JOURNALOF FERMENTATION ANDBIOENGINEERING Vol. 85, No. 4, 447-450. 1998
The Importance YOSHIO Department
of Ionic Strength as a Parameter in Screening Peptide Ligands from a Phage Display Library
EUN TAE LIM, SETSUO TSUJII, TAKESHI OMASA, AND KEN-ICHI SUGA
Graduate School of Engineering, Osaka University, Suita, Osaka 565-0871, Japan Received 21 November 1997/Accepted 20 January 1998
Peptide ligands which bound to a model monomeric protein, bovine pancreatic ribonuclease A, could be isolated from a constrained random hexapeptide phage library. Selection was successful in a low ionic strength buffer (10 mlW sodium phosphate, pH 6.0), whereas it failed in TBS (50 mM T&-Cl, 150 mM NaCl, pH 7.5). Two of the displayed amino acid sequences from among the clones isolated were AEGACEQLDYNC and AEGACLWHDQLC. Electrostatic interaction appeared to play an important role in the binding because these phages could not bind to RNase A at a high ionic strength. The results suggest that selection in low ionic strength buffers could mahe possible the isolation of peptide ligands against proteins of interest which do not originally interact with another peptide or protein. [Key words: phage peptide library, biopanning,
Specific peptide ligands against proteins of interest are useful not only in fundamental research but also in applications such as separation, identification, detection, and drag development (1). Phage peptide libraries are a powerful tool for identifying the sequences of peptides that specifically recognize proteins of interest. To create a phage peptide library, a random oligonucleotide is inserted into gene 3 of Ml3 or fd phage. Each inserted gene is expressed at the N-terminal of the gene 3 product, a minor coat protein of the phage. As a result, peptide libraries that contain more than a billion kinds of peptide can be constructed. The phage library is then affinity screened against immobilized proteins of interest, and specifically bound phages are recovered and amplified by infection into Escherichia coli host cells. This procedure is called biopanning. Finally, amplified phages can be sequenced for deduction of the specific peptide sequences. Using phage libraries, peptide ligands have been obtained for proteins such as antibodies (2-4), peptide hormones and their receptors (2, 5, 6), proteases (7), the core protein of a virus (8), and a chaperon protein (9). However, all of these proteins originally interact with other peptides or proteins. To our knowledge, there has been no report on peptide ligands against monomeric soluble proteins that do not originally interact with another peptide or protein. With a few exceptions (10, ll), TBS (Tris-buffered saline) (2, 5-8) or PBS (phosphate-buffered saline) (3, 4, 9), which have an ionic strength of about 0.2M, have been used for biopanning. These high ionic strength buffers are thought to prevent electrostatic binding of the phage to the target protein in biopanning, because the binding of proteins to ion exchange resin is known to be fairly weak under the condition of a relatively high ionic strength. Accordingly, we used a low ionic strength buffer for the screening of phage clones. As a result,
peptide ligands against a model monomeric protein, bovine pancreatic ribonuclease A (RNase A) (12), could be successfully obtained. A phage library was constructed according to the method of Smith (2) except that the sequence (NNK)6 was replaced by TGC(NNK),TGC, where N is an equimolar mixture of all 4 nucleotides, and K is an equimolar mixture of G and T. A synthesized template-5’-CTAT TCTCACTCGGCCGACGGGGCTTGC(NNK)6TGCGG GGCCGCTGGGGCCGAAACTGTTGAA-was amplified using the 5’ biotinylated primers 5’-CTATTCTCACTC GGCCGACG and 5’-TTCAACAGTTTCGGCCCCAG. The PCR fragment was digested with BgfI which recognizes GCCNNNNNGGC. After removal of the biotinylated fragments by streptavidin-agarose, the Bgn fragment containing the degenerated sequence was ligated with SJiI-treated fUSE5. The ligated DNA was then introduced into Escherichia coli MC1061 by electroporation. The resultant library should contain 3 x lo7 independent phage clones. RNase A (Sigma, type XIIA) was labeled with biotin using a three molar excess of biotin-Nhydroxysuccinimide ester (Gibco BRL, Tokyo) according to the manufacture’s instruction and dialyzed against TBS (50 mM Tris-Cl, 150 mM NaCI, pH 7.5) containing 0.5 (w/v)% Tween 20. Biopanning was done following the method of Smith (2), using TBS or LIB (low ionic strength buffer: 10mM sodium phosphate, pH 6.0). When LIB was used as a buffer, it was employed instead of TBS throughout the protocol. Ten micrograms streptavidin in 1 ml 0.1 M NaHC03 was incubated at 4°C overnight in a 35-mm polystyrene dish (code 25060-60; Corning, New York, NY, USA). The streptavidin-coated dish was blocked for 2 h at 4°C with 0.5% BSA in TBS containing 0.1 ,ug/ ml streptavidin, and then washed 6 times with TBS containing 0.5% Tween 20 (TBS/T). In the first round of biopanning, 10,ug biotinylated RNase A in 400 ~1 TBS/T containing 0.1% BSA was added to the blocked dish. After incubation for 2 h at 4”C, 4 ~1 10 mM biotin was added and incubated for 1 h at 4°C to block vacant streptavidin in the dish. After washing 6 times with TBS/T, the phage library (1011-1012 transducing units) in 400,ul TBS/T was added, followed by the addition of
* Corresponding author. Abbreviations: TBS, Tris-buffered saline; PBS, phosphate-buffered saline; LIB, low ionic strength buffer: BSA. bovine serum albumin: ELISA, enzyme-linked immunosorbent assay; RNase A, bovine pancreatic ribonuclease A. 447
KATAIWRA ET AL. TABLE 1.
Summary of biopanning and ELISA LIB
Biotinylated RNase A
1 2 3
10 Pg 1 zg 0.1 ng
5 x 10” 3 x 10” 2x10”
5XlW 3XlW 41
5 x 10” 3 x 10” 2x10”
6x105 4x105 14
Positive control Negative control
Positive in Method A Positive in Method B
4 ~1 10mM biotin to eliminate phages that recognized biotin and incubation for 4 h at 4°C. After washing 6 times with TBS/T containing 0.1% BSA, bound phages were eluted with 400 ~1 0.1 M HCl @H adjusted to 2.2 with glycine). The host cell (E. coli K9lkan) was added to the neutralized eluent for infection and amplification of the phages. In the second and third rounds of biopanning, the phage library and biotinylated RNase A were premixed in TBS and the mixture was added to the streptavidin-coated dish to trap the phages that bound to the biotinylated RNase A. As shown in Table 1, 41 and 14 clones were obtained from the LIB and TBS systems, respectively. The isolated phage clones were assayed by two kinds of ELISA (Fig. 1). Method A followed that of Smith (2)-the phage was adsorbed onto a microtiter plate (code 655051; Grainer, Bad Hallerstr, Austria) and blocked with 5% non-fat dry milk; biotinylated RNase A bound to the phage was then detected by alkaline phosphatase conjugated streptavidin (Gibco BRL, Tokyo) using pnitrophenylphosphate as a substrate in a l-h reaction at room temperature. Method B was as follows. A microtiter plate (code 655061; Grainer) was coated with 100 pg/ml native RNase A (in 0.1 M sodium carbonate, pH 9.6) at 4°C overnight. The plate was then washed 3 times with LIB containing 0.1% (w/v) Tween 20 (LIB/T), and blocked with 5% non-fat dry milk containing 0.02% sodium azide at 37°C for 2 h. After washing 3 times with LIB/T, 100 ~1 phage in LIB prepared by the method of Smith (2) was added and incubated at 37°C for 1 h. After washing 3 times with LIB/T, phage bound to RNase A was detected by peroxidase conjugated anti-Ml3 antibody (Pharmacia, Uppsala, Sweden) using 2,2’-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) Method A
APase-streptavidin Biotinylated RNaseA
Native RNaseA FIG. 1. Concepts of the two kinds of ELISA. Method A, Biotinylated RNase A bound to immobilized phages was detected by alkaline phosphatase conjugated streptavidin; Method B, phages bound to native RNase A were detected by peroxidase conjugated anti-Ml3 phage antibody.
Results of ELISA Method A Biotinvlated
L5 L22 L35 L39
0.409 0.454 0.421 0.425
0.399 0.455 0.525 0.459
T2 T3 T6 TlO
0.436 0.484 0.366 0.308
0.020 0.021 0.012 0.017
The numbers of input and output phages are expressed as transducing units (2). Forty-one and fourteen clones from the LIB and TBS systems were assayed by two kinds of ELISA, Methods A and B, respectively. The number of positive clones obtained by each method is shown. The amount of biotinylated RNase A was decreased for each round to reduce the possibility that multivalent recognition phages would be selected (18).
at 410 nm (Method A) and 492 nm (Method B) are
diammonium salt as a substrate in a l-h reaction at room temperature. In both methods, anti-RNase A single chain Fv phage antibody (13) and wild-type fd-tet phage (2) were used as positive and negative controls, respectively. Thirty-three of the LIB clones and 10 from the TBS system tested positive in ELISA Method A. However, in Method B, all the 10 TBS clones were negative, whereas 31 of the 33 LIB clones that were positive in Method A were also positive in Method B. The results of ELISA for 4 typical clones from each system are shown in Table 2. These results indicate that the majority of the clones from the LIB system recognized both native and biotinylated RNase A, whereas most of those from the TBS system did not recognize native RNase A (see Fig. 1). The N-terminal amino acid sequences of the gene 3 products of L5 and L22 deduced from their DNA sequences were AIJGACNYDLQEC and a_GACLQDH WLC, respectively (acidic residues are underlined). The effect of ionic strength on the binding of L5 and L22 to native RNase A was investigated using ELISA Method B. As shown in Fig. 2, the absorbance of both clones was reduced as the ionic strength of the buffer was increased. These results together with the amino acid sequences deduced for the clones indicate that electrostatic interaction between acidic amino acid residues of displayed peptides and basic amino acid residues of RNase A (isoelectric point 9.6 (12)) plays an important role in the binding. A possible explanation why true clones such as LS could not be obtained using the TBS system is as follows. Phage clones from the TBS system tested positive in ELISA A but negative in ELISA B. This indicates that the phages recognized the biotin moiety of RNase A, because one RNase A molecule could be modified by more than two biotin molecules. Alternatively, alkaline phosphatase conjugated streptavidin could bind directly to the peptide displayed on the phages. In either case, the dominant force in the binding of clones from the TBS system to their target should be non-electrostatic interaction, because biotin has no charge and is a hydrophobic molecule, and also amino acid sequences displayed on phages which bind to streptavidin have been reported to have no charge (3, 14). In contrast, as mentioned above, the dominant force in the binding of clones from the LIB system to native RNase A appears to be electrostatic interaction. Non-electrostatic interactions, such as hydrophobic interaction or hydrogen bond-
VOL. 85. 1998
O’T-----l 04 0.3 I 0.2 1 0.1
0 i 0
0.3 0.2 Ionic strength (M)
FIG. 2. Effect of ionic strength on the binding of isolated phages to RNase A. The binding of L5 and L22 phages to native RNase A was assayed by ELISA Method B in the presence of various concentrations of NaCI.
ing, are scarcely affected by ionic strength, although electrostatic interactions are weakened by increasing ionic strength. As a result, in a low ionic strength buffer like LIB, the binding of clones such as LS to RNase A is stronger than that of clones such as T3 to their targets; however, in a high ionic strength buffer like TBS, the relative binding strength between clones from the LIB and TBS systems is reversed. In the case of RNase A, it should be possible to avoid false clones such as T3 by coating native RNase A directly onto the polystylene dish even when the TBS is used as a buffer. However, this method is not generally used because proteins have been reported to undergo denaturation when coated directly onto polystyrene plates (15). This is probably the main reason why proteins of interest are usually immobilized via streptavidin-biotin in biopanning. In the case of RNase A, the tertiary structure on a polystylene plate is thought to be as same as that of native one for the following reasons: (i) RNase A is a rigid and extremely stable protein (12); (ii) the monoclonal antibody 3A21, which recognizes native RNase A but not reduced RNase A (indicating that this antibody recognizes the tertiary structure of RNase A), recognizes RNase A coated on the polystyrene plate (13, 16). If proteins of interest retain their native structure on the plate, it is better to coat the protein directly onto the plate in order to eliminate false clones such as T3. Although the binding specificity of the LIB clones to RNase A was not confirmed, the binding is thought to be specific because (i) when phages from the LIB system were incubated previously with an excess amount of native RNase A and assayed by Method B, the absorbances were reduced to almost the same level as that of the negative control (data not shown) and (ii) if the binding was due to a non-specific electrostatic interaction between acidic residues of the peptides and a basic protein RNase A, phage clones displaying more acidic peptides, such as polyglutamic acid, should be isolated instead of these clones. Though the generality of this method needs to be further investigated, it is a fact that phage clones against native RNase A were obtained using the LIB system but not with the TBS system. Accordingly, our findings show that the ionic strength of the buffer is one of the important parameters affecting the results of biopan-
ning. Oldenberg et al. used a buffer of 20 mM 3-morpholinopropanesulfonic acid (PH 64, 50 mM NaCl, 2 mM MgCls, 2 mM CaC12, and 0.2 mM EDTA to isolate phages binding to a tetrameric sugar-binding protein, concanavalin A (10). Their success in isolating the phages might be due to the employment of this buffer, which has a lower ionic strength than TBS or PBS. The use of high ionic strength buffers such as TBS or PBS might be the main reason why, as mentioned earlier, there has been no report on peptide ligands against monomeric soluble proteins. Charged amino acid residues, such as Arg, Lys, His, Glu, and Asp residues have been reported to account for 29% of the surface area of monomeric proteins (17), indicating that electrostatic interaction between proteins and displayed peptides should be taken more seriously in the biopanning of peptide libraries. This work was supported in part by a Grant-in-Aid from the Japanese Ministry of Education, Science, Sports and Culture. REFERENCES 1. Lowman,
H.B.: Bacteriophage display and discovery of peptide ligands for drug development. Annu. Rev. Biomol. Struct., 26, 401-424 (1997). Smith, G. P. and Scott, J. K.: Libraries of peptides and proteins displayed on fllamentous phage. Methods in Enzymol., 217, 228-257 (1993). MaeLafferty, M. A., Kent, R. B., Landner, R. C., and Markland, W.: Ml3 bacteriophage displaying disulfide constrained microproteins. Gene, 128, 29-36 (1993). Roberts, D., Guegler, K., and Winter, J.: Antibody as a surrogate receptor in the screening of a phage display library. Gene, 128, 67-69 (1993). Pasqualini, R., Koivunen, E., and Ruoslahti, E.: A peptide isolated from phage display libraries is a structural and functional mimic of an RGD-binding site on integrins. J. Cell Biol., 130, 1189-1196 (1995). Koivunen, E., Wang, B., Dickinson, C. D., and Ruoslahti, E.: Peptides in cell adhesion research. Methods in Enzymol., 245, 346-369 (1994). Fang, R., Qi, J., Lu, Z., Zhou, H., Li, W., and Shen, J.: Selection of tripsin inhibitors in phage peptide library. Biochem. Biophys. Res. Commun., 220, 53-56 (1996). Dyson, M.R. and Murray, K.: Selection of peptide inhibitors of interactions involved in complex protein assemblies: association of the core and surface antigens of hepatitis virus. Proc. Nat]. Acad. Sci. USA. 92. 2194-2198 (1995). Blond Elguindi, S., ‘Cwtrla, S. E., Dower, W. J., Lipshutz, R. J., Sprang, S. R., Sambrook, J. F., and Gething, M. J.: Affinity panning of a library of peptides displayed on bacteriophages reveals the binding specificity of Bip. Cell, 75, 717728 (1993). Oldenburg, K. R., Loganathan, D., Goldstein, I. J., Schultz, P. G., and Gallop, M. A.: Peptide ligands for a sugar-binding protein isolated from a random peptide libraries. Proc. Natl. Acad. Sci. USA, 89, 5393-5397 (1992). Krook, M., Mosbach, K., and Lindbladh, C.: Selection of peptides with aflinity for single stranded DNA using a phage display library. Biochem. Biophys. Res. Commun., 204, 849-854 (1994). Richards, F. M. and Wyckoff, H. W.: Bovine pancreatic ribonuclease, p. 647-806. In Boyer, P. D. (ed.), The enzyme, vol. 4, 3rd ed. Academic Press, New York (1971). Katakura, Y., Kobayashi, E., Kurokawa, Y., Omasa, T., Fujiyama, K., and Suga, K.: cDNA cloning and characterization of anti-RNase A monoclonal antibody 3A21. J. Ferment. Bioeng., 82, 312-314 (1996). Kay, B. K., Adey, N. B., He, Y., Manfredi, J. P., Mataragnon, A. H., and Fowlkes, D. M.: An Ml3 phage library displaying
KATAKURA ET AL.
random 38-amino-acid peptides as a source of novel sequences with affinity to selected targets. Gene, 128, 59-65 (1993). 15. Frlgnet, B. L., Djavadt-Ohantance, L., and Goldberg, M. E.: Some monoclonal antibodies raised with a native protein bind preferentially to the denatured antigen. Mol. Immunol., 21, 673-677 (1984). 16. Katakura, Y., Kumarnoto, T., Iwal, Y., Kurokawa, Y., Omasa, T., and Suga, K.: Fluorescence polarization study of a salt bridge between a single-chain Fv and its antigen ribonuclease
J. FERMENT. BIOENCX, A. Mol. Immunol., 34, 731-734 (1997). 17. Miller, S., Janin, J., Lesk, A.M., and Chotbta, C.: Interior and surface of monomeric proteins. J. Mol. Biol., 196, 641656 (1987). 18. Barrett, R. W., Cwirla, S. E., Ackerman, M. S., Olson, A.M., Peters, E. A., and Dower, W. J.: Selective enrichment and characterization of high-alfinity ligands from collections of random peptides on tilamentous phage. Anal. Biochem., 204, 357364 (1992).