Strain Engineering to Prevent Norleucine Incorporation During Recombinant Protein Production in Escherichia coli Karthik Veeravalli and Michael W. Laird Dept. of Late Stage Cell Culture, Genentech Inc., South San Francisco, CA 94080
Mark Fedesco Dept. of Purification Development, Genentech Inc., South San Francisco, CA 94080
Yu Zhang and X. Christopher Yu Dept. of Protein Analytical Chemistry, Genentech Inc., South San Francisco, CA 94080 DOI 10.1002/btpr.1999 Published online 00 Month 2014 in Wiley Online Library (wileyonlinelibrary.com)
Incorporation of norleucine in place of methionine residues during recombinant protein production in Escherichia coli is well known. Continuous feeding of methionine is commonly used in E. coli recombinant protein production processes to prevent norleucine incorporation. Although this strategy is effective in preventing norleucine incorporation, there are several disadvantages associated with continuous feeding. Continuous feeding increases the operational complexity and the overall cost of the fermentation process. In addition, the continuous feed leads to undesirable dilution of the fermentation medium possibly resulting in lower cell densities and recombinant protein yields. In this work, the genomes of three E. coli hosts were engineered by introducing chromosomal mutations that result in methionine overproduction in the cell. The recombinant protein purified from the fermentations using the methionine overproducing hosts had no norleucine incorporation. Furthermore, these studies demonstrated that the fermentations using one of the methionine overproducing hosts exhibited comparable fermentation performance as the control host in three different C 2014 American Institute of Chemical Engineers recombinant protein production processes. V Biotechnol. Prog., 000:000–000, 2014 Keywords: norleucine, sequence variant, methionine, incorporation, recombinant protein, fermentation
Introduction With the advancement in mass spectrometry-based methods to detect sequence variants at very low levels,1,2 amino acid incorporation in recombinant protein therapeutics produced in E. coli and other production systems have gained considerable attention in the last several years. Sequence variants, especially noncanonical amino acids such as norleucine or norvaline, in the final drug product is highly undesirable and may result in issues such as sensitivity to proteolysis, diminished biological activity, and possibly immunogenicity. In addition, these undesirable substitutions may require intensive analytical characterization, which could result in delays in product approvals by the regulatory health authorities. Hence, strategies should be implemented to prevent norleucine incorporation during recombinant protein production. Norleucine is a structural methionine analog and an isomer of branched chain amino acids, leucine, and isoleucine. Norleucine is synthesized at low levels in E. coli during production of leucine-containing recombinant proteins.3 Biosynthesis of norleucine under these conditions occurs from a-ketobutyrate via two passes through the leucine biosynthetic enzymes (LeuABCD) to Correspondence concerning this article should be addressed to K. Veeravalli at [email protected]. C 2014 American Institute of Chemical Engineers V
form a-ketocaproate, which is subsequently converted to norleucine by a transaminase.4 The production of leucine-containing recombinant proteins in E. coli results in a large demand for leucine during the fermentation process and as a result cells adapt to this higher demand by increasing the expression of leucine biosynthetic enzymes. High levels of leucine biosynthetic enzymes lead to increased levels of leucine as well as norleucine. Incorporation of norleucine at methionine positions during recombinant protein production in E. coli has been known for over 50 years.3,5,6 For example, about 14% of methionine residues in methionyl bovine somatotropin (MBS) exhibited norleucine incorporation during production of this protein in E. coli.3 Interestingly, in the same study, it was discovered that 6% of the methionine residues in native E. coli proteins were also substituted by norleucine. Production of a therapeutically relevant protein, Interleukin-2, in a minimal medium E. coli fermentation resulted in 19% of the methionine residues in the recombinant protein substituted with norleucine.5 Norleucine competes with methionine for incorporation into proteins due to the promiscuity of the methionyl tRNA synthetase (MetG).7,8 The norleucine residue that is incorporated at the N-terminal methionine residue of the protein can be formylated after the acylation reaction and the formylnorleucyl-tRNA can initiate protein synthesis.9 Norleucine incorporation into proteins can occur both at 1
2
Biotechnol. Prog., 2014, Vol. 00, No. 00
Table 1. Strains and Plasmids Used in This Study Strain or Plasmid W3110 JW3909-1 60E4 66G6 66H6 66H8 67B8 67B9 67C2 67C3 66F8 67C5 64B4 67C4 pS1080 pS1080-metA pS1080-metK pS1080-metA(R27C) pS1080-metA(Q64E) pS1080-metA(Y294C) pS1080-metA(I296SP298L) pS1080-metK(V185E) pS1080-metK(c1132del)
Genotype or Description
Ref. or Source
F2 k2 IN(rrnD-rrnE)1 rph-1 (DaraD-araB)567 DlacZ4787(::rrnB-3) k2 rph-1 D(rhaD-rhaB)568 DmetJ725::kanr hsdR514 W3110 DfhuA (DtonA) Dptr DompT DdegP DphoA ilvG2096 (IlvG1; Valr) Dfuc Drha 60E4 DmetJ725::kanr 60E4 metA(R27C) 60E4 metA(Y294C) 60E4 metA(I296SP298L) 60E4 metA(Q64E) 60E4 metA(Y294C) metK(V185E) 60E4 metA(Y294C) metK(c1132del) W3110 DfhuA (DtonA) DphoA ilvG2096 (IlvG1; Valr) DmanA Dprc spr43H1 DdegP lacIq DompT DmenE 66F8 metA(Y294C) W3110 DfhuA (DtonA) DphoA ilvG2096 (IlvG1; Valr) DmanA Dprc spr43H1 DdegP lacIq DompT 64B4 metA(Y294C) Counter-selectable allele-exchange suicide vector, R6Kc origin, SacB, Carbr MetA in pS1080 MetK in pS1080 MetA with R27C mutation cloned in pS1080 MetA with Q64E mutation cloned in pS1080 MetA with Y294C mutation cloned in pS1080 MetA with I296S and P298L mutations cloned in pS1080 MetK with V185E mutation cloned in pS1080 metK gene with cytosine deletion at position 1132 cloned in pS1080
Laboratory collection CGSC
internal residues as well as the amino terminus and the incorporation at the methionine loci in the protein is random.3 One method to reduce norleucine incorporation is to delete the genes involved in its biosynthetic pathway. Deleting the genes coding for leucine biosynthesis (leuA, leuB, leuC, and leuD) and/or the transaminases (ilvE or tyrB) have been successfully used to prevent norleucine incorporation.3,5 However, deleting the leucine biosynthetic pathway genes to prevent norleucine incorporation may require supplementation of high levels of leucine and other branched chain amino acids to the culture medium to support recombinant protein production. Other methods such as altering the protein amino acid sequence10 to remove the methionine codons and coexpressing enzymes to degrade norleucine11 have been successfully used to prevent norleucine incorporation. However, these methods may not be desirable for the production of human therapeutic drugs as they may potentially decrease the therapeutic activity and/or product yields during high cell density fermentations. A commonly used method to prevent norleucine incorporation in E. coli is to supplement the culture medium with methionine during the fermentation process. This nutrient supplementation ensures that there is excess methionine available to the cells thus reducing the probability of incorrect charging of the methionyl tRNA with norleucine. While continuous or bolus additions of methionine or other amino acids mitigates the risk of norleucine substitution, it increases the operational complexity and cost of the fermentation process. Moreover, the incorrect execution of the methionine feed due to delays in feed initiation, flow control valve issues, leaks in the transfer line, etc., during a manufacturing process could result in norleucine incorporation. The product obtained from fermentations with incorrect execution of methionine feed may require extensive analytical characterization, which could result in delays to product release or even lost product. In addition, continuous feeding or bolus additions during the fermentation process could lead to undesirable dilution of the fermentor contents resulting in lower cell densities and possibly lower recombinant protein yields. A novel approach, which has not
Laboratory collection This study This study This study This study This study This study This study Laboratory collection This study Laboratory collection This study Laboratory collection This study This study This study This study This study This study This study This study
yet been evaluated to our knowledge, to prevent norleucine incorporation during recombinant protein production in E. coli would be to utilize a methionyl tRNA synthetase from a different organism that cannot use norleucine as a substrate and incorporate it into E. coli. However, there is not much evidence on the specificity of methionyl tRNA synthetases from other organisms towards norleucine. Along these lines, engineering the E. coli methionyl tRNA synthetase to eliminate its activity towards norleucine in charging the methionyl tRNA could be a promising approach to prevent norleucine incorporation into recombinant proteins produced in E. coli. The existing methods to prevent norleucine incorporation described above exhibit several disadvantages. In this work, the advancements over the last several years in the wealth of genetic tools available were employed to engineer E. coli strains to prevent norleucine incorporation. Three E. coli production hosts were engineered by introducing chromosomal mutations in the genes involved in methionine biosynthesis and regulation (metA, metK, and metJ) that result in methionine overproduction in the cell. As expected, the fermentations using these mutant strains accumulated higher levels of intracellular and extracellular methionine as compared to their parent host fermentations. No norleucine incorporation was detected in the recombinant protein purified from the fermentations using the methionine overproducing hosts even when the fermentation was carried out in the absence of a methionine feed. Fermentations using the methionine overproducing hosts carrying the Y294C mutation in the MetA protein showed comparable fermentation performance as the metA1 host fermentations in three different recombinant protein production processes.
Materials and Methods Bacterial strains, plasmids, and growth conditions All strains in this study are derivatives of W3110.12 All strains and plasmids used in this study are listed in Table 1. Antibiotic selection was maintained for all markers at the
Biotechnol. Prog., 2014, Vol. 00, No. 00
3
Table 2. Oligonucleotides Used in This Study Primer Name SacI-metAflank-F SalI-metAflank-R SacI-metKflank-F SalI-metKflank-R metAflankmid-F metAflankmid-R metKflankmid-F metKflankmid-R pS1080-P pS1080-T QC-metAR27C-F QC-metAR27C-R QC-metAQ64E-F QC-metAQ64E-R QC-metAY294C-F QC-metAY294C-R QC-metAI296SP298L-F QC-metAI296SP298L-R QC-metKV185E-F QC-metKV185E-R QC-metKc1132del-F QC-metKc1132del-R
Sequence (50 -30 )* CACACGAGCTCCTCATTTTGCTCATTAACGTTGG
CACACGTCGACGCGAATGGAAGCTG CACACGAGCTCGTATGCAAAGCAGAGATGC CACACGTCGACCGTCATTGCCTTGTTTG GTTCTGATCCTTAACCTGATGCCGAAGAAG CCAGCGTTTGCGCATCATATTCGG GGCAAAACACCTTTTTACGTCCGAGTCC GAACTCACGTACCAGCAGGGTCAGTTG CCAGTCACGACGTTGTAAAACGACGG AGTGAACGGCAGGTATATGTGATGG GTGATGACAACTTCTtGTGCGTCTGGTCAGG CCTGACCAGACGCACaAGAAGTTGTCATCAC CAAACTCACCTTTGgAGGTCGATATTCAGC GCTGAATATCGACCTcCAAAGGTGAGTTTG GCTCAACTATTACGTCTgCCAGATCACGCCATACG CGTATGGCGTGATCTGGcAGACGTAATAGTTGAGC CGTCTACCAGAgCACGCtATACGATCTACG CGTAGATCGTATaGCGTGcTCTGGTAGACG ATCGATGCTGTCGaGCTTTCCACTCAG CTGAGTGGAAAGCtCGACAGCATCGAT GCGCAGCTGCTGGCGATGCTGCCG CGGCAGCATCGCCAGCAGCTGCGC
*Underlined residues introduce amino acid mutations. Lowercase residues are those that are different from the wild-type sequence.
following concentrations: Carbenicillin (plasmid or chromosomal), 50 lg/mL; Kanamycin (chromosomal), 30 lg/mL; Tetracycline (plasmid or chromosomal), 10 lg/mL. Strain and plasmid construction Oligonucleotides used in the construction of strains and plasmids are listed in Table 2. Standard techniques were used for cloning, DNA analysis, PCR, transformation, electroporation, and P1 transduction. Allele exchange was carried out using the methods described before.13–15 All allele replacements were confirmed by PCR and DNA sequencing. The metA gene was PCR amplified from the W3110 strain using primers SacI-metAflank-F and SalI-metAflank-R, digested with SacI and SalI and ligated into SacI and SalI digested pS1080 to obtain the plasmid pS1080-metAflank. Plasmids pS1080-metAflank(R27C), pS1080-metAflank (Q64E), pS1080-metAflank(Y294C), pS1080-metAflank (I296SP298L) were constructed by mutagenizing the plasmid pS1080-metAflank using the QuikChange kit (Stratagene) and the following set of primers (QC-metAR27C-F;QCmetAR27C-R), (QC-metAQ64E-F;QC-metAQ64E-R), (QCmetAY294C-F;QC-metAY294C-R), and (QC-metAI296SP298L-F;QC-metAI296SP298L-R), respectively. The metK gene was PCR amplified from the W3110 strain using primers SacI-metKflank-F and SalI-metKflank-R, digested with SacI and SalI and ligated into SacI and SalI digested pS1080 to obtain the plasmid pS1080-metKflank. Plasmids pS1080-metKflank(V185E) and pS1080-metKflank(c1132del) were constructed by mutagenizing the plasmid pS1080-metKflank using the QuikChange kit and the following set of primers (QC-metKV185E-F;QC-metKV185E-R), and (QCmetKc1132del-F;QC-metKc1132del-R), respectively. Fermentation Fermentations at the 10 L working volume were performed as described previously.16 Differences between the three antibody-fragment (AF) producing fermentation processes AF1, AF2, and AF3 described in this study are shown in Table 3. Fermentations were either performed with a con-
Table 3. AF1, AF2, and AF3 Fermentation Process Differences
Control host pH Agitation (rpm) Culture duration (hours) Feed (methionine or water) initiation (OD550)
AF1
AF2
AF3
60E4 7.0 850 72 200
66F8 7.0 650 50 150
64B4 7.3 650 72 150
tinuous methionine feed (stock concentration, 30 mg/mL), continuous water feed or no feed. To understand the effect of norleucine in the fermentation medium on sequence variant levels, a bolus of norleucine (final concentration, 0.15 mM) was added when the cells reached an OD550 of 200 during the fermentation. Metabolite analyses For analysis of intracellular methionine, fermentation broth samples containing 40 3 109 cells were pelleted at 17,000g for 5 min at 4 C and suspended in extraction buffer (10 mM Tris, 5 mM EDTA, 5 mM iodoacetamide (IAM), 0.2 mg/mL lysozyme, pH 6.8). Cells were then lysed by sonication and centrifuged for 20 min at 17,000g. The supernatants were transferred to 0.22 mm microcentrifuge tube filters (Bio Rad) and centrifuged at 17,000g for 5 min at 4 C. The filtrates were analyzed for methionine using Liquid Chromatography/ Mass Spectrometry (LC/MS). For analysis of extracellular methionine, supernatant samples obtained after pelleting the cells in the fermentation broth were analyzed using LC/MS. Phosphate levels were measured using the COBAS Integra 400 (Roche Diagnostics) using published methods.17 Recombinant protein yield measurements Fermentation broth samples were diluted as appropriate with extraction buffer (10 mM Tris, 5 mM EDTA, 5 mM IAM, 0.2 mg/mL lysozyme, pH 6.8), lysed by sonication and centrifuged at 17,000g for 20 min at 4 C. Soluble folded protein was determined from supernatants using reversedphase high performance liquid chromatography (HPLC). For total soluble protein measurements, samples were diluted 10-
4
Biotechnol. Prog., 2014, Vol. 00, No. 00
Table 4. Norleucine Quantification (in percent) Using LC-MS Analyses in the Recombinant Protein Purified from (a) AF1, (b) AF2, and (c) AF3 Processes Using Control and metA(Y294C) Hosts 60E4 1 Met Feed (a) Tryptic Peptides STAYLQMNSLR LSCAASGYDFTHYMGNWVR
60E4 metA(Y294C)
–*
1 nle bolus
60E4
–*
1 nle bolus
Mark Fedesco Dept. of Purification Development, Genentech Inc., South San Francisco, CA 94080
Yu Zhang and X. Christopher Yu Dept. of Protein Analytical Chemistry, Genentech Inc., South San Francisco, CA 94080 DOI 10.1002/btpr.1999 Published online 00 Month 2014 in Wiley Online Library (wileyonlinelibrary.com)
Incorporation of norleucine in place of methionine residues during recombinant protein production in Escherichia coli is well known. Continuous feeding of methionine is commonly used in E. coli recombinant protein production processes to prevent norleucine incorporation. Although this strategy is effective in preventing norleucine incorporation, there are several disadvantages associated with continuous feeding. Continuous feeding increases the operational complexity and the overall cost of the fermentation process. In addition, the continuous feed leads to undesirable dilution of the fermentation medium possibly resulting in lower cell densities and recombinant protein yields. In this work, the genomes of three E. coli hosts were engineered by introducing chromosomal mutations that result in methionine overproduction in the cell. The recombinant protein purified from the fermentations using the methionine overproducing hosts had no norleucine incorporation. Furthermore, these studies demonstrated that the fermentations using one of the methionine overproducing hosts exhibited comparable fermentation performance as the control host in three different C 2014 American Institute of Chemical Engineers recombinant protein production processes. V Biotechnol. Prog., 000:000–000, 2014 Keywords: norleucine, sequence variant, methionine, incorporation, recombinant protein, fermentation
Introduction With the advancement in mass spectrometry-based methods to detect sequence variants at very low levels,1,2 amino acid incorporation in recombinant protein therapeutics produced in E. coli and other production systems have gained considerable attention in the last several years. Sequence variants, especially noncanonical amino acids such as norleucine or norvaline, in the final drug product is highly undesirable and may result in issues such as sensitivity to proteolysis, diminished biological activity, and possibly immunogenicity. In addition, these undesirable substitutions may require intensive analytical characterization, which could result in delays in product approvals by the regulatory health authorities. Hence, strategies should be implemented to prevent norleucine incorporation during recombinant protein production. Norleucine is a structural methionine analog and an isomer of branched chain amino acids, leucine, and isoleucine. Norleucine is synthesized at low levels in E. coli during production of leucine-containing recombinant proteins.3 Biosynthesis of norleucine under these conditions occurs from a-ketobutyrate via two passes through the leucine biosynthetic enzymes (LeuABCD) to Correspondence concerning this article should be addressed to K. Veeravalli at [email protected]. C 2014 American Institute of Chemical Engineers V
form a-ketocaproate, which is subsequently converted to norleucine by a transaminase.4 The production of leucine-containing recombinant proteins in E. coli results in a large demand for leucine during the fermentation process and as a result cells adapt to this higher demand by increasing the expression of leucine biosynthetic enzymes. High levels of leucine biosynthetic enzymes lead to increased levels of leucine as well as norleucine. Incorporation of norleucine at methionine positions during recombinant protein production in E. coli has been known for over 50 years.3,5,6 For example, about 14% of methionine residues in methionyl bovine somatotropin (MBS) exhibited norleucine incorporation during production of this protein in E. coli.3 Interestingly, in the same study, it was discovered that 6% of the methionine residues in native E. coli proteins were also substituted by norleucine. Production of a therapeutically relevant protein, Interleukin-2, in a minimal medium E. coli fermentation resulted in 19% of the methionine residues in the recombinant protein substituted with norleucine.5 Norleucine competes with methionine for incorporation into proteins due to the promiscuity of the methionyl tRNA synthetase (MetG).7,8 The norleucine residue that is incorporated at the N-terminal methionine residue of the protein can be formylated after the acylation reaction and the formylnorleucyl-tRNA can initiate protein synthesis.9 Norleucine incorporation into proteins can occur both at 1
2
Biotechnol. Prog., 2014, Vol. 00, No. 00
Table 1. Strains and Plasmids Used in This Study Strain or Plasmid W3110 JW3909-1 60E4 66G6 66H6 66H8 67B8 67B9 67C2 67C3 66F8 67C5 64B4 67C4 pS1080 pS1080-metA pS1080-metK pS1080-metA(R27C) pS1080-metA(Q64E) pS1080-metA(Y294C) pS1080-metA(I296SP298L) pS1080-metK(V185E) pS1080-metK(c1132del)
Genotype or Description
Ref. or Source
F2 k2 IN(rrnD-rrnE)1 rph-1 (DaraD-araB)567 DlacZ4787(::rrnB-3) k2 rph-1 D(rhaD-rhaB)568 DmetJ725::kanr hsdR514 W3110 DfhuA (DtonA) Dptr DompT DdegP DphoA ilvG2096 (IlvG1; Valr) Dfuc Drha 60E4 DmetJ725::kanr 60E4 metA(R27C) 60E4 metA(Y294C) 60E4 metA(I296SP298L) 60E4 metA(Q64E) 60E4 metA(Y294C) metK(V185E) 60E4 metA(Y294C) metK(c1132del) W3110 DfhuA (DtonA) DphoA ilvG2096 (IlvG1; Valr) DmanA Dprc spr43H1 DdegP lacIq DompT DmenE 66F8 metA(Y294C) W3110 DfhuA (DtonA) DphoA ilvG2096 (IlvG1; Valr) DmanA Dprc spr43H1 DdegP lacIq DompT 64B4 metA(Y294C) Counter-selectable allele-exchange suicide vector, R6Kc origin, SacB, Carbr MetA in pS1080 MetK in pS1080 MetA with R27C mutation cloned in pS1080 MetA with Q64E mutation cloned in pS1080 MetA with Y294C mutation cloned in pS1080 MetA with I296S and P298L mutations cloned in pS1080 MetK with V185E mutation cloned in pS1080 metK gene with cytosine deletion at position 1132 cloned in pS1080
Laboratory collection CGSC
internal residues as well as the amino terminus and the incorporation at the methionine loci in the protein is random.3 One method to reduce norleucine incorporation is to delete the genes involved in its biosynthetic pathway. Deleting the genes coding for leucine biosynthesis (leuA, leuB, leuC, and leuD) and/or the transaminases (ilvE or tyrB) have been successfully used to prevent norleucine incorporation.3,5 However, deleting the leucine biosynthetic pathway genes to prevent norleucine incorporation may require supplementation of high levels of leucine and other branched chain amino acids to the culture medium to support recombinant protein production. Other methods such as altering the protein amino acid sequence10 to remove the methionine codons and coexpressing enzymes to degrade norleucine11 have been successfully used to prevent norleucine incorporation. However, these methods may not be desirable for the production of human therapeutic drugs as they may potentially decrease the therapeutic activity and/or product yields during high cell density fermentations. A commonly used method to prevent norleucine incorporation in E. coli is to supplement the culture medium with methionine during the fermentation process. This nutrient supplementation ensures that there is excess methionine available to the cells thus reducing the probability of incorrect charging of the methionyl tRNA with norleucine. While continuous or bolus additions of methionine or other amino acids mitigates the risk of norleucine substitution, it increases the operational complexity and cost of the fermentation process. Moreover, the incorrect execution of the methionine feed due to delays in feed initiation, flow control valve issues, leaks in the transfer line, etc., during a manufacturing process could result in norleucine incorporation. The product obtained from fermentations with incorrect execution of methionine feed may require extensive analytical characterization, which could result in delays to product release or even lost product. In addition, continuous feeding or bolus additions during the fermentation process could lead to undesirable dilution of the fermentor contents resulting in lower cell densities and possibly lower recombinant protein yields. A novel approach, which has not
Laboratory collection This study This study This study This study This study This study This study Laboratory collection This study Laboratory collection This study Laboratory collection This study This study This study This study This study This study This study This study
yet been evaluated to our knowledge, to prevent norleucine incorporation during recombinant protein production in E. coli would be to utilize a methionyl tRNA synthetase from a different organism that cannot use norleucine as a substrate and incorporate it into E. coli. However, there is not much evidence on the specificity of methionyl tRNA synthetases from other organisms towards norleucine. Along these lines, engineering the E. coli methionyl tRNA synthetase to eliminate its activity towards norleucine in charging the methionyl tRNA could be a promising approach to prevent norleucine incorporation into recombinant proteins produced in E. coli. The existing methods to prevent norleucine incorporation described above exhibit several disadvantages. In this work, the advancements over the last several years in the wealth of genetic tools available were employed to engineer E. coli strains to prevent norleucine incorporation. Three E. coli production hosts were engineered by introducing chromosomal mutations in the genes involved in methionine biosynthesis and regulation (metA, metK, and metJ) that result in methionine overproduction in the cell. As expected, the fermentations using these mutant strains accumulated higher levels of intracellular and extracellular methionine as compared to their parent host fermentations. No norleucine incorporation was detected in the recombinant protein purified from the fermentations using the methionine overproducing hosts even when the fermentation was carried out in the absence of a methionine feed. Fermentations using the methionine overproducing hosts carrying the Y294C mutation in the MetA protein showed comparable fermentation performance as the metA1 host fermentations in three different recombinant protein production processes.
Materials and Methods Bacterial strains, plasmids, and growth conditions All strains in this study are derivatives of W3110.12 All strains and plasmids used in this study are listed in Table 1. Antibiotic selection was maintained for all markers at the
Biotechnol. Prog., 2014, Vol. 00, No. 00
3
Table 2. Oligonucleotides Used in This Study Primer Name SacI-metAflank-F SalI-metAflank-R SacI-metKflank-F SalI-metKflank-R metAflankmid-F metAflankmid-R metKflankmid-F metKflankmid-R pS1080-P pS1080-T QC-metAR27C-F QC-metAR27C-R QC-metAQ64E-F QC-metAQ64E-R QC-metAY294C-F QC-metAY294C-R QC-metAI296SP298L-F QC-metAI296SP298L-R QC-metKV185E-F QC-metKV185E-R QC-metKc1132del-F QC-metKc1132del-R
Sequence (50 -30 )* CACACGAGCTCCTCATTTTGCTCATTAACGTTGG
CACACGTCGACGCGAATGGAAGCTG CACACGAGCTCGTATGCAAAGCAGAGATGC CACACGTCGACCGTCATTGCCTTGTTTG GTTCTGATCCTTAACCTGATGCCGAAGAAG CCAGCGTTTGCGCATCATATTCGG GGCAAAACACCTTTTTACGTCCGAGTCC GAACTCACGTACCAGCAGGGTCAGTTG CCAGTCACGACGTTGTAAAACGACGG AGTGAACGGCAGGTATATGTGATGG GTGATGACAACTTCTtGTGCGTCTGGTCAGG CCTGACCAGACGCACaAGAAGTTGTCATCAC CAAACTCACCTTTGgAGGTCGATATTCAGC GCTGAATATCGACCTcCAAAGGTGAGTTTG GCTCAACTATTACGTCTgCCAGATCACGCCATACG CGTATGGCGTGATCTGGcAGACGTAATAGTTGAGC CGTCTACCAGAgCACGCtATACGATCTACG CGTAGATCGTATaGCGTGcTCTGGTAGACG ATCGATGCTGTCGaGCTTTCCACTCAG CTGAGTGGAAAGCtCGACAGCATCGAT GCGCAGCTGCTGGCGATGCTGCCG CGGCAGCATCGCCAGCAGCTGCGC
*Underlined residues introduce amino acid mutations. Lowercase residues are those that are different from the wild-type sequence.
following concentrations: Carbenicillin (plasmid or chromosomal), 50 lg/mL; Kanamycin (chromosomal), 30 lg/mL; Tetracycline (plasmid or chromosomal), 10 lg/mL. Strain and plasmid construction Oligonucleotides used in the construction of strains and plasmids are listed in Table 2. Standard techniques were used for cloning, DNA analysis, PCR, transformation, electroporation, and P1 transduction. Allele exchange was carried out using the methods described before.13–15 All allele replacements were confirmed by PCR and DNA sequencing. The metA gene was PCR amplified from the W3110 strain using primers SacI-metAflank-F and SalI-metAflank-R, digested with SacI and SalI and ligated into SacI and SalI digested pS1080 to obtain the plasmid pS1080-metAflank. Plasmids pS1080-metAflank(R27C), pS1080-metAflank (Q64E), pS1080-metAflank(Y294C), pS1080-metAflank (I296SP298L) were constructed by mutagenizing the plasmid pS1080-metAflank using the QuikChange kit (Stratagene) and the following set of primers (QC-metAR27C-F;QCmetAR27C-R), (QC-metAQ64E-F;QC-metAQ64E-R), (QCmetAY294C-F;QC-metAY294C-R), and (QC-metAI296SP298L-F;QC-metAI296SP298L-R), respectively. The metK gene was PCR amplified from the W3110 strain using primers SacI-metKflank-F and SalI-metKflank-R, digested with SacI and SalI and ligated into SacI and SalI digested pS1080 to obtain the plasmid pS1080-metKflank. Plasmids pS1080-metKflank(V185E) and pS1080-metKflank(c1132del) were constructed by mutagenizing the plasmid pS1080-metKflank using the QuikChange kit and the following set of primers (QC-metKV185E-F;QC-metKV185E-R), and (QCmetKc1132del-F;QC-metKc1132del-R), respectively. Fermentation Fermentations at the 10 L working volume were performed as described previously.16 Differences between the three antibody-fragment (AF) producing fermentation processes AF1, AF2, and AF3 described in this study are shown in Table 3. Fermentations were either performed with a con-
Table 3. AF1, AF2, and AF3 Fermentation Process Differences
Control host pH Agitation (rpm) Culture duration (hours) Feed (methionine or water) initiation (OD550)
AF1
AF2
AF3
60E4 7.0 850 72 200
66F8 7.0 650 50 150
64B4 7.3 650 72 150
tinuous methionine feed (stock concentration, 30 mg/mL), continuous water feed or no feed. To understand the effect of norleucine in the fermentation medium on sequence variant levels, a bolus of norleucine (final concentration, 0.15 mM) was added when the cells reached an OD550 of 200 during the fermentation. Metabolite analyses For analysis of intracellular methionine, fermentation broth samples containing 40 3 109 cells were pelleted at 17,000g for 5 min at 4 C and suspended in extraction buffer (10 mM Tris, 5 mM EDTA, 5 mM iodoacetamide (IAM), 0.2 mg/mL lysozyme, pH 6.8). Cells were then lysed by sonication and centrifuged for 20 min at 17,000g. The supernatants were transferred to 0.22 mm microcentrifuge tube filters (Bio Rad) and centrifuged at 17,000g for 5 min at 4 C. The filtrates were analyzed for methionine using Liquid Chromatography/ Mass Spectrometry (LC/MS). For analysis of extracellular methionine, supernatant samples obtained after pelleting the cells in the fermentation broth were analyzed using LC/MS. Phosphate levels were measured using the COBAS Integra 400 (Roche Diagnostics) using published methods.17 Recombinant protein yield measurements Fermentation broth samples were diluted as appropriate with extraction buffer (10 mM Tris, 5 mM EDTA, 5 mM IAM, 0.2 mg/mL lysozyme, pH 6.8), lysed by sonication and centrifuged at 17,000g for 20 min at 4 C. Soluble folded protein was determined from supernatants using reversedphase high performance liquid chromatography (HPLC). For total soluble protein measurements, samples were diluted 10-
4
Biotechnol. Prog., 2014, Vol. 00, No. 00
Table 4. Norleucine Quantification (in percent) Using LC-MS Analyses in the Recombinant Protein Purified from (a) AF1, (b) AF2, and (c) AF3 Processes Using Control and metA(Y294C) Hosts 60E4 1 Met Feed (a) Tryptic Peptides STAYLQMNSLR LSCAASGYDFTHYMGNWVR
60E4 metA(Y294C)
–*
1 nle bolus
60E4
–*
1 nle bolus
