147

Molecular and Cellular Biochemistry 185: 147–152, 1998.© 1998 Kluwer Academic Publishers. Printed in the Netherlands.

Lithostathine messenger RNA expression indifferent types of chronic pancreatitis

Giorgio Cavallini,1 Paolo Bovo,1 Ercolina Bianchini,2 AntonellaCarsana,2 Chiara Costanzo,2 Marcello Merola,2 Daniela Sgarbi,1 LucaFrulloni,1 Vincenzo Di Francesco,1 Massimo Libonati2 and MartaPalmieri21Istituto di Clinica Medica, Cattedra di Gastroenterologia; 2Istituto di Chimica Biologica, Facoltà di Medicina e Chirurgia,Università degli Studi di Verona, Verona, Italy

Abstract

Lithostathine may play a physiological role in preventing the precipitation of excess calcium in the pancreatic juice. Thehypothesis has been advanced that in chronic calcifying pancreatitis the abnormal biosynthesis of lithostathine might be theoriginal defect to which genetic proneness to the disease may be ascribed. The aim of the present work was to studylithostathine messenger RNA expression in the pancreas of patients with different types of pancreatitis. Lithostathine andchymotrypsinogen mRNA were determined in surgical specimens obtained from the pancreases of the following subjects:(a) 13 patients with chronic alcoholic pancreatitis (84.6% calcified); (b) 4 patients with chronic hereditary pancreatitis (allcalcified); (c) 6 patients with chronic obstructive pancreatitis (4 calcified); and (d) 27 subjects suffering from pancreaticcancer. Significantly lower concentrations of both mRNAs were found in the pancreases of chronic pancreatitis patientsthan in non-cancerous tissue from pancreatic cancer subjects. However, about 70% of the pancreatic cancer subjects showedlithostathine and chymotrypsinogen mRNA levels comparable to those of chronic pancreatitis patients. These results indicatethat the decrease in the level of mRNA is not specific to lithostathine and it is unrelated to the presence of pancreatic stones.(Mol Cell Biochem 185: 147–152, 1998)

Key words: lithostathine, mRNA, chronic pancreatitis

CaCO3 (95% of stone weight) and an unsoluble polypeptide

component, identified as LS-H2, which is composed by the

C-terminal peptide of 133 amino acids of LS and it is one ofthe 2 trypsin cleavage products of naturally occurring LSdegradation [6]. It was reported that the amount of LS in thepancreatic secretion of CCP patients measured by ELISAusing a polyclonal antibody is markedly reduced [7]. Thedecrease in LS level may be responsible for the formation ofintraductal calcium stones due to calcium precipitation [6].These results, however, have not been confirmed by otherresearchers. Schmiegel et al. [8], in fact, using a fluorometricimmunoassay and a monoclonal antibody, found that the

Address for offprints: M. Palmieri, Istituto di Chimica Biologica, Facoltà di Medicina e Chirurgia, Università degli Studi di Verona, Strada Le Grazie 8,37134 Verona, Italy

Introduction

Lithostathine (LS) is a highly soluble secretory proteinpresent in mammalian pancreatic juice which inhibits nuclea-tion and growth of calcium carbonate crystals in vitro [1, 2,3, 4]. Therefore, in vivo LS should prevent the precipitationof any excess calcium present in pancreatic secretion [5].However, it has recently been demonstrated that also otherpancreatic or extra-pancreatic proteins show inhibitoryactivity against spontaneous calcium carbonate precipitationin the same concentration range of lithostathine [4]. Stonesin chronic calcifying pancreatitis (CCP) patients consist of

148

concentrations of LS in pancreatic secretions of patients withchronic or acute pancreatitis were identical to those ofcontrols. On the other hand, it was observed that the level oflithostathine mRNA (LS-mRNA) is decreased in the pan-creatic tissue of CCP patients, compared to controls orpatients with chronic obstructive pancreatitis (COP). This factshould give rise to a reduced synthesis of LS, and mayaccount for the genetic proneness to the disease [6, 9].

In this study we determined the level of lithostathine andchymotrypsinogen mRNAs in the pancreatic tissue of patientswith different types of pancreatitis and in portions of thepancreas not invaded by the tumour of subjects suffering frompancreatic cancer. The results obtained show that a signi-ficantly lower concentration of both messengers was foundin pancreases of chronic pancreatitis (CP) patients than incancer pancreases, though 70% of the last showed a level ofthe two mRNAs comparable to that of CP patients.

Materials and methods

Biological samples

Surgical samples were obtained from the pancreases of 4groups of patients. Group 1 comprised 13 patients (12 males,1 female) suffering from chronic alcoholic pancreatitis(CAP), all operated on for recurrent abdominal pain. Theirmean age was 40.8 ± 12.3 years. The diagnosis of CAP wasreached before surgery by means of medical history andclinical data (alcohol abuse, abdominal pain), biohumoraltests (increased amylase and lipase), instrumental tests (plainfilm of the abdomen, ultrasonography, computed tomo-graphy, ERCP and direct pancreatic tests). Eleven out of 13patients showed pancreatic calcifications. The mean alcoholintake of these patients was 129.2 ± 61.8 g per day, while themean number of cigarettes smoked by them was 32.8 ± 22.2per day. Group 2 consisted of 4 patients (3 females, 1 male)suffering from suspected chronic hereditary pancreatitis(CHP), with a mean age (at the beginning of the disease) of16.7 ± 3.9 years (range: 13–22 years). They were all sufferingfrom CCP and had been operated on by pancreaticojejuno-stomy because of intractable pain symptoms. All of thempresented with pancreatic calcifications. There was noevidence of alcohol abuse or chronic smoking, or other riskfactors such as primitive hyperparathyroidism, hyper-lipasemia and drug consumption. Group 3 comprised 6chronic obstructive pancreatitis (COP) patients (5 males, 1female) with a mean age of 44.8 ± 14.3 years. COP had beendiagnosed on the basis of an endoscopic examination (ERCPin all cases), which revealed a uniform Wirsung dilatationoccurring upstream of an obstacle that was endoscopicallyidentified in all cases (papillary stenosis [histological

diagnosis]: 3 cases; stenosis of Wirsung’s duct after acutenecrotizinghaemorrhagic pancreatitis: 1 case; adenomyosisof the papilla [histological diagnosis]: 2 cases). The meanalcohol intake of these patients was 100 ± 77.9 g per day.They smoked a mean number of 11 ± 5.6 cigarettes per day.Four out of 6 patients had radiologically detected stones inthe Wirsung duct. This was also confirmed by analysesperformed after the pancreaticojejunostomy which provednecessary because of persistent pain and/or a severe abnormaldigestion and absorption syndrome. Pancreatic tissue ofpatients of groups 1, 2 and 3 was subjected to histologicalanalysis. A certain degree of fibrosis was observed but nevertissue destruction. Group 4 consisted of 27 patients (12 males,15 females, with a mean age of 52.1 ± 15.3 years), allsuffering from pancreatic cancer. They had been operated onfor adenocarcinoma (17 patients), insulinoma (2 patients),cystadenoma (6 patients), endocrine malignancy (1), andpapillary carcinoma (1). Portions of the pancreas not invadedby the tumour were surgically resected and the absence oftumour tissue or chronic pancreatitis in them was confirmedby histological examinations performed by 2 independentobservers. Macro- and microscopic examinations showed theabsence of ductal stones. The mean alcohol intake of thesepatients was 21.1 ± 20.4 g per day. They smoked a meannumber of 3.7 ± 6.7 cigarettes a day.

RNA preparation

Surgical specimens of human pancreatic tissue were im-mediately frozen in liquid nitrogen and stored at –80°C. Eachspecimen was then pulverized in a precooled mortar. Theextraction of total RNA was performed as described pre-viously [10]. Two hundred mg of pulverized tissue werehomogenized in 3 ml of 4 M guanidine thiocyanate con-taining 0.1 M sodium acetate, pH 5, and 5 mM Na

2EDTA,

using an Ultra Turrax Antrieb T25 (Janke and Kunkel,Labortechnik, Staufen) for 30–50 sec at full speed. Thehomogenate was carefully layered over a 2 ml cushion of5.7 M cesium chloride containing 0.1 M sodium acetate,pH 5, and 5 mM Na

2EDTA and centrifuged at 33,000 rpm

for 18 h at 20°C in a Beckman SW 50.1 rotor. The pelletedRNA was suspended in a small volume of water and its con-centration was determined spectrophotometrically. Thequality of the RNA preparation was evaluated by 1.5%agarose/1 × TBE gel electrophoresis (5 × TBE = 0.45 MTris-HCl, 0.45 M boric acid, 0.01 M Na

2EDTA, pH 8) [11].

The quality of the RNA samples was assessed by measuringthe 28S/18S rRNA ratio which had to be as close as possibleto 2.

149

Slot-blot hybridization

Quantitative slot-blot analyses of RNA were performed asalready described [12]. Fifteen µg of RNA were resuspendedin 50 µl of ice-cold TE 1 × (10 mM Tris-HCl, 1 mM Na

2EDTA,

pH 8) and denatured in 7.4% w/w formaldehyde and 6 × SSC(20 × SSC = 3 M NaCl, 0.3 M trisodium citrate) at 65°C for15 min. For the analysis, a series of diluted aliquots of eachsample were spotted onto a Hybond N membrane using theMinifold II Slot Blot System (Schleicher and Schuell). TheRNA blots were prehybridized for 5 h at 42°C in a buffercontaining 50% (v/v) formamide, 5 × SSPE (20 × SSPE: 3.6M NaCl, 0.2 M NaH

2PO

4, 0.02 M Na

2EDTA, pH 7.4), 5 ×

Denhardt’s solution (0.1% Ficoll, 0.1% polyvinylpyrrolidone,0.1% BSA), 0.5% SDS, 80 µg/ml denatured salmon spermDNA. Hybridization was carried out at 42°C for 12 h in thesame buffer in the presence of a [32P] labelled probe. Theprobes used were: (1) a 560 bp fragment of human LS cDNA;(2) a 620 bp fragment of human chymotrypsinogen (CHT)cDNA; (3) the plasmid pXCR

7 containing the sequence of

human ribosomal RNA. The LS and CHT cDNAs were a kindgift from Dr. Dagorn [9]. The probes were labeled using aMultiprime DNA Labelling System Kit RPM 1600 (Amer-sham, specific activity 1–2.5 × 106 cpm/ng). After hybridi-zation, the nylon membranes were washed twice at roomtemperature with 2 × SSC, containing 0.1% SDS, for 10 min,then twice at 68°C with 0.5 × SSC, 0.1% SDS, for 20 min,and autoradiographed at –80°C using X-Omat S films(Kodak) and intensifying screens. The relative densitometricunits per g of RNA were determined by scanning the auto-radiograms with a laser densitometer (LKB).

Statistical analysis

The mRNAs concentrations in the pancreas of differentindividuals were non-parametric in distribution and thereforethe Mann-Whitney U test was used to compare results fordifferent patient groups.

Results

Figure 1 shows the results of a quantitative analysis of theLS-mRNA carried out on total RNA extracted from thepancreases of patients suffering from different types ofpancreatitis and from non-cancerous tissue of subjectssuffering from pancreatic cancer. All the results were nor-malized in relation to the level of ribosomal RNA, andexpressed as relative densitometric units (RDU) per µg oftotal RNA. It can be seen that the distribution of LS-mRNAvalues shows a considerable overlap (about 70%) among the

four diagnostic groups analyzed, while only 30% of pan-creatic cancer patients had a much higher level of LS-mRNA.However, when the mean value was determined in all patientswith chronic pancreatitis (CP) grouped together, a signi-ficantly lower LS-mRNA concentration was found in CP thanin subjects suffering from pancreatic cancer (C) (4.46 × 106

± 5.50 × 106 vs. 19.45 × 106 ± 24.9 × 106; p < 0.001). On theother hand, when CP patients were subdivided according todifferent forms of disease, a significant difference in the meanvalue was observed only between chronic alcoholic pan-creatitis (CAP) and C (p < 0.0008) and chronic obstructivepancreatitis (COP) and C (p < 0.002) but not between chronichereditary pancreatitis (CHP) and C (see Table 1). Nodifferences were discovered between the various CP groups.It should be noted that the lowest mean concentration of LS-mRNA was found in COP patients.

The chymotrypsinogen (CHT) mRNA pattern is similar tothat observed for LSmRNA (see Fig. 2 and Table 1). In fact,the mean CHT-mRNA value in CP was significantly lowerthan that observed in C (2.02 × 106 ± 2.16 × 106 vs. 5.06 ×106 ± 6.19 × 106; p < 0.01) and statistically significantdifferences were detected between CAP and C (p < 0.005)and between COP and C (p < 0.002), but not between CHPand C. A highly significant correlation between the levels ofLS-mRNA and CHT-mRNA was found both in CP (r = 0.69;p < 0.001) and in C (r = 0.82; p < 0.001), as shown in Figs 3and 4.

Fig. 1. Levels of LS-mRNA in the pancreas of subjects suffering frompancreatic cancer (C) and patients with chronic obstructive pancreatitis(COP), chronic alcoholic pancreatitis (CAP) and chronic hereditarypancreatitis (CHP). Results were normalized in relation to the level ofribosomal RNA and expressed as relative densitometric units (RDU) perµg of total RNA.

150

Table 1 also shows the mean values for alcohol and cigaretteconsumption in the different groups examined. On the basisof the means calculated one could argue that an increase inconsumption of alcohol and cigarettes is associated with adecreased concentration of LS-mRNA. However, no cor-relation was found between LS-mRNA concentration andalcohol or cigarettes consumption either in C or in CP patients.

Discussion

The results obtained in this study show that the concentrationsof lithostathine and chymotrypsinogen mRNAs in pancreatictissue of patients with CP is on average significantly lowerthan in portions of the pancreas not invaded by the tumourof subjects suffering from pancreatic cancer (C). However,

about 70% of the pancreatic cancer subjects showed litho-stathine and chymotrypsinogen mRNA levels comparable tothose of chronic pancreatitis patients. No significant dif-ferences in LS- and CHTmRNAs concentration were foundamong the different types of pancreatitis, and the lowestconcentrations were determined in pancreatic tissue of COPpatients. These data do not agree with those published byGiorgi et al. [9], who found low LS-mRNA concentrationsin CCP patients, normal LS-mRNA levels in COP patients,and no overlap between mRNA values measured in patientsand controls. Moreover, we demonstrate that the phenomenonis not specific to LS-mRNA which is also in contrast to whathas been shown by the same authors. In fact, we found thatCHT-mRNA levels were lower in CP patients than in C, witha significant correlation between LS-mRNA and CHT-mRNA. Our results, then, indicate that the occurrence of lowlevels of LS-mRNA is not a specific event and is not neces-sarily linked to the presence of stones. The low level oflithostathine and chymotrypsinogen mRNAs in C and CPsubjects would suggest that a reduced secretion of LS andCHT occurs in pancreatic juice (which actually was notdetermined in this study). In such case, as the low LS levelin C subjects would not seem to be sufficient to cause theformation of stones, other additional factors must be foundto explain the formation of stones observed in the majorityof CP subjects. The data in Table 1 show that a high alcoholintake in CP subjects, but not in C, as well as cigarettesmoking, may be among the factors involved. Alcohol, inparticular, is known to be capable of enhancing the litho-genicity of pancreatic secretion [13]. On the other hand, it isalso known that in animals alcohol alone is unable to inducechronic pancreatitis accompanied by lesions similar to thosedetected in humans [14, 15]. Studies in dogs and rats havealso shown that partial obstruction of the flux of pancreaticsecretion is capable of producing the formation of stones, thefrequency of which is a function of the duration of theobstruction [14–17]. Recently, the hypothesis has beenadvanced that, as in the case of primitive sclerosing cholan-gitis, the chronic calcifying pancreatitis may be due toinflammatory obliterating fibrosis of the pancreatic ducts asthe primary lesion [18]. Such an event may trigger a litho-genic process both by decreasing pancreatic flux and by

Table 1. Mean values of lithostathine and chymotrypsinogen mRNAs, alcohol intake and number of cigarettes smokedper day, in patients suffering from three types of chronic pancreatitis and subjects suffering from pancreatic cancer

CAP (n = 13) CHP (n = 4) COP (n = 6) C (n = 27)

alcohol (g/day) 129.20 ± 61.8* 0 100.00 ± 77.9 21.10 ± 20.4Cigarettes (cig/day) 32.80 ± 22.2** 0 11.00 ± 5.6 3.70 ± 6.7LS-mRNA (RDU × 106) 5.03 ± 5.98 7.27 ± 6.71 1.37 ± 1.28 19.40 ± 24.9†

CHT-mRNA (RDU × 106) 1.92 ± 2.11 2.71 ± 1.29 1.81 ± 2.92 5.06 ± 6.19††

Mann-Whitney U test: *p < 0.003 vs. C; p < 0.01 vs. CHP. **p < 0.001 vs. C; p < 0.01 vs. CHP. †p < 0.0008 vs. CAP;p < 0.002 vs. COP. ††p < 0.005 vs. CAP; p < 0.002 vs. COP.

Fig. 2. Levels of CHT-mRNA in the pancreas of subjects suffering frompancreatic cancer (C) and patients with chronic obstructive pancreatitis(COP), chronic alcoholic pancreatitis (CAP) and chronic hereditarypancreatitis (CHP). Results were normalized in relation to the level ofribosomal RNA and expressed as relative densitometric units (RDU) perµg of total RNA.

151

reducing the pancreatic proteins concentration as a con-sequence of functional damage and/or acinar cell atrophy dueto flux stasis. This would be favoured by the action ofexogenous toxic substances. This hypothesis may be sup-ported by the finding of low levels of LS-mRNA and CHT-mRNA in patients with chronic obstructive pancreatitis. Inthese subjects the pathogenetic factor is the chronic obstacleto pancreatic flux and, in contrast to previous ideas, stoneformation is a frequent event which depends on the durationof the obstruction and the presence of toxic substancescapable of interfering with LS secretion [19].

In conclusion, our present data indicate that a decreasedLS-mRNA concentration is a fairly common event in the

pancreas, which may be unrelated to the presence of chronicpancreatitis. The phenomenon does not appear to be specific,and should not be regarded as a genetically determinedprimary event . The finding that the LS-mRNA concentrationdecreases even in the absence of stones may suggest that thereduction in the LS concentration is not sufficient to triggerthe lithogenesis process, and that it must necessarily beassociated with other extrapancreatic (alcohol, smoking, etc.)or intrapancreatic (flux alteration, secretory proteins con-centration) factors. In agreement with our results, Bimmleret al. have recently [4] demonstrated that lithostathine is notthe only protein present in the pancreatic secretion which canexhibit calcite crystal inhibitor activity in vitro. Based on this

Fig. 3. Correlation between LS-mRNA and CHT-mRNA in 23 chronic pancreatitis patients.

Fig. 4. Correlation between LS-mRNA and CHT-mRNA in 27 subjects suffering from pancreatic cancer.

152

finding the authors conclude that it seems improbable thatlithostathine in vivo is a key factor in the prevention of stoneformation in pancreatic ducts.

Acknowledgements

This work was supported by grants from MURST and CISMI.

References

1. Bernard JP, Adrich Z, Montalto G, Multigner L, Dagorn JC, Sarles H,De Caro A: Immunoreactive forms of pancreatic stone protein in sixmammalian species. Pancreas 6: 162–167, 1991

2. De Caro A, Lohse J, Sarles H: Characterization of a protein isolatedfrom pancreatic calculi of men suffering from chronic calcifyingpancreatitis. Biochem Biophys Res Commun 87: 1176–1182, 1979

3. Bernard JP, Adrich Z, Montalto G, De Caro A, De Reggi M, Sarles H,Dagorn JC: Inhibition of nucleation and crystal growth of calciumcarbonate by human Lithostathine. Gastroenterology 103: 1277–1284,1992

4. Bimmler D, Graf R, Scheele GA, Frick TW: Pancreatic stone protein(Lithostathine), a physiologically relevant pancreatic calcium carbonatecrystal inhibitor? J Biol Chem 272: 3073–3082, 1997

5. Sarles H, Bernard JP, Johnson C, Chir M: Pathogenesis and epidemio-logy of chronic pancreatitis. Ann Rev Med 40: 453–468, 1989

6. Sarles H, Bernard JP: Lithostathine and pancreatic lithogenesis.Gastroenterol Intl 4: 130–134, 1991

7. Provansal-Cheylan M, Mariani A, Bernard JP, Sarles H, Dupuy P:Pancreatic stone protein: quantification in pancreatic juice by enzyme-linked immunosorbent assay and comparison with other methods.Pancreas 4: 680–689, 1989

8. Schmiegel W, Buchert M, Kalthoff H, Roeder C, Butzow G, GrimmH, Kremer B, Soehendra N, Schreiber H-W, Thiele H-G, Greten H:Immunochemical characterization and quantitative distribution ofpancreatic stone protein in sera and pancreatic secretions in pancreaticdisorders. Gastroenterology 99: 1421–1430, 1990

9. Giorgi D, Bernard JP, Rouquier S, Iovanna J, Sarles H, Dagorn JC:Secretory pancreatic stone protein messenger RNA. Nucleotidesequence and expression in chronic calcifying pancreatitis. J Clin Invest84: 100–106, 1989

10. Chirgwin JM, Przybyla AK, MacDonald RJ, Rutter WJ: Isolation ofbiologically active ribonucleic acid from sources enriched in ribo-nuclease. Am Chem Soc 18: 5294–5299, 1979

11. Sambrook J, Fritsch EF, Maniatis T: Molecular Cloning. Cold SpringHarbor Laboratory Press, 1989

12. White BA, Carter Bancroft F: Cytoplasmic dot hybridization. Simpleanalysis of relative mRNA levels in multiple small cell or tissuesamples. J Biol Chem 257: 8569–8572, 1982

13. Sarles H: Etiopathogenesis and definition of chronic pancreatitis. DigDis Sci 31: 91S–107S, 1986

14. Pap A, Boros L: Alcohol-induced chronic pancreatitis in rats aftertemporary occlusion of biliopancreatic ducts with ethibloc. Pancreas4: 249–255, 1989

15. Tanaka T, Ichiba Y, Miura Y, Itoh H, Dohi K: Pathogenesis of chronicalcoholic pancreatitis. Am J Gastroenterol 85: 1536, 1990

16. Konishi K, Izumi R, Kato O, Yamaguchi A, Miyazaki I: Experimentalpancreatolithiasis in the dog. Surgery 89: 687–691, 1987

17. Sakakibara A, Okumura N, Hayakawa T, Kanzaki M: Ultrastructuralchanges in the exocrine pancreas of experimental pancreatolithiasisin dogs. Am J Gastroent 77: 498–503, 1982

18. Cavallini G: Is chronic pancreatitis a primary disease of the pancreaticducts? A new pathogenetic hypothesis. Ital J Gastroenterol 25: 391–396, 1993

19. Cavallini G, Bovo P, Filippini M, Vaona B, Frulloni L, Di FrancescoV, Rigo L, Brunori MP, Ferrarini M, Andreaus MC, Tebaldi M,Angelini G, Procacci C, Pederzoli P: Chronic obstructive pancreatitis(COP) is a lithiasic disease. Pancreas 8: 751, 1993