56 Gut 1993; 34: 56-62 Anti-lactoferrin antibodies and other types ofANCA in ulcerative colitis, primary sclerosing cholangitis, Gut: first published as 10.1136/gut.34.1.56 on 1 January 1993. Downloaded from and Crohn's disease
E Peen, S Almer, G Bodemar, B-O Ryden, C Sjolin, K Tejle, T Skogh
Abstract staining pattern (C-ANCA), whereas anti- Fifty two serum samples from patients with myeloperoxidase and anti-elastase antibodies Crohn's disease, 24 from patients with ulcera- give rise to a blurry perinuclear staining pattern tive colitis, and 12 from patients with primary (P-ANCA) or staining of the PMNL nuclei sclerosing cholangitis were analysed for the (granulocyte specific anti-nuclear antibodies, presence of anti-neutrophil cytoplasm anti- GS-ANA).24' A detergent extract of isolated bodies (ANCA) of IgG and IgA class by means azurophil granules (a antigen)-can be used for the of enzyme linked immunosorbent assays using detection of anti-proteinase 3 antibodies lactoferrin, myeloperoxidase, and antigen (C-ANCA) by means ofenzyme linked immuno- extracted from azurophil granules, 'a antigen' sorbent assay (ELISA) but does not allow detec- (that is, an antigen preparation containing tion of anti-myeloperoxidase.56 Also, antibodies proteinase 3) as substrates. A high frequency against lactoferrin, an iron binding protein resid- of positive tests for IgG anti-lactoferrin anti- ing in specific granules of PNML,7 produce a bodies was found in sera from patients with P-ANCA pattern owing to perinuclear/nuclear ulcerative colitis (50%) and primary sclerosing localisation of lactoferrin after ethanol fixation.8 cholangitis (50%). In Crohn's disease only 4 of Apart from the artifactual GS-ANA staining 52 (8%) sera had anti-lactoferrin antibodies - in pattern caused by antibodies directed against all four instances the sera belonged to patients nucleophilic cytoplasmic antigens, it is possible with disease involving the colon. AU patients that true GS-ANAs also exist.9 with sclerosing cholangitis and positive tests Apart from the occurrence in primary for anti-lactoferrin had ulcerative colitis. Low systemic vasculitides and rapidly progressive levels of IgG antibodies against myeloperoxi- glomerulonephritis, P-ANCA/GS-ANA may be
dase or a antigen were also found occasionally seen in other disease states, for example rheuma- http://gut.bmj.com/ in sera from patients with ulcerative colitis and toid arthritis without signs ofvasculitis,'0 inflam- primary sclerosing cholangitis. IgA antibodies matory bowel disease, and primary sclerosing directed against lactoferrin and a antigen (but cholangitis."-'5 Recent evidence favours the idea not myeloperoxidase) were seen in all three that inflammatory bowel disease may be caused conditions. by mesenteric vasculitis.I'l8 (Gut 1993; 34: 56-62) In a preliminary study of 16 frozen sera from
patients with Crohn's disease we found low levels on September 29, 2021 by guest. Protected copyright. of anti-a antigen antibodies, anti-myeloperoxi- Since the description of anti-neutrophil cyto- dase antibodies, and anti-lactoferrin antibodies plasm antibodies (ANCA) in active Wegener's in some sera.'9 The present study was done to granulomatosis,' the interest in anti-granulocyte extend these observations and to include sera antibodies has increased vastly. ANCA is now from patients with ulcerative colitis and primary Department of Internal recognised as a family ofautoantibodies directed sclerosing cholangitis. Medicine, against cytoplasmic antigens, mainly lysosomal Gastroenterology and enzymes, in Rheumatology Clinics polymorphonuclear neutrophil and Department of leukocytes (PMNL).2 In Wegener's granulo- Patients and methods Medical Microbiology, matosis, ANCA are typically directed against Transfusion Medicine proteinase 3, a serine proteinase located in and Clinical Immunology CROHN S DISEASE Laboratory, Faculty of azurophilic granules of human PMNL.3 Anti- Fifty two patients, 27 men aged 21-71 years Health Sciences, proteinase 3 antibodies may also occur in (mean 48 years) and 25 women aged 17-55 years University Hospital, isolated, rapidly progressive glomerulone- (mean 38 years) were enrolled in the S-581 85 Linkoping, study. The Sweden phritis and occasionally in systemic vasculitic mean duration of the disease was 17 years (range E Peen conditions other than Wegener's granulomato- 4-33 years). Eight patients had disease limited to S Almer sis.2 Antibodies directed against other lysosomal the colon and/or rectum, 19 patients had G Bodemar disease B-O Ryden enzymes in PMNL azurophilic granules, for in the small bowel alone, and 25 patients had C Sjolin example myeloperoxidase and elastase, also disease manifestations in both small bowel and K Teile occur in primary systemic vasculitic diseases and colon/rectum. Forty patients had been operated T Skogh rapidly progressive glomerulonephritis.34 After on, 14 of whom had been subjected to only Correspondence to: Dr Thomas Skogh, ethanol fixation of PMNL, proteinase 3 remains ileocoecal resection. Three patients had Department of Internal within the cytoplasmic granules, whereas myel- ileorectal anastomosis, and six patients had Medicine, Rheumatology Clinic, Faculty of Health operoxidase and elastase become extracted and ileostomy after proctocolectomy. Sciences, University locate close to or on the cell nucleus.24 At indirect Hospital, Linkoping, S-581 85, Sweden. immunofluorescent microscopical determina- Accepted for publication tion of ANCA, anti-proteinase 3 antibodies ULCERATIVE COLITIS 12 June 1991 produce a typical pancytoplasmatic granular Twenty four patients, 11 men agedl9-65 years Anti-lactoferrin antibodies andother types ofANCA in ulcerative colitis, primary sclerosing cholangitis, and Crohn's disease 57
IgG anti-a antigen diagnosis had been made by endoscopic retro- grade cholangiography 3 years (range 1-6 years) Crohn's Ulcerative Primary before the study. Two patients had only intra- disease colitis sclerosing hepatic cholangitis and 10 had both extra- and Gut: first published as 10.1136/gut.34.1.56 on 1 January 1993. Downloaded from 100 cholangitis intrahepatic cholangitis. In four patients cirrho- sis of the liver had been diagnosed at histopatho- logical examination ofliver biopsies. One patient had undergone liver transplantation 6 years earlier. Nine of the 12 patients had also suffered 50 from ulcerative colitis for 8 years (range 0-21 years) and one had non-specific colitis. In the two remaining patients no accompanying disease had been diagnosed. None of the patients had Crohn's disease. Z_~~~oooooooo pp...... One of the patients originally participating in the study had to be excluded as she proved to have suffered from acute Campylobacter jejuni IgG anti-myeloperoxidase colitis.
Crohn's Ulcerative Primary ,o disease colitis sclerosing ANTIGEN PREPARATIONS 4)0 co cholangitis Azurophil granules were prepared from isolated e. 100 normal peripheral granulocytes after low pres- 0 sure homogenisation of the cells and centrifuga- C.) .0> tion of the homogenate in a Percoll (Pharmacia 01)Co Fine Chemicals, Uppsala, Sweden) gradient as described elsewhere.5 The granules were lysed .-0 50 with 0-01% (final concentration) Triton X-100 Q (Merck, Darmstadt, Germany).5 The extract
co thus achieved (a antigen) was diluted in n0L carbonate-bicarbonate buffer pH 9 5 and used as substrate for ELISA.5 0@@-@@@@::@t- 0 * 0 0 Lyophilised preparations of myeloperoxidase (Calbiochem, La Jolla, CA, USA), human milk
lactoferrin (Sigma Chemical Co, St Louis, MO, http://gut.bmj.com/ IgG anti-lactoferrin USA), and bovine milk lactoferrin (Sigma) were dissolved in carbonate-bicarbonate buffer (10 Crohn's Ulcerative Primary [tg antigen/ml) and used for ELISA. disease colitis sclerosing .P ..cholangitis 100 ELISA
High binding plastic microtitre plates (Nunc on September 29, 2021 by guest. Protected copyright. .00 gO~~~~~ Immunoplate, Roskilde, Denmark) were coated with a antigen, myeloperoxidase, or human lactoferrin solutions by incubation at 40C for 12 50 hours. After washing with buffer (phosphate buffered saline, PBS, with 0 05% Tween 20), *g @0 0 0@ serum samples (diluted 1:10 in PBS-Tween) *v@@@@@@@@@@@** ~ ~ 0 g were applied for 30 minutes at room tempera- 00000000000000e ture. Normal human serum diluted 1:10 served as a blank. The microtitre plates were washed Figure 1: Diagram to illustrate the ELISA results ofIgG autoantibody directed against thoroughly with PBS-Tween and incubated for a antigen, myeloperoxidase, and lactoferrin in sera from patients with Crohn's disease (n=52), another 30 minutes with alkaline-phospatase ulcerative colitis (n=24), andprimary sclerosing cholangitis (n= 12). The lower limitfor positive ELISA (that is, +2 SD ofthe mean value of218 normal reference sera) is indicated. (ALP) conjungated rabbit anti-human y chain or rabbit anti-human a chain antisera (Dako, Glostrup, Denmark) diluted 1:400 in PBS- (mean 46 years), and 13 women aged 20-68 years Tween. After washing, the substrate buffer was (mean 42 years) participated in the study. Nine- applied and the optical density (OD) read at teen patients had left sided colitis and five 405 nm when the positive reference samples had patients had more extensive or total colitis. reached OD 1.0. OD values exceeding 2 SDs of Fourteen of the patients had active disease - that the reference material from healthy blood donors is at least five loose stools with visible blood (all (n=218 for IgG tests; n=211 for IgA tests) were had visible inflammation at endoscopy) - and the considered positive. The IgG autoantibody remaining 10 patients were in clinical remission. analyses were performed on all sera from patients with Crohn's disease (n=52), ulcerative colitis (n=24), and primary sclerosing cholangitis PRIMARY SCLEROSING CHOLANGITIS (n=12). IgA autoantibody tests were done on Twelve patients, seven men aged 20-59 years sera from 51 patients with Crohn's disease, 21 (mean 38 years) and five women aged 35-63 years ulcerative colitis sera, and 11 sera from patients (mean 48 years), took part in the study. The with primary sclerosing cholangitis. 58 Peen, Almer, Bodemar, Ryden, Sjdlin, Tejile, Skogh
The rabbit anti-human leukocyte lactoferrin kD 1 2 3 4 5 served as a reference. ALP conjugated anti- human IgG and anti-rabbit IgG (Dako) were
used as secondary antibodies. Gut: first published as 10.1136/gut.34.1.56 on 1 January 1993. Downloaded from
- Wi 7- INDIRECT IMMUNOFLUORESCENCE (IIF) MICROSCOPY To study the distribution of lactoferrin in 45- ethanol fixed granulocytes cytocentrifuged onto microscope slides, the slides were incubated with 31- rabbit anti-human milk lactoferrin in a moist chamber for 30 minutes. After washing with PBS 1421 and incubation for another 30 minutes with sheep fluorescein isothiocyanate (FITC) con- jugated anti-rabbit Ig (Wellcome Diagnostics, MW Lf rabbit Patient seru Temple Hill, Dartford, UK), the slides were anti-Lf again washed with PBS, mounted with PBS- glycerin, and inspected under a fluorescence Figure 2: Sodium dodecyl sulphate polyacrylamide gels (SDS-PAGE) electrophoresis ofthe microscope with a mercury lamp (HBO 50) epi- human milk lactoferrin preparation (lane 1) showed only material ofmolecular weight about 80 kD when stained with Comassie blue. Lane 2 shows molecular weight markersfor comparison. illumination and filters for FITC activation/ Immunoblotting, using polyclonal rabbit anti-human leukocyte lactoferrin, showed slight emission. reaction with antigens oflower molecular weight, apartfrom the strong reaction with the 80 kD lactoferrin band (lane 3). Patient sera positivefor IgG anti-human milk lactoferrin, as tested by ELISA, however, reacted only with lactoferin. STATISTICS Differences in ELISA results between the con- IgG anti-bovine lactoferrin antibodies were trol sera and the patient sera were evaluated by measured essentially as described above, except the X2 test, and divided into four groups: p.-005 that no negative or positive serum references =not significant (NS); p<0O05 (*); p<0-01 (**); were used. OD values were read after 30 minutes p
IgA anti-a antigen IgG anti-lactoferrin antibodies were found in 50% of the sera (p<0-001 in both instances). All Crohn's Ulcerative Primary of the anti-lactoferrin positive cholangitis sera disease colitis sclerosing to Gut: first published as 10.1136/gut.34.1.56 on 1 January 1993. Downloaded from cholangitis belonged patients suffering from concomitant 100 ulcerative colitis - that is, 6 of 9 (67%) sera from patients with sclerosing cholangitis and con- comitant ulcerative colitis contained IgG anti- lactoferrin antibodies. Two of the sera from patients with primary sclerosing cholangitis and 50 one serum from a patient with ulcerative colitis were positive in all three ELISA tests. However, in all three instances the ELISA tests gave quite different OD values in the different assays, indicating true positive ELISA test results. Positive anti-lactoferrin ELISA results were blocked by preincubation of the lactoferrin- coated microtitre plates with rabbit IgG anti- lactoferrin (not illustrated). The occurrence of IgA anti-myeloperoxidase IgG anti-lactoferrin was also confirmed by western blotting, revealing a single reaction band Crohn's Ulcerative Primary with an antigen ofmolecular weight about 80000 disease colitis sclerosing and corresponding to rabbit anti-human leuko- 0 cholangitis 0) 100 cyte lactoferrin (Fig 2). 0a) One patient originally included in the study had acute colitis and a high level of IgG anti- lactoferrin at the first sampling occasion, and a further raised anti-lactoferrin level in a later .00) CL serum sample (not illustrated). This patient was, -0) 50 however, excluded from the study since her C.) symptoms were explained by infection with 0) Campylobacterjejuni. o . . 3 illustrates the lack of n S....~~~~~~~~~~~~~~~5... _... Figure correlation 0 0 0 0 0 *@S@:::::::SSSS@::: SSSS : :**.***.S.. *@ @-@-@ @ between levels ofIgG anti-human lactoferrin and IgG anti-bovine lactoferrin as tested by ELISA with 10 sera positive in the anti-human lacto- http://gut.bmj.com/ ferrin and 12 sera negative in the same test. IgA anti-lactoferrin Figure 4 shows the occurrence of IgA anti- bodies directed against a antigen, myeloperoxi- Crohn's Ulcerative Primary dase, and human lactoferrin. disease colitis sclerosing Raised levels of IgA anti-a antigen were seen cholangitis in 12 of 51 (24%) sera from Crohn's disease 100 (p<0-001), in 5 of 11 (45%) sera from patients on September 29, 2021 by guest. Protected copyright. with primary sclerosing cholangitis (p<0001), but in only 2 of 21 (10%) sera from patients with ulcerative colitis (NS). The levels of IgA anti- myeloperoxidase did not differ from the control 50 0~~~~~~ 0 group in any of the disease states, whereas IgA anti-lactoferrin was significantly (p<005) more .0 common in Crohn's disease, although it was seen only in a minority of the patient sera (5 of 51 = *.*SS****S**.*S S***-**.S..S *@S0000050 10%). IgA anti-lactoferrin was found in signific- antly (p<0-001) increased frequency also in Figure 4: Results ofthe ELISA testsfor IgA autoantibodies directed against a antigen, ulcerative colitis (7 of 21=33%), but not in myeloperoxidase, and lactoferrin. Fifty one sera from patients with Crohn's disease, 21 sera sclerosing cholangitis (1 of 11=9%). from patients with ulcerative colitis, and 1I sera from patients with primary sclerosing The levels of were cholangitis were analysed. The cut oflevels (that is, values exceeding 2 SD ofthe mean ofa IgA-ANCAs not explained reference material of21 I normal sera) are indicated. by the occurrence of agglutinating rheumatoid factors (not illustrated). Figure 5 shows the frequency of positive IgG the control group), and in 7 of24 ofthe ulcerative and/or IgA tests for either anti-a antigen, or anti- colitis sera (p<0001). Four of 12 sera from myeloperoxidase, or anti-lactoferrin in sera from patients with primary sclerosing cholangitis were blood donors, Crohn's disease, ulcerative colitis, positive (p<0001). and primary sclerosing cholangitis. Taken Four of 52 (8%) of the sera from patients with together, positive IgG tests for one or more ofthe Crohn's disease contained IgG anti-lactoferrin three granulocyte antigens were found in 11% antibodies (not significantly different from the (24 of 218) control sera, in 19% (10 of 52) of control group). All of these sera proved to come Crohn's disease (NS), in 67% (16 of 24) of from patients with Crohn colitis. However, ulcerative colitis (p<0-001), and in 67% (8 of 12) colitis also occurred in 33 of 48 (69%) without of primary sclerosing cholangitis sera; IgA anti- IgG anti-lactoferrin antibodies. Both in ulcera- granulocyte antibodies were found in 7% (15 of tive colitis and primary sclerosing cholangitis, 211) of the controls, in 29% (15 of 51) ofCrohn's 60 Peen, Almer, Bodemar, Ryden, Sjolin, Tejile, Skogh
antibodies. The levels of IgG and IgA anti- * Control granulocyte antibodies did not correlate to Cl Crohn's disease the total serum levels of IgG and IgA (not
illustrated). Gut: first published as 10.1136/gut.34.1.56 on 1 January 1993. Downloaded from O] Ulcerative colitis Figure 6 shows the GS-ANA/P-ANCA B Primary sclerosing cholangitis immunofluorescence pattern of cytocentrifuged 100 -ItL ethanol-fixed human polymorphonuclear neutrophil granulocytes after staining with rabbit anti-human milk lactoferrin. This, taken together with the western blot analysis (Fig 2), * * 80 - * * * * shows that antibodies against human milk lacto- ferrin recognise human leukocyte lactoferrin and * * vice versa. * U) * 60 - * (0) 0) * Discussion
(n * In this study we report high frequencies of IgG 0 40- anti-lactoferrin antibodies (corresponding to -0 P-ANCA) in sera from patients with ulcerative colitis and primary sclerosing cholangitis. In serum 20 Crohn's disease, however, anti-lactoferrin antibodies of IgG were rarely detected, and then only in patients with colonic disease manifesta- tions. IgA anti-lactoferrin antibodies were, how- ever, found in some cases of both ulcerative lgG IgA IgG+lgA 19G colitis and Crohn's disease. Several different types of autoantibodies have Figure 5: Diagram summarising the ELISA testsforIgG andlor IgA autoantibodies granulocyte antigens tested - that is, a antigen, myeloperoxidase, and human lactof been described in inflammatory bowel disease and primary sclerosing cholangitis."- 1921-27 Although none of the autoantibodies have been proved to be of pathogenetic significance, it has disease patients (p
statistically from the control group, although in a cross immunisation against human lactoferrin,34 few instances the antibody levels were surpris- and the lack ofcorrelation between levels ofanti- ingly high. In both ulcerative colitis and primary human lactoferrin and anti-bovine lactoferrin
sclerosing cholangitis the frequency of positive shown in this study contradicts the hypothesis Gut: first published as 10.1136/gut.34.1.56 on 1 January 1993. Downloaded from IgG anti-a/anti-myeloperoxidase tests differed that cross immunisation against dietary bovine statistically from the control group, but in lactoferrin explains the appearance of anti- general the levels were low and, considering the human lactoferrin autoantibodies. An exciting small number of patients with positive anti-cc/ possibility is the antigenic mimicry between the anti-myeloperoxidase tests, the relevance of 65 kD mycobacterial heat-shock protein (hsp- these results is uncertain. However, it is interest- 65) and human lactoferrin, and the fact that ing that these types ofANCAs were found at all, immunisation against hsp-65 results in the pro- since they are considered to be reliable markers duction of anti-lactoferrin antibodies.32 Myco- of primary systemic vasculitis.2 ANCAs of IgA bacterial infection and immunisation against class have been reported in Hennoch-Schonleins mycobacterial antigens/hsp-65 have been impli- purpura and IgA nephropathy, although these cated in several autoimmune disease states, both findings are sometimes explained by the experimental and clinical.20"3 435 Considering presence of rheumatoid factor.33 The appearance the occurrence of IgA anti-lactoferrin antibodies of antibodies directed against a antigen in some in some cases of Crohn's disease in the present cases of ulcerative colitis (IgG) and in Crohn's study, we find it interesting that IgA (but not disease (IgA) could possibly favour the hypo- IgG) anti-hsp-65 antibodies have been reported thesis of primary vasculitis as a pathogenetic in Crohn's disease.42 Furthermore, the registra- factor in these diseases.""-8 tion of anti-lactoferrin antibodies in a case of Lactoferrin is an iron binding protein that Campylobacter colitis in this study is interesting occurs abundantly not only in the specific with regard to the possibility of cross immunisa- granules ofgranulocytes,7 but also in tears, milk, tion against microbial antigen(s). We therefore and secretions at mucosal surfaces.3"36 Raised intend to analyse additional sera from patients levels of circulating lactoferrin are seen during with infectious gastroenteritis for the presence of active inflammatory disease.37 It exerts anti- anti-lactoferrin antibodies. bacterial effects by depriving bacteria of iron In conclusion, we have shown high frequen- required for growth, and it has anti-inflamma- cies of IgG anti-lactoferrin antibodies in ulcera- tory properties - for example, by preventing tive colitis, and primary sclerosing cholangitis, complement activation through inhibition of but not in Crohn's disease, whereas IgA auto- classical C3 convertase.38 Lactoferrin can also antibodies directed against lactoferrin were prevent the formation of hydroxyl radicals by found in some cases ofboth ulcerative colitis and iron binding.3 139 Lactoferrin may thus be of Crohn's disease. We suggest that anti-lactoferrin http://gut.bmj.com/ great importance as a non-specific anti-phlogistic may be of pathogenetic significance by counter- defence factor at the primary immunological acting the antiphlogistic properties oflactoferrin barriers. It has been shown experimentally that at mucosal surfaces, thereby aggravating and/or binding of lactoferrin by anti-lactoferrin anti- sustaining mucosal inflammation initiated by bodies increases the amount and duration of other factors. hydroxyl radical formation by granulocytes.39 on September 29, 2021 by guest. Protected copyright. Hypothetically, anti-lactoferrin autoantibodies This study was financed by grants from Professor Nanna Svatrz could, by counteracting the anti-inflammatory Foundation, the Swedish Association against Rheumatism, King Gustaf Vth 80-Year Foundation, and the Swedish Medical effects of lactoferrin, aggravate and prolong Research Council. mucosal inflammation induced by several differ- ent mechanisms, and the antibodies may there- 1 van der Woude FJ, Rasmussen N, Lobatto S, et al. Auto- fore have pathogenetic significance even though antibodies against neutrophils and monocytes: tool for their occurrence does not seem to correlate with diagnosis and marker of disease activity in Wegener's granulomatosis. Lancet 1985; i: 425-9. disease activity. Anti-lactoferrin antibodies may 2 Kallenberg CGM, Cohen Tervaert JW, van der Woude FJ, also have pathogenetic effects by activation of et al. Autoimmunity to lysosomal enzymes: a new clue to vasculitis and glomerulonephritis? Immunol Today 1991; 12: primed granulocytes infiltrating the gut mucosa 61-4. or adhering to vessel walls, in analogy with the 3 Jenne DE, Tschopp J, Ludemann J, Utecht B, Gross WL. I Wegener's auto-antigen decoded. Nature 1990; 346: 520. effects of other types of ANCA.5 `' In addition, 4 Falk RJ, Jenette JC. Anti-neutrophil-cytoplasmic auto- mucosal lactoferrin/anti-lactoferrin complexes antibodies with specificity for myeloperoxidase in patients stimulate intestinal cells to excessive with systemic vasculitis and idiopathic necrotizing and may goblet crescentic glomerulonephritis. N Engl J Med 1988; 318: mucus secretion, as do other immune com- 1651-7. 5 Skogh T, Dahlgren C, Holmgren K, et al. Anti-granulocyte plexes.42 On the other hand, the occurrence of antibodies (C-ANC, P-ANCA, GS-ANA) studied by con- anti-lactoferrin autoantibodies may, of course focal scanning laser fluorescence microscopy, ELISA, and chemiluminescence techniques. Scand J Immunol 1991; 34: merely be an epiphenomenon without patho- 137-45. genetic significance. The fact that the individual 6 Rasmussen N, Sjolin C, Isaksson B, et al. An ELISA for the detection of anti-neutrophil cytoplasm antibodies (ANCA). types of ANCA were found only in a minority J Immunol Meth 1990; 127: 139-45. (.50%) of the sera tested is an argument in 7 Furmanski P, Li Z-P. Multiple forms of Lactoferrin in Normal and Leukemic Human Granulocytes. Exp Hematol favour of this. On the other hand, consumption 1990; 18: 932-5. of circulating antibodies cannot be ruled out in 8 Briggs RC, Glass II WF, Montiel MM, Hnilica LS. Lactoferrin: nuclear localization in the human neutrophilic the seronegative instances. granulocyte?JT Histochem Cytochem 1981; 29: 1128-36. The origin of anti-lactoferrin autoantibodies is 9 Wiik A. Granulocyte-specific antinuclear antibodies. Allergy 1980; 35: 263-89. unknown. Although bovine and human lacto- 10 Wiik A, Munthe E. Complement fixing granulocyte-specific ferrin have molecular and antigenic similari- antinuclear factors in neutropenic cases of rheumatoid arthritis. Immunology 1974; 26: 1127-34. ties,35 immunisation of experimental animals 11 Nielsen H, Wiik A, Elmgreen J. Granulocyte specific anti- with bovine lactoferrin seldom seems to result in nuclear antibodies in ulcerative colitis. Aid in differential 62 Peen, Almer, Bodemar, Rydin, Sjolin, Tejile, Skogh
diagnosis of inflammatory bowel disease. Acta Pathol antibodies to colonic epithelium in ulcerative colitis. Gut Microbiol Scand[Sect C] 1983; 91: 23-6. 1991;32: 163-6. 12 Snook JA, Chapman RW, Fleming K, Jewell DP. Anti- 28 Green FHY, Fox H. The distribution of mucosal antibodies in neutrophil nuclear antibody in ulcerative colitis, Crohn's the bowel of patients with Crohn's disease. Gut 1975; 16: disease and primary sclerosing cholangitis. Clin Exp 125-31.
Immunol 1989; 76: 30-3. 29 Hallstensen TS, Mollnes TE, Garred P, et al. Epithelial Gut: first published as 10.1136/gut.34.1.56 on 1 January 1993. Downloaded from 13 Saxon A, Shanahan F, Landers C, et al. A distinct subset of deposition of immunoglobulin GI and complement (C3b antineutrophil cytoplasmic antibodies is associated with and terminal complement complex) in ulcerative colitis. inflammatory bowel disease. J Allergy Clin Immunol 1990; Gastroenterology 1990; 98: 1264-71. 86: 202-10. 30 Davidsson L, Lonnerdal B. Specific binding of lactoferrin to 14 Duerr RH, Targan SR, Landers CJ, et al. Anti-neutrophil brush-border membrane. AmJ3 Physiol 1988; 254: G580-5. cytoplasmic antibodies in ulcerative colitis. Gastroenterology 31 Barnes RMR, Allan S, Taylor-Robinson CH, et al. Serum 1991; 100: 1590-6. antibodies to Saccharomyces servisiae in inflammatory bowel 15 Duerr RH, Targan SR, Landers CJ, et al. Neutrophil cyto- disease: Is IGA antibody a marker for Crohn's disease? plasmic antibodies: a link between primary sclerosing Int Archs AllergyApply Immunol 1990; 92: 9-15. cholangitis and ulcerative colitis. Gastroenterology 1991; 32 Winfield J, Jariour W. Do stress proteins play a role in arthritis 100;: 1385-91. and autoimmunity? Immunol Rev 1991; 121: 193-220. 16 Wakefield AJ, Sawyerr AM, Dhillon AP, et al. Pathogenesis of 33 Daha MR, Falk RJ. Ig-class specificity of ANCA. NethJ3 Med Crohn's disease: multifocal gastrointestinal infarction. 1990; 36: 126-7. Lancet 1989; ii: 1057-62. 34 Wang C-S, Chan W-Y, Kloer UH. Comparative studies on the 17 Sawyerr AM, Wakefield AJ, Rowles P, Dhillon AP, Pittilo M, chemical and immunochemical properties of human milk, Hudson M, et al. Microvascular changes in Crohn's disease human pancreatic juice, and bovine milk lactoferrin. Comp and ulcerative colitis: examination by vascular casting and Biochem Physiol[B] 1984; 78: 575-80. scanning electron microscopy. Gut 1990; 31: A623. 35 Magnuson JS, Henry JF, Yip T-T, Hutchens TW. Structural 18 Wakefield AJ, Sankey EA, Dhillon AP, et al. Granulomatous homology of human, bovine, and porcine milk lactoferrins: vasculitis in Crohn's disease. Gastroenterology 1991; 100: evidence for shared antigenic determinants. Pediatr Res 1279-87. 1990; 28: 176-81. 19 Peen E, Teile K, Skogh T. Anti-granulocyte antibodies in 36 Kuizenga A, van Haeringen NJ, Kijlstra A. Inhibition of Crohn's disease. APMIS 1990; suppl 19: 34. hydroxyl radical formation by human tears. Invest 20 Esaguy N, Aguas AP, van Embden JDA, Silva MT. Myco- Ophthalmol VisSci 1987; 28: 205-13. bacteria and human autoimmune disease: direct evidence of 37 Adeyami EG, Campos LB, Loizou G, et al. Plasma lactoferrin cross-reactivity between human lactoferrin and the 65- and neutrophil elastase in rheumatoid arthritis and systemic kilodalton protein of tubercle and leprosy bacilli. Infect lupus erythematosus. BrJ3 Rheumatol 1990; 29: 15-20. Immun 1991; 59: 1117-25. 38 Kievits F, Kijlstra A. Inhibition of C3 deposition on solid- 21 Broberger 0, Perlmann P. Autoantibodies in human ulcera- phase bound immune complexes by lactoferrin. Immunology tive colitis. J Exp Med 1959; 110: 657-74. 1985; 54: 449-57. 22 Perlmann P, Hammarstrom S, Lagercrantz R, Campbell D. 39 Britgan BE, Hassett DJ, Rosen GM, et al. Neutrophil Autoantibodies to colon in rats and human ulcerative colitis: degranulation inhibits potential hydroxyl radical formation. cross reactivity with Escherichia coli 0:14 antigen. Proc Soc BiochemJr 1989; 264: 447-55. Exp Biol Med 1967; 125: 975-80. 40 Falk RJ, Terrell RS, Charles LA, Jenette JC. Anti-neutrophil cytoplasmic autoantibodies induce neutrophils to 23 Das KM, Kadono Y, Fleischner GM. Antibody-dependent degranulate and produce oxygen radicals in vitro. Proc Natl cell-mediated cytotoxicity in serum samples from patients Acad Sci USA 1990; 87: 4115-9. with ulcerative colitis. Am3 Med 1984; 77: 791-6. 41 Charles LA, Caldas MLR, Falk R, Terell RS, Jenette JC. 24 Skogh T, Heuman R, Tagesson C. Anti-brush border anti- Antibodies against granule proteins activate neutrophils in bodies (ABBA) in Crohn's disease. J ClGn Lab Immunol vitro. J Leukoc Biol 1991; 50: 539-46. 1982; 9:147-50. 42 Walker WA, Wu M, Bloch KJ. Stimulation by Immune 25 Skogh T, Bodemar G, Kihlstrom E, Ljunghusen 0. Anti- Complexes of Mucus Release from Goblet Cells of the Rat brush border antibodies (ABBA) in sera from patients with Small Intestine. Science 1977; 197: 370-2. ulcerative proctocolitis and in sera with antibodies against 43 Lydyard PM, Rook GA, Tsoulfa G, Sharif M, Smith M. Is Yersinia enterocolitica 0:3. J Clin Lab Immunol 1986; 19: there a role for mycobacteria in the etiopathogenesis of 117-8. rheumatoid arthritis? Immunol Rev 1991; 121: 137-54. 26 Chapman RW, Cottone M, Selby WS, et al. Serum auto- 44 Shoenfeld Y, Isenberg DA. Mycobacteria and autoimmunity. http://gut.bmj.com/ antibodies, ulcerative colitis and primary sclerosing Immunol Today 1988; 9: 178-82. cholangitis. Gut 1986; 27: 86-91. 45 Hampson SJ, McFadden JJ, Hermon-Taylor J. Mycobacteria 27 Snook JA, Lowes JR, Wu KC, et al. Serum and tissue and Crohn's disease. Gut 1988; 29: 1017-9. on September 29, 2021 by guest. Protected copyright.