Statistics from Altmetric.com
Nishimori et al have recently reported the presence of autoantibodies against carbonic anhydrase IV (anti-CA IV) in patients with autoimmune pancreatitis (Gut 2005;54:274–81). Furthermore, serum antibodies to CA II (anti-CA II) were observed in several autoimmune conditions. We have now investigated the presence of anti-CA IV and anti-CA II in a large series of sera from patients with primary biliary cirrhosis (PBC) and controls.
CA II is known to be expressed in the cytosol of various types of epithelial cells, including those lining bile ducts, renal tubules, and salivary ducts. For this reason, CA II was suggested as a common antigen in conditions characterised by an autoimmune aggression against epithelia. In autoimmune pancreatitis, serum anti-CA II are useful diagnostic tools while in PBC they were first detected by Gordon et al in 5/6 sera from patients with antimitochondrial antibody (AMA) negative PBC.1 Subsequent studies however demonstrated prevalence rates as high as 46% in PBC sera but failed to confirm their specificity for AMA negative sera.2,3 Interestingly, anti-CA II were also shown to inhibit enzyme activity.4
Apart from cytosolic CA II, the CA family also includes a highly active membrane bound enzyme that was coined CA IV.5 Both CA II and CA IV are abundantly expressed in human bile duct epithelial cells. Interestingly, mainly due to the sequence homology between CA II and CA IV and CA IV localisation on cell membranes, Nishimori et al hypothesised that the exposed CA IV might be more immunogenic than cytosolic CA II.
Seventy sera from patients with PBC (60 AMA positive; all anti-hepatitis C virus negative; 63 women; mean age 60 (SD 10) years) who attended our tertiary referral centre were consecutively enrolled in the study. Control sera were obtained from 50 healthy subjects matched with patients for sex and age class (⩽50 v >50 years). All sera were tested by immunoblotting for anti-CA IV and anti-CA II, as previously described.2 Briefly, proteins were denaturated and separated (10 μg/lane) on a 1.5 mm sodium dodecyl sulphate-12%. polyacrylamide gel. Proteins were then transferred onto nitrocellulose (pore size 0.45 mm) using a semi-dry transfer system. The nitrocellulose membrane was cut into 4 mm strips and, after blocking with 5% non-fat milk, all strips were incubated with serum samples diluted 1:100 and 1:200 for anti-CA IV and 1:100 for anti-CA II. Rabbit horseradish peroxidase conjugated antibodies against human immunoglobulins G, A, and M (Dako, Glostrup, Denmark) was diluted 1:1000 and used as secondary antibody. Peroxidase development was obtained with 0.05% 4-chloro-1-naphthol in Tris buffered saline containing 20% methanol and 0.05% H2O2. A rabbit polyclonal antihuman CA IV antisera6 was used as a positive control throughout the study. CA IV antigen and anti-CA IV antisera were provided by Dr William S Sly (St. Louis, Missouri, USA).
Results demonstrated no reactivity against CA IV in any of the PBC or healthy control sera. In contrast, similar to previous reports, anti-CA II antibodies were detected in 6/70 (9%) sera from patients with PBC but were absent in control sera.
In summary, we submit that the hypothesis that antibodies against the membrane bound CA IV may play a role in PBC should be rejected, based on our experimental data on a large series of sera. Our finding may be secondary to a different cellular expression of CA IV in the target organ (that is, pancreatic and bile ducts) but only specific tissue studies can provide these answers. At present, therefore, anti-CA IV should be regarded as specific to autoimmune pancreatitis and research should focus on better defining their possible role in this condition.
Conflict of interest: None declared.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.