Expression of inhibitor of apoptosis proteins in B-cell non-Hodgkin and Hodgkin lymphomas

Abstract

BACKGROUND.

Inhibitor of apoptosis proteins (IAPs) inhibit apoptosis by binding specific caspases, and possibly by other mechanisms. Eight IAPs have been identified in humans, of which cIAP1, cIAP2, and XIAP are well known. IAPs are being investigated as potential treatment targets in cancer patients.

METHODS.

cIAP1, cIAP2, and XIAP were assessed in lymphoma cell lines, 240 B-cell non-Hodgkin lymphoma (NHL) tumors, and 40 Hodgkin lymphoma (HL) tumors.

RESULTS.

All IAPs were expressed in most NHL and all HL cell lines. In NHL tumors, cIAP1 was expressed in 174 (73%), cIAP2 in 115 (48%), and XIAP in 37 (15%). cIAP1 was positive in all precursor B-cell lymphoblastic lymphoma/leukemia (LBL) and nodal marginal zone B-cell lymphoma (MZL), over 90% of follicular lymphoma and diffuse large B-cell lymphoma (DLBCL), and approximately 50% to 60% of myeloma, Burkitt lymphoma (BL), lymphoplasmacytic lymphoma/Waldenstrom macroglobulinemia (LPL/WM), small lymphocytic lymphoma/ chronic lymphocytic leukemia (SLL/CLL), extranodal marginal zone B-cell lymphoma of mucosa associated lymphoid tissue (MALT-lymphoma), splenic MZL, and mantle cell lymphoma. cIAP2 was positive in all MALT-lymphoma, over 90% of precursor B-cell LBL (94%), most BL (75%), LPL/WM (71%), and SLL/CLL (67%), and approximately 40% to 60% of follicular lymphoma, myeloma, and DLBCL. XIAP was positive most cases of precursor B-cell LBL (57%) and approximately 30% to 40% of nodal MZL, BL, and DLBCL. In HL tumors, cIAP1 was positive in 30 (75%), cIAP2 in 27 (68%), and XIAP in 23 (58%), and did not correlate with histologic type.

CONCLUSIONS.

Differential expression of IAPs in B-cell lymphomas suggests differences in pathogenesis that may have implications for novel treatment strategies targeting IAPs. Cancer 2006. © 2006 American Cancer Society.

Apoptosis (programmed cell death) is a tightly regulated and selective physiologic process that governs the removal of supernumerary or defective cells. There are 2 major pathways of apoptosis, known as the intrinsic and extrinsic pathways, that allow for induction of apoptosis by diverse stimuli, including chemotherapy.1 Despite a variety of apoptosis-initiating events, apoptosis signaling pathways converge on a common pathway mediated by caspases, a family of intracellular cysteine proteases that cleave substrates at an aspartic acid residue. Caspases are regulated by a number of pro- and antiapoptotic proteins.2

Increased cell resistance to apoptotic stimuli appears to be involved in the pathogenesis of many hematologic neoplasms.3 Members of the inhibitor of apoptosis protein (IAP) family are known to inhibit apoptosis.4, 5 IAP proteins are evolutionarily conserved and are grouped together on the basis of the presence of 1 to 3 baculovirus IAP repeat (BIR) domains. IAP proteins function by binding to specific caspases, directly inhibiting their function, and serving as endogenous antagonists.6 Growing evidence indicates that IAPs also may inhibit apoptosis by inhibiting other proteins, such as second mitochondrial activator of caspases/direct IAP binding protein with low pI (SMAC/DIABLO),7 and have additional roles in cell division, cell cycle progression, and signal transduction pathways.8–10 At least 3 classes of IAPs composed of a total of 8 proteins have been identified in humans, including cIAP1, cIAP2, XIAP, NAIP, livin, ILP2, BRKCK, and survivin. cIAP1, cIAP2, XIAP (all Class 1) and survivin (Class 3) are the best-known members of the IAP family in humans.11

In malignant lymphomas, cIAP2 and survivin have received the most attention in the literature. The cIAP2 gene, also known as API2, is involved in the t(11;18)(q21;q21) detected in 20% to 30% of cases of extranodal marginal zone B-cell lymphoma of mucosa associated lymphoid tissue (MALT-lymphoma), most often in neoplasms arising in the stomach and lung.12 Survivin has been assessed in a wide variety of malignant neoplasms, including malignant lymphomas, and its expression is associated with a poorer prognosis.13, 14

Other than survivin, little is known about the expression of IAPs in human lymphomas and lymphoma cell lines. Nevertheless, in many other types of cancer IAPs are being employed as potential therapeutic targets using either antisense oligonucleotides or small molecule inhibitors.15, 16 In anticipation that these strategies may be considered in patients with lymphoma, in this study we assessed the expression of 3 of the better-known IAP proteins, cIAP1, cIAP2, and XIAP, in a wide variety of cases of B-cell non-Hodgkin lymphoma (NHL) and Hodgkin lymphoma (HL).

MATERIALS AND METHODS

Study Group

The study group included 240 cases of B-cell NHL and 40 cases of HL accessed in the Department of Hematopathology at The University of Texas M. D. Anderson Cancer Center. The diagnosis of all tumors was based on morphologic, immunophenotypic, and molecular criteria as specified in the World Health Organization Classification of lymphoid neoplasms.17 The group of B-cell NHL included 53 cases of plasma cell myeloma, 38 diffuse large B-cell lymphoma (DLBCL), 38 mantle cell lymphoma, 30 follicular lymphoma, 28 precursor B-cell lymphoblastic lymphoma/leukemia (LBL), 12 Burkitt lymphoma, 10 extranodal marginal zone B-cell lymphoma (MZL) (MALT-lymphoma), 9 small lymphocytic lymphoma/chronic lymphocytic leukemia (SLL/CLL), 8 splenic MZL, 7 nodal MZL, and 7 lymphoplasmacytic lymphoma/Waldenstrom macroglobulinemia (LPL/WM). The group of HL included 6 nodular lymphocyte predominant and 34 classical HL (28 nodular sclerosis and 6 mixed cellularity). Six cases of reactive follicular hyperplasia were also studied.

Cell Lines

Ten cell lines derived from B-cell NHL were assessed, including 3 mantle cell lymphoma (Mino, SP53, and JMP-1), 2 DLBCL (DB and DOHH-2), 2 Burkitt lymphoma (Daudi and Raji), 2 precursor B-cell lymphoblastic lymphoma/acute lymphoblastic leukemia (Nalm6 and Reh), and 1 primary effusion lymphoma (BC-1). Four additional cell lines derived from HL (L428, HDLM2, MDA-E, and MDA-V) were also assessed. Cell lines were maintained in RPMI 1640 medium supplemented with 1% nonessential amino acids, 10% fetal calf serum (Life Technologies, Bethesda, MD), and 1% streptomycin-penicillin as previously described.18 Cells were incubated at 37°C in a humidified atmosphere containing 5% carbon dioxide.

Protein Extraction and Western Blot Analysis

Cells in log phase growth were collected, washed twice in cold phosphate-buffered saline (PBS), and lysed at 4°C in lysis buffer as previously described.18 Protein concentrations were determined using the Bio-Rad (Hercules, CA) protein assay. Aliquots of cell lysates (50 μg of protein) were fractionated in 10% to 12% SDS-polyacrylamide gel electrophoresis (PAGE), transferred to nitrocellulose polyvinylidene difluoride (PVDF) membranes (Bio-Rad), and incubated with primary antibodies as follows: cIAP1 (rabbit polyclonal antibody, 73 kDa, 1:1000 dilution, R&D Systems, Minneapolis, MN), cIAP2 (rabbit polyclonal antibody, clone H-85, 66 kDa, 1:500 dilution, Santa Cruz Biotechnology, Santa Cruz, CA), and XIAP (mouse monoclonal antibody, clone 48, 57 kDa, 1:250 dilution, BD Transduction Laboratories, Lexington, KY) overnight at 4°C. The membranes were then incubated with secondary antibody conjugated with horseradish peroxidase (Bio-Rad) and immunodetection was performed with enhanced chemiluminescence reagents (Amersham Pharmacia, Piscataway, NJ). Polyclonal β-actin antibodies (Sigma, St Louis, MO) served as an internal positive control for all cell lines.

Immunohistochemical Methods

Tumors were analyzed using either tissue microarrays (n = 232) or full tissue sections (n = 48). The tissue microarrays included 3 or 4 tumor cores obtained from all tumors. Cell pellets from the lymphoma cell lines were fixed in 4% buffered formalin overnight and embedded in paraffin. These cell blocks were then used to construct a cell line microarray. A manual tissue microarrayer (Beecher Instruments, Silver Springs, MD) was used to construct the microarrays as described previously.19

The immunohistochemical methods have been described previously.20 Formalin-fixed, paraffin-embedded sections were deparaffinized in xylene and rehydrated in a graded series of ethanols. For all antibodies, heat-induced epitope retrieval was performed using a Target Retrieval Solution (DakoCytomation, Carpinteria, CA). The sections were incubated with the same primary antibodies used for Western blot analysis specific for cIAP1 (rabbit polyclonal antibody, 1:50 dilution, R&D Systems), cIAP2 (rabbit polyclonal antibody, clone H-85, 1:50 dilution, Santa Cruz Biotechnology), and XIAP (mouse monoclonal antibody, clone 48, 1:30 dilution, BD Transduction Laboratories, San Jose, CA) overnight at 4°C. Detection of the immunoreaction was performed using the biotin-streptavidin-horseradish peroxidase method (LSAB+ kit, DakoCytomation). We used 3,3′-diaminobenzidine/H₂O₂ (DakoCytomation) as chromogen.

All 3 antibodies showed cytoplasmic staining in positive cases. Positivity was essentially an all-or-none phenomenon. In other words, in positive cases usually all of the neoplastic cells (and at least 50% or more) were positive.

Statistical Analysis

The nonparametric Spearman rank correlation coefficient was applied to evaluate possible correlations between cIAP1, cIAP2, and XIAP. Fisher exact test was used to compare the frequency of expression of cIAP1, cIAP2, and XIAP in various subsets of B-cell NHL and HL. A P-value <.05 was considered statistically significant.

RESULTS

Expression of cIAP1, cIAP2, and XIAP in Non-Hodgkin Lymphoma and Hodgkin Lymphoma Cell Lines

In NHL cell lines, Western blot analysis showed high cIAP1 levels in DB, SP53, JMP-1, BC-1, Raji, Nalm6, and Reh, lesser levels in Mino and Daudi, and no detectable protein in DOHH-1. cIAP2 was detected at high levels in BC-1, Daudi, Nalm 6, and Reh, lesser levels in SP53, DB, and Raji, was very weakly positive in JMO-1 and DOHH-1, and was absent in Mino. XIAP was present at high levels in all NHL cell lines tested (Fig. 1). In 4 HL cell lines, Western blot analysis demonstrated high expression levels of all 3 IAP proteins in all cell lines tested (Fig. 1).

Figure 1. Western blot analysis of non-Hodgkin lymphoma (NHL) and Hodgkin lymphoma (HL) cell lines. The specificity of the cIAP1, cIAP2, and XIAP was confirmed by the presence of appropriately sized bands. As shown, in NHL strong cIAP1 expression was detected in 2 mantle cell lymphoma cell lines (SP53 and JMP-1), 1 primary effusion lymphoma cell line (BC-1), 1 diffuse large B-cell lymphoma cell line (DB), 1 Burkitt lymphoma cell line (Raji), and 2 precursor B-cell lymphoblastic lymphoma/acute lymphoblastic leukemia cell lines (Nalm6 and Reh). Strong cIAP2 was detected in BC-1, Daudi (Burkitt), Nalm 6, and Reh. Strong XIAP expression was observed in all cell lines tested. In HL, all IAPs were strongly expressed in all 4 cell lines tested (L428, HDLM2, MDA-E, and MDA-V).

The specificity of the cIAP1, cIAP2, and XIAP antibodies was confirmed by the presence of appropriately sized bands in cell lines shown by Western blot analysis. The Western blot results correlated well with the results of immunohistochemistry performed on paraffin-embedded blocks of pellets of cell lines using a cell line microarray (Table 1).

Table 1. Expression Profile of cIAP1, cIAP2, and XIAP in B-cell Non-Hodgkin and Hodgkin Lymphoma Cell Lines

Cell line	Lymphoma type	cIAP1 WB/IHC	cIAP2 WB/IHC	XIAP WB/IHC
WB indicates, Western blot analysis; IHC, immunohistochemistry; MCL, mantle cell lymphoma; PEL, primary effusion lymphoma; DLBCL; diffuse large B-cell lymphoma; ND, not done; BL, Burkitt lymphoma; B-LBL/ALL, precursor B-cell lymphoblastic lymphoma/acute lymphoblastic leukemia; HL, Hodgkin lymphoma. −/+, faint; + moderate; ++, high; −, negative.
Mino	MCL	+/++	−/++	++/++
SP53	MCL	++/++	+/+	++/++
JMP-1	MCL	++/+	+/−	++/+
BC-1	PEL	++/+	++/++	++/−
DB	DLBCL	++/ND	+/ND	++/ND
DOHH-1	DLBCL	−/ND	+/ND	++/ND
Daudi	BL	+/+	++/+	++/−
Raji	BL	++/ND	+/ND	++/ND
Nalm6	B-LBL/ALL	++/++	++/++	++/++
Reh	B-LBL/ALL	++/++	++/+	++/+
L428	HL	++/++	++/++	++/++
HD2	HL	++/++	++/++	++/+
MDA-E	HL	++/ND	++/ND	++/ND
MDA-V	HL	++/ND	++/ND	++/ND

Expression of cIAP1, cIAP2, and XIAP in Benign Lymphoid Tissues and B-cell Non-Hodgkin Lymphomas

cIAP1 and cIAP2 expression were detected predominantly as diffuse cytoplasmic staining, whereas expression of XIAP was observed as diffuse and granular cytoplasmic staining. In reactive lymph nodes, cIAP1 and cIAP2 staining was observed in plasma cells and histiocytes; lymphocytes were negative. XIAP staining was observed in germinal center centroblasts and histiocytes.

cIAP1, cIAP2, and XIAP expression in various types of B-cell NHL are summarized in Table 2. cIAP1 was positive in 174 of 240 (73%) cases including all histologic types. All cases of precursor B-cell LBL and nodal MZL were positive as well as over 90% of cases of follicular lymphoma and DLBCL (Fig. 2). Approximately half of other lymphoma types were positive, including plasma cell myeloma, Burkitt lymphoma, LPL/WM, SLL/CLL, splenic MZL, MALT-lymphoma, and mantle cell lymphoma.

Figure 2. Expression of cIAP1, cIAP2, and XIAP in B-cell non-Hodgkin lymphoma tumors. Strong expression of cIAP1 was observed in most cases of (A) diffuse large B-cell lymphoma and (B) follicular lymphoma and all cases of nodal marginal zone B-cell lymphoma (not shown). Strong expression of cIAP2 was detected in a subset of (C) Burkitt lymphoma and (D) MALT lymphoma. Moderate-to-strong expression of XIAP was observed in subsets of (E) precursor B-cell lymphoblastic lymphoma and (F) nodal marginal zone B-cell lymphoma. Immunohistochemistry with hematoxylin counterstain. Original magnification × 500 (A–F).

Table 2. Expression of cIAP1, cIAP2, and XIAP in B-Cell Non-Hodgkin Lymphoma

	No. of cases	cIAP1	cIAP2	XIAP
B-NHL indicates B-cell non-Hodgkin lymphoma; FL, follicular lymphoma; SLL/CLL, small lymphocytic lymphoma/chronic lymphocytic leukemia; LPL/WM, lymphoplasmacytic lymphoma/Waldenstrom macroglobulinemia; NMZL, nodal marginal zone lymphoma; MALT lymphoma, extranodal marginal zone B-cell lymphoma of mucosa-associated lymphoid tissue; SMZL, splenic marginal zone lymphoma; DLBCL; diffuse large B-cell lymphoma; BL, Burkitt lymphoma; B-LBL/ALL, precursor B-cell lymphoblastic lymphoma/acute lymphoblastic leukemia.
Clinically indolent B-NHL
	109	72 (66%)	40 (37%)	5 (6%)
FL	30	29 (97%)	17 (57%)	0
SLL/CLL	9	5 (56%)	6 (67%)	0
LPL/WM	7	4 (57%)	5 (71%)	1 (14%)
NMZL	7	7 (100%)	0	3 (43%)
MALT lymphoma	10	5 (50%)	10 (100%)	0
SMZL	8	4 (50%)	1 (13%)	1 (13%)
Mantle cell lymphoma	38	18 (47%)	1 (3%)	0
Diffuse aggressive B-NHL
	78	69 (89%)	49 (63%)	30 (39%)
DLBCL	38	34 (92%)	14 (37%)	10 (26%)
BL	12	7 (58%)	9 (75%)	4 (33%)
Pre-LBL/ALL	28	28 (100%)	26 (93%)	16 (57%)
Plasma cell myeloma	53	34 (64%)	26 (49%)	2 (4%)

cIAP2 was detected in 115 of 240 (48%) lymphomas, including all cases of MALT-lymphoma (Fig. 2) and most cases of precursor B-cell LBL. Other NHL types in which an appreciable subset of cases were also positive included 9 of 12 (75%) Burkitt lymphoma (Fig. 2), 5 of 7 (71%) LPL/WM, 6 of 9 (67%) CLL/SLL, 17 of 30 (57%) follicular lymphoma, and 14 of 38 (37%) DLBCL. Single cases of mantle cell lymphoma (3%) and splenic MZL (13%) were also positive for cIAP2.

XIAP was detected in 37 of 240 (15%) of B-cell NHL, including 16 of 28 (57%) precursor B-cell LBL (Fig. 2), 3 of 7 (43%) nodal MZL (Fig. 2), 10 of 38 (26%) DLBCL, and single cases of LPL/WM (14%) and splenic MZL (13%).

Expression of cIAP1, cIAP2, and XIAP in Hodgkin Lymphomas

Expression of IAP in 40 cases of HL is summarized in Table 3. cIAP1 was expressed in 30 (75%), cIAP2 in 25 (63%), and XIAP in 23 (58%) cases. Nodular lymphocyte predominant HL was positive for all 3 proteins in 5 of 6 cases. In nodular sclerosis HL, between 50% to 75% of cases were positive for cIAP1 (21/28), cIAP2 (14/28), or XIAP (14/28) (Table 3, Fig. 3). In mixed cellularity HL, between 50% to 67% of cases were positive for cIAP (4/6), cIAP2 (3/6), or XIAP (4/6).

Figure 3. Expression of cIAP1, cIAP2, and XIAP in Hodgkin lymphoma (HL). Strong expression of (A) IAP1 and (B) IAP2 in nodular lymphocyte predominant HL. (C) Moderate staining of XIAP in mixed cellularity HL. (D) Strong and granular staining of XIAP in nodular sclerosis HL. Immunohistochemistry with hematoxylin counterstain. Original magnification × 500 (A–C); × 1000 (D).

Table 3. Expression of cIAP1, cIAP2, and XIAP in Hodgkin Lymphoma

	No. of cases	cIAP1	cIAP2	XIAP
NLPHL indicates, nodular lymphocyte predominant Hodgkin lymphoma; NSHL, nodular sclerosis Hodgkin lymphoma; MCHL, mixed cellularity Hodgkin lymphoma.
Hodgkin lymphoma	40	30 (75%)	25 (63%)	23 (58%)
NLPHL	6	5 (83%)	5 (83%)	5 (83%)
NSHL	28	21 (75%)	17 (60%)	14 (50%)
MCHL	6	4 (67%)	3 (50%)	4 (67%)

Associations Between Expression of cIAP1, cIAP2, and XIAP in B-cell Non-Hodgkin Lymphomas and Hodgkin Lymphomas

There was a significant positive correlation between cIAP1 and cIAP2 (P < .0001), cIAP1 and XIAP (P = .0006), and cIAP2 and XIAP (P = .0014). Diffuse aggressive B-cell NHL and HL had the highest frequency of IAP expression, with the only difference between these groups being a borderline significant (P = .0538) higher frequency of XIAP in HL than in diffuse aggressive B-cell NHL. Diffuse aggressive NHL more often expressed cIAP1 and XIAP than clinically indolent B-cell NHL (P = .0538 and P < .0001, respectively) or plasma cell myeloma (P = .0011 and P < .0001, respectively). Diffuse aggressive B-cell NHL also expressed cIAP1 (P < .0001), cIAP2 (P < .0001), and XIAP (P < .0001) more frequently than mantle cell lymphoma.

Of all tumor types examined, mantle cell lymphomas expressed IAPs least often. Similar to the comparison with diffuse aggressive B-cell NHL, HL more often expressed cIAP1 (P = .0193), cIAP2 (P < .0001), and XIAP (P < .0001) than mantle cell lymphoma. Clinically indolent B-cell NHL more often expressed cIAP1 (P = .0033) and cIAP2 (P < .0001) than mantle cell lymphoma, and plasma cell myeloma were more often positive for cIAP2 than mantle cell lymphoma (P < .0001).

The only other significant comparisons were for XIAP expression, more common in HL than clinically indolent B-cell NHL (P < .0001) or plasma cell myeloma (P < .0001).

In the MZL group of tumors, all nodal MZL were positive for cIAP1 compared with a subset of MALT-lymphoma (P < .05) or splenic MZL (P = .05). cIAP2 was detected in all MALT-lymphoma, compared with a small subset of splenic MZL (P < .0001) and no cases of nodal MZL (P < .0001). XIAP also showed a significant difference in frequency of expression between MALT-lymphoma and nodal MZL (P = .05).

DISCUSSION

Increased resistance to apoptotic stimuli appears to be important in the pathogenesis of hematologic neoplasms.3 IAP family proteins are able to suppress apoptosis through directly binding to and inhibiting specific caspases, and also appear to act by inhibiting SMAC/DIABLO.5–7 Preclinical studies have suggested that IAP family members are potential molecular targets for therapy.15, 16 Either antisense oligonucleotides or small molecular inhibitors directed against XIAP or survivin can induce apoptosis as well as sensitize cells to chemotherapy and γ-irradiation in vitro and are being employed in Phase I clinical trials.15, 21, 22 In anticipation that these approaches may be applied to lymphoma patients, we were interested in assessing for IAP protein expression in B-cell NHL and HL. Although there are 8 known IAP family members, only 4 members are well known: cIAP1, cIAP2, XIAP, and survivin. Of these, survivin has been relatively well studied in lymphomas.13, 14 Thus, in this study we focused on cIAP1, cIAP2, and XIAP.

Our results indicate that IAP family proteins are commonly and often simultaneously expressed in lymphoma cell lines. In 4 HL lymphoma cell lines assessed, we found that cIAP1, cIAP2, and XIAP were expressed at high levels in all cell lines tested. In B-cell NHL cell lines, XIAP was also strongly expressed in all cell lines examined. In contrast, cIAP1 and cIAP2 were more variably expressed in NHL cell lines, but were positive in most cell lines and more often present with stronger intensity in cell lines derived from diffuse aggressive B-cell NHLs.

In B-cell NHL tumors there were some differences with the cell line results. XIAP, strongly positive in all cell lines, was expressed in only 15% of tumors. cIAP1 and cIAP2 generally paralleled the cell line data, positive in 73% and 48% of all lymphomas, respectively. All 3 IAP proteins were expressed frequently in diffuse aggressive B-cell NHL, although XIAP was the least common, in approximately one-third of tumors. In clinically indolent B-cell NHL, only cIAP1 and cIAP2 were relatively common, with XIAP being rarely expressed, in 6% of tumors. The higher frequency of cIAP1 and cIAP2 expression compared with XIAP in B-cell NHL, particularly in indolent B-cell NHL, suggests that cIAP1 and cIAP2 may be more important in pathogenesis or resistance to therapy than XIAP. This differs from the pattern of IAP expression in solid tumors, in which XIAP is frequently expressed at high levels.16, 21, 22

The results show a number of differences in the pattern of expression of IAP proteins in various types of B-cell lymphoma (Table 2). For example, cIAP1 (100%), cIAP2 (93%), and XIAP (57%) were expressed frequently in precursor B-cell LBL. Follicular lymphomas were often positive for cIAP1 (97%) and cIAP2 (57%) but were negative for XIAP. By contrast, mantle cell lymphoma showed the lowest overall expression of IAPs. cIAP1 was positive in almost half of tumors, whereas cIAP2 and XIAP were virtually absent. One other study, which used reverse-transcriptase polymerase chain reaction (RT-PCR) methods to assess for cIAP1 and cIAP2 and other apoptotic molecules, also showed differences of expression in various lymphoma types, although the differences in methods used do not allow easy comparison of their data with our results.23 Nevertheless, both studies support the concept that apoptotic mechanisms are differentially engaged in various lymphoma types and may have implications for predicting potential response rates to therapies that target IAPs.

Although the numbers in the 3 MZL groups are very small, and therefore these results are preliminary, we found differences in the expression pattern of cIAP1, cIAP2, and XIAP among nodal MZL, MALT-lymphoma, and splenic MZL. These data suggest differences in apoptotic mechanisms in the 3 types of MZL that may influence potential therapeutic strategies. Of particular interest are the results for cIAP2, positive in 10 cases of MALT-lymphoma, negative in 7 cases of nodal MZL, and rarely positive in splenic MZL. The t(11;18)(q21;q21) known to be associated with a subset of MALT-lymphomas results in fusion of cIAP2 gene (also known as API2) at 11q21 and the MALT1 gene at 18q21. cIAP2 has been reported to be a ubiquitin ligase of BCL-10 targeting it for degradation, whereas BCL-10 is critically involved in antigen receptor-mediated NF-κB activation.12

The differential pattern of expression of cIAP2 in MZLs, if confirmed, is also of interest because it could potentially be exploited for differential diagnosis. Both nodal MZL and MALT-lymphoma can involve lymph nodes and closely resemble each other histologically. Currently, the distinction is made by convention—the presence of extranodal disease supports the diagnosis of MALT-lymphoma.17 Theoretically, however, nodal MZL can involve all extranodal sites or vice versa, particularly at the time of recurrence. Immunostaining for cIAP2 may be helpful to sort out these issues, as cIAP2 expression suggests the diagnosis of MALT-lymphoma. However, this suggestion needs to be tested in a larger number of cases, including MALT-lymphomas of various anatomic sites and associated with each of the 4 known chromosomal translocations associated with MALT-lymphoma.24, 25

In HL of all types, all 3 IAP proteins are commonly expressed, ranging from 75% for cIAP1 to 58% for XIAP. In particular, expression of XIAP is higher in HL as compared with most types of B-cell NHL. CD30 signaling is known to enhance proliferation of Hodgkin and Reed-Sternberg cells and protect them from cell death, particularly by activation of NF-κB, which can up-regulate XIAP expression.26 Kashkar et al.27 reported that a defect in caspase-3 activation is present in HL-derived B-cell lines, secondary to high-level expression of XIAP. Our observation that cIAP1, cIAP2, and XIAP are consistently overexpressed in the neoplastic cells of HL provides further evidence that resistance to apoptotic stimuli is important for Hodgkin and Reed-Sternberg cell survival.

In conclusion, IAPs are commonly expressed in many types of B-cell NHL and HL, suggesting that apoptotic mechanisms mediated by IAPs are involved in the pathogenesis of lymphomas. The differential expression pattern in various types of B-cell NHL and HL suggests that mechanisms of apoptosis resistance differ in various histologic types of B-cell lymphoma. The heterogeneous patterns of IAP expression B-cell lymphomas may have implications for designing potential therapeutic trials that target particular IAPs.

Acknowledgements

Dr. Nalan Akyurek was supported by a grant from the NATO Science Fellowship Programme by The Scientific and Technical Research Council of Turkey (TUBITAK).