A sensitive fluorometric assay for quantitatively measuring specific peptide binding to HLA class I and class II molecules

JOIJMAL OF IMMU#OlOGlCAL METHOOS ELSEVIER

Journal of Immunological

Methods 200 (1997) 89-97

A sensitive fluorometric assay for quantitatively measuring specific peptide binding to HLA class I and class II molecules Tom H.M. Ottenhoff *, Annemieke Geluk, Mireille Toebes, Willemien E. Benckhuijsen, Krista E. van Meijgaarden, Jan Wouter Drijfhout Department of Immunohematology and Blood Bank, Leiden Unirersig Hospital. Leiderz. Nether1atld.c Received

13 June 1996; revised 6 September

1996: accepted 23 September

1996

Abstract A sensitive, highly reproducible assay was developed for measuring binding of peptides to various HLA class I and II alleles. The assay is based on competition for binding to HLA between a peptide of interest and a fluorescent labelled standard peptide. This mixture is incubated with HLA to obtain equilibrium binding, and subsequently separated on an HPLC size-exclusion column in (8 a protein fraction containing HLA and bound peptide and (ii) a free peptide fraction. Each assay uses only 100 fmol labelled peptide and approximately 10 pmol of HLA. The analytical system contains an autosampler that samples from 96-well microtiter plates. Injections and data recording/evaluation is fully automated. Typical analysis time is IO-12 min per sample. The fluorescence in the HLA-bound peptide and free peptide containing fractions is measured on-line. The ratios of fluorescence signal in protein and peptide fractions at various concentrations of the peptide of interest are determined. IC,, values are calculated from the binding curve as obtained by curve fitting of the data. Here we show results for peptide binding to HLA-DRI and -DR17 molecules purified from detergent solubilized cell lysates, and for recombinant HLA-A* 0201 and HLA-A* 0301 expressed in E. coli. The assay reported is sensitive and reproducible. It is non-radioactive and is non-labor intensive due IO the high degree of automation. Keword.cc Fluorometric

assay: Peptide binding:

HLA class I; HLA class II

1. Introduction MHC molecules bind and present processed peptide antigens to T cells. MHC class I molecules usually bind peptides derived from cytoplasmic, ‘in-

’ Corresponding

author. At: Department of Immunohematology I, E3-Q. Leiden University Hospital, P.O. Box 9600.2300 RC Leiden, Netherlands. Tel.: ( + 31)-71-5263800; Fax: (+31)-71.5216751: e-mail: [email protected] and Blood Bank, Bldg.

0022-1759/97/$17.00 Copyright PII soo22- I759(96)00 190.

tracellular’ proteins and present these to CD8+ T cells. The latter are thus able to survey intracellular compartments for the presence of virally encoded or tumour associated antigens. By contrast, the majority of MHC class II molecules present peptides from extracellular proteins or proteins residing in intracellular vacuolar compartments that are topographically related to the extracellular milieu. These peptides are presented to CD4+ T cells which therefore scrutinize the ‘extracellular’ milieu for the presence of foreign target antigens (Germain, 19951.

0 1997 Elsevier Science B.V. All rights reserved

I

T.H.M.

90

Ottenhoff

et al. / Jourml

of Immunological

Peptides interact with MHC molecules through contact of particular peptide ‘anchor’ residues with highly polymorphic MHC pockets as well as through hydrogen bonding and salt bridging with conserved MHC structures. Depending on the relative contribution of these interactions to the total binding energy, some peptide-MHC interactions are highly allele specific, e.g. in the case of peptide-HLA DR17 interactions (Geluk et al., 1992), whereas others are much more promiscuous (Panina-Bordignon et al., 1989). Precise understanding of peptide/MHC/T cell receptor interactions in terms of affinity, stability and specificity is important with regard to the selection and design of therapeutically effective peptides. Proper MHC binding peptides can be used to vaccinate against microbes or tumour cells whereas, alternatively, altered peptide ligands (APL) can be designed that are capable of downmodulating undesired autoaggressive responses in autoimmune disease (Allen, 1994; Sette et al., 1994a). Peptide/MHC class II on-rates are usually slow but can be strikingly enhanced at low pH (Jensen, 1991) probably reflecting the acidic nature of the endosome/lysosome like compartments where peptide-class II binding occurs in vivo. Once peptide/MHC complexes are formed, off rates are typically low (Buus et al., 1986). Various assays to measure peptide/MHC interactions have been described but most of these are either nonor only semi-quantitative (Geluk et al., 19921, rely on the use of radiolabelled isotopes (Kast et al.. 1994; O’Sullivan et al., 1990). permit the analysis of peptide binding to only class I or class II molecules

Methods

200 (1997189-97

(Kast et al., 1994; O’Sullivan et al.. 19901, or consume relatively large amounts of HLA. For these reasons we have developed a sensitive, fluorescence based assay that automatically quantifies interactions between peptide antigens and HLA class I as well as HLA class II molecules.

2. Materials

and methods

2.1. Peptide synthesis Synthetic peptides were made on an Abimed 422 multiple peptide synthesizer (Abimed, Langenfeld, Germany) at 10 pmol scale (Gausepohl et al., 1990a.b). TentagelS AC resins (Rapp et al., 1990; Sheppard and Williams, 1982) (Rapp, Tubingen, Germany) were used in combination with Fmoc-protected amino acids carrying TFA-labile side chain protecting groups where needed (Fields and Noble, 1990). Acylations were carried out with a six-fold excess amino acid using PyBOP/NMM activation in NMP (Caste et al., 1990; Gausepohl et al., 1990b). Deprotection was performed with piperidine/DMA l/4 (v/v>. Cleavage of the peptides and removal of the side chain protecting groups was performed with TFA/water 19/l (v/v) for 2.5 h. For W containing peptides ethanethiol (5%) and for C containing peptides triethylsilane (2%) was added to the cleavage cocktail (Pearson et al., 1989). Peptides were isolated and purified by repeated ether precipitations, dissolved in 10% acetic acid and lyophilized.

Table 1 Peptides used in this study Peptide Peptide Peptide Peptide Peptide Peptide Peptide Peptide Peptide Peptide Peptide

Sequence

1 2 3 4 5 6 7 8 9 10

Fl = fluorescein

FLPSDC(Fl)FPSV YIGEVLVSV KVFPC(FI)ALINK KVFPYALINK YMLDLQPETT QVPLRPHTYK FI-Ahx-PKYVKQNTLKLAT Fl-Ahx-KTIAYDEEARR PKYVKQNTLKLAT KTIAYDEEARR label. Ahx = h-aminohexoyl

Allele

Source

Ref.

HLA-A* 0201 HLA-A” 020 1 HLA-A” 0301 HLA-A* 0301 HLA-A” 0201 HLA-A” 030 1 HLA-DR 1 HLA-DR 17 (HLA-DR3) HLA-DR 1 HLA-DR17 (HLA-DR3)

HBV core ant. (47-56) Class I myosin Consensus sequence Consensus sequence HPV 16E7( 1 l-20) HIV-l NeX73-82) Influenza hemaglutinin Heat shock protein 65 Influenza hemaglutinin Heat shock protein 65

Naumann et al. (1993) Den Haan et al. (1995) Sette et al. (1994b) Sette et al. (1994b) Seedorf et al. (1985) Arya and Gallo (19861 Geluk et al. (19921 Geluk et al. t 1992) Geluk et al. (1992) Geluk et al. (1992)

spacer between the label and the peptide.

T.H.M. Ottenhoffet

2.2. Introduction peptides

of fluorescent

label

al./Journal

of Immunological

in synthetic

2.2.1. HLA class II birlding peptides Epitope sequences were N-terminally elongated with an 6-aminohexanoyl spacer (Ahx. see Table 11 in order to reduce possible steric hindrance of peptide binding to HLA due to the presence of the fluorescent label. The fluorescent label (Fl) was attached to the elongated HLA class II binding peptides at the N-terminus of the peptide. At the end of the synthesis an N-terminal S-acetyl mercaptoacetyl group (SAMA) (Drijfhout et al., 1990) was introduced using an equimolar mixture of SAMA-pentafluorophenyl ester and 1-hydroxybenzatriazole (both 60 pmol in 200 ~1 NMP). Cleavage and isolation of the SAMA peptides was performed as described above. A SAMA peptide (about 3 mg> was dissolved in a mixture of 250 ~1 sodium phosphate buffer 150 mM. pH 8.0 and 150 ~1 acetonitrile. 10 ~1 hydroxylamine. HCI (1 M solution in water) was added, immediately followed by the addition of 1 mg solid 4-(iodoacetamido) fluorescein. The reaction mixture was vortexed for 3 min to dissolve the labelling reagent and stirred for 2 h in the dark. The reaction mixture was applied to a Sephadex GlO column (30 X 1 cm) in 100 mM sodium phosphate pH 7.5. The yellow fluorescent peptide fraction was isolated and further purified by analytical RP-HPLC (water/acetonitrile containing 0.1% TFA). Concentrations were determined in sodium phosphate 0.1 M pH 7.5 by OD,,,, measurements (E~,+(~ = 8.2 X 10’ 1 . molp’ . cm _ ’ 1. The fluorescent peptides 7 and 8 (see Table 11 were analyzed by MALDI-MS (LaserMat. Finnigan, UK) using internal peptide standards. 2.2.2. HLA class I binding peptides Labelling of HLA class I binding peptides was performed by attachment of the fluorescent label to a cysteine which was introduced into known HLAA” 0201 and HLA-A* 0301 binding standard sequences (Kast et al., 1994). The position of the cysteine was chosen such that it was at a position where the labelled cysteine side chain projected out the peptide binding groove, thus excluding interference with binding to HLA. FLPSDC(Fl)FPSV, modified HBV core antigen 47-56 (Naumann et al.,

Method:, 200 (19971 89-97

91

19931 was used for HLA-A* 0201 and the KVFPC(Fl)ALINK, modified HLA-A3-binding consensus peptide (Sette et al., 1994b) for HLA-A” 0301. The cysteine peptides were synthesized. labelled. purified and analyzed as described above with the exception that the hydroxylamine treatment for removing the S-acetyl was omitted. 2.3. Pur$ication

of DR molecules

As a source of DR molecules EBV-BLCL homozygous for DR were used: LG2.1 (DRB 1 * 01011 and MAT (DRB1*030l/DRB3” 01011. Cells were cultured in RPM1 1640 (Gibco, Paisley, UK), supplemented with 2 mM t_-glutamine (Gibco), 100 U/ 100 pg/ml penicillin/ streptomycin solution (Gibcol, and 10% heat-inactivated FCS (Gibcol. Cells were lysed at a concentration of 10’ cells/ml in 50 mM Tris-HCl, pH 8.5, containing 2% Renex (Accurate Chemicals and Scientific Corp., Westbury, NY), 1.50 mM NaCl. 5 mM EDTA, and 2 mM PMSF. The lysates were cleared of nuclear and other debris by centrifugation at 10000 X g for 20 min. HLA class II molecules were purified essentially as described before (Gorga et al., 1987). The lysates were passed through the following columns, using a flow rate of 30 ml/h: Sepharose CL-4B (10 ml), protein A-Sepharose (5 ml). W6/32 (anti-class I>protein A-Sepharose (10 ml), and B8.1 1.2 (antiDRl-protein A-Sepharose (10 ml). The columns were washed with IO-column volumes of 10 mM Tris-HCI (pH 8.01 with 0.1% Renex, two-column volumes of PBS, and two-column volumes of PBS-l% octylglucoside. Bound DR was eluted from the B8.11.2 column with 50 mM diethylamine in 150 mM NaCl containing 1% octylglucoside and 0.02% NaN, (pH 11.51. The eluate was immediately neutralized with 2 M glycine (pH 2.51 and concentrated through an Amicon 8050 YM30 membrane under Nz pressure. Protein content was evaluated by a BCA protein assay (Pierce Chemical Co.. Rockford, IL) and the purity confirmed by SDS-PAGE. 2.4. Purification of recnmbinarlt HLA class 1 molecules and ~z-microglobldin. complex ,formation Recombinant HLA class I (A2. A31 molecules and µglobulin were overexpressed and puri-

92

T.H.M. Ottenhoff er al. / Joumul of Immunologiccd Methods 200

fied essentially as described (Garboczi et al., 1992). Complexes were formed and isolated as described (Garboczi et al., 1992). For HLA-A* 0201 folding peptide 5 was used (YMLDLQPETT, HPV16E7(11/20) (Seedorf et al., 1985>, for HLAA* 0301 folding peptide 6 was used (QVPLRPHTYK, HIV- I Nef(73-82) (Arya and Gallo, 1986). Isolation of the complexes was performed on a Superdex 75 16/60 column (Pharmacia, Sweden). 2.5. HLA-peptide

binding assay

The HLA preparations were first titrated in the presence of 100 fmol (6.7 nM) standard peptide to determine the HLA concentration necessary to obtain IO-40% binding of the fluorescent peptide. All subsequent (competition) assays were performed at these predetermined concentrations. HLA protein was incubated for 48 h at room temperature at pH 4.5 for class II and pH 7.0 for class I with 100 fmol (6.7 nM) fluorescent standard peptide in 15 ~1 final fluid volume containing a buffer consisting of 100 mM sodium phosphate, 75 mM NaCl, 1 mM CHAPS (Merck, Darmstadt, Germany) and either 15% (v/v) CH,CN for HLA class II or no CH,CN for HLA class I. All experiments were performed in the presence of a protease inhibitor mixture (1 PM chymostatin, 5 PM leupeptin, 10 PM pepstatin A, 1 mM EDTA, 200 /IM pefablock, Boehringer Mannheim). In the case of class I, exogenous P,-microglobulin (30 pmol/l5 ~1 (1 PM) Sigma) was added as well. For the determination of binding capacities the test peptides were added to HLA molecules simultaneously with the fluorescent standard peptide. After 48 h incubation the mixture was diluted 6-fold with running buffer (150 mM sodium phosphate, 75 mM NaCl, 1 mM CHAPS 15% CH,CN (v/v>. pH 7.0) and applied to a size-exclusion HPLC column, (Synchropak GPClOO (250 mm X 4.6 mm: Synchrom, Lafayette, IN). Samples were introduced into the HPLC system by means of a Jasco AS950 autosampler (B&L Systems, Zoetermeer, Netherlands, sampling from a microtiter plate, U-bottom 96 well serocluster polypropylene, Costar, Cambridge. MA) which makes it possible to analyze samples every 12 min with full automation. The HPLC was run isocratically with running

C1997) 89-97

buffer at a flow of 0.7 ml/min. to separate the HLA protein fraction from the free peptide fraction. The fluorescence of the column effluent was monitored with a Jasco FP-920 fluorescence monitor (B&L Systems, Zoetermeer, Netherlands): excitation was at 490 nm (bandwidth 18 nm) and emission was at 528 nm (bandwidth 18 nm). The percentage of standard peptide bound to HLA was calculated as the ratio of the fluorescence signal of the HLA protein containing fraction and that of the sum of the peptide and the protein containing fraction. Peptides were typically tested at concentrations ranging from 6 nM to 60 PM. Ratios of the fluorescence signal measured in the HLA protein versus free peptide were expressed as a percentage of the binding of the labelled peptide. This was done by setting the ratio obtained in the absence of the competitor peptide to ‘100% binding’ and the ratio obtained in the absence of HLA to 0% binding. The concentration of competitor peptide yielding 50% inhibition of binding of the standard peptide was deduced from the dose-response curve as obtained by curve fitting with TableCurve 2D software (Jandel Scientific) using the dose-response equation: .v = CI+ b( 1 + ( x/c)~). Each peptide was tested in at least three separate experiments and curve fitting was done with the average binding value at each concentration.

3. Results 3.1. Analysis of peptide binding to puriJied class II CHLI-DRI and HLA-DR17) molecules

HU

In order to analyze binding to purified HLA class two HLA-DR alleles, DRI II molecules, (DRBl*OlOl) and DR17 (DRB1*0301), were chosen as model class II molecules because they have largely non-overlapping binding specificities. Binding of the influenza hemagglutinin peptide HA(307/3 19) PKYVKQNTLKLAT to DR 1 molecules has been extensively studied in in vitro peptide-HLA binding studies (Geluk et al., 1992; O’Sullivan et al., 1990; Rothbard et al., 1989; Stem et al., 1994). HA(307/319) is known to bind well to with the exception of DR17 all DR alleles, (DRB I * 0301) and DR9 (DRB 1 * 090 I) (Geluk et al.,

T.H.M.

Ottenhqff

Table 2 Binding of Fl-HA(307/319)

to HLA-DRI,

Concentration

(nM)

HA(?07/319)

0 7 70 700 7 000 70 000

et al. / Joumul

inhibition

oj”lmmunologicol

by HLA-DRI

% binding of Fl-HA(307/3

binder HA(307/319)

19)

39 34 ?I 3.2

Concentration

200

(I 997189-97

and HLA-DR17

HSP65(3/

13) (nM)

0 7 70 700 7 000 70 000

1.A 0

L)3

binder HSP65(3/13) % binding of FI-HA(307/3

19)

38 36 39 38 37 n.t

ICC0 z-. 1 mM

IC,,, = 77 nM

1992). Peptide binding to DR17 was measured using peptide 8 (Table I> as the standard fluorescent peptide. Thus HSP65(3/ 13) binds well to DR17, the most frequently occurring subtype of DR3, but not to any other DR molecule (Geluk et al., 1994). Binding affinities of HA(307/319) for DRl and of HSP65(3/ 13) for DR17 were analyzed by the simultaneous addition of the unlabelled peptides in several concentrations (ranging from 7 nM to 7 PM) together with 6.7 nM of the labelled standard peptides 7 and 8 to purified HLA-DR. The results of these experiments are shown in Tables 2 and 3 and Fig. 1 and Fig. 2. In the absence of inhibitor peptide the fluorescence pattern for both DRl and DR17 showed two major peaks with retention times of approximately 3 and 5 min (Fig. 1). Analysis of the fluorescent peptide without addition of HLA resulted in only one peak with a retention time of about 5 min. As HLA molecules by themselves did not result in any fluorescent signal (data not shown) the peak with a retention time of 3 min was assumed to contain fluorescent labelled peptide complexed to HLA-DR. Addition of unlabelled peptide to the mix-

Table 3 Binding of Fl-HSP65(3/13) Concentration

Methods

HSP65(3/13)

0 7 70 700 7 000

IC,,, = 102 nM

to HLA-DRi7, (nM)

inhibition

by HLA-DR17

R binding of Fl-HSP65(3/13) 36 31.1 11.7 3.6 0.8

ture of labelled peptide and HLA-DR led to a dosedependent, peptide specific decrease of the peak at 3 min, indicating that the unlabelled peptide was able to compete with the fluorescent peptide for binding to HLA-DR (Fig. 1). No inhibition was seen when non-binding test peptides were added, demonstrating the specificity of the peptide/MHC binding as detected by the assay. For the interaction between DRI and HA(307/319) the IC,, value was 77 nM. The IC,, value for HSP65(3/ 13) and DR17 was 102 nM. Similar data have been obtained for DR2 (DRBl* 1501), DR4 (DRB1*0401) and DR7 (DRB I* 0701) (data not shown). 3.2. Analysis of peptide binding to purified recombiplant HLA class I 1HL4A2

and HLA-ASI molecules

Peptide binding by class I molecules was analyzed for recombinant HLA-A* 0201 and HLAA* 0301. Fluorescent peptides 1 and 3 (Table 1) were used as standard peptides. Binding of the fluorescent peptides resulted in two peaks with retention times of approximately 4 and 7 min for both alleles.

binder HSP65(3/13) Concentration

and HLA-DRI

HA(307/319)

0 7 70 700 7 000

IC,, >-) 100 /*M

(nM)

binder HA(307/319) % binding of Fl-HSP65(3/ 32 33 36 35 31

13)

T.H.M.

94

Ottenhoff et al./Joumal

qf lmrnunotogical

Methods 200 (1997) KY-97

10-

51 oi_

_~

0.01

Fig. 7. Curve chromatograms nM.

2

6

1

8

2

n&l

e

1

8

Id”

Fig. 1. Chromatograms of mixtures of HLA-DRI and Fl-AhxPKYVKQNTLKLAT containing increasing concentrations of competitor PKYVKQNTLKLAT. A = 0 nM, B = 7 nM. C = 70 nM, D = 700 nM. E = 7 PM, F = 70 WM. The HPLC was run isocratically with running buffer at a flow rate of 0.7 ml/min to separate the HLA protein fraction fat about 3 min) from the peptide fraction (at around 5.5 min). The fluorescence of the column effluent was monitored: excitation was at 490 nm (bandwidth 18 nm) and emission was at 528 nm (bandwidth 18 nm).

Table 4 Binding of FLPSDC(FI)FPSV RRIKEIVKK Concentration

YIGEVLVSV

0 7 23 70 230 700 2300 7000 23 000 70 000 230 000

IC 50 = 147 nM

to HLA-A’O201.

(nM)

inhibition

by

‘%binding of FLPSDC(FI)FPSV 56 55 55 35 16 5 3

I 0 0 0

~~_.

-_-__

___.

1 10 100 le+03 inhibitor concentration (nM)

0.1

le+o4

1e+05

fitting of the binding data as obtained from the in Fig. I using TableCurve 2D software. IC,, = 77

The latter peak corresponded to the free labelled peptide since labelled peptide alone showed a retention time of 7 min. Equimolar amounts HLA-A* 0201 and HLAA* 0301 (700 nM) incubated with excess &m and fluorescent peptide 1 or 3 (7 nM) did not result in identical percentages of binding (Tables 2 and 3). This indicates that the recombinant HLA-A” 0201peptide complexes used allow easier substitution by standard peptide than the HLA-A* 0301-peptide complexes. This might be a reflection of the higher affinity of peptide 5 for HLA-A* 0201 compared to the affinity of peptide 6 for HLA-A* 0301. and/or

HLA-A* 0201 Concentration

YIGEVLVSV

RRIKEIVKK

(nM)

and

I mM

HLA-A- 0201

non-binder

% binding of FLPSDC(FI)FPSV 56 nt nt nt nt nt nt nt 52 54 51

0 7 23 70 230 700 2300 7000 23 000 70 000 230000

IC,, >

binder

T.H.M. Ottenhqfet

al. / Jounurl

of Immunoiogicnl

95

Methods 200 (19Y71 89-97

Table 5 Binding

of

KVFPC(FI)ALINK

KTIAYDEEAR Concentration

(HSP65 3/12, KVFF’YALINK

to HLA-A’0301. HLA-DRl7

inhibition

by HLA-A”

0301

binder

KVFPYALINK

and HLA-A’0301

(nM) c/r binding of KVFPC(FI)ALINK

Concentration

KTIAYDEEAR

(nM) % binding of KVFPC(FI)ALINK

0

19

0

7

18.3

7

nt

70

10.1

70

nt

700

4.2

700

20

7 000

2.9

IC,,

non-binder

binder)

= 85 nM

by differences in on-rates of the labelled peptides related to off-rates of the complexed peptides. Complete inhibition of the interaction between fluorescent peptide 1 and HLA-A” 0201 was reached using a lOO-fold excess of unlabelled peptide (700 nM). whereas complete inhibition of binding of fluorescent peptide 3 to HLA-A” 0301 could not be attained using a lOOO-fold excess of unlabelled peptide. The IC,, value of peptide 2 for binding to HLAA* 0201 was determined to be 440 nM, that of peptide 4 for binding to HLA-A* 0301 was 85 nM (see Tables 4 and 5 respectively).

4. Discussion In this paper we describe a novel, sensitive, quantitative and automated fluorescence based peptide/MHC binding assay for both HLA class I and class II molecules. The assay comprises a number of advantages compared to existing assays. The most important features are the excellent reproducibility and the high sensitivity. Good reproducibility is a key parameter beacause this type of assay is normally used to screen extensive sets of peptides for their affinity to HLA and crucial information is derived from comparisons of various peptides. The fact that binding of lo-20 fmol of fluorescent peptide already yields significant and reproducible results (signal to noise ratio > 25) suggests that relatively small quantities of HLA preparations

19

7000

18.6

70 000

19.2

700 000

18.3

IC,,, >

I

mM

suffice in the assay. As preparations of this kind are hard to obtain in large quantities. the high sensitivity of the assay is important from a practical point of view. In addition the assay is relatively quick: although the samples are measured sequentially, analysis time is short (lo- 12 min) which assures a reasonable through-put. Furthermore, since sample analysis has been automated completely, continuous data acquisition is possible. It is worthwhile to note that the assay is non-radioactive and can thus be performed in virtually any laboratory without the risk of radioactive contamination. The assay requires investment in an HPLC with autosampler, a highly sensitive fluorescence monitor and a suitable HPLC size-exclusion column. Furthermore, although the synthesis of the proper fluorescein labelled standard peptides is straightforward, our laboratory is willing to make these labelled peptides described available to potential users at cost price. The assay should be useful when monitoring affinities of peptides for HLA class I and II alleles. Binding of peptides with lengths ranging from 9 to 25 amino acids can be measured. Using the assay we have determined peptide affinities for various HLA class II alleles (DRl , 2, 17, 4, 7) and for HLA class I alleles (HLA-A” 0301 and HLA-A’ 0301). Class I alleles appeared to be more susceptible to the percentage acetonitrile in the assay buffer. We routinely use 0% acetonitrile in class I assays and 15% acetonitrile in class 11 assays. Peptide binding to HLA-

96

T.H.M.

Ottenhoff

et al. /Journal

qf Immunological

DR17 was tested at various pHs and for some peptides the binding depended on the pH of the incubation buffer, but not on the pH of the elution buffer (Geluk et al., 1995). In the case of HLA alleles which have not been previously tested, the conditions for the incubations and the composition of the elution buffer may have to be optimized. In addition it is clear that the presence of 1 mM CHAPS in the elution buffer not only reduces non-specific interactions of the HLA molecules with the column, but also increases the fluorescence signal of the fluorescent label about three-fold. In addition to HLA-peptide binding, the assay may also be useful when studying other kinds of binding, for example antigen-antibody screening.

Acknowledgements We thank Prof. Dr. C.J.M. Melief and Dr. F. Koning for critically reading the manuscript. We thank Dr. J.E. Coligan and Dr. K. Parker for kindly providing HLA-A2 and HLA-A3 constructs. The authors’ work receives financial support from the Dutch Leprosy Relief Association (NSL), WHO, CEC, the Netherlands Organisation for Scientific Research (NWO), the Royal Nederlands Academy of Arts and Sciences (KNAW), the Macropa Foundation and the University of Leiden.

References Allen, P.M. (1994) Peptides in positive and negative selection: a delicate balance. Cell 76, 593. Arya, SK. and Gallo, R.C. (1986) Three novel genes of human T-lymphotropic virus type III: immune reactivity of their products with sera from acquired immune deficiency syndrome patients. Proc. Natl. Acad. Sci. USA 83, 2209. Buus, S., Sette, A., Colon, S.M., Jenis. D.M. and Grey H.M. (1986) Isolation and characterization of antigen-Ia complexes involved in T cell recognition. Cell 47, 1071. Coste, J., Dufour, M.-N.. Nguyen, D. and Castro, B. (1990) BOP and congenes: present status and new developments. In: J.E. Rivier and G.R. Marshall (Eds.), Peptides: Chemistry, Structure and Biology. Escom, Leiden, p, 885. Den Haan, J.M.M., Sherman, N.E., Blokland, E., Huczko, E.. Koning, F.. Drijfhout, J.W., Skipper, .I.. Shabanowitz, J., Hunt, D.F., Engelhard, V.H. and Goulmy, E. (1995) Identifica-

Methods

200 C1997) 89-97

tion of a graft versus host disease-associated human minor histocompatibility antigen. Science 268, 1476. Drijfhout, J.W., Bloemhoff. W.. Poolman, J.T. and Hoogerhout, P. (1990) Solid-phase synthesis and applications of N-(Sacetylmercaptoacetyl)peptides. Anal. Biochem. 37. 349. Fields, G.B. and Noble. R.L. (1990) Solid phase peptide synthesis utilizing 9-fluorenylmethoxycarbonyl amino acids. Int. J. Peptide Protein Res. 35. 161. Garboczi, D.N.. Hung. D.T. and Wiley, D.C. (1992) HLA-A2peptide complexes: Refolding and crystallization of molecules expressed in Escherichia roli and complexed with single antigenic peptides. Proc. Nat]. Acad. Sci. USA 89. 3429. Gausepohl. H.. Kraft, M.. Boulin, C. and Frank, R.W. (1990a) A multiple reaction system for automated simultaneous peptide synthesis. In: E. Giralt and D. Andreu (Eds.). Peptides 1990. Escom. Leiden, p. 206-207 Gausepohl. H.. Kraft. M.. Boulin. Ch. and Frank, R.W. (1990b) Automated multiple peptide synthesis with BOP activation. In: J.E. Rivier and G.R. Marshall (Eds.). Peptides: Chemistry, Structure and Biology. Escom. Leiden. p. 1003. Geluk, A.. Bloemhoff, W.. de Vries, R.R.P. and Ottenhoff, T.H.M. (1993-j Binding of a major T cell epitope of mycobacteria to a specific pocket within HLA-DRw 17(DR3) molecules. Eur. J. Immunol. 22. 107. Geluk. A., van Meijgaarden, K.E., Southwood, S., Oseroff. C., Drijfhout, J.W., de Vries. R.R.P.. Ottenhoff, T.H.M. and Sette. A. (1994) HLA-DR3 molecules can bind peptides carrying two alternative specific motifs. J. Immunol. 152, 5742. Geluk. A.. van Meijgaarden, K.E., Drijfhout. J.W. and Ottenhoff, T.H.M. (1995) CLIP binds to HLA class II using methionine based, allele dependent motifs as well as allele independent supermotifs. Mol. Immunol. 32. 975. Germain, R.N. (1995) MHC-associated antigen processing, presentation, and recognition. Adolescence, maturity and beyond. Immunologist 3. 5. Gorga, J.C.. Horejsi. V.. Johnson. D.R., Raghupathy. R. and Strominger, J.L. (1987) Purification and characterization of class II histocompatibility antigens from a homozygous human B cell line. J. Biol. Chem. 263, 16087. Jensen. P.E. (1991) Enhanced binding of peptide antigen to purified class II major histocompatibility glycoproteins at acidic pH. J. Exp. Med. 174, 1 I I I. Kast. M.W.. Brandt, R.M.P.. Sydney, J., Drijfhout, J.W., Kubo. R.T.. Grcy, H.M., Melief, C.J.M. and Sette. A. (1994) Role of HLA-A motifs in identification of potential CTL epitopes in human papillomavirus type 16 E6 and E7 proteins. J. Immunol. 152. 3904. Naumann. H.. Schaefer. S.. Yoshida. C.F.T., Gaspar, A.M.C., Repp. R. and Gerlich, W.H. (1993) Identification of a new hepatitis B virus (HBV) genotype from Brazil that expresses HBV surface antigen subtype adw4. J. Gen. Viral. 74, 1627. O’Sullivan, D., Arrhenius. T., Sidney, J.. Del Guercio. M.-F., Albertson. Wall, M., Oseroff, C., Southwood, S., Colon, S.M., Gaeta. F.C.A. and Sette. A. (1990) On the interaction of promiscuous antigenic peptides with different DR alleles. Identification of common structural motifs. J. Immunol. 147. 2663.

T.H.M.

OttenhofJ’et

ul. / Journal

of Immunological

Panina-Bordignon, P.. Tan. A., Termijtelen. A., Demotz, S., Corradin. G. and Lanzavecchia, A. (19891 Universally immunogenic T cell epitopes: promiscuous binding to human MHC class II and promiscuous recognition by T cells. Eur. J. Immunol. 19. 2237. Pearson, D.A., Blanchette, M.. Baker. M.L. and Guindon, C.A. ( 1989) Trialkylsilanes as scavengers for the trifluaroacetic acid deblocking of protecting groups in peptide synthesis. Tet. Lett. 30. 2739. Rapp, W.. Zhang, L. and Bayer, E. (19901 Continuous flow peptide synthesis on PSPOE-graft copolymers. In: R. Epton (Ed.), Innovation and Perspectives in Solid Phase Peptide Synthesis. SPCC (UK). Birmingham. p. 205. Rothbard, J.B., Busch, R.. Howland. K., Bal. V.. Fenton, C., Taylor, W.R. and Lamb, J.R. (1989) Structural analysis of a peptide-HLA class II complex: identification of critical interactions for its formation and recognition by T cell receptor. Int. Immunol. I. 479. Seedorf. K.. Krammer. Cl.. Durst. M.. Suhai. S. and Rowekamp.

Methods

200 (1997) 89-97

97

W.G. (19851 Human papillomavirus type 16 DNA sequence. Virology 14.5. 181. Sette, A.. Alexander, J., Ruppert, J.. Snoke. K., France, A.. Ishioka. G. and Grey, H.M. (1994al Antigen analogs/MHC complexes as specific T cell receptor antagonists. Annu. Rev. Immunol. 12. 413. Sette, A.. Sidney, J., de1 Guercio. M.F., Southwood, S., Ruppert, J.. Dahlberg. C.. Grey, H.M. and Kubo, R.T. (I 994b) Peptide binding to the most frequent HLA-A class I alleles measured by quantitative molecular binding assays. Mol. Immunol. 31. 813. Sheppard, R.C. and Williams. B.J. (1982) Acid-labile resin linkage agents for use in solid phase peptide synthesis. Int. J. Peptide Protein Res. 20. 45 1. Stern. L.J., Brown. J.H., Jardetzky. T.S.. Gorga, J.C., Urban, R.G., Strominger, J.L. and Wiley, D.C. (1994) Crystal structure of the human class II MHC protein HLA-DRl complexed with an influenza virus peptide. Nature 368. 215.