CELIAC DISEASE ASSOCIATES WITH TWO DIFFERENT PATTERNS OF ANTIRETICULIN ANTIBODIES

This is an adaptation from the contents of a poster presented at the 24th National Congress of the Spanish Society of Immunology held in Murcia in the year 1998. An abstract in Spanish was published at the same event by J Rosique Román and R Alvarez López.

SUMMARY

Introduction: Rizzetto M and Doniach D (J Clin Pathol 1973 26:841-51. PMID:4587940) reported in 1973 five types of anti-reticulin antibodies (RA) that could be detected in human sera (R1, R2, Rs, Rac and Rkc) mostly through their indirect immunofluorescence (IIF) staining patterns in rat liver, but now other tissues are of common clinical use as well. It had already been established that RA-R1 appear specifically in celiac disease, but the significance of the others remained unclear. In recent times a related specific marker of celiac disease has been described based on testing on primate oesophagus, the IgA antiendomysial antibody (EMA).

Methods: The study includes all sera submitted to our laboratory for study of non-organ specific autoantibodies or RA since January 1996. IIF testing was performed through standard procedures on three rat tissues (stomach, kidney cortex and liver). An ELISA method was used for the measurement of IgA wheat gliadin (AGA) antibodies. IgA antiendomysial antibodies (EMA) were also tested where appropriate.

Results: We describe a pattern of RA antibody different from R1, R2, Rs or Rkc in all three rat tissues. We have found it at 1/160 serum dilution in 44 patients. Of these, 21 (48%) were positive for IgA gliadin antibodies, and 20 (45%) for IgA antiendomysial antibodies.

Conclusion: RA antibodies can be an unexpected finding in routine autoimmunity testing. The pattern of RA antibodies described here is different from R1. If present in high titer it has a definite predictive value for celiac disease, so further diagnostic workup is recommended.

INTRODUCTION

Reticulin antibodies are a very heterogeneous group of antibodies defined by their staining patterns in rodent tissues. Their name comes out of their similarity to some previously described connective tissue fibres, but the term actually encompasses most antibodies to components around renal tubules, muscle fibers, neural or vascular elements, or hepatic sinusoids. Other authors (Storch W B) may describe “antiendothelial” antibodies as a separate group.

RA-R1 make a pattern of coarse reticulin fibre staining around renal tubules, hepatic portal areas and central lobular veins even in low-titer sera. Hepatic sinusoid-associated fibres of characteristic “curly hair” appearance may be observed and is a unique feature. It also stains the stomach submucosal spaces, and the endomysium of smooth muscle fibers. If these are cut transversally a “honeycomb” pattern becomes apparent. These antibodies are specifically associated with untreated celiac disease.

RA-R2 main characteristic is the thinness or delicacy of the fibrils, and their distribution in muscular layers, or blood vessels. They can be seen sparsely within stomach smooth muscle tissue, or within arterioles, thus becoming visible in the kidney or the hepatic portal areas as well. Hepatic central lobular veins may display a fine “radial” pattern.

RA-Rs make the whole staining of hepatic sinusoids, and typically come in association with strong heterophile proximal renal tubular “brush border” and gastric parietal cell antibodies. Honeycombing and granular staining of the myenteric plexus ganglia of the stomach may also occur.

RA-Rkc stain hepatic sinusoid kuppfer cells. Our impression is that this pattern also comes characteristically with “dots” inside the kidney glomeruli. At other vascular spaces we may find “intravascular clumps” or endothelial linings.

MATERIAL AND METHODS

Patients:

From January 1996 to May 1998, a total of 13500 sera were sent to our laboratory for non-organ specific autoantibodies or RA testing.

We scored the following significant significant RA positivities: RA-R1 in 85 patients, RA-R2 in 130 patients, RA-Rs in 37 patients. We found 57 patients that displayed at 1/160 dilution a pattern that stained the vascular endothelia between the renal tubules, making an overall R1-like appearance there, whereas the stomach showed no honeycombing at the circular smooth muscle layer but a positive staining of the myenteric plexus ganglia and other neural elements. Forty-four of these patients were further tested for gliadin and EMA antibodies.

Antibodies assessments:

RA: IIF procedure was performed on rat stomach, renal cortex and liver tissue slides (Biosystems, Barcelona). The initial dilution of sera for screening was 1/40. Bound antibodies were then stained with FITC-conjugated anti-human G,A,M, kappa, lambda antibody.

 

EMA were assessed by IIF testing on monkey oesophagus slides (Biosystems, Barcelona). Sera were diluted 1/10, incubated, washed and finally revealed by FITC-conjugated anti-human A antibody.

Gliadin antibodies: Each Maxisorp (Nunc) microtiter well was sensitized overnight with 1 µg wheat gliadin (Sigma). After blocking with bovine albumin, the wells were incubated with 100 µl of samples at 1/400 dilution. After washing, bound antibodies were revealed with peroxidase-conjugated anti-human IgA and OPD-based color development for 2-3 minutes. For the calculation of  final concentrations a standard calibration curve was constructed by running in parallel 8 serial dilutions of a stored serum, and the relative units of each sample were then calculated out of their color density readings.

Computation of results: the amount of these antibodies in celiac disease rapidly vanishes depending on treatment or adherence to the diet. For the present study we enter the highest recorded value for each patient.

RESULTS

Description of our reticulin antibody “R1-variant”:

In rat liver you can observe staining of central lobular veins endothelia and adjacent sinusoids, as well as a few round cells within the sinusoids.

In rat renal cortex, there is staining of endothelia of (roughly triangular) vascular spaces between the tubules, as well as small intense dots within glomeruli. Some distal tubules may show cytoplasmic staining with the most intense sera.

The rat stomach displays granular fluorescence of the myenteric plexus ganglia between the smooth muscle circular and longitudinal layers, along with “stars” within the circular layer and “worms” between the mucosal glands which might correspond to neural elements too. There is absolutely no honeycombing unless other RA patterns coexist.

Endothelial images can also be seen within the loose connective tissue between the muscularis mucosae and the circular smooth muscle layer.

Associated antibodies

Of 57 patients with RA-R1, 43 were significantly positive for AGA, 14 negative. Of 60 patients with RA-R1, 58 were EMA positive and 2 negative, a fact that may be due to concomitant IgA deficiency.

Of 44 patients with RA-R1-variant, 21 were significantly positive for AGA, of which 19 were EMA positive. Of the remaining 23, only one was EMA positive.

Of a control group of 14 patients with RA-R2, only one was AGA positive, and EMA negative.

CONCLUSION

This might prove to be a novel antibody response pathway associated with celiac disease. Correlation with genetic or diet modifiers may be worth studying.

The incidental finding in a patient of either R1 or R1-variant antibodies should prompt further diagnostic workup aimed at testing more specific markers of celiac disease. The rat tissue IIF test in our setting has so far proved very effective in detecting unsuspected cases of this disease the prevalence of which is much higher than was previously thought.

RA-variant in rat kidney



RA-variant in rat liver









RA-variant in rat stomach

















RA-R1 in rat kidney



RA-R1 in rat liver



 

RA-R1 in rat stomach