- Email: [email protected]
δ T-cell receptors in recurrent aphthous ulceration (RAU)
Copyright © Munksgaard 2000
Int. J. Oral Maxillofac. Surg. 2000; 29:375-380 Printed in Denmark. All rights reserved
rmtemationalJot~alof
Oral& MaxillofaciaISurgery I S S N 0901-5027
Medicine
Increased densil of
lymph0c es bearing T-cell receptors recurrent aphthous ulceration (RAU)
Sirajedin S. Natah 1-3,s, Ritva H~iyrinen-lmmonen 2, Jarkko Hietanen 3, Pertti Patinen 2-4, Maria Malmstr6m 2, Erkki Savilahti s, Yrj6 T. Konttinen 1,2,8 1Department of Medicine, Helsinki University Central Hospital; 2Department of Oral Medicine, SDepartment of Oral Pathology, Institute of Dentistry; '*Department of Surgery, and 5Children's Hospital, University Hospital of Helsinki; Slnstitute of Biomedicine, University of Helsinki, Finland
S. S. Natah, R. Hiiyrinen-Immonen, J. Hietanen, P. Patinen, M. Malmstrdm, E. Savilahti, Y. T. Konttinen: Increased density of Iymphocytes bearing 7l~ T-cell receptors in recurrent aphthous ulceration (RA U). Int. J. Oral Maxillofac. Surg. 2000," 29." 375-380. © Munksgaard, 2000 Abstract. Lymphocytes bearing the T-cell :receptor (TCR) 7/c~are increased in the peripheral blood of patients with recurrent aphthous ulcers (RAU) and Behget's disease. In this study, we examined whether the density of TCR-7/c~ bearing lymphocytes was also increased locally in RAU lesions. Ten RAU lesions from ten patients were compared with ulcer-free mucosa from sites contralateral to the lesions, and with 10 samples of clinically healthy oral mucosa taken from 10 healthy volunteers. Samples were labeled with a panel of monoclonal antibodies specific to CD3, cd~ TCR and 7/6 TCR in avidin-biotin-peroxidase complex (ABC) staining. Lymphocytes expressing ,v/6 TCRs were very low in non-lesional mucosa and clinically healthy mucosa. By contrast, 7/c~T-cells were numerous and observed in all RAU lesions especially within the epithelium, inflammatory infiltrates and at perivascular locations. The count of 7/6 T-cells was high in connective tissue of RAU (200--- 126 cellshum 2) compared with connective tissue of controls (4---4 cells/mm2; P<0.0001) or non-lesional mucosa (5---7 cells/mm2). Interestingly, the density of 7/c~T-cells was also high in the epithelium of RAU (70--_34 cells/mm 2) compared with the epithelium of non-lesional mucosa (2.8-+06 cells/mm2; P<0.0001) or epithelium of healthy controls (1.2--_1.5 cells/mm2; P<0.0001). Moreover, the mean percentage of 7/&+ T-cells among total CD3+ lymphocytes was increased in the connective tissue area from 4% and 5% in controls and non-lesional mucosa, respectively, to 19% in RAU. In epithelial areas, the average percentage was increased from 2% and 6% in controls and non-lesional mucosa, respectively, to 36% in RAU. These data showed that 7/6 Tcells are more numerous in RAU lesions and such an increase was purely restricted to RAU inflammatory areas.
Recurrent aphthous ulceration (RAU) is an inflammatory condition characterized by painful, recurrent, single or multiple ulcerations of mucosal tissues 17. The definitive cause of RAU remains obscure and there is no curative therapy. Histopathologically, the lesions of RAU are characterized by superficial tissue necrosis with destruction of the
epithelium and an infiltration of activated T-lymphocytes and other inflammatory cells at the site of the lesion 12. Phenotypic investigations have revealed the preferential distribution of ),/ T-cells in epithelial tissue 2. T-cell receptor (TCR) 7lfi+ cells can be activated to produce interleukin-2 and/or
Key words: aphthous ulcers; T-cell receptors; ~'/8 T-cells. Accepted for publication 30 May 2000
cytotoxicity by bacterial heat-shock proteins, mycobacterial antigens and a variety of bacteria 1°,15,19. Earlier studies have shown that y/& T-cells proliferate in response to virus-infected cells and mediate their destructions 3, and their activity correlated with disease severity in HIV infection 21. High percentages of ?/& T-cells were
376
Natah et al.
observed in and around necrotic lesions of tuberculous lymphadenitis 5. As well, demonstrations o f an increased number of 7/6 T-cells in the active synovitis of rheumatoid arthritis 14 seem to support the proposition that )46 T-cells participate in inflammatory diseases. In certain diseases like coeliac disease, an increase of intestinal intraepithelial 7/6 T-cells has been demonstrated 8, whereas 50% of patients with Beh~et's disease have increased y/d Tcells in their peripheral blood 26. Heatshock protein 65 K D a , which is one of the targets of 7/6 T-cells, was reported to be increased in sera from patients with R A U TM, and increased levels of peripheral 7/6 T-cells have been reported in patients suffering from R A U 2a. This study aimed to examine the hypothesis that 7/6 T-cells also increase locally at the sites of R A U lesions. By analogy to the increased proportion of 7/6 T-cells in the peripheral blood o f patients with R A U during both active and inactive R A U 2a, we hypothesized that such increases are present in both active R A U lesions as well as in lesional-free mucosa. Using a panel of antibodies to CD3, a/fl and 7/d T-cells, we studied the frequency and distribution of T C R phenotypes o f lymphocytes in R A U lesions, ulcer-free mucosa from contralateral sites to the lesions, and clinically healthy mucosa from healthy volunteers.
Material and methods
Ten aphthous ulcers were excisionally biopsied from 10 patients with RAU (five women and five men; the mean age of patients was 32 years, ranging from 21 to 49 years). In addition, specimens from the clinically unaffected area at the corresponding site opposite to the ulcer were obtained from all patients with RAU to demonstrate differences between clinically unaffected mucosa and RAU lesions. For comparison, biopsies of clinically healthy oral mncosa were taken from non-keratinized cheek mucosa of ten healthy volunteers (six women and four men; the mean age was 36 years, ranging from 25 to 52 years) with a negative history of RAU. All patients included in this study had been investigated in our clinic during previous episodes of ulceration. The results of full blood examination, serum B12 vitamin and folate, were within the normal level. None of the patients or controls had a history of any gastrointestinal signs or symptoms other than those associated with minor illness. Entry requirements were that the patients should have classical minor RAU with at least one episode per month on average, so
that the patient had aphthae severe enough to motivate them to participate in the study. Any patients with features suggesting an extra oral or systemic disease were excluded from the study. Potential subjects were also excluded if they had treated their current ulcer with any form of topical or systemic medication. Diagnosis of RAU was made from clinical presentation (discrete, painful, shallow, recurrent ulcers, covered by a yellow-grey pseudomembrane and surrounded by an erythematous halo). At any given time, ulcers numbered one to five in the nonkeratinized mucosa of the mouth, and each measured less than 10 mm. The histological picture was that of a nonspecific ulcer with the absence of pathogens or pathogenic processes that cause other oral ulcers. All RAU lesions studied were 2-7 days old, minor RAU, and painful. RAU lesions were obtained from the cheek and/or labial mucosa. The study protocol was approved by The Ethics Committee of the Institute of Dentistry, University of Helsinki, Finland, and informed consent was obtained from each participant of the study. The biopsies were done under local anaesthesia (20 mg/rul xylocaine+12.5 #g/ml adrenalin, Astra Co., S6dertglje, Sweden), and the collected samples were snap-frozen in liquid nitrogen and embedded in a mounting media (OCT, Tissue Tek®; Miles, Elkhart, IN, USA) and stored at - 7 0 °C until cryostat sectioning. Monoclonal antibodies
Monoclonal antibody TCR 6 1 (T cell Sciences Inc., Cambridge, MA, USA) was used to recognize ?~ldTCR, and antibody fl F1 (T cell Sciences Inc.) to identify a/fl TCR positive lymphocytes, and both antibodies were used at a dilution 1:100. Monoclonal antibody anti-Leu4 (anti-CD3; BecktonDickinson, Mountain View, CA, USA) was used at dilution 1:400. Immunohistochemistry
An avidin-biotin peroxidase method was used for staining of the cells as previously described26. In brief, serial cryostat sections (45/~m) were cut, fixed in acetone for 20 rain at 4°C, then in chloroform for 20 min at 20°C and finally washed three times in phosphate-buffered saline (PBS, pH 7.2). After washing, intrinsic peroxidase activity was abolished by incubating tissue sections in 0.3% H202 for 30 min. Nonspecific binding sites were blocked by incubation in normal serum, and the sections covered with a dilution of monoclonal antibodies in PBS for 1 h. A Vectastain Elite ABC kit (PK-6102; Vector Laboratories; Burlingame, CA, USA) was used to show the binding of the monoclonal antibodies according to the manufacturer's instruction. Monoclonal mouse IgG with irrelevant specificity (Aspergillus niger glucose oxidase [Mab (IgG1) Anti-aspergillus niger, X0931, Dako A/S, Glostrup,
Denmark]) or normal mouse serum (Dako A/S) was used at the same concentration as and instead of the primary antibodies as a staining control. Tissue sections were examined under an Olympus BH-2 light microscope (Olympus Corp., Tokyo, Japan), using an eyepiece graticule and ×40 objective. The immunoreactivity of adjacent sections was compared and the number of positively stained lymphocytes was calculated in 20 consecutive high power fields (0.22×0.32 mm) from the intact epithelium of RAU, non-lesional mucosa and controls (as mean number of positive cells per mm 2 of surface epithelium), and from five different fields in underlying connective tissue (as mean number of cells/mm2 connective tissue). The results are given as mean and standard deviation (mean_+SD) to describe the dispersion of the data. The means of the three groups were compared by using the Kruskal-Wallis rank test, and the comparisons between the pairs of means by using the Mann-Whitney U-test, with a downward adjustment of the a level to compensate for multiple comparisons. Linear correlation was calculated between the densities of intraepithelial 7/6 T-cells and the ages of RAU lesions. A P-value of 0.01 or less was considered statistically significant. BMDP-PC, version 7.01 (BMDP statistical software, Cork, Ireland) was used for all calculations.
Results Histopathological features
R A U lesions reveal a mononuclear infiltrate at the subepithelial compartment with variable numbers of intraepithelial mononuclear cells that tend to be closely associated with basal keratinocytes. Other features include destruction of epithelium and accumulation of neutrophils at areas adjacent to the ulcer. Sections of non-lesional mucosa and healthy controls had very few intraepithelial inflammatory cells (mainly lymphocytes and p o l y m o r p h o nuclear leukocytes). Occasional infiltrating mononuclear inflammatory cells and resident cell population (like mast cells, and fibroblasts) were also seen scattered in the connective tissue beneath the epithelium in every specimen. Lymphocyte populations in the connective tissue
The anti-CD3, anti-cdfl T C R , and antiT C R 7/6 antibodies staining resulted in a ring-shaped m e m b r a n o u s staining of lymphocytes. R A U samples revealed many C D 3 + T-cells within the inflammatory infiltrate areas at the subepithelial compart-
7/0 T cells in recurrent aphthous ulcers
377
Table 1. Density of CD3, ~/fl, and 7/c~T-cell lymphocytes in the connective tissue areas in clinically healthy oral mucosa (CHM), non-lesional mucosa (NLM), and RAU lesions t
CD3+ T-cells
Subject no. 1 2 3 4 5 6 7 8 9 10 Mean_+SD K-W test
CHM 138_+70 185_+89 198_+66 150-+33 65_+44 161_+100 85_+08 265_+75 166_+58 228_+90 164_+60 -
NLM 178+35 125_+19 78_+11 130_+46 131_+25 80-+10 71_+08 145_+17 89_+06 257_+26 128_+57 NS
a/fl T-cells RAU
CHM
1625_+517 2318_+321 754_+299 1206_+325 787_+282 997_+1~94 1028_+294 1201_+392 758_+89 990_+274 1166_+483 <0.001"
137_+84 197_+144 158-+40 153-+52 137_+25 125_+100 52_+10 220_+48 98_+46 214-+85 149_+52
?/&
NLM
RAU
170+19 116_+14 83_+11 128_+34 123_+24 88_+08 60_+07 140_+13 86_+41 234_+15 123_+51 NS
1162_+239 1891_+614 645_+278 1139-+234 433_+75 944_+297 1000_+98 906_+279 417_+125 314_+155 885_+469 <0.001"
T-cells
CHM
NLM
05_+12 08_+12 0 03_+06 05_+12 05_+07 13_+01 0 0 0 04_+04 -
05_+07 14+06 0 0 0 0 21_+17 0 02_+03 06_+08 05_+07 NS
Age of RAU (days)
RAU 60_+41 372+_115 131-+58 96-+66 433_+75 160_+40 138_+78 137_+43 160+66 314_+155 200_+126 <0.001"
5 2 2 7 3 3~4 6 4 3 3
NS=non significant vs clinically healthy oral mucosa. * Significant vs non-lesional mucosa or vs clinically healthy oral mucosa. K-W test=Kruskal-Wallis rank test.
ment (1166-+483 cells/mm 2) compared with non-lesional mucosa (128-+57 cells/ram 2) and healthy controls (164-+60 cells/mm2, Table 1). In all of the samples examined, T-cells expressing the alfl T C R formed the d o m i n a n t subset of C D 3 + T-cells. The n u m b e r of alfl T-cells was much higher in R A U (885-+469 cells/mm ~) than that found in both the clinically n o r m a l mucosa of patients with R A U (123-+ 51 cells/mm 2) and mucosa of healthy controls (149-+52 cells/mm2; Kruskal-Wallis test, df=2, P<0.0001). ~16 T-cells in the n o n lesional mucosa and the mucosa of healthy controls were rare or absent, comprising 4-5% of the CD3 + population (Fig. 1). By contrast, 7/8 T-cells
were more numerous and were observed in all R A U lesions especially within the inflammatory infiltrates and at the perivascular locations (Fig. 2). They represented 19% of CD3 + lymphocytes (Fig. 1). Counts per n l l T l 2 w e r e 4-+4 in controls, 5-+7 in non-lesional mucosa, and 200-+126 in R A U (Kruskal-Wallis test, df=2, P<0.0001). Lymphocyte populations in the epithelium
For C D 3 + cells in R A U lesions, nonlesional mucosa, and healthy controls the counts were 210+80, 80-+30 and 73-+31 per m m 2 epithelium, respectively. T-cells bearing alfl T C R and 7/& T C R were scattered singly between
60
~P+++m 50
ca I,,,-. +
CHM NLM RAU
basal and supra-basal epithelial cells of R A U lesions and control specimens. A n analysis of intraepithelial cell counts demonstrated increases in the numbers of afl-TCR-positive lymphocytes in R A U lesions (118-+33 cells/mm ~) compared with the non-lesional mucosa of patients with R A U (72-+24) and the mucosa of healthy controls (63-+27 cells/mm~; Kruskal-Wallis test, df=2, P<0.001). 7/c~+ T-cells were seen in 5 of the 10 non-lesional mucosa, and in 4 of the 10 control samples (Table 2), and represented 6% and 2% of the C D 3 + lymphocytes in non-lesional and healthy controls epithelium, respectively (Fig. 1). In R A U lesions, the increase of 7/c~ T-cells tended to be focal (affecting some epithelial ridges more than others). The high density of 7/c~ T-cells was mostly seen at the lower epithelial layers and basement m e m b r a n e zone (Fig. 3). A wide individual variation of ~/6 T-cell count was observed in the epithelium of the R A U lesions (Table 2)
40
•
30
" ~,-~"
. 4~
~
~¢~
~'~+a.
~+~,,
.......
~a 20
++,.++ ~+m+ ",,,;Lp+++ ~ V '"i +~'~" ++:+,+ \"
+++ +. 5
10
+'++
+
m , +' . ~ ++ • , ++p+.~:;-+~ -,.+++>: ++++, .
• ~:'~ I+,++: ,~~'++~' ++ ¢,,~
++,,+
+++ ,++~~ + .
++,++:
Fig. 2. T-cells with y/6 TCR in the subepith-
Epithelium
Connective tissue
Fig. l. Clinically healthy mucosa, non-lesional mucosa, and RAU: 716 T-cell counts expressed as a proportion of CD3+ T-cells.
elial inflammatory infiltrate of an RAU lesion. Note the increased density of ~/6 Tcells at the perivascular locations (V=blood vessel). Original magnification ×250.
378
Natah et aL
Table 2. Density of CD3, a/fl, and 7/~ T-cell lymphocytes in the epithelium of clinically healthy oral mucosa (CHM), non-lesional mucosa (NLM), and RAU lesions a/fl+ T-cells
CD3+ T-cells Subject no. 1 2 3 4 5 6 7 8 9 10 Mean-+SD K-W test
CHM 73-+30 65--+21 112-+29 75-+55 36-+31 22-+16 65+-13 128-+61 80-+40 70-+35 73+31 -
NLM
RAU
CHM
2 6 - + 2 0 221-+44 94_+22 325-+74 59+17 63-+22 9 7 - + 2 5 191-+60 6 4 - + 3 3 286-+65 7 0 - + 4 0 208-+99 57+10 245-+85 119+_68 1 4 7 - + 4 7 110-+62 282-+42 108---24 130-+26 80_+30 210-+80 NS <0.00l*
NLM
73+24 25-+10 80-+32 97-+26 42+30 60-+12 87-+66 77-+41 36-+30 58±30 69-+54 54-+21 35+28 63-+12 113-+50 92-+36 33-+29 87-+65 62-+27 103-+34 63-+27 72+24 NS
7/3+ T-cells RAU 104-+50 195-+26 111+37 106±51 104-+21 125-+69 111-+22 119-+49 139--_50 63_+20 118-+33 <0.001"
CHM
NLM
RAU
03-+06 0 5 - + 0 8 35_+18 03-+06 0 128-+38 0 0 40-+15 03-+05 0 32-+21 0 01-+02 104---21 0 0 68-+56 03-+06 19-+08 111-+31 0 0 67-+55 0 02+03 49-+30 0 01-+02 63_+20 1.2-+1.5 2.8_+06 70_+34 NS <0.001"
Age of RAU (days) 5 2 2 7 3 34 6 4 3 3
NS=non significant vs clinically healthy oral mucosa. * Significant vs non-lesional mucosa or vs clinically healthy oral mucosa. K-W test=Kruskal-Wallis rank test.
with a mean percentage of 36% of intraepithelial CD3+ population (Fig. 1). A significant increase in the numbers of 7/ O-TCR-positive cells in RAU lesions was observed (Figs. 14). The count of 716 T-cells was much higher in the epi-
thelium of RAU (70-+34 cells/mm2) compared with the epithelium of nonlesional mucosa from patients with RAU (2.8-+06 cells/mm2; KruskalWallis test, df=2, P<0.0001), and the epithelium of normal controls (1.2-+ 1.5 cells/ram2; Kruskal-Wallis test, df=2, P<0.0001). No significant increases in 7/fi T-cells were found in non-lesional mucosa when compared with healthy controls. A weak and non-significant negative correlation was found between the densities of intraepithelial 7lfi T-cells and the duration of RAU lesions (r= -0.25).
Discussion
Fig. 3. Numerous t 6 T-cells were demonstrated in the epithelium and subepithelial compartments of a RAU lesion. The increased number of intraepithelial y/6 Tlymphocytes is more evident in the lower epithelial layers (small arrows). Original magnification ×250.
Fig. 4. Only a few fl3-TCR-positive cells are seen in the non-lesional mucosa from the countersite of RAU. Original magnification ×160.
In this discussion (unless specified), we used the term controls to denote both non-lesional mucosa and clinically healthy oral mucosa. Our data showed that the epithelium of RAU lesions contained three times as many CD3+ lymphocytes compared with the epithelium of controls. These findings seem to agree with most former pathological studies that have found frequent infiltration of lymphocytes to the lower epithelial layers in RAU lesions x7. In the present study, most of the CD3 lymphocytes in all samples expressed the a/fl TCR. Although we found that the absolute count of TCR-afl+ was increased in RAU lesions, their proportion to CD3 T cells was reduced compared with the mucosa of controls, apparently because of the relative increase of 716 T-cells. Such a relative reduction of the fraction of TCR-afl+ cells was also found in the peripheral
blood of patients with active RAU compared with patients in the inactive disease stage and the healthy controls22. Lymphocytes expressing t 6 TCRs were found in a small proportion ( - 2 % of CD3+ T-cells) in the healthy control mucosa, a value similar to that previously reported 23. By contrast to the results obtained in mice 13, 7/6 T-cells were uncommon in the epithelium and in the connective tissue of normal human oral mucosa. Our results demonstrated that the high density of 7/6 T-cells was restricted to RAU inflammatory lesions. In nonlesional mucosa from opposite sites of RAU, 7/6 T-cells were rarely seen in the epithelium or in the connective tissue areas. Although we were unable to determine in the present retrospective study if the increased percentage of f16 cells in RAU lesions simply reflected the numbers in the peripheral blood, previous work showed an increased proportion of 7/0 T-cells in the peripheral blood of patients with RAU during both inactive and active aphthous ulceration22. Furthermore, SUZUKI et al. 27 have shown that the concentration of circulating fl 6+ T-cells was higher in patients suffering from Behqet's disease with mucocutaneous RAU compared to patients without mucocutaneous lesions. The requirements for 7/6 T-cell activation are less stringent than those for a/fl T-cell activation. For instance, 7/0 Tcells can recognize antigens directly, without the need for specialized antigen-presenting cells 26. Furthermore, 7/ O T-cells can also react to non-peptidic antigens such as a lipid fraction of M.
713 T cells in recurrent aphthous ulcers tuberculosis 28, and various phosphorylated metabolites 24. Such characteristic features of 7/3 T-cells would probably enable them to respond quickly to a variety of antigens before the a/fl lymphocyte population begins to expand 4. The bowel affected by coeliac disease contains higher than n o r m a l numbers of 7/6 T-cells in either active or "silent" periods of the disease (when the pathology is mildest). Furthermore, the density of such cells is higher in the epithelium than in the lamina propria 16. Such expansion of T C R 7/3+ intraepithelial lymphocytes might be driven by an intestinal allergy to cereal proteins associated with autoimmunity directed against tissue transglutaminase. In aphthae, the high density of 713 T-cells was present exclusively in R A U lesions and in both the epithelium and the subepithelial connective tissue. Thus, it is possible that some of 713 T-cells were responding in an antigen-specific fashion and such antigenic stimuli for 713 T-cells were expressed only during the development/progression of RAU. The exact biological function of 7/3 T-cells in the epithelium of R A U lesions remains unclear. However, BOISMENU& HAVRAN (1994) have shown that activated 7/d T-cells produce the epithelial cell-specific fibroblast growth factor (FGF)-71. The effects of F G F - 7 on the proliferation and differentiation of epithelial cells may help to reduce mucosal damage following tissue injury 29. This is mostly in line with our recent study 2° on expression of the cell proliferationassociated nuclear antigen (Ki-67) in R A U lesions that showed higher proliferation activity in the epithelium of R A U compared with that in the epithelium of controls. M a n y reports have shown that different populations of )43 T-cells have different functions and different patterns of cytokine secretion. For instance, FERRICK et al. 6 found that 7/d T-cells can produce either T-helper type-1 ( T h l ) or Th2 cytokines depending on the pathogen they are exposed to. However, the role of 7/3 T-cells may not be limited to merely enhancing epithelial proliferation or modulating i m m u n e responses of alfl T cells 7. For instance, other investigators 9,11 have suggested that fld T-cells may play a role in immunological damage of aphthae. The "Janus-like" coexistence of cytotoxic and protective potential of fld T-cells might enable them to play a dual func-
tion in the pathogenesis of RAU, thereby acting as immunoregulator to maintain the homeostasis and functional integrity of the epithelial barrier. It should be pointed out that there is a wide variation in the density of 7/3 Tcells between R A U samples in this study. Similar wide individual variations were also noted in the peripheral blood of Behget's disease 27 and R A U patients 22. The reasons for these variations are currently unknown. Although our results provide no answers regarding the function of ),/d Tcells in R A U lesions, they showed that 7/3 T-cells are increased in R A U lesions and such an increase was concomitant with the R A U inflammatory process and did n o t extend to mucosal areas other than the inflammation sites.
Acknowledgements. We are grateful to Mrs. Sirkku Kristiansen and Mrs. Marjatta Kivek~is for skillful technical assistance. This research was supported by the Finnish Academy, and Helsinki University Central Hospital.
References 1. BOISMENUR, HAVRANWL. Modulation of epithelial cell growth by intraepithelial 73 t cells. Science 1994: 266:1253 5. 2. Bos JD, TEUNISSENMBM, CAIROI, et al. T-cell receptor gamma delta bearing cells in normal human skin. J Invest Dermatol 1990: 94: 37-42. 3. BUKOWSKIJF, MORITACT, BRENNERMB. Recognition and destruction of virus-infected cells by human 73 CTL. J Immunol 1994: 153: 5133-40. 4. DE LIBEROG. Sentinel function of broadly reactive human ?,d T cells. Immunol Today 1997: 18: 22-6. 5. FALINIB, FLENGHIL, PILERIS, et al. Distribution of T cells bearing different forms of the T cell receptor 7/d in normal and pathological human tissues. J Immunol 1989: 143: 2480-8. 6. FERRICKDA, SCHRENZELME), MULVANIA T, HSIEH B, FERLIN WG, LEPPER H. Differential production of interferon-7 and interleukin-4 in response to Thl- and Th2- stimulating pathogens by 7/3 T cells in vivo. Nature 1995: 373: 255-7. 7. FUJIHASI-IIK, McGHEE JR, KWEONMN, et al. 7/3 T cell-deficient mice have impaired mucosal immunoglobulin A responses. J Exp Med 1996: 183: 1929-35. 8. HALSTENSENTS, SCOTTH, BRANDTZAEG P. Intraepithelial T cells of the TCR 7/ 3+ CD8- and V31/J31 + phenotypes are increased in coeliac disease. Scand J Immunol 1989: 30: 665-72. 9. HAMZAOUIK, HAMZAOOIA, HENTATIF, et al. Phenotype and functional profile of T cells expressing 73 receptor from pa-
379
tients with active Behqet's disease. J Rheumatol 1994: 21: 2301-6. 10. HAREGEWOIN A, SOMAN G, HOM RC, FINBERG RW Human 7fi+ T cells respond to mycobacterial heat-shock protein. Nature 1989: 340: 309-12. 11. HASANA, FORTUNEF, WILSONA, et al. Role of 7/6 T cells in pathogenesis and diagnosis of Behqet's disease. Lancet 1996: 347: 789-94. 12. H)~YRINEN-IMMONEN R. Immune-activation in recurrent oral ulcers ROU. Scand J Dent Res 1992: 100: 222-7. 13. ITOHARAS, FARRAG, LAFAILLEJJ, et al. Homing of a gamma delta thymocyte subset with homogenous T-cell receptors to mucosal epithelia. Nature 1990: 343: 754-7. 14. JACOBS MR, HAYNES BE Increase in TCR 73 T lymphocytes in synovia from rheumatoid arthritis patients with active synovitis. J Clin Immunol 1992: 12: 1308. 15. JANIS EM, KAUFMANNSHE, SCHWARTZ RH, PARDOLL DM. Activation of 73 T cells in the primary immune response to Mycobacterium tuberculosis. Science 1989: 244: 713-6. 16. KUTLU T, BROUSSEN, RAMBAUDC, LE DEIST F, SCHMITZJ, CERF-BENSUSSANN. Numbers of T cell receptor (TCR) aft+ but not of TcR ~3+ intraepithelial lymphocytes correlate with the grade of villous atrophy in coeliac patients on a long term normal diet. Gut 1993: 34: 208-14. 17. LEHNERT. Pathology of recurrent oral ulceration and oral ulceration in Beh~et's syndrome: light, electron and fluorescence microscopy. J Pathol 1969: 97: 481-94. 18. LEHNERT, LAVERYE, SMITHR, VAN DER ZEE R, MIZUSHIMAY, SHINNICI(T Association between the 65-kilodalton heat shock protein, streptococcus sanguis, and the corresponding antibodies in Beh~et's syndrome. Infect Immun 1991: 59: 143441. 19. MUNK ME, GATRILL AJ, KAUFMANN SHE. Target cell lysis and IL-2 secretion by 73 T lymphocytes after activation with bacteria. J Immunol 1990: 145: 2434-9. 20. NATAH SS, HIETANEN J, H.;~YRINEN-IMMONEN R, JUNGELL P, MALMSTROMM, KONTTINENYT. Expression of cell proliferation-associated nuclear antigen (Ki67) in recurrent aphthous ulcers (RAU). Oral Med Pathol 1998: 3: 29-34. 21. NORAZMIM-N, ARIFIN H, JAMARUDDIN MA. Increased level of activated 73 lymphocytes correlates with disease severity in HIV infection. Immunol Cell Biol 1995: 73:245 8. 22. PEDERSENA, RYDERLE 73 T-cell fraction of peripheral blood is increased in recurrent aphthous ulceration. Clin Immunol Immunopathol 1994: 72: 98-104. 23. PEPIN LF, ROGERT, MORISSETJ, SEMAN M. Preferential V31 expression among TcR fl6-bearing T cells in human oral
380
N a t a h et al.
epithelium. Scand J Immunol 1993: 37: 28%94. 24. PORCELLI SA, MORITA CT, MODLIN RL. T-cell recognition of non-peptide antigens. Curr Opin Immunol 1996: 8: 5106. 25. SAVILAHTI E, REUNALA T, M•KI M. Increase of lymphocytes bearing the f13 T cell receptor in the jejunum of patients with dermatitis herpetiformis. Gut 1992: 33:206-11. 26. SCHILD H, MAVADDATN, LITZENBERGER C, et al. The nature of major histocompatibility complex recognition by 73 T cells. Cell 1994: 76: 29-37. 27. SuzuKI Y, HOSHI K, MATSUDA T, MI-
ZUSI-IIMA Y. Increased peripheral blood ~c~+ T cells and natural killer cells in Behget's disease. J Rheumatol 1992: 19: 588-92. 28. TSUYUGUCHI I, KAWASUMI H, UETA C, YANO I, KISHIMOTO S. Increase of T-cell receptor gamma/delta-bearing T cells in cord blood of newborn babies obtained by in vitro stimulation with mycobacterial cord factor. Infect Immun 1991: 59: 3053-9. 29. ZEEHJM, PROCACCINOF, HOFFMANNP, et al. Keratinocyte growth factor ameliorates mucosal injury in an experimental model of colitis in rats. Gastroenterology 1996: 110: 1077--83.
Address: Dr Sirajedin S. Natah Inflammation Research Group ( Tules Ryhmgi) Institute o f Biomedicine P.O. Box 9 Siltavuorenpenger 20A FIN-O0014 University o f Helsinki Helsinki Finland Tel." +35891918576 Fax: +358 91918499
Int. J. Oral Maxillofac. Surg. 2000; 29:375-380 Printed in Denmark. All rights reserved
rmtemationalJot~alof
Oral& MaxillofaciaISurgery I S S N 0901-5027
Medicine
Increased densil of
lymph0c es bearing T-cell receptors recurrent aphthous ulceration (RAU)
Sirajedin S. Natah 1-3,s, Ritva H~iyrinen-lmmonen 2, Jarkko Hietanen 3, Pertti Patinen 2-4, Maria Malmstr6m 2, Erkki Savilahti s, Yrj6 T. Konttinen 1,2,8 1Department of Medicine, Helsinki University Central Hospital; 2Department of Oral Medicine, SDepartment of Oral Pathology, Institute of Dentistry; '*Department of Surgery, and 5Children's Hospital, University Hospital of Helsinki; Slnstitute of Biomedicine, University of Helsinki, Finland
S. S. Natah, R. Hiiyrinen-Immonen, J. Hietanen, P. Patinen, M. Malmstrdm, E. Savilahti, Y. T. Konttinen: Increased density of Iymphocytes bearing 7l~ T-cell receptors in recurrent aphthous ulceration (RA U). Int. J. Oral Maxillofac. Surg. 2000," 29." 375-380. © Munksgaard, 2000 Abstract. Lymphocytes bearing the T-cell :receptor (TCR) 7/c~are increased in the peripheral blood of patients with recurrent aphthous ulcers (RAU) and Behget's disease. In this study, we examined whether the density of TCR-7/c~ bearing lymphocytes was also increased locally in RAU lesions. Ten RAU lesions from ten patients were compared with ulcer-free mucosa from sites contralateral to the lesions, and with 10 samples of clinically healthy oral mucosa taken from 10 healthy volunteers. Samples were labeled with a panel of monoclonal antibodies specific to CD3, cd~ TCR and 7/6 TCR in avidin-biotin-peroxidase complex (ABC) staining. Lymphocytes expressing ,v/6 TCRs were very low in non-lesional mucosa and clinically healthy mucosa. By contrast, 7/c~T-cells were numerous and observed in all RAU lesions especially within the epithelium, inflammatory infiltrates and at perivascular locations. The count of 7/6 T-cells was high in connective tissue of RAU (200--- 126 cellshum 2) compared with connective tissue of controls (4---4 cells/mm2; P<0.0001) or non-lesional mucosa (5---7 cells/mm2). Interestingly, the density of 7/c~T-cells was also high in the epithelium of RAU (70--_34 cells/mm 2) compared with the epithelium of non-lesional mucosa (2.8-+06 cells/mm2; P<0.0001) or epithelium of healthy controls (1.2--_1.5 cells/mm2; P<0.0001). Moreover, the mean percentage of 7/&+ T-cells among total CD3+ lymphocytes was increased in the connective tissue area from 4% and 5% in controls and non-lesional mucosa, respectively, to 19% in RAU. In epithelial areas, the average percentage was increased from 2% and 6% in controls and non-lesional mucosa, respectively, to 36% in RAU. These data showed that 7/6 Tcells are more numerous in RAU lesions and such an increase was purely restricted to RAU inflammatory areas.
Recurrent aphthous ulceration (RAU) is an inflammatory condition characterized by painful, recurrent, single or multiple ulcerations of mucosal tissues 17. The definitive cause of RAU remains obscure and there is no curative therapy. Histopathologically, the lesions of RAU are characterized by superficial tissue necrosis with destruction of the
epithelium and an infiltration of activated T-lymphocytes and other inflammatory cells at the site of the lesion 12. Phenotypic investigations have revealed the preferential distribution of ),/ T-cells in epithelial tissue 2. T-cell receptor (TCR) 7lfi+ cells can be activated to produce interleukin-2 and/or
Key words: aphthous ulcers; T-cell receptors; ~'/8 T-cells. Accepted for publication 30 May 2000
cytotoxicity by bacterial heat-shock proteins, mycobacterial antigens and a variety of bacteria 1°,15,19. Earlier studies have shown that y/& T-cells proliferate in response to virus-infected cells and mediate their destructions 3, and their activity correlated with disease severity in HIV infection 21. High percentages of ?/& T-cells were
376
Natah et al.
observed in and around necrotic lesions of tuberculous lymphadenitis 5. As well, demonstrations o f an increased number of 7/6 T-cells in the active synovitis of rheumatoid arthritis 14 seem to support the proposition that )46 T-cells participate in inflammatory diseases. In certain diseases like coeliac disease, an increase of intestinal intraepithelial 7/6 T-cells has been demonstrated 8, whereas 50% of patients with Beh~et's disease have increased y/d Tcells in their peripheral blood 26. Heatshock protein 65 K D a , which is one of the targets of 7/6 T-cells, was reported to be increased in sera from patients with R A U TM, and increased levels of peripheral 7/6 T-cells have been reported in patients suffering from R A U 2a. This study aimed to examine the hypothesis that 7/6 T-cells also increase locally at the sites of R A U lesions. By analogy to the increased proportion of 7/6 T-cells in the peripheral blood o f patients with R A U during both active and inactive R A U 2a, we hypothesized that such increases are present in both active R A U lesions as well as in lesional-free mucosa. Using a panel of antibodies to CD3, a/fl and 7/d T-cells, we studied the frequency and distribution of T C R phenotypes o f lymphocytes in R A U lesions, ulcer-free mucosa from contralateral sites to the lesions, and clinically healthy mucosa from healthy volunteers.
Material and methods
Ten aphthous ulcers were excisionally biopsied from 10 patients with RAU (five women and five men; the mean age of patients was 32 years, ranging from 21 to 49 years). In addition, specimens from the clinically unaffected area at the corresponding site opposite to the ulcer were obtained from all patients with RAU to demonstrate differences between clinically unaffected mucosa and RAU lesions. For comparison, biopsies of clinically healthy oral mncosa were taken from non-keratinized cheek mucosa of ten healthy volunteers (six women and four men; the mean age was 36 years, ranging from 25 to 52 years) with a negative history of RAU. All patients included in this study had been investigated in our clinic during previous episodes of ulceration. The results of full blood examination, serum B12 vitamin and folate, were within the normal level. None of the patients or controls had a history of any gastrointestinal signs or symptoms other than those associated with minor illness. Entry requirements were that the patients should have classical minor RAU with at least one episode per month on average, so
that the patient had aphthae severe enough to motivate them to participate in the study. Any patients with features suggesting an extra oral or systemic disease were excluded from the study. Potential subjects were also excluded if they had treated their current ulcer with any form of topical or systemic medication. Diagnosis of RAU was made from clinical presentation (discrete, painful, shallow, recurrent ulcers, covered by a yellow-grey pseudomembrane and surrounded by an erythematous halo). At any given time, ulcers numbered one to five in the nonkeratinized mucosa of the mouth, and each measured less than 10 mm. The histological picture was that of a nonspecific ulcer with the absence of pathogens or pathogenic processes that cause other oral ulcers. All RAU lesions studied were 2-7 days old, minor RAU, and painful. RAU lesions were obtained from the cheek and/or labial mucosa. The study protocol was approved by The Ethics Committee of the Institute of Dentistry, University of Helsinki, Finland, and informed consent was obtained from each participant of the study. The biopsies were done under local anaesthesia (20 mg/rul xylocaine+12.5 #g/ml adrenalin, Astra Co., S6dertglje, Sweden), and the collected samples were snap-frozen in liquid nitrogen and embedded in a mounting media (OCT, Tissue Tek®; Miles, Elkhart, IN, USA) and stored at - 7 0 °C until cryostat sectioning. Monoclonal antibodies
Monoclonal antibody TCR 6 1 (T cell Sciences Inc., Cambridge, MA, USA) was used to recognize ?~ldTCR, and antibody fl F1 (T cell Sciences Inc.) to identify a/fl TCR positive lymphocytes, and both antibodies were used at a dilution 1:100. Monoclonal antibody anti-Leu4 (anti-CD3; BecktonDickinson, Mountain View, CA, USA) was used at dilution 1:400. Immunohistochemistry
An avidin-biotin peroxidase method was used for staining of the cells as previously described26. In brief, serial cryostat sections (45/~m) were cut, fixed in acetone for 20 rain at 4°C, then in chloroform for 20 min at 20°C and finally washed three times in phosphate-buffered saline (PBS, pH 7.2). After washing, intrinsic peroxidase activity was abolished by incubating tissue sections in 0.3% H202 for 30 min. Nonspecific binding sites were blocked by incubation in normal serum, and the sections covered with a dilution of monoclonal antibodies in PBS for 1 h. A Vectastain Elite ABC kit (PK-6102; Vector Laboratories; Burlingame, CA, USA) was used to show the binding of the monoclonal antibodies according to the manufacturer's instruction. Monoclonal mouse IgG with irrelevant specificity (Aspergillus niger glucose oxidase [Mab (IgG1) Anti-aspergillus niger, X0931, Dako A/S, Glostrup,
Denmark]) or normal mouse serum (Dako A/S) was used at the same concentration as and instead of the primary antibodies as a staining control. Tissue sections were examined under an Olympus BH-2 light microscope (Olympus Corp., Tokyo, Japan), using an eyepiece graticule and ×40 objective. The immunoreactivity of adjacent sections was compared and the number of positively stained lymphocytes was calculated in 20 consecutive high power fields (0.22×0.32 mm) from the intact epithelium of RAU, non-lesional mucosa and controls (as mean number of positive cells per mm 2 of surface epithelium), and from five different fields in underlying connective tissue (as mean number of cells/mm2 connective tissue). The results are given as mean and standard deviation (mean_+SD) to describe the dispersion of the data. The means of the three groups were compared by using the Kruskal-Wallis rank test, and the comparisons between the pairs of means by using the Mann-Whitney U-test, with a downward adjustment of the a level to compensate for multiple comparisons. Linear correlation was calculated between the densities of intraepithelial 7/6 T-cells and the ages of RAU lesions. A P-value of 0.01 or less was considered statistically significant. BMDP-PC, version 7.01 (BMDP statistical software, Cork, Ireland) was used for all calculations.
Results Histopathological features
R A U lesions reveal a mononuclear infiltrate at the subepithelial compartment with variable numbers of intraepithelial mononuclear cells that tend to be closely associated with basal keratinocytes. Other features include destruction of epithelium and accumulation of neutrophils at areas adjacent to the ulcer. Sections of non-lesional mucosa and healthy controls had very few intraepithelial inflammatory cells (mainly lymphocytes and p o l y m o r p h o nuclear leukocytes). Occasional infiltrating mononuclear inflammatory cells and resident cell population (like mast cells, and fibroblasts) were also seen scattered in the connective tissue beneath the epithelium in every specimen. Lymphocyte populations in the connective tissue
The anti-CD3, anti-cdfl T C R , and antiT C R 7/6 antibodies staining resulted in a ring-shaped m e m b r a n o u s staining of lymphocytes. R A U samples revealed many C D 3 + T-cells within the inflammatory infiltrate areas at the subepithelial compart-
7/0 T cells in recurrent aphthous ulcers
377
Table 1. Density of CD3, ~/fl, and 7/c~T-cell lymphocytes in the connective tissue areas in clinically healthy oral mucosa (CHM), non-lesional mucosa (NLM), and RAU lesions t
CD3+ T-cells
Subject no. 1 2 3 4 5 6 7 8 9 10 Mean_+SD K-W test
CHM 138_+70 185_+89 198_+66 150-+33 65_+44 161_+100 85_+08 265_+75 166_+58 228_+90 164_+60 -
NLM 178+35 125_+19 78_+11 130_+46 131_+25 80-+10 71_+08 145_+17 89_+06 257_+26 128_+57 NS
a/fl T-cells RAU
CHM
1625_+517 2318_+321 754_+299 1206_+325 787_+282 997_+1~94 1028_+294 1201_+392 758_+89 990_+274 1166_+483 <0.001"
137_+84 197_+144 158-+40 153-+52 137_+25 125_+100 52_+10 220_+48 98_+46 214-+85 149_+52
?/&
NLM
RAU
170+19 116_+14 83_+11 128_+34 123_+24 88_+08 60_+07 140_+13 86_+41 234_+15 123_+51 NS
1162_+239 1891_+614 645_+278 1139-+234 433_+75 944_+297 1000_+98 906_+279 417_+125 314_+155 885_+469 <0.001"
T-cells
CHM
NLM
05_+12 08_+12 0 03_+06 05_+12 05_+07 13_+01 0 0 0 04_+04 -
05_+07 14+06 0 0 0 0 21_+17 0 02_+03 06_+08 05_+07 NS
Age of RAU (days)
RAU 60_+41 372+_115 131-+58 96-+66 433_+75 160_+40 138_+78 137_+43 160+66 314_+155 200_+126 <0.001"
5 2 2 7 3 3~4 6 4 3 3
NS=non significant vs clinically healthy oral mucosa. * Significant vs non-lesional mucosa or vs clinically healthy oral mucosa. K-W test=Kruskal-Wallis rank test.
ment (1166-+483 cells/mm 2) compared with non-lesional mucosa (128-+57 cells/ram 2) and healthy controls (164-+60 cells/mm2, Table 1). In all of the samples examined, T-cells expressing the alfl T C R formed the d o m i n a n t subset of C D 3 + T-cells. The n u m b e r of alfl T-cells was much higher in R A U (885-+469 cells/mm ~) than that found in both the clinically n o r m a l mucosa of patients with R A U (123-+ 51 cells/mm 2) and mucosa of healthy controls (149-+52 cells/mm2; Kruskal-Wallis test, df=2, P<0.0001). ~16 T-cells in the n o n lesional mucosa and the mucosa of healthy controls were rare or absent, comprising 4-5% of the CD3 + population (Fig. 1). By contrast, 7/8 T-cells
were more numerous and were observed in all R A U lesions especially within the inflammatory infiltrates and at the perivascular locations (Fig. 2). They represented 19% of CD3 + lymphocytes (Fig. 1). Counts per n l l T l 2 w e r e 4-+4 in controls, 5-+7 in non-lesional mucosa, and 200-+126 in R A U (Kruskal-Wallis test, df=2, P<0.0001). Lymphocyte populations in the epithelium
For C D 3 + cells in R A U lesions, nonlesional mucosa, and healthy controls the counts were 210+80, 80-+30 and 73-+31 per m m 2 epithelium, respectively. T-cells bearing alfl T C R and 7/& T C R were scattered singly between
60
~P+++m 50
ca I,,,-. +
CHM NLM RAU
basal and supra-basal epithelial cells of R A U lesions and control specimens. A n analysis of intraepithelial cell counts demonstrated increases in the numbers of afl-TCR-positive lymphocytes in R A U lesions (118-+33 cells/mm ~) compared with the non-lesional mucosa of patients with R A U (72-+24) and the mucosa of healthy controls (63-+27 cells/mm~; Kruskal-Wallis test, df=2, P<0.001). 7/c~+ T-cells were seen in 5 of the 10 non-lesional mucosa, and in 4 of the 10 control samples (Table 2), and represented 6% and 2% of the C D 3 + lymphocytes in non-lesional and healthy controls epithelium, respectively (Fig. 1). In R A U lesions, the increase of 7/c~ T-cells tended to be focal (affecting some epithelial ridges more than others). The high density of 7/c~ T-cells was mostly seen at the lower epithelial layers and basement m e m b r a n e zone (Fig. 3). A wide individual variation of ~/6 T-cell count was observed in the epithelium of the R A U lesions (Table 2)
40
•
30
" ~,-~"
. 4~
~
~¢~
~'~+a.
~+~,,
.......
~a 20
++,.++ ~+m+ ",,,;Lp+++ ~ V '"i +~'~" ++:+,+ \"
+++ +. 5
10
+'++
+
m , +' . ~ ++ • , ++p+.~:;-+~ -,.+++>: ++++, .
• ~:'~ I+,++: ,~~'++~' ++ ¢,,~
++,,+
+++ ,++~~ + .
++,++:
Fig. 2. T-cells with y/6 TCR in the subepith-
Epithelium
Connective tissue
Fig. l. Clinically healthy mucosa, non-lesional mucosa, and RAU: 716 T-cell counts expressed as a proportion of CD3+ T-cells.
elial inflammatory infiltrate of an RAU lesion. Note the increased density of ~/6 Tcells at the perivascular locations (V=blood vessel). Original magnification ×250.
378
Natah et aL
Table 2. Density of CD3, a/fl, and 7/~ T-cell lymphocytes in the epithelium of clinically healthy oral mucosa (CHM), non-lesional mucosa (NLM), and RAU lesions a/fl+ T-cells
CD3+ T-cells Subject no. 1 2 3 4 5 6 7 8 9 10 Mean-+SD K-W test
CHM 73-+30 65--+21 112-+29 75-+55 36-+31 22-+16 65+-13 128-+61 80-+40 70-+35 73+31 -
NLM
RAU
CHM
2 6 - + 2 0 221-+44 94_+22 325-+74 59+17 63-+22 9 7 - + 2 5 191-+60 6 4 - + 3 3 286-+65 7 0 - + 4 0 208-+99 57+10 245-+85 119+_68 1 4 7 - + 4 7 110-+62 282-+42 108---24 130-+26 80_+30 210-+80 NS <0.00l*
NLM
73+24 25-+10 80-+32 97-+26 42+30 60-+12 87-+66 77-+41 36-+30 58±30 69-+54 54-+21 35+28 63-+12 113-+50 92-+36 33-+29 87-+65 62-+27 103-+34 63-+27 72+24 NS
7/3+ T-cells RAU 104-+50 195-+26 111+37 106±51 104-+21 125-+69 111-+22 119-+49 139--_50 63_+20 118-+33 <0.001"
CHM
NLM
RAU
03-+06 0 5 - + 0 8 35_+18 03-+06 0 128-+38 0 0 40-+15 03-+05 0 32-+21 0 01-+02 104---21 0 0 68-+56 03-+06 19-+08 111-+31 0 0 67-+55 0 02+03 49-+30 0 01-+02 63_+20 1.2-+1.5 2.8_+06 70_+34 NS <0.001"
Age of RAU (days) 5 2 2 7 3 34 6 4 3 3
NS=non significant vs clinically healthy oral mucosa. * Significant vs non-lesional mucosa or vs clinically healthy oral mucosa. K-W test=Kruskal-Wallis rank test.
with a mean percentage of 36% of intraepithelial CD3+ population (Fig. 1). A significant increase in the numbers of 7/ O-TCR-positive cells in RAU lesions was observed (Figs. 14). The count of 716 T-cells was much higher in the epi-
thelium of RAU (70-+34 cells/mm2) compared with the epithelium of nonlesional mucosa from patients with RAU (2.8-+06 cells/mm2; KruskalWallis test, df=2, P<0.0001), and the epithelium of normal controls (1.2-+ 1.5 cells/ram2; Kruskal-Wallis test, df=2, P<0.0001). No significant increases in 7/fi T-cells were found in non-lesional mucosa when compared with healthy controls. A weak and non-significant negative correlation was found between the densities of intraepithelial 7lfi T-cells and the duration of RAU lesions (r= -0.25).
Discussion
Fig. 3. Numerous t 6 T-cells were demonstrated in the epithelium and subepithelial compartments of a RAU lesion. The increased number of intraepithelial y/6 Tlymphocytes is more evident in the lower epithelial layers (small arrows). Original magnification ×250.
Fig. 4. Only a few fl3-TCR-positive cells are seen in the non-lesional mucosa from the countersite of RAU. Original magnification ×160.
In this discussion (unless specified), we used the term controls to denote both non-lesional mucosa and clinically healthy oral mucosa. Our data showed that the epithelium of RAU lesions contained three times as many CD3+ lymphocytes compared with the epithelium of controls. These findings seem to agree with most former pathological studies that have found frequent infiltration of lymphocytes to the lower epithelial layers in RAU lesions x7. In the present study, most of the CD3 lymphocytes in all samples expressed the a/fl TCR. Although we found that the absolute count of TCR-afl+ was increased in RAU lesions, their proportion to CD3 T cells was reduced compared with the mucosa of controls, apparently because of the relative increase of 716 T-cells. Such a relative reduction of the fraction of TCR-afl+ cells was also found in the peripheral
blood of patients with active RAU compared with patients in the inactive disease stage and the healthy controls22. Lymphocytes expressing t 6 TCRs were found in a small proportion ( - 2 % of CD3+ T-cells) in the healthy control mucosa, a value similar to that previously reported 23. By contrast to the results obtained in mice 13, 7/6 T-cells were uncommon in the epithelium and in the connective tissue of normal human oral mucosa. Our results demonstrated that the high density of 7/6 T-cells was restricted to RAU inflammatory lesions. In nonlesional mucosa from opposite sites of RAU, 7/6 T-cells were rarely seen in the epithelium or in the connective tissue areas. Although we were unable to determine in the present retrospective study if the increased percentage of f16 cells in RAU lesions simply reflected the numbers in the peripheral blood, previous work showed an increased proportion of 7/0 T-cells in the peripheral blood of patients with RAU during both inactive and active aphthous ulceration22. Furthermore, SUZUKI et al. 27 have shown that the concentration of circulating fl 6+ T-cells was higher in patients suffering from Behqet's disease with mucocutaneous RAU compared to patients without mucocutaneous lesions. The requirements for 7/6 T-cell activation are less stringent than those for a/fl T-cell activation. For instance, 7/0 Tcells can recognize antigens directly, without the need for specialized antigen-presenting cells 26. Furthermore, 7/ O T-cells can also react to non-peptidic antigens such as a lipid fraction of M.
713 T cells in recurrent aphthous ulcers tuberculosis 28, and various phosphorylated metabolites 24. Such characteristic features of 7/3 T-cells would probably enable them to respond quickly to a variety of antigens before the a/fl lymphocyte population begins to expand 4. The bowel affected by coeliac disease contains higher than n o r m a l numbers of 7/6 T-cells in either active or "silent" periods of the disease (when the pathology is mildest). Furthermore, the density of such cells is higher in the epithelium than in the lamina propria 16. Such expansion of T C R 7/3+ intraepithelial lymphocytes might be driven by an intestinal allergy to cereal proteins associated with autoimmunity directed against tissue transglutaminase. In aphthae, the high density of 713 T-cells was present exclusively in R A U lesions and in both the epithelium and the subepithelial connective tissue. Thus, it is possible that some of 713 T-cells were responding in an antigen-specific fashion and such antigenic stimuli for 713 T-cells were expressed only during the development/progression of RAU. The exact biological function of 7/3 T-cells in the epithelium of R A U lesions remains unclear. However, BOISMENU& HAVRAN (1994) have shown that activated 7/d T-cells produce the epithelial cell-specific fibroblast growth factor (FGF)-71. The effects of F G F - 7 on the proliferation and differentiation of epithelial cells may help to reduce mucosal damage following tissue injury 29. This is mostly in line with our recent study 2° on expression of the cell proliferationassociated nuclear antigen (Ki-67) in R A U lesions that showed higher proliferation activity in the epithelium of R A U compared with that in the epithelium of controls. M a n y reports have shown that different populations of )43 T-cells have different functions and different patterns of cytokine secretion. For instance, FERRICK et al. 6 found that 7/d T-cells can produce either T-helper type-1 ( T h l ) or Th2 cytokines depending on the pathogen they are exposed to. However, the role of 7/3 T-cells may not be limited to merely enhancing epithelial proliferation or modulating i m m u n e responses of alfl T cells 7. For instance, other investigators 9,11 have suggested that fld T-cells may play a role in immunological damage of aphthae. The "Janus-like" coexistence of cytotoxic and protective potential of fld T-cells might enable them to play a dual func-
tion in the pathogenesis of RAU, thereby acting as immunoregulator to maintain the homeostasis and functional integrity of the epithelial barrier. It should be pointed out that there is a wide variation in the density of 7/3 Tcells between R A U samples in this study. Similar wide individual variations were also noted in the peripheral blood of Behget's disease 27 and R A U patients 22. The reasons for these variations are currently unknown. Although our results provide no answers regarding the function of ),/d Tcells in R A U lesions, they showed that 7/3 T-cells are increased in R A U lesions and such an increase was concomitant with the R A U inflammatory process and did n o t extend to mucosal areas other than the inflammation sites.
Acknowledgements. We are grateful to Mrs. Sirkku Kristiansen and Mrs. Marjatta Kivek~is for skillful technical assistance. This research was supported by the Finnish Academy, and Helsinki University Central Hospital.
References 1. BOISMENUR, HAVRANWL. Modulation of epithelial cell growth by intraepithelial 73 t cells. Science 1994: 266:1253 5. 2. Bos JD, TEUNISSENMBM, CAIROI, et al. T-cell receptor gamma delta bearing cells in normal human skin. J Invest Dermatol 1990: 94: 37-42. 3. BUKOWSKIJF, MORITACT, BRENNERMB. Recognition and destruction of virus-infected cells by human 73 CTL. J Immunol 1994: 153: 5133-40. 4. DE LIBEROG. Sentinel function of broadly reactive human ?,d T cells. Immunol Today 1997: 18: 22-6. 5. FALINIB, FLENGHIL, PILERIS, et al. Distribution of T cells bearing different forms of the T cell receptor 7/d in normal and pathological human tissues. J Immunol 1989: 143: 2480-8. 6. FERRICKDA, SCHRENZELME), MULVANIA T, HSIEH B, FERLIN WG, LEPPER H. Differential production of interferon-7 and interleukin-4 in response to Thl- and Th2- stimulating pathogens by 7/3 T cells in vivo. Nature 1995: 373: 255-7. 7. FUJIHASI-IIK, McGHEE JR, KWEONMN, et al. 7/3 T cell-deficient mice have impaired mucosal immunoglobulin A responses. J Exp Med 1996: 183: 1929-35. 8. HALSTENSENTS, SCOTTH, BRANDTZAEG P. Intraepithelial T cells of the TCR 7/ 3+ CD8- and V31/J31 + phenotypes are increased in coeliac disease. Scand J Immunol 1989: 30: 665-72. 9. HAMZAOUIK, HAMZAOOIA, HENTATIF, et al. Phenotype and functional profile of T cells expressing 73 receptor from pa-
379
tients with active Behqet's disease. J Rheumatol 1994: 21: 2301-6. 10. HAREGEWOIN A, SOMAN G, HOM RC, FINBERG RW Human 7fi+ T cells respond to mycobacterial heat-shock protein. Nature 1989: 340: 309-12. 11. HASANA, FORTUNEF, WILSONA, et al. Role of 7/6 T cells in pathogenesis and diagnosis of Behqet's disease. Lancet 1996: 347: 789-94. 12. H)~YRINEN-IMMONEN R. Immune-activation in recurrent oral ulcers ROU. Scand J Dent Res 1992: 100: 222-7. 13. ITOHARAS, FARRAG, LAFAILLEJJ, et al. Homing of a gamma delta thymocyte subset with homogenous T-cell receptors to mucosal epithelia. Nature 1990: 343: 754-7. 14. JACOBS MR, HAYNES BE Increase in TCR 73 T lymphocytes in synovia from rheumatoid arthritis patients with active synovitis. J Clin Immunol 1992: 12: 1308. 15. JANIS EM, KAUFMANNSHE, SCHWARTZ RH, PARDOLL DM. Activation of 73 T cells in the primary immune response to Mycobacterium tuberculosis. Science 1989: 244: 713-6. 16. KUTLU T, BROUSSEN, RAMBAUDC, LE DEIST F, SCHMITZJ, CERF-BENSUSSANN. Numbers of T cell receptor (TCR) aft+ but not of TcR ~3+ intraepithelial lymphocytes correlate with the grade of villous atrophy in coeliac patients on a long term normal diet. Gut 1993: 34: 208-14. 17. LEHNERT. Pathology of recurrent oral ulceration and oral ulceration in Beh~et's syndrome: light, electron and fluorescence microscopy. J Pathol 1969: 97: 481-94. 18. LEHNERT, LAVERYE, SMITHR, VAN DER ZEE R, MIZUSHIMAY, SHINNICI(T Association between the 65-kilodalton heat shock protein, streptococcus sanguis, and the corresponding antibodies in Beh~et's syndrome. Infect Immun 1991: 59: 143441. 19. MUNK ME, GATRILL AJ, KAUFMANN SHE. Target cell lysis and IL-2 secretion by 73 T lymphocytes after activation with bacteria. J Immunol 1990: 145: 2434-9. 20. NATAH SS, HIETANEN J, H.;~YRINEN-IMMONEN R, JUNGELL P, MALMSTROMM, KONTTINENYT. Expression of cell proliferation-associated nuclear antigen (Ki67) in recurrent aphthous ulcers (RAU). Oral Med Pathol 1998: 3: 29-34. 21. NORAZMIM-N, ARIFIN H, JAMARUDDIN MA. Increased level of activated 73 lymphocytes correlates with disease severity in HIV infection. Immunol Cell Biol 1995: 73:245 8. 22. PEDERSENA, RYDERLE 73 T-cell fraction of peripheral blood is increased in recurrent aphthous ulceration. Clin Immunol Immunopathol 1994: 72: 98-104. 23. PEPIN LF, ROGERT, MORISSETJ, SEMAN M. Preferential V31 expression among TcR fl6-bearing T cells in human oral
380
N a t a h et al.
epithelium. Scand J Immunol 1993: 37: 28%94. 24. PORCELLI SA, MORITA CT, MODLIN RL. T-cell recognition of non-peptide antigens. Curr Opin Immunol 1996: 8: 5106. 25. SAVILAHTI E, REUNALA T, M•KI M. Increase of lymphocytes bearing the f13 T cell receptor in the jejunum of patients with dermatitis herpetiformis. Gut 1992: 33:206-11. 26. SCHILD H, MAVADDATN, LITZENBERGER C, et al. The nature of major histocompatibility complex recognition by 73 T cells. Cell 1994: 76: 29-37. 27. SuzuKI Y, HOSHI K, MATSUDA T, MI-
ZUSI-IIMA Y. Increased peripheral blood ~c~+ T cells and natural killer cells in Behget's disease. J Rheumatol 1992: 19: 588-92. 28. TSUYUGUCHI I, KAWASUMI H, UETA C, YANO I, KISHIMOTO S. Increase of T-cell receptor gamma/delta-bearing T cells in cord blood of newborn babies obtained by in vitro stimulation with mycobacterial cord factor. Infect Immun 1991: 59: 3053-9. 29. ZEEHJM, PROCACCINOF, HOFFMANNP, et al. Keratinocyte growth factor ameliorates mucosal injury in an experimental model of colitis in rats. Gastroenterology 1996: 110: 1077--83.
Address: Dr Sirajedin S. Natah Inflammation Research Group ( Tules Ryhmgi) Institute o f Biomedicine P.O. Box 9 Siltavuorenpenger 20A FIN-O0014 University o f Helsinki Helsinki Finland Tel." +35891918576 Fax: +358 91918499