Antibody Fab display and selection through fusion to the pIX coat protein of filamentous phage

Antibody Fab display and selection through fusion to the pIX coat protein of filamentous phage

Journal of Immunological Methods 360 (2010) 39–46 Contents lists available at ScienceDirect Journal of Immunological Methods j o u r n a l h o m e p...

586KB Sizes 0 Downloads 86 Views

Journal of Immunological Methods 360 (2010) 39–46

Contents lists available at ScienceDirect

Journal of Immunological Methods j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j i m

Research paper

Antibody Fab display and selection through fusion to the pIX coat protein of filamentous phage Mark Tornetta a,⁎, Scott Baker b, Brian Whitaker a, Jin Lu a, Qiang Chen b, Eileen Pisors a, Lei Shi a, Jinquan Luo a, Raymond Sweet a, Ping Tsui a,c a b c

Centocor Research and Development Inc., Radnor, PA, United States Centocor Research and Development Inc., San Diego, CA, United States Department of Antibody Discovery and Protein Engineering, MedImmune Gaithersburg, MD, United States

a r t i c l e

i n f o

Article history: Received 7 January 2010 Received in revised form 24 May 2010 Accepted 2 June 2010 Available online 17 June 2010 Keywords: Phage display pIX Fab Antibody engineering Respiratory syncytia virus RSV

a b s t r a c t Fab antibody display on filamentous phage is widely applied to de novo antibody discovery and engineering. Here we describe a phagemid system for the efficient display and affinity selection of Fabs through linkage to the minor coat protein pIX. Display was successful by fusion of either Fd or Lc through a short linker to the amino terminus of pIX and co-expression of the counter Lc or Fd as a secreted, soluble fragment. Assembly of functional Fab was confirmed by demonstration of antigen-specific binding using antibodies of known specificity. Phage displaying a Fab specific for RSV-F protein with Fd linked to pIX showed efficient, antigen-specific enrichment when mixed with phage displaying a different specificity. The functionality of this system for antibody engineering was evaluated in an optimization study. A RSV-F protein specific antibody with an affinity of about 2 nM was randomized at 4 positions in light chain CDR1. Three rounds of selection with decreasing antigen concentration yielded Fabs with an affinity improvement up to 70-fold and showed a general correlation between enrichment frequency and affinity. We conclude that the pIX coat protein complements other display systems in filamentous phage as an efficient vehicle for low copy display and selection of Fab proteins. © 2010 Elsevier B.V. All rights reserved.

1. Introduction Filamentous phage display is widely used for protein engineering, most notably for antibody discovery and maturation (Hoogenboom, 2005). The external surface of the phage is dominated by the major coat protein, pVIII, that encapsulates the positive single-stranded DNA genome or surrogate phagemid. The termini of the core are capped by pairs of minor coat proteins, pIII and pVI at one end and pVII and pIX at the other (Clackson and Lowman, 2000). For antibody engineering, the pIII protein, or an N-terminal truncated variant, has been the preferred fusion protein, in part because its low copy number is suited for affinity selection (Garrard et al., 1991; Bradbury and

⁎ Corresponding author. Centocor Research and Development Inc., 145 King of Prussia Rd, Radnor, PA 19087, United States. Tel.: + 1 610 651 6158. E-mail address: [email protected] (M. Tornetta). 0022-1759/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.jim.2010.06.001

Marks, 2004). However, the pIII protein is critically involved in the attachment and entry process and proteins displayed on pIII may interfere with the efficiency of infection (Krebber et al., 1995; Marzari et al., 1997; Malmborg et al., 1997 and Spada and Plückthun, 1997). Counter to prior conclusions (Endemann and Model, 1995), the pIX protein was found to localize on the phage surface in an amino-terminal exposed orientation suitable for protein and peptide display (Gao et al., 1999). Like pIII, pIX is present at low copy number on the phage and it has been applied in a phagemid format for selection of specific antibodies from a scFv fusion library (Gao et al., 2002). While scFv are useful in their own right, the unnatural Vh–Vl configuration sometimes confounds conversion to a full antibody format, a process more readily accomplished with Fabs. Here we describe the efficient display of Fab proteins of known specificity through fusion of either the Fd or Lc to pIX coupled with soluble expression of the complementary chain, and demonstrate the utility of this system in affinity maturation from libraries of Fab variants.

40

M. Tornetta et al. / Journal of Immunological Methods 360 (2010) 39–46

Fig. 1. The pCNTO phagemid vector. (a) A vector map for the construct with Ch (CH1) fused to pIX. (b) Amino acid sequences showing the linkage of either Hc (Fd) or Lc (λ or κ) to pIX starting on the left with the C-terminal sequence of Ch or Cl. Stop codon is represented by an *.

2. Results 2.1. Fab display through the pIX protein The pCGMT9 phagemid (Gao et al., 1999) served as the backbone for development of a phagemid vector suitable for insertion of heavy (Vh) and light chain (Vl) variable region segments in a Fab format for display through fusion of the CH1 or CL fragments to pIX. Fig. 1 shows the features of the dual display and expression vector pCNTO-Fab-pIX, in the format for display of CH1 (Ch) linked to pIX and Lc (Vl–Cl) as a secreted fragment. For display through light chain, the pIX gene was fused to the C-terminus of the CL region through a short linker and Fd (Vh–CH1 region) was expressed as the soluble secreted fragment. Both vectors can be modified to secrete soluble Fab upon excision of the pIX gene via restriction digest with SpeI and NheI followed by self-ligation (see Materials and methods). For initial studies, we constructed a Fab derivative of T56, a scFv specific for TNFα. The Vh and Vl genes were placed into the pCNTO phagemid vector via conventional cloning to create the Lambda light chain (λ Lc) Fab. This Fab was used to evaluate display through linkage of Fd or Lc to the N-terminus of pIX. Phage particles for both constructs were generated by

transforming the pCNTO-T56 phagemid vector into TG1 cells and subsequent infection with helper phage. The recovered phage were evaluated by phage ELISA for surface display of the Fd or λ Lc via the pIX protein and for co-association of the corresponding soluble Lc or Fd fragment, respectively (Fig. 2a). As shown in Fig. 2b, serial dilutions of phage with the Fab linked through either Fd or Lc showed dose response binding to both anti-Fd or anti-Lambda antibody coated wells, indicating the display of both chains. The capture appeared to be specific because helper phage alone gave no signal in these assays. To further evaluate display via Fd or Lc, two different Fabs specific for the respiratory syncytial virus (RSV) F protein, Fab T40 (Tsui et al., 2002) and Fab B23 (Tsui, et al. manuscript in preparation), were cloned into the pCNTO vectors. Both Fabs have Kappa light chains (κ Lc). Phage ELISA results (Fig. 2c), show both T40 and B23 phage were captured by anti-Fd or anti-Kappa antibodies in either format of Fd or Lc fusion to pIX. These results further demonstrate that either Fd or Lc can be displayed by linkage to pIX and that the respective soluble Lc or Fd fragment coassociates with the displayed chain on the surface of the phagemid particle. Both the T40 and B23 Fab phage specifically bound to the RSV-F antigen in either Fd or Lc

Fig. 2. Fab display through pIX. a) Schematic showing the format of the Phage ELISA. The phage pictured on the left has Fd (Hc) linked to pIX and on the right has Lc linked to pIX. Display of the Fd or the Lc on the phage was measured by capture of the phage with a Fd or Lc specific antibody, respectively, coated on the ELISA well and detection with a HRP labeled M13 antibody. Display of functional Fab was detected by capture of the phage on plate wells coated with antigen followed by detection with the M13 antibody. b) Fd and Lc display for the T56 TNF Fab. The T56 Fab was displayed by linking either Fd (T56pCNTO Hc-pIX) or Lc (T56pCNTO Lc-pIX) to pIX and expressing the counter chain as a secreted, soluble fragment. Phage produced from both constructs were captured on the anti-Fd (Hc specific) or anti-Lambda (Lc specific) coated wells. The graph bars show the luminescence signal (RLU) for serial 1/5 dilutions of phage (left to right), starting with a neat concentration. Helper phage (“M13”) was used as a negative control. c) Display and antigen-specific binding for the T40 and B23 RSV Fabs. Both Fabs were displayed by linkage of Fd (“HcpIX” ) or Lc (“LcpIX”) to pIX. Serial dilutions of phage were captured on wells coated with anti-Fd antibody, anti-Kappa antibody, RSV-F protein, or TNF protein (negative control) and detected as in b. In b and c, serial dilution samples were measured in single wells.

M. Tornetta et al. / Journal of Immunological Methods 360 (2010) 39–46

41

42

M. Tornetta et al. / Journal of Immunological Methods 360 (2010) 39–46

display formats and did not bind to the irrelevant antigen, human TNFα (Fig. 2c). As an additional negative control, M13 helper phage, lacking any recombinant protein fused to its surface, showed no binding to the antibody or antigen coated wells (not shown). Correlative antigen selectivity was observed for the T56 Fab phage, although the binding to its specific antigen, TNFα, was much lower than observed with the RSV-F Fab phage, consistent with the rather low affinity of this antibody (not shown). We also evaluated Fab expression and purification using the pCNTO-Fab-pIX vector. The pIX gene was excised with NheI and SpeI restriction endonucleases and the vector was self-ligated to yield a B23 Fab expression cassette with the Fd segment linked to a hexa-histidine tag to enable purification. The Fab was purified to about 80% homogeneity following a single-step batch-procedure with IMAC resin and yielded ∼ 1 mg Fab from a 1 L induction. Similar yield and purity were achieved for many Fabs expressed using this vector and purification method (data not shown).

three rounds of panning, phagemid was prepared and the pIX gene was removed by digestion with NheI and SpeI restriction endonucleases and self-ligation. Individual colonies were grown in 96 well plates and spent media supernatants were screened for binding to RSV-F protein by ELISA. Unique positive clones were identified by DNA sequence analysis (Table 2). The dominant substitutions recovered after phage selection were Y31L and N32F. These clones were selected for production of purified Fab for further characterization. Purified Fabs containing the combination of Y31L and N32F had KDs ranging from 23 to 170 pM as measured by Biacore, a 9–68 fold improvement over the parental B23 Fab (Table 2). In contrast, I34 was tolerant to many substitutions with a bias towards charged and polar amino acids, while no substitutions were selected at Y36, indicating that Y was essential at this position. The efficacy of low copy display on pIX was further indicated by a general correlation between the affinity of the Fab and its frequency of occurrence in the selected population (Table 2).

2.2. Enrichment of antigen-specific displayed Fab

3. Discussion

To evaluate antigen enrichment of pIX-displayed Fab, T40 phagemid particles with Fd linked to pIX were diluted into a T56 scFv phage preparation at ratios of 1/100 and 1/10,000. This mixture was then selected for binding to RSV-F protein coated onto wells of an ELISA plate through one or two rounds of panning. Recovered phage were plated out and 96 colonies were screened by PCR for the T40 or T56 inserts that differed in size by 0.7 kb. As shown in Table 1, a 1/100 dilution of T40 was enriched 28-fold and 90-fold after one and two rounds of selection, respectively. The selective advantage was better revealed in panning with a 1/10,000 dilution in which the T40 phagemid was enriched 9000-fold after 2 rounds.

Previous reports demonstrated the ability to display Fv and scFv on the M13 surface via fusion to the minor coat protein pIX and to select antigen-specific scFv from a de novo library (Gao et al., 1999, 2002). The scFv format has some limitations for antibody discovery, including occasional instability and difficulty in conversion and production as full IgGs. To avoid these issues, we investigated pIX display of the Fab fragment as a better representation of full length antibodies. We reconfigured the pCGMT9 vector including revision of the linker sequence between the displayed Fab chain and the pIX protein. In this vector, pCNTO, the TNF Fab displayed well with either the Fd or Lc fused to pIX. For additional characterization of pIX Fab display, we studied two RSV Fabs, T40 and B23. Both showed efficient display with either Fd or Lc fused to pIX and expressed well as soluble Fabs. Phage displaying the T40 Fab with the Fd linked to pIX were shown to enrich many fold in panning against immobilized RSV-F protein in the presence of phage displaying an unrelated antibody. The efficient display and antigen-specific enrichment of Fabs linked through pIX encouraged us to apply this system to an affinity maturation library for Fab B23. Using Fd display through pIX, selection of a 106 random library across four residues in Vκ CDR1 identified Fabs with a 10–70 fold improvement in binding affinity. These results indicate that the efficiency of pIX display is comparable to the well established display through pIII. This and other successful applications to libraries for Fab affinity maturation (Tsui, et al., manuscript in preparation) provided a rationale for extension of this pIX methodology to de novo Fab libraries for isolation of antibodies to a variety of antigens (Shi et al., 2010). One potential advantage of display through pIX is that unlike pIII, it is not involved in the initial infection process. Phage bound to the target antigen can be recovered efficiently by direct addition of host bacteria which avoids potential bias against recovery of phage displaying Fabs with high affinity. The demonstration of Fab display now doubles the size of proteins reported to display through pIX to about 50 kDa.

2.3. Fab library display and selection with pIX Demonstration of display and efficient antigen-based enrichment of Fabs indicates that the pIX system could be applied to select variants from antibody libraries. Thus, Fab display using Fd linked to pIX was employed to search for variants of the B23 Fab with improved binding affinity to the RSV-F protein. The sequence of the B23 Fd and Lc are shown in Fig. 3. A combinatorial library randomizing residues Y31, N32, I34 and Y36 in CDR1 of the Lc (Tsui, et al., manuscript in preparation) was generated by mutating these four positions to all 20 amino acids using NNK codons. The library contained 1.1 × 106 possible DNA variants encoding 1.6 × 105 different amino acid sequences. A modified single-stranded mutagenesis method (Kunkel, 1985; Kunkel et al., 1987, 1991) subsequently developed for library applications (Sidhu et al., 2004; Fellouse et al., 2004) was used to construct the library, designated P1 (see Materials and methods). The transformation efficiency was 2 × 109 cfu/μg DNA. The total number of transformants was more than 6 × 109, well above the designed diversity of 1 × 106. Among the 189 colonies sequenced through the mutated area, 69% had unique mutations. Three rounds of panning were performed with the P-1 Fab-pIX phage library using biotinylated RSV-F protein followed by capture on strepavidin magnetic beads. After

M. Tornetta et al. / Journal of Immunological Methods 360 (2010) 39–46 Table 1 Antigen-specific enrichment of T40 Fab displayed as an Fd fusion on pIX. T40 Fab TG1

a

XL-1 b

Dilutions c

R1 d

R2 d

1/100 1/10,000 1/100 1/10,000

28% 0% 26% 0%

90% 90% 95% 80%

a

Phage produced in TG1 cells. Phage produced in XL-1 blue cells. Dilution ratio of T40 Fab phage to T56 scFv phage. d Percent of colonies with T40 Fab insert after selection rounds 1 and 2, as determined by a PCR screening assay. b c

Further studies are required to determine the practical size limitation for this system. Surface display of foreign proteins is described for fusion to the amino termini of the pIII, pVII, pVIII, and pIX coat proteins and the carboxy-terminus of the pVI protein of the M13/Fd filamentous phage (Gao et al., 2002; Smith and Petrenko, 1997; Kwaśnikowski et al., 2005). This opens the possibility of parallel selection of multiple proteins or binding sites on a single phage particle. In this regard, pIII and pIX are a well-matched pair because both occur at low copy number and they are positioned at opposite ends of the phage particle. 4. Materials and methods 4.1. Reagents T56, an antibody reactive to TNFα, was provided by Bin Zhou and Kim Janda (Scripps Research Institute) as a scFv construct in pCGMT9. RSV-F protein was provided by Jose Melero (Instituto de Salud Carlos III, Madrid, Spain). Human TNFα protein produced in Sp2/0 cells at Centocor. Other antibodies and detection reagents were: sheep anti-human (Fd) (The Binding Site Limited), goat anti-human(κLC) (Southern Biotechnology), goat anti-human(λLC) (Southern Biotechnology), goat anti-human(F(ab)’2) (Jackson Immunological), Bt-mouse-anti-HIS (R&D Systems), and mouse-antiM13-HRP (Amersham), Streptavidin-HRP (Zymed), and Chemiluminescence POD ELISA substrate (Roche Applied Sciences). Culture media (2xYT), buffers, and other general reagents were purchased from Teknova, Sigma, or Invitrogen. 4.2. Construction of pCNTO Phagemid vector, pCGMT9 provided by Scripps Research Institute (Gao et al., 2002), served as the backbone for the development of pCNTO (Fig. 1). The backbone includes the

43

colE1, f1, and the β-lactamase gene conferring resistance to ampicillin. Both oligo and gene synthesis procedures along with standard cloning procedures were used to construct a designed cassette for Fab molecules. The Fab-pIX fusion cassette contains the lac promoter for expression, Fd (VhCH1) fragment fused to the amino terminus of the pIX gene, and Lc sequences. The Fd-pIX cassette has a pelB signal sequence and the light chain cassette has an OmpA signal sequence. A lacI repressor gene is present in the phagemid to enhance the regulation of expression from the lac operon. To express soluble Fab protein, the pIX gene is excised from the display vector by SpeI and NheI restriction enzyme digestion, followed by the re-ligation of the digested vector fragment. The ligation brings a His-tag sequence to the C-terminus of Fd for soluble Fab purification. A Lc display phagemid was constructed in a similar manner to that of Fd whereby the Lc was fused through a linker to the N-terminus of pIX (Fig. 1b).

4.3. Phage-display of Fab The pCNTO Fab phagemids were transformed into TG1 cells (Stratagene). A single colony was used to inoculate 2xYT (2x yeast tryptone broth) containing 100 μg/mL carbenicillin (carb) plus 1% glucose and grown overnight with shaking at 37 °C. The overnight (O/N) culture was used to inoculate fresh 2xYT/carb at a ratio of 1:500 and incubated with shaking at 37 °C until the culture reached an OD600nm of 0.5– 0.6. The cultures were infected with helper phage (N1011 pfu/ mL) and incubated for 30 min at 37 °C without shaking. The infected culture was centrifuged and the cell pellet was resuspended in 2xYT/carb containing kanamycin (kan) at 50 μg/mL and 0.5 mM isopropyl-beta-D-thiogalactopyanocide (IPTG). This induced culture was incubated at 30 °C O/N with shaking. The phage cultures were centrifuged and the media supernatant containing the phage was collected in fresh tubes. PEG and NaCl (PEG/NaCl solution) were added to final concentrations of 2% and 0.25 M, respectively, and each sample was incubated on ice for 2 h with occasional mixing. The samples where then centrifuged and the supernatant discarded. The precipitated phage was re-suspended in 2 mL 1xPBS. The re-suspended phage was centrifuged again, placed into new vials, and used in experiments or stored at −80 °C. To assess the titer (spot titration), the phage sample was serially diluted at 1/10 intervals in 2xYT and the dilution series was mixed in plate wells with TG1 cells grown to an OD600nm of 0.5-0.8 and incubated for 30 min at 37 °C without shaking. A 2 μL sample from each infected well was spotted onto a dry LB/carb/1% glucose and LB/Kan agar plates and

Fig. 3. Amino acid sequences of B23 Fab Fd and Lc. Residues in κLc CDR1 that were randomized in the P1 library are underlined.

44

M. Tornetta et al. / Journal of Immunological Methods 360 (2010) 39–46

Table 2 Fabs selected from the P1 library. Fab Hits

B23 A7 H8 F8 A2 F5 H4 C11 B8 A11 a b c d e

Enrichment From round 3 a Parent 17/176 8/176 6/176 13/176 1/176 1/176 2/176 1/176 1/176

Sequence within

Activity

V κ CDR1 (NNK at each site)

b

Biacore c

Y31

N32

I34

Y36

Ka (×106 M−1 s−1)

Kd (×10−4 s−1)

KD (pM)

– L L L L V L – L –

– F F F F F F L F S

– R K Q V R T R S R

– – – – – – – – – –

0.96 2.42 4.19 1.66 2.5 e 1.4 2.02 1.4 e 1.24 1.39

6.0 0.87 0.95 0.86 2.2 e 2.4 0.87 2.0 e 1.17 2.34

667 d 36 23 52 87 170 43 138 94 168

(0.01) (0.01) (0.02) (0.03) (0.1) (0.04) (0.03) (0.01)

(0.3) (0.01) (0.02) (0.01) (0.2) (0.01) (0.01) (0.02)

Frequency of each Fab hit that was found after phage panning and single colony sequencing. Parental (top row) and variant amino acid sequences at the indicated positions in Vκ CDR1. A dash (-) indicates the same amino acid as in the parent. The number in ( ) represents the standard error in the last significant digit of the rate values from each experiment (Canziani et al., 2004). The KD for B23 from several experiments ranged from 667–1700 pM. Statistics were not obtained for these samples.

incubated O/N at 37 °C. The spots containing clearly separated single colonies were counted (less than 10 colonies per spot). 4.4. Phage ELISA For Fd detection on phage, Maxisorp plates (Nunc) were coated with 1 μg/mL of sheep anti-human (Fd) antibody (anti-Fd) in 1xPBS at 4 °C O/N. To determine if the Lc was associated with the Fd, Maxisorp plates were coated with 1 μg/mL of goat anti-human (κLC) antibody (anti-Kappa) or goat anti-human (λLC) antibody (anti-Lambda) in 1xPBS at 4 °C O/N. To determine if the displayed Fabs bound to their relevant Ag, Maxisorp plates were coated with either 0.1 μg/ mL of RSV-F protein (5) or 0.5 μg/mL TNFα protein in 1xPBS at 4 °C O/N. All of the coated plates were washed 3 times with 1xTBST. The plates were then blocked with 5% milk/1xTBS for 1 h at RT. The plates were washed 1 time with 1xTBST. Varying amounts of cloned phage were serial diluted in 0.5% milk/1xTBST and incubated at RT for 1 h. The plates were washed 5 times with 1xTBST. To detect the bound Fab phage, HRP labeled mouse anti-M13 antibody (anti-M13-HRP) was added at a 1/5000 dilution in 0.5% milk/1xTBST and incubated for 1 h at RT. The plates were washed 5 times with 1xTBST. Chemiluminescence POD ELISA substrate was added and the plates were read immediately on a Tecan FluorMax plate reader using the luminescence optimal gain settings. 4.5. Antigen-specific phage enrichment A well of a Maxisorp plate was coated with 300 μL of 0.1 μg/mL of RSV-F protein in 1xPBS at 4 °C O/N. The solution originally containing the antigen was removed and rinsed once with 1xPBS. The whole well was then blocked with 5% milk/1xPBS for 1 h at RT. After removal of the blocking solution, a dilution ratio of 1/100 and 1/10,000 of T40 phage (with Fd linked to pIX) to T56 scFv phage in 0.5%milk/1xPBST was added to the well and incubated at RT for 1 h. The unbound phage solution was removed and 10 successive washes were applied using 1xPBST. After the washes a rinse with 1xPBS was performed. The bound phage was then eluted with 300ul of TG1 cells at a density of OD600nm of 0.5–0.8 and

incubated at 37 °C. The elution was repeated one more time and the two aliquots of infected cells were combined and plated on 150 mm agar plates containing LB/carb (100 μg/mL). A second panning round was performed as above. Single colonies from the first and second rounds were isolated and analyzed in a PCR screening assay with Taq polymerase and primer oligomers that positioned at either end of the cassette, between EcoRI and NotI (Fig. 1a). Purified phagemid DNA containing T56 scFv and the T40 Fab were used as positive controls. 4.6. Library P1 A phagemid template was created that contained stop codons in the region of the antibody to be mutated so that any template contamination in the library could not produce parental Fab and decrease the efficiency of the downstream selection process. The primers used to make the stop template for P-1 library were (plus strand)5′ GCG TCT CAG TCT GTT GAC TAA TAA GGT TAA TCT TAA ATG CAC TGG TAC CAG CAG and (minus strand) 5′ CTG CTG GTA CCA GTG CAT TTA AGA TTA ACC TTA TTA GTC AAC AGA CTG AGA CGC. The underlined bases indicate stop codons. The reaction was carried out with pCNTO-B23-pIX using a QuickChange kit (Stratagene, CA). To generate random mutation of the 4 amino acid residues within the variable region of the Vκ of the B23 Fab, an oligonucleotide was synthesized that had a 12 base pair (bp) degenerate NNK codon at each amino acid position. This oligonucleotide was flanked by two 18 bp nucleotide sequences identical to the regions preceding and following the region to be mutagenized. These primers were phosphorylated at the 5′ end. The sequences of the primers used for the P-1 library are: (Lc CDR1 plus NNK) 5′-Phos-GCG TCT CAG TCT GTT GAC NNK NNK GGT NNK TCT NNK ATG CAC TGG TAC CAG CAG and (Lc CDR1 plus VNN) 5′ phos-GCG TCT CAG TCT GTT GAC NNK VNN GGT VNN TCT NNK ATG CAC TGG TAC CAG CAG-3. To purify a single-stranded DNA template, a single colony of E. coli CJ236 harboring the template phagemid pCNTO-B23pIX with a stop codon was picked and put into 5 mL of 2xYT/ carb and chloramphenicol (10 μg/mL). The culture was

M. Tornetta et al. / Journal of Immunological Methods 360 (2010) 39–46

shaken at 200 rpm at 37 °C for about 6 h. VCSM13 helper phage was added to a final concentration of 1010 pfu/mL without shaking, for 10 min. The culture was then transferred to 150 mL of 2xYT with carb (10 μg/mL) and uridine (0.25 μg/ mL) and incubated at 37 °C with shaking O/N. The culture was centrifuged at 2 °C. The supernatant was poured into a fresh tube, followed by addition of PEG/NaCl solution and incubated for 5 min at room temperature. The solution was then centrifuged at 2 °C. The phage pellet was re-suspended in 2 mL of PBS and re-centrifuged for 5 min at 2 °C. Singlestranded DNA was purified using a Qiagen QIAprep Spin M13 kit according to the manufacturer's instructions. The dUssDNA was quantified by UV absorbance. The yield was approximately 10 μg. To anneal the degenerate oligonucleotide to the template, dU-ssDNA (8 μg) template was combined with phosphorylated oligonucleotide at a molar ratio of 1:10 (template:oligo) in a buffer containing Tris–HCl (50 mM, pH 7.5) and MgCl2 (10 mM). The 250 μL reaction was incubated at 90 °C for 2 min, 50 °C for 3 min, and 20 °C for 5 min. After the annealing reaction, 10 μL ATP (10 mM), 10 μL dNTPs (25 mM each), 15 μL DTT (100 mM), T4 ligase (30 units), and T7 DNA polymerase (30 units) were added and the reaction mixture was incubated at room temperature for 4 h. The resulting DNA was purified, desalted, and dissolved in 35 μL of water. The dsDNA product was electroporated into E. coli TG-1 cells and the transformants were allowed to grow O/N on agar plates containing ampicillin (100 μg/mL) and glucose (1%). The colonies were scraped off the plates in 2xYT/15%glycerol medium and stored at −80 °C. Based on the colony count, the library contained 6×109 members. To construct the phage library, a 50 μL aliquot of the FabpIX phagemid library glycerol stock was inoculated into 25 mL 2xYT/carb (100 μg/mL) and incubated at 37 °C with shaking at 250 rpm until the culture reached an OD600nm of 0.5–0.6. Cells were infected with 1 mL of VCSM13 helper phage (N1011 cfu final concentration) and incubated for 45 min at 37 °C without shaking. Subsequently, the cells were centrifuged and re-suspended in 25 mL of 2xYT/carb/ kan (50 μg/mL)/0.5 mM IPTG and incubated at 30 °C with shaking at 250 rpm for 12–16 h. The phage library was harvested by precipitation using PEG/NaCl solution. The titer of the library was measured by spot titration. The library was aliquoted and stored at −80 °C. 4.7. Library selection and screening Three rounds of panning were performed with the P-1 FabpIX phage library. Briefly, a 100 μL volume of phage library was blocked with 100 μL ChemiBlocker at room temperature for 1 h in a pre-blocked 1.5 mL micro tube. Biotinylated RSV-F protein (10 nM) was added and the mixture was incubated at room temperature for 1 h with occasional mixing, 10 μL of streptavidin coated magnetic beads (Dynabeads M280) were added and the mixture was incubated with occasional stirring for 30 min at room temperature. A magnetic separator (Dynal) was used for separating beads from the supernatants between each of the following steps: 10 washes with incremental volumes of 1xTBST from 1 mL to 10 mL, a 10 mL rinse with 1xTBS and a final rinse with 1 mL of 1xTBS. After the washing, bound phage were recovered by addition of 1 mL of freshly grown TG1 cells (OD600nm 0.5–0.8) and incubation at 37 °C for 40 min without

45

shaking. Then the cells were plated and grown O/N. The next day, the colonies on the plates were scraped using 2 mL of 2xYT/carb/ 20% glycerol for each plate, and the sample was frozen at −80 °C. An aliquot was used to initiate the culture for making a fresh phage library from these recovered colonies. The new phage library was used for additional rounds of panning as described above except the antigen concentration was decreased to 1 nM in round 2 and 0.1 nM in round 3. After three rounds of panning, phagemid was prepared and the pIX gene was removed by digestion with NheI and SpeI restriction endonucleases. The linearized vector was purified by gel electrophoresis, extracted, self-ligated, and transformed into TG1 cells. Individual colonies from overnight growth on LB/carb agar plates were picked into 96 deep well plates containing 0.5 mL of 2xYT/carb per well. The plate was incubated at 37 °C shaking until the culture reached an OD600nm of 0.7–1.0. IPTG was added to 0.5 mM and the plate was incubated at 30 °C with shaking O/N. The cultures were centrifuged and spent culture media was used in Fab ELISA to screen for hits that gave binding activities to RSV-F protein better than the parent Fab, B23. Fabs that showed binding to RSV-F protein were then sequenced by using primers that hybridized upstream of either the Fd or Lc. 4.8. Fab production and affinity characterization A single colony was inoculated into 2xYT/carb/1% glucose and grown O/N with agitation at 37 °C. The culture was used to inoculate fresh 2xYT/carb/0.1% glucose at a ratio of 1:100. This culture was grown to an OD600nm of 1.0. Fab expression was induced by addition of IPTG to a final concentration of 0.5 mM, and the sample shaken O/N at 30 °C. The cells were harvested from the expression culture (1 L) by centrifugation. The cell pellet was re-suspended in 100 mL of 1xPBS/350 mM NaCl/7.5 mM imidazole and complete protease inhibitor without EDTA (Roche) and lysed with a microfluidizer (Microfluidics) by running it through 3 times. The lysate was centrifuged 2 separate times. Talon® resin (Clonetech), equilibrated with 1xPBS, was added to the supernatant and mixed gently for 2 h. The Fab-bound resin was collected by centrifugation and loaded onto a column. The resin was washed 2 times with 1xPBS/350 mM NaCl/7.5 mM imidazole and eluted with 150 mM EDTA/1xPBS. The eluate was dialyzed overnight against 1xPBS. The sample was concentrated by filter centrifugation (Centiprep-20, Amicon). Concentration was determined by OD280nm and purity was determined by SDS-PAGE and staining with Coommassie Blue. Briefly, the Biacore experiments were formatted to be able to rank each Fab candidate against each other and the parent molecule, B23. The antigen, RSV-F protein, was coupled to the sensor directly at a density of 80RU. Each purified Fab at 90 nM was passed over the RSV-F protein. All experiments were performed on a Biacore 3000 and analyzed as described by Canziani et al., 2004. Acknowledgements The pIX display vector, pCGMT9, and a derivative construct for display of the T56 scFv, specific for human TNFα, were kindly provided by Bin Zhou and Kim Janda at The Scripps Research Institute (TSRI). Bin Zhou (TSRI) provided extensive

46

M. Tornetta et al. / Journal of Immunological Methods 360 (2010) 39–46

experimental assistance in the transfer of the pIX technology to Centocor. At Centocor the Biacore experiments were carried out by Gabriela Canziani. Ben Amegadzie and Karen O'Neil provided advice during the project and John Wheeler reviewed the manuscript. References Bradbury, A.R.M., Marks, J.D., 2004. J. Immunol. Methods 290, 29. Canziani, G.A., Klakamp, S., Myszka, D.G., 2004. Anal. Biochem. 325 (2), 301. Clackson and Lowman “Phage Display” Oxford, August 2000. Endemann, H., Model, P., 1995. J. Mol. Biol. 250, 496. Fellouse, F., Wiesmann, C., Sidhu, S., 2004. Proc. Natl. Acad. Sci. U. S. A. 101, 12,467. Gao, C., Mao, S., Lo, C.-H.L., Wirsching, P., Lerner, R.A., Janda, K.D., 1999. Proc. Natl. Acad. Sci. U. S. A. 96, 6025. Gao, C., Mao, S., Kaufmann, G., Wirsching, P., Lerner, R.A., Janda, K.D., 2002. Proc. Natl. Acad. Sci. U. S. A. 99 (20), 12,612. Garrard, L.J., Yang, M., O'Connell, M.P., Kelley, R.F., Henner, D.J., 1991. Nat. Biotechnol. 9, 1373.

Hoogenboom, H.R., 2005. Nat. Biotechnol. 23, 1105. Krebber, C., Spada, S., Desplancq, D., Plückthun, A., 1995. FEBS Lett. 377, 227. Kunkel, T.A., 1985. Proc. Natl Acad. Sci. U. S. A. 82, 488. Kunkel, T.A., Roberts, J., Zakoour, R., 1987. Methods Enzymol. 154, 367. Kunkel, T.A., Bebenek, K., McClary, J., 1991. Methods Enzymol. 204, 125. Kwaśnikowski, P., Kristensen, P., Markiewicz, W., 2005. J. Immunol. Methods 307 (1–2), 135. Malmborg, A.C., Söderlind, E., Frost, L., Borrebaeck, C.A., 1997. J. Mol. Biol. 222, 544. Marzari, R., Sblattero, D., Righi, M., Bradbury, A., 1997. Gene 185, 27. Shi, L., Wheeler, J.C., Sweet, R.W., Lu, J., Luo, J., Tornetta, M., Whitaker, B., Reddy, R., Brittingham, R., Borozdina, L., Chen, Q., Amegadzie, B., Knight, D.M., Almagro, J.C., Tsui, P., 2010. J. Mol. Biol. U. S. A. 327 (2), 385. Sidhu, S.S., Li, B., Chen, Y., Fellouse, F.A., Eigenbrot, C., Fuh, G., 2004. J. Mol. Biol. 338, 299. Smith, G., Petrenko, V., 1997. Chem. Rev. 97 (2), 391. Spada, S., Plückthun, A., 1997. Nat. Med. 6, 694. Tsui, P., Tornetta, M.A., Ames, R.S., Silverman, C., Porter, T., Weston, C., Griego, S., Sweet, R.W., 2002. J. Immunol. Methods 263, 123.