Radioimmunoassay for human pancreatic secretory trypsin inhibitor: measurement of serum pancreatic secretory trypsin inhibitor in normal subjects and subjects with pancreatic diseases

135

Clinica Chimica Acta, 103 (1980) 135-143 @ Elsevier/North-Holland Biomedical Press

CCA 1318

RADIOIMMUNOASSAY FOR HUMAN PANCREATIC SECRETORY TRYPSIN INHIBITOR: MEASUREMENT OF SERUM PANCREATIC SECRETORY TRYPSIN INHIBITOR IN NORMAL SUBJECTS AND SUBJECTS WITH PANCREATIC DISEASES

TAKESHI KITAHARA *, YUICHI TAKATSUKA, KEN-ICHI FUJIMOTO, SHIGENORI TANAKA, MICHIO OGAWA and GORO KOSAKI Second

Department

of Surgery,

Osaka University

Medical School,

Osaka (Japan)

(Received August lOth, 1979)

summary A reliable radioimmunoassay (RIA) for human pancreatic secretory trypsin inhibitor (PSTI) has been developed. The method is highly sensitive (0.4 ng/ml), reproducible and specific. A good parallel relationship was observed between the standard curve and dilution curves for serum and urine. The PST1 bound to trypsin-ar,-macroglobulin complexes was found not to be immunoreactive, whereas a part of the PSTI-trypsin complex was immunoreactive. In healthy individuals, serum PST1 level ranged from 5.4 ng/ml to 16.0 ng/ ml, the average being 11.3 ng/ml (S.D. + 2.7). Elevated values were observed in patients with acute pancreatitis (highest value 3200 ng/ml), and in some patients with chronic relapsing pancreatitis.

Introduction It has been known for some time that inappropriate trypsin-induced activation of proteases in the pancreas and the pancreatic duct plays an important role in the etiology of pancreatitis [ 1,2,3]. Numerous attempts have been made to quantitate activated proteases, including elastase [4], carboxypeptidase B [5] and trypsin [6] in serum, but they have all failed because of insufficient sensitivity of the method used, and the presence of large amounts of endogenous protease inhibitors, (Y,antitrypsin and czl-macroglobulin. In normal pancreatic juice, trypsinogen is protected from activation by a cer* Correspondence University

should be addressed to Takeshi Kitaham, The Second Department Medical School, l-l-50 Fukushima. Fukushima-ku. Osaka 553, Japan.

of Surgery,

Osaka

136

tain level of protease inhibitors known as Kazal inhibitor [7] or pancreatic secretory trypsin inhibitor (PSTI) [S]; the latter is a specific trypsin inhibitor secreted by the acinar cells of the pancreas into the pancreatic ducts [9,10]. Decrease or absence of this inhibitor in the ducts was thought to be the trigger factor in the development of pancreatitis [ 11,121. Thus, the measurement of PSTI, instead of proteases, in pancreatic juice and serum seems to be advisable when studying the pathophysiology of pancreatitis. Earlier approaches, using enzymological [ 11,12,13] and immunochemical [ 141 methods, however, were not sensitive nor specific enough to determine the PST1 level in biological fluids. In this paper we report the development of a radioimmunoassay (RIA) for the measurement of PST1 in human biological fluids. Materials and methods Special materials Nalz51 and Na1311 were purchased from New England Nuclear (N.E.N.), Boston, MA, U.S.A. Sephadex G-50 and G-200 were purchased from Pharmacia Fine Chemicals AB, Uppsala, Sweden. Soybean trypsin inhibitor was obtained from Boehringer, Mannheim, F.R.G. Trasylol@ was provided by Bayer AG, Leverkussen, F.R.G. Human al-antitrypsin and bovine trypsin were purchased from Sigma, St. Louis, MO, U.S.A. Human pancreatic amylase [15] and elastase-2 (K. Fujimoto et al. Biochim. Biophys. Acta, in press) were available in our laboratory. Other materials were of pure grade and commercially available. l’wification of human pancreatic secretory trypsin inhibitor Human PST1 was purified from pancreatic juice according to the modified methcd of Greene [ 131 as reported previously [ 161. The preparation proved to be homogeneous by the criteria of analytical polyacrylamide gel electrophoresis, electrophoretic analysis in the presence of sodium dodecyl sulfate (S.D.S.), and immunoelectrophoresis. Its molecular weight was calculated to be approximately 6500. Antiserum Antisera against human PST1 were produced in rabbits, and used at a dilution of 1 : 100 000, which resulted in a maximal binding of the labelled antigen of approximately 50%. Radioiodination 5 pug of the purified PST1 was radioiodinated with 1.2 mCi 1251 by the chloramine-T method as described previously [ 171. Purification of the labelled PST1 was accomplished by gel filtration on a Sephadex G-50 column. The specific activity obtained was approximately 250 pCi/pg. The labelled inhibitor retained both its antigenicity and its ability to inhibit trypsin. ‘3’I-labelled PST1 was also prepared as described above. Radioimmunoassay The radioimmunoassay

procedure

for PST1 was essentially the same as

137

described previously [ 171, with the following modification: the reaction mixture contained (a) 0.4 ml of 0.01 mol/l phosphate-buffered saline (PBS), pH 7.6, with 0.2% bovine serum albumin (BSA), 0.1% sodium azide (NaN,), 20 pmol/l Trasylol @ and 5 mmol/l ethylenediamine tetraacetic acid (EDTA); (b) 0.1 ml of standard solution or samples; (c) 0.1 ml of PST1 antiserum; (d) 0.1 ml of labelled PST1 (lo4 cpm) as tracer antigen. At the end of the first incubation, the free and the bound forms were separated by the double antibody method as described previously [ 171. The percentage binding of the tracer antigen was expressed as the ratio of the radioactivity of the bound fraction to the activity of the same fraction in the absence of unlabelled antigen. All samples were measured in duplicate. Immunoreactivity of PSTI and trypsin complex in serum 3 pg of [1311]PSTI was incubated for 15 min with 200 ,ul of serum, followed by the addition of 30 pg of trypsin. After further incubation for 1 h at room temperature, the sample was subjected to gel filtration on Sephadex G-200 (column 2.5 cm X 90 cm), equilibrated with 0.05 mol/l Tris-HCl, pH 7.6, containing 0.05 mol/l NaCl and 2 mmol/l CaC$. Flow rate was 10 ml/h, and fraction volume 3.0 ml. a,-Macroglobulin (cu,-M) and aI-antitrypsin (al-AT) peaks were identified by the radial immunodiffusion technique, using specific antibodies to each inhibitor. Serum samples The serum samples were obtained from healthy individuals and hospitalized patients with or without pancreatic disorders. Chronic pancreatitis was diagnosed by X-ray, pancreozymin-secretin test, and histological findings. The serum samples were stored at -20°C until used. Results Standard curve In Fig. 1, the standard curve for the radioimmunoassay of human PST1 is shown together with the dilution curves for normal serum and urine. These dilution curves were parallel to the standard curve, indicating that the protein detected in human sera and urine can be measured immunologically at different dilutions. Using this standard curve, PST1 concentration could be measured accurately in the range 0.4-100 ng/ml (Fig. 1). As this assay range seemed to be suitable for the determination of PST1 concentration in serum and urine, no further effort was made to increase the sensitivity of the assay. Specificity The cross-reactivities of bovine Kunitz inhibitor (Trasylol @), soybean trypsin inhibitor (SBTI), human cY1-antitrypsin and human pancreatic proteases were studied (see Fig. 2). No cross-reactivity with human PST1 was observed in any of the protease inhibitors or human pancreatic proteases tested. In sera from subjects with acute pancreatitis, however, PST1 may exist free

DILUTION 164

132

1

116

18

OF 1:4

SAMPLE 1.2

10 PST1

11

100 “g/ml

Fig. 1. PST1 standard curve and dilution curves of serum and urine. (The conditions for incubation and separation of bound antigens are described in “Materials and methods”.) The data are presented in semilog plot; .-, PST1 standard curve; o-----0, dilution curve of serum; X -X, dilution curve of urine.

0 PROTEIN

Fig. 2.

l---4, X-X,

ng/ml

Cross-reactivities of various proteases and protease inhibitors with antiserum to human PSTI. human PST1 standard curve; 0 -0, human pancreatic amylase; A-, human elastase-2; human al-antitrypsin; A-. soybean trypsin inhibitor. Trasylol @‘;0 -0,

139

or bound to trypsin, and/or trypsin-c2-macroglobulin complexes [ 18,191. Such complexes were easily obtained by incubating [ 1311]PSTI and trypsin with serum as described in “Materials and methods”. Fig. 3 shows the distribution of both 1311-labelled PST1 and PST1 measurable by radioimmunoassay in these reaction mixtures, after gel filtration on Sephadex G-200. The labelled PST1 was eluted in three peaks, corresponding to the fractions containing a2-macroglobulin, PSTI-trypsin complexes and free PSTI, respectively. The minor peak of immunoreactive PST1 was found in the elution volume of the PSTI-trypsin complex, and the major peak was in that of free PSTI. No immunoreactive PST1 was, however, determined in the fraction containing cr2macroglobulin. These observations demonstrated that PST1 bound to trypsin cross-reacts to a small extent with the standard PST1 in this RIA, while the cross-reactivity of PST1 bound to trypsin-cu2-macroglobulin complex is negligible. Immuno-cross-reactivities of PST1 between human and other mammalian species were also studied by comparing the serum dilution curves of these mammals to the standard curve of human PST1 (Fig. 4). No immunological identity was observed between human and other mammalian sera. Reproducibility and accuracy The reproducibility of the assay was studied by estimating intra- and interassay variances of 20 measurements. On intra-assay precision, the coefficients of variation (C.V.), using two pooled sera with a concentration of 9.1 ng/ml and 14.3 ng/ml, were 4.5% and 4.2%, respectively. On inter-assay precision, the

Fig. 3. Sephadex G-200 gel filtration of a mixture of 1 3 1 I-PSTI, trypsin and pooled normal serum. The condition for incubation of 1 3 1 I-PST1 and trypsln with serum were described in “Materials and methods”. The arrows showed the peak elution volumes of orz-macroglobulin (a~-M) and al-antitrypsin ((Al-AT) assayed in separate experiments.

140 SERUM 1’128

1:64

1:X?

DILUTION 1:16

1 :a

1:4

1:2

1:1

I

1

1

100

10 PST I

ro

ng s’rn,

Fig. 4. Comparative effects of antiserum to human PST1 on sera of several mammalian species. O------e, bovine serum; X-X. cat serum; a-----0, rat serum; human PST1 standard curve; O-0, A-, sheep serum; A------% canine serum.

C.V. was 5.6% for 7.6 ng/ml, and 4.2% for 12.2 ng/ml of PSTI. The percentage of the exogenous PST1 from pooled normal sera was measured by RIA (see Table I). An average recovery of 94% was obtained, suggesting that substances present in normal serum do not interfere with the assay. This fact was also supported by the good parallel relationship observed between the standard curve and the dilution curve of serum (Fig. 1). Serum PST1 con tent The serum immunoreactive PST1 content from 20 healthy individuals, aged 20 to 40 years, was measured to be 11.3 rk 2.7 ng/ml (mean + SD.), with a range of 5.4 to 16.0 ng/ml. Fig. 5 shows the levels of serum PST1 in hospitalized patients with or without pancreatic diseases.

TABLE I RECOVERY

OF PST1 ADDED TO POOLED NORMAL SERUM, MEASURED

BY RIA

Endogenous PST1 @g/ml)

Erogenous PST1 @g/ml)

Radioimmunoassayable PST1 Mean f S.D. (n&x/ml)

Percenta6e recovery exogenous PST1 (96)

6.21

1.00 2.00 5.00

7.16 f 0.44 8.06 * 0.32 10.94 * 0.33

96 92 96

of

141

-

-I-

!‘I

=w.:.. . . . .. . . .

:

. 1:;

!:

I .

Chic

\

RelapG

Pancreatitii

*

Fig. 5. Serum PST1 content of healthy controls and of patients with various diseases. Serum PST1 content was measured by the present RIA. Each bar represents the mean and standard deviation (S.D.).

Markedly elevated PST1 levels (highest value 3200 ng/ml) were observed in all patients with acute pancreatitis examined. In some patients with chronic relapsing pancreatitis, a striking increase in the level was observed during the active phase, although serum amylase activities remained within normal range. In hospital controls (free from pancreatic, liver, renal, and gastrointestinal diseases), patients suffering from cancer of the gastrointestinal tract and pancreatic cancer, PST1 content was within normal range, although two patients with gastric cancer showed slightly elevated values. They were found to have cancerous involvement of the pancreas at operation.

Discussion PST1 was first isolated by Kazal et al. [7] in 1948. Since then, its chemical structure and biological functions have been studied in great detail [13,18,20]. In contrast, information concerning the PST1 level in various biological fluids is scanty. PST1 has previously been measured by enzymological techniques [11,12,13] which, by their nature, lack specificity. In 1978, Eddeland and Ohlsson [14 ]

142

developed a specific immunochemical method to determine the level of PST1 in human pancreatic juice, but this method was not sensitive enough to measure PST1 in serum or urine. Fink and Greene [21] were first to measure successfully a picogram quantity of bovine PST1 by radioimmunoassay. This heterogeneous RIA could not, however, be used for serum from humans, because no cross-reactivity between human and bovine PST1 could be demonstrated. Very recently Eddeland and Ohlsson reported a RIA [22] for human PSTI. To date, however, neither the interaction of PST1 with trypsin and serum protease inhibitors nor the immuno-reactive of PST1 bound to trypsin or trypsin-cllz-macroglobulin complexes has been investigated in the RIA for PSTI. The RIA described here showed that PST1 bound to trypsin-cz-macroglobulin did not cross-react with the standard PST1 at all, but the cross-reactivity with PSTI-trypsin complex was approximately 25% M/M. It has been reported that the PSTI-trypsin complex, when incubated with serum in vitro or injected intravenously in dogs, dissociated rapidly, and that a large amount of PST1 existed as free inhibitor; but a small amount remains complexed with trypsin and with trypsin-cl12-macroglobulin to a limited extent

[W. Moreover, the PST1 bound to trypsiniy2-macroglobulin complex, even if its was present in the circulation, was reported to be rapidly eliminated from blood, with a half-life of 5 min [ 231. Therefore, the amount of PST1 bound to trypsin-cr2-macroglobulin in the circulation would seem to be quite negligible compared to the amount of free PSTI. This fact was also supported by the good parallel relationship observed in this report between the curves for the standard PST1 and diluted serum. Thus, the present RIA was considered to be capable of measuring more or less the total amount of PST1 in serum. In our present study, serum immunoreactive PST1 levels in acute pancreatitis were significantly higher than in normal subjects. Moreover, in some cases of acute exacerbation of chronic relapsing pancreatitis, the PST1 level was elevated, whereas amylase activity and other tests for pancreatic functions remained within the normal range. Our observations suggest that the determination of immunoreactive PST1 in serum could be of great value in the diagnosis not only of acute pancreatitis, but perhaps also of chronic relapsing pancreatitis. Further studies on the latter will be reported in the following paper. Acknowledgement This work was partly supported by a research grant (457315) Ministry of Education, Culture and Science of Japan.

from

References 1 2 3 4 5 6

Geokas, M.C. and Rinderknecht. H. (1974) Am. J. Digest. Dis. 19, 591-598 Allan, B.J. and White, T.T. (1974) Digest 11.428-431 Rinderknecht. H. and Renner, LG. (1976) Gestroenterology 70.929 Geokas. M.C.. Brodrick. J.W.. Johnson, J.H. and Largman, C. (1977) 3. Biol. Chem. 252,61-67 Geokas, M.C., Wollesen. F. and Rinderknecht, H. (1974) J.‘Lab. Clin. Med. 84, 574-683 Temler. R.S. and Felber. J-P (1976) Biochim. Biophys. Acta 445. 720-728

the

143 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

Kazal, L.A., Spicer. D.S. and Brahinsky. R.A. (1948) J. Am. Chem. Sot. 79, 3034-3040 Greene, L.J. and Giordans. Jr., J.S. (1969) J. Biol. Chem. 244, 285-298 Greene, L.J. (1968) Ann. N.Y. Acad. Sci. 146. 386-387 Renner, I.G., Rinderknecht. H. and Douglas. A.P. (1978) Gastroenterology 75. 1090-1098 Morgan, A., Robinson, L.A. and White, T.T. (1968) Am. J. Surg. 115,131-135 Geokas, M.C. and Rinderknecht. H. (1974) Digest. Dis. 19, 591-598 Pubols, M.H., Bartelt. D.C. and Greene, L.J. (1974) J. Biol. Chem. 249.1974 Eddeland, A. and Ohisson, K. (1978) Scand. J. CIin. Lab. Invest. 38. 261-267 Matsuura, K., Ogawa, M., Kosaki, G., Minamiura, N. and Yamamoto. T. (1978) J. Biochem. 83, 329332 Tanaka, S.. Ogawa. M., Iwaki, K.. Kosaki, G., Takahashi, S. and Nomoto, M. (1979) Proc. Jap. Sot. Clin. Biochem. Metabol. 16. 61-64 Takatsuka. Y.. Kitahara. T.. Matsuura, K., Ogawa. M.. Azukizawa. M., Miyai. K. and Kosaki, G. (1979) Clin. Chim. Acta 97, 261-268 Eddeland, A. and Ohisson, K. (1978) Stand. J. Clin. Lab. Invest. 38, 507-515 Eddeland. A. (1979) Howe-Seyler’s 2. Physiol. Chem. 360. 145-149 Bartelt, D.C.. Shapanka, R. and Greene, L.J. (1977) Arch. Biochem. Biophys. 179.189-199 Fink, E. and Greene, L.J. (1974) in Bayer-Symposium V “Proteinase Inhibitors”, 243-249 Eddeland, A. and Ohlsson, K. (1978) Hoppe-Seyler’s Z. Physiol. Chem. 359, 671-675 Eddeland. A. and Ohlsson, K. (1978) Hoppe-Seyler’s Z. Physiol. Chem. 359. 379-384