to form a space
in which to pour the gel.

0 Make sure that spacers have the same thickness as the comb that is to be used to
create the sample reservoirs.

Adjust the thumb screws on the front of the caster to create room for the pouch

between the front pressure plate and the rear wall of the caster.

Place the pouch into the caster so that it rests against the rear wall. Lightly tighten the

thumb screws to hold the pouch upright.

Straighten the spacers near the edges of the two gel casting plates.

0 This is easily done by sliding a third spacer between the two plates and using it to
press out against the edges the other two spacers.

Finish tightening the thumbscrews until the spacers and the bag have formed a ﬁrm

seal with the plates.

109

0 The plates can crack if too much pressure is used. Stop tightening when the
screws require extra effort to turn them.

9) Level the gel caster using the leveling screws at the four comers of the gel caster.

0 Adjust the front two leveling screws until the air pocket is centered in the plastic
bubble in the middle of the gel. Most likely, the caster will be balancing on only
three of the four leveling screws. Lower the fourth leveling screw until it also
begins to support a portion of the caster’s weight.

10) Mark the desired height of the separating gel.

0 Slide the sample well comb into the slot in the top of the gel plates until the tabs
are completely below the notch in the metal plate. Measure one centimeter down
from the bottom of the wells and mark the spot.

11) Prepare the separating gel according to the Table given above.

0 Wear gloves.

12) Add the required amount of ammonium persulfate and TEMED to the separating gel
solution.

0 These components are used to initiate the polymerization. Only add them when
the gel is ready to be poured.

13) Thoroughly mix the separating gel solution

0 Cap the resolving solution container and gently mix the polymerization initiators
by ﬂipping the container upside down and back again. Do not shake the container
to mix it. The SDS will cause a foam to be created that cannot be easily poured.
The solution can also be mixed by gently stirring with a pipette tip or spatula.

l4) Pour the separating gel solution into the space between the plates.

0 This can be tricky. The easiest method is to use a pipette to add separating gel
solution until the top of the liquid reaches the marked line. Otherwise, if the
container has a small diameter opening, the solution can be slowly poured into the
space between the plates. When done, check to make sure that solution is not
leaking around the edges of the spacers and into the sides of the pouch. If the
level drops, use the thumb screws to add more pressure to the plates and add more
solution if needed.

0 Do not wait to pour the gels. Polymerization begins immediately and the solution
will become too viscous to pour.

15) Immediately cover the polymerizing separating gel with water-saturated butanol.

Ethanol will also work if butanol is not available.

0 The butanol must be saturated with water or it will dehydrate the acrylimide
solution and shrink and concentrate the gel. The alcohol layer serves three
purposes. One, it removes any bubbles that formed while pouring the gel. Air
pockets in the gel will stop the migration of the protein and ruin the results. Two,
the butanol protects the solution from the air while it is polymerizing. Oxygen
that is present in the solution will interfere with the free radical polymerization
used to link the acrylimide. Three, the pressure on the separating solution helps to
keep it level.

16) Allow the gel to sit until polymerized.

o The expected time requirement is 15 minutes. If the gel has not hardened within
20 minutes, a mistake has probably occurred in the preparation of the gelling
solution.

110

o It can be difﬁcult to determine if the gel has polymerized due to the liquid layer
on top. A plug of solution can be kept in the pipette tip or sealed in the
acrylimide mixing container. When this has hardened, so has the gel in the stand.

17) Pour off the butanol solution from the top of the gel.

18) Gently wash the top of the gel with water to remove any remaining alcohol or
unpolymerized separating gel solution.

19) Pour off the water from the top of the gel.

0 The surface tension of the liquid can make it difﬁcult to pour all of it out from
between the plates. The edge of a KimiWipe held against the lip of the plates
while they are turned upside down will draw the rest of the liquid out.

20) Prepare the stacking gel according to the table given above.

21) Add the required amount of ammonium persulfate and TEMED to the stacking gel
solution.

22) Pour the stacking gel solution onto the separating gel until the solution just reaches
the opening created by the notch in the metal plate.

23) Insert the sample reservoir comb into the resolving gel solution in the space between
the plates.

0 The friction of the plates will hold the comb in place.

0 Do not allow air bubbles to be trapped between the comb and the separating gel.
They will block the movement of the sample through the gel or allow adjoining
sample reservoirs to mix. Bubbles can sometimes be removed by slightly tilitng
the gel stand to one side and tapping on the glass. If all else fails, remove the
comb before the gel has polymerized, add more solution if any leaked out, and
reinsert the comb.

24) Allow the plates to sit for 15 min or until the gel has polymerized.
25) Unscrew the thumbscrews and carefully remove the gel pouch.
26) Gently slide the gel plates out of the pouch.

0 First, check if gel leaked out during the polymerization and bonded the pouch to
the outside of the plates. If a seal has formed, gently separate the plate from the
pouch in order to keep the plates from pulling apart as they are being removed.

27) Scrape away any gel that has stuck to the outside walls or edges of the plates.
28) Inspect the gels.

0 The gel should not have any air bubbles and the spacers and comb should be ﬂush
with the edges of the plates. It is best if the bottom of the wells are approximately
1 cm above the interface of the stacking and the separating gel layers. If a visible
bubble exists in the gel, the well directly above the air pocket should not be used
to separate a protein sample. Instead, ﬁll it with 1x Laemmli buffer during
sample loading. '

29) If the gels are not going to be used immediately, place them back into the pouch or
some other bag or container. Seal the pouch shut and store it in a refrigerator.

0 Do not leave the gel Open to the environment for long periods of time or it can
become dried out.

Cell Fractionation

The following protocol only pertains to experiments in which the operator wishes to
determine if inclusion body formation during protein expression is an issue. If the

protein is known to be soluble or the operator only wants to determine the overall extent
to which the protein is being expressed, then this section can be skipped.

A common problem with expression of proteins in recombinant E. coli is the production
of insoluble protein aggregates called inclusion bodies. Inclusion bodies are usually
formed by heterologous proteins that have a number of hydrophobic regions that make
them difﬁcult to maintain in solution. During expression, the protein might come into
contact with another protein with a similar hydrophobic region and the two proteins
might denature and fold together to shelter the hydrophobic region from the aqueous
cytoplasm of the cell. This is the seed for an inclusion body and it continues to grow as
more hydrophobic proteins are drawn into it. Due to the denaturing that occurs when the
proteins form aggregates, proteins that are recovered from inclusion bodies must be
refolded to be active. This can be a very complicated procedure.

Inclusion body formation is especially a problem for eukaryotic, integral membrane
proteins that are extremely hydrophobic and are only stable in the hydrophobic regions of
a cellular membrane. In their native environments, these types of proteins generally
require chaperone proteins to protect them from denaturing conditions as they are
transported to the cell membrane. When expressed in prokaryotic hosts such as E. coli,
the proteins generally rely on random diffusion for transport. At the high expression
levels and protein concentrations that characterize recombinant protein expression
strategies, it is very likely that the proteins collide and form inclusion bodies before they
reach the cell membrane. Furthermore, the cell membrane might become saturated with
the expressed protein and either kill the cell or increase the rate of inclusion body
formation. Therefore, the most common ﬁrst step in reducing insoluble protein levels in
the cell is to reduce the rate of formation of the expressed protein. This is usually
accomplished by reducing the temperature and the concentration of the inducer.

SDS-PAGE can be used to determine the extent of inclusion body formation during
expression in recombinant hosts. The inclusion bodies are very dense compared to the
soluble components of the cell. Therefore, normal centrifugation can be used to separate
the inclusion bodies and the rest of the insoluble fraction from the soluble fraction of the
cell. By comparing the electrophoresis bands of the expressed protein from both
fractions, the operator can determine whether a change needs to be made in the
expression strategy in order to produce more of the soluble protein.

The following protocol is a simple method of fractionating a 1 mL cell pellet sample that
does not require any specialized equipment. The freeze/thaw method lyses the cell by
quickly freezing the cell samples to create ice crystals that fracture the cell membrane.
The procedure is more effective the quicker the sample freezes, so liquid nitrogen should
be used in place of the dry ice/ethanol combination speciﬁed in this protocol if it is
available. A tip sonicator can also be used.

Supplies
1 mL cell culture samples

Sturdy plastic bowl

112

Styrofoam ice box

1 cup of dry ice
Ethanol

Metal tongs or tweezers
Microcentrifuge tubes
Marker

Beaker with water

Time: 30 minutes

Procedure
1) Thaw the 1 mL cell culture sample pellets.
2) Spoon the dry ice into the plastic bowl and place the bowl in the Styrofoam box.

0 Use enough ice to cover most of the bottom of the bowl while leaving some open
space between the nuggets.

3) If applicable, return the dry ice to the freezer from which it was obtained.
4) Add ethanol to the bowl until the liquid level is just above the level of the dry ice.

0 If available, liquid nitrogen works much better for quickly freezing the culture
samples.

5) Make sure that the culture samples are well marked.

0 The ethanol will tend to strip permanent marker from the outside of the centrifuge
tubes. It is best to mark both the top and the side of the tubes in case the tube
becomes submerged. An even better strategy is to create a harness using a
microcentrifuge tube rack that can be dipped in and out of the ethanol bath.

6) Vortex the samples to resuspend the cell material.
0 Suggested: 30 seconds
7) Use the tongs, to dip the centrifuge tubes in the dry ice/ethanol bath until they are

completely frozen. 1

0 Suggested: 2 minutes

0 Try to position the microcentrifuge tubes in the dry ice so that the top is not
submerged.

8) After the cell samples in the tubes have frozen, remove them from the dry ice bath
and place them in the water bath.
9) When the samples have thawed, remove them from the water bath and wipe them dry.

0 Do not allow the samples to reach room temperature. It reduces the speed with
which the samples can be frozen and increases the rate of protein degradation in
the sample.

0 If proteolytic activity is causing the appearance of bands of degraded protein, a
protease inhibitor can be added to the buffer.

10) Vortex the samples again to resuspend the cells.

11) Repeat the freeze/thaw procedure multiple times.
0 Suggested: 5 freeze/thaw cycles

12) After the ﬁnal thaw, place the samples in a microcentrifuge and pellet the insoluble
fraction;

0 Suggested: 15,000 RPM; 20 minutes

113

0 Use a microcentrifuge that is kept in refrigerated conditions if one is available. It
reduces protein degradation. One can be found in the Protein Expression
Laboratory.

0 Remember to balance the centrifuge.

13) Label empty microcentrifuge tubes in a way that indicates that they will contain the
soluble fractions of the samples that have just been fractionated.
14) When the centrifugation is complete, remove the sample tubes.

0 Be careful not to unsettle the pellet by shaking or dropping the tubes.

0 A small pellet should be visible on the bottom of each tube. If one cannot be seen
or the insoluble material appears to have formed a viscous liquid layer instead of a
hard pellet, the sample may need to be centrifuged for a longer time period.

15) Pipette the supernatant from the sample tube to the newly marked microcentrifuge
tube.

0 Be careful not to suck up a portion of the pellet. It is unlikely that the pellet is
fully compacted against the bottom of the centrifuge tube. Use a small volume
pipette whose intake rate can be easily controller. Keep the pipette tip as far away
from the pellet as possible while removing the supernatant. When only a small
volume is left in the tube, it can be rotated to horizontal to open up more area of
the supernatant to the pipette. Remove as much of the supernatant as possible.

16) Add the same amount of 1X PBS buffer to the pellet as was added to the cell sample
initially.

17) Vortex the pellet to suspend the insoluble material
0 Suggested: 30 seconds

18) If the samples are not going to be analyzed immediately, place them in the freezer.
- Suggested: -80 °C

Sample Preparation

As a ﬁrst step in determining expression levels in recombinant E. coli, it is often easiest
to solubilize an entire 1 mL culture sample from the fermentation and examine the total
cell protein (T CP) by SDS-PAGE. Comparing the band of the recombinant protein to the
bands from the rest of the proteins gives the operator a good idea of how well the desired
product is being overexpressed. If the product band cannot be seen when comparing the
uninduced sample to the induced sample, then the expression levels are either poor or
nonexistent and a new expression host might be required.

The following sample preparation procedure is based on the assumption that the sample
is a cell pellet taken directly from a fermentation culture. The sample can, in fact, be a
solution of puriﬁed protein, lysate from a cell fractionation, or the insoluble portion of a
cell fractionation. A method for fractionating and analyzing a cell sample based on its
insoluble and soluble components is given in Cell Fractionation on pg. 111.

The primary consideration for sample preparation is having the correct amount of total
protein per SDS PAGE Sample. Too much protein causes the gel lane to be a large blue
streak that expands outward against the other lanes. Too little protein and the operator
might be unable to view the desired protein even if it had a reasonable expression level.
Here are some common principles:

114

At least 1 p. g of protein is required to be visible by SDS-PAGE.

4 - 10 n g is the optimum quantity of protein for viewing of the band.
The total amount of protein per well should not exceed 100 p. g.

The sample well can hold between 10 — 20 uL of prepared sample.

A ﬁrst estimate of the volume of buffer required for solubilizing a full cell pellet for
SDS-PAGE can be obtained from the optical density of cell sample. For every unit of
ODwo, a 1 mL E. coli pellet should be solubilized in 200 uL of SDS-PAGE buffer. For
example, if the OD500 of the culture was 2.00 at the time that the 1 mL culture sample
was pelleted, the ﬁnal SDS-PAGE sample should have a volume of 400 uL: 200 uL of
PBS and 200 uL of 2X Laemmli sample buffer. The sample buffer is named after the
researcher who developed the SDS-PAGE technique (Laemmli, 1970).

Laemmli sample buffer contains:
A buffering reagent: Tris-HCl
A charged ionic component for bonding to the protein: SDS
A reagent for increasing density so that the sample will sink into the wells:
glycerol
A chemical that cleaves disulﬁde bonds between proteins: 2-mercaptoethanol
A low MW dye that signiﬁes the protein front during diffusion: bromophenol blue
A denaturing reagent for untangling insoluble components: urea

SDS is an anionic surfactant with very strong ability to denature and solubilize proteins.
The amount of SDS that becomes attached to the protein is, in most cases, directly
proportional to the weight of the protein. SDS is the component of the buffer that causes
the protein to migrate towards the cathode of the SDS-PAGE stand. Standard Laemmli
buffer does not require urea. It has been added to this protocol to help solubilize
inclusion bodies and other normally insoluble components that can be found in an
unpuriﬁed cell sample. Urea is often not included in the Laemrrrli buffer if the sample is
culture lysate or puriﬁed protein. All of the chemicals that make up the buffer can be
premixed and stored frozen except for the 2-mercaptoethanol. 2-mercaptoethanol should
be added directly before use.

Su lies

2X Laemmli sample buffer
PBS

TEMED

Pelleted culture samples
Heated water bath
Microcentrifuge

Vortex

Time: 20 minutes

115

Procedure

1) Remove the 1 mL culture sample pellets or protein samples from the freezer and
allow them to thaw.

2) Remove a 9 mL aliquot of 2X Laemmli sample buffer from the freezer and allow it to
thaw.

0 The urea in the buffer is near its saturation limit. Therefore, it is like that crystals
will from when the buffer is thawed. This does not cause problems.

3) Add 1 mL of 2-mercaptoethanol to the 9 mL aliquot of sample buffer.

0 Wear gloves.

0 The Laemmli buffer can still be used for a week after the 2-mercaptoethanol is
added without noticeable degradation in the results. It can be stored at room
temperature. .

4) Begin heating a water bath for boiling the samples.
5) Add PBS to all of the sample pellets if the culture samples have not yet been
resuspended in buffer or puriﬁed.

0 Suggested: 0.1 mL per unit OD600 of the culture sample.

6) Add equal volumes of 2x Laemmli buffer to the volumes of buffer already present in
the sample tubes.

7) Vortex the microcentrifuge tubes to suspend the samples.
0 Suggested: 30 seconds

8) Check to see if the water bath has begun to boil.
0 The temperature can range between 85 to 100 °C.

9) Boil the samples for ﬁve minutes.

0 Do not heat the samples for too long or signiﬁcant cleavage of peptide bonds can

occur.
10) Centrifuge the samples to pellet any remaining insoluble cellular components.
0 Suggested: 14,000 RPM; 5 minutes.
- Insoluble material that is added with the sample will cause streaking in the gel.
11) If the sample is not going to be immediately run, label it with name and date and store
it in the refrigerator. Heat it back up to 37 ° C and centrifuge it again before use.

Electrophoresis Procedure

In accordance with the instruction manual for the Fischer vertical electrophoresis stand,
the upper buffer chamber (UBC) is the vertical stand to which the gel plates, power
chords, and cooling water lines are attached, and the lower buffer chamber (LBC) is the
electrophoresis buffer pool in which the UBC sits. With the gel plates ﬁrmly sealed
against the sides of the UBC, the top of the gel is only open to the buffer that is in contact
with the anode and bottom of the gel is only open to the buffer in contact with the
cathode. This allows a voltage source to be used to create a potential across the length of
the gel as long as a leak is not present between the buffers in the LBC and UBC.

The sample experiences more random diffusion the longer it remains in the gel.
Therefore, the quicker and more efﬁciently that the samples can be loaded and run, the
better the results will be. The migration speed of the protein being analyzed is dependant
on the voltage. The optimum voltage is empirically determined and it is bound by two
considerations. If the voltage is too high, a large amount of current is forced through the

116

gel and the gel heats up. Heat dissipation in the electrophoresis stand is aided by the use
of the cooling water and the use of metal plates to increase transport. If the gel becomes
too hot, however, the gel and the samples deform and the resulting dye front looks like its
smiling. If this occurs, lower the voltage. In contrast, if the voltage is too low, the length
of time required for the proteins to travel through the gel is too long and undirected
diffusion allows the bands to spread. When running a gel for the ﬁrst time, it is good
practice to monitor the gel regularly to determine whether the dye front and samples are
distorting.

The following protocol is an adaptation of the instructions supplied with the gel stand.

Su lies

Upper buffer chamber (UBC)

Lower buffer chamber (LBC)

Water

5X electrophoresis buffer

500 mL or larger volumetric ﬂask or beaker
Prepped samples

Protein ladder

Small pipette and tips or Hamilton syringe
Two prepped gel plates or one gel plate and a plastic plate
Voltage regulator

Power chords

3/8” cooling water tubing

Time: 2 hrs
Procedure '
1) Inspect the white rubber gaskets that line the outer walls of the UBC.

0 These rubber gaskets seal the plates against the UBC and separate the
electrophoresis buffer in the UBC from the buffer in the LBC. If a gasket is
damaged or improperly seated in the UBC, buffer can leak around the gel plates
and cause drainage of the UBC buffer reservoir or allow electric current to pass
directly from the UBC to the LBC electrodes without passing through the gel.

2) Unscrew the black knobs on the sides of the UBC to loosen the black clamps that seal
the gel plates against the UBC.
3) Seal the plate or plates against the sides of the UBC.

0 Slide the plate down between the black clamps and the UBC. The metallic plate
with the notch in the top should face the inside. Make sure that the plates are
centered and both of the lower edges are settled into the well cut into the lower
section of the UBC. Position the clamps so that the notch between the thinner
section of the clamp and the base of the clamp cradles the edge of the plates.
Tighten the knobs until the white gasket between the plate and the UBC wall
begins to compress and form a seal. It is very easy to break the clamps or crack
the plates with too much pressure. Do not over tighten.

117

4)

5)

6)

7)

8)

9)

o In order to function, both sides of the UBC must be sealed. If two gels are not
going to be run, one of the sides can be sealed with the thick glass plate that is
supplied with the electrophoresis stand.

Place the UBC into the LBC so that openings in the slots of the LBC are facing the

pegs extruding from the lower sides of the UBC. Slide the UBC pegs into the slots on

the LBC.

Mix 100 mL of 5X electrophoresis buffer with 400 mL of DI water to make 1X

electrophoresis buffer.

Add 1X electrophoresis buffer to the UBC.

0 Suggested: 200 mL

0 The buffer should reach to about 5 mm from the top of the glass slide while
completely covering the opening to the gel created by the notch in the metal slide.
If the comb used to create the wells has not yet been removed after pouring the
plates, this opening in the gel will still be covered.

0 Inspect the seal between the UBC and gel plates to make sure that buffer is not
leaking around the gasket and pooling in the LBC. If a leak exists, pour off the
buffer, remove the gel plates and start over.

0 Foaming of the electrophoresis buffer can impede sample loading. If any bubbles
have formed in the buffer, use a pipette to suck them off.

Add 1X electrophoresis buffer to the LBC to approximately 5 mm

0 Suggested: 200 mL

0 The buffer should reach to about 5 nun above the opening in the bottom of the gel
plates.

0 If the electrophoresis buffer level drops during a run and contact is lost with the
gel, more buffer can be added and the voltage potential reestablished

Remove the comb from the gel.

0 It can be difﬁcult to view the sample well positions after the comb is removed. A
marker can be used to outline the positions of the wells on the glass plate.

0 The comb should be removed slowly and carefully. A vacuum tends to form
between the notches in the comb and the gel as they are separated from each
other. If the comb is gently twisted and bent back and forth while being removed,
buffer from the UBC is allowed to drain between the plates and the comb and ﬁll
the space that is created. After the comb has been removed, use a thin pipette tip
to reposition any of the sample well walls that have been bent or displaced. If the
gel of the wall cracks or tears, the two wells that it separates cannot be used for
running samples.

Attach one end of a piece of 3/8" ID clear ﬂexible lab tubing to the nozzle marked as

“in” on the UBC. Attach the other end to the spigot on a cold water tap.

10) Attach another piece of tubing to the nozzle marked as “out” on the UBC and allow

the outlet of the tubing to drain into a sink.

11) Open the water tap and allow cooling water to ﬂow through the UBC.

0 The pressure and ﬂow rate of the water does not need to be high. Water ﬂow
should be below 2 Umin

12) Load samples onto the gel.

0 Suggested: 10 to 20 uL per sample. Only 10 uL of the protein ladder is required.
0 Remember to include a protein ladder for comparison to known MW.

118

0 Use a Hamilton syringe and rinse the syringe with buffer between each sample.
Otherwise, use a 200 uL pipette and replace the pipette tip between each sample.

0 Place the tip of the syringe or pipette against the back wall of the glass slide and
above the opening to the well. Slowly add the sample so that it sinks through the
buffer and into the well.

0 Voltage must be applied immediately after the samples are added to keep the
protein from diffusing out of the well.

0 Load empty unused wells with 1X sample buffer to prevent the spreading of
neighboring samples.

13) Set the safety lid onto the unit on so that power chords connect with the electrodes.

0 Make sure that the colors of the electrodes and power chords align. If the voltage
is reversed, the samples will be lost.

14) Plug the power chords into the corresponding electrodes on the voltage supply.
15) Turn on the voltage supply and set the desired voltage.

0 Suggested: 170 V - 200 V

0 When voltage is initially applied to the gel, the current transfer can be quite high.
A common practice is to use a lower voltage than 170 V to pull the sample
through the stacking gel. An 80V stacking gel voltage loads the protein into a
single band at the top of the resolving gel and takes about 15 min. The current
will drop and the voltage can be increased to its full running value.

0 If the electrophoresis buffer level drops during a run and contact is lost with the
gel, more buffer can be added and the voltage potential reestablished.

0 The samples can be saved if a leak develops between the gasket and the gel plate
after the samples have already entered the separating gel. Quickly pour off the
buffers, dismantle the apparatus and reset the seals. Then, add the buffers back
and reestablish the voltage

16) Maintain the voltage until the dye front reaches the bottom of the gel.

0 Suggested: 90 minutes — 120 nrinutes

17) Turn off the power to the voltage supply.

0 The gels need to be processed immediately after the voltage is removed or the
protein bands will spread due to diffusion.

18) Unhook the voltage supply, remove the cover to the UBC, and dump the buffer.
19) Remove the plates or plates from the UBC.

0 Immediately begin the staining procedure or the separated protein bands will

become diffuse and difﬁcult to analyze.

Gel Staining and Analysis

After the proteins are separated in the gel, they must be stained to be made visible. The
dye that is often used is Coomassie Brilliant Blue which bonds speciﬁcally to certain
amino acids in the protein. Other dyes can be used that are either much more speciﬁc to
certain proteins or are more sensitive to lower concentrations. These stains are either
more expensive or toxic, however.

The mixture of acid and methanol in the staining solution acts to shrink the gel and

precipitate the protein. This keeps the protein from diffusing out of the gel and maintains
the sharpness of the bands. The speciﬁc staining and destaining solution compositions

119

given in this protocol are not essential to the success of the protocol. In addition, a large
number of different staining and destaining procedures can be used. The only ﬁrm rule is
that the operator must not allow the gel to sit in pure water for extended periods of time.
Pure water causes the gel to swell and release the protein.

Analysis of the gel is a subjective measurement. The visible intensity of the protein
bands is proportional to the mass of the protein in the band. Certain results can make
analysis difﬁcult, however. If too much protein is included in the sample, the bands
widen and run into each other. If the sample contains too little protein, it can be difﬁcult
to see the protein. In addition, large amounts of DNA or RNA in the sample or an
especially large concentration of one speciﬁc protein size can distort or cause streaking in
the gel lane. Multiple gels and sample concentrations might need to be run to get a good
sense of the relative protein concentrations. If the expressed protein band is difﬁcult to
view due to the extraneous bands created by the other proteins in the sample, increasing
the sample size does not help. It just causes the bands to become oversaturated.
Puriﬁcation procedures might be required to remove some of the extraneous protein and
cellular components.

Su lies

Coomassie staining solution
Destaining solution

Razor

Glass staining container with cover
Water bottle

Plastic sheets or clear overhead slides

Time: 13 hours

Procedure
19) Separate the gel from between the plates.

0 Use gloves.

0 It is very easy to tear the gel, especially at low acrylimide concentrations. The gel
will be more tightly bonded to the metal plate than to the glass. First, use a razor
to separate the outer edges of the gel from the surface of the glass plate. Use the
razor as a lever to very slowly pry the glass plate away from the gel. Then, gently
separate the bottom of the gel from the bottom surface of the metal plate. Lay the
bottom of the plate against the ﬂoor of the glass staining container so that the
bottom of the gel can ﬂop against the container. Slowly lift up on the plate so that
gravity and surface tension will cause the gel to slide into the container.

20) Rinse the gel with a small quantity of water and pour off the water.
21) Add enough coomassie stain to the glass staining container to just cover the surface of
the gel.

0 Suggested: 100 mL

0 Use gloves. The stain cannot be washed off.

22) Cover the container and place it on a rocker to allow the stain to set.

0 Suggested: 1 hr

120

o Staining times are not especially important. As long as the staining and solutions
contains a large percentage of methanol, the protein bands will become ﬁxed as
the gel becomes dehydrated and shrinks. If the gel is allowed to sit in a high
concentration of methanol for too long of a time, the gel might shrink too much.

23) Pour out the stain and rinse the gel with a small amount of RO water.

0 Save the stain. It can be reused multiple times.

24) Add destaining solution to coat the gel and place it back on the rocker.

0 Suggested: 100 mL; 12 hours

0 The destaining solution has the same general composition as the staining solution.
The low concentration of coomassie blue in the destaining solution acts to draw
out the dye from the gel that has not bonded to the protein.

0 The destaining process can be sped up by replacing the destaining solution at
regular intervals.

0 A folded paper towel placed in the destaining bath will soak up excess stain and
allow the re-use of the destaining solution.

0 The methanol concentration in the destaining solution can be reduced to 20% if
the destaining process is causing the gel to shrink too much.

25) Dump out the destaining solution and wash with R0 water.

0 If the destaining solution does not have a high concentration of coomassie blue, it
can be used as an initial destaining wash for another gel. Save it.

26) Cut two pieces of thin plastic laminating sheet or obtain two overhead slides.
27) Add a couple of drops of water to the fronts of both sheets.
28) Trap the gel between the two sheets.

0 Make sure that air bubbles are not caught between the sheets and the gel. Air
bubbles distort the images.

0 If desired, the two sheets can be taped closed.

29) Pictures of the gel can be captured by a scanner and analyzed using a computer.

Stock Solutions
80% Glycerol (100 mL)

0 20 mL of RO water in 100 mL graduated cylinder
0 Add glycerol to 100 mL by volume displacement

0.2 % Bromophenol Blue (100 mL)
0 Dissolve 0.2 g of bromophenol blue in 5 mL of EtOH
0 Add water to 100 mL

20% SDS (Sodium Dodecyl Sulfate) (50 mL)
0 Stir 30 mL of water

0 Add 20 g of SDS

0 Add water to 50 mL

05 M Tris (pH: 6.8) (100 mL)

0 Stir 60 mL of water
0 Add 6.1 g of Tris base

121

0 Adjust pH to 6.8 with 5M or concentrated HCl
0 Add water to a ﬁnal volume of 100 mL

Stacking Buffer (100 mL)
0 98 mL of 0.5 M Tris (pH 6.8)
0 Add 2 mL of 20% SDS

Separating Buffer (100 mL)
0 98 mL of 1.5 M Tris (pH 8.8)
0 Add 2 mL of 20% SDS

1.5 M Tris (pH=8.8)

Stir 60 mL of water

Add 18.2 g of tris base

Adjust pH to 8.8 with 5M or concentrated HCl
Add water to a ﬁnal volume of 100 mL

2x Laemmli Sample Buffer (90 mL)

Stir 20 mL 0.5 M Tris pH6.8

Add 25 mL 80% Glycerol

Add 48 g Urea

Add 2 g SDS

Add 1 mL 0.2% Bromophenol Blue

Add water to 90 mL

Separate into 9 mL aliquots and store at -20 °C

5x electrophoresis buffer (100 mL)

Stir 80 mL of water

Add 1.5 g Tris base

Add 7.2 g glycine

Add 0.5 g SDS

Add water to 100 mL

Conﬁrm that pH is near 8.3 but do not adjust
Store in a glass container

10% Ammonium Persulfate (10 mL)
0 10 mL of water

0 Add 1 g of ammonium persulfate
0 Store in at 4 °C

Butanol
0 Mix water and butanol
0 Allow to separate

Destaining Solution (100 mL)

122

Add 40 mL methanol
Add 10 mL acetic acid
Add 50 mL water

Coomassie Blue Stain (100 mL)

Stir 100 mL Destaining Solution

Add 0.1 g Coomasie Brilliant Blue R250
Filter the solution to remove undissolved dye
The stain can be saved and reused

Phosphate Buffer Saline (PBS) Solution (100 mL)
Stir 80 mL of water

Add 0.26 g sodium phosphate monobasic

Add 1.28 g sodium phosphate dibasic heptahydrate
Add 0.7 g N aCl

Conﬁrm that pH is near 7.2.

123

APPENDIX B

PRESTO CONTROL MANUAL

124

APPENDIX B: TABLE OF CONTENTS

Introduction ..................................................................................................................... 126
Control System Overview ........................................................................................... 126
Bioreactor Control Overview ...................................................................................... 128

Recipes ............................................................................................................................ 134
Standard Fed Batch Control ........................................................................................ 135
Reduced Fed Batch Control ........................................................................................ 142
Maximum DO Control ................................................................................................ 142
Exponential Corrected Feed Forward Control ............................................................ 143
Minimum Feed Rate Control ....................................................................................... 143
Maximum Feed Rate Control ...................................................................................... 144
Saturation Pulse Test Control ...................................................................................... 144
Feed PID 1 & 2 ............................................................................................................ 145

125

Introduction

Control System Overview

“Presto Control” is the name given to the controller and display programs designed using
Opto 22 that allow knowledge based (KB) control of the BioFlo IIc. Opto 22 is a set of
software programs and processing units designed for industrial automation applications.
Due to the limited control capabilities built into the BioFlo, Opto 22 has been used to add
functionality to the bioreactor system. Presto Control utilizes a number of interconnected
components to monitor the bioreactor environment and implement control schemes.
These components are the OptoDisplay program, the OptoControl program, the control
unit, the recipe data ﬁle, the individual I/O modules, and the BioFlo unit.

OptoDisplay

OptoDisplay is the visual interface program that allows the operator to interact with the
control unit. It displays the monitored fermentation variables and transmits the values
that the operator enters into the computer. The computer and OptoDisplay, however, do
not actually make any process calculations or controller decisions. In fact, the control
unit can control the bioreactor and implement a control scheme without being connected
to the computer. The majority of this manual is dedicated to describing the graphical
interface that is governed by the OptoDisplay program.

OptoControl

OptoControl is a computer program that is used to design control strategies. It uses a
graphical programming interface to represent multiple command loops running in
parallel. The program also includes built in control processes such as PID calculations
and timers. After an OptoControl strategy is created and compiled, it must be
downloaded into the control unit before it can be run.

Control Unit

The control unit is a stand alone processor that implements the control strategy created
using OptoControl. It calculates all of the process values and sends and receives signals
from the computer and the I/O modules.

Recipe

All of the parameters that deﬁne the stages of a control scheme in Presto Control are
represented as tables of values within the OptoControl strategy. For example, a table
titled “Stage_Time_Trigger_t” holds a list of 1’s and 0’s that signiﬁes for the controller
whether the time trigger is active for the corresponding stage. A feature of Opt022 is its
ability to use text ﬁles called recipes to store data for table variables in control strategies.
Presto Control utilizes these recipes to save a control scheme so that it can be used in
future fermentations by the operator.

IlO Modules
The output from the probes that monitor the bioreactor conditions are analog signals.
The U0 modules convert the analog signals to digital signals that the control unit can

126

read. The 1/0 modules also convert the digital signals that the control unit sends to the
BioFlo to analog. The calibration of the signals is a function of how the variables are
deﬁned in the OptoControl strategy.

BioFlo Unit

A description of the BioFlo He can be found in the Protocols for Protein Expressi_op. The
BioFlo controls the agitation, temperature, and pH in the bioreactor and transmits data on
those parameters to the control unit. The control unit handles air ﬂow and feed rate using
its own processes and individual control modules.

127

Bioreactor Control Overview

The goal of fermentation control is to establish environmental conditions within the
bioreactor that optimize cell growth and/or product accumulation. The optimum
environment can vary greatly from one project to the next and it depends on factors such
as the host organism, the type of product being produced, and the media and expression
system being employed. Also, the amount of knowledge that an operator has of the
bioreactor environment is limited by the probes available for monitoring it. Less
information results in fewer variables available for modeling and control of the cellular
state. Therefore, a large number of possible control schemes exist. Advanced methods
include the use of artiﬁcial neural networks (ANN), physiological state (PS) variables,
and modeling of the enzymatic pathway ﬂuxes. Presto Control does not utilize any direct
mathematical modeling. Instead, it relies on the operator’s expectations of how a culture
should react to changes in the control parameters. These expectations are used to set up a
knowledge based (KB) control scheme.

Advanced modeling of a bioreactor requires a wide variety of probes. For example, a
ﬂow through spectrophotometer can be used to calculate cell growth, a glucose analyzer
can monitor substrate consumption, and off gas analysis of C02 production can indicate
the respiration rate. In order to reduce the cost and complexity of the system, Presto
Control employs a relatively small number of probes to measure environmental
conditions. The monitored variables are temperature, pH, and dissolved oxygen
concentration (DO).

Most fermentations are run at a constant temperature and constant pH. The BioFlo
controls the temperature through the use of a water jacket and the pH through the use of
acid and base pumps. They are neither controlled nor monitored by Presto Control. If a
pH or temperature change needs to be made to increase protein production, the
adjustments must be manually implemented by the operator.

The primary environmental condition monitored by Presto Control is the DO. Even
without other information such as substrate concentration, the DO can be used to
determine growth rate, substrate consumption and depletion, and cell concentration. The
fermentation parameters that Presto Control can use to control the D0 are the agitation
rate, the air ﬂow rate, and the feed rate. Most control schemes are designed to keep the
oxygen concentration constant throughout the growth of the culture because anoxic
conditions can irreversibly harm the growth of certain host organisms such as E. coli.

During exponential growth, the rate of increase in cell mass can be deﬁned by the Monod
Equation: '
—dX p x ——S x X
= 1

dt max K S + S ( )
Where X is the cell concentration, pm is the maximum speciﬁc growth rate, S is the
substrate concentration, K5 is the kinetic constant for the substrate, and t is the time. The
value of pm is heavily dependent on temperature, pH, media composition, and the host

128

strain. Note that the model does not take into account volume increases due to feed
addition during fed batch growth. Also, the growth rate can be inhibited by low oxygen
concentrations, but the relationship is much more complex than the standard Monod
model can accurately predict. A measured D0 of 10% in a well mixed lab-scale
bioreactor can assure that no portion of the cell culture is starved of oxygen. A graphical
representation of how the growth rate varies with the substrate concentration when K,

equals 0.001 g/L (kinetic constant for E. coli growing on glucose) can be seen in Figure
1 1.

Figure 11: Glucose Limited Growth
1

 

 

 

/
/

 

 

 

 

 

totalizinba'xrboio

 

 

Fraction of Maximum Growth Rate
:3 o o o o o o o o

 

 

o —L
_‘

I I T r I

0.02 0.04 0.06 0.08 0.1 0.12
Glucose (gIL)

0

Except in specialized cases, neither the oxygen nor the substrate concentrations should be
growth rate limiting. The cellular uptake mechanisms for both of the molecules should
be saturated to produce optimum growth kinetics.

Oxygen is consumed by the cell culture as a result of cellular growth, cellular
maintenance, and protein production. The accumulation of oxygen in the bioreactor is
equal to the difference between the rate of oxygen transport and the consumption of
oxygen by the culture. It can be mathematically represented by:

d0

I = kLa(0 * '0)- 40 (2)

Where k La is the mass transfer coefﬁcient of the sparged air, 0* is the concentration of

oxygen in the media in equilibrium with the sparged air, 0 is the concentration of oxygen
in the vessel, and qo is the rate of consumption of oxygen by the cells per unit volume.

129

Equation 2 assumes that the system is well mixed and the concentration of oxygen in the
sparged gas does not decrease as the bubble travels through the bioreactor. This
assumption is acceptable for lab-scale fermentations.

A simple representation of qo is:
1 (IX

= Y X / 0 dt
Where Y X/O is the cell/oxygen yield coefﬁcient for growth, and mo is the rate of oxygen
consumption by the cells for maintenance. Combining Equations 1 and 3 produces:

 

qo +m0X (3)

 

1 S

40 [YXloﬂmaxxKS+S+m0]X (4)
Equation 4 demonstrates that the rate of oxygen consumption is proportional to the cell
concentration during exponential growth. Therefore, in order to maintain a constant
oxygen concentration within the bioreactor, the rate of oxygen transport, kLa (O*-O) will
have to increase in relation to the cell concentration. The oxygen concentration gradient,
(O*—O), can be increased by increasing the oxygen concentration in the sparged gas.
Presto Control cannot control the composition of the gas. Therefore, the area available
for oxygen mass transfer must be increased, and Presto Control has two methods for
inﬂuencing the value of km. One is to increase the volume of air entering the bioreactor
by increasing the air ﬂow rate. The other method is to increase the agitation rate, which
acts to break up the sparged air into smaller bubbles and increase the overall surface area.
Increasing the agitation rate also increases the rate of mixing which helps to maintain a
consistent oxygen concentration throughout the bioreactor. A common practice in
fermentation control is to ﬁrst increase the agitation to a maximum value and then ramp
up the air ﬂow rate.

Figure 12 is a graph of kLa for the BioFlo for various air ﬂow and agitation rates. The
values were determined based on experimental data and the experimental protocol
outlined in the Exmriment Manual for the Biochemical Engineering Laboratogy (Worden
et al., 2002). Equation 2 can be integrated to obtain:

21:-
In L—Q— =kLaXt (5)
0*-00

Where 00 is the measured oxygen concentration at time equals 0. Equation 5 is obtained
by assuming no oxygen consumption and constant 0*. The latter assumption is valid
when the rate of oxygen transfer is small compared to the response of the DO probe and
the composition of the air bubbles in the bioreactor is uniform. Mass transfer coefﬁcients
were obtained by forcing step changes in the composition of the sparged gas and using
least squares analysis to ﬁt the resulting changes in the measured oxygen concentration to
Equation 5. In order to allow the assumption of a constant 0* to be valid, only the data
obtained as the concentration approached steady state were used. 0* was then assumed
to be the equilibrium solubility of oxygen in contact with the inlet gas stream. Bioreactor
conditions were as described in the BioFlo Protocol. The bioreactor contained 1 L of LB
media with 0.05 mL of added antifoam agent. Agitation was created by two six spoke,
ﬂat bladed impellers with a diameter of 2 1/8”. The lower impeller was placed just above

130

the sparging ring at the end of the impeller shaft. The second impeller was completely
submerged just below the surface of the liquid. The measured kLa values were very much
dependent on these conditions. Increasing the media volume or changing the positions of
the impeller could increase the maximum kLa rates.

Figure 12: kLa for the BioFlo IIc

 

 

20° " +100 RPM
130 4 +425 RPM
+750 RPM
160 /
140

 

 

 

 

 

.5
N
O

 

 

kLa (hr4-1)
8
\

 

 

 

 

 

 

 

 

 

60
40
20 /
0 9* . . + . a
0 0.2 0.4 0.5 0.8 1 1 .2

Air Flow (lein)

The maximum agitation and air ﬂow rates are limited by the foaming and frothing they
create. If either is too great, the broth can be forced out the seals in the head plate or into
the air ﬁlters. This may cause pressure to build up or contamination to occur. In
addition, shear forces from bubble coalescence and agitation can kill certain cell types.
Therefore, the respiration rate of the culture is limited by the oxygen transfer restrictions
as indicated by the maximum km of the bioreactor. When the airﬂow and agitation have
reached their maximum rates, the rate of oxygen consumption by the cells needs to be
limited so that the cells can continue to grow in an aerobic state. This is done by
reducing the growth rate of the cells. According to Equation 1, this can be accomplished
by limiting the substrate concentration in the bioreactor until the cells are essentially
starved of food and can no longer grow at their maximum speciﬁc growth rate. In most
microbes, the stress response to oxygen starvation is much more harmful than the
response to substrate depletion.

Combining Equations 2 and 4 produces:

131

 

 

d0 1
— = k a 0 * -0 - x

dl L( ) [YX/oﬂmax KS-l-S
During substrate limited growth, the DO can be maintained by control of the feed rate.
Substrate accumulates in the bioreactor at a rate that is equal to the difference between its
feed rate and it consumption. Its representative Equation is:

d5

7; = D5 - 45 (7)
Where BS is the feed rate of the substrate per unit volume of the culture, and qs is the rate
of substrate consumption by the cells per unit volume of the culture. If the controller
needs to maintain a steady, growth limiting concentration of substrate, then the feed rate
must equal the rate of feed consumption. When the dissolved oxygen increases above its
desired value, the feed rate is increased in order to increase the substrate concentration
and allow a higher rate of cell growth. When the DO drops below its set-point, the feed
rate is decreased in order to limit growth.

'1' m0 ]X (6)

A simple representation of qs is:

1 (IX
4 s = ‘1' m s X (8)
Y X / 5 dl
Where Yx/s is the cell/substrate yield coefﬁcient, and ms is the rate of substrate
consumption by the cells for maintenance. As with oxygen consumption, combining

Equations 1 and 8 produces:

1 S
As with the respiration rate, the overall rate of substrate consumption is proportional to
the cell concentration.

 

 

It is often assumed that the rate of oxygen consumption is proportional to the rate of
substrate consumption during exponential growth. Oxygen and substrate consumption
are linked by their yield coefﬁcients and the molecular composition of the cell. Overall
yield coefﬁcients, Y ’m and Y 'X/o. can be obtained when the individual yield coefﬁcients
and maintenance requirements of the cell are constant during exponential growth.

 

 

 

1 l S
= —— ————+ (10)
Y'XIS [YXIS #mx sz+5 ms)
1 l S .
. = __,, x mm] (11)
YX/o [YX/O max Ks+5

Substituting Equation 10 into Equation 9 and Equation 11 into Equation 4 and dividing
the rate of substrate consumption by the rate of oxygen consumption, Equation 12 is
obtained.

ﬁzl'xmody-S/O (12)
40 Y x /0
Where Y’s/o is the mass of substrate consumed per mass of oxygen consumed by the
culture. Assuming a value of 0.5 g/g for Y 'X/s and a value of 1 g/g for Y 'oxs and a
maintained D0 of 10% of saturated air in the bioreactor, the rate of glucose and air

132

consumption for various air ﬂow and agitation rates can be obtained by comparison to the
graph of km values. Figure 13 shows a graph of the expected consumption rates. This
knowledge can be useful in designing control schemes for bioreactor control.

Figure 13: Mass Transfer and Consumption in the BioFlo IIc

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1.4
-+- 100 RPM

1 2 .. +425 RPM / -~ 2.5 A
A +750 RPM / g,
i. 1 2 .5
3 ‘6.
.g E
0- 0.8 :1
E //I —— 1.5 g
3 o
9 0
ﬂ
0 0.6 o
0 111'
C 9 1 .3
O In
S 0.4 .g
C (D

9 0.5
0.2
+_ —o
0 I I I I I O
0 0.2 0.4 0.6 0.8 1 1.2

Air Flow lein

The above description is a simpliﬁed model of cell growth and metabolism. The
exponential growth patterns described by Monod kinetics cannot account for lag phases,
steady state growth, or inhibition by products or media components. Also, the yield
coefﬁcients and maintenance needs vary depending on the metabolic state of the culture.
The main points to remember are that, in the absence of inhibitor, oxygen transfer in the
bioreactor is the limiting factor in high cell density culture growth, and the rate of oxygen
consumption is dependent on the substrate concentration during substrate limited growth.

In recombinant engineering, product expression in the cell is often designed to be induced
by a change in the environmental conditions in the bioreactor. For example, induction
could be triggered by the addition of a media component such as IPTG, a change in
temperature, or a change in pH. Rather than maximizing cell growth, the goal of the
bioreactor controller during induction is to maximize product expression. For example,
the culture may need to be kept in oxygen starved conditions if the product is sensitive
DO levels. Therefore, the control scheme will need to consist of two stages. One stage
will be active during culture growth while the other stage will be initiated at the time of
induction.

133

Recipes
Presto Control is fundamentally a simpliﬁed programming tool built on the framework of
the Opto 22 software. Whereas the OptoControl software offers a generic method of
programming a controller, Presto Control limits the variables available for making
controller decisions to just those parameters that characterize a fermentation. These
variables are the dissolved oxygen, the agitation rate, the air ﬂow rate, the feed rate, the
age of the fermentation, and the amount of feed added to the bioreactor. When a speciﬁc
value of one of these variables is reached, a controller action is triggered. This action is
the result of a new stage being called into operation that is deﬁned by different control
parameters and trigger values. Therefore, the control scheme must be broken up into
specific stages in order to program it into Presto Control. Every controller action requires
its own stage. For example, the control scheme may call for PID control of D0 to be
transferred from agitation to the air ﬂow rate when the agitation reaches 750 RPM. In
Presto Control, this is accomplished by ﬁrst deﬁning a stage in which D0 is controlled by
agitation and the stage’s trigger is an RPM greater than 750. The ﬁrst stage’s trigger
calls a second stage in which D0 is controlled by air ﬂow rate. Since two sets of triggers
exist, the controller has a choice of two paths leading out of a stage. Each of these paths
can also branch out, thereby creating a virtually limitless number of choices and control
schemes.

Seven example recipes can be directly loaded from the Recipe Loader Menu of the
Fermentation Setup Window. These recipes were developed in order to allow the operator
to test the ability of different feed control schemes to deal with each expression system’s
individual requirements. Due to their complexity and unreliability, some of these
schemes have more of a conceptual rather than practical value. The recipe parameters
may also need to be customized depending on the growth characteristics or induction
strategy of the strain.

Each of the predeﬁned recipes is an adaptation of a control strategy that has been used by
other researchers to stop the accumulation of inhibitory concentrations of fermentation
byproducts. Other than the toxic effects of induction and recombinant protein expression,
the most commonly reported inhibitor of high cell density culture growth in bioreactors
are the cellular byproducts lactic and acetic acid that are formed via anaerobic metabolic
pathways. The culture creates these byproducts when it consumes more substrate than it
can fully oxygenate in the respiratory cycle. Various models for this phenomenon have
been proposed, but the environmental indicators are a relatively high cell growth rate, and
saturated substrate and oxygen uptake rates.

Culture growth can also be inhibited by the stringent response of the cells due to
prolonged starvation. A very brief substrate depletion event, as indicated by a quick
spike in the DO due to a reduction in the respiration rate of the cells, is not overly
harmful, however. The goal of many control schemes is to keep the feed below the
uptake saturation concentration while also keeping the culture from experiencing
starvation for long periods of time.

134

These control schemes all have some characteristics in common. All assume a working
volume of l L. Also, the required initial feed mass is supplied in the ﬁrst stage of the
control scheme so substrate does not need to be added to the media during bioreactor
preparation. In all but the “Maximum DO Control” recipe, this ﬁrst stage also waits for
the oxygen uptake rate of the culture to meet the minimum oxygen transfer rate of the
bioreactor. Then, once the dissolved oxygen has dropped to its set-point value, the
agitation rate is allowed to ramp up to its maximum value before the air ﬂow ramp is
initiated. If both the agitation and the air ﬂow reach their maximum values, the dissolved
oxygen is maintained by PID control of feed with checks to determine that excessive
feeding does not occur.

Feed depletion in the media is assumed to have occurred when the DO increases a
speciﬁed amount above its current value or set-point. In most recipes, the feed depletion
event is conﬁrmed if the DO drops in response to the feeding of substrate into the
bioreactor. If feed does not cause a response, a preset, open loop feed control scheme is
used to make sure that the culture does not suffer from starvation before the dissolved
oxygen returns to its set-point.

Each recipe description includes the reference from which the control scheme was
devised. In addition, a table lists the major features of the recipe and some subjective
judgments of its reliability and robustness:

0 Dissolved Oxygen Set-Point: The DO set-point can be changed In the ﬁrst stage and it
will carry through the entire recipe for the PID control, but the trigger values In each
stage for the detection of feed depletion will also need to be adjusted

0 Air Flow Fate Range: Describes the minimum and maximum air ﬂows spanned
before DO control is transferred to PID control by feed.

Agitation Range: The values that need to be input into the BioFlo.
Initial Feed Concentration: The amount of feed added in the ﬁrst stage. It can be
adjusted with little affect on the performance of the control recipe.

0 Momentary Feed Depletion: The expected number of DO spikes caused by feed
starvation that will be allowed in the recipe per hour.

0 Chance of Excessive Feed: The likelihood that the scheme will fail and dump in more
feed than intended by the design of the recipe.

0 Chance of Prolonged Starvation: The likelihood that the control scheme will fail and
allow the culture to starve for extended periods of time without adding feed.

0 Growth Rate Limitations by Feed: The lower the substrate concentration is below its
saturation level, the slower the culture will grow.

0 Average Feed Concentration: A comparison of how strict the recipe is in maintaining
low substrate concentrations relative to the other recipes.

0 Reliability of Feed Scheme: Subjective measurement of how well the scheme will
function without adaptation or constant monitoring by the operator.

Standard Fed Batch Control
Reference: Knop, 2002

135

This is an adaptation of a control strategy created by David Knop for growing high cell
density cultures in the BioFlo. This type of staged control scheme is common for
controlling laboratory scale batch fermentations. A large feed mass is added initially
with the intention of sustaining culture growth until both the agitation and the airﬂow
have ramped to their maximum values and DO control by feed rate is initiated. If feed
depletion occurs before this point, a measured quantity of feed is added to support growth
to the desired density without overshooting. This scheme allows the least chance of feed
depletion or starvation during the exponential growth phase.

Table 4: Standard Fed Batch Control

 

Dissolved Oxygen Set-Point (%): 10

 

Air Flow Fate Range (Umin): 0.1 — 1.0

 

Agitation Range (RPM): 50- 750

 

 

 

 

 

Initial Feed Concentration (g/L): 10
Momentary Feed Depletion (events/hr): 0
Chance of Excessive Feed: very low
Chance of Prolonged Starvation: low
Growth Rate Limitations by Feed: none

 

Average Feed Concentration: very high

 

 

 

 

Reliability of Feed Scheme: veryhggh

 

The following is a walk through of the recipe that deﬁnes the Standard Fed Batch control
scheme. This discussion and the included recipe maps give a better understanding of
some of the features of recipe creation in Presto Control.

Stage 1: DO Drop

Control: This stage sets up the initial environmental conditions of the bioreactor. The
set-point for the D0 is speciﬁed as 10% and this value is carried over throughout all of
the rest of the stages. In addition, the air ﬂow rate is set to 0.1 lJmin and held constant
until PID control is speciﬁed in stage 5. Stage 1 also adds the initial feed charge of 10 g
with the feed pump set to its maximum rate. This mass can be modiﬁed without
seriously affecting the control scheme.

Trigger A (Stage 2): When the bioreactor is ﬁrst inoculated, the culture usually cannot
consume enough oxygen to cause the dissolved oxygen concentration to drop below its
set-point of 10%. At the minimum air ﬂow and agitation rates of the bioreactor, the mass
transfer rate of oxygen at its set-point concentration is greater than the culture’s oxygen
uptake rate. Subsequent stages in the control scheme detect a feed depletion event by the
DO rising above 15%. Therefore, trigger A does not call the next stage until the DO has
dropped below the set-point. An additional constraint of trigger A is that the feed rate
must drop below 0.001 g/hr. This indicates that the initial 30 g of feed has been added
and the feed pump has shutoff. Trigger A is also set to wait for a fermentation time of 5
hours before activating. The stage called by the trigger checks for both feed depletion
and a maximum agitation rate, neither of which is likely to occur within ﬁve hours of
inoculation. The Fermentation Time Trigger keeps the controller from responding to a
false feed depletion event, but it can be turned off without affecting the basic operation of
the control scheme. When all of these trigger values are met, trigger A calls stage 2.

136

Trigger B (Stage 14): Trigger B calls stage 14 after 24 hours of fermentation time. This
trigger can only be met if the DO does not drop below 10% within a day of inoculation.
The only likely reasons for this to happen are culture death or a malfunctioning of the DO
probe. In case of a DO probe malfunction, the called stage adds feed at a preset rate in
order to keep the culture from starving. The time value can be reduced depending on the
operator’s expectations of culture growth.

Stage 2: RPM Ramp

Control: This stage is intended to allow the agitation to ramp up to its maximum value of
750 RPM before the air ﬂow ramp is enacted. The air ﬂow rate is again set to 0.1 Umin
because stage 4 calls this stage if the air ﬂow drops too low during Air PID Control.

Feed control is set to Full On even though no mass of feed is directly speciﬁed. When
this stage is called by stage 1, the feed control immediately sets itself to Full 013‘ since the
feed mass is 0. Stage 10, however, also calls stage 2, and the control scheme has been
designed to allow stage 2 to ﬁnish feeding the feed mass speciﬁed by stage 10.

Trigger A (Stage 3): When the BioFlo has ramped up the agitation rate past 750 RPM,
stage 3 is called so that PID control by the air ﬂow rate can be initiated.

Trigger B (Stage 10): Feed depletion is indicated by a spike in the DO. The culture
stops growing and consuming oxygen when the available substrate is completely
consumed. The result is such a quick accumulation of oxygen in the broth that the
controller cannot decrease the agitation rate quick enough to keep the DO from
increasing at least 5% above its set-point. Trigger B is used to detect this feed depletion
event. When the DO increases above 15%, stage 10 is called to add more feed to the
bioreactor.

Stage 3: RPM Check

Control: A malfunction in the impeller motor might cause the agitation to accidentally
spike above 750 RPM and activate the triggering of this stage by stage 2. Stage 3
attempts to detect this type of problem before PID control by air ﬂow is initiated in the
subsequent stages.

Trigger A (Stage 4): After 30 seconds with an agitation rate greater 750 RPM, it is
assumed that the agitation has ramped up to its maximum value due to oxygen
consumption by the culture. Trigger A calls stage 4, which initiates the air ﬂow ramp.
Trigger B (Stage 2): If, within the 30 seconds allowed by Trigger A, the agitation drops
back below 750 RPM, Trigger B sends the control scheme back to stage 2.

Stage 4: Air Ramp

Control: Stage 4 is designed to allow the PID controller to ramp up the air ﬂow rate to its
maximum value. As in stage 2, the feed control is set to Full On.

Trigger A (Stage 5): As in stage 2, an increase in the DO above 15% signals that all of
the feed has been consumed. This event triggers stage 5.

Trigger B (Stage 2): In case the oxygen uptake rate of the culture substantially decreases
during stage 3, trigger B stops the PID controller from forcing the air ﬂow to drop below
its minimum value. The trigger sends control back to stage 2 where the air ﬂow rate is
reset to 0.1 Umin. This trigger assumes that control of agitation by DO set-point is still

137

 

active on the BioFlo. If desired, Trigger B can be inactivated without harming the
control scheme.

Stage 5: Feed Depletion

Control: This stage begins to feed substrate into the bioreactor in response to the
detection of a feed depletion event by stage 4. It also detects whether the air ﬂow has
reached its maximum rate and determines whether PID control by feed should be
attempted. This check of the air ﬂow rate is only done after the detection of a feed
depletion event because PID control by feed is ineffective if the culture is not substrate
limited.

Trigger A (Stage 6): After 5 s with the air ﬂow rate being maintained above 1 Umin,
stage 6 is called to transfer PID control by the DO from air ﬂow to feed rate.

Trigger B (Stage 15): If the air ﬂow rate is below 1 Umin, a series of stages is called
that determines how much feed should be added to allow the culture to grow until the
maximum air ﬂow rate is reached. The ﬁrst stage of this series is stage 14.

Stage 6: PID check

Control: If PID control by feed rate is attempted too soon in a fermentation, a glucose
dump might occur. Stage 6 determines whether PID control by feed has been initiated
before the culture has become substrate limited. The stage drops the air ﬂow rate down
to l Umin and begins feeding 1 g of substrate at an initial rate of 8 g/hr by PID control.
Trigger A (Stage 29): This trigger is met if PID control by feed is not effective in
bringing the DO back down to a value below its set point. This would occur if the DO
spike in stage 4 was not caused by feed depletion and the bioreactor has not yet reached
its maximum oxygen transfer capacity. A gram of feed is allowed to be added before
stage 29 is called. The triggering event is a drop in the feed rate below 0.01 g/hr due to
the feed control switching to Full 01?.

Trigger B (Stage 30): If PID control by feed causes the D0 to drop below its set-point,
the culture is substrate limited. Stage 30 is called to establish the required PID controlled
feed rate to support cell growth.

Stage 10: Start Feed

Control: This stage begins feeding substrate at the maximum rate in response to the feed
depletion event detected in stage 2. The stage speciﬁes that 10 g should be added to
supplement growth. This mass can be changed by the operator depending on the needs of
the culture and the goal of the fermentation.

Trigger A (Stage 2): If the addition of the feed mass causes the D0 to drop back below
14% within 30 seconds of the triggering of the stage, the assumption built into the control
scheme is that the DO spike was a result of a feed depletion event. Therefore, stage 2 is
called and the rest of the feed mass is added.

Trigger B (Stage 11): The DO might rise above 15 % in stage 2 due to failure of the DO
probe or an accidental spike in the air ﬂow or agitation. Therefore, if the feed pulse is
ineffective in reducing the DO, trigger B calls stage 11 in order to halt the addition of
feed. Stage 11 is also called if more than 41 g of feed has been added to the
fermentation. Signiﬁcant amounts of feed should not be required to support the culture
during the RPM ramp portion of the control scheme. The addition of over 41 g indicates

138

that the DO might be ﬂuctuating due to controller oscillations or malfunctioning
equipment. The speciﬁed trigger mass can be changed according to the needs of the
operator.

Stage 11: Wait Response

Control: After it is determined that the DO has not spiked above 15% due to feed
depletion, stage 11 stops the feed and removes the excess mass speciﬁed in stage 10.
This keeps stage 2 from adding the rest of the feed when it is ultimately called. Stage 12
also calls stage 11 to have its Maximum Feed Mass reset. Stage 11 waits for the D0 to
return to normal as the bioreactor adapts to whatever caused the DO spike.

Trigger A (Stage 2): When the DO drops below 14%, this trigger calls stage 2 so that
normal control operations can be reinstated.

Trigger B (Stage 12): If, after 5 minutes, the DO still has not decreased, it is assumed
that the control system has failed and stage 12 is called in order to implement a safety
procedure.

Stage 12: Safety

Control: This stage is intended to keep the culture from starving until the control system
is corrected by the operator. Feed is added at a constant rate of 4 g/hr. The air ﬂow rate
is speciﬁed to be 0.1 Umin in order to return the air ﬂow to this minimum value when
stage 12 is called by stage 22.

Trigger A (Stage 11): Control is returned to stage 11 and then stage 2 when the DO
returns to the expected value.

Trigger B (No Stage): Manual hold.

Stage 15: 0.8 — 1 Flow

Control: Stages 15 through 19 are used to ﬁnd the current air ﬂow rate and determine the
correct quantity of feed to add to support growth until PID control by feed is initiated.
Each air ﬂow rate range (i.e. 0.8 to 1.0 Umin) allows a different amount of feed to be
added to the bioreactor. The speciﬁed mass can be changed by the Operator depending on
the expectations of the cultures growth rate and oxygen uptake rate. If the correct air
ﬂow rate range is represented by the stage, a time delay is allowed before the next stage
is automatically called in order to determine if the DO responds to the addition of
substrate. Stage 15 allows l g of feed to be added.

Trigger A (Stage 20): Trigger A is the same for stages 15 through 19. It gives the DO
30 seconds to drop below 14.9% in response to the addition of the feed. Stage 20 is then
called to determine how the culture responded to the feed pulse.

Trigger B (Stage 16): If the air ﬂow rate is not between 0.8 to 1.0 Umin, stage 16 is
called in order to ﬁnd the correct amount of feed to add.

Stage 16: 0.6 — 0.8 Flow

Control: The function is the same as stage 15 except that stage 16 adds an additional 2 g
of feed over the previous stage.

Trigger A (Stage 20): Same as stage 15.

Trigger B (Stage 17): This trigger calls stage 17 to check for air ﬂow rates below 0.6
Umin.

139

Stage 17: 0.4 - 0.6 Flow

Control: The function is the same as stage 15 except that stage 16 adds an additional 2 g
of feed over the previous stage.

Trigger A (Stage 20): Same as stage 15.

Trigger B (Stage 18): This trigger calls stage 18 to check for air ﬂow rates below 0.4
Umin.

Stage 18: 0.2 - 0.4 Flow

Control: The function is the same as stage 15 except that stage 16 adds an additional 2 g
of feed over the previous stage.

Trigger A (Stage 20): Same as stage 15.

Trigger B (Stage 19): This trigger calls stage 19 if the air ﬂow rate is below 0.2 Umin.

Stage 19: 0.1 - 0.2 lein

Control: Stage 19 is the same as the rest of the stages in the series except that it allows
PID control by air ﬂow. Also, Trigger B is modiﬁed.

Trigger A (Stage 20): Same as stage 16.

Trigger B (Stage 10): Since 0.1 Umin is the lowest allowable air ﬂow rate, this trigger
sends the control scheme back to PID control by agitation if the air ﬂow rate drops below

0.1 Umin. Stage 10 is called initially in order to allow the control scheme to respond to
the DO spike.

Stage 20: Response Check

Control: Stage 20 is called by all of the DO spike response stages to determine how the
DO has responded to the addition of feed. The stage checks to see whether the D0 is
dropping as quickly as would be expected if the feed had become depleted. If the D0 is
oscillating with a slow period in response to poor PID tuning, this stage keeps excessive
feeding from occurring. A check of the total feed added to the bioreactor is also made.
Trigger A (Stage 21): Trigger A gives the DO 10 seconds to drop below 14% before
calling stage 21, which cuts off feed. Stage 21 is also automatically called if 60 grams of
feed has already been added during the fermentation. The reasoning for this controller
response is that the respiration rate of the culture should have exceeded the maximum
oxygen transfer rate of the bioreactor by the time 60 grams of feed has been consumed.
This mass value can be changed according to the needs of the fermentation.

Trigger B (Stage 4): If the DO drops an additional 0.9% from the calling stage, trigger B
calls stage 4 and allows the rest of the feed mass to be added during normal ramping of
the air ﬂow rate.

Stage 21: Wait Response

Control: This stage has the same function as stage 11 of the agitation ramp portion of the
control scheme. It waits for the D0 to approach it set-point value without adding
additional feed.

Trigger A (Stage 22): If it takes over ﬁve minutes for the D0 to return to normal, stage
22 is triggered in order to implement the safety response. This safety response is also
called if the DO increases by 5%, indicating that another DO spike has occurred due to

140

either feed depletion or continuing failure of the control scheme. Another trigger for the
safety stage is a drop in the air ﬂow to value below its minimum rate of 0.1 Umin.
Trigger B (Stage 4): When the DO drops back to its set-point, this trigger calls stage 4 to
return to normal ramping of the air ﬂow rate.

Stage 22: Safety

Control: Stage 22 has the same purpose as stage 12. It feeds substrate at 8 g/hr to keep
the culture from starving. If required, it also calls stage 12 in order to keep the PID
controller from starving the cells of air.

Trigger A (Stage 29): When the DO has regained its expected value, stage 29 is called to
reset the control scheme.

Trigger B (Stage 12): Stage 12 is called if the high DO that led to the triggering of stage
22 causes the PID controller to drop the air ﬂow rate below its minimum value of 0.1
Umin. Stage 12 reestablishes an air ﬂow rate of 0.1 Umin while waiting for the D0 to
return to its set point.

Stage 29: Reset

Control: This stage is called for the sole purpose of removing the Maximum Feed values
speciﬁed by stages 22, 29, and 30.

Trigger A (Stage 4): The stage is given 1 s to complete its task before stage 4 is called.
The time lag speciﬁed by the trigger is not actually required for all of the stage
parameters to be loaded.

Trigger B (No Stage): Manual hold.

Stage 30: P11) SS

Control: The maximum oxygen uptake rate of the culture must be signiﬁcantly higher
than the maximum oxygen transfer capacity of the bioreactor before PID control by the
air ﬂow rate is discontinued. Otherwise, the DO might rise above its set-point
irrespective of the feed rate and a glucose dump can occur. Stage 30 is meant to make
sure that PID control by feed rate is necessary for keeping the culture from consuming
too much oxygen. After the DO drops below 10% in stage 6, PID control by feed is
maintained in stage 30 until a gram of feed has been added. At the end of this period, it is
expected that the controller has dampened the feed rate oscillations caused by the
initiation of PID control. Control of the fermentation is then sent to one of the Feed PID
control schemes. If the control scheme detects that a glucose dump is occurring because
oxygen transfer is greater than its uptake, control is instead returned to PID control by air
ﬂow.

Trigger A (Stage 29): Trigger A attempts to detect whether a glucose dump is occurring.
A glucose dump is a result of the PID controller ramping up the feed rate in response to
the DO not dropping down to its set-point. Therefore, if the feed rate exceeds 30 g/hr,
stage 29 is called in order to return to PID control by air ﬂow. A rise in the DO above
14% also causes the triggering of this response. Both of these indicators of a glucose
dump are heavily dependant on the PID control parameters and the maximum oxygen
transfer rate of the bioreactor. Therefore, the PID and/or trigger values might need to be
tuned.

141

Trigger B (Stage 31): If a gram of feed has been successfully added without a glucose
dump being detected, the ﬁrst stage (stage 31) of the Feed PID control scheme is
triggered.

Reduced Fed Batch Control

Reference: Shiloach, 1996

This scheme is the same as the Standard Fed Batch Control except that the initial feed
concentration is smaller. In addition, only a small amount of feed is added when a feed
depletion event is detected. This scheme is intended to reduce acetate production due to
high glucose concentration in the media, but it allows a high rate of feed depletion events.

Table 5: Reduced Fed Batch Control

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Dissolved Oxygen Set-Point (%): 30
Air Flow Fate Range (Umin): 0.1 — 1.0
Agitation Range (RPM): 50- 750
Initial Feed Concentration (g/L): 2
Momentary Feed Depletion (events/hr): up to 60
Chance of Excessive Feed: low
Chance of Prolonged Starvation: low
Growth Rate Limitations by Feed: low
Average Feed Concentration: moderate
Reliability of Feed Scheme: high
Maxim um DO Control

Reference: Jun Sun, unpublished

This scheme is similar to a strategy described by Jun for controlling the BioFlo 3000 in
the PEL. Unlike the other recipes, this scheme does not employ a ramp of agitation and
airﬂow to maintain the DO at a set-point. Instead, the oxygen transfer rate is maximized
by setting the airﬂow and agitation at their maximum rates and the D0 is allowed to ramp
down as cell growth increases. When the DO spikes due to feed depletion, a small pulse
of feed is added to maintain a low average feed concentration. This is the least
complicated recipe.

Table 6: Maximum DO Control

 

 

 

 

 

 

 

 

 

 

 

Dissolved Oxygen Set-Point (%): maximum
Air Flow Fate Range (L/min): l
Agitation Range (RPM): 750
Initial Feed Concentration (g/L): 10
Momentary Feed Depletion (events/hr): 3O
Chance of Excessive Feed: very low
Chance of Prolonged Starvation: low
Growth Rate Limitations by Feed: low
Averzge Feed Concentration: moderate
Reliabilityof Feed Scheme: high

 

 

 

I42

Exponential Corrected Feed Forward Control

Reference: Nor, 2001

This is a relatively complicated recipe intended to keep the substrate concentration close
to the uptake saturation limit by maintaining the feed rate at the maximum substrate
uptake rate (MSUR). When the initial feed mass is depleted, a speciﬁed quantity of feed
is added to the bioreactor. The time required for this feed to be consumed and a DO
spike to occur is used to estimate the MSUR. This calculated MSUR is used as the basis
for an open loop, exponentially increasing feed rate. After 30 minutes, feed is stopped
and the exponential feed ramp rate is either increased or decreased depending on the time
required for feed depletion to occur. Then, the MSUR is recalculated as before and the
process is repeated.

Table 7: Exponential Corrected Feed Forward Control

 

 

 

 

 

 

 

 

 

 

 

Dissolved Oxygen Set-Point (%): 10
Air Flow Fate Range (Umin): 0.1 — 1.0
Agitation Range (RPM): 50- 750
Initial Feed Concentration (g/L): 2
Momentary Feed Depletion (events/hr): 4
Chance of Excessive Feed: low
Chance of Prolonged Starvation: moderate
Growth Rate Limitations by Feed: low
Average Feed Concentration: moderate
Reliability of Feed Scheme: moderate

 

 

 

Minimum Feed Rate Control
Reference: Suzuki et al., 2001

This recipe adds feed at a constant rate while attempting to detect relatively small but
signiﬁcant increases in the DO as a signal that the current feed rate is too low. The feed
rate is then increased and the process is repeated. This recipe can lead to excessive feed

rate limitations on growth if it is not tuned properly for detection of the DO increase.

Table 8: Minimum Feed Rate Control

 

 

 

 

 

 

 

 

 

 

 

Dissolved Oxygen Set-Point (%): 50
Air Flow Fate Range (Umin): 0.1 - 1.0
Agitation Range (RPM): 50- 750
Initial Feed Concentration (g/L): 5
MomentarLFeed Depletion (events/hr): minimal
Chance of Excessive Feed: moderate
Chance of Prolongd Starvation: very low
Growth Rate Limitations by Feed: high
Average Feed Concentration: low
Reliability of Feed Scheme: low

 

 

 

143

Maximum Feed Rate Control

Reference: Suzuki et al., 2001

This recipe includes all of the same features as the Minimum Feed Rate Control recipe.
In addition, after a gram of feed is added without the DO rising to signify the need for an
increase in the feed rate, the feed rate is doubled for a short period of time. The extent of
the decrease in the DO as a result of this step increase in feed is used to determine
whether the feed should be increased, decreased, or maintained at the current rate. The
result is a more robust control scheme that can detect if the feed rate is too high or if the
DO rises are not being detected.

Table 9: Maximum Feed Rate Control

 

Dissolved Oxygen Set-Point (%): 50

 

Air Flow Fate Range (L/min): 0.1 — 1.0

 

Agitation Range (RPM): 50— 750

 

Initial Feed Concentration (g/L): 5

 

Momentary Feed Depletion (events/hr): minimal

 

Chance of Excessive Feed: moderate

 

Chance of Prolonged Starvation: very low

 

Growth Rate Limitations by Feed: moderate

 

Average Feed Concentration: low

 

 

 

 

Reliability of Feed Scheme: low

 

Saturation Pulse Test Control

Reference: Akesson et al., 1999

This control scheme is based on the use of step changes in the feed rate to determine
whether the substrate and oxygen uptake mechanisms of the host are saturated. The
recipe attempts to maintain the feed concentration below the uptake saturation level in
order to reduce fermentation byproducts. If a step increase in the feed rate causes a drop
in the DO, then neither the oxygen nor the substrate cellular uptake mechanisms are
saturated and the nominal feed rate is increased. If a step decrease in the feed rate does
not induce an increase in the DO, then the substrate uptake mechanism is saturated and
the nominal feed rate is decreased. The step rate changes occur on a four minute cycle.
This allows 30 seconds for a feed rate step change and 90 seconds to recover from the
response. This recipe attempts to detect transient changes in the bioreactor environment
due to transient changes in a very complex system, and it can therefore be considered
relatively unreliable. Furthermore, at high substrate uptake rates, a step change in the
feed rate has a disproportionate effect on the rate of change in the substrate concentration
as compared to low substrate uptake rates. The parameters may need to be retuned for
different substrate uptake rates in order to be effective.

144

 

Table 10: Saturation Pulse Test Control

 

Dissolved Oxygen Set—Point (%): 30

 

Air Flow Fate Range (Umin): 0.1 — 1.0

 

Agitation Range (RPM): 50- 750

 

 

 

 

 

Initial Feed Concentration (g/L): 5
Momentary Feed Depletion (events/hr): minimal
Chance of Excessive Feed: low
Chance of Prolonged Starvation: moderate
Growth Rate Limitations by Feed: high

 

Average Feed Concentration: very low

 

 

 

 

Reliability of Feed Scheme: very low

 

Feed PID I & 2

The Feed PID recipes are supplementary control schemes for maintaining the dissolved
oxygen at its set point by PID control of the feed rate. They are intended to be loaded in
conjunction with one of the primary control schemes listed above. All of the primary
control schemes trigger the ﬁrst stage (stage 31) of the Feed PID control scheme when
the maximum air ﬂow and agitation rates are reached. If a Feed PID recipe has not yet
been downloaded, the controller will become trapped in an undeﬁned stage.

The Feed PID control schemes are designed to be universal; they maintain the DO set-
point, air ﬂow rate, and initial feed rate of the primary control scheme. The initial stages
of the Feed PID recipe are designed to wait for the PID controller to reach a near steady
state feed rate. They also make sure that the DO does not drop far below 10%. When a
steady feed rate is attained, the control scheme monitors the feed rate to make sure that
the feed concentration is the growth limiting factor. If the feed concentration is not
growth rate limiting, the respiration rate of the cells cannot be controlled by the feed rate.
Therefore, the DO might rise above its set-point irrespective of the feed concentration,
and the controller will respond by ramping up the feed rate in an attempt to reduce the
DO. This response is called a “glucose dump” and it can lead to toxic feed
concentrations and overﬁlling of the bioreactor.

If a glucose dump is detected, Feed PID 1 and Feed PID 2 respond in different ways.
Feed PID 1 responds by dropping the air ﬂow rate in order to reduce oxygen transfer and
bring culture growth back to feed limiting conditions. PII) control by feed is maintained.
Feed PID 2 responds by switching to DO control by air ﬂow rate. The controller then
waits for a DO spike to indicate feed depletion and adds a short pulse of feed. This
scheme is similar to that used in the later stages of Reduced Fed Batch Control.

The recipes for both Feed PID l and Feed PID 2 contain stages that assist in D0 control
during induction. Induction almost always causes a reduction in the respiration rate of
the cells. This leads to a glucose dump if PID control of the feed rate is active. To stop
this from happening, the controller sets the feed rate to be 20% greater than the average
PID feed rate and transfers DO control to the air ﬂow rate. This is accomplished if the
operator manually triggers stage 50 at the time of induction. The operator should only do

145

 

this if PID control of feed is active and at a near steady state rate. If the optimum
induction strategy does requires low substrate concentrations, an alternate control scheme
should be developed.

Table 11: Feed PID 1 & 2

 

Dissolved Oxygen Set-Point (%): undeﬁned

 

Air Flow Fate Range (L/min): undeﬁned

 

Agitation Range (RPM): undeﬁned

 

 

 

Initial Feed Concentration (g/L): 0
Momentary Feed Depletion (events/hr): minimal
Chance of Excessive Feed: high

 

Chance of Prolonged Starvation: very low

 

Growth Rate Limitations by Feed: very high I,

 

Average Feed Concentration: very low

 

 

 

 

Reliability of Feed Scheme: moderate

 

 

146

REFERENCES

147

 

REFERENCES

Akesson M., Karlsson E. N., Hagander P., Axelsson J. P. and Tocaj A. (1999). "On-Line
Detection of Acetate Formation in Escherichia coli Cultures Using Dissolved Oxygen
Responses to Feed Transients." Biotechnology and Bioengineering 64(5): 590-598.

Alltenback—Rehm J ., Nell C. and Arnold M. (1999). "Parallel Bubble Columns with Fed-
Batch Technique for Microbial Process Development on a Small Scale." Chemical
Engineering Technology 22: 1051-1058.

Bailey J. E. and Ollis D. F. (1986). Biochemical Engineering Fundamentals. New York,
NY, McGraw Hill.

Broadwater J. A. and Fox B. G. (1996). "Lactose Fed-Batch Fermentation: A High-Yield
Method Suitable for Use with the pET System." InNovations 3: 8-9.

Chen Y., Krol J ., Sterkin V., Weimin F., Xikang Y., Huang W., Cino J. and Julien C.
(1999). "New process control strategy used in a rapamycin fermentation." Process
Biochemistry 34: 383-389.

Epand R. F., Macosko J. C., Russel C. J ., Shin Y. K. and Epand R. M. (1999). "The
Ectodomain of HA2 of Inﬂuenza Virus Promotes Rapid pH Dependent Membrane
Fusion." Journal of Molecular Biology 286: 489-503.

Epand R. F., Yip C. M., Chemomordik L. V., LeDuc D. L., Shin Y. K. and Epand R. M.
(2001). "Self-assembly of inﬂuenza hemagglutinin: studies of ectodomain aggregation by
in situ atomic force microscopy." Biochimica et Biophysica Acta 1513: 167-175.

Garrett R. H. and Grisham C. M. (1999). Biochemistry. Fort Worth, TX, Saunders
College Publishing.

Glick B. R. (1995). "Metabolic load and heterologous gene expression." Biotechnology
Advances 13: 247-261.

Goto N. K. and Kay L. E. (2000). "New developments in isotope labeling strategies for
protein solution NMR spectroscopy." Current Opinion in Structural Biology 10: 585-592.

Grabski A., Mehler M. and Drott D. (2003). "Unattended high-density cell growth and

induction of protein expression with the Overnight ExpressTM Autoinduction System."
InNovations 17: 3-6.

Grabski A. C. and Burgess R. R. (2001). "Preparation of protein samples for SDS-
polyacrylamide gel electrophoresis: procedures and tips." InNovations 13: 10-12.

148

 

Grossman T. H., Kawasaki E. S., Punreddy S. R. and Osbume M. S. (1998).
"Spontaneous CAMP-dependent derepression of gene expression in stationary phase plays
a role in recombinant expression instability." Gene 209: 95-103.

Hernandez L. D., Hoffman L. R., Wolfsberg T. G. and White J. M. (1996). "Virus-Cell
and Cell-Cell Fusion." Annual Reviews in Cell Development Biology 12: 627-661.

Hirschel B. J., Shen V. and Schlessinger D. (1980). "Lactose operon transcription from
wild-type and L8-UV5 lac promoters in Escherichia coli treated with chloramphenicol."
Journal of Bacteriology 143: 1534-1537.

Hunke S. and Betton J. M. (2003). "Temperature effect on inclusion body formation and
stress response in the periplasm of Escherichia coli." Molecular Microbiology 50: 1579-
1589.

J aradat A. W. and Bhunia A. K. (2002). "Glucose and Nutrient Concentrations Affect the
Expression of a 104-Kilodalton Listeria Adhesion Protein in Listeria monocytogenes."
Applied and Environmental Microbiology 68: 4876—4883.

Julka S. and Regnier F. (2004). "Quantiﬁcation in Proteomics through Stable Isotope
Codinng Review." Journal of Proteome Research 3: 350-363.

Kim C. H., Macosko J. C. and Shin Y. K. (1998). "The Mechanism for Low-pH-Induced
Clustering of Phospholipid Vesicles Carrying the HA2 Ectodomain of Inﬂuenza
Hemagglutinin." Biochemistry 37: 137-144.

Kim C. H., Macosko J. C., Yu Y. G. and Shin Y. K. (1996). "On the Dynamics and
Conformation of the HA2 Domain of the Inﬂuenza Virus Hemagglutinin." Biochemistry
35: 5359-5365.

Knop D. (2002). Hydroaromatic Quilibration During Shikimic Acid and Ouinic Acid
Biosynthesis. Chemical Engineering and Materials Science. East Lansing, Michigan State
University.

Konstantinov K. B., Kishimoto M., Seki T. and Yoshida T. (1990). "A balanced DO-stat
and its application to the control of acetic acid expression by recombinant Escherichia
coli." Biotechnology and Bioengineering 36: 750-758.

Konz J. 0., King J. and Cooney C. L. (1998). "Effects of Oxygen on Recombinant
Protein Expression." Biotechnology Progress 14: 393-409.

Laage R. and Langosch D. (2001). "Strategies for Prokaryotic Expression of Eukaryotic
Membrane Proteins." Trafﬁc 2: 99-104.

149

 

Lab R. D. Fundamentals of Fermentation: Techniques For Benchtop Ferrnentors. Edison,
NJ, New Brunswick Scientiﬁc Co., Inc: 1-15.

Laemmli U. K. (1970). "Cleavage of structural proteins during the assembly of the head
of bacteriophage T4." Nature 227: 680-685.

LeDuc D. L., Shin Y. K., Epand R. F. and Epand R. M. (2000). "Factors Determining
Vesicular Lipid Mixing Induced by Shortened Constructs of Inﬂuenza Hemagglutinin."
Biochemistry 39: 2733-2739.

Lee S. Y. (1996). "High Cell-Density Culture of Escherichia Coli." Trends in
Biotechnology 14: 98-105.

Leikina E., LeDuc D. L., Macosko J. C., Epand R. F., Epand R. M., Shin Y. K. and
Chemomordik L. V. (2001). "The 1-127 HA2 Construct of Inﬂuenza Virus
Hemagglutinin Induces Cell-Cell Hemifusion." Biochemistry 40: 8378-8386.

Lian L. Y. and Middleton D. A. (2001). "Labelling Approaches for Protein Structural
Studies by Solution-State and Solid-State NMR." Progress in Nuclear Magnetic
Resonance Spectroscopy 39: 171-190.

Liden G. (2002). "Understanding the Bioreactor." Bioprocess and Biosystems
Engineering 24: 273-279.

Liu Y. C., Wang F. S. and Lee W. C. (2001). "On-line monitoring and controlling system
for fermentation processes." Biochemical Engineering Journal 7: 17-25.

Macosoko J. C., Kim C. H. and K S. Y. (1997). "The Membrane Topology of the Fusion
Peptide Region of Inﬂuenza Hemagglutinin Determined by
Spin-labeling EPR." Journal of Molecular Biology 267: 1139-1148.

Medina M. G., Carbonell X. and Villaverde A. (2002). "Connection Between Gene
Dosage and Protein Stability Revealed by a High-Yield Production of Recombinant

Proteins in an E. coli LexAl(Ind-) Background." Biotechnology and Bioengineering 78:
722-730.

Muchmore D. C., McIntosh L. P., Russell C. B., Anderson D. E. and Dahlquist F. W.
(1989). "Expression and Nitrogen-15 Labeling of Proteins for Proton and Nitrogen-15
Nuclear Magnetic Resonance." Methods in Enzymology 177: 44-73.

Murphy 0. J ., Kovacs F. A., Sicard E. L. and Thompson L. K. (2001). "Site-Directed
Solid-State NMR Measurement of a Li gand-Induced Conformational Change in the
Serine Bacterial Chemoreceptor." Biochemistry 40: 1358-1366.

Nor Z. M., Tamer M. 1., Scharer J. M., Moo-Young M. and Jervis E. J. (2001).
"Automated fed-batch culture of Kluyveromyces fragilis based on a novel method for on-

150

line estimation of cell speciﬁc growth rate." Biochemical Engineering Journal 9: 221-
231.

Notley L. and Ferenci T. (1996). "Induction of RpoS-Dependent Functions in Glucose—
Limited Continuous Culture: What Level of Nutrient Limitation Induces the Stationary
Phase of Escherichia coli?" Journal of Bacteriology 178: 1465-1468.

Novagen (2002). pET System Manual, Novagen.

Novy R. and Morris B. (2001). "Use of glucose to control basal expression in the pET
System." InNovations 13: 8-10.

Ochocka A. M., Czyzewska M. and Pawelczyk T. (2003). "Expression in Escherichia coli
of human ARHGAP6 gene and puriﬁcation of His-tagged recombinant protein." Acta
Biochimica Polonica 50: 239-247.

Oh G., Moo-Young M. and Chisti Y. (1998). "Automated fed-batch culture of
recombinant Saccharrnyces cerevisiae based on on-line monitored maximum substrate
uptake rate." Biochemical Engineering Journal 1: 211-217.

Pumphrey B. and Julien C. (1996). An Introduction to Fermentation: Fermentation
Basics, New Brunswick Scientiﬁc: 1-24.

Rani K. Y. and Rao V. S. R. (1999). "Control of ferrnenters- a review." Bioprocess
Engineering 21: 77-88.

Riesenberg D. and Guthke R. (1999). "High-Cell-Density Cultivation of
Microorganisms." Applied Microbiology and Biotechnology 52: 422-430.

Sambrook J. and Russel D. (2001). Molecular Cloning: A Laboratog Manual. Cold
Spring Harbor, New York, Cold Spring Harbor Laboratory Press.

Shiloach J ., Kaufman J., Guillard A. S. and Fass R. (1996). "Effect of Glucose Supply
Strategy on Acetate Accumulation, Growth, and Recombinant Protein Production by
Escherichia coli BL21 (hDE3) and Escherichia coli JM 109." Biotechnology and
Bioengineering 49: 421-428.

Shuler M. L. and Kargi F. (2002). Bioprocess Engineering: Basic Concepts. Upper
Saddle River, NJ, Prentice Hall.

Suzuki H., Kishimoto M., Kamoshita Y., Omasa T., Katakura Y. and Suga K.-i. (2000).
"On-line control of feeding of medium components to attain high cell density."

Bioprocess Engineering 22: 433-440.

Schweder T., Lin H. Y., Jurgen B., Breitenstein A., Riemschneider S., Khalameyze r. V.,
Gupta A., Buttner K. and Neubauer P. (2002). "Role of the general stress response during

151

 

strong overexpression of a heterologous gene in Escherichia coli." Applied Microbiology
and Biotechnology 58: 330-337.

Van de Walle M. and Shiloach J. (1998). "Proposed Mechanism of Acetate Accumulation
in Two Recombinant Escherichia coli Strains During High Density Fermentation."
Biotechnology and Bioengineering 57: 71-78.

Verdemato P. E., Brannigan J. A., Darnblon C., Fabio Z., Moody P. C. E. and Lian L. Y.
(2000). "DNA-binding mechanism of the Escherichia coli Ada O6-alkulguanin-DNA
alkyltransferase." Nucleic Acids Research 28: 3719-3718.

Wang J., Balazs Y. S. and Thompson L. K. (1997). "Solid-State REDOR NMR Distance
Measurements at the Ligand Site of a Bacterial Chemotaxis Membrane Receptor. "
Biochemistry 36: 1699-1703.

Wiley D. C. and Skehel J. J. (1987). "The Structure and Function of the Hemagglutinin
Membrane Glycoprotein of Inﬂuenza Virus." Annual Reviews in Biochemistry 56: 365-
394.

Wilson I. A., Skehel J. J. and Wiley D. C. (1981). "Structure of the Hemagglutinin
Membrane Glycoprotein of Inﬂuenza Virus at 3 A Resolution." Nature 289: 366-373.

Worden R. M., Kinney A. B. and Bender T. (2002). Exmriment Manual for the
Biochemical Engineering Laboratog, Michigan State University.

Yang J ., Parkanzky P. D., Bodner M. L., Duskin C. A. and Weliky D. P. (2002).
"Application of REDOR subtraction for ﬁltered MAS observation of labeled backbone

carbons of membrane-bound fusion peptides." Journal of Magnetic Resonance 87: 1951-
1963.

Yanisch-Perron C., Vieira J. and Messing J. (1985). "Improved M13 Phage Cloning
Vectors and Host Strains - Nucleotide-Sequences of the Ml3mpl8 and Pucl9 Vectors."
Gene 33(1): 103-119.

Zhang J. and Greasham R. (1999). "Chemically Deﬁned Media for Commercial
Fermentations." Applied Microbiology and Biotechnology 51: 407-421.

Zhu H., Pan 8., GU S. and Bradbury E. M. (2002). "Amino Acid Residue Speciﬁc Stable
Isotope Labeling for Quantitative Proteomics." Rapid Communications in Mass
Spectrometry 16: 2115-2123.

152

 

l91999999