Experimental autoimmune dacryoadenitis was produced in Lewis rats by ... 10-55K) induced marked dacryoadenitis following a single injection of 2.0 mg protein ...
Experimental Autoimmune Dacryoadenitis /. Lacrimal Gland Disease in the Rat Sammy H. Liu, Robert A. Prendergasf, and Arthur M. Silverstein Experimental autoimmune dacryoadenitis was produced in Lewis rats by immunization with a single intradermal administration of a 3M KC1 extract of exorbital lacrimal gland in CFA, when enhanced by simultaneous i.v. injection of killed Bordetella pertussis. No significant lacrimal lesions were observed in control animals immunized with the extracts of Harderian or salivary glands. Gel filtration of the 3M KC1 extract on Sephacryl S-300 column yielded three protein fractions. Only fraction III (MW = 10-55K) induced marked dacryoadenitis following a single injection of 2.0 mg protein in CFA plus pertussis. The infiltrates in the exorbital lacrimal lesions were first apparent around the ducts and associated vasculature. From this area, the infiltrates appeared to spread to the acini drained by these ducts, ultimately involving as much as 30-50% of the gland. The affected glands most commonly showed a diffuse nongranulomatous infiltrate of small lymphocytes, macrophages, and plasma cells; this was focal in nature, involving acinar atrophy and breakdown, and replaced the normal architecture in extreme cases. The Harderian and salivary glands were uninvolved in these animals, suggesting a restricted specificity of this response. Lewis rats immunized with exorbital lacrimal gland fractions I or II in CFA plus pertussis showed only minimal lesions, similar to controls receiving CFA and pertussis without antigen. These findings suggest that an autoantigen exists in the lacrimal gland of the rat that is capable of inducing a specific lymphoproliferative dacryoadenitis. Invest Ophthalmol Vis Sci 28:270-275,1987
junctivitis sicca and xerostomia and is often accompanied by rheumatoid arthritis. The New Zealand Black, the New Zealand Black/ New Zealand White (NZB/NZW) Fl-hybrid, and the MRL/Mp strains of mice are subject to many autoimmune diseases, and they have been shown to develop spontaneous lesions in the lacrimal and salivary glands that mimic those seen in patients with Sjogren's syndrome.2"5 Furthermore, the NZB/NZW mouse has been found to develop cytotoxic autoantibody to lacrimal gland cells with increasing age.6 These studies hint that an autoimmune mechanism may also be involved in the lacrimal gland pathology. However, attempts to induce dacryoadenitis in experimental animals by sensitization with extracts of lacrimal gland have not been consistently successful. Waterhouse7 produced mild lesions in low frequency (17%) in rats injected with lacrimal gland homogenate in complete Freund's adjuvant (CFA). Mizejewski8 was able to produce severe dacryoadenitis in rats, but the incidence of the disease induced was not reported. The etiologies and the pathogenetic mechanisms responsible for most lesions of the lacrimal gland are still unknown. The development of experimental animal models may help to elucidate the pathogenesis of some of these human disease processes. In the present studies, we have succeeded in producing a severe dacryoadenitis in Lewis rats by taking a new approach to the isolation
The lacrimal gland is the site of a wide variety of both inflammatory and neoplastic orbital diseases. Although chronic idiopathic dacryoadenitis represents the most common inflammatory disease (reviewed in ref. 1), analogous lesions of the lacrimal gland may accompany more widespread disorders, such as Sjogren's syndrome or macroglobulinemia. Indeed, the histopathology of Sjogren's syndrome, a well-known autoallergic process, is similar in many respects to that of idiopathic dacryoadenitis.1 One of the chief features of both is chronic inflammation, with accumulations of lymphocytes, macrophages, and some plasma cells, accompanied by acinar atrophy. Although the histopathology of the two conditions may be similar, they are readily distinguishable, because idiopathic dacryoadenitis is a local disease unassociated with other systemic disorders, whereas Sjogren's syndrome includes the characteristic clinical picture of keratocon-
From The Wilmer Ophthalmological Institute, The Johns Hopkins University School of Medicine, Baltimore, Maryland. Supported in part by U.S. Public Health Service Research Grants EY-04444, EY-02650, and EY-03521 from the National Institutes of Health, Bethesda, Maryland, and by an Independent Order of Odd Fellows Research Professorship. Submitted for publication: March 5, 1986. Reprint requests: Sammy H. Liu, PhD, The Wilmer Institute, Johns Hopkins Hospital, 600 N. Wolfe Street, Baltimore, MD 21205.
270
271
EXPERIMENTAL AUTOIMMUNE DACRYOADENITIS / Liu er ol.
No. 2
of organ-specific antigen(s) and to the mode of immunization. We describe here the method by which autoimmune dacryoadenitis may be induced and the histopathologic features of the resulting disease.
1-2
r
1.0
Materials and Methods Animals Female Lewis rats weighing 150-175 g were obtained from Charles River Lab., Wilmington, MA, and were used in all experiments. Preparation of Lacrimal Gland Antigens KCl extracts: Exorbital lacrimal glands were excised from normal Lewis rats, trimmed of fat and connective tissue, and stored at — 20°C until use. The glands were subsequently fragmented into small pieces and suspended in 0.01 M Tris-HCl buffer at pH 7.8 containing 3M KCl (3 ml solvent per gland). The suspension was stirred slowly for 24 hr at 4°C, and centrifuged for 30 min at 40,000 g. The supernatant was decanted and saved. The pellet was reextracted with 3M KCl buffered solution for an additional 24 hr and recentrifuged. The two supernatants were pooled, dialyzed in the cold against several changes of phosphate-buffered saline (PBS) at pH 7.4, centrifuged at 40,000 g for 30 min at 4°C, and the supernatant retained. Partial purification of antigens: The 3M KCl extract of lacrimal gland was concentrated and adjusted to 50 mg/ml and applied to a 2.5 X 120 cm Sephacryl S-300 column (Pharmacia; Piscataway, NJ). The column was equilibrated, eluted with PBS, and calibrated with a mixture of known molecular weight markers, all at 4°C. The material was eluted at a flow rate of 60 ml/hr. Ten-milliliter fractions were collected and monitored for absorbance at 280 nm. The fractions were pooled according to the 280-nm profile (Fig. 1). The fractions obtained from the chromatographic columns were concentrated and adjusted to 10 mg/ml. All concentrations were done on an Amicon YM-10 membrane (Amicon Corp.; Danvers, MA). The dye-binding BioRad protein microassay (Bio-Rad Lab.; Richmond, CA), using bovine serum albumin as a standard, was used for all protein determinations. Control antigens: Fresh submaxillary glands and Harderian glands were extracted with 3M KCl as described above, dialyzed, and concentrated for use as control antigens. Sensitization of Animals Different groups of four or eight Lewis rats were sensitized with either 10 mg of the crude KCl extract or 2.0 mg of each of the 3 fractions obtained from the chromatographic column. In each case, the protein so-
0.8
S 0.6
0.4
0.2
500
Fig. 1. Gel nitration pattern of the 3M KCl extract of rat exorbital lacrimal gland on Sephacryl S-300 column. Fractions were eluted with isotonic phosphate buffered saline and monitored for adsorbance at 280 nm. Molecular weight standards used were bovine gamma globulin, 150,000; bovine serum albumin, 65,000; ovalbumin, 45,000; and lysozyme, 15,000 daltons. The eluates pooled for fractions I, II, and III are indicated.
lution was emulsified with an equal volume of CFA supplemented with 2 mg/ml Mycobacterium tuberculosis H37Ra strain (Difco; Detroit, MI). The emulsion (0.4 ml) was distributed equally among the four footpads of the rat. Control groups were similarly immunized with KCl extracts of salivary and Harderian glands. In some experiments, the rats were injected intravenously with 5 X 1010 Bordetella pertussis bacteria at the time of immunization. Killed pertussis vaccine (lot SS-9IB) was obtained from the Michigan Department of Health, Lansing, MI. Histologic Analysis of Inflammatory Response Animals were killed at intervals from 15-60 days after sensitization, and their lacrimal glands removed and processed for histologic evaluation of the degree of dacryoadenitis induced. Harderian and submaxillary salivary glands from these animals were also examined histologically. All glands were fixed in 10% buffered formalin, and embedded in paraffin. Histologic step sections were made often different areas of each gland, and stained with hemotoxylin and eosin. The severity of dacryoadenitis was graded on a scale of 1-3+, according to the following criteria: 1 + dacryoadenitis was defined as an infiltrate of at least 50 cells in one or more foci of inflammation, in excess of those seen in control animals; 2+ dacryoadenitis was defined as
272
INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / Februory 1987
Table 1. Elicitation of autoimmune lacrimal dacryoadenitis using crude antigen extracts Mode of sensitizalionf
Antigen*
Group
Lacrimal gland Lacrimal gland Harderian gland Submaxillary gland
CFA CFA CFA CFA CFA
Incidence]:
+ pertussis + pertussis + pertussis + pertussis
0/4 1/8 5/8 0/8 0/8
* All tissue antigens employed 10 mg of a crude 3 M KG extract. t CFA with or without antigen was injected in footpads, while pertussis was administrated i.v. % Numbers indicate number of animals positive/total number tested. Histologic assessment of lacrimal gland pathology was made 30 days after sensitization, and was recorded as positive when multiple foci occupied more than 10% of the gland (grade 2+ dacryoadenitis or greater).
multiple foci of cellular infiltration occupying 10-25% of the gland; 3+ dacryoadenitis indicated that the lesions occupied 25-50% of the gland. These investigations conform to the ARVO Resolution on the Use of Animals in Research.
Results
Table 1 summarizes the data on the incidence and specificity of autoimmune dacryoadenitis obtained in Lewis rats after a single immunization with lacrimal gland or control antigens. All rats were killed on day 30 after sensitization and their lacrimal glands processed for histological evaluation of dacryoadenitis. Immunization of rats with the crude 3M KC1 extract of lacrimal gland in CFA alone (group 2), caused one of eight to show significant pathologic changes in its lacrimal glands. However, the majority of rats (five of eight) immunized with the same extract in CFA, together with i.v. injection of pertussis (group 3), developed severe dacryoadenitis with lymphocytic infiltraTable 2. Dacryoadenitis induced in rats after sensitization with peak III exorbital lacrimal gland antigens
15 20 30 40 50 60
Exorbital lacrimal gland
Intraorbital lacrimal gland
Harderian gland
Salivary gland
1/3*
1/3 2/4 3/4 3/4 3/4 3/4
0/3 0/4 0/4 0/4 0/4 0/4
0/3 0/4 0/4 0/4 0/4 0/4
2/4 3/4 3/4 3/4 3/4
tion in both exorbital and intraorbital lacrimal glands. The pattern of destruction of glandular tissue was dramatic, and the inflammatory cell infiltration was greatly enhanced and multifocal. None of these rats showed significant lesions in their Harderian glands. In contrast, the lacrimal glands of those rats receiving the 3M KC1 extract of Harderian gland in CFA plus pertussis (group 4) contained only occasional minute foci of chronic inflammatory cells. These foci usually numbered one to two, never in excess of three or four, and appeared to occupy less than 5% of the gland (grade 1 + dacryoadenitis or less). Other control rats immunized with either saline (group 1) or the 3M KC1 extract of submaxillary gland (group 5) in CFA plus pertussis did not develop any pathologic changes in their lacrimal glands. Therefore, we selected a score of 2+ or greater as being positive dacryoadenitis for the experimental groups. It is important to point out that all rats immunized with the KC1 extract of Harderian gland in CFA plus pertussis (group 4), although developing no lacrimal gland disease, did show severe Harderian dacryoadenitis, a finding to be reported in an accompanying report. Sephacryl S-300 Chromatography of the KC1 Lacrimal Gland Extract and Sensitization With Fractionated Proteins
Elicitation of Autoimmune Dacryoadenitis With Crude Antigen Extracts
Days after immunization
Vol. 28
* Numbers indicate number of animals positive/total number tested. A lesion score of grade 2+ or greater was judged to be positive.
To obtain partially purified antigens, the 3M K G extract of the lacrimal gland was subjected to gel-filtration on a Sephacryl S-300 column, and separated into three protein fractions (Fig. 1). The molecular weight of the elutes was estimated after calibration of the column with known marker proteins. Peak I contained the highest MW proteins (> 100,000), found nearest the void volume; peak II contained proteins of approximately 60,000-90,000 daltons; and peak III was composed of proteins of approximately 10,000-55,000 daltons. Each of the three fractions was used to immunize separate groups of Lewis rats in order to test the partially purified material for the presence of active autoantigen. Sensitization with peak III proteins: Lewis rats were immunized with a single dose of 2.0 mg of peak III proteins in CFA followed by i.v. injection of pertussis, and were killed at intervals from 15-60 days. The development and the course of autoimmune dacryoadenitis were followed by histologic analysis, and are summarized in Table 2. Chronic inflammatory cellular infiltration was first noted 15-20 days after immunization. The area surrounding small collecting ducts appeared to be the primary inflammatory site, with involvement of the bulk of the gland in a diffuse distribution. The infiltrate was composed primarily of lymphocytes, with macrophages
No. 2
EXPERIMENTAL AUTOIMMUNE DACRYOADENITI5 / Liu er ol.
273
Fig. 2. Experimental autoimmune dacryoadenitis in the rat, 20day-old-Iesion. Small areas of the exorbital gland show diffuse inflammatory infiltrates, with early changes in some acini (arrows). The remainder of the gland appears normal. Dacryoadenitis grade 2+ (H & E, X 190).
Fig. 3. Experimental autoimmune dacryoadenitis in the rat, 30day-old-lesion, A relatively large proportion of the gland is involved in a diffuse inflammatory process. There is extensive acinar atrophy and some destruction. Dacryoadenitis grade 3+ (H & E, X 300).
and scattered plasma cells. In general there was little evidence of acinar distraction at this time, although a decrease in epithelial cell size suggestive of acinar atrophy was noted in the areas of inflammation, as shown in Figure 2. At 30-40 days postimmunization, there was more extensive inflammation, and glandular destruction was present in the most highly involved of these sites. At this time, the glandular epithelium showed vacuolar degeneration, and chronic inflammatory cells, principally lymphocytes, were infiltrating the acini. Small accumulations of inflammatory cells could be demonstrated within the acinar lumens, as shown in Figure 3. This level of inflammation appeared to spread throughout the gland, with lesions in most animals well advanced and the majority of rats showing grade 24- to 3+ dacryoadenitis. The histologic appearance of the exorbital lacrimal gland at this time was considered to represent the fully developed experimental inflammatory disease. In the most severe cases, acini were replaced by foci of inflammatory cells adjacent to small collecting ducts in the fibrous tissue stroma of the gland (Fig. 4). Although the inflammatory process was present throughout the gland, its distribution was still spotty and areas of intact acini were still present adjacent to inflammatory foci, as demonstrated in Figure 4. Among the more striking alterations in the normal glandular architecture at this time was the accumulation of large numbers of lymphocytes and macrophages within distended lymphatic channels adjacent to collecting ducts within the fibrous tissue stroma. Desquamated epithelium could also be demonstrated in large collecting ducts, in animals killed at days 50 and 60.
glands appeared to progress in parallel fashion for the entire period of observation, as did the extent and severity of the disease. The Harderian glands of those rats with severe lacrimal gland dacryoadenitis showed only minimal involvement. These were single, small focal collections of histocytes, involving less than 5% of the gland, and were similar to the mild changes seen in control animals injected with CFA and pertussis alone, without glandular extracts. The submaxillary, sublingual, and parotid glands of these animals showed no inflammatory changes above control levels at any time during the study. Sensitization with peak I and peak II proteins: Groups of rats were sensitized with the proteins from peak I and II, and killed on the same schedule as those receiving peak III proteins. Only the rare animal sen-
The intraorbital lacrimal gland presented the same incidence of infiltrative lesions as those seen in the exorbital gland. The histopathologic changes in both
Fig. 4. Experimental autoimmune dacryoadenitis in the rat, 40day-old-lesion. There is a widespread and severe focal and diffuse inflammatory process, with considerableinflammatory cell infiltration and destruction of acinar structures. Note the adjacent area of relatively normal acini. Dacryoadenitis grade 3+ (H & E, X 300).
274
INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / February 1987
sitized with peak II proteins showed mild histologic changes. These lesions consisted of small foci of inflammatory cells scattered through the gland, involving approximately 5-10% of the gland, unlike the extensive dacryoadenitis seen at same time after immunization with peak III proteins. It is possible that this was due to minor contamination of the peak II material by lower molecular weight proteins. No animals that received peak I proteins showed any sign of inflammatory disease. Discussion Experimental autoimmune dacryoadenitis was produced in Lewis rats by sensitization with a single dose of lacrimal gland extract emulsified in CFA. The incidence and severity of this experimental disease was markedly enhanced by the simultaneous administration of dead Bordetella pertussis, a bacterium known to exert a strong adjuvant effect in a number of different model systems for the production of autoimmune disease.9"1 ' The histopathologic features varied, depending on the time of death after immunization. Lesions in the lacrimal gland appeared 2 wk after sensitization with the active lacrimal gland fraction, but starting at 3 wk the extent and the severity of the disease progressed rapidly and reached a maximum at about 4-6 wk. Relatively severe dacryoadenitis continued for the full 60-day period of observation after immunization. The early lesions presented as small, perivascular accumulations of lymphocytes admixed with macrophages and plasma cells. As the process continued, the nongranulomatous lymphoproliferative nature of the inflammation became more apparent: small foci of inflammation became confluent and thus affected more of the glandular tissue, ultimately replacing much of the lacrimal gland, but still spared large areas. Preliminary efforts to purify the active lacrimal gland antigen by gelfiltrationof the KC1 extract showed that the active component, as assessed by the appearance of inflammation after immunization, was localized in a fraction containing proteins with molecular weight in the range of 10,000-55,000 daltons. Sensitization of rats with this fraction produced severe dacryoadenitis in most animals, whereas use of column fractions containing higher molecular weight material failed to produce disease. The lesions induced were tissue-specific, in that rats immunized with the active lacrimal gland fraction had intense lymphocyte infiltration in both exorbital and intraorbital lacrimal glands, but no significant lesions in their Harderian or salivary glands. The specificity of the response was further demonstrated by the inability to stimulate lacrimal gland lesions by immunization with analogous extracts of ei-
Vol. 28
ther Harderian or salivary glands. Indeed, we found that Lewis rats immunized with the active fraction of Harderian gland showed dacryoadenitis only in that tissue.12 Our data suggest that a common autoantigen exists in both exorbital and intraorbital lacrimal glands that is different from the one responsible for Harderian gland dacryoadenitis. The present studies appear to offer a conceptual framework for an understanding of the immunopathogenesis of certain forms of dacryoadenitis, since the histopathologic pictures seen in both the experimental as well as clinical disease are strikingly similar. We suggest that the pathogenesis of some forms of chronic dacryoadenitis depends upon autoimmunization against constituents of the patient's own lacrimal gland. An initial, nonspecific local inflammatory reaction may lead to tissue breakdown and the release of normally sequestered antigens. Systemic sensitization to these antigens would then set the stage for a chronic and progressive disease with many of the histologic and cytologic features seen not only in this experimental model of dacryoadenitis in the rat, but also in such autoallergic diseases as Hashimoto's thyroiditis, allergic encephalomyelitis, or the autoallergic retinouveitis induced by S-antigen.9"11 This infiltrative process causes the progressive destruction of tissue and eventually results in functional abnormalities. Failure of the lacrimal gland to secrete fluid of sufficient quantity and proper composition would result in dry eyes. Although conflicting information exists on Schirmer tests in patients with Sjogren's syndrome, the tear lysozyme content has been shown to diminish in patients with Sjogren's or sicca syndromes.1314 The present model system may provide a unique opportunity to study the relationship between the severity of these organ-specific lesions and the function of the lacrimal gland. Key words: dacryoadenitis, autoimmune dacryoadenitis, lacrimal gland autoimmunity, Harderian gland autoimmunity
References 1. Jakobiec FA and Jones IS: Orbital inflammations. In Diseases of the Orbit, Jones IS and Jakobiec FA, editors. Hagerstown, MD, Harper and Row, 1979, pp. 205-255. 2. Kessler HS, Cubberly M, and Manski W: Eye changes in autoimmune NZB and NZB X NZW mice. Arch Ophthalmol 85: 211, 1971. 3. DeLuise VP and Tabbara KF: NZB/NZW F, hybrid mice, an animal model of Sjogren's syndrome. In Animal Models of Ocular Disease, Tabbara KF and Cello RM, editors, Springfield, IL, Thomas, 1984, pp. 237-246. 4. Hoffman RW, Alspaug MA, Waggie KS, Durham JB, and Walker SE: Sjogren's syndrome in MRL/1 and MRL/n mice. Arthritis Rheum 27:157, 1984. 5. Jabs DA, Alexander EL, and Green WR: Ocular inflammation
No. 2
EXPERIMENTAL AUTOIMMUNE DACRYOADENITIS / Liu er ol.
in autoimmune MRL/Mp mice. Invest Ophthalmol Vis Sci 26: 1223, 1985. 6. Ohashi Y, So SK, Minaski PN, and Tabbara KF: The presence of cytotoxic autoantibody to lacrimal gland cells in NZB/W mice. Invest Ophthalmol Vis Sci 26:214, 1985. 7. Waterhouse JP: Focal adenitis in salivary and lacrimal glands. ProcRSocMed 56:911, 1963. 8. Mizejewski GJ: Studies of autoimmune induction in the rat lacrimal gland. Experientia 34:1093, 1978. 9. Levine S, Wenk EJ, Devlin HB, Pieroni RE, and Levine L: Hyperacute allergic encephalomyelitis: adjuvant effect of pertussis vaccines and extracts. J Immunol 97:363, 1966. 10. Rose NR, Twarog FJ, and Crowie AJ: Murine thyroiditis: im-
11.
12.
13.
14.
275
portance of adjuvant and mouse strain for the induction of thyroid lesions. J Immunol 106:698, 1971. Mochizuki M, Charley J, Kuwabara T, Nussenblatt RB, and Gery I: Involvement of the pineal gland in rats with experimental autoimmune uveitis. Invest Ophthalmol Vis Sci 24:1333, 1983. Liu SH, Sakai F, Prendergast RA, and Silverstein AM: Experimental autoimmune dacryoadenitis: II. Harderian gland disease in the rat. Invest Ophthalmol Vis Sci 28:72, 1987. Avisar R, Menache R, Shaked P, Rubinstein J, Machtey I, and Savir H: Lysozyme content of tears in patients with Sjogren's syndrome and rheumatoid arthritis. Am J Ophthalmol 87:148, 1979. DeLuise VP and Tabbara KF: Quantitation of tear lysozyme levels in dry-eye disorders. Arch Ophthalmol 101:634, 1983.
