Studies on the immunological cross-reactivity of various pancreatic colipases

Biochimica et Bwphysica Acta, 742 (1983) 39-48 Elsevier Biomedical Press

39

BBA 31418

STUDIES ON T H E I M M U N O L O G I C A L CROSS-REACTIVITY OF VARIOUS PANCREATIC COLIPASES I S O L A T I O N BY IMMUNOAFFINITY C H R O M A T O G R A P H Y OF A S I N G L E FORM O F P R O C O L I P A S E FROM PORCINE PANCREAS J. R A T H E L O T a, p. DELORI b and L. S A R D A a " Laboratoire de Biochimie, Facult~ St-Charles, Place V. Hugo, 13003 Marseille and h Laboratoire de Biochimie, Facult~ de Mbdecine Nord, Boulevard P. Dramard, 13015 Marseille (France) (Received July 19th, 1982)

Key words: Immunological cross-reactivity," Colipase precursor," (Pancreas)

Antibodies against porcine procolipase B were produced in rabbits. The antiserum was used to immunoinactivate various forms of native and trypsin-treated porcine colipase. Our results indicate that all forms of the porcine cofactor bind to anti-porcine procolipase B antibodies. Human colipase showed lower affinity for the antibodies than porcine colipase. No cross-reactivity was observed between pig and horse, cow, dog or chicken colipases. Immunological studies on porcine colipase, carried out in the presence of lipid, provided evidence that antibodies bind to colipase at or near the lipase binding site. The binding of antibodies to colipase is not affected by the adsorption of the cofactor at a lipid interface. Using a predictive method for identification of the antigenic determinants, it was found that, in pig colipase, regions at positions 42-48 and 70-74 might represent antigenic sites. In the horse protein, the peptide segment 42-48 was also recognized as a possible antigenic site. An immunoadsorbent gel column was prepared for a one-step isolation of porcine colipase. In contrast to purification methods described so far, immunoaffinity chromatography yielded only one form of the porcine cofactor when starting from a pancreatic extract. This protein preparation has structural, biochemical and immunochemical properties similar to that of porcine procolipase A previously isolated from pancreas in the presence of detergent.

Introduction Inhibition of pancreatic lipase by bile salt is specifically reversed by colipase, a polypeptide of low molecular weight also found in the pancreatic secretion [1,2]. Colipase acts as a cofactor by anchoring lipase to emulsified lipid substrates [3]. It also protects lipase against surface denaturation [4-6]. There is evidence showing that the colipaselipase binding occurs at the interface in a i to ! molar ratio with a dissociation constant as low as 10 - 9 M [7,8]. Pancreatic colipase has been purified from various sources but, to date, most of the 0167-4838/83/0000-0000/$03.00 © 1983 Elsevier Biomedical Press

structural and functional investigations have been carried out on the porcine and the equine proteins [9]. Several forms of active colipase have been isolated from pig pancreas or pancreatic juice [10,11]. Evidence was obtained showing that the occurrence of multiple forms of porcine colipase could result from partial proteolysis at both ends of the polypeptide chain. Purification of colipase in the presence of a non-ionic detergent, Triton X-100, allowed us to prepare two forms of porcine colipase with the same specific activity, namely colipases A and B, having undergone no proteolytic

40 degradation in the amino-terminal region [12]. Removal of the N-terminal pentapeptide by specific tryptic cleavage at the Arg-5-Gly-6 bond yielded activated colipase with better lipid-binding properties [13,14]. Trypsinolysis is accompanied by a discrete local conformational change in the region of the cofactor molecule recognized as the lipid binding site [15]. According to BorgstrOm et al. [13] removal of the N-terminal pentapeptide upon limited tryptic proteolysis is of physiological significance and corresponds to the activation of a precursor form of the cofactor (procolipase). Sequence analysis in the C-terminal region of porcine colipases indicates differences between the two forms of the cofactor. The carboxy-terminal sequence of colipase B (93 residues) prepared in the presence of Triton X-100, was found to be Arg-Ser. An additional dipeptide (Asp-Ser) was found in colipase A obtained by the same method [15]. This finding is in accordance with the results of Chapus et al. [16], who purified two forms of pig colipase by a common procedure using trypsin and carboxypeptidase inhibitors. Larsson and Erlanson-Albertsson [17] have reported the isolation of two forms of colipase containing 101 residues whose C-terminal sequence was recently identified as Leu-Ser. Horse colipase has been shown to exist in two forms which were isolated in different conditions from pancreas homogenate. From partial sequence analysis it was found that these polypeptides are isocolipases differing by several amino acids [18]. This conclusion was confirmed later by comparing the entire sequences of the two proteins [19. 20]. Much attention has been given to the species specificity of the lipase-colipase interaction. As previously demonstrated, the pancreatic cofactor does not activate bile-salt-inhibited fungal lipases [4]. Studies on the activation of pancreatic lipase from one species by colipase from another species have shown no specificity for the cofactor [8,21]. This result indicates that the specific site on the surface of the cofactor, where the pancreatic lipase-colipase interaction takes place, should have identical characteristics within animal species to ensure mutual recognition and binding. Here we report immunological studies on coiipase. Cross-reactivity of an antiserum raised against porcine colipase B was assayed with several

forms of native and trypsin-treated porcine colipases and with cofactors from other species. All forms of porcine colipase show similar antigenic properties, while differences are found among other colipases. In the second part of this communication we report the isolation by immunoaffinity chromatography of porcine colipase from a pancreatic extract. A single form of porcine procoiipase has been obtained by this method and characterized. Materials and Methods

Chemicals. CNBr-activated Sepharose 4B was obtained from Pharmacia (Uppsala, Sweden). Materials for polyacrylamide gel electrophoresis were from BDH Chemicals Ltd (Poole, U.K.). Sodium deoxycholate and benzamidine were from Fluka (Lucerne, Switzerland). Phenylmethylsulfonyl fluoride, o-phenanthroline and bovine serum albumin were from Sigma (St. Louis, U.S.A.). Carboxypeptidases A and B were obtained from Boehringer (Mannheim, F.R.G.). Other chemicals used were of reagent grade. All solutions were made with deionized quartz-redistilled water. Colipase samples. Pure colipase was prepared in the laboratory from pig, cow, horse and chicken pancreas homogenates according to previously described methods [12,22-24]. Porcine colipases A and B (spec. act. 20000 units, m g - i ) were treated with trypsin to split off the N-terminal pentapeptide as recently reported [15]. All forms of trypsin-treated porcine colipase (Atl, A t . and Bt) had an N-terminal glycine and different C-terminal sequences. The carboxy-terminal peptide sequences, Arg-92-Ser-Asp-Ser, Arg-92 and Arg-92Ser were found in colipase At I, At N and Bt, respectively. Experiments were performed using horse colipase B and bovine colipase A. Crude samples of human and canine colipases were obtained from pancreatic juice (a generous gift from Dr. O. Guy, Marseille) by acidification to pH 3. Insoluble material was removed by centrifugation. Assay of colipase activity. Colipase was assayed at 25°C and pH 9 with the potentiometric method using emulsified triolein in the presence of 18 mM sodium deoxycholate [25]. Pure horse lipase [8] was used in the assay. One colipase unit corre-

41 sponds to the liberation of 1 #equiv. fatty acid under the standard conditions. Immunization. Antigens used for immunization were either porcine colipase B or its trypsin-treated form, Bt. Female rabbits (Blanc de Bouscat strain, Evic Ceba, 33290 Blanquefort, France) weighing 2.5-3 kg were injected with 3.8 mg protein per animal. Three to four animals were used for each antigen. Before injection, antigens were dissolved in 0.5 ml of 150 mM NaCi and emulsified with an equal volume of complete Freund's adjuvant (M~rieux 69260 Marcy l'Etoile, France). Injections were given first intradermally at multiple sites on the shaved back skin at days I, 8, 14 and, later, subcutaneously at days 36, 39, 43, 60, 62, 68. At day 79, all rabbits were bled by cardiac puncture and sera were stored at - 3 0 ° C or freeze-dried. The preimmune sera and 53-day punctures were also collected. Serum was reconstituted by dissolving 80 mg freeze-dried powder in 1 ml bi-distilled water and filtered through Millipore filters (0.45 ~m pores). Immune serum against colipase B from the rabbit showing the best neutralizing titre was used in all experiments. Ouchterlony immunodiffusion analysis. The double immunodiffusion assay was performed according to Ouchterlony [26] using 2% Agarose in 0.15 M NaCI, pH 7. Colipase samples (5 mg/ml) and antiporcine colipase B antiserum were applied, and diffusion was allowed to proceed for 48 h at 20°C. Gels were stained with Coomassie brilliant blue and destained.

Quantitative precipitation of pure porcine colipase B. Increasing amounts of colipase up to 4 nmol were added to 0.15 ml of specific immune serum. The final volume was adjusted to 450 pl with 50 mM Tris-HCl buffer, pH 7.4, containing 150 mM NaCi and bovine serum albumin at a concentration of 1 mg per mi. The mixtures were incubated for I h at 37°C and kept at 0°C for 14 h. The immunoprecipitates were collected by centrifugation (9000 × g 5 min), washed twice with 450 ~1 of 150 mM NaCI, and dissociated in 1 ml of 0.1 M NaOH. Colipase activity was assayed immediately and the protein content of the solubilized precipitates was measured spectrophotometrically at 280 nm and by the colorimetric method of Lowry et al. [27]. Non-precipitated colipase was determined after elimination of the immunopre-

cipitates in the supernatants. Supernatants were then acidified to pH 3 with CH3COOH to dissociate possible soluble complexes.

Immunoinactivation of colipase by antiporcine colipase B antiserum. To test the specificity and the cross-reactivity of the porcine antiserum for the various colipases, aliquots (50 pl) of 8 #M colipase solutions in 50 mM Tris-HCl buffer, pH 7.4, (containing 150 mM NaCI and bovine serum albumin at a concentration of i mg per ml) were incubated with increasing volumes (0 to 100 ~1) of serum during 1 h at 37°C followed by 3 h at 0°C. The immunoprecipitates were centrifuged and supernatants were immediately assayed for residual colipase activity. Immunoactivation of colipase was also studied in the presence of lipid substrate under typical cofactor assay conditions [25]. Increasing amounts of porcine colipase B antiserum (5 to 200 ~1) were added to the assay system which contained 25 pmol porcine colipase. After 5 min incubation, 0.1 nmol pure horse lipase was added and activity was measured.

Antigenic determinant prediction by hydrophilicity analysis. The sequence of residues of porcine colipase B and equine colipase B used for antigenic determinant prediction is presented in Table I. Analysis of the sequence of porcine colipase B and equine colipase B was carried out according to the predictive method proposed by Hopp and Woods [30] to locate the antigenic determinants of proteins. This method is based upon the study of 12 proteins of known sequence for which immunochemical analysis has been carried out. Average hydrophilicities of hexapeptide sequences were computed over the length of the colipase polypeptide chains using the value of relative hydrophilicity assigned to each amino acid by Hopp and Woods. According to them, the point of highest local average hydrophilicity lies invariably in or very close to one of the protein's antigenic determinants. Purification of specific antibodies. Porcine colipase (6 mg) was covalently coupled to 1.6 g CNBr-activated Sepharose as recommended by the manufacturer [31]. The gel was then poured in a 5 ml bed column volume. 15 ml of antiporcine colipase B serum were submitted to chromatography over this column at a rate of 5 m i / h at 4°C.

42 TABLE I A M I N O A C I D S E Q U E N C E OF EQU1NE COLIPASE B A N D P O R C I N E COLIPASE S U B M I T T E D TO A N T I G E N I C PREDICTIONS The sequence of horse colipase B is taken from Chapus et al. [19]. The sequence of porcine colipase was established by Charles et al. [28] and by Erlanson et al. [29]. The C-terminal sequence reported here is that corresponding to colipase B. Porcine colipase A has an additionnal dipeptide (Asp-94-Ser-95) at the C-terminal end.

Horse B

5 10 15 20 25 Val -Pro-r, s p - P r o I f ~ r g - G l y - V a l - l i e - l le-Asn-L.e,~-,Glu-Ala-G1y-Glu- i le-Cys-Yet-A~n-S:_~r-A ' , a - G l n - C y s - L y s - S e r

Pig

Val-Pro-Asp-Pro-Arg-Gly- :le-ile-

Horse B

S(i 35 43 ;5 50 G~u~Cys~Cys-His~A~g-G~u~Se~Ser~Leu~Ser~Leu~A~a~Ar~Cys-A~a~A~a~Lys~A1a~Se~G]u~As~:~c~G~u~Cys~Ser

Pig

Asn-Cys-Cys-Gln-His-~so-Thr-ile-!eu-Ser-l.eu

Horse B

•";5 60 65 z~75 AI a- Trp-Thr-I..-_.u- iyr-Giy-Va'. - [ y r - T y r - L y s - C y s - P r o - C y s - G l u - A ~ g - G l y - L e u - rhr-C'is-Gl n-'. a l -,',s;)-', ys- ~hr-Leu

Pig

A!,~-Ph'_',-:hr-: c:;-T'..'r-S:L.-.',~:-Tyr-Tyr-Lys-Cys-Pro-Cys-Gl u-A~g-Gly-l.eu-, hr-,r..~ -i~',..,-Sly-: -,~-~ y s - S ~ " - I eu

Horse B

80 85 gO .; : V a l - G l y - S e r - l l e - H e L - . ' ~ s n - , ~ n r - A s n - P h e - G l y - I l e - C y s - : ; h e - A s p - A l a A!a-Arg.-Ser G'... S::,-.:.rg

Pig

Val - G i y - S e r - i]e-ih,'-,'~r~-:~ :'-Asn-Phe-Gly- l l e - C y s - H i s - A s n - V a l - G ] y - A r g - S e r

I le-Asn-Leu-Asp-G1u-G1y-Glu-i e u - C y s - l e u - . , ' ~ n - S e r - A l a - G l n - C y s - L y s - S e r

Elution of purified antibodies was performed under the same conditions as previously used in snake cardiotoxin studies [32]. 60 ml purified immunoglobulins were recovered as estimated by absorbance at 280 nm.

Immobilization of anticolipase antibodies on CNBr-activated Sepharose 4B. Purified antibodies (105 mg obtained from 25 ml serum) were concentrated to 35 ml (Diaflo x M50 membrane), dialysed against 0.1 M NaHCO 3, pH 8.3, containing 0.5 M NaCI, and coupled to 10 g CNBr-activated Sepharose 4B (yield: 90%). The immunoadsorbent gel column (I.6 x 20 cm) was then tested with pure porcine colipase B. Colipase (15 mg in 20 ml 0.1 M NaHCO 3, pH 7.45, containing 0.150 M NaCI) was passed through the column. Non-adsorbed colipase (8.8 mg) was eluted by passage of bicarbonate buffer. Elution of adsorbed colipase was performed with formic acid according to a previously described method [32]. Under these conditions, 5.8 mg of colipase were recovered in eluted fractions.

Isolation of porcine colipase by immunoaffinity chromatography. Porcine pancreas (125 g) collected at the slaughter-house was defatted and sliced in 360 ml ice-cold 0.25 M H2SO 4 containing 0.2% Triton X-100, 2 mM benzamidine, 0.15 M phenyimethylsuifonyl fluoride, 1 mM diisopropylphosphofluoridate and 2 mM o-phenanthroline. After

Ala-Arg-Cys-Ala-teu-',.ys-A:a-7,rc-G1u-f..sr.-S,:-r-Giu-Cys-Ser

30 min, the pancreatic tissue was homogenized in a Waring blender at +4°C. Homogeneization (30 s) was repeated four times. The homogenate was then adjusted to pH 2 with 5 M HzSO 4 and maintained under gentle stirring for 30 min. pH was then brought to 7.4 and insoluble material was removed by centrifugation (4°C; 60000 x g 2 h). A fraction (80 ml containing 400000 colipase units) of the supernatant was immediately percolated (5 m l / h ) through the immunoadsorbent gel column. Unretarded proteins were washed by running 300 ml of bicarbonate buffer through the column at a flow rate of 50 ml/h. This fraction contained excess colipase not adsorbed on the column. Adsorbed proteins were eluted and collected as described [32]. Pooled fractions containing colipase were dialyzed against distilled water (18 h) and freeze-dried. Analytical methods. Protein concentration was determined spectrophotometrically at 280 nm. Absorbance coefficients, EI¢,,, of pig, horse and chicken colipases and of immunoglobulins are 3.6, 10, and 13.5, respectively. The colorimetric method of Lowry et al. [27] was also used with bovine serum albumin as standard. Amino acid analysis and N-terminal residue determination were performed as previously described [18]. C-terminal residues were studied by following the rate of release of free amino acids

43

during digestion by carboxypeptidase. Digestions were performed with 1/20 (mol/mol) of carboxypeptidase A and with a 2:1 mixture of carboxypeptidases A and B. The procedure described by Ambler [33] was followed. Digestions were carried out in 0.2 M N-ethylmorpholine acetate at pH 8 and 37°C. Amino acids released after 30, 60, 120 and 180 min were analyzed on the amino acid analyzer (Biotronik LC 200). Polyacrylamide slab gel electrophoreses were performed at pH 8.4 according to Laemmli [34]. Results

Quantitative precipitation of porcine colipase B by anticolipase serum Fig. 1 shows typical precipitin curves obtained when colipase B was assayed with anticolipase B antiserum. One can observe that in the equivalence zone the precipitate obtained from 150 #1 antiserum contains 0.63 mg of specific antibodies for 26/zg antigen. The antiserum thus contains 4.2 mg

_t 5'

400

E /.

*

ii /

o51

200

if

I/

10

20

., I "~

30

410

~10

collpmse ],i.g Fig. 1. Quantitative precipitation of porcine colipase B by specific antibodies. Increasing amounts (4 to 40 t.tg) of pure porcine colipase B was incubated with 150 #1 homologous antiserum in a final volume of 450 #1. After incubation for 30 min at 37°C and 14 h at 4°C, the precipitate was separated by centrifugation and resuspended in 0.1 M NaOH. O . . . . . . O , colipase activity in the supernatant; A. . . . . . zx, colipase activity in the acidified supernatant; • @, colipase activity in the immunoprecipitate; • • , precipitated immunoglobulins as measured spectrophotometrically at 280 nm; • II, precipitated immunoglobulins determined by the colorimetric method of Lowry et al.

of specific antibodies per ml. In the precipitate, the molar ratio between antigen and antibodies is ! to 2, assuming an average molecular weight of 150000 for immunoglobulins. For a higher coiipase concentration, the existence of soluble antigen-antibody complexes explains why colipase activity could not be totally recovered by summing activity found in the supernatant and in the precipitate. Treatment of the supernatants at acidic pH dissociates these complexes and actually restores full colipase activity. Nonspecific serum has no effect on colipase precipitation.

Immunoinactivation of porcine colipase B by antiserum The activity of colipase B was determined under standard assay conditions in the presence of increasing amounts of specific antiserum. Fig. 2 shows that the addition of antiserum results in the progressive inhibition of colipase activity. Half inhibition of cofactor activity (total colipase concentration 0.7. 10 - 9 M) was obtained with about 20 #i antiserum, corresponding to a specific antibody concentration of 1.4.10-8 M. The inhibition curve presented in Fig. 2 allows the evaluation, by a Scatchard plot, of an apparent dissociation constant of the colipase-immunoglobulin complex (Ko = 2 . 3 . 1 0 -8 M). For this evaluation, it can be noticed that, under the experimental conditions used, free colipase is fully titrated in presence of an excess of lipase [25]. Under the same conditions, addition of increasing amounts (up to 1 ml) of nonspecific serum had no effect on colipase activity. In alternative experiments, the order of addition of antiserum and lipase was changed. When antiserum was added after the enzyme, no inhibition of colipase activity was found. The same general observations were made in parallel experiments carried out with tributyrin as substrate, in the absence or presence of deoxycholate.

Immunological cross-reactivity of various colipases Immunodiffusion studies. Double diffusion analyses were carried out on pure native and trypsinolysed forms of porcine colipases A and B by the Ouchterlony method. All forms of porcine colipase give precipitin lines of complete identity with those of specific antiserum raised against porcine colipase B. Under the same conditions,

44

.,.[

b

(b}

(G)

lo~

-~r--

r -o-

~--e-

-~i

- -- ~I~

!

i I

i

i

i

i 50

i

i

i

I

i I00

Fig. 2. (left-hand figure) Immunoinactivation of porcine pancreatic colipase B in the presence of substrate. Pure colipase (25 pmol) was mixed with emulsified triolein and added with specific or nonspecific antiserum. After 5 min incubation, pure lipase (150 pmol) was added (final concentration 5. | 0 - 9 M) and activity was determined. Experiments carried out with (0 0) specific and (zx ,',) nonspecific serum. The dotted line curve ( O . . . . . . ©) represents activity measured when antiserum is added to the assay system after colipase and lipase. Fig. 3. (a) Cross-reactivity of antiporcine colipase B antiserum. Increasing volumes of rabbit antiserum raised against porcine colipase B were added to equal amounts (5 #g) of various forms of porcine colipase. O ©, colipase A; ,', zx, colipase B; • O, colipase Atu; • • , colipase Bt. (b). Same experiments performed with colipases from various species. © . . . . . . O, bovine colipase A; t3- . . . . -O, equine colipase B; • . . . . . . A, chicken colipase; • . . . . . . O, canine colipase; • •, human colipase. The inactivation curve of porcine colipase B (zx ,',) is also reported for comparison. For experimental data. see Materials and Methods.

neither horse colipase B nor chicken colipase showed any reaction in the immunodiffusion test (results not shown). lmmunoinactivation studies. Samples of the different forms of porcine colipase and of bovine, equine, human and chicken colipases were treated with increasing amounts of antiporcine colipase B antiserum under the experimental conditions described in Materials and Methods. Following incubation and centrifugation, aliquots were removed for determination of residual colipase activity. Identical inactivation curves were obtained for all different forms of native and trypsin-treated porcine colipases (Fig. 3a). Fig. 3b shows that human colipase is inactivated but with a lower affinity for porcine colipase B antibodies than porcine colipase, while canine, bovine, equine and chicken colipases are not inactivated by this antiserum.

Prediction of antigenic determinants of porcine and equine colipases Hexapeptide hydrophilicity profiles are presented in Fig. 4. From the porcine colipase profile

! .,----

i

-~ ...... ~ ..... -I:, ..... ~___,L'~ ....... :~ ....

I

:',2~:~ ~

!~, II

:

i

J Jr,"

::

/i!l:"

",

l!ii ,l :/

l

l

I n I

i

I

n [

I I

n I I

I

t I

J I I

Fig. 4. Hexapeptide hydrophilicity profiles of porcine and equine colipases for the prediction of antigenic determinants. Average hydrophilicity values of hexapeptides were plotted versus position along the amino acid sequences. Data points are plotted at the center of the averaging group from which they are computed. - - . profile obtained for porcine colipase B; . . . . . . , profile obtained for equine colipase B.

45

it can be observed that the point of highest local hydrophilicity lies at positions 44.5 and 45.5, corresponding to the peptide segment at positions 42 to 48. The second and third highest points fall at positions 72.5 and 12.5, corresponding to peptide segments 70-75 and 10-15, respectively. It is worth noticing that the peptide sequence from residue 49 to residue 69, located between the two regions of highest hydrophilicity, contains the aromatic hydrophobic domain (residues 51-59) recognized as the specific lipid binding site of the colipase molecules [35,36]. Comparison of the sequence in the region 49-69 in pig and horse colipases actually indicates only one amino acid substitution, which is highly conservative (Phe-52 ~ Trp-52). In horse colipase, the peptide segment 42-48 is highly hydrophilic, as in the porcine protein. Large differences between the two proteins are found in regions corresponding to hexapeptides at position 24.5 to position 32.5, and to hexapeptides at position 70.5 to position 72.5. Actually, these hexapeptides correspond to highly substituted peptide segments at residues 26-33 and 70-74, respectively.

Isolation and chemical characterization of a single form of procolipase from porcine pancreas by immunoaffinity chromatography A typical affinity chromatography carried out on a porcine pancreatic extract is shown in Fig. 5. Colipase activity is associated with the single protein peak eluted from the column by passage of a solution of formic acid. The specific activity of the colipase obtained is 20000 u n i t s . m g - t a value which is identical to that previously found for porcine colipases A and B. The amount of colipase obtained from one chromatography is about 5.6 mg or 0.55 ~tmol colipase for 0.6 #mol purified anticolipase B antibodies immobilized on the column. Gel electrophoresis analysis of this fraction (Fig. 6) reveals the presence of one single protein constituent whose mobility is identical to that of porcine colipase A. No protein with the mobility of porcine colipase B was detected. Analysis by the dansylation method indicates only valine as Nterminal residue of the protein. Digestion for 3 h with carboxypeptidase A liberates 0.64 mol serine and 0.4 mol aspartic acid per mol colipase. Several other amino acids were liberated in trace amounts.

w

1

i

i

....... J 20

40

L 60

I 80 Fraction ~ m b e r

Fig. 5. Isolation of colipase by immunoaffinity chromatography. The supernatant (80 ml) from porcine pancreas homogenate containing 400000 colipase units was percolated through an immunoadsorbent gel column containing 90 mg of antibodies raised against porcine colipase B. The excess of colipase and nonadsorbed proteins were eluted from the column by passage of 300 ml 0.1 M NaHCOj, pH 7.45, containing 0.150 M NaC1. Adsorbed colipase (6 mg), corresponding to the saturation of the immunoadsorbent column, was eluted with formic acid (arrow). Volume of each fraction: 5 ml. • 0, colipase; • . . . . . . o, protein determined by the method of Lowry et al.

Digestion during the same incubation time with a mixture of carboxypeptidases A and B yields 2 mol serine, 0.6 mol aspartic acid and 0.6 mol arginine per mol colipase. Under the same conditions, digestion of porcine colipase A yielded 1.8 mol serine, 0.7 tool aspartic acid and 0.5 mol arginine per mol of a colipase A sample prepared by the detergent method [12]. The amino acid

®

a

b

c

d

•

f

Fig. 6. Polyacrylamide gel electrophoresis of porcine colipase: (a) bovine colipase A; (b) porcine colipase B; (c) porcine colipase A; (d) and (e) samples of porcine colipase obtained from the immunoaffinity chromatography column; (f) mixture of equal amounts of porcine colipase A and colipase 'eluted from the immunoadsorbent column.

46 TABLE II A M I N O A C I D ANALYSIS OF T H E SINGLE F O R M OF P A N C R E A T I C COLIPASE ISOLATED F R O M PORCINE P A N C R E A S BY I M M U N O A F F I N I T Y C H R O M A T O G R A PHY The amino acid composition of porcine colipases A and B prepared by the detergent method are included for comparison. Amino acid

Porcine colipase isolated by immunoaffinity chromatography Experimental

Nearest integer

3.95 4.71 1 1.46 9,6 7,57 8,38 2.05 5.82 8.90 4.49 . 2.02 3.04 9.15 5.53 . 2.89 4.29

4 5 12 10 8 8 2 6 9 4- 5 . 2 3 9 5-6 . . 3 4 94-96

Ala Arg Asx Cys Glx Gly His lie Leu Lys Met Phe Pro Ser Thr Trp Tyr Val Total

.

.

Porcine colipase A

Porcine colipase B

4 5 12 10 8 8 2 6 9 4

4 5 I1 10 8 8 2 6 9 4

2 3 9 5 . 3 4 94

2 3 8 5

.

3 4 92

composition of porcine colipase isolated by immunoaffinity chromatography is presented in Table II. As observed from the data of Table II, this composition is very similar to that of colipase A prepared in the laboratory. Porcine colipase obtained from the immunoadsorbent gel column is inactivated by antiserum raised against porcine colipase B. The inactivation curve is identical to that of porcine colipase A (Fig. 3a). Pure bovine colipase was not retarded on the immunoadsorbent column. In contrast, human colipase was retained under the same conditions. Discussion

Immunological properties of various forms of porcine colipase Immunological comparison of various forms of

native and trypsin-treated porcine colipase, differing by their amino- o r / a n d their carboxy-terminal end, indicates that they all bind to antibodies raised against porcine colipase B. As a result of this binding, cofactors are inactivated. Immunoinactivation studies were also performed on porcine colipase B in the presence of lipase substrate (triolein) and bile salt (sodium deoxycholate) in conditions where colipase binds to the lipid-water interface [3]. Addition of increasing amounts of the antiserum resulted in a progressive inactivation of colipase. Nonspecific serum had not effect when added to the heterogeneous system in the same conditions. These experiments show that protein-protein interaction between colipase and lipase is prevented by the binding of the antibodies to colipase and, further, that this binding is not affected by colipase adsorption at the triglyceride interface. When the antiserum was mixed with emulsified substrate to which colipase and lipase had been added, no inactivation could be detected. Thus it can be speculated that antibodies bind to colipase in a region located close to the surface domain of the cofactor that interacts specifically with the enzyme. This region is at a distinct position with respect to the previously characterized lipid binding site of colipase [35-37]. Cross-reactivity of the antiserum raised against porcine colipase B was also studied with colipases from various mammalian and avian species. It was found that human colipase binds to the porcine antibodies but with a lower affinity than the homologous antigen. No indication of cross-reaction was obtained with the cow, horse, dog and chicken colipases. Since the amino acid sequences of horse and pig colipases have been established, it seemed to us of interest to attempt to characterize better the antigenic regions of the two proteins by using the predictive method recently introduced by Hopp and Woods [30]. Results indicate that, in porcine colipase, the two regions of highest hydrophilicity are located at positions 42-48 and 70-74. Speculation can thus be made that these regions lie within or are adjacent to one of the antigenic determinants of colipase. Furthermore, it is worth considering the possible importance of the aspartic acid residue at position 72. This residue was proposed as being one of the negatively charged

47

carboxylic groups of porcine colipase that might participate in the specific binding between pancreatic colipase and lipase [38]. We find that, in porcine colipase, this residue is located in a predicted antigenic region and, as already mentioned, that the binding of antibodies to the cofactor prevents colipase-lipase interaction. These results agree well with the hypothesis according to which Asp-72 might play a critical role in colipase-lipase binding. This residue is actually conserved in pig and horse colipases, which have been shown to activate horse lipase to the same degree [8]. It can be further noted that, in horse colipase, as in the pig, the peptide segment at positions 42-48 is a region of high hydrophilicity. Since porcine and equine colipases, which do not cross-react, show about 80% homology in their sequence, antigenic determinants of the proteins should possibly be found in peptide segments which are highly substituted [39], such as regions at positions 26-33, 70-74 and 88-91 (see Table I).

Isolation of a single form of porcine pancreatic procolipase by immunoaffinity chromatography The obtention of an anti-pig colipase B serum containing 4 m g / m l of specific antibodies against porcine procolipase B allowed us to prepare an immunoadsorbent gel column by coupling the purified antibodies to Sepharose. This column should retain the various forms of native or partially degraded porcine colipase that all bind to porcine colipase B antibodies. 6 mg colipase were obtained from a supernatant of porcine pancreatic tissue homogenized in the presence of trypsin and carboxypeptidase inhibitors. All operations were completed within 30 h after collection of the gland. These conditions were aimed at minimizing proteolytic degradation of colipase during extraction and isolation. Polyacrylamide gel electrophoresis and N- and C-terminal analyses of the porcine cofactor obtained from the column indicated that a single form of porcine procolipase had been isolated, in contrast to previously described methods that yielded two forms of the cofactor. Structural, biochemical and immunological properties of this preparation were identical to those of porcine procolipase A isolated by the detergent method [12 ]. The amino-terminal residue was a valine and the C-terminal sequence was found to be Arg-Ser-

Asp-Ser. The finding that anti-porcine procolipase B antibodies bound to the column of immunoadsorption only retained procolipase A from pancreas homogenate supports the conclusion that procolipase B is not present in the porcine pancreatic tissue. Procolipase B previously isolated from pancreas homogenized in the presence of Triton X-100 is likely formed by the action of carboxypeptidases on procolipase A. In conclusion, it appears that, in contrast to the horse, where two isocolipases were identified, only one from of the cofactor is present in the pig.

Acknowledgements The authors are grateful to Professor H. Rochat • for stimulating discussions. Thanks are due to Dr. M. El Ayeb for valuable advice regarding the immunological experiments. It is a pleasure to acknowledge Dr. R.. Julien for his help, Miss. 1. Bosc-Bierne for her active collaboration and Dr. O. Guy for a gift of human and canine pancreatic juice. This work was supported, in part, by grants from the C.N.R.S. (L.A. 202) and from D.G.R.S.T. (M.R.M. 78.7.0331 and 80.7.0151).

References I Maylie, M.F., Charles, M., Gache, C. and Desnuelle, P. (1971) Biochim. Biophys. Acta 229, 286-289 2 BorgstriSm, B. and Erlanson, C. (1971) Biochim. Biophys. Acta 242, 509-513 3 Chapus, C., Sari, H., S~meriva, M. and Desnuelle, P. (1975) FEBS Left. 58, 155-158 4 Canionk P., Julien, R., Rathelot, J. and Sarda, L. (1977) Lipids 12, 393-397 5 Rietsch, J., Pattus, F., Desnuelle, P. and Verger, R. (1977) J. Biol. Chem. 252, 4313-4318 6 Granon, S. and Semeriva, M. (1980) Eur. J. Biochem. I11, 117-124 7 Borgstrism, B. (1975)J. Lipid Res. 16, 411-417 8 Rathelot, J., Julien, R., Bosc-Bierne, I., Gargouri, Y., Canioni, P. and Sarda, L. (1981) Biochimie 63, 227-234 9 Borgstr~m, B., Erlanson-Albertsson, C. and Wieloch, T. (1979) J. Lipid Res. 20, 805-816 10 Maylib, M.F., Charles, M., Astier, M. and Desnuelle, P. (1973) Biochem. Biophys. Res. Commun. 52, 291-297 11 Erlanson, C., Fernlund, P. and Borgstrbm, B. (1973) Biochim. Biophys. Acta 310, 437-445 12 Canioni, P., Julien, R., Rathelot, J., Rochat, H. and Sarda, L. (1977) Biochimie 59, (19-925 13 Borgstr6m, B., Wieloch, T. and Erlanson-Albertsson, C. (1979) FEBS Lett. 108, 407-410

48

14 Wieloch, T., Borgstrom, B., Pieroni, G., Pattus, F. and Verger, R. (1981) FEBS Lett. 128, 217-220. 15 Rathelot, J., Canioni. P., Bosc-Bierne, 1., Sarda, L., Kamoun, A., Kaptein, R. and Cozzone. P.J. (1981) Biochim. Biophys. Acta 671, 155-163. 16 Chapus, C., Desnuelle, P. and Foglizzo, E. (1981) Eur. J. Biochem. 115, 99-105 17 Larsson, A. and Erlanson-Albertsson, C. (1981) Biochim. Biophys. Acta 664, 538-548 18 Julien, R., Bechis, G., Gregoire, J., Rathelot, J., Rochat, H. and Sarda, L. (1980) Biochem. Biophys. Res. Commun. 95, 1245-1252 19 Bonicel, J., Couchoud, P., Foglizzo, E., Desnuelle, P. and Chapus, C. (1981) Biochim. Biophys. Acta 669, 39-45 20 Pierrot, M., Astier, J.P., Astier, M., Charles, M. and Drenth, J. (1982) Eur. J. Biochem. 123, 347-354 21 Sternby, B. and Borgstrism, B. (1981) Comp. Biochem. Physiol. 68, 15-18 22 Rathelot, J., Julien, R., Canioni, P. and Sarda, L. (1975) Biochimie 57, 1123-1130 23 Julien, R., Rathelot, J., Canioni, P., Sarda, L., Gregoire, J. and Rochat, H. (1978) Biochimie 60, 103-107 24 Bosc-Bierne, I., Rathelot, H. Canioni, P., Julien, R., Bechis, G., Gregoire, J., Rochat, H. and Sarda, L. (1981) Biochim. Biophys. Acta 667, 225-232 25 Rathelot, J., Julien, R., Canioni, P., Coeroli, C. and Sarda, L. (1975) Biochimie 57, 1117-1122

26 Ouchterlony, O. (1968) Prog. Allerg. 5, 1-78 27 Lowry, O.H., Rosebrough, N.J., Farr.. A.L. and Randall, R.J. (1951) J. Biol. Chem. 193, 265-275 28 Charles, M., Erlanson, C., Bianchetta, J., Joffre, A., Guidoni, A. and Rovery, M. (1974) Biochim. Biophys. Acta 359, 186-197 29 Erlanson, C., Fernlund, P. and Borgstr~m, B. (1973) Biochim. Biophys. Acta 310, 437-445 30 Hopp, T.P. and Woods, K.R. (1981) Proc. Natl. Acad. Sci. U.S.A. 78, 3824-3828 31 Affinity Chromatography: Principles and Methods (1979) Pharmacia Fine Chemicals, Uppsala 1. Sweden 32 Delori. P. and Tessier, M. (1980) Biochimie, 62, 287-288 33 Ambler, R.P. (1967) Methods Enzymol. 1I, 155-156 34 Laemli, V.K. (1970) Nature 227, 680-685 35 Wieloch, T., BorgstriSm, B., Falk, K.E. and Forsen. S. (1979) Biochemistry 18, 1622-1628 36 Canioni, P., Cozzone, P.J. and Sarda, L. (1980) Biochim. Biophys. Acta 621.29-42 37 Erlanson-Albersson, C. and Larsson, A. (1981) Biochim. Biophys. Acta 665, 250-255 38 Erlanson, C., Barrowman, J.A. and Borgstrom, B. (1977) Biochim. Biophys. Acta 489, 150-162. 39 Wilson, A.C. and Prager, E.M. (1974) in Lysozyme (Osserman, E.G., Canfield, R.E. and Beychok, S., eds.), pp. 127-141, Academic Press, New York