Antibodies to trypsinogen activation peptides recognize both Ca2+ ...

625th MEETING, LONDON and/or due to changes beyond neuromuscular junction, i.e. changes within the muscle (Merton, 1954; Bigland-Ritchie et a/. 1978; Dawson e t a / . 1980; Edwards, 1981). Bigland-Ritchie, B., Jones, D. A,, Hosking, G. P. & Edwards, R. H. T. (1978) Clin. Sci. 54,609-6 14 Bradshaw, E. G. & Maddison, S. (1979) Rr. J . Anaesrh.S1,955-960 Dawson, M. J., Gadian, D. G. & Wilkie, D. R. (1980) J. I’hysiol. (London) 299,465-484

337 Edwards, R. H. T. (1981) in Human Muscle Fatigue: I’hysiolo@caical Mechanisms (Porter, R. & Whelan, J., eds.1, PP. 1-18, Ciba b u n dation Symposium no. 82, Pitman Medical, London Merton, P. A. (1954)J. Physiol. (London) 123, 553-564 Payne, J. P. & Hughes, R. ( 1 981) Br. J. Anaesfh. 53,45-54

Received 20 October 1987

Antibodies to trypsinogen activation peptides recognize both Ca2+-dependent and Ca2 independent epitopes +

PAUL R. HURLEY, ALfSTAlR J. COOK, BRIAN M. AUSTEN and JOHN HERMON-TAYLOR Department of Surgery, St George S Hospital Medical School, Cranmer Terrace, Tooting, London SW17 ORE, U.K. The activation peptides of the pancreatic zymogen trypsinogen contain the sequence Asp-Asp-Asp-Asp-Lys [(Asp),Lys], which has been highly conserved in the course of vertebrate evolution (De Haen et al., 1975). Recognition and cleavage of trypsinogen at the C-terminal Lys residue of these peptides is normally catalysed in the proximal small intestine by small gut enteropeptidase ( E C 3.4.21.9) during digestion. To investigate the extent of release of the activation peptides into the bloodstream by inappropriate activation of pancreatic trypsinogen in the course of some pathological states, such as acute pancreatitis, we have developed immunoassays to the activation peptides (Hurley et al., 19aS). Methods of chemical synthesis of the peptides and haptenization were adopted to give antibodies recognizing only those forms of (Asp),-Lys which possessed a free Cterminal Lys residue, and thus did not bind the precursor trypsinogen. In the course of development of the assay we discovered that the haptenized peptide stimulated the production of two populations of antibodies, one dependent upon Ca?+for binding to the peptide, and one not. The Cys thiol of the synthetic peptide Cys-Ala-Pro-Phe(Asp),-Lys was coupled to bovine thyroglobulin via the heterobifunctional linker m-maleimidobenzoyl-N-hydroxisuccinimide (Green et al., 1982). Rabbits were immunized with 1 ml of of this hapten ( 1 mg/ml) plus 1 ml of complete Freund’s adjuvant, and boosted every month, using incomplete Freund’s adjuvant. After 20 weeks, blood (16 ml) was taken and immunoglobulins precipitated by addition of an equal volume of saturated ammonium sulphate at 4”C, collected by centrifugation, and dialysed against 50 mM-Tris/ HCI, 150 mM-NaCI, 20 mM-CaC12 (pH 7.4) (TSC buffer). The dialysate was applied to a column of Tyr-(Asp),-Lys coupled to activated CH-Sepharose (Pharmacia Ltd) at a substitution of 1.16 pmol per ml of gel, and eluted with TSC buffer at 24 ml/h until the absorbance at 280 nm of the eluate was zero. The affinity column was then eluted with 50 mM-Tris/HCI, 150 mM-NaC1, 20 mM-EDTA (pH 7.4) which eluted one population of antibodies (3.3 mg protein in 11 mi), followed by 1 M-propionic acid, which eluted a second population of antibodies (9 mg in 37 ml) which were immediately neutralized with 1 M-Tris after elution from the column. Antibodies present in the eluate were assayed by a solidphase immunoradiometric assay in which polyvinylchloride microtitre plates, were coated overnight with 50 pl of Tyr(Asp),-Lys coupled to bovine serum albumin by benzidine hydrochloride (Bassiri & Utiger, 1972) at 50 pg/ml either in 50 mM-Tris/HCI (pH 7.4), 0.l0/, (w/v) sodium azide (TCA Abbreviations used: TSC, TCA, PBE, TCAHS, PBEHS, buffers, for the composition of which see text.

Vol. 16

buffer) or in phosphate-buffered saline containing EDTA (PBE buffer), and then blocked for excess protein-binding sites and washed with TCA buffer containing 10% (v/v) horse serum (TCAHS buffer) or with PBE-buffer containing 10% (v/v) horse serum (PBEHS-buffer). Wells were then incubated with eluted antibodies in either TCAHS buffer or PBEHS buffer, either in the presence or absence of lo-’ M(Asp),-Lys to correct for non-specific binding. Bound antibodies were detected by adding 50 000 c.p.m. of 12sI-labelled affinity-purified goat (anti-rabbit Ig) antibodies. No counts per minute in the assay were displaced by the peptide when the antibodies eluted by EDTA from the affinity column were assayed in the presence of EDTA, but up to 33% of added goat (anti-rabbit) was bound when these antibodies had first been dialysed against TCA buffer containing 20 mM-CaCl,, and re-assayed in the presence of 20 mM-CaC1, (Fig. 1). Other divalent metal ions Mg2+,Z n 2 + , Ba2+, Hg2+ and Cd2+ did not substitute for Ca2+. In

40

r

2

10

25

50

100

Dilution of antibodies (by vol.)

Fig. 1. Competitive immunoradiometric assay in the presence and absence of Ca’+ of antibodies afinity-purified from immobilized (Asp),-Lys

Shown is the percentage of IZSI-goat(anti-rabbit Ig) antibodies displaced by excess (Asp),-Lys from wells coated with (Asp),-Lys linked to BSA after incubation with (-) EDTA-eluted antibodies or ( A-A ) propionicacid-eluted antibodies assayed in the presence of Ca”, and (M EDTA-eluted ) antibodies or (-) propnicacid-eluted antibodies assayed in the absence of Ca and presence of EDTA.

338

BIOCHEMICAL SOCIETY TRANSACTIONS

contrast, antibodies not eluted from the affinity column by EDTA, but eluted later with 1 M-propionic acid, gave up to 36% binding either in the presence of Ca‘+, or absence of Ca” and presence of 20 mM-EDTA (Fig. 1). The presence of light and heavy chains in both preparations was confirmed by SDS/polyacrylamide-gel electrophoresis. The (Asp), sequence in trypsinogen activation peptides has been shown to bind Ca” (Delaage & Lazdunski, l967), and dissociation constants of 7 x 1 0 - 3 M have been reported (Cliffe & Grant, 198 1). The demonstration that antibodies raised to haptenized I ( Asp),-Lys contain populations that bind to the peptide only in the presence of Ca?+,while other populations bind in both the presence and absence of Ca?+ indicate that two distinct epitopes are present. One of these may be a chelate in which two or more of the side-chain carboxyls chelate to one o r two Ca2+ions, while the other epitope may consist of the peptide without Ca?+. The

presence of Ca?+ does not interfere with recognition o f the second epitope. We gratefully acknowledge the support o f the M.R.C Bassiri, R. M. & Utiger, R. D. ( 1972) Endocrino/ogy90, 722-727 Cliffe. S. G. R. & Grant, D. A. W.( 1981) Riochem. J. 193,655-658 De Haen, C., Neurath, H. & Teller, I). C. ( 1 975) J. Mol. Niol. 92. 225-259 Delaage, M. & Lazdunski, M. ( 1967) Biocht,m. Hiophys. Kes. Commun. 28, 390-394 Green, N., Alexander, H., Olson, A.. Alexander. S.. Shinnick. T., Sutcliffe. J. & Lerner, R. ( 1 982) C ’ d 28.477-484 Hurley, P. R., Cook, A. J., Austen, B. M. & Hermon-Taylor, J. ( 1988)J. Immitnol. Meihods in the press Received 7 December I987

Allergenicity of the major genetic variants of bovine /hctoglobulin ZAHIDA MALIK,* ROBIN BOTTOMLEYt and BRlAN AUSTEN* *Department of Surgery, St George’s Hospital Medical School, Cranmer Terrace, London S W1?/ORE, U.K. and t Research arid Development, Express Foods Group, Victoria Road, South Ruislip, Middlesex HA4 OHF, U.K. ‘The reported incidence of cows’ milk allergy (CMA) in young infants up to one year old has been reported to range from 0.05 to 3% and its occurrence probably depends on genetic factors, degree of breast feeding in the infant population, and the type of milk formulae used. Infants, or adults, can present with the following unpleasant symptoms in any combination, diarrhoea, vomitting, rhinorrhoea, recurrent croup, bronchitis, wheezing, eczema, colic, fretfulness and failure to thrive. Wholly breast fed babies can also present with the classic symptoms of CMA and these are generally thought to be due to extrinsic ingredients in the mother’s milk (Atherton, 1983). Cows’ milk contains more than 25 distinct proteins which in themselves contain potential triggers of allergy. The three most abundant proteins in cows’ milk are casein, Plactoglobulin (p-LG) and alpha-lactalbumin, of which, P-LG has been found to be the most frequently involved in infant CMA (Fallstrom et al., 1978). Six genetic variants of P-LG have been identified, of which the two most abundant of these are the A and B variants (Aschaffenburg & Drewry, 1957). Preliminary studies on the nature of the allergenic determinants in B-LG mixtures of the genetic variants have been reported (Haddad et al., 1979). The purpose of this study was to test the allergenicity of the B-LG A and B variants separately in a rat model, to understand more about the allergenicity of these proteins, and also to see if the hypoallergenic milk formula based on one of the variants could be produced. Groups of Wistar rats ( 3 - 5 animals/group) weighing approximately 150 g were immunized i.p. with 100 p g amounts of either pure B-LG A or @-LGB in the presence of Al(OH), gels or Freund’s complete adjuvant or Bordetella pertussis vaccine. Blood samples were taken from the tail vein before immunization and then at weekly intervals for a further three weeks. The serum anti-B-LG specific IgE antibody levels were assayed by ELISA (enzyme-linked Abbreviations used: CMA, cows’ milk allergy; B-LG, Plactoglobulin; ELISA, enzyme-linked immunosorbent assay; 5-HT, 5-hydroxytryptamine.

immunosorbent assay) (Coleman et id., 1983; Bos et ul., 1981) against antigen (pure P-LG A or B at SO pg/ml, 50 pl/well)-coated wells. Concentrations of P-LG specific IgE (immunoglobulin E ) were expressed as net absorbance after subtracting readings obtained with wells not coated with PLG. There was no P-LG A or B specific IgE in the sera of immunized rats on day 7, but circulating anti-LG antibodies were detected at day 14, for rats immunized to P-LG A and B, and higher levels were detected at day 2 1 and day 28. At day 35, IgE levels had receded to those estimated at day 14. Three adjuvants were tested, and IgE levels in responding animals to P-LG A in the presence of Al(OH), gels or Bordetella pertussis were similar, and 2.5-fold higher than Freund’s complete adjuvant. With P-LG B, the overall IgE levels were 5-fold greater with AI(OH), gels than Freund’s complete adjuvant and 2-fold greater than Bordetellu pertussis. Hooded Lister rats were immunized i.p. with differing amounts ( 100 pg, 10 p g and 1 pg) of either P-LG A or B and blood samples obtained and assayed by ELISA as described previously. Both genetic variants of B-LG promoted a transient IgE response and circulating levels of the antibody were detected at day 14 and day 21 in rats sensitized to /3-LG A and B, with 100 pg, 50 p g and 10 p g amounts. However, IgE titres were higher in rats sensitized i.p. with #?-LG B than those sensitized to B-LG A, at 100 pg, 5 0 pg and 10 pg concentrations of the sensitizing protein (Fig. 1). For each variant of B-LG, eight rats were immunized; five rats responded to P-LG A and seven rats responded to P-LG B. IgE levels were 5-fold greater in animals sensitized to the B variant. N o rats immunized with l p g amounts of either P-LG variant exhibited an antibody response. Animals from the previous experiment were killed on day 21 and the release of [3H]5-hydroxytryptamine([3H]S-HT) from preloaded peritoneal mast cells was measured (Coleman et al., 1981), see Table 1. Release of [jHH]S-HT was stimulated by challenge to mast cells obtained only from those rats sensitized with 100 p g amounts of B-LG, the response to 10 p~ of challenging protein being optimal in each case. Cells from rats sensitized with smaller amounts of B-LG did not release significant amounts of [3H]5-HT.N o difference could be discerned in the response of cells from animals sensitized to B-LG A compared to those sensitized to B-LG B. Both the A and B variants of P-LG may thus be presented to rats in such a way that raised transient levels of circulating, specific IgE, result, P-LG B giving somewhat higher levels 1988