- Email: [email protected]
Serum antibodies to carbonic anhydrase I and II in patients with idiopathic chronic pancreatitis and Sjogren's syndrome
GASTROENTEROLOGY 1996;110:1579–1586
Serum Antibodies to Carbonic Anhydrase I and II in Patients With Idiopathic Chronic Pancreatitis and Sjo¨gren’s Syndrome JUNKO KINO–OHSAKI,* ISAO NISHIMORI,* MASANORI MORITA,* KAZUICHI OKAZAKI,* YASURO YAMAMOTO,* SABURO ONISHI,* and MICHAEL A. HOLLINGSWORTH‡ *First Department of Internal Medicine, Kochi Medical School, Kochi, Japan; and ‡Eppley Cancer Institute, University of Nebraska Medical Center, Omaha, Nebraska
Background & Aims: Patients with idiopathic chronic pancreatitis and Sjo¨gren’s syndrome show immune responses against a 60-kilodalton protein isolated from human pancreas extracts. Monoclonal antibody SP3-1, raised against the 60-kilodalton protein, reacts with ductal cells in several exocrine organs and cross-reacts with human carbonic anhydrase II. The present study evaluated serum from patients with these diseases for antibodies to human carbonic anhydrase I and II. Methods: An enzyme-linked immunosorbent assay was used to quantify serum antibody against carbonic anhydrase I and II. Results: Serum antibodies against carbonic anhydrase I and II were detected in 7 and 11 of 33 patients with idiopathic chronic pancreatitis, respectively, and in 8 and 13 of 21 patients with Sjo¨gren’s syndrome, respectively. The positive prevalence rates of patients with antibodies to carbonic anhydrase II were significantly different among patients with idiopathic chronic pancreatitis and Sjo¨gren’s syndrome compared with normal patients (1 of 21). There were no significant differences in the prevalence rates of patients with alcoholic chronic pancreatitis (3 of 20), gallstone-related chronic pancreatitis (0 of 7), and primary biliary cirrhosis (1 of 11). Conclusions: The results suggest that carbonic anhydrase II is one candidate target antigen recognized during the autoimmune pathophysiology observed in idiopathic chronic pancreatitis and Sjo¨gren’s syndrome.
B
etween 30% and 40% of patients with chronic pancreatitis (CP) have no apparent underlying cause of their disease.1 These patients are considered to have idiopathic CP, and several etiologic factors have been proposed, such as dysfunction of the sphincter of Oddi and ischemia.2 Autoimmunity has also been suggested as a possible etiologic factor of idiopathic CP, and various immunologic alterations have been described in patients with CP.3 – 6 However, the autoimmune mechanisms that contribute to the pathogenesis of CP remain poorly understood. / 5e0d$$0019
04-18-96 16:54:11
gasas
It has been well documented that CP is occasionally observed as a complication in patients with Sjo¨gren’s syndrome (SjS) and primary biliary cirrhosis (PBC).7,8 Based on clinical findings, the concept of a disease complex, autoimmune exocrinopathy9 or dry gland syndrome,10 has been put forth to describe patients with the above-mentioned diseases. It is hypothesized that these diseases may be manifestations in distinct organs of an autoimmune reaction against a common antigen that is expressed by the ductal epithelial cells of exocrine organs. It has been reported that patients with SjS and CP have serum antibodies that react with duct cells in salivary glands and pancreas11 – 15; however, the molecular nature of this antigen has not been determined. We reported previously that patients with SjS and idiopathic CP showed cellular and humoral immune responses against a 60-kilodalton protein that was isolated from human and porcine pancreas extracts.16 – 18 Monoclonal antibody SP3-1 was raised against the 60-kilodalton protein and was found to react with the ductal cells in several exocrine organs, including pancreas, salivary gland, liver (biliary duct), kidney (distal renal tubule), and esophageal glands.19 Recently, we determined that monoclonal antibody SP3-1 reacts with human carbonic anhydrase II (CAII), a finding that is consistent with the distribution of SP3-1 reactive tissues described above.20 Moreover, we reported that sialoadenitis was successfully induced in mice bearing the H-2s and H-2u major histocompatibility complex haplotypes by intradermal immunization with human CAII and adjuvant.20 Recently, it has been reported by several investigators that patients with SjS, systemic lupus erythematosus (SLE), and autoimmune cholangitis have serum autoantibodies to Abbreviations used in this paper: BT-PABA, N-benzoyl-L-tyrosyl-pamino-benzoic acid; CAI, carbonic anhydrase I; CAII, carbonic anhydrase II; CP, chronic pancreatitis; ERP, endoscopic retrograde pancreatography; PBC, primary biliary cirrhosis; SjS, Sjo¨gren’s syndrome; SLE, systemic lupus erythematosus. 䉷 1996 by the American Gastroenterological Association 0016-5085/96/$3.00
WBS-Gastro
1580 KINO–OHSAKI ET AL.
GASTROENTEROLOGY Vol. 110, No. 5
CAII.21 – 23 These observations prompted us to determine whether patients with idiopathic CP have autoantibodies to CAII. In the present study, sera from patients with CP and with SjS, PBC, and SLE were evaluated for autoantibodies to CAII and carbonic anhydrase I (CAI).
Materials and Methods Patients Sera from 74 patients with CP were studied. The diagnosis of CP was based on one of the following findings: decreased pancreatic excretory function as shown by the secretin test; pancreatic stones; dilatation, stenosis, or cyst formation of the main pancreatic duct and its branches shown by endoscopic retrograde pancreatography (ERP); or histologically proven CP. Patients with CP had no history of acute pancreatitis and were divided into the following five groups according to etiology: 20 patients with alcoholic CP, 7 patients with gallstone-related CP, 7 patients with SjS-associated CP, 4 patients with PBC-associated CP, and 33 patients with idiopathic CP. Patients with idiopathic CP had no obvious cause of CP, such as alcoholism, gallstones, or other identifiable etiologic factors including associated autoimmune diseases. In addition, 21 patients with SjS and 11 patients with PBC, who were not complicated by CP, were studied. SjS and PBC were diagnosed according to the previously defined criteria, respectively.24,25 Twelve patients with SLE were diagnosed according to the American Rheumatism Association 1982 revised criteria for the classification of SLE,26 and 21 normal controls were also studied.
Enzyme-Linked Immunosorbent Assay for CAI and CAII Antibody Human CAI and CAII (electrophoretically purified from erythrocytes) were purchased from Sigma Chemical Co. (St. Louis, MO). Serum antibodies to CAI and CAII were quantified by enzyme-linked immunosorbent assay using a standard method with minor modifications.27 Briefly, microtiter plates were coated with 50 mL of 5 mg/mL CAI or CAII overnight at 4⬚C. After five washes with phosphate-buffered saline, the plates were coated with 2% bovine serum albumin and 5% goat serum in phosphate-buffered saline overnight at 4⬚C and then incubated with 50 mL of diluted patient serum (1:25) overnight at 4⬚C. After five washes with phosphatebuffered saline containing 0.05% Tween 20 (Nacalai, Kyoto, Japan), the plates were incubated with 50 mL of goat antihuman immunoglobulin G antibody conjugated with horseradish peroxidase (Sigma Chemical Co.) at room temperature for 2 hours. After five washes with phosphate-buffered saline containing 0.05% Tween 20, bound reactants were detected by 15-minute incubation with 50 mL of 50 mg/mL 2.2-azinobis diammonium salt (Nacalai). The absorbance was determined at 405 nm. Control wells that were not coated with CAI or CAII were also used for enzyme-linked immunosorbent assay of each serum studied. All assays were performed in triplicate,
/ 5e0d$$0019
04-18-96 16:54:11
gasas
and the specific binding of serum antibody to CAI or CAII was calculated as follows: the average absorbance of the antigen-coating wells minus the average absorbance of control wells.
Specific Inhibition of Serum Anti-CAI or Anti-CAII Antibody The specificity of the enzyme-linked immunosorbent assay system was assessed by absorption of serum antibody with CAI or CAII. Briefly, microtiter wells were precoated with 50 mL of 0–80 mg/mL CAII overnight at 4⬚C and further incubated with 2% bovine serum albumin and 5% goat serum in phosphate-buffered saline overnight at 4⬚C. Aliquots of sera diluted to 1:25 in phosphate-buffered saline were incubated in the wells at room temperature for 2 hours. The absorbed samples were tested for anti-CAI or anti-CAII antibody by enzyme-linked immunosorbent assay as described above.
Assessment of Clinical Findings in Patients With CP With Serum Anti-CAII Antibody Clinical findings in patients with CP with serum antiCAII antibodies were evaluated and compared with patients with CP without anti-CAII antibodies. The following factors were assessed: age, sex, serum pancreatic enzyme levels (amylase, elastase I, and trypsin), pancreatic exocrine function studied by N-benzoyl-L-tyrosyl-p-amino-benzoic acid (BT-PABA) excretion test, and ERP. Non–organ specific immunologic parameters, including antinuclear antibody and rheumatic factor, and serum concentrations of g-globulin and immunoglobulins G, A, and M were also evaluated. Serum amylase levels were measured using a commercially available kit with ethylidene-(G7)-para-nitrophenylmaltoheptaoside as a substrate (Amylase EPS; Boehringer, Tokyo, Japan; normal range, 40– 130 IU/L). Elastase I and trypsin were quantified by using commercially available kits (Dinabot, Tokyo, Japan, and Hoechst, Hounslow, England; normal range, 80–410 ng/dL and 110–460 ng/mL, respectively). BT-PABA excretion tests were performed according to previously described methods.28 ERP findings were graded into the following four groups according to the criteria defined by Kasugai et al.29: normal, minimal-change pancreatitis, moderate-change pancreatitis, and advanced-change pancreatitis. Antinuclear antibody and rheumatic factor were studied by immunofluorescence methods using commercially available kits as follows: antinuclear antibody, indirect immunofluorescence using HEp-2 cell as substrate (Fluoro Hepana Test; MBL, Nagoya, Japan); and rheumatic factor, indirect agglutination using rabbit denatured immunoglobulin G as antigen (Se-Rodia-RA, Fujirebio, Tokyo, Japan). Serum g-globulin concentration was estimated based on serum total protein concentration and whole serum electrophoresis quantified on a densitometer. Serum concentrations of immunoglobulin subclass were measured by rate nephelometry (Beckman, Galway, Ireland).
WBS-Gastro
May 1996
ANTI–CAII ANTIBODY IN CHRONIC PANCREATITIS 1581
Figure 1. Enzyme-linked immunosorbent assay of serum antibody against CAII. Absorbance was determined at 405 nm (A405). Dotted line indicates mean absorbance / 2SD of normal controls (0.341). P õ 0.05, significant difference in comparisons of mean { SD value between patient groups and normal controls.
Statistical Analysis Fisher’s Exact Probability Test with Yates’ correction was used for statistical analyses of the antibody prevalence rate, patient’s profile, ERP finding, antinuclear antibody, and rheumatic factor. Student’s t test was performed in comparisons of mean values of serum anti-CAII and anti-CAI antibody titers, pancreatitis enzyme levels, and g-globulin and immunoglobulin subclass concentrations. Pearson’s correlation coefficient and Fisher’s Exact Test were used for an analysis of correlation between serum anti-CAII and anti-CAI antibody titer. Statistical significance was defined as P values of õ0.05. All data were expressed as mean { SD.
body-positive prevalence rate in patients with idiopathic CP and SjS (P õ 0.05). Significant differences in mean values of serum antibody titers were also observed in patients with idiopathic CP (0.314 { 0.125), SjS (0.360 { 0.068), and SjS-associated CP (0.313 { 0.052) compared with normal controls (0.247 { 0.047) (P õ 0.05). There was no significant difference in mean { SD values between normal controls and patients with alcoholic CP (0.272 { 0.094), gallstone-related CP (0.237 { 0.040), PBC (0.273 { 0.134), PBC-associated CP (0.313 { 0.053), and SLE (0.287 { 0.089). Absorption of serum anti-CAII antibody in 2 patients with idiopathic CP and 3 patients with SjS reduced absorbance values to less than the mean / 2SD values determined for normal controls (0.341) (Figure 2). Preabsorption had no effect on absorbance for normal controls, suggesting that these normal controls had no detectable anti-CAII antibody. This finding confirms the specificity of the enzyme-linked immunosorbent assay system used in the present study for measuring anti-CAII antibodies in sera and supports our use of a cutoff point that is 2SD greater than mean values determined for normal controls. Anti-CAI Antibodies Anti-CAI antibody was also quantified in those disease groups that showed a relatively high prevalence
Results Anti-CAII Antibodies Levels of serum anti-CAII antibody in patients and normal controls were quantified by enzyme-linked immunosorbent assay (Figure 1). The mean { SD of absorbance values for normal controls was 0.247 { 0.047 (n Å 21). In the present study, positive results were arbitrarily defined as absorbance values of ú0.341, which corresponds to a value that is the mean / 2SD of normal control values. Among patients with CP, positive results were obtained in 11 of 33 patients with idiopathic CP (33%), 3 of 20 patients with alcoholic CP (15%), 2 of 7 patients with SjS-associated CP (29%), 1 of 4 patients with PBC-associated CP (25%), and 0 of 7 patients with gallstone-related CP (0%). Thirteen of 21 patients with SjS (62%), 1 of 11 patients with PBC (9%), 3 of 12 patients with SLE (25%), and 1 of 21 normal controls (4.8%) showed positive results. Compared with normal controls, there was a significant difference in the anti/ 5e0d$$0019
04-18-96 16:54:11
gasas
Figure 2. Absorption of serum antibody with CAII in enzyme-linked immunosorbent assay as described in Materials and Methods. Patients with idiopathic CP (●; n Å 2) and with SjS (䊊; n Å 3) showed inhibition of the anti-CAII antibody down to the level of normal controls (䊏; n Å 2) in an antigen concentration–dependent manner. A405, Absorbance at 405 nm.
WBS-Gastro
1582 KINO–OHSAKI ET AL.
GASTROENTEROLOGY Vol. 110, No. 5
antibody titer was evaluated using Pearson’s Correlation Coefficient and Fisher’s Tests. In patients with SjS, the presence of anti-CAII antibody showed a fair correlation with the presence of anti-CAI antibody (r Å 0.55627; P õ 0.01; Figure 4A). In patients with idiopathic CP, however, there was no significant correlation between anti-CAI antibody and anti-CAII antibody titers (Figure 4B). The patients could be divided into the following three groups: anti-CAI antibody positive and anti-CAII antibody negative (anti-CAI Ab/), anti-CAI antibody negative and anti-CAII antibody positive (anti-CAII Ab/), and negative for both antibodies. One patient had both antibodies but with marginally positive values. Two pairs of patients were sampled among anti-CAI Ab/ and anti-CAII Ab/ groups: idiopathic CP-1 and -2 and idiopathic CP-3 and -4, respectively. The specificity of serum antibody against CAI or CAII in these patients was further studied as described below (Figure 5). Specificity of Serum Antibodies Against CAI or CAII in Patients With Idiopathic CP Figure 3. Enzyme-linked immunosorbent assay of serum antibody against CAI. Absorbance was determined at 405 nm (A405). Dotted line indicates mean absorbance / 2SD of normal controls (0.146). P õ 0.05, significant difference in comparisons of mean { SD value between patient groups and normal controls.
rate of anti-CAII antibody, including idiopathic CP, SjS, and SjS-associated CP (Figure 3). Positive results were arbitrarily set as an absorbance of ú0.146, corresponding to the mean { 2SD of values for normal controls. AntiCAI antibody was detected in 7 of 33 patients with idiopathic CP (21%), 8 of 21 patients with SjS (38%), and 2 of 7 patients with SjS-associated CP (29%). There was a significant difference in the antibody-positive prevalence rate in patients with SjS compared with normal controls (1 of 20; 5%) (P õ 0.05). Significant differences in mean values of the serum antibody titers were observed in patients with SjS (0.147 { 0.061) and SjS-associated CP (0.125 { 0.063) compared with normal controls (0.090 { 0.028) (P õ 0.05). There was no significant difference in mean values between normal controls and patients with idiopathic CP (0.119 { 0.084). It is noteworthy that 2 patients with idiopathic CP showed the highest antibody titers among all patients studied, including patients with SjS. Correlation Between Anti-CAII Antibodies and Anti-CAI Antibodies In patients with idiopathic CP and SjS, the correlation between anti-CAII antibody titer and anti-CAI / 5e0d$$0019
04-18-96 16:54:11
gasas
The specificity of serum antibodies in patients from the anti-CAI Ab/ and anti-CAII Ab/ groups was studied by absorption of the sera with either CAI or CAII as described in Materials and Methods (Figure 5). Serum anti-CAI antibodies in ICP-1 and ICP-2 patients (antiCAI Ab/ group; see Figure 4B) were completely absorbed by incubation with CAI, but they were not absorbed with CAII (Figure 5A). In contrast, serum anti-CAII
Figure 4. Correlation between anti-CAI antibody titer and anti-CAII antibody titer in patients with SjS and with idiopathic CP. (A ) In patients with SjS, anti-CAII antibody showed a relatively good correlation to anti-CAI antibody (r Å 0.55627; P õ 0.01). (B ) In patients with idiopathic CP (ICP), there is no significant correlation between two antibodies, and it was possible to divide patients into the following three groups: anti-CAI antibody positive and anti-CAII antibody negative, anti-CAI antibody negative and anti-CAII antibody positive, and negative for both antibodies. Only 1 patient showed both antibody positive with a marginally positive value. Dotted lines indicate cutoff value of absorbance (A405) in both antibodies as described in Results and in the legends to Figures 1 and 3.
WBS-Gastro
May 1996
ANTI–CAII ANTIBODY IN CHRONIC PANCREATITIS 1583
Figure 5. Absorption of serum antibody with CAI (䊏) or CAII (䊊) in enzyme-linked immunosorbent assay as described in Materials and Methods. (A ) Serum anti-CAI antibodies in ICP-1 and ICP-2 patients (see Figure 4B ) were absorbed by an incubation with CAI down to the level of normal controls (NC) but were not absorbed by CAII at all. (B ) Serum anti-CAII antibodies in ICP-3 and ICP-4 patients (see Figure 4B ) absorbed by an incubation with CAII, but these were not absorbed by CAI. A405, Absorbance at 405 nm.
antibodies in ICP-3 and ICP-4 patients (anti-CAII Ab/ group; see Figure 4B) were absorbed by an incubation with CAII but not with CAI (Figure 5B). Although absorption of anti-CAII antibody with CAII in ICP-3 and ICP-4 patients did not reduce the absorbances to background values, the absorbances were less than the mean / 2SD of values for normal controls (0.341). These findings show that patients in each group, anti-CAII Ab/ and anti-CAI Ab/ groups, have antibodies specific for CAI or CAII, respectively. Clinical Findings in Patients With CP With Serum Anti-CAII Antibodies Regarding clinical parameters described in Materials and Methods, patients with CP with serum antiCAII antibodies were compared with patients with CP without the antibody (Table 1). There was no significant difference in age or sex among both groups. No significant differences were observed in serum pancreatic enzyme levels, exocrine pancreatic function studied by the BT-PABA excretion test, and pancreatic ductal morphology classified by ERP among the two groups. Antinuclear antibody was observed with a significantly higher prevalence rate in patients with CP with anti-CAII antibody, but rheumatic factor was not observed with a higher rate. Regarding total and subclass immunoglobulin concentrations, only immunoglobulin A levels were signifi/ 5e0d$$0019
04-18-96 16:54:11
gasas
cantly greater in patients with CP with anti-CAII antibody; however, the values were within the normal range. Despite a finding of no significant difference in pancreatic ductal morphology when a classical grading of pancreatography was used, it is noteworthy that patients with CP with anti-CAII antibody often showed an interesting pancreatic ductal configuration as shown in Figure 6. The main pancreatic duct and branch of the duct were narrowed in segments and were irregularly shaped. This abnormality was rarely observed in patients with CP without anti-CAII antibodies, especially among patients with alcoholic CP. It was difficult to draw conclusions from clinical findings of patients with idiopathic CP with anti-CAI antibody because of the small number of patients; only 2 patients were obviously positive for anti-CAI antibodies.
Discussion One significant finding of this study is the prevalence rate of anti-CAII antibodies in sera of patients with idiopathic CP (11 of 33 patients). In contrast, only 3 of 20 patients with alcoholic CP and 0 of 7 patients with gallstone-related CP had anti-CAII antibodies in their sera. These findings suggest that a significant percentage of patients with idiopathic CP may have some form of autoimmune CP. The fact that significantly lower percentages of patients with alcoholic CP and gallstonerelated CP had anti-CAII antibodies implies that production of anti-CAII antibody is not a result of pancreas tissue injury. This suggests that autoimmune responses to CAII should be further investigated as a possible etiologic factor of CP, along with long-term alcohol consumption and the presence of gallstones. A second significant finding is that 62% of patients with SjS without CP (13 of 21 patients) and 29% of patients with SjS-associated CP (2 of 7 patients) have anti-CAII antibodies in their sera. Together, 54% of patients with SjS show anti-CAII antibodies. Thus, antiCAII antibodies are detected in some but not all patients with SjS. This finding is consistent with previous studies by Inagaki et al.21 and Itoh and Reichlin,22 who reported prevalence rates for serum anti-CAII antibody in patients with SjS of 12.5% and 20.8%, respectively. Interestingly, antibodies reactive with pancreatic ductal cells have been reported in patients with SjS.11,12 It has been proposed that a disease complex exists that includes patients with SjS, CP, PBC, primary sclerosing cholangitis, and renal tubular acidosis, and this disease complex has been given various names over the past 15 years, including autoimmune exocrinopathy,9 dry gland syndrome,10 or autoimmune epithelitis.30 It is hypotheWBS-Gastro
1584 KINO–OHSAKI ET AL.
GASTROENTEROLOGY Vol. 110, No. 5
Table 1. Clinical Findings in Patients With CP With or Without Serum Anti-CAII Antibody Anti-CAII antibody
Age (yr) Sex (M/F) Serum pancreatic enzyme Amylase (IU/L) Elastase I (ng/dL) Trypsin (ng/mL) BT-PABA excretion test (%) ERP Advanced-change pancreatitis Moderate-change pancreatitis Minimal-change pancreatitis Normal Antinuclear antibody Rheumatic factor g-Globulin (g/dL) Immunoglobulin G (mg/dL) Immunoglobulin A (mg/dL) Immunoglobulin M (mg/dL)
Positive
Negative
Significance of differencea
65.5 { 11.1 11/4
59.6 { 11.5 33/15
NS NS
150.0 { 49.1 344 { 297 508.7 { 232.3 80.8 { 16.1
200 { 150.7 437 { 263.8 579 { 279.7 79.2 { 15.3
NS NS NS NS
2/12 (16.7%) 6/12 (50.0%) 3/12 (25.0%) 1/12 (8.3%) 6/13 (46.2%) 1/12 (8.3%) 1.50 { 0.25 1465.9 { 329.0 319.8 { 130.0 135.4 { 56.2
2/29 (6.9%) 10/29 (34.5%) 12/29 (41.3%) 5/29 (97.2%) 2/24 (8.3%) 0/26 (0%) 1.27 { 0.40 1303.0 { 531.9 232.5 { 83.4 162.8 { 96.5
NS NS NS NS P õ 0.05 NS NS NS P õ 0.05 NS
a
Significance of difference in comparison of mean value or of positive prevalence rate studied in patients with CP with and without serum anti-CAII antibody.
sized that one common mechanism of pathogenesis of these diseases is an autoimmune reaction against a common antigen that is expressed in the ductal epithelial cells of exocrine organs. CAII is a reasonable target anti-
Figure 6. ERP of 2 patients with idiopathic CP with anti-CAII antibody. Note irregular configuration including segmental narrowing but no or mild dilatation of main pancreatic duct and its branches (arrowhead ).
/ 5e0d$$0019
04-18-96 16:54:11
gasas
gen for such an autoimmune disease complex because it is expressed by duct cells of exocrine organs included in these syndromes, including pancreas, salivary gland, liver (bile duct), and kidney (renal tubule).31,32 In addition to the data described above, several experimental findings support the hypothesis that autoimmune responses to CAII may be involved in the pathogenesis of this complex of related diseases. Recently, experimental autoimmune sialoadenitis was successfully induced in SJL/JCr (H-2s) and PL/J (H-2u) mice by immunization with CAII,20 suggesting that major histocompatibility complex–restricted immune responses to CAII are associated with the pathogenesis in this model mouse for SjS. In addition, patients with other autoimmune disease have been reported to have anti-CAII antibodies in their sera.21 – 23 These include patients with SLE21,22 and autoimmune cholangitis.23 In the study of patients with autoimmune cholangitis, 83.3% of patients had anti-CAII antibodies. In contrast, only 8.3% of patients with PBC had anti-CAII antibodies.23 It was proposed that autoimmune cholangitis is a novel disease entity in patients with PBC-compatible histological findings in liver specimens who are negative for antimitochondrial antibodies and positive for antinuclear antibody.33 The present study also failed to detect anti-CAII antibodies in patients with PBC (1 of 11 patients; 9.1%). One interpretation of these results is that the liver phenotype that presents during autoimmune exocrinopathies directed toward CAII is autoimmune cholangitis instead of PBC. To further investigate the specificity of the anti-CAII WBS-Gastro
May 1996
ANTI–CAII ANTIBODY IN CHRONIC PANCREATITIS 1585
antibodies observed in patients with CP and SjS, we investigated the reactivity of these patients’ sera with CAI. Carbonic anhydrases are zinc metalloenzymes that catalyze the reversible reaction CO2 / H2O Å HCO30 / H/, the hydration of certain aldehydes, and the hydrolysis of certain esters.34 There are at least eight isozymes of carbonic anhydrase that have been studied.35 The similarity in amino acid sequence between CAI and CAII is 60.2%. In contrast to CAII, which is widely expressed in human cells including erythrocytes, exocrine glands described above, and the gastrointestinal tract including stomach, small intestine, and colon,36 – 38 CAI is expressed mainly in erythrocytes. In the present study, 7 of 33 patients with idiopathic CP, 8 of 21 patients with SjS, and 2 of 7 patients with SjS-associated CP had anti-CAI antibodies in their sera. It seemed likely that the anti-CAI antibodies detected in patients with SjS were cross-reactive with the anti-CAII antibodies because the anti-CAI antibody titer in these patients showed good correlation with anti-CAII antibody titer. In patients with idiopathic CP, however, anti-CAII antibodies and anti-CAI antibodies were independently observed in different patients. Moreover, the explanation that some antiCAI antibodies are cross-reactive with similar amino acid sequences on CAII, whereas others are not, is not supported by the data presented in Figure 5. In one previous report, serum anti-CAI antibody and anti-CAII antibody were concomitantly or independently observed in patients with SjS and SLE.22 The exact reason for the independent occurrence of both antibodies is unknown, because expression of CAI is mostly restricted to erythrocytes in human tissues as described above. The fact that anti-CAII antibodies are detected in only a portion of patients with SjS, CP, and related diseases makes it unlikely that CAII is the only antigen recognized in these diseases. Instead, it is likely that there are several target antigens and that different combinations of diseases could result from immune reactions against the different antigens. In this light, it is noteworthy that some patients with CP with anti-CAII antibody were positive for antinuclear antibody. The prevalence rate of antinuclear antibody (46.2%) for these patients was significantly greater than patients without the antibody (8.3%). Interestingly, all 3 patients with alcoholic CP with anti-CAII antibody were also positive for antinuclear antibody. It remains possible that autoimmune reactions to CAII have some effect on the pathophysiology of these patients during acceleration or modification of pancreatic tissue injury by alcohol. Pancreatic ductal morphology evaluated by ERP is also noteworthy as a clinical finding in patients with CP with anti-CAII anti/ 5e0d$$0019
04-18-96 16:54:11
gasas
body. An irregular configuration, including segmental narrowing but no or mild dilatation of the pancreatic duct system, that was observed in anti-CAII–positive patients with CP (Figure 6) is rarely observed in patients with alcoholic CP. In patients with alcoholic CP, the pancreatic duct usually shows a marked dilatation and occasional calcification. We previously reported that patients with SjS and PBC showed similar ERP findings.39 This difference in ERP findings between patients with alcoholic CP and patients with anti-CAII antibody–positive CP may suggest that different pathophysiology contributes to the disease process in these patients. We previously reported that patients with idiopathic CP have antibodies and show lymphocyte-proliferative responses against a 60-kilodalton molecule isolated from human pancreas extracts.16 – 18 The 60-kilodalton pancreatic molecule is hypothesized to be a homodimer of CAII based on the following previous results: a monoclonal antibody against this antigen (SP3-1) reacts with CAII; the histological distribution of the antigen is consistent with the distribution of CAII; and other findings.19,20 The present study provides further support for this hypothesis by showing the presence of serum antibodies against human CAII in a significant percentage of patients with idiopathic CP and SjS. These results indicate that immune sensitization against CAII or its derived peptides has occurred in some patients with idiopathic CP and SjS. Further studies, including an analysis of T-cell responses against CAII, are warranted to further investigate the pathogenesis of these diseases.
References 1. Steer ML, Waxman I, Freedman S. Chronic pancreatitis. N Engl J Med 1995;332:1482–1490. 2. Markes IN, Bank S. Chronic pancreatitis: etiology, clinical aspects, and medical management. In: Berk JE, ed. Bockus’ gastroenterology. Volume 6. 4th ed. Philadelphia: Saunders, 1985: 4020–4040. 3. Bank S, Novis BH, Petersen E, Dowdle E, Marks IN. Serum immunoglobulins in calcific pancreatitis. Gut 1973;14:723–725. 4. Antal L, Ka´vai M, Szabo´ G, Sonkoly I, Pa´lo´czi K, Szegedi G, Sa´py P, Va´rhelyi I. Immunological investigations in acute and chronic human pancreatitis. Digestion 1980;20:100–105. 5. Lankisch PG, Koop H, Seelig R, Seelig HP. Antinuclear and pancreatic acinar cell antibodies in pancreatic diseases. Digestion 1981;21:65–68. 6. Bovo P, Mirakian R, Merigo F, Angelini G, Cavallini G, Rizzini P, Bottazzo GF, Scuro LA. HLA molecule expression on chronic pancreatitis specimens: is there a role for autoimmunity? A preliminary study. Pancreas 1987;2:350–356. 7. Sheikh SH, Shaw-Stiffel TA. The gastrointestinal manifestations of Sjo¨gren’s syndrome. Am J Gastroenterol 1995;90:9–14. 8. Schaffner F, Bach N. Gastrointestinal syndromes in primary biliary cirrhosis. Semin Liver Dis 1988;8:263–271. 9. Strand V, Talal N. Advances in the diagnosis and concept of Sjo¨gren’s syndrome (autoimmune exocrinopathy). Bull Rheum Dis 1980;30:1046–1052.
WBS-Gastro
1586 KINO–OHSAKI ET AL.
GASTROENTEROLOGY Vol. 110, No. 5
10. Epstein O, Chapman RWG, Lake-Bakaar G, Foo AY, Rosalki SB, Sherlock S. The pancreas in primary biliary cirrhosis and primary sclerosing cholangitis. Gastroenterology 1982;83:1177–1182. 11. Ludwig H, Schernthaner G, Scherak O, Kolarz G. Antibodies to pancreatic duct cells in Sjo¨gren’s syndrome and rheumatoid arthritis. Gut 1977;18:311–315. 12. Sundkvist G, Lindahl G, Koskinen P, Bolinder J. Pancreatic autoantibodies and pancreatic function in Sjo¨gren’s syndrome. J Intern Med 1991;229:61–66. 13. Macsween RNM, Goudie RB, Anderson JR, Armstrong E, Murray MA, Mason DK, Jasani MK, Boyle JA, Buchanan WW, Williamson J. Occurrence of antibody to salivary duct epithelium in Sjo¨gren’s disease, rheumatoid arthritis, and other arthritides: a clinical and laboratory study. Ann Rheum Dis 1967;26:402–411. 14. Anderson LG, Tarpley TM, Talal N, Cummings NA, Wolf RO, Schall GL. Cellular-versus-humoral autoimmune responses to salivary gland in Sjo¨gren’s syndrome. Clin Exp Immunol 1973;13:335– 342. 15. Rumessen JJ, Marner B, Thorsgaard Pedersen N, Permin H. Autoantibodies in chronic pancreatitis. Scand J Gastroentorol 1985; 20:966–970. 16. Nishimori I, Okazaki K, Yamamoto Y, Morita M, Tamura S, Yamamoto Y. Sensitization against pancreatic antigen in Sjo¨gren’s syndrome and chronic pancreatitis. Digestion 1992;51:71–74. 17. Nishimori I, Okazaki K, Yamamoto Y, Morita M, Tamura S, Yamamoto Y. Specific cellular immune responses to pancreatic antigen in chronic pancreatitis and Sjo¨gren’s syndrome. J Clin Immunol 1993;13:265–271. 18. Nishimori I, Yamamoto Y, Okazaki K, Morita M, Onodera M, Kino J, Tamura S, Yamamoto Y. Identification of autoantibodies to a pancreatic antigen in patients with idiopathic chronic pancreatitis and Sjo¨gren’s syndrome. Pancreas 1994;9:374–381. 19. Okazaki K, Tamura S, Morita M, Nishimori I, Yamamoto Y, Yamamoto Y. Interspecies crossreactive antigen of the pancreatic duct cell prepared by monoclonal antibody. Int J Pancreatol 1989;5: 359–377. 20. Nishimori I, Bratanova T, Toshkov I, Caffrey T, Mogaki M, Shibata Y, Hollingsworth MA. Induction of experimental autoimmune sialoadenitis by immunization of PL/J mice with carbonic anhydrase II. J Immunol 1995;154:4865–4873. 21. Inagaki Y, Jinno-Yoshida Y, Hamasaki Y, Ueki H. A novel autoantibody reactive with carbonic anhydrase in sera from patients with systemic lupus erythematosus and Sjo¨gren’s syndrome. J Dermatol Sci 1991;2:147–154. 22. Itoh Y, Reichlin M. Antibodies to carbonic anhydrase in systemic lupus erythematosus and other rheumatic diseases. Arthritis Rheum 1992;35:73–82. 23. Gordon SC, Quattrociocchi-Longe TM, Khan BA, Kodali VP, Chen J, Silverman AL, Kiechle FL. Antibodies to carbonic anhydrase in patients with immune cholangiopathies. Gastroenterology 1995; 108:1802–1809. 24. Homma M, Tojo T, Akizuki M, Yamagata H. Criteria for Sjo¨gren’s syndrome in Japan. Scand J Rheumatol 1986;61:26–27. 25. Schaffner F. Primary biliary cirrhosis. In: Berk JE, ed. Bockus’
/ 5e0d$$0019
04-18-96 16:54:11
gasas
26.
27. 28.
29. 30. 31.
32.
33.
34.
35.
36.
37. 38.
39.
gastroenterology. Volume 5. 4th ed. Philadelphia: Saunders, 1985:3150–3176. Tan EM, Cohen AS, Fries JF, Masi AT, Mcshane DJ, Rothfield NF, Schaller JG, Talal N, Winchester RJ. The 1982 revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum 1982;25:1271–1277. Engvall E. Enzyme immunoassay ELISA and EMIT. Methods Enzymol 1980;70:419–439. Smith HW, Finkelstein N, Aliminosa L, Crawford B, Graber M. The renal clearances of substituted hippuric acid derivatives and other aromatic acids in dog and man. J Clin Invest 1945;24: 388–404. Kasugai T, Kano N, Kizu M. Manometric endoscopic retrograde pancreatocholangiography. Am J Dig Dis 1974;19:485–502. Moutsopoulos HM. Sjo¨gren’s syndrome: autoimmune epithelitis. Clin Immunol Immunopathol 1994;72:162–165. Tashian RE, Hewett-Emmett D. The value of deficiencies of human carbonic anhydrase isozymes in understanding their cellular roles. Ann NY Acad Sci 1984;429:262–275. Parkkila S, Parkkila A-K, Juvonen T, Rajaniemi H. Distribution of the carbonic anhydrase isozymes I, II, and VI in the human alimentary tract. Gut 1994;35:646–650. Ben-Ari Z, Dhillon AP, Sherlock S. Autoimmune cholangiopathy: part of the spectrum of autoimmune chronic active hepatitis. Hepatology 1993;18:10–15. Kannan KK, Notstrand B, Fridborg K, Lo¨vgren S, Ohlsson A, Petef M. Crystal structure of human erythrocyte carbonic anhydrase B: three-dimensional structure at a nominal-2.2-A˚ resolution. Proc Natl Acad Sci USA 1975;72:51–55. Foster RE II. Carbonic anhydrase and the cartid body. In: Data PG, Acker H, Lahiri S, eds. Neurobiology and cell physiology of chemoreception. New York: Plenum, 1993:137–147. Spicer SS, Sens MA, Tashian RE. Immunocytochemical demonstration of carbonic anhydrase in human epithelial cells. J Histochem Cytochem 1982;30:864–873. Carter MJ. Carbonic anhydrase: isozyme, properties, distribution, and functional significance. Biol Rev 1972;47:465–513. Carter MJ, Parsons DS. The isozymes of carbonic anhydrase: tissue, subcellular distribution and functional significance, with particular reference to the intestinal tract. J Physiol 1971;215: 71–94. Nishimori I, Morita M, Kino J, Onodera M, Nakazawa Y, Okazaki K, Yamamoto Y, Yamamoto Y. Pancreatic involvement in patients with Sjo¨gren’s syndrome and primary biliary cirrhosis. Int J Pancreatol 1995;17:47–54.
Received September 25, 1995. Accepted January 4, 1996. Address requests for reprints to: Isao Nishimori, M.D., First Department of Internal Medicine, Kochi Medical School, Nankoku, Kochi 783, Japan. e-mail: [email protected]. Fax: (81) 888-80-2244. Supported in part by a grant from the Japanese Ministry of Health and Welfare for study of intractable pancreatic disease and by grants DK 46589 and CA 36727 from the National Institutes of Health. The authors thank Drs. Yasutake Yamamoto and Satoru Tamura for their help and advice.
WBS-Gastro
Serum Antibodies to Carbonic Anhydrase I and II in Patients With Idiopathic Chronic Pancreatitis and Sjo¨gren’s Syndrome JUNKO KINO–OHSAKI,* ISAO NISHIMORI,* MASANORI MORITA,* KAZUICHI OKAZAKI,* YASURO YAMAMOTO,* SABURO ONISHI,* and MICHAEL A. HOLLINGSWORTH‡ *First Department of Internal Medicine, Kochi Medical School, Kochi, Japan; and ‡Eppley Cancer Institute, University of Nebraska Medical Center, Omaha, Nebraska
Background & Aims: Patients with idiopathic chronic pancreatitis and Sjo¨gren’s syndrome show immune responses against a 60-kilodalton protein isolated from human pancreas extracts. Monoclonal antibody SP3-1, raised against the 60-kilodalton protein, reacts with ductal cells in several exocrine organs and cross-reacts with human carbonic anhydrase II. The present study evaluated serum from patients with these diseases for antibodies to human carbonic anhydrase I and II. Methods: An enzyme-linked immunosorbent assay was used to quantify serum antibody against carbonic anhydrase I and II. Results: Serum antibodies against carbonic anhydrase I and II were detected in 7 and 11 of 33 patients with idiopathic chronic pancreatitis, respectively, and in 8 and 13 of 21 patients with Sjo¨gren’s syndrome, respectively. The positive prevalence rates of patients with antibodies to carbonic anhydrase II were significantly different among patients with idiopathic chronic pancreatitis and Sjo¨gren’s syndrome compared with normal patients (1 of 21). There were no significant differences in the prevalence rates of patients with alcoholic chronic pancreatitis (3 of 20), gallstone-related chronic pancreatitis (0 of 7), and primary biliary cirrhosis (1 of 11). Conclusions: The results suggest that carbonic anhydrase II is one candidate target antigen recognized during the autoimmune pathophysiology observed in idiopathic chronic pancreatitis and Sjo¨gren’s syndrome.
B
etween 30% and 40% of patients with chronic pancreatitis (CP) have no apparent underlying cause of their disease.1 These patients are considered to have idiopathic CP, and several etiologic factors have been proposed, such as dysfunction of the sphincter of Oddi and ischemia.2 Autoimmunity has also been suggested as a possible etiologic factor of idiopathic CP, and various immunologic alterations have been described in patients with CP.3 – 6 However, the autoimmune mechanisms that contribute to the pathogenesis of CP remain poorly understood. / 5e0d$$0019
04-18-96 16:54:11
gasas
It has been well documented that CP is occasionally observed as a complication in patients with Sjo¨gren’s syndrome (SjS) and primary biliary cirrhosis (PBC).7,8 Based on clinical findings, the concept of a disease complex, autoimmune exocrinopathy9 or dry gland syndrome,10 has been put forth to describe patients with the above-mentioned diseases. It is hypothesized that these diseases may be manifestations in distinct organs of an autoimmune reaction against a common antigen that is expressed by the ductal epithelial cells of exocrine organs. It has been reported that patients with SjS and CP have serum antibodies that react with duct cells in salivary glands and pancreas11 – 15; however, the molecular nature of this antigen has not been determined. We reported previously that patients with SjS and idiopathic CP showed cellular and humoral immune responses against a 60-kilodalton protein that was isolated from human and porcine pancreas extracts.16 – 18 Monoclonal antibody SP3-1 was raised against the 60-kilodalton protein and was found to react with the ductal cells in several exocrine organs, including pancreas, salivary gland, liver (biliary duct), kidney (distal renal tubule), and esophageal glands.19 Recently, we determined that monoclonal antibody SP3-1 reacts with human carbonic anhydrase II (CAII), a finding that is consistent with the distribution of SP3-1 reactive tissues described above.20 Moreover, we reported that sialoadenitis was successfully induced in mice bearing the H-2s and H-2u major histocompatibility complex haplotypes by intradermal immunization with human CAII and adjuvant.20 Recently, it has been reported by several investigators that patients with SjS, systemic lupus erythematosus (SLE), and autoimmune cholangitis have serum autoantibodies to Abbreviations used in this paper: BT-PABA, N-benzoyl-L-tyrosyl-pamino-benzoic acid; CAI, carbonic anhydrase I; CAII, carbonic anhydrase II; CP, chronic pancreatitis; ERP, endoscopic retrograde pancreatography; PBC, primary biliary cirrhosis; SjS, Sjo¨gren’s syndrome; SLE, systemic lupus erythematosus. 䉷 1996 by the American Gastroenterological Association 0016-5085/96/$3.00
WBS-Gastro
1580 KINO–OHSAKI ET AL.
GASTROENTEROLOGY Vol. 110, No. 5
CAII.21 – 23 These observations prompted us to determine whether patients with idiopathic CP have autoantibodies to CAII. In the present study, sera from patients with CP and with SjS, PBC, and SLE were evaluated for autoantibodies to CAII and carbonic anhydrase I (CAI).
Materials and Methods Patients Sera from 74 patients with CP were studied. The diagnosis of CP was based on one of the following findings: decreased pancreatic excretory function as shown by the secretin test; pancreatic stones; dilatation, stenosis, or cyst formation of the main pancreatic duct and its branches shown by endoscopic retrograde pancreatography (ERP); or histologically proven CP. Patients with CP had no history of acute pancreatitis and were divided into the following five groups according to etiology: 20 patients with alcoholic CP, 7 patients with gallstone-related CP, 7 patients with SjS-associated CP, 4 patients with PBC-associated CP, and 33 patients with idiopathic CP. Patients with idiopathic CP had no obvious cause of CP, such as alcoholism, gallstones, or other identifiable etiologic factors including associated autoimmune diseases. In addition, 21 patients with SjS and 11 patients with PBC, who were not complicated by CP, were studied. SjS and PBC were diagnosed according to the previously defined criteria, respectively.24,25 Twelve patients with SLE were diagnosed according to the American Rheumatism Association 1982 revised criteria for the classification of SLE,26 and 21 normal controls were also studied.
Enzyme-Linked Immunosorbent Assay for CAI and CAII Antibody Human CAI and CAII (electrophoretically purified from erythrocytes) were purchased from Sigma Chemical Co. (St. Louis, MO). Serum antibodies to CAI and CAII were quantified by enzyme-linked immunosorbent assay using a standard method with minor modifications.27 Briefly, microtiter plates were coated with 50 mL of 5 mg/mL CAI or CAII overnight at 4⬚C. After five washes with phosphate-buffered saline, the plates were coated with 2% bovine serum albumin and 5% goat serum in phosphate-buffered saline overnight at 4⬚C and then incubated with 50 mL of diluted patient serum (1:25) overnight at 4⬚C. After five washes with phosphatebuffered saline containing 0.05% Tween 20 (Nacalai, Kyoto, Japan), the plates were incubated with 50 mL of goat antihuman immunoglobulin G antibody conjugated with horseradish peroxidase (Sigma Chemical Co.) at room temperature for 2 hours. After five washes with phosphate-buffered saline containing 0.05% Tween 20, bound reactants were detected by 15-minute incubation with 50 mL of 50 mg/mL 2.2-azinobis diammonium salt (Nacalai). The absorbance was determined at 405 nm. Control wells that were not coated with CAI or CAII were also used for enzyme-linked immunosorbent assay of each serum studied. All assays were performed in triplicate,
/ 5e0d$$0019
04-18-96 16:54:11
gasas
and the specific binding of serum antibody to CAI or CAII was calculated as follows: the average absorbance of the antigen-coating wells minus the average absorbance of control wells.
Specific Inhibition of Serum Anti-CAI or Anti-CAII Antibody The specificity of the enzyme-linked immunosorbent assay system was assessed by absorption of serum antibody with CAI or CAII. Briefly, microtiter wells were precoated with 50 mL of 0–80 mg/mL CAII overnight at 4⬚C and further incubated with 2% bovine serum albumin and 5% goat serum in phosphate-buffered saline overnight at 4⬚C. Aliquots of sera diluted to 1:25 in phosphate-buffered saline were incubated in the wells at room temperature for 2 hours. The absorbed samples were tested for anti-CAI or anti-CAII antibody by enzyme-linked immunosorbent assay as described above.
Assessment of Clinical Findings in Patients With CP With Serum Anti-CAII Antibody Clinical findings in patients with CP with serum antiCAII antibodies were evaluated and compared with patients with CP without anti-CAII antibodies. The following factors were assessed: age, sex, serum pancreatic enzyme levels (amylase, elastase I, and trypsin), pancreatic exocrine function studied by N-benzoyl-L-tyrosyl-p-amino-benzoic acid (BT-PABA) excretion test, and ERP. Non–organ specific immunologic parameters, including antinuclear antibody and rheumatic factor, and serum concentrations of g-globulin and immunoglobulins G, A, and M were also evaluated. Serum amylase levels were measured using a commercially available kit with ethylidene-(G7)-para-nitrophenylmaltoheptaoside as a substrate (Amylase EPS; Boehringer, Tokyo, Japan; normal range, 40– 130 IU/L). Elastase I and trypsin were quantified by using commercially available kits (Dinabot, Tokyo, Japan, and Hoechst, Hounslow, England; normal range, 80–410 ng/dL and 110–460 ng/mL, respectively). BT-PABA excretion tests were performed according to previously described methods.28 ERP findings were graded into the following four groups according to the criteria defined by Kasugai et al.29: normal, minimal-change pancreatitis, moderate-change pancreatitis, and advanced-change pancreatitis. Antinuclear antibody and rheumatic factor were studied by immunofluorescence methods using commercially available kits as follows: antinuclear antibody, indirect immunofluorescence using HEp-2 cell as substrate (Fluoro Hepana Test; MBL, Nagoya, Japan); and rheumatic factor, indirect agglutination using rabbit denatured immunoglobulin G as antigen (Se-Rodia-RA, Fujirebio, Tokyo, Japan). Serum g-globulin concentration was estimated based on serum total protein concentration and whole serum electrophoresis quantified on a densitometer. Serum concentrations of immunoglobulin subclass were measured by rate nephelometry (Beckman, Galway, Ireland).
WBS-Gastro
May 1996
ANTI–CAII ANTIBODY IN CHRONIC PANCREATITIS 1581
Figure 1. Enzyme-linked immunosorbent assay of serum antibody against CAII. Absorbance was determined at 405 nm (A405). Dotted line indicates mean absorbance / 2SD of normal controls (0.341). P õ 0.05, significant difference in comparisons of mean { SD value between patient groups and normal controls.
Statistical Analysis Fisher’s Exact Probability Test with Yates’ correction was used for statistical analyses of the antibody prevalence rate, patient’s profile, ERP finding, antinuclear antibody, and rheumatic factor. Student’s t test was performed in comparisons of mean values of serum anti-CAII and anti-CAI antibody titers, pancreatitis enzyme levels, and g-globulin and immunoglobulin subclass concentrations. Pearson’s correlation coefficient and Fisher’s Exact Test were used for an analysis of correlation between serum anti-CAII and anti-CAI antibody titer. Statistical significance was defined as P values of õ0.05. All data were expressed as mean { SD.
body-positive prevalence rate in patients with idiopathic CP and SjS (P õ 0.05). Significant differences in mean values of serum antibody titers were also observed in patients with idiopathic CP (0.314 { 0.125), SjS (0.360 { 0.068), and SjS-associated CP (0.313 { 0.052) compared with normal controls (0.247 { 0.047) (P õ 0.05). There was no significant difference in mean { SD values between normal controls and patients with alcoholic CP (0.272 { 0.094), gallstone-related CP (0.237 { 0.040), PBC (0.273 { 0.134), PBC-associated CP (0.313 { 0.053), and SLE (0.287 { 0.089). Absorption of serum anti-CAII antibody in 2 patients with idiopathic CP and 3 patients with SjS reduced absorbance values to less than the mean / 2SD values determined for normal controls (0.341) (Figure 2). Preabsorption had no effect on absorbance for normal controls, suggesting that these normal controls had no detectable anti-CAII antibody. This finding confirms the specificity of the enzyme-linked immunosorbent assay system used in the present study for measuring anti-CAII antibodies in sera and supports our use of a cutoff point that is 2SD greater than mean values determined for normal controls. Anti-CAI Antibodies Anti-CAI antibody was also quantified in those disease groups that showed a relatively high prevalence
Results Anti-CAII Antibodies Levels of serum anti-CAII antibody in patients and normal controls were quantified by enzyme-linked immunosorbent assay (Figure 1). The mean { SD of absorbance values for normal controls was 0.247 { 0.047 (n Å 21). In the present study, positive results were arbitrarily defined as absorbance values of ú0.341, which corresponds to a value that is the mean / 2SD of normal control values. Among patients with CP, positive results were obtained in 11 of 33 patients with idiopathic CP (33%), 3 of 20 patients with alcoholic CP (15%), 2 of 7 patients with SjS-associated CP (29%), 1 of 4 patients with PBC-associated CP (25%), and 0 of 7 patients with gallstone-related CP (0%). Thirteen of 21 patients with SjS (62%), 1 of 11 patients with PBC (9%), 3 of 12 patients with SLE (25%), and 1 of 21 normal controls (4.8%) showed positive results. Compared with normal controls, there was a significant difference in the anti/ 5e0d$$0019
04-18-96 16:54:11
gasas
Figure 2. Absorption of serum antibody with CAII in enzyme-linked immunosorbent assay as described in Materials and Methods. Patients with idiopathic CP (●; n Å 2) and with SjS (䊊; n Å 3) showed inhibition of the anti-CAII antibody down to the level of normal controls (䊏; n Å 2) in an antigen concentration–dependent manner. A405, Absorbance at 405 nm.
WBS-Gastro
1582 KINO–OHSAKI ET AL.
GASTROENTEROLOGY Vol. 110, No. 5
antibody titer was evaluated using Pearson’s Correlation Coefficient and Fisher’s Tests. In patients with SjS, the presence of anti-CAII antibody showed a fair correlation with the presence of anti-CAI antibody (r Å 0.55627; P õ 0.01; Figure 4A). In patients with idiopathic CP, however, there was no significant correlation between anti-CAI antibody and anti-CAII antibody titers (Figure 4B). The patients could be divided into the following three groups: anti-CAI antibody positive and anti-CAII antibody negative (anti-CAI Ab/), anti-CAI antibody negative and anti-CAII antibody positive (anti-CAII Ab/), and negative for both antibodies. One patient had both antibodies but with marginally positive values. Two pairs of patients were sampled among anti-CAI Ab/ and anti-CAII Ab/ groups: idiopathic CP-1 and -2 and idiopathic CP-3 and -4, respectively. The specificity of serum antibody against CAI or CAII in these patients was further studied as described below (Figure 5). Specificity of Serum Antibodies Against CAI or CAII in Patients With Idiopathic CP Figure 3. Enzyme-linked immunosorbent assay of serum antibody against CAI. Absorbance was determined at 405 nm (A405). Dotted line indicates mean absorbance / 2SD of normal controls (0.146). P õ 0.05, significant difference in comparisons of mean { SD value between patient groups and normal controls.
rate of anti-CAII antibody, including idiopathic CP, SjS, and SjS-associated CP (Figure 3). Positive results were arbitrarily set as an absorbance of ú0.146, corresponding to the mean { 2SD of values for normal controls. AntiCAI antibody was detected in 7 of 33 patients with idiopathic CP (21%), 8 of 21 patients with SjS (38%), and 2 of 7 patients with SjS-associated CP (29%). There was a significant difference in the antibody-positive prevalence rate in patients with SjS compared with normal controls (1 of 20; 5%) (P õ 0.05). Significant differences in mean values of the serum antibody titers were observed in patients with SjS (0.147 { 0.061) and SjS-associated CP (0.125 { 0.063) compared with normal controls (0.090 { 0.028) (P õ 0.05). There was no significant difference in mean values between normal controls and patients with idiopathic CP (0.119 { 0.084). It is noteworthy that 2 patients with idiopathic CP showed the highest antibody titers among all patients studied, including patients with SjS. Correlation Between Anti-CAII Antibodies and Anti-CAI Antibodies In patients with idiopathic CP and SjS, the correlation between anti-CAII antibody titer and anti-CAI / 5e0d$$0019
04-18-96 16:54:11
gasas
The specificity of serum antibodies in patients from the anti-CAI Ab/ and anti-CAII Ab/ groups was studied by absorption of the sera with either CAI or CAII as described in Materials and Methods (Figure 5). Serum anti-CAI antibodies in ICP-1 and ICP-2 patients (antiCAI Ab/ group; see Figure 4B) were completely absorbed by incubation with CAI, but they were not absorbed with CAII (Figure 5A). In contrast, serum anti-CAII
Figure 4. Correlation between anti-CAI antibody titer and anti-CAII antibody titer in patients with SjS and with idiopathic CP. (A ) In patients with SjS, anti-CAII antibody showed a relatively good correlation to anti-CAI antibody (r Å 0.55627; P õ 0.01). (B ) In patients with idiopathic CP (ICP), there is no significant correlation between two antibodies, and it was possible to divide patients into the following three groups: anti-CAI antibody positive and anti-CAII antibody negative, anti-CAI antibody negative and anti-CAII antibody positive, and negative for both antibodies. Only 1 patient showed both antibody positive with a marginally positive value. Dotted lines indicate cutoff value of absorbance (A405) in both antibodies as described in Results and in the legends to Figures 1 and 3.
WBS-Gastro
May 1996
ANTI–CAII ANTIBODY IN CHRONIC PANCREATITIS 1583
Figure 5. Absorption of serum antibody with CAI (䊏) or CAII (䊊) in enzyme-linked immunosorbent assay as described in Materials and Methods. (A ) Serum anti-CAI antibodies in ICP-1 and ICP-2 patients (see Figure 4B ) were absorbed by an incubation with CAI down to the level of normal controls (NC) but were not absorbed by CAII at all. (B ) Serum anti-CAII antibodies in ICP-3 and ICP-4 patients (see Figure 4B ) absorbed by an incubation with CAII, but these were not absorbed by CAI. A405, Absorbance at 405 nm.
antibodies in ICP-3 and ICP-4 patients (anti-CAII Ab/ group; see Figure 4B) were absorbed by an incubation with CAII but not with CAI (Figure 5B). Although absorption of anti-CAII antibody with CAII in ICP-3 and ICP-4 patients did not reduce the absorbances to background values, the absorbances were less than the mean / 2SD of values for normal controls (0.341). These findings show that patients in each group, anti-CAII Ab/ and anti-CAI Ab/ groups, have antibodies specific for CAI or CAII, respectively. Clinical Findings in Patients With CP With Serum Anti-CAII Antibodies Regarding clinical parameters described in Materials and Methods, patients with CP with serum antiCAII antibodies were compared with patients with CP without the antibody (Table 1). There was no significant difference in age or sex among both groups. No significant differences were observed in serum pancreatic enzyme levels, exocrine pancreatic function studied by the BT-PABA excretion test, and pancreatic ductal morphology classified by ERP among the two groups. Antinuclear antibody was observed with a significantly higher prevalence rate in patients with CP with anti-CAII antibody, but rheumatic factor was not observed with a higher rate. Regarding total and subclass immunoglobulin concentrations, only immunoglobulin A levels were signifi/ 5e0d$$0019
04-18-96 16:54:11
gasas
cantly greater in patients with CP with anti-CAII antibody; however, the values were within the normal range. Despite a finding of no significant difference in pancreatic ductal morphology when a classical grading of pancreatography was used, it is noteworthy that patients with CP with anti-CAII antibody often showed an interesting pancreatic ductal configuration as shown in Figure 6. The main pancreatic duct and branch of the duct were narrowed in segments and were irregularly shaped. This abnormality was rarely observed in patients with CP without anti-CAII antibodies, especially among patients with alcoholic CP. It was difficult to draw conclusions from clinical findings of patients with idiopathic CP with anti-CAI antibody because of the small number of patients; only 2 patients were obviously positive for anti-CAI antibodies.
Discussion One significant finding of this study is the prevalence rate of anti-CAII antibodies in sera of patients with idiopathic CP (11 of 33 patients). In contrast, only 3 of 20 patients with alcoholic CP and 0 of 7 patients with gallstone-related CP had anti-CAII antibodies in their sera. These findings suggest that a significant percentage of patients with idiopathic CP may have some form of autoimmune CP. The fact that significantly lower percentages of patients with alcoholic CP and gallstonerelated CP had anti-CAII antibodies implies that production of anti-CAII antibody is not a result of pancreas tissue injury. This suggests that autoimmune responses to CAII should be further investigated as a possible etiologic factor of CP, along with long-term alcohol consumption and the presence of gallstones. A second significant finding is that 62% of patients with SjS without CP (13 of 21 patients) and 29% of patients with SjS-associated CP (2 of 7 patients) have anti-CAII antibodies in their sera. Together, 54% of patients with SjS show anti-CAII antibodies. Thus, antiCAII antibodies are detected in some but not all patients with SjS. This finding is consistent with previous studies by Inagaki et al.21 and Itoh and Reichlin,22 who reported prevalence rates for serum anti-CAII antibody in patients with SjS of 12.5% and 20.8%, respectively. Interestingly, antibodies reactive with pancreatic ductal cells have been reported in patients with SjS.11,12 It has been proposed that a disease complex exists that includes patients with SjS, CP, PBC, primary sclerosing cholangitis, and renal tubular acidosis, and this disease complex has been given various names over the past 15 years, including autoimmune exocrinopathy,9 dry gland syndrome,10 or autoimmune epithelitis.30 It is hypotheWBS-Gastro
1584 KINO–OHSAKI ET AL.
GASTROENTEROLOGY Vol. 110, No. 5
Table 1. Clinical Findings in Patients With CP With or Without Serum Anti-CAII Antibody Anti-CAII antibody
Age (yr) Sex (M/F) Serum pancreatic enzyme Amylase (IU/L) Elastase I (ng/dL) Trypsin (ng/mL) BT-PABA excretion test (%) ERP Advanced-change pancreatitis Moderate-change pancreatitis Minimal-change pancreatitis Normal Antinuclear antibody Rheumatic factor g-Globulin (g/dL) Immunoglobulin G (mg/dL) Immunoglobulin A (mg/dL) Immunoglobulin M (mg/dL)
Positive
Negative
Significance of differencea
65.5 { 11.1 11/4
59.6 { 11.5 33/15
NS NS
150.0 { 49.1 344 { 297 508.7 { 232.3 80.8 { 16.1
200 { 150.7 437 { 263.8 579 { 279.7 79.2 { 15.3
NS NS NS NS
2/12 (16.7%) 6/12 (50.0%) 3/12 (25.0%) 1/12 (8.3%) 6/13 (46.2%) 1/12 (8.3%) 1.50 { 0.25 1465.9 { 329.0 319.8 { 130.0 135.4 { 56.2
2/29 (6.9%) 10/29 (34.5%) 12/29 (41.3%) 5/29 (97.2%) 2/24 (8.3%) 0/26 (0%) 1.27 { 0.40 1303.0 { 531.9 232.5 { 83.4 162.8 { 96.5
NS NS NS NS P õ 0.05 NS NS NS P õ 0.05 NS
a
Significance of difference in comparison of mean value or of positive prevalence rate studied in patients with CP with and without serum anti-CAII antibody.
sized that one common mechanism of pathogenesis of these diseases is an autoimmune reaction against a common antigen that is expressed in the ductal epithelial cells of exocrine organs. CAII is a reasonable target anti-
Figure 6. ERP of 2 patients with idiopathic CP with anti-CAII antibody. Note irregular configuration including segmental narrowing but no or mild dilatation of main pancreatic duct and its branches (arrowhead ).
/ 5e0d$$0019
04-18-96 16:54:11
gasas
gen for such an autoimmune disease complex because it is expressed by duct cells of exocrine organs included in these syndromes, including pancreas, salivary gland, liver (bile duct), and kidney (renal tubule).31,32 In addition to the data described above, several experimental findings support the hypothesis that autoimmune responses to CAII may be involved in the pathogenesis of this complex of related diseases. Recently, experimental autoimmune sialoadenitis was successfully induced in SJL/JCr (H-2s) and PL/J (H-2u) mice by immunization with CAII,20 suggesting that major histocompatibility complex–restricted immune responses to CAII are associated with the pathogenesis in this model mouse for SjS. In addition, patients with other autoimmune disease have been reported to have anti-CAII antibodies in their sera.21 – 23 These include patients with SLE21,22 and autoimmune cholangitis.23 In the study of patients with autoimmune cholangitis, 83.3% of patients had anti-CAII antibodies. In contrast, only 8.3% of patients with PBC had anti-CAII antibodies.23 It was proposed that autoimmune cholangitis is a novel disease entity in patients with PBC-compatible histological findings in liver specimens who are negative for antimitochondrial antibodies and positive for antinuclear antibody.33 The present study also failed to detect anti-CAII antibodies in patients with PBC (1 of 11 patients; 9.1%). One interpretation of these results is that the liver phenotype that presents during autoimmune exocrinopathies directed toward CAII is autoimmune cholangitis instead of PBC. To further investigate the specificity of the anti-CAII WBS-Gastro
May 1996
ANTI–CAII ANTIBODY IN CHRONIC PANCREATITIS 1585
antibodies observed in patients with CP and SjS, we investigated the reactivity of these patients’ sera with CAI. Carbonic anhydrases are zinc metalloenzymes that catalyze the reversible reaction CO2 / H2O Å HCO30 / H/, the hydration of certain aldehydes, and the hydrolysis of certain esters.34 There are at least eight isozymes of carbonic anhydrase that have been studied.35 The similarity in amino acid sequence between CAI and CAII is 60.2%. In contrast to CAII, which is widely expressed in human cells including erythrocytes, exocrine glands described above, and the gastrointestinal tract including stomach, small intestine, and colon,36 – 38 CAI is expressed mainly in erythrocytes. In the present study, 7 of 33 patients with idiopathic CP, 8 of 21 patients with SjS, and 2 of 7 patients with SjS-associated CP had anti-CAI antibodies in their sera. It seemed likely that the anti-CAI antibodies detected in patients with SjS were cross-reactive with the anti-CAII antibodies because the anti-CAI antibody titer in these patients showed good correlation with anti-CAII antibody titer. In patients with idiopathic CP, however, anti-CAII antibodies and anti-CAI antibodies were independently observed in different patients. Moreover, the explanation that some antiCAI antibodies are cross-reactive with similar amino acid sequences on CAII, whereas others are not, is not supported by the data presented in Figure 5. In one previous report, serum anti-CAI antibody and anti-CAII antibody were concomitantly or independently observed in patients with SjS and SLE.22 The exact reason for the independent occurrence of both antibodies is unknown, because expression of CAI is mostly restricted to erythrocytes in human tissues as described above. The fact that anti-CAII antibodies are detected in only a portion of patients with SjS, CP, and related diseases makes it unlikely that CAII is the only antigen recognized in these diseases. Instead, it is likely that there are several target antigens and that different combinations of diseases could result from immune reactions against the different antigens. In this light, it is noteworthy that some patients with CP with anti-CAII antibody were positive for antinuclear antibody. The prevalence rate of antinuclear antibody (46.2%) for these patients was significantly greater than patients without the antibody (8.3%). Interestingly, all 3 patients with alcoholic CP with anti-CAII antibody were also positive for antinuclear antibody. It remains possible that autoimmune reactions to CAII have some effect on the pathophysiology of these patients during acceleration or modification of pancreatic tissue injury by alcohol. Pancreatic ductal morphology evaluated by ERP is also noteworthy as a clinical finding in patients with CP with anti-CAII anti/ 5e0d$$0019
04-18-96 16:54:11
gasas
body. An irregular configuration, including segmental narrowing but no or mild dilatation of the pancreatic duct system, that was observed in anti-CAII–positive patients with CP (Figure 6) is rarely observed in patients with alcoholic CP. In patients with alcoholic CP, the pancreatic duct usually shows a marked dilatation and occasional calcification. We previously reported that patients with SjS and PBC showed similar ERP findings.39 This difference in ERP findings between patients with alcoholic CP and patients with anti-CAII antibody–positive CP may suggest that different pathophysiology contributes to the disease process in these patients. We previously reported that patients with idiopathic CP have antibodies and show lymphocyte-proliferative responses against a 60-kilodalton molecule isolated from human pancreas extracts.16 – 18 The 60-kilodalton pancreatic molecule is hypothesized to be a homodimer of CAII based on the following previous results: a monoclonal antibody against this antigen (SP3-1) reacts with CAII; the histological distribution of the antigen is consistent with the distribution of CAII; and other findings.19,20 The present study provides further support for this hypothesis by showing the presence of serum antibodies against human CAII in a significant percentage of patients with idiopathic CP and SjS. These results indicate that immune sensitization against CAII or its derived peptides has occurred in some patients with idiopathic CP and SjS. Further studies, including an analysis of T-cell responses against CAII, are warranted to further investigate the pathogenesis of these diseases.
References 1. Steer ML, Waxman I, Freedman S. Chronic pancreatitis. N Engl J Med 1995;332:1482–1490. 2. Markes IN, Bank S. Chronic pancreatitis: etiology, clinical aspects, and medical management. In: Berk JE, ed. Bockus’ gastroenterology. Volume 6. 4th ed. Philadelphia: Saunders, 1985: 4020–4040. 3. Bank S, Novis BH, Petersen E, Dowdle E, Marks IN. Serum immunoglobulins in calcific pancreatitis. Gut 1973;14:723–725. 4. Antal L, Ka´vai M, Szabo´ G, Sonkoly I, Pa´lo´czi K, Szegedi G, Sa´py P, Va´rhelyi I. Immunological investigations in acute and chronic human pancreatitis. Digestion 1980;20:100–105. 5. Lankisch PG, Koop H, Seelig R, Seelig HP. Antinuclear and pancreatic acinar cell antibodies in pancreatic diseases. Digestion 1981;21:65–68. 6. Bovo P, Mirakian R, Merigo F, Angelini G, Cavallini G, Rizzini P, Bottazzo GF, Scuro LA. HLA molecule expression on chronic pancreatitis specimens: is there a role for autoimmunity? A preliminary study. Pancreas 1987;2:350–356. 7. Sheikh SH, Shaw-Stiffel TA. The gastrointestinal manifestations of Sjo¨gren’s syndrome. Am J Gastroenterol 1995;90:9–14. 8. Schaffner F, Bach N. Gastrointestinal syndromes in primary biliary cirrhosis. Semin Liver Dis 1988;8:263–271. 9. Strand V, Talal N. Advances in the diagnosis and concept of Sjo¨gren’s syndrome (autoimmune exocrinopathy). Bull Rheum Dis 1980;30:1046–1052.
WBS-Gastro
1586 KINO–OHSAKI ET AL.
GASTROENTEROLOGY Vol. 110, No. 5
10. Epstein O, Chapman RWG, Lake-Bakaar G, Foo AY, Rosalki SB, Sherlock S. The pancreas in primary biliary cirrhosis and primary sclerosing cholangitis. Gastroenterology 1982;83:1177–1182. 11. Ludwig H, Schernthaner G, Scherak O, Kolarz G. Antibodies to pancreatic duct cells in Sjo¨gren’s syndrome and rheumatoid arthritis. Gut 1977;18:311–315. 12. Sundkvist G, Lindahl G, Koskinen P, Bolinder J. Pancreatic autoantibodies and pancreatic function in Sjo¨gren’s syndrome. J Intern Med 1991;229:61–66. 13. Macsween RNM, Goudie RB, Anderson JR, Armstrong E, Murray MA, Mason DK, Jasani MK, Boyle JA, Buchanan WW, Williamson J. Occurrence of antibody to salivary duct epithelium in Sjo¨gren’s disease, rheumatoid arthritis, and other arthritides: a clinical and laboratory study. Ann Rheum Dis 1967;26:402–411. 14. Anderson LG, Tarpley TM, Talal N, Cummings NA, Wolf RO, Schall GL. Cellular-versus-humoral autoimmune responses to salivary gland in Sjo¨gren’s syndrome. Clin Exp Immunol 1973;13:335– 342. 15. Rumessen JJ, Marner B, Thorsgaard Pedersen N, Permin H. Autoantibodies in chronic pancreatitis. Scand J Gastroentorol 1985; 20:966–970. 16. Nishimori I, Okazaki K, Yamamoto Y, Morita M, Tamura S, Yamamoto Y. Sensitization against pancreatic antigen in Sjo¨gren’s syndrome and chronic pancreatitis. Digestion 1992;51:71–74. 17. Nishimori I, Okazaki K, Yamamoto Y, Morita M, Tamura S, Yamamoto Y. Specific cellular immune responses to pancreatic antigen in chronic pancreatitis and Sjo¨gren’s syndrome. J Clin Immunol 1993;13:265–271. 18. Nishimori I, Yamamoto Y, Okazaki K, Morita M, Onodera M, Kino J, Tamura S, Yamamoto Y. Identification of autoantibodies to a pancreatic antigen in patients with idiopathic chronic pancreatitis and Sjo¨gren’s syndrome. Pancreas 1994;9:374–381. 19. Okazaki K, Tamura S, Morita M, Nishimori I, Yamamoto Y, Yamamoto Y. Interspecies crossreactive antigen of the pancreatic duct cell prepared by monoclonal antibody. Int J Pancreatol 1989;5: 359–377. 20. Nishimori I, Bratanova T, Toshkov I, Caffrey T, Mogaki M, Shibata Y, Hollingsworth MA. Induction of experimental autoimmune sialoadenitis by immunization of PL/J mice with carbonic anhydrase II. J Immunol 1995;154:4865–4873. 21. Inagaki Y, Jinno-Yoshida Y, Hamasaki Y, Ueki H. A novel autoantibody reactive with carbonic anhydrase in sera from patients with systemic lupus erythematosus and Sjo¨gren’s syndrome. J Dermatol Sci 1991;2:147–154. 22. Itoh Y, Reichlin M. Antibodies to carbonic anhydrase in systemic lupus erythematosus and other rheumatic diseases. Arthritis Rheum 1992;35:73–82. 23. Gordon SC, Quattrociocchi-Longe TM, Khan BA, Kodali VP, Chen J, Silverman AL, Kiechle FL. Antibodies to carbonic anhydrase in patients with immune cholangiopathies. Gastroenterology 1995; 108:1802–1809. 24. Homma M, Tojo T, Akizuki M, Yamagata H. Criteria for Sjo¨gren’s syndrome in Japan. Scand J Rheumatol 1986;61:26–27. 25. Schaffner F. Primary biliary cirrhosis. In: Berk JE, ed. Bockus’
/ 5e0d$$0019
04-18-96 16:54:11
gasas
26.
27. 28.
29. 30. 31.
32.
33.
34.
35.
36.
37. 38.
39.
gastroenterology. Volume 5. 4th ed. Philadelphia: Saunders, 1985:3150–3176. Tan EM, Cohen AS, Fries JF, Masi AT, Mcshane DJ, Rothfield NF, Schaller JG, Talal N, Winchester RJ. The 1982 revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum 1982;25:1271–1277. Engvall E. Enzyme immunoassay ELISA and EMIT. Methods Enzymol 1980;70:419–439. Smith HW, Finkelstein N, Aliminosa L, Crawford B, Graber M. The renal clearances of substituted hippuric acid derivatives and other aromatic acids in dog and man. J Clin Invest 1945;24: 388–404. Kasugai T, Kano N, Kizu M. Manometric endoscopic retrograde pancreatocholangiography. Am J Dig Dis 1974;19:485–502. Moutsopoulos HM. Sjo¨gren’s syndrome: autoimmune epithelitis. Clin Immunol Immunopathol 1994;72:162–165. Tashian RE, Hewett-Emmett D. The value of deficiencies of human carbonic anhydrase isozymes in understanding their cellular roles. Ann NY Acad Sci 1984;429:262–275. Parkkila S, Parkkila A-K, Juvonen T, Rajaniemi H. Distribution of the carbonic anhydrase isozymes I, II, and VI in the human alimentary tract. Gut 1994;35:646–650. Ben-Ari Z, Dhillon AP, Sherlock S. Autoimmune cholangiopathy: part of the spectrum of autoimmune chronic active hepatitis. Hepatology 1993;18:10–15. Kannan KK, Notstrand B, Fridborg K, Lo¨vgren S, Ohlsson A, Petef M. Crystal structure of human erythrocyte carbonic anhydrase B: three-dimensional structure at a nominal-2.2-A˚ resolution. Proc Natl Acad Sci USA 1975;72:51–55. Foster RE II. Carbonic anhydrase and the cartid body. In: Data PG, Acker H, Lahiri S, eds. Neurobiology and cell physiology of chemoreception. New York: Plenum, 1993:137–147. Spicer SS, Sens MA, Tashian RE. Immunocytochemical demonstration of carbonic anhydrase in human epithelial cells. J Histochem Cytochem 1982;30:864–873. Carter MJ. Carbonic anhydrase: isozyme, properties, distribution, and functional significance. Biol Rev 1972;47:465–513. Carter MJ, Parsons DS. The isozymes of carbonic anhydrase: tissue, subcellular distribution and functional significance, with particular reference to the intestinal tract. J Physiol 1971;215: 71–94. Nishimori I, Morita M, Kino J, Onodera M, Nakazawa Y, Okazaki K, Yamamoto Y, Yamamoto Y. Pancreatic involvement in patients with Sjo¨gren’s syndrome and primary biliary cirrhosis. Int J Pancreatol 1995;17:47–54.
Received September 25, 1995. Accepted January 4, 1996. Address requests for reprints to: Isao Nishimori, M.D., First Department of Internal Medicine, Kochi Medical School, Nankoku, Kochi 783, Japan. e-mail: [email protected]. Fax: (81) 888-80-2244. Supported in part by a grant from the Japanese Ministry of Health and Welfare for study of intractable pancreatic disease and by grants DK 46589 and CA 36727 from the National Institutes of Health. The authors thank Drs. Yasutake Yamamoto and Satoru Tamura for their help and advice.
WBS-Gastro