Synthesis of a One-Bead One-Compound
Combinatorial Peptide Library
Kit S. Lam and Michal Lebl
1. Introduction
The four general methods to generate and screen a huge combinatorial pep-
tide library +-lo7 peptides) are: biological libraries such as filamentous phage
(I), plasmid (2)) or polysome (3) libraries; the “one-bead one-compound” syn-
thetic combmatonal library method or the “Selectlde process” (4-6); synthetic
peptide library methods that require deconvolution, such as an iterative
approach (7,8), positional scanning (9); orthogonal partition approach (JO), or
recurse deconvolution (II); and synthetic library using affinity column
selection method (12,13).
There are advantages and disadvantages m each of these methods. In gen-
eral, the main advantages of the biological library method are that large pep-
tides can be displayed on a filamentous phage library, and that large protein
folds can be mcorporated into the library. However, the main disadvantage is
that biological libraries, in general, are restricted to all L-amino acids. In con-
trast, the remaining three methods all use synthetic libraries; therefore,
o-amino acids, unnatural ammo acids, nonpeptide components, and small rigid
scaffoldings can all be incorporated into these libraries.
The “one-bead one-compound” library is based on the concept (4,5) that
when a solid-phase split synthesis method (4,8,14) is used, each solid-phase
particle (bead) displays only one peptide entity although there are approx 1013
copies of the same peptide in the same bead. The resulting peptide-bead library
(e.g., lo7 beads) is then screened in parallel using either “on-bead” binding
assays (15) or “solution phase-releasable” assays (16) to identify peptide-beads
with the desired biologic, biochemical, chemical, or physical properties. The
From Methods m Molecular Biology, vol 87 Combmatonal Peptrde Library Protocols
Edlted by S CablIly 0 Humana Press Inc , Totowa, NJ
2
Lam and Lebl

positive peptide-beads are then physically isolated for microsequencing with
an automatic protein sequencer. In this chapter, detailed methods for the syn-
thesis of a random “one-bead one-compound” combinatorial peptide library
will be described. Chapters 2 and 10 give examples of two general screening
methods for such libraries.
2. Materials
2.1. Chemicals
1. Tenta-Gel Resin S-NH, (90-100 pm) resin may be obtained from Rapp Polymere,
Tubmgen, Germany (see Note 1).
2. Fmoc amino acids with standard side chain-protectmg groups, N-hydroxy-
benzotriazole (HOBt), benzotriazolyl-oxy-trisdimethylammo-phosphonmm
hexafluorophosphate (BOP), diisopropylethylamme (DIEA), diisopropyl-
carbodumide (DIC), piperidme, trifluoroacetic acid (TFA), nmhydrm, may be
obtained from many different suppliers, such as Bachem (Torrance, CA), Bio-
science (King of Prussia, PA), Advanced ChemTech (Louisville, KY),
Novabiochem (San Diego, CA), and Peptides International (Louisville, KY)
3. Technical grade solvents such as dimethylformamide (DMF) or dichloromethane
(DCM) may be obtained from many different chemical suppliers HPLC-grade
DMF for the coupling may be obtamed from Burdock and Jackson, Muskegon,
MI. Ethanol, phenol, p-cresole, thioamsole, ethanedithiol, pyndme, and potas-
sium cyanide may be obtained from many different chemical suppliers.
4 0 1 g/mL Nmhydrm in ethanol
5 4 g/mL Phenol m ethanol.
6 10 mM Potassmm cyanide, stock solution.
7. 50% Piperidme m DMF
8. Reagent K: TFAlp-cresolelwaterlthioamsole/ethanedithiol, 82 5*5:5:5*2.5. (v/v/
v/v/v)
9 10% DIEA m DMF.
10 Dimethylsulfoxide (DMSO)/Amsole/TFA, 10:5:85
2.2. Apparatus

1. Polypropylene vials (5-lo-mL) may be purchased from Baxter Scientific Prod-
ucts, McGaw Park, IL. Polyethylene disposable transfer pipets may be purchased
from Elkay Products, Shrewsbury, MA.
2 Motorized rockmg platform.
3 Randomization glass vessel (chromatography column 5-6
x
18 cm) fitted with a
medmm glass smtered frit connected to vacuum and nitrogen via a two-way valve
from below The three positions of the valve are “off,” “vacuum,” or “nitrogen.”
4 Recnculatmg water aspirator or a solvent-resistant vacuum pump with cold trap
5 Nitrogen tank.
One-Bead One-Compound
3. Methods
3.1. Synthesis of a Linear Pentapeptide Library
As indicated earlier, a solid-phase split synthesis method (4,8,14) is used to
generate a random peptide library. The composition and final structure of the
peptide library depends on the number of amino acids (one or more) used m
each coupling cycle and the number of coupling cycles used. The final peptide
library may be linear or cyclic, or have specific secondary structures. For sim-
plicity, the method for the synthesis of a linear pentapeptide library with all 19
eukaryotic amino acids except cysteine is given below:
1. Swell 10 g TentaGel Resin S-NH, beads (- 0 25 mEq/g, see Notes 1 and 2) for at
least 2 h m HPLC-grade DMF with gentle shaking in a silicomzed flask.
2 Wash the beads twice with HPLC-grade DMF in the slllcomzed randomlzatlon
vessel as follows* add 75 mL DMF from the top, gently bubble nitrogen from
below through the smtered glass for 2 min, then remove the DMF by vacuum
from below (see Note 3).
3 Transfer all the beads to a slllcomzed flask in HPLC-grade DMF Then dlstrlbute
the beads into 19 equal allquots. A disposable polyethylene transfer plpet IS
extremely useful m the even distribution of the beads mto each polypropylene

vial (see Note 4).
4 Allow the beads to settle and remove most of the DMF above the settled bead
surface from each polypropylene reaction vial.
5 Add threefold molar excess of each of the 19 Fmoc-protected ammo acids (see
Note 5) and threefold molar excess of HOBt to each reaction vial using a mml-
ma1 volume of HPLC-grade DMF.
6. Add threefold molar excess each of BOP and DIEA to each reaction vial to ml-
tlate the coupling reaction.
7. Cap the reaction vials tightly and rock them gently for 1 h at room temperature
8. To confirm the completion of couplmg reaction, plpet a minute amount of resin
from each reaction vial into small borosilicate glass tubes (6
x
50-mm) and per-
form ninhydrm test (17) as follows:
Wash the minute quantity of resin m the small glass tubes (6 x 50-mm)
sequentially with the following solvents* DMF, t-amyl alcohol
(2-methylbutan-2-ol), acetic acid, t-amyl alcohol, DMF, and ether Add to
each tube one drop of each of the following three reagents, (ninhydrin m etha-
nol (0.1 g/mL), phenol m ethanol (4 g/mL), and potassium cyanide stock
solution diluted 50 times with pyridme. Place the tubes m a heating block at
120°C for 2 min. Observe the color intensity of the beads under a microscope.
To ensure complete couplmg, every bead from the minute quantity of sample
beads should be nmhydrin negative, I e , straw yellow color.
4 Lam and Lebl
9 If the couplmg IS mcomplete (some beads remamed purple or brown with nmhy-
drm test), remove the supernatant from those reaction vials and add fresh Fmoc-
protected ammo acids, BOP, DIEA, and HOBt mto the reaction vial for addmonal
coupling
10. If the couplmg 1s complete (beads remained straw yellow color with nmhydrm
test) discard the supernatants of each reaction vial, and transfer and wash all the

beads to the randomtzation vessel with technical grade DMF
11 After all the 19 couplmg reactrons are completed, all the beads are transferred to
the randomizatton vessel Wash the beads (8 times, 2 mm each) with technical
grade DMF
12 Add 75 mL 50% ptpertdme (m DMF) to the randomtzatton vessel to remove the
Fmoc protectmg group After 10 mm, remove the ptpertdme and add 75 mL fresh
50% prpertdme. After another 10 mm, wash the beads 8 times wtth techmcal
grade DMF and twtce with HPLC-grade DMF
13 Distribute the beads mto each of the 19 reaction vials and carry out the next
couplmg reaction as described above
14. After all the randomtzatton steps are completed, remove the Fmoc protectmg
group with prpertdine as described above
15 After thorough washing with technical grade DMF (5X) followed by DCM (3X), add
10 mL of reagent K (18) to the randomrzatron vessel for 3 h at room temperature
16. Wash the deprotected resms thoroughly with DCM (3X), followed by technical
grade DMF (5X), then once with 10% DIEA to neutralize the resin
17. After thorough washing with technical grade DMF, store the bead library m
HPLC-grade DMF at 4°C. Alternatively, the bead library can be washed thor-
oughly with water and stored in 0.1
M
HCl or 0.1 Mphosphate buffer with 0 05%
sodmm azide.
3.2. Synthesis of a Cyclic Peptide Library
The synthesis of a cyclic peptide library (disulfide bond formation) is essen-
tially the same as that of the linear library except that Fmoc-Cys (Trt) is added
at the carboxyl as well as amino terminus of the linear random peptide After
deprotectton, add a mixture of DMSO/Anisole/TFA
(see Subheading 2.1.,
item 10)
into the resin; incubate overnight at room temperature. After thor-

ough washing, store the library at 4°C as described above.
4. Notes
1 We have tested several commerctally available resins for our library synthesis
The two satisfactory resins are TentaGel (polyethylene grafted polystyrene beads)
and Pepsyn gel (polydimethylacrylamtde beads) Overall, the TentaGel 1s prefer-
able as it is nonsticky and mechanically more stable However, unlike Pepsyn
gel, the level of substrtutron of each TentaGel bead is far from uniform Wtth the
advent of combmatorral chemistry, we anticipate newer resins entering the market
m the near future
One-Bead One-Compound 5
2. TentaGel already has a long polyethylene linker and we do not routmely add
additional linker for our library synthesis In contrast, a linker (preferably a
hydrophilic lmker) is necessary for the synthesis of a peptide library with
polydimethylacrylamide beads. We have used Fmoc-P-alanme and/or Fmoc-
aminocaprorc acid as linkers in the past. However, aminocaproic acid is rather
hydrophobic A polyethyleneglycol-based amino acid (Shearwater, Polymers,
Huntsville, AL) is probably preferable.
3 All glass vessels should be sdiconized thoroughly prior to use Besides using
nitrogen bubbling through the randomization vessel to mix and wash the beads,
we have also prepared libraries in hourglass reaction vessels (Peptides Interna-
tional, Louisville, KY), usmg rocking motion to mix the resins.
4. Each polypropylene reaction vial should be engraved with a letter correspondmg
to a specific amino acid to ensure no mix-up during the synthesis
5. We often omit cysteines from the synthesis of linear peptide libraries to avoid the
complication of intracham and/or interchain crosslinking
Acknowledgments
This work was partially supported by NIH grants CA23074 and CA17094.
Kit S. Lam is a scholar of the Leukemia Society of America.
References
1 Scott, J K. and Smith, G. P. (1990) Searchmg for peptide ligands with an epitope

library. Science 249,386-390.
2. Schatz, P. (1993) Use of peptide libraries to map the substrate specificity of a
peptide-modifying enzyme A 13 residue consensus peptide specifies biotinylation
m Escherichia cob Biotechnology 11,1138-l 143.
3. Kawasaki, G. (199 1) Cell-free synthesis and isolation of novel genes and polypep-
tides. PCT International Patent Application W09 l/05058.
4 Lam, K. S., Salmon, S. E , Hersh, E M., Hruby, V. J , Kazmierski, W. M , and
Knapp, R. J. (1991) One-bead, one-peptide: a new type of synthetic peptide library
for identifymg bgand-bmdmg activity. Nature 354,82-84
5 Lebl, M., Krchnak, V., Sepetov, N F., Seligmann, B., Strop, P., Felder, S and
Lam, K S (1995) One-bead-one structure combmatorial libraries. Bzopolymers
37,177-198.
6 Lam, K S , Lebl, M , and Krchnak, V. (1997) The “one-bead-one-compound”
combinatorial library method. Chem. Rev. 97,41 l-448
7 Geysen, H. M , Rodda, S J., and Mason, T J. (1986) A prior-z delmeation of a
peptide which mimics a discontmuous antigenic determinant. Mol. Immunol. 23,
709-715.
8. Houghten, R. A , et al. (199 1) Generation and use of synthetic peptide combmato-
rial libraries for basic research and drug discovery. Nature 354,84-86
9. Dooley, C. T and Houghten, R A (1993) The use of positional scanning syn-
thetic peptide combinatorial libraries for the rapid determination of opioid recep-
tor ligands Life Scl. 56, 1509-1517
6 Lam and Lebl
10. Deprez, B , Willard, X , Bourel, L , Coste, H , Hyafil, F., and Tartar, A (1995)
Orthogonal combmatorial chemical libraries J Am. Chem. Sot. 117,5405-5408
11 Erb, E , Janda, K., and Brenner, S. (1994) Recenstve deconvolutton of combma-
torial chemtcal ltbraries Proc. Nutl Acad. Scz USA 91, 11,422-l 1,425
12. Zuckermann, R. N , Kerr, J. M , Slam, M A , Banvtlle, S. C., and Santa, D V
(1992) Identification of highest-affinity ligands by affinity selection from eqmmo-
lar pepttde mixtures generated by robotic synthesis. Proc Natl. Acad. Scz. USA

89,4505-4509.
13 Songyang, Z., Carraway, K L , Eck, M. J , Harrtson, S C., Feldman, R. A ,
Mohammadi, M , Schlessmger, J , Hubbard, S. R , Smith, D P , Eng, C., Lorenzo,
M. J., Ponder, B. A J , Mayer, B J , and Cantley, L. C (1995) Catalytic spectfrc-
tty of protein-tyrosme kmases 1s crmcal for selecttve stgnallmg Nature 373,
536-539
14. Furka, A., Sebestyen, F., Asgedom, M., and Dtbo, G (1991) General method for
rapid synthesis of multicomponent peptide mixtures. Int J Peptzde Protein Res
37,487+93.
15 Lam, K. S and Lebl, M (1994) Selectide technology-bead bmdmg screening
Methods f&372-380
16 Lebl, M , Krchnak, V , Salmon, S E., and Lam, K. S (1994) Screenmg of com-
pletely random one-bead-one-pepttde libraries for activmes m solution MethodA
f&381-387.
17 Kaiser, E., Colescott, R. L , Bossmger, C D., and Cook, P. I. (1970) Color test for
detection of free terminal ammo groups m the solid-phase synthesis of pepttdes
Anal. Blochem. 34,595-602.
18 King, D. S., Fields, C G , and Fields, G. B. (1990) A cleavage method which
mmtmtzes side reactions followmg Fmoc solid phase pepttde synthesis Znt. J.
Peptlde Protein Res. 36,255-266.
Enzyme-Linked Calorimetric Screening
of a One-Bead One-Compound Combinatorial Library
Kit S. Lam
1. Introduction
In the “one-bead one-compound” combinatorial library method, each bead
displays only one chemical compound although there are approx 1013 copies of
the same compound in and on the same bead (I-3). With an appropriate detec-
tion scheme, compound-beads with specific biological, physical, or chemical
properties can be identified, and physically isolated, and then their chemical
structure can be determined. In biological systems, one important property that

is of interest is the binding property between a ligand and a ligate. The hgate or
acceptor molecule could be an enzyme (4-6)) an antibody (1,7,8), a receptor
(9,10), a structural protein, or even small molecules (II). Furthermore, the
“one-bead one-compound” library method can also be applied to the discovery
of ligands that bind to the whole viral particle, bacteria, or mammalian cell by
screening for compound-beads that bind to intact cells.
When we mix a ligate with an “one-bead one-compound library,” some com-
pound-beads may be coated by the ligate. This interaction can be detected by
either a labeled ligate or a labeled secondary probe that recognizes the ligate.
Common labels are enzyme, fluorescent probe, color dye, or radionuclide.
There are advantages and disadvantages to each of these methods. The choice
of detection scheme depends largely on the nature and availability of specific
labeled ligates. From our experience, enzyme-linked calorimetric assay is prob-
ably the most convenient, economical, and rapid screening method that does
not require any elaborate equipment (12). Methods for the preparation of the
peptide-bead library are detailed in Chapter 7 of this volume. Details on the
enzyme-linked calorimetric screening method will be given in the next
sections,
From Methods m Molecular Bology, vol 87 Combmatonal Pep/de Library Protocols
Edlted by S CablIly 0 Humana Press Inc , Totowa, NJ
7
8
Lam
2. Materials
All
the reagents needed are standard enzyme-linked immunosorbent assay
(ELISA) reagents and are readily available
from many biochemical and chemi-
cal companies. The following buffers are needed for the screening:
1. PBS-Tween. 8 mM,Na2HP04, 1.5 mMKH2P04, 137mMNaC1,2.7mMKCl,pH

7.2, with 0.1% Tween-20 (v/v)
2 Binding Buffer 16 m&Z Na2HP04, 3 mM KH,PO,, 274 mM NaCl, 5 4 mM KCl,
pH 7 2, with 0.1% Tween-20 (v/v) and 0.1% gelatm (w/v).
3 TBS. 2 5 n&Z Trts-HCl, 13 7 mM NaCl, and 0 27 mM KCl, pH 8 0
4. BCIP/Alkalme phosphatase buffer. 1.65 5-Bromo-4-chloro-3-mdolyl- mg
phosphate (BCIP) m 10 mL of 0 lMTris-HCl, 0 lMNaC1 with 2.34 mMMgCl,,
pH 8.5-9 0
5 Gelatin. 0 1% in water
6 6M Guamdme HCl, pH 1 0
3. Methods
3.1. Screening with an Enzyme-Linked Ligate
Common enzymes used in
ELISA are alkaline phosphatase, horseradish per-
oxrdase, P-galactosrdase, and glucose oxidase. From our experience the alka-
line phosphatase system is more specific and tends to produce the least artifact
when we screen a “one-bead one-compound” library.
1 If ligate-alkaline phosphatase complex is not commercially available, one may
conmgate the ligate to alkaline phosphatase using bifunctional crosslmkmg
reagents Many such reagents are commercially available (e g , Pierce Chemical,
Rockford, IL) and standard coupling procedures are supplied by the manufactur-
ers Before screening a library, one has to make sure that the coqugation method
does not impair the bmdmg property of the ligate This can usually be accom-
plished by an ELISA assay using a 96-well plate coated with a known hgand (see
Note 1)
2 Transfer l-10 mL of the bead-library (200,000 to 2 million beads) to a 50-
lOO-mL polypropylene container Slowly dilute the dimethylformamtde (DMF)
by adding an incremental amount of double-distilled water. Wash the bead-
library thoroughly with double-distilled water m a column (e g., Econo column,
Bio-Rad, Hercules, CA) Coat the bead-library with 0 1% gelatin (w/v) m water
for at least 1 h. Wash the bead-library with PBS-Tween Transfer the library back

into the polypropylene contamer with the bmdmg buffer (see Note 2) Add the
ligate-alkaline phosphatase coqugate into the library with gentle mixing for 1 to
24 h at room temperature (see Note 3).
3. Transfer the bead-library to the column and wash the beads thoroughly with PBS-
Tween. Then wash the bead-library one last time with TBS.
Enzyme-Linked Calorimetric Library Screening 9
Fig. 1. (A) Photomicrograph of a typical enzyme-linked calorimetric bead-library
screen; a positive bead is noted in the middle of the micrograph. (B) Single positive
beads can easily be retrieved with a handheld micropipet under a dissecting
microscope.
4. Transfer and wash the bead-library to lo-20 polystyrene Petri dishes (100
x
20
mm) with the BCIP/alkaline phosphatase buffer (see Notes 4 and 5). More dishes
may be needed if the beads are too crowded and there are too many positive
beads. Let the enzyme-linked color reaction develop for 30 min to 2 h. Stop the
reaction by acidifying the BCIP/alkaline phosphatase buffer with several drops
of 1
M
HCl. Figure 1A shows the photomicrograph of a typical bead-library
screen.
5. With the aid of a light box and a micropipet (e.g., Pipetman PlO, Gilson), transfer
the turquoise beads into a small Petri dish. Many colorless beads will also be
transferred during this process.
6. Place the small Petri dish of positive beads under a dissecting microscope and
pipet individual turquoise beads to a small Petri dish of 6
M
guanidine-HCI, pH
1 .O (Fig. 1B). At this stage, transfer only the positive beads (see Note 6). After
10 Lam

20-30 mm at room temperature m 6 Mguanidme-HCI, transfer the posmve beads
to a dish of double-dtstrlled water. Then prpet each posrttve bead onto a glass
filter and msert mto the protein sequencer cartrtdge for mtcrosequencmg (see
Notes 7 and 8)
3.2. Screening with an Unlabeled Ligate by Probing
with an Enzyme-Linked Secondary Antibody
1 Prepare the library as m Subheading 3.1., item 2.
2 Add the alkaline phosphatase-linked anti-ligate antibody to the bead library and
incubate m bmdmg buffer for l-2 h at room temperature
3 Wash the bead-library thoroughly with PBS-Tween and finally once with TBS
4. Add BCIP substrate to the library as described m Subheading 3.1., item 4
5 After 30 mm to 2 h, stop the colortmetrrc reaction by adding several drops of 1 M
HCl to each Petri dish. Remove all the color beads from the library over a light
box with a mrcropipet These color beads interact with the secondary antibody
alone and may be discarded.
6 Recycle the remaining colorless library with the following steps: Incubate the
library with 6 M guamdine-HCl, pH 1 .O, 20-30 min, wash 5 times with double-
distilled water, mix the library with DMF for 1 h, wash 5 times wrth double-distilled
water, followed by PBS-Tween
7 Add the unlabeled ligate to the bead-library and incubate l-24 h at room
temperature
8 Wash the bead-library thoroughly with PBS-Tween
9. Add the alkaline phosphatase-linked antrligate antibody to the bead-library and
incubate l-2 h (see Note 3) Then wash the bead-library thoroughly wtth PBS-
Tween and finally once with TBS.
10 Add BCIP substrate to the library as described m Subheading 3.1., item 4. After
30 min to 2 h, stop the colorimetrtc reaction by adding several drops of 1
M
HCl
to each Petri dish. Since the library has been prescreened with the secondary

antibody alone, the posrttve beads Identified at this time should be a result of
bmdmg to the ligate and not to the secondary antibody.
11. Isolate those individual posrtive beads for microsequencmg as descrrbed m
Subheading 3.1., items 5,6.
4. Notes
1 For the two-step screening process, instead of using the lrgate/antr-ligate-enzyme
system, one may use a biotmylated-ligate/streptavrdm-enzyme system
2. Most of the methods employed m Western blot or ELISA for lowering the back-
ground can be applied to the screening of the bead-library We routmely add high
salt (2X PBS), 0 1% Tween-20, and 0 1% gelatin to the bmdmg buffer Bovme
serum albumin instead of gelatin has also been used successfully.
3 In order to mmimtze the background and false posmves, the concentration of
ligate, ligate-enzyme conjugate, or antibody-enzyme conjugate used m the
screening should be as dilute as possible Sometrmes lt is advantageous to use a
Enzyme-Linked Colorimetnc Library Screemng II
small sample of resm (e g., 0.1 mL) to test several levels of reagent concentration
before screening a large library It is not uncommon that the concentration of a
reagent can be lo-fold more dilute than the optimal concentration recommended
for standard ELISA.
4. Although a combmatlon of BCIP and mtroblue tetrazohum (NBT) 1s commonly
used m Western blot, we prefer to use BCIP alone. The BCIP/NBT substrate 1s
much more sensitive However, NBT can be reduced to formazon and form a
dark purple preclpltate on the bead if there IS a trace amount of residual reducmg
agent left in the bead-library. Additionally, certain ammo acid sequences such as
Asn-Asn-Asn can reduce NBT to formazon m the absence of alkaline phos-
phatase Furthermore, the formazon deposit on the surface of the bead 1s msoluble
in many of the common solvents that we have tested. Therefore, If BCIP/NBT
substrates are used, we will not be able to recycle the library for subsequent use
or recycle a specific positive bead for confirmatory testmg before sequencing
However, under certain circumstances, the tetrazohum salts are useful as a

substrate as different tetrazolmm salt generates different colors upon reduction
Therefore, a multicolor detection system can be designed for such applxatlons
(13) Neither the formazon (when BCIPlNBT are used) nor the indigo (when
BCIP alone 1s used) products ~111 affect the microsequencmg results
5 Alkalme phosphatase works best under alkaline condltlons (e g , pH 9 5) How-
ever, there 1s a concern about the stability of the llgand-ligate interaction under
such condltlon Therefore, depending on the ligate, we routmely adJust the BCIP/
alkahne phosphatase buffer to pH 8.5 to 9.0.
6. In some instances, a dual-color colorlmetrlc detection scheme may be helpful m
selectmg the true posltlve beads (13).
7 Since the rate-llmltmg step of the “one-bead one-compound” hbrary method IS
the mlcrosequencmg step, one needs to ensure that most of the positive beads
submltted to mlcrosequencmg are “true positives.”
8 To further improve the probability of true positlvlty, one may decolorize the posl-
tive beads with DMF and restam the beads m the presence or absence of a com-
peting llgand.
Acknowledgments
This work was partially supported by NIH grants CA23074 and CA17094.
Kit S. Lam is a scholar of the Leukemia Society of America
References
1 Lam, K S., Salmon, S. E , Hersh, E M , Hruby, V , Kazmlerskl, W M , and
Knapp, R. J. (1991) A new type of synthetic peptlde hbrary for ldentlfymg hgand-
bmdmg actlvlty.
Nature 354,82-84
2 Lebl, M , Krchnak, V , Sepetov, N F , Seligmann, B , Strop, P , Felder, S , and
Lam, K S (1995) One-bead-one structure combmatorlal libraries
Bzopolymen
37,177-198
12 Lam
3

4.
5.
6
7
8
9
Lam, K S , Lebl, M , and Krchnak, V. (1997) The “One-Bead-One-Compound”
combinatorial library method. Chem. Rev 97,41 l-448
Wu, .I , Ma, Q N , and Lam, K. S (1994) Identtfymg substrate motifs of protein
kinases by a random lrbrary approach Biochemistry 33,14,825-14,833
Lam, K. S , Wu, J S , and Lou, Q (1995) Identtfication and characterization of a
novel peptide substrate specific for src-family tyrosme kinase Znti. J. Protean
Peptlde Res. Q&587-592
Lou, Q., Leftwich, M., and Lam, K. S (1996) Identification of GIYWHHY as a
novel peptide substrate for human p60c-src protein tyrosme kmase. Bloorg. Med
Chem., 4,677-682
Lam, K S., Lebl, M., Krchnak, V , Wade, S , Abdul-Lattf, F , Ferguson, R ,
Cuzzocrea, C , and Wertman, K. (1993) Discovery of D-ammo acid contammg
ligands with Selectide Technology Gene 137,13-16
Lam, K. S , Lake, D , Salmon, S E , Smtth, J., Chen, M-L., Wade, S., Abdul-
Latrf, F , Leblova, Z , Ferguson, R. D., Krchnak, V , Sepetov, N. F., and Lebl, M
(1996) A one-bead, one-pepttde combmatorial library method for B-cell epitope
mapping Methods. A Compamon to Methods m Enzymology 9,482-493
Smtth,M H.,Lam,K.S.,Hersh,E M ,andGrtmes,W.(1994)Peptidesequences
bmdmg to MHC class I proteins using a synthetic peptide hbrary approach. Mol
Immunol 31,1431-1437.
10. Salmon, S E , Lam,K S , Lebl, M , Kandola, A., Khattrt, P , Wade, S., Patek, M ,
Kocis, P., and Krchnak, V (1993) An orthogonal partial cleavage approach for
solution-phase rdenttficatton of biologically active peptides from large chemtcal-
synthesized peptide libraries Proc. Nat1 Acad. Scz USA 90, 11,708-l 1,7 12

11. Lam, K. S , Zhao, Z G., Wade, S , Krchnak, V , and Lebl, M (1994) Identtfica-
tion of small peptides that interact specifically with a small organic dye Drug
Dev. Res 33,157-160
12 Lam, K. S and Lebl, M (1994) Selectide Technology-Bead bmdmg screenmg.
Methods 6,372-380.
13 Lam, K. S., Wade, S., Abdul-Latif, F , and Lebl, M. (1995) Application of a dual
color detection scheme m the screening of a random combmatorial peptide library
J Immunol Methods 180,219-223
3
Synthesis and Screening of Positional Scanning
Combinatorial Libraries
Colette T. Dooley and Richard A. Houghten
1. Introduction
Synthetic combinatorial libraries (SCLs) are collections of very large num-
bers of synthetic compounds, in which all possible combinations of the burld-
ing blocks used are represented. The development and verification of the utility
of combinatorial libraries represent a dramatic advance in the drug discovery
process by greatly reducing the time needed to identify new drug leads. Posi-
tional scanning (PS) SCLs (Z,2) represent a modified format of the origmal
synthetic combmatorial libraries described by this laboratory (3). In contrast to
the original libraries, which required several iterative syntheses to identify
individual active compounds, this library format provides mformation on the
substituent responsible for activity at each varied position withm a structure.
Therefore, only a single subsequent synthesis is required. The screening of
PS-SCLs, in most instances, permits the identification of the most active sub-
stituents at each position of a compound in a single assay. Thus PS-SCLs serve
to reduce further the time required to identify new drug leads. PS-SCLs are
composed of individual positional SCLs, in which a single position is defined
with one substituent while the remaining positions are composed of mixtures
of substituents. The defined position is “walked” through the entire sequence

of the PS-SCL. Therefore, the number of positional SCLs is equal to the num-
ber of residues in each compound of the PS-SCL. It should also be noted that
each posrtional SCL, although addressing a single positron of the sequence,
represents the same collection of individual compounds For example, a
hexapeptide, or a compound with six positions, can be represented as:
OIXXXXX, XO,XXXX, XX03XXX, XXO,XX, XXXX05X, or XXXXX06.
From Methods m Molecular Biology, vol 87 Combmatonal Peptrde Library Protocols
Edlted by S Cablily 0 Humana Press Inc , Totowa, NJ
13
14 Dooley and Houghten
Using 20 amino acids (this represents 120 mixtures in total), each peptide mix-
ture contains 3.2 million (205) different sequences, the six positional libraries
each contam 64 million hexamers. Peptide PS-SCLs can be prepared with an
acetylated N-terminus, as well as with a C-terminal amide or carboxylate. One
highly advantageous characterlstlc of the PS-SCLs prepared in this laboratory
is that they are free to interact m solution, i.e., they are not bound to any sup-
port (beads, glass, phage, and so on), and therefore can be readily screened in
any assay system. When used in concert, the data derived from each positional
SCL yield Information about the most important substituents for every posl-
tlon. The information IS then used to synthesize individual compounds repre-
senting all possible combmations of the most active substituents at each
position. This serves to confirm the PS-SCL screening results, as well as to
Identify the individual compounds with the highest activities.
The preparation of a PS-SCL composed of L-ammo acid hexapeptides is
described. This library consists of six separate positional SCLs, each composed
of 20 different peptlde mixtures having a single posltlon defined with one of
the 20 natural ammo acids (represented as 0)) and the remaining five positions
are composed of mixtures of 19 amino acids (represented as X; cysteine is
omitted, see Note 1). The six posltlonal SCLs differ only in the location of the
defined position A description IS given for the preparation of the library using

either Boc or Fmoc chemistries. The choices of procedure depends on the labo-
ratory facilities available, safety, and financial considerations. Methods for
screening such a library in a radioreceptor assay are given as an example.
Although the methods described here involve the use of peptides, the posi-
tlonal scanning concept may be equally applied to a library of any class of
compounds m which there are a number of positions that may be systemati-
cally altered. PS-SCLs have been prepared composed of decapeptides (4),
hexapeptides comprised solely of D-ammo acids (5)) of tetrapeptides comprised
of more than 50
L-, D-,
and unnatural amino acids (6) and of heterocycles (24).
PS-SCLs have been used successfully to identify antigenic determinants rec-
ognized by monoclonal antibodies, trypsm inhibitors, opioid receptor llgands,
and antlmicroblal compounds. A series of papers on the use of peptlde,
peptidomlmetic, polyamme and heterocyclic PS-SCLs in the aforementioned
assays have been published by our laboratory (5-15,24).
2. Materials
2.1. Library Synthesis
2.1.1. General Requrements for Synthesis
1 Resin packets (T-bags) are made with polypropylene mesh (74 pm, Spectrum,
Houston TX), using an impulse sealer.
Synthesis and Screenmg
15
2. T-bags are filled with polystyrene resin, 200 mg, 0.2 mEq.
3. Solvents for all synthetic procedures are dimethylformamide (DMF) and/or
dtchloromethane (DCM)
4 A lyophihzer and somcator are used m the final preparations of the peptide
mixtures.
2.1.2 t-Boc Synthesis
1. Methylbenzhydrylamine (MBHA) polystyrene resm

2 This synthesis employs N-a-Boc-amino acids with the followmg side cham-
protecting groups: benzyl is used as the side chain protection for Asp, Glu,
Ser, and Thr; 2,4-dmitrophenyl for HIS; Z,chloro-benzyloxycarbonyl (CBZ) for
Lys; formyl for Trp, sulfoxide for Met, p-tosyl for Arg; and 2,bromo-
benzyloxycarbonyl for Tyr
3. DCM and isopropanol (IPA) are used alternatively m wash steps.
4. Trifluoroacettc acid (TFA) is used to remove Boc protecting groups.
5. Dusopropylethylamme (DIEA) is used as a base m neutralization steps
6. Dnsopropylcarbodumide (DIC) and 1-hydroxybenzotnazole (HOBt) are used as
coupling reagents
7 Thiophenol, dimethyl sulfide (DMS), ethylenedithiol (EDT), p-cresol, and
hydrogen fluoride (HF) gas are used for side cham deprotection, and methanol is
used m the wash procedure
8 HF gas and a 24-vessel cleavage apparatus are used for peptide cleavage.
2.1.3. Fmoc Synthesis
1 Polyoxyethylene-grafted polystyrene resin (TentaGel).
2. Fmoc-2,4,dimethoxy-4’(carboxymethyloxy)-benzhydrylam~ne, TFA cleavable
linker
3 This synthesis employs Fmoc protected ammo acids with the following side chain
protection groups: t-butyl for Ser, Thr, Tyr, Asp, and Glu, trityl for Cys, His,
Asn, and Gln, Boc for Lys, and 2,2,5,7,8-pentamethylchroman-6-sulfonyl (Pmc)
for Arg.
4. DMF is used in the wash steps
5, Pipendme is used to remove Fmoc protecting groups.
6 DIC and HOBt are used as coupling reagents
7 TFA, trusobutylsilane, water, and “Quick Snap” plastic tubes equipped with a
smtered bottom disc (Isolab, Akron, OH) are used for side chain deprotectlon/
cleavage
8 Tert-butylmethylether, hexane and a benchtop centrifuge are required to precipi-
tate and collect peptides

2.2. Screening
2.2.7. Receptor Assay
1 Aqueous buffer, e g ,50 mMTris, pH 7 4
2 Receptor preparation
76
Dooley and Houghten
3 1-mL Polypropylene tubes wtth caps (Contmental Laboratory Products, San
Diego, CA) and 96-well trays (Costar, Pleasanton, CA)
4. Radtoligand.
5. GF/B filtermats, Tomtec harvester, Beta Plate Liquid Scmttllatton Counter, Beta
Plate scmtillation fluid (Wallac, Garthersburg, MD).
3. Methods
3.1. Library Synthesis
For general procedures on solid phase peptide synthesis, readers are referred
to (16) and (17). The peptide mixtures making up the PS-SCLs are synthesized
by simultaneous multiple peptide synthesis (SMPS) (18). Mixture posrtions
(X) are incorporated by couplmg mixtures of protected ammo acids for pep-
ttdes, or aldehydes, carboxylic acids, and so forth for nonpeptides, using
isokmetic coupling of an excess of a predetermined molar ratio, which com-
pensates for the different couplmg rates of the various amino acid derivattves
(Table 1). The advantage of using SMPS (also referred to as the T-Bag tech-
nology, US. Patent No. 4,63 1,211) is that all wash and deprotection steps may
be carried out in a common vessel. For the hexapeptide library, 120 T-bags
(resin packets) are made and labeled (see Note 2)
3.1.1. t-Boc Synthesis
1 All bags are washed (approx 4 ml/bag for l-m-square bags) 1X DCM, 2X IPA,
2X DCM This 1s to ensure that the bags do not leak
2 Neutralize bags* 3
x
5% DIEA/DCM, 2 mm; 2X DCM, 2X DMF, 1 mm each

3 Activatton/couplmg (see Subheading 3.1.3. for couplmg procedure and Note 3)
4 Wash bags 1X DMF, 2X DCM
5. Deprotect using a 55% TFA/DCM solution for 30 mm
6 Wash bags 1X DCM, 2X IPA, 2X DCM.
7 Steps 2-6 are repeated for the required number of couplmgs
8 Deprotect the peptide srde chains using (a) DNP removal, 2 5% thtophenol/DMF
for 1 h Wash bags 3X DMF, 12X alternating washes of IPA and DCM; and (b)
low-HF 60% DMS, 5% EDT, 10% p-cresol, and 25% HF for 2 h at 0” C. Wash
bags, 8X alternating washes of IPA and DCM, 4X DMF, 3X DCM, 1X methanol
(MeOH).
9 Cleave peptrdes from the resm using 7 5% amsole/HF for 1 h at 0” C
10. Extract pepttdes with water or drlute acetic acid Lyophthze peptide solutions
twice and reconstitute in water at lo-20 mg/mL (see Note 4) Mixtures may be
stored for prolonged pertods at -20” C (see Note 5).
3.1.2. Fmoc syn thesrs
1 Wash bags. 3X DMF, 3X DCM
2 Couple TFA cleavable lmker to resin (100 mEq), shake overmght
Synthesis and Screenmg
17
Table 1
Molar Ratios of Amino Acids Used for Coupling Mixture Positions (X)
Ammo acid Molar ratio
Three letter code Single letter code Boc synthesis Fmoc synthesis
Ala A 0.55 0.75
ASP D 0 67 1.20
Glu E 0.70 1 .oo
Phe F 0 49 0.60
GUY G 0 55 0.70
His H
0.69 0.78

Be I 3 34 2.29
LYS K 1 20 1 .oo
Leu L 0 96 0 77
Met M 0.44 0.71
Asn N 1.03 1.20
Pro 0 83 0.86
Gln
i
102 1 .oo
Arg R 1 26 1.10
Ser S
0 54 0 72
Thr T 0 92 0.98
Val V
2.17 1 80
Trp W 0.73 0 70
T r
y
Y 0.80 071
18.99 18.87
3. Wash bags 5X DMF, 1 min
4 Deprotect using 20% prperidme/DMF, 20 mm.
5. Wash bags 5X DMF (1 mm)
6. Acttvatlon/Couplmg (see Subheading 3.1.3. for coupling procedure). Fmoc-
ammo acld/DIC/HOBt (5 Eq solution m DMF), 90 mm. Test for coupling comple-
tion (see Note 3).
7. Wash bags 5X DMF, 1 mm.
8. Repeat steps 4-7 as necessary
9 Remove resin from bags and place m “Qmck Snap” plastic tubes.
10. Cleave peptides using 1 5 mL of TFA/DCM/H,O/triisobutylsilane 70:20*5.5, for

3 h at room temperature
11 Snap off the tips of the “Quick Snap” tubes and add cleavage solutton to centrt-
fuge tubes. Precipitate pepttdes with cold (4” C) tert-butylmethylether (30 mL)
Centrifuge at 3000g for 10 mm. Dissolve peptides m 15 mL of water, lyophllize
pepttde soluttons twice, and reconstitute m water at lo-20 mg/mL (see Note 4).
Mtxtures may be stored for prolonged periods at -20” C (see Note 5)
18 Dooley and Houghten
Table 2
Coupling Procedure for a Hexapeptide PS-SCL
Couple as O” Counle as Xb
Coupling step 1 Bags 101-120 l-100
Coupling step 2 Bags 81-100 l-80,101-120
Coupling step 3 Bags 61-80 l-60,81-120
Coupling step 4 Bags 41-60 l-40,61-120
Coupling step 5 Bags 21-40 l-20,41-120
Coupling step 6 Bags l-20 21-120
“where 0 = A, or C, or D . or Y See Subheading 3.1.3.
bwhere X = A, and D, and E See Subheading 3.1.3.
3.1.3. Couplmg Procedure
1 The coupling procedure described here is for a hexapeptide PS-SCL. A similar
procedure would be used for a library of any peptide length. Bags are labeled 1 to
120, each set of 20 bags represents a particular position m the hexapepttde At
each couplmg stage of the synthesis, the bags are separated mto vessels as
described m Table 2
2. The 20 bags bemg coupled as 0 (1 of the 20 ammo acids) are separated mto 20
vessels and mdividually coupled to each of the 20 ammo acids (1 = A, 2 = C,
3 = D, and so on) using 0 2 A4 of ammo actd/DCM (6 mL) (with an equimolar
concentration of HOBt for asparagine and glutamme), and 0.2 M DIC/DCM
(6 mL) for 1 h (6 Eq)
3. Bags being coupled as X are combined m a single vessel and coupled using a

mixture of 19 amino acids in ratios described in Table 1 Solutions of ammo acid
mixture, DIC, and HOBt (0.5 M, solubihzed m DMF) are mixed to yield a final
concentration of 0.167 M
4 The N-terminus of the peptides may be acetylated if desired usmg 0 2 M acetyl
tmidazole in DMF (20 Eq).
3.2. Screening and Analysis
3.2.1. Radioreceptor Assay
1. Screening of SCLs requires a high-throughput assay. A 96-well or ELISA format
is recommended. Depending on the type of receptor or radiolabel used, adaptmg
a known protocol to a 96-well format may require several experiments to opti-
mize assay conditions. It is important to have good separation between total bmd-
mg and nonspecific bmdmg (ZlOOO cpm) and httle variation between replicates
(l-5%).
2. Prepare receptor preparation, a membrane-bound receptor m tissue homogenate
is described here as an example Protein content of crude homogenates should be
determined using the methods described by Bradford (19) or Lowry (20), as
appropriate (see Note 6).
Synthesis and Screenmg
Plates 1 and 2
79
I
1 2 3 4 5 6 7 8 9
10 II 12
A
B
C
D
~
E
F

G
H
Plate 3
1 2 3 4 5 6 7 8 9
10 11 12
A
B
C
D
E
F
G
H
Fig 1. Layout of plates for screening the posttlonal scanning library m a
radloreceptor assay. NSB, nonspecific binding, SCI-6, dtluttons of cold hgand for
standard curve, TB, total bmdmg Plates 1 and 2 are duplicates.
3 Perform the bmdmg assays m 1 -mL polypropylene tubes. A sample plan for this
assay IS given m Fig. 1 Usually two replicates are sufftclent for screenmg stud-
ies Minimtze ptpetmg by using multichannel pipets or computertzed ptpetmg
stations,
4. Determine mterassay and intra-assay variation using standard curves, incubate the
radlohgand m the presence of a range of concentrations of an unlabeled hgand (Stan-
dard Curve) Reserve one column of each plate for a standard curve (see Fig. 1).
20 Dooley and Houghten
5. Add peptide mixture (50 pL of 5 mg/mL solution) and the appropriate volumes
of buffer, radtohgand, and receptor preparation to each tube. Because of the large
size of most assays, it IS recommended that the receptors be added last (see
Note 7).
6. Incubate assay tubes until equtlibrmm is reached This is generally longer than
the time needed for the unlabeled ligand to reach equillbrmm, and needs to be

determmed before the assay (see Note 8)
7 Termmate the reaction by filtration through GF-B filters on a Tomtec Harvester
(see Note 9). Wash each sample on the filter with 6 mL of Tris-HCl buffer, at
4” C Count bound radioactivity on a LKB Beta Plate Liqurd Scrntillation
Counter, the counts are expressed m counts per mmute (CPM).
8 Process the raw data using spreadsheet software (lotus, Excel, and so forth)
Average replicates and express as percent mhtbition* 100 - [(Mean - NSB)/
(TB - NSB) * 1001 Graph data such that there are six graphs (one for each posl-
tton of the hexamer); each graph should contain 20 bars (one for each ammo
acid) For certam assays, this data 1s sufficient to identify one, two, or three ammo
acids from each of the SIX posittons, which can then be combined to make mdi-
vidual peptides (Jee step 9 below). For receptor assays, often too many mixtures
are active. Calculate ICsD values (concentration of mixture that mhiblts 50%
radioltgand binding) in order to determine the most active mixtures.
9 Perform competitive mhibition assays as above using serial dtlutions of the pep-
tide mixtures. Prepare five threefold dilutions and use the peptide mixture such
that six concentrations are tested (e g , 1X mixture, 0.3X mixture, 0.1X mixture,
0 037X mixture, 0.012X mixture, and 0.004X mixture). Calculate IC5a values
using the curve-fittmg software, e g , GRAPHPAD (ISI, San Diego, CA) For
small combmatorial libraries such as the hexapepttde PS-SCL, IC,, values can
easily be determined for all 120 mixtures Rank order the IC,, values for each of
the SIX positions, and use these values to choose ammo acids for individual
peptides
10 Synthesize all combinations of the most active mixture(s) for each of the six
positrons as individual peptides (Table 3) The numbers of individual peptides to
be synthesized rises exponentially with the number of amino acids chosen (1 e ,
one ammo acid from each posmon generates one peptide, two amino acids from
each position generates 64 [26] peptides, and three ammo acids for all SIX posi-
tions generates 729 [36] peptides)
11 It is important to note that the activity of the individual pepttdes either supports

or disproves the connectivity of the most active ammo acid at each position found
from the screenmg of the library For example, one of the most active peptide
mixtures from the screenmg of a PS-SCL usmg ELISA was AC-XXXPXX-NH,,
although none of the resulting individual peptides containing prolme at the fourth
position were found to be acttve. However, completron of the iterative process
for this pepttde mixture yielded an active indtvtdual pepttde that was completely
different from the sequences derived from the PS-SCL (21)
Synthesis and Screenmg
Table 3
21
Example of Individual Sequences Derived from Screening
a Hexapeptide Positional Scanning Library
Posltlon
Amino acids chosen Pepttde sequences
oxxxxx
F 1
xoxxxx AT 2
xxoxxx LV 2
xxxoxx DE
2
xxxxox W 1
xxxxxo M 1
8
FALDWM
FALEWM
FAVDWM
FAVEWM
FTLDWM
FTLEWM
FTVDWM

FTVEWM
4. Notes
1
2
3.
4
5
6
Care must be taken when analyzmg data of mixtures containmg cysteme, whtch
is mcluded only in the 0 posttton (i.e., CXXXXX, XCXXXX, XXCXXX,
XXXCXX, XXXXCX, and XXXXXC) If one of these mixtures IS found to have
activity, tt 1s best to iterate (sequentially define the X posmons) or prepare a
posittonal scanning hbrary of that mixture
The resm swells and shrinks depending on solvent used. Bags should be made
large enough to accommodate swelling
It 1s important to ensure that all couplmg reactions go to completion. It is htghly
recommended that ninhydrm (22) or bromophenol blue (23) monitormg be car-
rted out on control resins after each coupling.
Somcation is used to solubdize mixtures containing hydrophobic amino acids m
the defined positions It 1s important to keep the water m the somcator cool, add
ice if necessary
It 1s recommended that the library be altquoted before storage to avoid freeze and
thaw damage. Addmonally, if the library is allquoted mto a 96-well format, the
ptpeting required for screening 1s substantially reduced
Lowry Method: To 0.1 mL of sample add 0.1 mL of 2N NaOH. Boll at 100” C for
10 min. Cool and add 1 mL of reagant (100 ~012% w/v Na,COa m water, 1 vol
1% w/v CuSo4*5Hz0 m water, and 1 ~012% sodium potassium tartrate m water).
Incubate for 10 mm. Add 0.1 mL of Folm reagant, mix, and mcubate for 30 mm
Read absorbance at 750 and/or 550 nm. Prepare a standard curve using bovine
serum albumin (BSA) and use tt to determine the concentratron of the sample.

Bradford Method. Dissolve 100 mg Coomassie Blue G250 in 50 mL of 95%
ethanol, mix with 100 mL of 85% phosphoric acid, and make up to 1 L with water.
Filter the reagent Ptpet 0 1 mL of sample into test tube, add 5 mL of reagent, and
mix Measure absorbance at 595 nm lo-60 min after mtxmg. Prepare a standard
curve using BSA and use it to determine the concentration of the sample.
22
Dooley and Houghten
7. Screenmg of this hbrary m an optold radioreceptor assay was optimal at a fmal
concentration of 0 4 mg/mL. When screening a new receptor, it is advisable to
screen at a high concentration (l-5 mg/mL) and subsequently decrease the con-
centration if too many mixtures are found to be active
8 The l-mL polypropylene tubes come with plugs m strips of eight We have found
that the only way to ensure adequate mtxmg of components 1s to cap all tubes and
invert the tray two or three times, tapping both ends to dislodge any solvent from
the top or bottom of the tube
9 We have found that soaking filters in polyethylemmme (PEI), as 1s often recom-
mended to reduce nonspecific binding, causes problems when using the filters
obtained from Wallac. PEI causes the ink and portions of the filter to stick to the
harvester A 5-mg/mL BSA/buffer solution has generally sufficed to muumlze
nonspecific bmdmg
Acknowledgments
This work was funded in part by Trega Biosciences, Inc. (formerly Houghten
Pharmaceuticals), San Diego, CA.
References
1 Pnulla, C., Appel, J R , Blanc, P , and Houghten, R A. (1992) Rapid identiftca-
tion of high affmny peptide hgands using positional scanning synthetic peptide
combinatorial libraries. BloTeclznzques
13,901-905
2. Dooley, C. T. and Houghten, R. A (1993) The use of postttonal scanning syn-
thetic peptide combmatorial libraries for the rapid determmation of opioid recep-

tor hgands Lzfe Scz 52, 1509-15 17
3 Houghten, R. A., Pnulla, C., Blondelle, S E., Appel, J R., Dooley, C T., and
Cuervo, J H. (199 1) Generation and use of synthetic peptide combmatorial librar-
ies for basic research and drug discovery. Nature 354,84-86
4 Pmllla, C., Appel, J R , and Houghten, R A (1994) Investigation of antigen-
antibody mteractions using a soluble nonsupport-bound synthetic decapeptide
library composed of four trillion sequences Bzochem J.
301,847-853.
5. Pimlla, C , Appel, J R , Blondelle, S. E., Dooley, C T., Elchler, J , Ostresh, J M.,
and Houghten, R. A (1994) Versatility of positional scannmg synthetic combma-
torial libraries for the JdentifJcatlon of mdtvidual compounds Drug Dev. Res. 33,
133-145
6 Dooley, C T , Bower, A. N , and Houghten, R A (1996) Identification of
mu-selective tetrapeptides using a positional scannmg combmatorial library con-
taming
L-,
D- and unnatural ammo acids, m Peptldes Chemutry, Structure and
Biology (Proceedmgs of the Fourteenth American Peptzde Symposium (Kaumaya,
P T P and Hodges, R S., eds ), Escom, Leaden, Germany, pp 623,624
7 Pmilla, C , Buencammo, J., Appel, J R., Houghten, R A , Brassard, J A., and
Ruggeri, Z M (1995) Two antipeptide monoclonal antibodies that recogmze
adhesive sequences in fibrinogen Identification of antlgenlc determinants and
Synthesis and Screenmg
23
unrelated sequences using synthetic combmatorial libraries Blamed. Pept. Pro-
terns Nucletc Actds 1, 199-204.
8 Pimlla, C Buencamino, J. Appel, J. R., Hopp, T. P , and Houghten, R A (1995)
Mapping the detailed speciftctty of a calcmm-dependent monoclonal antibody
through the use of soluble positional scanning combmatonal libraries identifica-
tion of potent calcium-independent antigens. Mol Diversity 1,21-28.

9. Ostresh, J. M., Husar, G. M., Blondelle, S. E., Dorner, B , Weber, P A., and
Houghten, R. A. (1994) “Ltbrartes from libraries”. Chemical transformatton of
combmatorial libraries to extend the range and repertoire of chemical diversity.
Proc. Nut1 Acad Scl. USA 91,11,138-l 1,142
10. Perez-Pay&E , Takahashi, E., Mmgarro, I., Houghten, R. A., and Blondelle, S. E.
(1996) Use of synthetic combmational libraries to identify pepttde mhtbitors of
Ca2’-complexed calmodulm, m Peptides: Chemistry, Structure and Btology (Pro-
ceedings of the Fourteenth American Peptzde Symposium) (Kaumaya, P T P and
Hodges, R S., eds ), Escom, Leiden, Germany, pp. 303,304
11 Perez-Pay& E., Houghten, R. A., and Blondelle, S E. (1996) The destgn of self-
assemblmg pepttde complexes usmg conformattonally defined ltbrartes, m Tech-
niques tn Protem Chemtstry VII (Marshak, D., ed ), Academic Press, San Diego,
CA, pp. 65-7 1.
12. Blondelle, S. E , Houghten, R. A , and Perez-Pay& E (1996). All D-ammo acrd
hexapepttde mhlbttors of meltttrn’s cytolyttc activity derived from synthetic com-
bmatonal bbraries J. Mol. Recog. 9, 163-168.
13 Blondelle, S E.,Takahasht, E , Houghten, R. A., and Perez-Pay& E. (1996) Rapid
rdentifrcation of compounds having enhanced antimicrobial activity using confor-
mattonally defmed combmatortal ltbrartes. Bzochem. J. 313, 141-147
14. Dooley , C. T. and Houghten, R A (1995) Identtficatton of mu-selecttve poly-
amine antagonists from a synthetic combmatorial library, Analgesza, 1,400404
15 Eichler, J., Lucka, A. W , and Houghten, R A. (1994) Cyclic pepttde template
combmatorial libraries Synthesis and identification of chymotrypsin inhibitors,
Pept. Res 7,300-307
16 Steward, J M and Young, J D. (1984) Soled Phase Pepttde Syntheses 2nd ed ,
U.S.A Pierce Chemtcal Company, Rockford, IL
17. Pennmgton, M. W and Dunn, B. M., eds. (1994) Pepttde Synthesis Protocols
Methods in MoEeculur Biology. Humana, Totowa, NJ.
18 Houghten, R A. (1985) General method for the raptd solid-phase synthesis of
large numbers of peptides. specificity of antigen-antibody mteraction at the level

of mdividual ammo actds Proc. Nat/. Acad. Set. USA, 82,5 13 1-5 135
19 Bradford, M M. (1976) A rapid and sensitive method for the quantitation of
mrcrogram quanttties of protem utibzmg the princtple of protein-dye binding.
Anal. Btochem 72,248-254
20. Lowry, 0. H , Rosebrough, N. J , Farr, A L., and Randall, R J. (195 I) Protein
measurement with the Folm phenol reagent J. Btol. Chem. 193,265-275
21. Pimlla, C , Buencammo, J., Houghten, R A , and Appel, J R (1995) Detailed
studies of antibody specificity using synthetic combinatorial libraries, m Vuccznes
24
Dooley and Houghten
1995: Molecular Approaches to the Control of Infectlow Diseases (Brown, F ,
Chanock, R., Ginsberg, H , and Non-by, E., eds.), Cold Sprmg Harbor Laboratory
Press, Cold Spring Harbor, pp 13-17
22 Kaiser, E T., Colescott, R. L., Blossmger, C D , and Cook, P. I. (1970) Color test
for detection of free terminal ammo groups m the solid-phase synthesis of pep-
tides. AnaE. Blochem. 34,595-598
23 Krchnak, V , Vagner, J., Safir, P,, and Lebl, M (1988) Nonmvasive contmuous
momtormg of solid phase peptide synthesis by acid-base mdtcator. Collect. Czech.
Chem Commun. 53,2542-2548.
24 Nefzi, A., Ostresh, J M , Meyer, J -P , and Houghten, R A (1997) Solid phase
synthesis of heterocycltc compounds from linear pepttdes. cyclic ureas and
thioureas Tetrahedron Lett. 38,93 l-934
4
Synthesis and Screening of Peptide Libraries
on Continuous Cellulose Membrane Supports
Achim Kramer and Jens Schneider-Mergener
1. Introduction
There are different strategies for the construction of soluble and solid phase-
bound chemrcal peptide libraries. These libraries have been used for the detec-
tion of epitopes as well as for the identification of peptides that act as

antagonists of medically relevant proteins. We have prepared different types of
cellulose-bound peptide libraries by spot synthesis (I), which is a powerful
tool for the simultaneous and positronally addressable synthesis of thousands
of peptrdes or peptide mixtures bound to continuous cellulose membrane sup-
ports. Presently up to 8000 different spots (peptides or peptide mixtures) can
be automatically synthesized on a 20 x 30-cm cellulose membrane and screened
for ligand binding within l-2 wk.
These libraries can be used for the detection of peptides that bind to pro-
teins, metals, and nucleic acids (2). We have mapped several linear and nonlm-
ear antibody epitopes (3-9), and also used this approach for the detection of
receptor-ligand interactions For instance, cellulose-bound peptide scanning
libraries allowed the detection of the contact sites between tumor necrosis fac-
tor-o, and interleukin-6 with its receptors (4,IO). Furthermore, this method
proved to be valuable to characterize heat shock protein-peptide mteractlons
(II). As another biologically relevant application, these libraries were applied
for the study of metal-peptide interactions. For example, we identified tech-
netium-99m binding peptides important for tumor diagnosis (12) and nickel-
binding peptides that can be used for the purification of recombinant proteins (3).
The synthesis of cellulose-bound peptide libraries is not restricted to
L-amino acids. Other building blocks, such as o-amino acids, unnatural ammo
acids, and organic compounds, can also be used. Furthermore, the synthesis of
From Methods m Molecular Bology, vol 87 Combmatonal Pepbde Library Protocols
Edlted by S CablIly 0 Humana Press Inc , Totowa, NJ
25