Immunologic and clinical observations of granulocyte-specific antinuclear antibodies.код для вставкиСкачать
Arthritis and Rheumatism official Journal of the ARA Secfion of fheArfhrifisFoundafion JUNE, 1969 VOL. XII, NO. 3 Immunologic and Clinical Observations of Granulocyte-Specific Antinuclear Antibodies By KATHRYN H. SVEC, M.D. The presence of granulocyte-specific antinucleoprotein antibodies was established by absorption tests in 31 sera and unmasked in three additional sera. The majority of patients demonstrated rheumatoid arthritis, but 2 patients had lupoid hepatitis. Differences in fluorescent patterns of staining and degrees of heterologous reactivity delineated two separate granulocyte-specific antinuclear antibodies, one of which showed a significant association with granulocytopenia. A FACTORS associated with systemic lupus in general show diffuse reactivity by immunofluorescence with nuclei of various mammalian tissues. Sera from patients with rheumatoid arthritis, however, tend to show variable reactivity with nuclei from different sources although peripheral blood leukocytes frequently provide the required nuclear antigens.l Faber and associates identified an antibody which gave evidence of reactivity only with granulocytic nuclei and occurred primarily in sera of rheumatoid arthritis patient^.^^^ To establish the specificity of granulocyte-reactive antinuclear antibodies, absorption tests have been found necessary. Elling demonstrated that the granulocyte reactivity in occasional sera was decreased or abolished by absorption with nongranulocytic c e k 4 Although it was suggested that these results might reflect contamination of the absorbent with granulocytes, similar experiments in this laboratory have indicated that apparent granulocyte specifk reactivity may be abolished subsequent to absorption with calf thymus nucleoprotein in the absence of granulocytes. Accordingly, in the following studies this absorptive technic was employed to substantiate the occurrence of granulocyte specifh (GS) antinuclear antibodies in 31 sera. Immunologic characteristics of the GS antibodies and associated clinical aspects were investigated. Based on differences in fluorescent patterns of staining and degrees of heterologous nuclear reactivity, evidence is From the Department of Medicine, Western Reserve University School of Medicine, and University Hospitals, Cleveland, Ohio. Supported b y the Hankins Foundation and the Arthritis Foundation of Northeast Ohio. KATHRYN H. SVEC,M.D.: Assistant Professor of Medicine, Western Reserve University School of Medicine, and University Hospitals, Cleveland, Ohio 44106. Reprint requests should be addressed to Dr. Svec at University Hospitals, 2065 Adelbert Road, Cleveland, Ohio 44106. "UCLEAR ARTHRITIS AND RHEUMATISM, VOL. 12, No. 3 (JUNE 1969) 165 166 KATHRYN H. SVEC presented for the occurrence of two GS from fresh calf thymus at low ionic strength by antibodies, one of which has a significant the method of Chargaff.8 Sera were absorbed with an equal volume of nucleoprotein for 1 hour at association with granulocytopenia. 37OC and 18 hours at 4OC. Control sera were diluted 1/2 with physiologic saline solution. Normal human granulocytes obtained in heparinized peripheral blood were sedimented in 5 per Clinical m a t e d . Sera were obtained from staff patients in the University Hospitals Arthritis Clinic cent dextran, molecular weight 232,000, and sepaand from private patients of rheumatologists in the rated by differential centrifugation.9 Only final Arthritis Unit. Complete clinical and laboratory preparations consisting of 90 per cent or more granulocytes were used. Neutrophils were also obdata were available. All sera were stored at -22OC. The American Rheumatism Association's revised tained from synovial fluid collected by aspiration criteria5 were used for the diagnosis of rheumatoid and open drainage from a patient with arthritis arthritis. Six patients who showed 2 to 10 LE cells and ulcerative colitis. The fluid contained over on two cover glass smears were included in the 60,000 leukocytes per cubic millimeter, 98 to 100 rheumatoid arthritis group because they exhibited per cent of which were polymorphonuclear leukodeforming arthritis with rheumatoid nodules and cytes. Cultures and gram and acid-fast stains of the no evidence of rash, serositis, vasculitis, or renal fluid were negative. A 100 per cent pure preparation of normal huinvolvement. No clinical differences were apparent between these patients and the rheumatoid arthri- man lymphocytes was provided in chyle obtained tis patients who had negative LE cell preparations. from thoracic duct cannulation by Dr. William Systemic lupus was diagnosed on the basis of typi- Falor, Akron, Ohio,lo and made available by Dr. cal multisystem involvement and positive tests for Rune Stjernholm, Cleveland. Prior to storage, the antinuclear factors, frequently including LE cells granulocytes from synovial fluid and the lymphoand/or anti-DNA. Lupoid hepatitis was defined as cytes from chyle were washed three and five times, chronic liver disease with associated antinuclear respectively, with physiologic saline solution. Washed granulocytes from synovial fluid were factors and variable lupus-like extrahepatic sympemployed to prepare subcellular fractions. Nuclei toms. Zmmunofluorescent methods. Antinuclear factors were isolated with 0.2 per cent citric acid by the were detected by the indirect fluorescent antibody method of Nathan and Snapper.11 Staining of the technic with nuclei provided by normal human resulting preparation with Wright's stain revealed spleen imprints. The preparation of the imprints that 95 per cent of the nuclei were devoid of and the sensitivity of these nuclei have been previ- cytoplasm. Nucleoprotein was extracted at low ously described.6 For studies of heterologous re- ionic strength.8 In contrast to the nucleoprotein activity, imprints were similarly prepared from prepared from calf thymus, this product was less rat, rabbit, cow, and dog spleens. When human readily soluble in distilled water and was stored liver and kidney sections were employed, tis- as a suspension. Final preparations of human cells and cell fracsues were frozen and cut in the cryostat at 4 microns. Unfixed air-dried imprints or sections were tions were packed by centrifugation as small washed with phosphate-buffered saline, 0.01 M, measured aliquots in Lusteroid tubes and stored pH 7.2,for 10 min. and incubated for 60 min. at at -22OC. In absorption tests, equal volumes of abroom temperature with sera and fluorescent anti- sorbent and serum at 1/2 dilution were incubated. sera, respectively. The fluorescent conjugates spe- In control tests, sera were diluted 1/4 with phoscific for y globulin and for yG, yM, and yA globu- phate-buffered saline. Treatment of nuclei with enzymes. Serum nulins have been previously described.7 yA antinuclear factors were tested in a triple layer system. clear reactivity was tested with human spleen imTwofold serum dilutions were employed for prints which had been treated with DNA-ase or quantitation of reactivity. Prior to studies of trypsin for 60 min. at room temperature and species specificity and immunoglobulin composition washed twice with buffered saline. DNA-ase ( 1x of antinuclear factors, sera were routinely ab- recrystallized),* 0.25 mg. per milliliter of 0.1 M sorbed with calf thymus nucleoprotein. Fluorescent KH,PO,, pH 7, with 1 &M MgCl, per 0.1 mg. microscopy was performed with a Leitz micro- DNA-ase and trypsin ( 2 x recrystallized)," 0.01 scope, used with an Osram HBO 200 watt mer- mg. per milliliter of phosphate-buffered saline, cury lamp, and fitted with a UG-1 exciter filter 0.01 M, pH 7.8, were employed. and a colorless ultraviolet absorbing bamer filter. Absorption tests. Nucleoprotein was isolated *Worthington Biochemicals, Freehold, N. J. MATERIALSAND METHODS 167 GRANULOCYTE-SPECIFIC ANTINUCLEAR ANTBODIES hSULTS Definition of study sera containing granulocyte specific ( G S ) antibodies. In the course of routine immunofluorescent testing of over 5000 sera for antinuclear activity with spleen imprints, 41 sera were encountered which gave prominent staining of the nuclei of granulocytes and lesser or undetectable reactivity with the majority of cell nuclei, i.e., lymphocytes. In 31 sera, the occurrence of GS antibodies was demonstrated when absorption with calf thymus nucleoprotein resulted in the abolition of lymphocyte staining while reactivity with granulocytic nuclei persisted. Subsequent to absorption with calf thymus nucleoprotein, however, all nuclear reactivity was abolished from 10 sera. The presence of GS antibodies was most often verified in sera which showed a marked difference between granulocyte and lymphocyte nuclear reactivity prior to absorption (Table 1, Fig. 1). Fluorescent patterns of GS reactiuity. In 25 of the study sera, GS nuclear staining was homogeneous (type I staining, Fig. 2 ) . Six sera reacted with granulocyte nuclei to give a bizarre pattern of staining consisting of irregular fluorescent streaks and globules which appeared to be intranuclear although nuclear outlines were indistinct (type I1 staining, Figs. 3A and B). The same two types of reactivity were demonstrable with granulocyte nuclei when present in human liver and kidney sections, while lesser or undetectable reactivity occurred with parenchymal nuclei. Absorption with calf thymus nucleoprotein did not alter either pattern of GS staining. Further tests of antigenic specificities. Absorption of selected sera with human peripheral cell preparations gave additional evidence that both types of reactivity were granulocyte specific. When six type I and four type I1 sera were absorbed with human lymphocytes, no diminution in granulocytic nuclear reactivity occurred. Absorption of the same sera with either whole granulocytes or isolated nuclei of granulocytes completely abolished all nuclear reactivity. The diffuse nuclear reactivity in two lupus sera with both lymphocytes and granulocytes was abolished following incubation with any of the above absorbents. Species specificities. When tested with heterologous spleen imprints, all 25 sera containing type I antibodies demonstrated Table 1.-Effect of Absorption of Sera with Calf Thymus Nucleoprotein on Fluorescent Nuclear Reactivity with Spleen Imprints 1 After Absorption Number of Sera Tested Granulocyte8 10 9 5 8 4+ 4+ 4+ 3 f 7 3+ 2+ 2 - Nuclear Reactivity* Number+ LympaOcytes Neg. 1+ 2+ Neg. shoftnnn GS Anbbody 10 9 2 7 + Number of Sera Showiq no Nuclear Reachvlty t 0 0 3 1 5 1 1+ 2 Neg. - 1 - 31 10 41 'Graded 1 to 44- on the basis of intensity of fluorescence. f Staining of granulocytic nuclei unchanged in intensity but reaction with lymphocytic nuclei abolished. 4 No staining of either lymphocytic or granulocytic nuclei. 168 KATHRYN H. SVEC Fig. 1.-Intense homogeneous nuclear reactivity with granulocytes in human spleen imprint demonstrated by immunofluorescence. Nuclei of lymphocytes show only faint staining. X 430. rabbit spleen. The six sera containing type I1 antibodies showed no reactivity with any of the heterologous spleen nuclei. The presence of granulocytes in imprints from all species was verified with Wright’s stain. Immunoglobulin composition. Type I GS factors were consistently found present in the yG immunoglobulin class with titers of reactivity ranging from 1/16 to 11256. In only five sera were low titers of yA and/or yM reactivity detected. Some of these sera also contained rheumatoid factor, and it is recognized that apparent macroglobulin antinuclear activity may occasionally be the Fig. 2.-Homogeneous type I immunofluo- result of interaction between rheumatoid rescent staining pattern with nucleus of granu- factor and yG antinuclear factors.12 Type locyte. Smaller rounded nuclei of lymphocytes I1 GS antinuclear antibodies were restricted in background show no fluorescence and are to the yG immunoglobulin class and were barely discernible. X 1250. detected at serum dilutions of 118 to 11256. Clinical obseruations. As shown in Table reactivity with granulocyte nuclei in dog 2, the majority of patients with either type spleen imprints, 10 sera reacted with gran- I or I1 antibodies had rheumatoid arthritis. ulocyte nuclei in both dog and cow spleen, One patient with type I1 antibody demonbut no serum reacted with nuclei of rat or strated an ill-defined connective tissue dis- GRANULOCYTE-SPECIFIC ANTINUCLEAR ANTIBODIES Fig. 3.-(A) nuclei. 169 and (B): Irregular nonhomogeneous type I1 reactivity with granulocyte X 1250. ease diagnosed as possible systemic lupus, characterized by chronic motor neuropathy and rare episodes of arthralgias. Statistically signscant clinical differences between patients exhibiting the two types of GS antibodies are summarized in Table 3. Group I patients more frequently had definite or classical rheumatoid arthritis and were on the average 16 years older than group I1 patients. These differences were not a reflection of duration of disease since rheumatic symptoms had been present an average of 12 years in both groups. The two groups also differed in that patients with type I1 GS antibodies more commonly showed leukopenia, 4500 white blood cellslcu. mm. or less, with granulocytes decreased to 2300/cu. mm. or less. Both group I granulocytopenic patients also exhibited splenomegaly, while granulocytopenia occurred as an isolated finding in the group I1 patients. There was no correlation between leukopenia and the administration of drugs, including chloroquine. No significant differences were observed between patients with type I or I1 antibodies with respect to sex or race, presence of nodules, or occurrence of rheuma- Table 2.4linical Diagnoses of Patients Exhibiting Granulocyte-Specific Antinuclear Antibodies Granulocyte-S-c Diagnosis Rheumatoid arthritis: Classical Definite Probable Possible systemic lupus * Numbers of patients. -1 10 * 13 2 - Antibody Type II 1 1 3 1 toid factor. LE cells were detected in 6 of 16 (37.5 per cent) group I patients and in none of the 4 group I1 patients tested ( p > 0.05). Serial studies. In 14 group I patients, follow-up sera were obtained at 4 to 6 month intervals for 1 to 4 years. Homogeneous reactivity restricted to granulocytic nuclei persisted without significant changes in immunoglobulin composition. Fluctuaations in titers of reactivity occasionally occurred but did not appear to parallel rheumatic disease activity. Serial sera obtained from 2 group I1 patients over 1 and 2 year periods, respectively, continued to show the same irregular pattern of staining 170 KATHRYN H. SVEC Table 3.-Signihant Clinical Differences Between Patients with Types I or II Granulocyte-Specific Antinuclear Antibodies Clinical Feature Definite or classical rheumatoid arthritis Average age Leukopenia with granulocytopenia $ Type1 Patieats. 25 23 pts. (92%) 55.7 years 2 pts. (8%) Pat~ents.6 2 pts. (33%) 39 years 5 pts. (83%) P < 0.01 * < 0.01 t < 0.001 * Fisher’s exact test. + T test for groups of unequal distribution. $4500 leukocytes or less and 2300 granulocytes or less per cu. mm. of granulocytic nuclei without change in either titer of reactivity or immunoglobulin composition. Granulocytopenia which was present in both patients also persisted. No patient in either group showed any increase in nongranulocytic nuclear reactivity. Detection of masked GS antibodies. Sera showing strong reactivity with all nuclei of human spleen imprints were absorbed with calf thymus nucleoprotein to determine whether GS antibodies were present but obscured by other antinuclear antibodies. Sera from 15 rheumatoid arthritis patients, 55 systemic lupus patients, and 5 lupoid hepatitis patients were studied. In three instances, absorption revealed GS antinuclear antibodies which were verified by additional absorption tests with human granulocytes and lymphocytes. The GS antibody in one serum from a patient with definite rheumatoid arthritis showed the pattern of staining and species cross-reactivity characteristic of type I antibodies. The other two sera were obtained from young lupoid hepatitis patients, who exhibited splenomegaly and pancytopenia including granulocytopenia. GS antibodies in these sera gave a pattern of nuclear staining identical to type I1 reactivity and did not react with nuclei in the four heterologous spleens tested. Although many sera prior to absorption appeared to react with slightly increased intensity with granulocytes, absorption with calf thymus nucleoprotein resulted in complete abolition of all nuclear staining with the exception of occasional sera from lupus patients which demonstrated persistence of fine speckling in nuclei, presumably antibody to buffer extract antigen.13 It was considered unlikely that this residual reactivity could obscure the more obvious staining given by GS antibodies. DI~CUSSION The present report confirms the existence of nuclear reactivity restricted to granulocytes and presents evidence for two GS antibodies which can be differentiated on the basis of patterns of nuclear staining and degree of species specificity. Although GS antibodies may occasionally be masked and the true frequency of these antibodies is not known, evidence of selective reactivity with granulocytic nuclei is not commonly encountered in the course of routine testing for antinuclear antibodies. On the other hand, GS antibodies do not appear to be a rare occurrence in established rheumatoid arthritis. Among a group of 40 patients with stage I1 to IV rheumatoid arthritis who were selected for rehabilitation studies,14 type I GS reactivity was detected and verified by absorption tests in 8 (20 per cent) of these patients. In addition to the high degree of antigenic specificity of GS antinuclear antibodies, certain other characteristics are worthy of note. Although rheumatoid arthritis was by far the most common diagnosis for patients exhibiting either type I or I1 antibodies, both antibodies were con- 171 GRANULOCYTE-SPECIFIC ANTINUCLEAR ANTIBODIES sistently yG globulins, and only occasional type I antibodies demonstrated additional weak reactivity in the yM and yA immunoglobulin classes. Faber and Elling similarly observed that GS antibodies were not sensitive to penicillamine.KThese findings are at variance with the general prevalence of high molecular weight antinuclear factors in rheumatoid arthritis as established by the use of specific immunofluorescent antitests of mercaptoethanol sensitivity,lK and studies of antinuclear activity in serum fractions separated by chromatography and ultracentrifugation.l6J2 The reactivity of both GS antibodies with nucleoprotein was demonstrated by enzyme and absorption tests. The irregular pattern of staining given by the type I1 antibody is in contrast to the homogeneous pattern which has been previously related to antibody to nucleoprotein.l’J8 The recent report of Ritchie, however, indicates that homogeneous patterns of staining in some instances may actually represent composite reactivities of antibodies which individually give nonhomogeneous patterns of staining. Two yG antibodies which gave “nodular” or “reticular” staining, respectively, with the 1 M NaCl extractable fraction of rat liver nuclei were detected in sera of patients with connective tissue disease~.~~ In contrast to the commonly observed serologic overlap among connective tissue diseases, type I GS antibodies occurred only in rheumatoid arthritis patients. Although the disease was usually well established in these patients (average, stage II- I11) , associated clinical and laboratory findings did not suggest any obvious features that would set this group apart from the general population of rheumatoid patients with comparably advanced disease. A frequent occurrence of leukopenia, often in association with Felty’s syndrome, was previously observed in a group of patients whose sera contained both GS and granulocyte reactive antibodies? In the present study, the relationship between type I1 GS antibody and granulocytopenia was striking. This was further emphasized by the unmasking of morphologically and immunologically identical antibodies in sera of 2 lupoid hepatitis patients who demonstrated granulocytopenia. The presence of this antibody was not revealed by absorption tests in any of the sera of patients with established systemic lupus. It may be relevant that although leukopenia is frequent in lupus, white blood cell differentials are usually either normal or show lymphopenia.20 These observations might suggest a possible pathogenic role for type I1 GS antibodies, but it is quite conceivable that these serum factors may be only reflectors of granulocytic injury produced by unrelated mechanisms. The currently available data are descriptive in nature, and clarification of the actual significance of GS antinuclear antibodies will depend on future investigation. ACKNOWLEDGMENTS Bruce Veit, Pauline Yanowitz, and Austin Ray provided excellent technical assistance. Dr. Robert Kellermeyer kindly reviewed the manuscript. SUMMARIO IN INTERLINGUA Le presentia de granulocyto-specific anticorpore anti nucleoproteina esseva establite per tests de absorption in 31 seros e demascate in 3 seros additional. Le majoritate del patientes in question demonstrava arthritis rheumatoide, sed 2 de illes habeva hepatitis lupoide. Le observate differentias in le configurationes fluorescentic del tincturation e le grados de reactivitate heterologe characterisava duo distincte granulocyto-specific anticorpores antinucleari, incluse un que monstrava un significative association con granulocytopenia. 172 KATHRYN H. SVEC REFERENCES 1. Barnett, E. V., North, A. F., Condemi, J. J.. Jacox, R. F., and Vaughan, J. H.: Antinuclear factors in systemic lupus erythematosus and rheumatoid arthritis. Ann. Intern. Med. 63:100, 1965. 2. Faber, V., Elling, P., Norup, G., Mansa, B., and Nissen, N. I.: An antinuclear factor specific for leukocytes. Lancet 2:344, 1964. 3. Faber, V., and Elling, P.: Leukocyte-specific anti-nuclear factors in patients with Felty’s syndrome, rheumatoid arthritis, systemic lupus erythematosus, and other diseases. Acta Med. Scand. 179:257, 1966. 4. Elling, P.: Reaction of antinuclear factors with polymorphonuclear leukocytes. 1. Absorption studies. Acta Path. Microbiol. Scand. 68:281, 1966. 5. Ropes, M. W., Bennett, G. A., Cobb, S., Jacox, R., and Jessar, R. A.: 1958 revision of diagnostic criteria for rheumatoid arthritis. Bull. Rheum. Dis. 9:175, 1958. 6. Svec, K. H.: The use of human spleen imprints for routine testing for serum antinuclear factors by immunofluorescence. Amer. J. Clin. Path. 47:432, 1967. 7. Svec, K. H., and Veit, B. C.: Age-related antinuclear factors: immunologic characteristics and associated clinical aspects. Arthritis Rheum. 10:509, 1967. 8. Chargaff, E.: Isolation and composition of the deoxypentose nucleic acids and of the corresponding nucleoproteins. In: Chargaff, E., and Davidson, J. N. (Eds.): The Nucleic Acids, Vol. I. New York, Academic Press, 1955. 9. Esmann, V.: Effect of cell concentration on the metabolism of normal and diabetic leukocytes in vitro. Metabolism 13:354, 1964. 10. Falor, W. H.: Bilateral scalene lymph node excision: bilateral thoracic duct cannulation. Pacif. Med. Surg. 72:350, 1964. 11. Nathan, D. J., and Snapper, I.: On the interaction of dead leukocytic nuclei, L. E. factor and living leukocytes in the L. E. phenomenon. Blood 13:883, 1958. 12. Barnett, E. V., Condemi, J. J., Leddy, J. P., and Vaughan, J. H.: Gamma2, gamma,* and gammalM antinuclear factors in human sera. J. Clin. Invest. 43:1104, 1964. 13. Beck, J. S.: Partial identification of the “speckled” nuclear antigen. Lancet 1:241, 1962. 14. Katz, S., Vignos, P. J., Jr., Moskowitz, R. W., Thompson, H. M., and Svec, K. H.: Comprehensive outpatient care in rheumatoid arthritis. J.A.M.A. 206:1249, 1968. 15. Weir, D. M., and Holborow, E. J.: Serum anti-nuclear factor. Laboratory studies. Ann. Rheum. Dis. 21:40, 1962. 16. Baum, J., and Ziff, M.: 7 s and macroglobulin antinuclear fluorescence factors in systemic lupus erythematosus and rheumatoid arthritis. Arthritis Rheum. 5:636, 1962. Abstract. 17. Beck, J. S.: Variations in the morphological patterns of “autoimmune” nuclear fluorescence. Lancet 1:1203, 1961. 18. Bonomo, L., Tursi, A., and Dammacco, F.: Characterization of the antibodies producing the homogeneous and the speckled fluorescence patterns of cell nuclei. J. Lab. Clin. Med. 6642, 1965. 19. Ritchie, R. F.: Two new antinuclear antibodies: their relationship to the homogeneous immunofluorescent pattern. Arthritis Rheum. 11:37, 1968. 20. Dubois, E. L.: Lupus Erythematosus. New York, McGraw-Hill, 1966.