- Email: [email protected]
Contribution of TIA-1+ and granzyme B+ cytotoxic T lymphocytes to cryptal apoptosis and ulceration in active inflammatory bowel disease
ARTICLE IN PRESS
Pathology – Research and Practice 203 (2007) 717–723 www.elsevier.de/prp
ORIGINAL ARTICLE
Contribution of TIA-1+ and granzyme B+ cytotoxic T lymphocytes to cryptal apoptosis and ulceration in active inflammatory bowel disease Hiroyuki Mitomia,, Yasuo Ohkurab, Kaoru Yokoyamac, Miwa Sadac, Kiyonori Kobayashic, Satoshi Tanabec, Naoshi Fukuid, Hideki Kanazawae, Ichiro Kishimotoe, Katsunori Saigenjic a Department of Clinical Research Laboratory (Pathology Division) National Hospital Organization Sagamihara Hospital, 18-1 Sakura-dai, Sagamihara, Kanagawa 228-8522, Japan b Department of Pathology, Kyorin University, School of Medicine, 6-20-2 Shinkawa, Mitaka, Tokyo 181-8611, Japan c Department of Gastroenterology, Kitasato University East Hospital, 2-1-1 Asamizo-dai, Sagamihara, Kanagawa 228-8520, Japan d Department of Pathomechanisms, Clinical Research Center, National Hospital Organization Sagamihara Hospital, 18-1 Sakura-dai, Sagamihara, Kanagawa 228-8522, Japan e Department of Surgery, National Hospital Organization Sagamihara Hospital, 18-1 Sakura-dai, Sagamihara, Kanagawa 228-8522, Japan
Received 10 April 2007; accepted 6 June 2007
Abstract In spite of the clinicopathological differences between Crohn’s disease (CD) and ulcerative colitis (UC), they share the fundamental feature of destructive inflammatory processes involving the intestinal wall. The aim of the present study was to investigate the contribution of cell-mediated cytotoxicity to mucosal damage in CD and UC. Colonic mucosal biopsy specimens from patients with active CD (n ¼ 25) and UC (n ¼ 26) and normal controls (n ¼ 12) were immunohistochemically analyzed for the expression of CD3, CD4, CD8, and T cell-restricted intracellular antigen (TIA)-1, which promotes apoptosis by alternative splicing of pre-messenger RNA of the Fas receptor, and granzyme B (GrB), which leads to apoptosis through induction of perforin. Histological scores for cryptal apoptosis and ulceration were assessed in hematoxylin- and eosin-stained sections. In patients with CD and UC, CD3+(Po0.001), CD4+(Po0.001), CD8+(Po0.01), TIA-1+(CD, Po0.01; UC, Po0.001), and GrB+(CD, Po0.01; UC, Po0.001) intraepithelial lymphocytes (IELs) were significantly increased as compared with controls. Positive relationships were found between the histological scores for apoptosis or ulceration and the numbers of CD8+or TIA-1+IELs. In conclusion, cytotoxic T lymphocytes are present in increased numbers in the mucosa of patients with active CD and UC, and local activation of IELs may contribute to mucosal damage with these diseases. r 2007 Elsevier GmbH. All rights reserved. Keywords: Cytotoxic lymphocytes; Apoptosis; Crohn’s disease; Ulcerative colitis
Introduction Corresponding author. Tel.: +81 42 742 8311;
fax: +81 42 742 5314. E-mail address: [email protected] (H. Mitomi). 0344-0338/$ - see front matter r 2007 Elsevier GmbH. All rights reserved. doi:10.1016/j.prp.2007.06.007
Under normal physiological conditions, T lymphocytes are present in the epithelium (intraepithelial lymphocytes [IELs]) and in the lamina propria
ARTICLE IN PRESS 718
H. Mitomi et al. / Pathology – Research and Practice 203 (2007) 717–723
(lamina propria lymphocytes [LPLs]) of the intestine. In both compartments, they display heterogeneous surface marker profiles, indicating functionally different T-cell subpopulations, with expression of several cytokines suggesting involvement in cell-mediated immune responses [4,18]. It is clear that cytotoxic T lymphocytes can promote apoptosis by secretion of the cytotoxic molecules such as granzymes and perforin [12,15]. Granzyme B (GrB), the neutral serine protease, leads to apoptosis through both caspase cascade-dependent and -independent pathways [12]. Perforin induces membrane damage and provides access for granzyme into target cells (i.e. the perforin-granzyme-mediated pathway) [12]. T cell-restricted intracellular antigen (TIA-1) can stimulate alternative splicing of premessenger RNA for the Fas receptor, resulting in promotion of apoptosis [6]. Crohn’s disease (CD) and ulcerative colitis (UC) constitute the two major idiopathic inflammatory bowel diseases (IBDs). CD may be manifested in all parts of the alimentary tract, but the ileocecal region is most commonly affected; the inflammation is granulomatous and transmural. By contrast, the inflammation in UC is restricted to the large intestine and localized mainly in the mucosa. In spite of these clinicopathological differences between CD and UC, they share the fundamental feature of destructive inflammatory processes involving the intestinal wall. In IBD, support has been provided for the notion of aberrant regulation of local immune responses [14,19,20,22], but limited data are available for the relationship between activated lymphocytes and tissue injury. The aim of the present study was thus to investigate the contribution of TIA-1+ and GrB+ LPLs and IELs to cryptal apoptosis and mucosal damage in cases of CD and UC.
and 3.0 (range 1–13), respectively. All patients underwent colonoscopy, and colonic biopsy specimens were taken from actively inflamed areas. Twelve colonic biopsy samples from 12 patients with colon polyps (9 male and 3 female, mean age 58 years [range 42–65 years]) served as controls. Macroscopically normal tissue was chosen in areas remote from colon polyps. Written informed consent was obtained from all patients.
Histological scores for mucosal damage and apoptotic indices Biopsy specimens were fixed in 10% buffered formalin, routinely processed for embedding in paraffin wax, and sectioned at 4 mm. The sections were stained with hematoxylin and eosin (H&E) and histological scores for mucosal damage were generated by applying a grading scale [9] as follows: (i) crypt destruction: none, 0; local excess of inflammatory cells in parts of crypts, 1; marked crypt attenuation, 2; unequivocal crypt destruction, 3; (ii) ulceration: no erosion, ulceration, or granulation tissue, 0; recovering epithelium+adjacent inflammation, 1; probable erosion-focally stripped, 2; unequivocal erosion, 3; ulceration or granulation tissue, 4. Detection of apoptotic cells in H&E-stained sections was performed under high power view, applying the standard morphological criteria [16]: overall shrinkage and homogenously dark basophilic nuclei, presence of nuclear fragments (apoptotic bodies), sharply delineated cell borders surrounded by empty spaces and homogenous eosinophilic cytoplasm (Fig. 1). Care was taken to differentiate apoptotic cells from polymorphonuclear leukocytes and IELs as detailed in our previous study [21]. We chose not to use the terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate-biotin nick end-labeling method because it detects not only
Materials and methods Patients and materials The study population comprised 25 patients with active CD (6 males and 19 females, mean age 32 years [range 15–60 years]) and 26 patients with active UC (9 males and 17 females, mean age 29 years [range 14–60 years]), their diagnoses based on clinical, endoscopic, radiological, and histological features. In the CD group, the sites of disease were ileocolonic (n ¼ 12) and colonic (n ¼ 13), with a median disease duration of 2.0 years (range 1 month to 21 years). Of the UC patients, 4 had proctitis, 2 had left-sided disease, 17 had extensive disease, and there were 3 other types with a median duration of 1.8 years (range 1 month to 14 years). The median activity index for CD [3] and the median clinical disease score for UC [27] were 188.9 (range, 19.0–371.4)
Fig. 1. An apoptotic body (arrow) in crypt epithelium of a CD case (H&E; original magnification, 100).
ARTICLE IN PRESS H. Mitomi et al. / Pathology – Research and Practice 203 (2007) 717–723
719
apoptotic but also necrotic cells [8]. In addition, our previous study [1] showed a significant, positive correlation between immunohistochemical analysis of singlestranded DNA, which detects early-stage apoptosis [7], and the conventional assessment used here. In the biopsy specimens, all crypts (range, 14–52 crypts/specimen) were evaluated, and apoptotic indices were calculated after counting the numbers of apoptotic cells in epithelium per crypt.
made using the Kruskal-Wallis nonparametric analysis of variance, and if results were significant, the Mann– Whitney U test was applied to compare pairs of groups. Associations of pairs of items were analyzed using the Spearman rank correlation. A P value less than 0.05 was considered to indicate statistical significance. Statview software (Abacus Concepts, Berkeley, CA) was employed for all statistical analyses.
Immunohistochemstry
Results
Sections serial to those for H&E staining were labeled using a streptavidin–biotin immunoperoxidase technique (Histofine SAB-PO [M] kit, Nichirei, Tokyo, Japan). Briefly, after deparaffinization, serial sections, where necessarily heated in 10 mM citrate buffer (pH 6.0) for 15 m using a microwave oven, were incubated overnight at 4 1C with optimal dilutions of the primary antibodies: monoclonal anti-human T cell, CD3 (1:50, DAKO, Glostrup, Denmark), monoclonal anti-human T cell, CD4 (prediluted, Nichirei), monoclonal antihuman T cell, CD8 (1:100, DAKO), monoclonal antihuman TIA-1 (1:400, Coulter, Marseille, France), and monoclonal anti-human GrB (GrB-7, 1:200, Monosan, Sanbio, the Netherlands). Negative controls were prepared by processing the sections in the same manner except for one point: omission of primary antibodies. Hydrogen peroxidase and diaminobenzidine were employed in the colorization step. Slides were faintly counterstained with hematoxylin to facilitate examination. Cells showing intense and distinctive immunoreactivity were considered positive.
Histological scores for mucosal damage and apoptotic indices
Quantification of labeled LPLs At least 500 LPLs were counted in five high-power fields per case, and the number of immunoreactive lymphocytes was expressed as the percentage of all stromal cells counted. Cells morphologically identified as polymorphonuclear leukocytes, endothelial cells, and fibroblasts were not included in the counts.
Quantification of labeled IELs The numbers of immunoreactive IELs were counted in at least 500 surface and crypt cells in the most severely involved areas, and the results were expressed per 100 epithelial cells. All slides were evaluated blind to any clinical information.
Statistical analysis Results are expressed as medians with interquartile ranges (IQRs). Comparisons among three groups were
Microscopically, moderate inflammatory cell infiltration was present in the colonic mucosa from all patients with UC and CD. No inflammatory reactions or crypt alterations were evident in the mucosa of normal controls. The histological scores for crypt destruction (UC: median, 1; IQRs, 0–2; CD: 0; 0–1) and ulceration (UC: 1; 1–2; CD: 0; 0–2) did not significantly differ between UC and CD. The apoptotic indices for UC (median, 0.21/crypt; IQRs, 0.15–0.3/crypt; Po0.001) and CD (0.15; 0.08–0.23; Po0.05) were significantly higher than in the control group (0.08; 0.05–0.12), the difference between the two diseases also being significant (Po0.05).
Subpopulations of LPLs In the normal colon, the number of CD3+LPLs was close to the sum of CD4+ and CD8+cells; the CD4+:CD8+ ratio was approximately 1:1. The percentages of CD8+ and TIA-1+LPLs were similar, and GrB+LPLs were less frequent. In UC, CD3+LPLs were often distributed diffusely in the mucosa, whereas in most CDs, they were heavily infiltrated in the subepithelial portions. Both CD8+ and TIA-1+LPLs were more often seen in the pericryptal zone in UC as compared with CD. The colonic mucosa in both diseases contained significantly greater numbers of CD3+(Po0.001), CD8+(CD, Po0.01; UC, Po0.001), and TIA-1+(CD, Po0.01; UC, Po0.001) LPLs when compared with the control group (Table 1).
Subpopulations of IELs In the normal colon, the number of CD3+IELs was close to that of CD8+cells; CD4+cells were a minority population. The numbers of CD8+ and TIA-1+ IELs were roughly equal, and only a few GrB+IELs were identified. CD3+(Po0.001), CD4+(Po0.001), CD8+ (Po0.01), TIA-1+ (CD, Po0.01; UC, Po0.001), and GrB+ (CD, Po0.01; UC, Po0.001) IELs were
ARTICLE IN PRESS 720
H. Mitomi et al. / Pathology – Research and Practice 203 (2007) 717–723
Subpopulations of LPLs in colonic mucosa
Table 1.
CD UC Normal
CD3
CD4
CD8
TIA-1
GrB
13.8 (11.3–18.0)* 12.5 (9.3–18.0)* 8.4 (5.3–9.1)
4.4 (2.7–7.4) 4.8 (3.1–7.1) 4.6 (2.7–6.3)
8.2 (5.3–10.9)y 10.0 (7.7–13.1)y 5.1 (3.4–6.4)
8.8 (6.0–16.2)z 9.6 (7.6–11.6)z 5.4 (3.9–6.2)
0.8 (0.4–1.5) 0.9 (0.8–1.5) 0.9 (0.6–1.7)
The results are expressed as median (interquartile range) percentages of lamina propria stromal cells.;*Po0.001 CD or UC vs. control; yPo0.01 CD vs. control, Po0.001 UC vs. control; zPo0.01 CD vs. control, Po0.001 UC vs. control; CD, Crohn’s disease; UC, ulcerative colitis.
Subpopulations of IELs in colonic mucosa
Table 2.
CD3 CD UC Normal
5.8(3.8–8.1)* 4.9(4.4–5.8)* 2.8(0.9–4.4)
CD4 2.6(2.0–3.8)y 3.7(3.2–4.2)y 0.2(0.2–0.2)
CD8 4.4(3.7–6.6)z 4.4(3.6–5.4)z 3.3(2.9–3.9)
TIA-1
GrB y
4.8(3.0–5.8) y 4.7(3.6–6.8) 2.3(0.7–3.7)
z
0.2(0.2–0.4) z 0.3(0.2–0.4) 0(0–0.2)
The results are expressed as median (interquartile range) numbers of intraepithelial lymphocytes/100 epithelial cells.;*Po0.001 CD or UC vs. y control, Po0.01 CD vs. UC; yPo0.001 CD or UC vs. control; zPo0.01 CD or UC vs. control; Po0.01 CD vs. control, Po0.001 UC vs. z
control; Po0.01 CD vs. control, Po0.001 UC vs. control; CD, Crohn’s disease; UC, ulcerative colitis. significantly increased in both CD and UC as compared with the control values. Furthermore, CD3+IELs were more prevalent (Po0.01) in CD than in UC (Table 2). Figs. 2 and 3 illustrate the immunoreactivity of LPLs and IELs to TIA-1 and GrB in an IBD patient. The staining pattern for both antibodies was granular and cytoplasmic.
Relationships between numbers of labeled lymphocytes and histological scores for mucosal damage and apoptotic indices In UC, the histological scores for ulceration were positively correlated with the numbers of TIA-1+IELs (r ¼ 0.434; Po0.05; Fig. 4). In CD, positive relationships were found between the apoptotic indices and the numbers of CD3+(r ¼ 0.463; Po0.05) and CD8+IELs (r ¼ 0.412; Po0.05; Fig. 5).
Fig. 2. Immunoreactivity of LPLs and IELs (arrows) for TIA-1 in a CD patient. Cells containing TIA-1 show a brownish, somewhat granular staining of the cytoplasm (immunoperoxidase; original magnification, 100).
Discussion The present study demonstrated: (i) increased frequencies of CD3+, CD4+, CD8+, TIA-1+ and GrB+ LELs and IELs in the inflamed mucosa of biopsy samples from active CD and UC patients, with no significant differences between the two disease entities; (ii) histological scores for ulceration that correlated positively with the numbers of TIA-1+ IELs in UC; (iii) a positive relationship between crypt apoptotic indices and the numbers of CD8+ IELs in CD. Flow cytometric analysis has shown that TIA-1 reacts strongly with CD8+ cells [2], and GrB mRNA is known
to be expressed in CD8+ LPLs [20]. In CD8+ cells, exocytosis of perforin- and granzyme-containing cytoplasmic granules appears to account for the cytotoxic activity observed [12]. TIA-1 is considered a marker of resting as well as of activated T cells possessing cytolytic potential [2], whereas GrB expression is absolutely restricted to activated T lymphocytes [11]. Prominent GrA-expressing T cells were observed earlier in a study of mucosa of resected specimens of patients undergoing surgery for IBD, with no obvious difference between CD and UC [22]. Similarly, in surgically resected CD ileum and UC colon, GrB mRNA was detected in T cells of inflamed intestine [20]. We also found an equal
ARTICLE IN PRESS H. Mitomi et al. / Pathology – Research and Practice 203 (2007) 717–723
721
CD8+IEL count (/100ECs)
12 10 8 6 4 2 0 0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
Apoptotic index in epithelium (/crypt)
Fig. 3. Immunoreactivity of LPLs and IELs (arrows) for GrB in a CD patient (immunoperoxidase; original magnification, 100).
TIA-1+ IEL count (/100ECs)
9
Fig. 5. Relationship between the CD8+IEL count and the apoptotic index in CD cases. The numbers of CD8 immunoreactive IELs were counted in at least 500 epithelial cells (ECs) in the most severely involved areas, and the results are expressed as numbers of stained lymphocytes per 100 ECs. The apoptotic index was calculated by counting the numbers of crypt cell apoptoses per crypt.
8 7 6 5 4 3 2 1 -0.5
0
0.5
1
1.5
2
2.5
3
3.5
Histological score for ulceration
Fig. 4. Relationship between the histological score for ulceration and the TIA-1+IEL count in UC cases. The histological scores were graded 0–4 as detailed in ‘‘Materials and Methods’’. The numbers of TIA-1 immunoreactive IELs were counted in at least 500 epithelial cells (ECs) in the most severely involved areas, and the results are expressed as numbers of stained lymphocytes per 100 ECs.
increase in GrB+IELs in both diseases. However, in another study, GrB mRNA was elevated in untreated CD, but not in naive UC mucosal biopsies [14]. The discrepancy might be explained by the severity of disease at the time of surgery or biopsy. Another possibility is that the methods (antibodies) used to detect cytotoxic T cells were different. GrB+ cells are known to be strongly concentrated in focal inflammatory lesions in the mucosa of CD, whereas regarding UC, smaller numbers are grouped preferentially around the sites of active crypt inflammation [14]. GrA mRNA-expressing cells are generally
local in the outermost regions of crypts in patients with IBD [22]. In our study, GrB+ and TIA-1+ cells were equally increased in CD and UC, but their localizations were different. In UC, the cells were found preferentially in the pericryptal zone where crypt abscesses occur more frequently. Although the results of our study and previous reports [14,22] do not allow a conclusive explanation, it appears likely that elevated access to antigens from the crypt lumen and/or enhanced antigenpresenting capacities of epithelial cells at these locations may lead to the observed distribution of activated cytotoxic cells within the intestinal mucosa. In UC, increased cryptal apoptosis is caused by Fas–Fas ligand interactions [13,26]. With regard to CD, increased matrix metalloproteinase-1 might be one of the possible mechanisms responsible for increased cryptal apoptosis [5]. Another known inducer or mediator of apoptosis is tumor necrosis factor-a [17], which has already been reported to be highly produced by LPLs in CD [19]. We have obtained evidence that in UC, TIA-1+ IELs might contribute to mucosal damage (ulceration) and that CD8+ IELs are associated with cryptal apoptosis in CD. In celiac disease, TIA-1+ and GrB+ IELs are significantly increased [23,24], and this correlates with the degree of mucosal destruction [24]. In addition, TIA-1, GrB, and perforin are known to be common participants in inflammatory skin diseases in which an apoptotic pathway mediated by cytotoxic granules is active [10,25]. The available findings thus suggest a similar process of cytotoxic lymphocyte recruitment despite different inflammatory conditions.
ARTICLE IN PRESS 722
H. Mitomi et al. / Pathology – Research and Practice 203 (2007) 717–723
In conclusion, cytotoxic T lymphocytes are present in increased numbers in inflamed mucosa from patients with both active CD and UC. Although local activation of IELs is still enigmatic, this appears to be a factor in the destructive inflammatory process of IBD.
Acknowledgments We thank K. Yamashita, M. Yokozawa, and T. Kuba, Kitasato University East Hospital for their expert technical assistance.
References [1] F. Akino, H. Mitomi, T. Nakamura, Y. Ohtani, M. Ichinoe, I. Okayasu, High apoptotic activity and low epithelial cell proliferation with underexpression of p21WAF1/CIP1 and p27Kip1 of mucinous carcinomas of the colorectum: comparison with well-differentiated type, Am. J. Clin. Pathol. 117 (2002) 908–915. [2] P. Anderson, C. Nagler-Anderson, C. O’Brien, H. Levine, S. Watkins, H.S. Slayter, M.-L. Blue, S.F. Schlossman, A monoclonal antibody reactive with a 15-kDa cytoplasmic granule-associated protein defines a subpopulation of CD8+ T lymphocytes, J. Immunol. 144 (1990) 574–582. [3] W.R. Best, J.M. Becktel, J.W. Singleton, F. Kern Jr, Development of a Crohn’s disease activity index: National Cooperative Crohn’s Disease Study, Gastroenterology 70 (1976) 439–444. [4] P. Brandtzaeg, I.N. Farstad, L. Helgeland, Phenotypes of T cells in the gut, in: T.T. MacDonald (Ed.), Mucosal T Cells Chemical Immunology, vol. 71, Karger, Basel, 1998, pp. 1–26. [5] A. Di Sabatino, R. Ciccocioppo, O. Luinetti, L. Ricevuti, R. Morera, M.G. Cifone, E. Solcia, G.R. Corazza, Increased enterocyte apoptosis in inflamed areas of Crohn’s disease, Dis. Colon Rectum 46 (2003) 1498–1507. [6] P. Fo¨rch P, J. Valca´rcel, Molecular mechanisms of gene expression regulation by the apoptosis-promoting protein TIA-1, Apoptosis 6 (2001) 463–468. [7] O.S. Frankfurt, J.A. Robb, E.V. Sugarbaker, L. Villa, Monoclonal antibody to single-stranded DNA is a specific and sensitive cellular marker of apoptosis, Exp. Cell Res. 226 (1996) 387–397. [8] Y. Gavrieli, Y. Sherman, S.A. Ben-Sasson, Identification of programmed cell death in situ via specific labeling of nuclear DNA fragmentation, J. Cell Biol. 119 (1992) 493–501. [9] K. Geboes, R. Riddell, A. O¨st, B. Jensfelt, T. Persson, R.A. Lo¨fberg, reproducible grading scale for histological assessment of inflammation in ulcerative colitis, Gut 47 (2000) 404–409. [10] T. Gross, A. Wagner, S. Ugurel, W. Tilgen, U. Reinhold, Identification of TIA-1+ and granzyme B+ cytotoxic T cells in lichen sclerosus et atrophicus, Dermatology 202 (2001) 198–202.
[11] R.D. Hanson, G.M. Sclar, O. Kanagawa, T.J. Ley, The 50 -flanking region of the human CGL-1/granzyme B gene targets expression of a reporter gene to activated T-lymphocytes in transgenic mice, J. Biol. Chem. 266 (1991) 24433–24438. [12] P.A. Henkart, Lymphocyte-mediated cytotoxicity: two pathways and multiple effector molecules, Immunity 1 (1994) 343–346. [13] M. Iwamoto, T. Koji, K. Makiyama, N. Kobayashi, P.K. Nakane, Apoptosis of crypt epithelial cells in ulcerative colitis, J. Pathol. 180 (1996) 152–159. [14] D. Jenkins, R. Seth, J.A. Kummer, B.B. Scott, C.J. Hawkey, R.A. Robins, Differential levels of granzyme B regulatory cytokines, and apoptosis in Crohn’s disease and ulcerative colitis at first presentation, J. Pathol. 190 (2000) 184–189. [15] C-M. Kam, D. Hudig, J.C. Powers, Granzymes (lymphocyte serine proteases): characterization with natural and synthetic substrates and inhibitors, Biochim. Biophys. Acta. 1477 (2000) 307–323. [16] J.F.R. Kerr, A.H. Wyllie, A.R. Currie, Apoptosis: a basic biological phenomenon with wide-ranging implications in tissue kinetics, Br. J. Cancer 26 (1972) 239–257. [17] S.M. Laster, J.G. Wood, L.R. Gooding, Tumor necrosis factor can induce both apoptic and necrotic forms of cells, J. Immunol. 141 (1998) 2629–2634. [18] C. Lundqvist, S. Melgar, M.M-W. Yeung, S. Hammarstro¨m, M-L. Hammarstro¨m, Intraepithelial lymphocytes in human gut have lytic potential and a cytokine profile that suggest T helper 1 and cytotoxic functions, J. Immunol. 157 (1996) 1926–1934. [19] P. Mariani, A. Bachetoni, M. D’Alessandro, D. Lomanto, P. Mazzocchi, V. Speranza, Effector Th-1 cells with cytotoxic function in the intestinal lamina propria of patients with Crohn’s disease, Dig. Dis. Sci. 45 (2000) 2029–2035. ( O¨berg, A. ( Danielsson, [20] S. Melgar, S. Hammarstro¨m, A. M-L. Hammarstro¨m, Cytolytic capabilities of lamina propria and intraepithelial lymphocytes in normal and chronically inflamed human intestine, Scand. J. Immunol. 60 (2004) 167–177. [21] H. Mitomi, I. Okayasu, M.P. Bronner, H. Kanazawa, Y. Nishiyama, Y. Otani, M. Sada, S. Tanabe, M. Igarashi, T. Katsumata, K. Saigenji, Comparative histologic assessment of proctocolectomy specimens from Japanese and American patients with ulcerative colitis with or without dysplasia, Int. J. Surg. Pathol. 13 (2005) 259–265. [22] S. Mu¨ller, J. Lory, N. Corazza, G.M. Griffiths, K. Z’graggen, L. Mazzucchelli, A. Kappeler, C. Mueller, Activated CD4+ and CD8+ cytotoxic cells are present in increased numbers in the intestinal mucosa from patients with active inflammatory bowel disease, Am. J. Pathol. 152 (1998) 261–268. [23] G. Oberhuber, H. Vogelsang, M. Stolte, S. Muthenthaler, A.J. Kummer, T. Radaszkiewicz, Evidence that intestinal intraepithelial lymphocytes are activated cytotoxic T cells in celiac disease but not in giardiasis, Am. J. Pathol. 148 (1996) 1351–1357. [24] G.J. Russell, C. Nagler-Anderson, P. Anderson, A.K. Bhan, Cytotoxic potential of intraepithelial lymphocytes
ARTICLE IN PRESS H. Mitomi et al. / Pathology – Research and Practice 203 (2007) 717–723
(IELs): presence of TIA-1, the cytolytic granuleassociated protein in human IELs in normal and diseased intestine, Am. J. Pathol. 143 (1993) 350–354. [25] A. Santoro, A. Majorana, E. Bardellini, F. Gentili, S. Festa, P. Sapelli, F. Facchetti, Cytotoxic molecule expression and epithelial cell apoptosis in oral and cutaneous lichen planus, Am. J. Clin. Pathol. 121 (2004) 758–764.
723
[26] J. Stra¨ter, I. Wellisch, S. Riedl, H. Walczak, K. Koretz, A. Tandara, P.H. Krammer, P. Mo¨ller, CD95 (APO-1/ Fas)-mediated apoptosis in colon epithelial cells: a possible role in ulcerative colitis, Gastroenterology 113 (1997) 160–167. [27] I. Talstad, E. Gjone, The disease activity of ulcerative colitis and Crohn’s disease, Scand. J. Gastroenterol. 11 (1976) 403–408.
Pathology – Research and Practice 203 (2007) 717–723 www.elsevier.de/prp
ORIGINAL ARTICLE
Contribution of TIA-1+ and granzyme B+ cytotoxic T lymphocytes to cryptal apoptosis and ulceration in active inflammatory bowel disease Hiroyuki Mitomia,, Yasuo Ohkurab, Kaoru Yokoyamac, Miwa Sadac, Kiyonori Kobayashic, Satoshi Tanabec, Naoshi Fukuid, Hideki Kanazawae, Ichiro Kishimotoe, Katsunori Saigenjic a Department of Clinical Research Laboratory (Pathology Division) National Hospital Organization Sagamihara Hospital, 18-1 Sakura-dai, Sagamihara, Kanagawa 228-8522, Japan b Department of Pathology, Kyorin University, School of Medicine, 6-20-2 Shinkawa, Mitaka, Tokyo 181-8611, Japan c Department of Gastroenterology, Kitasato University East Hospital, 2-1-1 Asamizo-dai, Sagamihara, Kanagawa 228-8520, Japan d Department of Pathomechanisms, Clinical Research Center, National Hospital Organization Sagamihara Hospital, 18-1 Sakura-dai, Sagamihara, Kanagawa 228-8522, Japan e Department of Surgery, National Hospital Organization Sagamihara Hospital, 18-1 Sakura-dai, Sagamihara, Kanagawa 228-8522, Japan
Received 10 April 2007; accepted 6 June 2007
Abstract In spite of the clinicopathological differences between Crohn’s disease (CD) and ulcerative colitis (UC), they share the fundamental feature of destructive inflammatory processes involving the intestinal wall. The aim of the present study was to investigate the contribution of cell-mediated cytotoxicity to mucosal damage in CD and UC. Colonic mucosal biopsy specimens from patients with active CD (n ¼ 25) and UC (n ¼ 26) and normal controls (n ¼ 12) were immunohistochemically analyzed for the expression of CD3, CD4, CD8, and T cell-restricted intracellular antigen (TIA)-1, which promotes apoptosis by alternative splicing of pre-messenger RNA of the Fas receptor, and granzyme B (GrB), which leads to apoptosis through induction of perforin. Histological scores for cryptal apoptosis and ulceration were assessed in hematoxylin- and eosin-stained sections. In patients with CD and UC, CD3+(Po0.001), CD4+(Po0.001), CD8+(Po0.01), TIA-1+(CD, Po0.01; UC, Po0.001), and GrB+(CD, Po0.01; UC, Po0.001) intraepithelial lymphocytes (IELs) were significantly increased as compared with controls. Positive relationships were found between the histological scores for apoptosis or ulceration and the numbers of CD8+or TIA-1+IELs. In conclusion, cytotoxic T lymphocytes are present in increased numbers in the mucosa of patients with active CD and UC, and local activation of IELs may contribute to mucosal damage with these diseases. r 2007 Elsevier GmbH. All rights reserved. Keywords: Cytotoxic lymphocytes; Apoptosis; Crohn’s disease; Ulcerative colitis
Introduction Corresponding author. Tel.: +81 42 742 8311;
fax: +81 42 742 5314. E-mail address: [email protected] (H. Mitomi). 0344-0338/$ - see front matter r 2007 Elsevier GmbH. All rights reserved. doi:10.1016/j.prp.2007.06.007
Under normal physiological conditions, T lymphocytes are present in the epithelium (intraepithelial lymphocytes [IELs]) and in the lamina propria
ARTICLE IN PRESS 718
H. Mitomi et al. / Pathology – Research and Practice 203 (2007) 717–723
(lamina propria lymphocytes [LPLs]) of the intestine. In both compartments, they display heterogeneous surface marker profiles, indicating functionally different T-cell subpopulations, with expression of several cytokines suggesting involvement in cell-mediated immune responses [4,18]. It is clear that cytotoxic T lymphocytes can promote apoptosis by secretion of the cytotoxic molecules such as granzymes and perforin [12,15]. Granzyme B (GrB), the neutral serine protease, leads to apoptosis through both caspase cascade-dependent and -independent pathways [12]. Perforin induces membrane damage and provides access for granzyme into target cells (i.e. the perforin-granzyme-mediated pathway) [12]. T cell-restricted intracellular antigen (TIA-1) can stimulate alternative splicing of premessenger RNA for the Fas receptor, resulting in promotion of apoptosis [6]. Crohn’s disease (CD) and ulcerative colitis (UC) constitute the two major idiopathic inflammatory bowel diseases (IBDs). CD may be manifested in all parts of the alimentary tract, but the ileocecal region is most commonly affected; the inflammation is granulomatous and transmural. By contrast, the inflammation in UC is restricted to the large intestine and localized mainly in the mucosa. In spite of these clinicopathological differences between CD and UC, they share the fundamental feature of destructive inflammatory processes involving the intestinal wall. In IBD, support has been provided for the notion of aberrant regulation of local immune responses [14,19,20,22], but limited data are available for the relationship between activated lymphocytes and tissue injury. The aim of the present study was thus to investigate the contribution of TIA-1+ and GrB+ LPLs and IELs to cryptal apoptosis and mucosal damage in cases of CD and UC.
and 3.0 (range 1–13), respectively. All patients underwent colonoscopy, and colonic biopsy specimens were taken from actively inflamed areas. Twelve colonic biopsy samples from 12 patients with colon polyps (9 male and 3 female, mean age 58 years [range 42–65 years]) served as controls. Macroscopically normal tissue was chosen in areas remote from colon polyps. Written informed consent was obtained from all patients.
Histological scores for mucosal damage and apoptotic indices Biopsy specimens were fixed in 10% buffered formalin, routinely processed for embedding in paraffin wax, and sectioned at 4 mm. The sections were stained with hematoxylin and eosin (H&E) and histological scores for mucosal damage were generated by applying a grading scale [9] as follows: (i) crypt destruction: none, 0; local excess of inflammatory cells in parts of crypts, 1; marked crypt attenuation, 2; unequivocal crypt destruction, 3; (ii) ulceration: no erosion, ulceration, or granulation tissue, 0; recovering epithelium+adjacent inflammation, 1; probable erosion-focally stripped, 2; unequivocal erosion, 3; ulceration or granulation tissue, 4. Detection of apoptotic cells in H&E-stained sections was performed under high power view, applying the standard morphological criteria [16]: overall shrinkage and homogenously dark basophilic nuclei, presence of nuclear fragments (apoptotic bodies), sharply delineated cell borders surrounded by empty spaces and homogenous eosinophilic cytoplasm (Fig. 1). Care was taken to differentiate apoptotic cells from polymorphonuclear leukocytes and IELs as detailed in our previous study [21]. We chose not to use the terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate-biotin nick end-labeling method because it detects not only
Materials and methods Patients and materials The study population comprised 25 patients with active CD (6 males and 19 females, mean age 32 years [range 15–60 years]) and 26 patients with active UC (9 males and 17 females, mean age 29 years [range 14–60 years]), their diagnoses based on clinical, endoscopic, radiological, and histological features. In the CD group, the sites of disease were ileocolonic (n ¼ 12) and colonic (n ¼ 13), with a median disease duration of 2.0 years (range 1 month to 21 years). Of the UC patients, 4 had proctitis, 2 had left-sided disease, 17 had extensive disease, and there were 3 other types with a median duration of 1.8 years (range 1 month to 14 years). The median activity index for CD [3] and the median clinical disease score for UC [27] were 188.9 (range, 19.0–371.4)
Fig. 1. An apoptotic body (arrow) in crypt epithelium of a CD case (H&E; original magnification, 100).
ARTICLE IN PRESS H. Mitomi et al. / Pathology – Research and Practice 203 (2007) 717–723
719
apoptotic but also necrotic cells [8]. In addition, our previous study [1] showed a significant, positive correlation between immunohistochemical analysis of singlestranded DNA, which detects early-stage apoptosis [7], and the conventional assessment used here. In the biopsy specimens, all crypts (range, 14–52 crypts/specimen) were evaluated, and apoptotic indices were calculated after counting the numbers of apoptotic cells in epithelium per crypt.
made using the Kruskal-Wallis nonparametric analysis of variance, and if results were significant, the Mann– Whitney U test was applied to compare pairs of groups. Associations of pairs of items were analyzed using the Spearman rank correlation. A P value less than 0.05 was considered to indicate statistical significance. Statview software (Abacus Concepts, Berkeley, CA) was employed for all statistical analyses.
Immunohistochemstry
Results
Sections serial to those for H&E staining were labeled using a streptavidin–biotin immunoperoxidase technique (Histofine SAB-PO [M] kit, Nichirei, Tokyo, Japan). Briefly, after deparaffinization, serial sections, where necessarily heated in 10 mM citrate buffer (pH 6.0) for 15 m using a microwave oven, were incubated overnight at 4 1C with optimal dilutions of the primary antibodies: monoclonal anti-human T cell, CD3 (1:50, DAKO, Glostrup, Denmark), monoclonal anti-human T cell, CD4 (prediluted, Nichirei), monoclonal antihuman T cell, CD8 (1:100, DAKO), monoclonal antihuman TIA-1 (1:400, Coulter, Marseille, France), and monoclonal anti-human GrB (GrB-7, 1:200, Monosan, Sanbio, the Netherlands). Negative controls were prepared by processing the sections in the same manner except for one point: omission of primary antibodies. Hydrogen peroxidase and diaminobenzidine were employed in the colorization step. Slides were faintly counterstained with hematoxylin to facilitate examination. Cells showing intense and distinctive immunoreactivity were considered positive.
Histological scores for mucosal damage and apoptotic indices
Quantification of labeled LPLs At least 500 LPLs were counted in five high-power fields per case, and the number of immunoreactive lymphocytes was expressed as the percentage of all stromal cells counted. Cells morphologically identified as polymorphonuclear leukocytes, endothelial cells, and fibroblasts were not included in the counts.
Quantification of labeled IELs The numbers of immunoreactive IELs were counted in at least 500 surface and crypt cells in the most severely involved areas, and the results were expressed per 100 epithelial cells. All slides were evaluated blind to any clinical information.
Statistical analysis Results are expressed as medians with interquartile ranges (IQRs). Comparisons among three groups were
Microscopically, moderate inflammatory cell infiltration was present in the colonic mucosa from all patients with UC and CD. No inflammatory reactions or crypt alterations were evident in the mucosa of normal controls. The histological scores for crypt destruction (UC: median, 1; IQRs, 0–2; CD: 0; 0–1) and ulceration (UC: 1; 1–2; CD: 0; 0–2) did not significantly differ between UC and CD. The apoptotic indices for UC (median, 0.21/crypt; IQRs, 0.15–0.3/crypt; Po0.001) and CD (0.15; 0.08–0.23; Po0.05) were significantly higher than in the control group (0.08; 0.05–0.12), the difference between the two diseases also being significant (Po0.05).
Subpopulations of LPLs In the normal colon, the number of CD3+LPLs was close to the sum of CD4+ and CD8+cells; the CD4+:CD8+ ratio was approximately 1:1. The percentages of CD8+ and TIA-1+LPLs were similar, and GrB+LPLs were less frequent. In UC, CD3+LPLs were often distributed diffusely in the mucosa, whereas in most CDs, they were heavily infiltrated in the subepithelial portions. Both CD8+ and TIA-1+LPLs were more often seen in the pericryptal zone in UC as compared with CD. The colonic mucosa in both diseases contained significantly greater numbers of CD3+(Po0.001), CD8+(CD, Po0.01; UC, Po0.001), and TIA-1+(CD, Po0.01; UC, Po0.001) LPLs when compared with the control group (Table 1).
Subpopulations of IELs In the normal colon, the number of CD3+IELs was close to that of CD8+cells; CD4+cells were a minority population. The numbers of CD8+ and TIA-1+ IELs were roughly equal, and only a few GrB+IELs were identified. CD3+(Po0.001), CD4+(Po0.001), CD8+ (Po0.01), TIA-1+ (CD, Po0.01; UC, Po0.001), and GrB+ (CD, Po0.01; UC, Po0.001) IELs were
ARTICLE IN PRESS 720
H. Mitomi et al. / Pathology – Research and Practice 203 (2007) 717–723
Subpopulations of LPLs in colonic mucosa
Table 1.
CD UC Normal
CD3
CD4
CD8
TIA-1
GrB
13.8 (11.3–18.0)* 12.5 (9.3–18.0)* 8.4 (5.3–9.1)
4.4 (2.7–7.4) 4.8 (3.1–7.1) 4.6 (2.7–6.3)
8.2 (5.3–10.9)y 10.0 (7.7–13.1)y 5.1 (3.4–6.4)
8.8 (6.0–16.2)z 9.6 (7.6–11.6)z 5.4 (3.9–6.2)
0.8 (0.4–1.5) 0.9 (0.8–1.5) 0.9 (0.6–1.7)
The results are expressed as median (interquartile range) percentages of lamina propria stromal cells.;*Po0.001 CD or UC vs. control; yPo0.01 CD vs. control, Po0.001 UC vs. control; zPo0.01 CD vs. control, Po0.001 UC vs. control; CD, Crohn’s disease; UC, ulcerative colitis.
Subpopulations of IELs in colonic mucosa
Table 2.
CD3 CD UC Normal
5.8(3.8–8.1)* 4.9(4.4–5.8)* 2.8(0.9–4.4)
CD4 2.6(2.0–3.8)y 3.7(3.2–4.2)y 0.2(0.2–0.2)
CD8 4.4(3.7–6.6)z 4.4(3.6–5.4)z 3.3(2.9–3.9)
TIA-1
GrB y
4.8(3.0–5.8) y 4.7(3.6–6.8) 2.3(0.7–3.7)
z
0.2(0.2–0.4) z 0.3(0.2–0.4) 0(0–0.2)
The results are expressed as median (interquartile range) numbers of intraepithelial lymphocytes/100 epithelial cells.;*Po0.001 CD or UC vs. y control, Po0.01 CD vs. UC; yPo0.001 CD or UC vs. control; zPo0.01 CD or UC vs. control; Po0.01 CD vs. control, Po0.001 UC vs. z
control; Po0.01 CD vs. control, Po0.001 UC vs. control; CD, Crohn’s disease; UC, ulcerative colitis. significantly increased in both CD and UC as compared with the control values. Furthermore, CD3+IELs were more prevalent (Po0.01) in CD than in UC (Table 2). Figs. 2 and 3 illustrate the immunoreactivity of LPLs and IELs to TIA-1 and GrB in an IBD patient. The staining pattern for both antibodies was granular and cytoplasmic.
Relationships between numbers of labeled lymphocytes and histological scores for mucosal damage and apoptotic indices In UC, the histological scores for ulceration were positively correlated with the numbers of TIA-1+IELs (r ¼ 0.434; Po0.05; Fig. 4). In CD, positive relationships were found between the apoptotic indices and the numbers of CD3+(r ¼ 0.463; Po0.05) and CD8+IELs (r ¼ 0.412; Po0.05; Fig. 5).
Fig. 2. Immunoreactivity of LPLs and IELs (arrows) for TIA-1 in a CD patient. Cells containing TIA-1 show a brownish, somewhat granular staining of the cytoplasm (immunoperoxidase; original magnification, 100).
Discussion The present study demonstrated: (i) increased frequencies of CD3+, CD4+, CD8+, TIA-1+ and GrB+ LELs and IELs in the inflamed mucosa of biopsy samples from active CD and UC patients, with no significant differences between the two disease entities; (ii) histological scores for ulceration that correlated positively with the numbers of TIA-1+ IELs in UC; (iii) a positive relationship between crypt apoptotic indices and the numbers of CD8+ IELs in CD. Flow cytometric analysis has shown that TIA-1 reacts strongly with CD8+ cells [2], and GrB mRNA is known
to be expressed in CD8+ LPLs [20]. In CD8+ cells, exocytosis of perforin- and granzyme-containing cytoplasmic granules appears to account for the cytotoxic activity observed [12]. TIA-1 is considered a marker of resting as well as of activated T cells possessing cytolytic potential [2], whereas GrB expression is absolutely restricted to activated T lymphocytes [11]. Prominent GrA-expressing T cells were observed earlier in a study of mucosa of resected specimens of patients undergoing surgery for IBD, with no obvious difference between CD and UC [22]. Similarly, in surgically resected CD ileum and UC colon, GrB mRNA was detected in T cells of inflamed intestine [20]. We also found an equal
ARTICLE IN PRESS H. Mitomi et al. / Pathology – Research and Practice 203 (2007) 717–723
721
CD8+IEL count (/100ECs)
12 10 8 6 4 2 0 0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
Apoptotic index in epithelium (/crypt)
Fig. 3. Immunoreactivity of LPLs and IELs (arrows) for GrB in a CD patient (immunoperoxidase; original magnification, 100).
TIA-1+ IEL count (/100ECs)
9
Fig. 5. Relationship between the CD8+IEL count and the apoptotic index in CD cases. The numbers of CD8 immunoreactive IELs were counted in at least 500 epithelial cells (ECs) in the most severely involved areas, and the results are expressed as numbers of stained lymphocytes per 100 ECs. The apoptotic index was calculated by counting the numbers of crypt cell apoptoses per crypt.
8 7 6 5 4 3 2 1 -0.5
0
0.5
1
1.5
2
2.5
3
3.5
Histological score for ulceration
Fig. 4. Relationship between the histological score for ulceration and the TIA-1+IEL count in UC cases. The histological scores were graded 0–4 as detailed in ‘‘Materials and Methods’’. The numbers of TIA-1 immunoreactive IELs were counted in at least 500 epithelial cells (ECs) in the most severely involved areas, and the results are expressed as numbers of stained lymphocytes per 100 ECs.
increase in GrB+IELs in both diseases. However, in another study, GrB mRNA was elevated in untreated CD, but not in naive UC mucosal biopsies [14]. The discrepancy might be explained by the severity of disease at the time of surgery or biopsy. Another possibility is that the methods (antibodies) used to detect cytotoxic T cells were different. GrB+ cells are known to be strongly concentrated in focal inflammatory lesions in the mucosa of CD, whereas regarding UC, smaller numbers are grouped preferentially around the sites of active crypt inflammation [14]. GrA mRNA-expressing cells are generally
local in the outermost regions of crypts in patients with IBD [22]. In our study, GrB+ and TIA-1+ cells were equally increased in CD and UC, but their localizations were different. In UC, the cells were found preferentially in the pericryptal zone where crypt abscesses occur more frequently. Although the results of our study and previous reports [14,22] do not allow a conclusive explanation, it appears likely that elevated access to antigens from the crypt lumen and/or enhanced antigenpresenting capacities of epithelial cells at these locations may lead to the observed distribution of activated cytotoxic cells within the intestinal mucosa. In UC, increased cryptal apoptosis is caused by Fas–Fas ligand interactions [13,26]. With regard to CD, increased matrix metalloproteinase-1 might be one of the possible mechanisms responsible for increased cryptal apoptosis [5]. Another known inducer or mediator of apoptosis is tumor necrosis factor-a [17], which has already been reported to be highly produced by LPLs in CD [19]. We have obtained evidence that in UC, TIA-1+ IELs might contribute to mucosal damage (ulceration) and that CD8+ IELs are associated with cryptal apoptosis in CD. In celiac disease, TIA-1+ and GrB+ IELs are significantly increased [23,24], and this correlates with the degree of mucosal destruction [24]. In addition, TIA-1, GrB, and perforin are known to be common participants in inflammatory skin diseases in which an apoptotic pathway mediated by cytotoxic granules is active [10,25]. The available findings thus suggest a similar process of cytotoxic lymphocyte recruitment despite different inflammatory conditions.
ARTICLE IN PRESS 722
H. Mitomi et al. / Pathology – Research and Practice 203 (2007) 717–723
In conclusion, cytotoxic T lymphocytes are present in increased numbers in inflamed mucosa from patients with both active CD and UC. Although local activation of IELs is still enigmatic, this appears to be a factor in the destructive inflammatory process of IBD.
Acknowledgments We thank K. Yamashita, M. Yokozawa, and T. Kuba, Kitasato University East Hospital for their expert technical assistance.
References [1] F. Akino, H. Mitomi, T. Nakamura, Y. Ohtani, M. Ichinoe, I. Okayasu, High apoptotic activity and low epithelial cell proliferation with underexpression of p21WAF1/CIP1 and p27Kip1 of mucinous carcinomas of the colorectum: comparison with well-differentiated type, Am. J. Clin. Pathol. 117 (2002) 908–915. [2] P. Anderson, C. Nagler-Anderson, C. O’Brien, H. Levine, S. Watkins, H.S. Slayter, M.-L. Blue, S.F. Schlossman, A monoclonal antibody reactive with a 15-kDa cytoplasmic granule-associated protein defines a subpopulation of CD8+ T lymphocytes, J. Immunol. 144 (1990) 574–582. [3] W.R. Best, J.M. Becktel, J.W. Singleton, F. Kern Jr, Development of a Crohn’s disease activity index: National Cooperative Crohn’s Disease Study, Gastroenterology 70 (1976) 439–444. [4] P. Brandtzaeg, I.N. Farstad, L. Helgeland, Phenotypes of T cells in the gut, in: T.T. MacDonald (Ed.), Mucosal T Cells Chemical Immunology, vol. 71, Karger, Basel, 1998, pp. 1–26. [5] A. Di Sabatino, R. Ciccocioppo, O. Luinetti, L. Ricevuti, R. Morera, M.G. Cifone, E. Solcia, G.R. Corazza, Increased enterocyte apoptosis in inflamed areas of Crohn’s disease, Dis. Colon Rectum 46 (2003) 1498–1507. [6] P. Fo¨rch P, J. Valca´rcel, Molecular mechanisms of gene expression regulation by the apoptosis-promoting protein TIA-1, Apoptosis 6 (2001) 463–468. [7] O.S. Frankfurt, J.A. Robb, E.V. Sugarbaker, L. Villa, Monoclonal antibody to single-stranded DNA is a specific and sensitive cellular marker of apoptosis, Exp. Cell Res. 226 (1996) 387–397. [8] Y. Gavrieli, Y. Sherman, S.A. Ben-Sasson, Identification of programmed cell death in situ via specific labeling of nuclear DNA fragmentation, J. Cell Biol. 119 (1992) 493–501. [9] K. Geboes, R. Riddell, A. O¨st, B. Jensfelt, T. Persson, R.A. Lo¨fberg, reproducible grading scale for histological assessment of inflammation in ulcerative colitis, Gut 47 (2000) 404–409. [10] T. Gross, A. Wagner, S. Ugurel, W. Tilgen, U. Reinhold, Identification of TIA-1+ and granzyme B+ cytotoxic T cells in lichen sclerosus et atrophicus, Dermatology 202 (2001) 198–202.
[11] R.D. Hanson, G.M. Sclar, O. Kanagawa, T.J. Ley, The 50 -flanking region of the human CGL-1/granzyme B gene targets expression of a reporter gene to activated T-lymphocytes in transgenic mice, J. Biol. Chem. 266 (1991) 24433–24438. [12] P.A. Henkart, Lymphocyte-mediated cytotoxicity: two pathways and multiple effector molecules, Immunity 1 (1994) 343–346. [13] M. Iwamoto, T. Koji, K. Makiyama, N. Kobayashi, P.K. Nakane, Apoptosis of crypt epithelial cells in ulcerative colitis, J. Pathol. 180 (1996) 152–159. [14] D. Jenkins, R. Seth, J.A. Kummer, B.B. Scott, C.J. Hawkey, R.A. Robins, Differential levels of granzyme B regulatory cytokines, and apoptosis in Crohn’s disease and ulcerative colitis at first presentation, J. Pathol. 190 (2000) 184–189. [15] C-M. Kam, D. Hudig, J.C. Powers, Granzymes (lymphocyte serine proteases): characterization with natural and synthetic substrates and inhibitors, Biochim. Biophys. Acta. 1477 (2000) 307–323. [16] J.F.R. Kerr, A.H. Wyllie, A.R. Currie, Apoptosis: a basic biological phenomenon with wide-ranging implications in tissue kinetics, Br. J. Cancer 26 (1972) 239–257. [17] S.M. Laster, J.G. Wood, L.R. Gooding, Tumor necrosis factor can induce both apoptic and necrotic forms of cells, J. Immunol. 141 (1998) 2629–2634. [18] C. Lundqvist, S. Melgar, M.M-W. Yeung, S. Hammarstro¨m, M-L. Hammarstro¨m, Intraepithelial lymphocytes in human gut have lytic potential and a cytokine profile that suggest T helper 1 and cytotoxic functions, J. Immunol. 157 (1996) 1926–1934. [19] P. Mariani, A. Bachetoni, M. D’Alessandro, D. Lomanto, P. Mazzocchi, V. Speranza, Effector Th-1 cells with cytotoxic function in the intestinal lamina propria of patients with Crohn’s disease, Dig. Dis. Sci. 45 (2000) 2029–2035. ( O¨berg, A. ( Danielsson, [20] S. Melgar, S. Hammarstro¨m, A. M-L. Hammarstro¨m, Cytolytic capabilities of lamina propria and intraepithelial lymphocytes in normal and chronically inflamed human intestine, Scand. J. Immunol. 60 (2004) 167–177. [21] H. Mitomi, I. Okayasu, M.P. Bronner, H. Kanazawa, Y. Nishiyama, Y. Otani, M. Sada, S. Tanabe, M. Igarashi, T. Katsumata, K. Saigenji, Comparative histologic assessment of proctocolectomy specimens from Japanese and American patients with ulcerative colitis with or without dysplasia, Int. J. Surg. Pathol. 13 (2005) 259–265. [22] S. Mu¨ller, J. Lory, N. Corazza, G.M. Griffiths, K. Z’graggen, L. Mazzucchelli, A. Kappeler, C. Mueller, Activated CD4+ and CD8+ cytotoxic cells are present in increased numbers in the intestinal mucosa from patients with active inflammatory bowel disease, Am. J. Pathol. 152 (1998) 261–268. [23] G. Oberhuber, H. Vogelsang, M. Stolte, S. Muthenthaler, A.J. Kummer, T. Radaszkiewicz, Evidence that intestinal intraepithelial lymphocytes are activated cytotoxic T cells in celiac disease but not in giardiasis, Am. J. Pathol. 148 (1996) 1351–1357. [24] G.J. Russell, C. Nagler-Anderson, P. Anderson, A.K. Bhan, Cytotoxic potential of intraepithelial lymphocytes
ARTICLE IN PRESS H. Mitomi et al. / Pathology – Research and Practice 203 (2007) 717–723
(IELs): presence of TIA-1, the cytolytic granuleassociated protein in human IELs in normal and diseased intestine, Am. J. Pathol. 143 (1993) 350–354. [25] A. Santoro, A. Majorana, E. Bardellini, F. Gentili, S. Festa, P. Sapelli, F. Facchetti, Cytotoxic molecule expression and epithelial cell apoptosis in oral and cutaneous lichen planus, Am. J. Clin. Pathol. 121 (2004) 758–764.
723
[26] J. Stra¨ter, I. Wellisch, S. Riedl, H. Walczak, K. Koretz, A. Tandara, P.H. Krammer, P. Mo¨ller, CD95 (APO-1/ Fas)-mediated apoptosis in colon epithelial cells: a possible role in ulcerative colitis, Gastroenterology 113 (1997) 160–167. [27] I. Talstad, E. Gjone, The disease activity of ulcerative colitis and Crohn’s disease, Scand. J. Gastroenterol. 11 (1976) 403–408.