The first two authors contributed equally to this work.
Overexpression of caspase recruitment domain (CARD) membrane-associated guanylate kinase 1 (CARMA1) and CARD9 in primary gastric B-cell lymphoma
Article first published online: 21 SEP 2005
Copyright © 2005 American Cancer Society
Volume 104, Issue 9, pages 1885–1893, 1 November 2005
How to Cite
Nakamura, S., Nakamura, S., Matsumoto, T., Yada, S., Hirahashi, M., Suekane, H., Yao, T., Goda, K. and Iida, M. (2005), Overexpression of caspase recruitment domain (CARD) membrane-associated guanylate kinase 1 (CARMA1) and CARD9 in primary gastric B-cell lymphoma. Cancer, 104: 1885–1893. doi: 10.1002/cncr.21421
- Issue published online: 17 OCT 2005
- Article first published online: 21 SEP 2005
- Manuscript Accepted: 16 MAY 2005
- Manuscript Revised: 29 MAR 2005
- Manuscript Received: 29 NOV 2004
- gastric lymphoma;
- mucosa-associated lymphoid tissue;
- caspase recruitment domain membrane-associated guanylate kinase 1;
- caspase recruitment domain 9;
- t(11;18) (q21;q21);
- Helicobacter pylori
Although caspase recruitment domain (CARD) membrane-associated guanylate kinase (MAGUK) protein 1 (CARMA1) and CARD9 play important roles in lymphocyte activation, the significance of CARMA1 and CARD9 in the pathogenesis of gastric mucosa-associated lymphoid tissue (MALT) lymphoma remains to be elucidated.
By using reverse transcription-polymerase chain reaction analysis, the expression levels of mRNA of CARMA1, CARD9, Bcl10, and the apoptosis inhibitor 2 (API2)-MALT1 chimeric transcript were determined in tissue specimens from 65 patients with primary gastric B-cell lymphoma (43 patients with low-grade MALT lymphoma, 16 patients with MALT lymphoma plus diffuse large B-cell lymphoma [DLBL], and 6 patients with DLBL without MALT lymphoma) and in tissue specimens from 18 patients with chronic gastritis. The expression levels of CARMA1 and BCL10 were examined immunohistochemically in 30 patients with lymphoma.
CARMA1 mRNA was detected in 55% of lymphoma patients but in only 17% of chronic gastritis patients. The positive rates for CARD9, Bcl10, and API2-MALT1 chimeric transcript in the lymphoma patients were 48%, 98%, and 8%, respectively, whereas the 3 molecules were not detected in any specimens from patients with chronic gastritis. The expression of CARMA1 and CARD9 was frequent in the Helicobacter pylori-negative patients (100% and 86%, respectively), in the API2-MALT1 chimeric transcript-positive patients (100% and 100%, respectively), and in the specimens from patients who did not respond to H. pylori eradication (76% and 71%, respectively). In addition, CARMA1 expression was positive more frequently in patients of DLBL without MALT lymphoma (100%) than in patients of MALT lymphoma (51%). CARMA1 protein expression was correlated significantly with the expression of CARMA1 mRNA and also with the expression of nuclear BCL10.
The overexpression of CARMA1 and CARD9 presumably is associated with the development or progression of gastric B-cell lymphoma, especially among patients who have disease in which the pathogenesis is not related to H. pylori. Cancer 2005. © 2005 American Cancer Society.
Extranodal marginal zone B-cell lymphoma of mucosa-associated lymphoid tissue (MALT lymphoma) has been regarded as a distinct clinicopathologic entity among non-Hodgkin lymphomas. It has been shown that Helicobacter pylori plays an important role in the development of gastric MALT lymphoma, and from 60% to 90% of patients with such disease show a complete remission after the eradication of H. pylori.1–4 Recently, chromosomal translocations t(11;18) (q21;q21)5 and t(1;14) (p22;q32)6 were found in a certain proportion of patients with low-grade MALT lymphoma. T(11;18) (q21;q21) results in a fusion of 2 genes, the apoptosis inhibitor-2 (API2) gene at 11q21 and the novel gene at 18q21, which has been referred to as MALT1,7MALT lymphoma-associated translocation (MLT),8 or human paracaspase.9 The API2-MALT1 chimeric transcript is specific to t(11;18) (q21;q21), and most patients who have gastric MALT lymphoma with this translocation do not respond to H. pylori eradication therapy.10–12
Conversely, t(1;14) (p22;q32) involves an apoptotic signaling gene Bcl10, which encodes a 32-kDa protein containing an amino-terminal caspase recruitment domain (CARD).6, 13, 14 The CARD membrane-associated guanylate kinase (MAGUK) protein 1 (CARMA1), also known as CARD11 or Bcl10-interacting MAGUK protein 3 (Bimp3), originally was identified as a 130-kDa CARD-containing protein of the MAGUK family.15–17 It has been suggested that CARMA1, Bcl10, and MALT1 all possess signal transductions in the antigen-receptor-mediated activation of T and B lymphocytes through the nuclear factor-κB (NF-κB) pathway.18–22 Another CARD-containing protein, CARD9, also binds to Bcl10 and activates NF-κB.23, 24 However, the significance of CARMA1 and CARD9 in the pathogenesis of gastric MALT lymphoma remains to be elucidated. In the current study, we analyzed the mRNA expression of CARMA1, CARD9, Bcl10, and API2-MALT1 chimeric transcript in 65 patients with gastric B-cell lymphoma using reverse-transcription polymerase chain reaction (RT-PCR) analysis, and we investigated any correlation between such expression and clinicopathologic features.
MATERIALS AND METHODS
Sixty-five patients with primary gastric B-cell lymphoma who were diagnosed at our institutions between October 1993 and January 2004 and from whom fresh or frozen tissue samples were available were enrolled in the current study. The patients included 34 men and 31 women, and they ranged in age from 16–84 years (mean age, 61.8 yrs). The histologic classification of lymphoma was done according to the World Health Organization criteria. Forty-three patients (66%) were diagnosed with low-grade MALT lymphoma, 16 patients (25%) were diagnosed with diffuse large B-cell lymphoma (DLBL) plus MALT lymphoma, and 6 patients (9%) were diagnosed with DLBL without MALT lymphoma. B-cell monoclonality was confirmed in all patients using PCR for detection of the immunoglobulin heavy-chain gene rearrangement.25 The depth of tumor invasion was determined by endoscopic ultrasonography (EUS) as follows: Eighteen lymphomas were restricted to the mucosa, 26 lymphomas had invaded the submucosa, 10 lymphomas involved the muscularis propria, and 11 lymphomas reached the subserosa or serosa. According to the Lugano International Conference classification for clinical staging of gastrointestinal lymphoma,26 45 tumors (69%) were classified as Stage I, 11 tumors (17%) were classified as Stage II1, 3 tumors (5%) were classified as Stage II2, and 6 tumors (9%) were classified as Stage IV. H. pylori infection was positive in 58 patients and negative in 7 patients, as assessed by the histology, culture, rapid urease test, the 13C urea breath test, and serology findings.
Fifty-one of 65 patients (79%) initially underwent the eradication of H. pylori with a proton pump inhibitor (omeprazole, lansoprazole, or rabeprazole) plus a combination of antibiotics (clarithromycin and amoxicillin with or without metronidazole). Thirty-two of those 51 patients underwent eradication therapy alone, whereas 9 patients received radiation after eradication, 7 patients received chemotherapy after eradication, and 3 patients underwent a surgical resection after eradication. Ten patients (15%) initially underwent surgical resection either with or without chemotherapy, and 4 patients (6%) received chemotherapy alone. We also enrolled 18 patients with H. pylori-associated chronic gastritis, including 11 men and 7 women who ranged in age from 31–83 years (mean age, 58.3 yrs) for the current investigation.
This study was performed in accordance with the Helsinki Declaration as revised in 1989. Informed consent was obtained from each patient regarding the study objectives and protocol.
Total RNA was extracted from frozen tissue specimens, which were obtained either by endoscopic biopsies or from endoscopically/surgically resected samples, using the acid guanidium thiocyanate-phenol-cloroform method. One microgram of total RNA was reverse-transcribed to complementary DNA (cDNA) by reverse transcriptase (Superscript II; Invitrogen, La Jolla, CA) using random primers (Invitrogen) in a total volume of 20 μL containing deoxyribonucleoside-5′-triphosphates (dNTP). One microliter of cDNA solution was processed with PCR for CARMA1, CARD9, Bcl10, and API2-MALT1 chimeric transcript using the primer pairs indicated in Table 1. The primers for CARMA1 and CARD9 were developed based on the sequences listed in GenBank under accession numbers AF322641 and AF311287, respectively. The primers used for the Bcl10 and API2-MALT1 chimeric transcript were the same as those described previously.27, 11 Each reaction mixture (30 μL) was composed of 0.75 units of AmpliTaq (PerkinElmer, Wellesley, MA); 10 mmol/L Tris, pH 8.3, at 25 °C; 1.5 mmol/L MgCl2; 0.2 mmol/L of dNTP (Invitrogen); and 20 pmole of each primer. The PCR reaction was conducted in a thermal cycler (PerkinElmer) under the following conditions: For CARMA1, 5 minutes at 94 °C were followed by 35 cycles of denaturation for 1 minute at 94 °C, annealing for 1 minute at 62 °C, and extension for 1 minute at 72 °C. For CARD9, 5 minutes at 94 °C were followed by 30 cycles of denaturation for 1 minute at 94 °C, annealing for 1 minute at 58 °C, and extension for 1 minute at 72 °C. For Bcl10, 5 minutes at 95 °C were followed by 35 cycles for 30 seconds at 94 °C, for 30 seconds at 50 °C, and extension for 1 minute at 72 °C.28 For API2-MALT1 chimeric transcript, 2 minutes at 94 °C were followed by 35 cycles of denaturation for 20 seconds at 94 °C, annealing for 20 seconds at 63 °C for 20 seconds, and extension for 30 seconds at 72 °C.10 After the reaction, 10 μL of the PCR products were electrophoresed on 1.6% agarose gels and stained with ethidium bromide. Glyceraldehydes-3-phosphate dehydrogenase (GAPDH) was amplified as a positive internal control. The reaction was performed under conditions of 94 °C for 5 minutes followed by 30 cycles of denaturation at 94 °C for 1 minute, annealing at 58 °C for 1 minute, and extension at 72 °C for 1 minute, followed by a final extension at 72 °C for 7 minutes.
|Target and primer||Primer sequence|
|Sense||5′-GAT GGA GGA GGA ATG TAA GC|
|Antisense||5′-GGA TGG TAG GGC AGG AAG TA|
|Sense||5′-AAC GGA AAG TGG GTG TGC TC|
|Antisense||5′-CGC CCT TCT CCT CAC TCT GG|
|Sense||5′-ACC TCA CTG AAG TGA AGA AG|
|Antisense||5′-ATC TGG AGA GGT TGT TCG TG|
|Sense||5′-CGG GAA ACT GTG GCG TGA TG|
|Antisense||5′-AAC TGT GAG GAG GGG AGA TT|
Immunohistochemistry for CARMA1 and BCL10 Proteins
Immunohistochemical staining for CARMA115 and BCL1028–31 proteins was performed in 30 lymphomas in which formalin fixed, paraffin embedded tissue specimens obtained by surgical or endoscopic mucosal resection were available. Tissue sections cut at a thickness of 4 μm were deparaffinized and rehydrated. After antigen retrieval by heat treatment in 0.01 M citrate buffer, pH 6.0, at 99 °C for 15 minutes, endogenous peroxidase activity was blocked by 3% hydrogen peroxide. A polyclonal antibody against CARMA1 (AL220; Alexis Biomedicals, Tokyo, Japan) and a monoclonal antibody against BCL10 (clone 151; Zymed Laboratory, Inc., South San Francisco, CA) were applied for the primary antibody at 4 °C for 12 hours, followed by incubation with secondary biotinylated antibody for 20 minutes, and then a streptavidin-biotin-peroxidase complex reagent for 20 minutes at room temperature (Histofine SAB-PO Kit; Nichirei, Tokyo, Japan) The sections were washed thoroughly with phosphate-buffered saline between the steps. Diaminobenzidine tetrahydrochloride was used as the chromogen, and hematoxylin was used for counterstaining.
Statistical differences were evaluated using either the Fisher exact-probability test or the chi-square test. Probabilities (P values) < 0.05 were regarded as statistically significant.
Expression of mRNA of CARMA1, CARD9, Bcl10, and API2-MALT1 in Gastric Lymphoma and Chronic Gastritis
The reference gene GAPDH was amplified successfully by RT-PCR in all specimens. Table 2 compares the mRNA expression of CARMA1, CARD9, Bcl10, and API2-MALT1 chimeric transcript between gastric lymphoma and chronic gastritis. CARMA1 mRNA was detected in 36 patients (55%) of gastric lymphoma but was detected in only 3 patients (17%) of chronic gastritis (P < 0.005) (Fig. 1A). CARD9 and Bcl10 were positive in 31 lymphomas (48%) and 64 lymphomas (98%), respectively, but in none of the specimens of chronic gastritis (P < 0.001) (Fig. 1B). The API2-MALT1 chimeric transcript was detected only in 5 patients of lymphoma. All lymphoma specimens that were positive for the chimeric transcript were low-grade MALT lymphoma.
|Factor||No. of patients (%)|
|Lymphoma (n = 65)||Gastritis (n = 18)||P valuea|
|CARMA1||36 (55)||3 (17)||0.003|
|CARD9||31 (48)||0 (0)||<0.001|
|Bcl10||64 (98)||0 (0)||<0.001|
|API2-MALT1||5 (8)||0 (0)||0.284|
CARMA1 mRNA Expression in Gastric Lymphoma
Table 3 summarizes the correlations between positivity for CARMA1 mRNA and the clinicopathologic findings for gastric lymphoma. CARMA1 mRNA was detected more frequently in DLBL without MALT lymphoma (100%) than in the other 2 types that contained areas of MALT lymphoma (30 of 59 patients; 51%; P < 0.05). There was no difference in the positive rate of CARMA1 between low-grade MALT lymphoma (51%) and DLBL plus MALT lymphoma (50%). When we compared CARMA1 expression in chronic gastritis, MALT lymphoma with or without DLBL, and DLBL without MALT lymphoma, the positive rate increased as the histologic grade advanced (Fig. 2A) (P < 0.005). The frequency of CARMA1 expression was significantly greater in the H. pylori-negative patients (100%) than in the positive patients (50%; P < 0.05) and in the API2-MALT1 chimeric transcript-positive patients (100%) than in the negative patients (52%; P < 0.05). There was a significant correlation between the expression of CARMA1 mRNA and CARD9 mRNA; CARMA1 was positive in 81% of the CARD9-positive patients, whereas it was positive in only 31% of the CARD9-negative patients (P < 0.001). In 51 patients who underwent H. pylori eradication, CARMA1 was detected more frequently in patients who did not respond to eradication (76%) than in patients who showed a complete remission after eradication therapy (37%; P < 0.05). The expression of CARMA1 was not correlated with gender, patient age, clinical stage, or the depth of tumor invasion determined by EUS.
|Factor||No. of patients (%)|
|CARMA1 positive||CARMA1 negative||P value|
|Male||18 (53)||16 (47)||0.435a|
|Female||18 (58)||13 (42)|
|≥ 62 yrs||21 (55)||17 (45)||0.591a|
|≤ 61 yrs||15 (56)||12 (44)|
|Low-grade MALT lymphoma/MALT lymphoma plus DLBL||30 (51)||29 (49)||0.024a|
|DLBL without MALT lymphoma||6 (100)||0 (0)|
|Stage I||24 (53)||21 (47)||0.411a|
|≥ Stage II1||12 (60)||8 (40)|
|Depth of invasion|
|Mucosa||8 (44)||10 (56)|
|Submucosa||15 (58)||11 (42)||0.524b|
|Muscularis propria or beyond||13 (62)||8 (38)|
|H. pylori status|
|Positive||29 (50)||29 (50)||0.012a|
|Negative||7 (100)||0 (0)|
|API2-MALT1 chimeric transcript|
|Positive||5 (100)||0 (0)||0.045a|
|Negative||31 (52)||29 (48)|
|Positive||25 (81)||6 (19)||<0.001a|
|Negative||11 (32)||23 (68)|
|Response to H. pylori eradication (n = 51)|
|Complete remission||11 (37)||19 (63)||0.006a|
|No response||16 (76)||5 (24)|
CARD9 mRNA Expression in Gastric Lymphoma
The correlations between CARD9 mRNA expression and clinicopathologic findings are indicated in Table 4. CARD9 mRNA expression was positive more frequently in H. pylori-negative patients than in H. pylori-positive patients (86% vs. 43%; P < 0.05) and was positive more frequently in API2-MALT1 chimeric transcript-positive patients (100%) than in API2-MALT1 chimeric transcript-negative patients (43%; P < 0.001). Among the patients who underwent H. pylori eradication therapy, CARD9 expression was detected more frequently in patients who did not respond to eradication therapy (71%) than in patients who achieved complete remission (33%; P < 0.01). No significant correlation was found between CARD9 expression and gender, patient age, histologic type, clinical stage, or depth of tumor invasion. When we compared CARD9 expression among chronic gastritis, MALT lymphoma with or without DLBL, and DLBL without MALT lymphoma, the rate of positive CARD9 expression increased as the histologic grade advanced (Fig. 2B) (P < 0.001). However, the difference between MALT lymphomas (27 of 59 patients; 46%) and DLBL (67%) did not reach statistical significance (P = 0.293).
|Factor||No. of patients (%)|
|CARD9 positive||CARD9 negative||P value|
|Male||14 (41)||20 (59)||0.197a|
|Female||17 (55)||14 (45)|
|≥62 yrs||20 (53)||18 (47)||0.245a|
|≤61 yrs||11 (41)||16 (59)|
|Low-grade MALT lymphoma/MALT lymphoma plus DLBL||27 (46)||32 (54)||0.293a|
|DLBL without MALT lymphoma||4 (67)||2 (33)|
|Stage I||18 (40)||27 (60)||0.055a|
|≥ Stage II1||13 (65)||7 (35)|
|Depth of invasion|
|Mucosa||6 (33)||12 (67)|
|Submucosa||12 (46)||14 (54)||0.200b|
|Muscularis propria or beyond||13 (62)||8 (48)|
|H. pylori status|
|Positive||25 (43)||33 (57)||0.039a|
|Negative||6 (86)||1 (14)|
|API2-MALT1 chimeric transcript|
|Positive||5 (100)||0 (0)||<0.001a|
|Negative||26 (43)||34 (57)|
|Response to H. pylori eradication (n = 51)|
|Complete remission||10 (33)||20 (67)||0.008a|
|No response||15 (71)||6 (29)|
Correlation between CARMA1 and CARD9 in Gastric Lymphoma
Gastric lymphomas were divided into 25 tumors that were positive for both CARMA1 mRNA and CARD9 mRNA, 23 tumors that were negative for both CARMA1 and CARD9, and 17 tumors that were positive for either 1 of the 2 caspases. H. pylori-negative tumors, API2-MALT1 chimeric transcript-positive tumors, and tumors that did not respond to H. pylori eradication were more frequent among patients with gastric lymphomas that were positive for both CARMA1 and CARD9 than among patients who had gastric lymphomas that were negative for the caspases (data not shown). When we assessed the efficacy of H. pylori eradication in 51 patients, 18 of 21 patients (85.7%) who showed no response to eradication were positive for either CARMA1, CARD9, or both, whereas 15 of 30 patients (50%) who achieved complete remission were negative for both molecules (P = 0.009). Positive and negative predictive values for the expression of either CARMA1 mRNA, CARD9 mRNA, or both for unresponsiveness to eradication therapy were 54.5% (18 of 33 patients) and 83.3% (15 of 18 patients), respectively.
Immunohistochemical Expression of CARMA1 and BCL10 Proteins in Gastric Lymphoma
Immunohistochemical staining showed weak CARMA1 protein expression in some of the nonneoplastic lymphoid cells in all patients. We regarded tumors in which obvious perinuclear or nuclear expression of CARMA1 was observed in ≥ 10% of neoplastic lymphoid cells as positive in this study (Fig. 3). Consequently, CARMA1 protein was positive in 18 of 30 tumors (60%). Although variable degrees of cytoplasmic staining for BCL10 were observed in all of the examined tumors (Fig. 4A), aberrant nuclear BCL10 expression28–31 was detected in 13 of 30 tumors (43%) (Fig. 4B).
Table 5 summarizes the correlations between the immunohistochemical expression of CARMA1 protein and the clinicopathologic findings. There was a significant correlation between the expression of CARMA1 protein and CARMA1 mRNA; CARMA1 protein was positive in 80% of the tumors that were positive for CARMA1 mRNA but was positive in only 20% of the tumors that were negative for CARMA1 mRNA (P < 0.005). A significant correlation also was observed between positive CARMA1 protein expression and nuclear BCL10 expression; CARMA1 protein was detected more frequently in tumors with nuclear BCL10 expression (92%) than in tumors without BCL10 expression (35%; P < 0.005). However, the expression of CARMA1 protein was not correlated with the positive expression of CARD9 mRNA or API2-MALT1 chimeric transcript.
|Factor||No. of patients (%)|
|Positive for CARMA1 protein||Negative for CARMA1 protein||P valuea|
|Positive||16 (80)||4 (20)||0.003|
|Negative||2 (20)||8 (80)|
|Positive||13 (72)||5 (28)||0.098|
|Negative||5 (42)||7 (58)|
|API2-MALT1 chimeric transcript|
|Positive||4 (100)||0 (0)||0.112|
|Negative||14 (54)||12 (46)|
|Nuclear BCL10 expression|
|Positive||12 (92)||1 (8)||0.002|
|Negative||6 (35)||11 (65)|
Recent studies have suggested that CARMA1, Bcl10, and MALT1 are signaling compounds that play crucial and specific roles in T-cell receptor-induced and B-cell receptor-induced NF-κB activation.18–21 In addition, it has been suggested that all three of these molecules play an essential role in the proliferation of B cells and T cells and even in the development of MALT lymphoma.22, 29 Thus, it is believed that Bcl10 is associated with the development or progression of MALT lymphoma mediated by the activation of NF-κB.6, 13, 14, 28–31 Although both CARMA1 and CARD9 are upstream activators of Bcl10 in the NF-κB signaling pathway, to our knowledge, little is known to date regarding the significance of these molecules in the pathogenesis of gastric MALT lymphoma.
Although the expression of CARMA1 protein has been demonstrated in cultured lymphoid cells,15 to our knowledge no previous studies have investigated its expression in gastric lymphoma. In the current immunohistochemical study, CARMA1 protein overexpression was observed in 60% of gastric lymphomas and was associated with both CARMA1 mRNA expression and nuclear expression of BCL10 protein (Table 5) (Figs. 4, 5). These observations suggest that CARMA1 contributes to the up-regulation of BCL10 in gastric lymphoma.
In the current study, CARMA1 mRNA was detected in 55% of primary gastric B-cell lymphomas and in only 17% of chronic gastritis samples. Figure 2A shows that the frequency of CARMA1 mRNA expression increased as the histologic grade advanced (P < 0.005). Immunohistochemical expression of CARMA1 protein also was associated with the overexpression of CARMA1 mRNA in specimens of lymphoma. These results suggest that CARMA1 may play a role in the development and progression of gastric MALT lymphoma and DLBL. In addition, CARMA1 may be associated more closely with gastric B-cell lymphomas that are unrelated to H. pylori, because CARMA1 was found frequently in H. pylori-negative patients, API2-MALT1 chimeric transcript-positive patients, and patients who did not respond to H. pylori eradication therapy (Table 3).
In addition to the expression of CARMA1, the frequent expression of CARD9 mRNA was associated with H. pylori negative patients, API2-MALT1 chimeric transcript-positive patients, and patients who did not respond to H. pylori eradication. In addition, the positive rate of CARD9 mRNA increased according to the histologic progression of lymphoma (Fig. 2B), although the difference in the positive rate between MALT lymphomas and DLBL was not significant. Furthermore, there was a significant correlation noted between the expression of CARD9 mRNA and CARMA1 mRNA (Table 3). It has been reported that CARD9 has a molecular structure similar to that of CARMA1, which contains a CARD domain with an additional coiled-coil motif to bind to Bcl10 and to activate NF-κB.24, 25 On the basis of these observations, it seems possible that CARD9 may contribute to the pathogenesis of gastric B-cell lymphoma, as well as CARMA1.
Bcl10 mRNA was detected frequently, as reported previously, whereas the API2-MALT1 chimeric transcript was found infrequently in our specimens of gastric B-cell lymphoma.11, 28 The API2-MALT1 chimeric transcript, which results from t(11;18) (q21;q21), is one of the genetic alterations that is specific to MALT lymphoma, and this fusion molecule has been detected in 15–30% of all gastric low-grade MALT lymphomas.10–12 It has been shown that the API2-MALT1 fusion protein activates NF-κB as strongly as the Bcl10-MALT1 complex.9, 29 Recently, Stoffel et al.32 demonstrated that the API2-MALT1 fusion protein inhibits p53-mediated apoptosis in an NF-κB-dependent manner, and it is considered that this inhibition contributes to the development of MALT lymphoma. In our current study, API2-MALT1 chimeric transcript was detected in 5 of 43 patients (12%) with low-grade MALT lymphoma, all of which expressed both CARMA1 and CARD9 mRNA (Tables 3, 4). On the basis of these findings, it is presumed that CARMA1 and CARD9 are molecules that play a key role in the enhanced signal transduction of the Bcl10-NF-κB pathway, especially in API2-MALT1-associated MALT lymphoma.
It also should be noted that the positive rate for CARMA1 mRNA or CARD9 mRNA (65%) was lower than that for Bcl10 (98%) in our specimens of gastric lymphoma. This observation suggests that other putative, upstream activators of Bcl10 are associated with the NF-κB signaling pathway for the development of gastric MALT lymphoma. This speculation awaits further detailed molecular analyses.
Although API2-MALT1 chimeric transcript reportedly is a specific marker for predicting unresponsiveness to H. pylori eradication in gastric MALT lymphoma,10–12 its sensitivity in our series was very low because of its low prevalence. Among 51 patients who received eradiation therapy, including only 4 patients with API2-MALT1 chimeric transcript, the specificity and positive predictive value of this genetic alteration were both 100%, whereas the sensitivity and negative predictive value were only 19.0% and 63.8%, respectively (detailed data not shown). Conversely, patients who were positive for either CARMA1 mRNA, CARD9 mRNA, or both were associated strongly with unresponsiveness to H. pylori eradication therapy (P = 0.009), with a sensitivity and a negative predictive value of 85.7% and 83.3%, respectively, although the specificity and positive predictive value were relatively low (50% and 54.5%, respectively). Thus, we believe that the evaluation of CARMA1 and CARD9 mRNA may be useful for predicting the efficacy of H. pylori eradication, especially in patients who are negative for the API2-MALT1 chimeric transcript.
In conclusion, there was a positive correlation noted between the expression levels of CARMA1 and of CARD9 mRNA in primary gastric B-cell lymphoma, and they were associated with H. pylori-negative tumors, API2-MALT1 chimeric transcript-positive tumors, and tumors that did not respond to H. pylori eradication therapy. Furthermore, the positive rate of these two molecules increased as the histologic grade advanced. These results suggest that CARMA1 and CARD9 may play some roles in the development or progression of gastric B-cell lymphoma, especially in patients with disease that is not related to H. pylori. Therefore, further investigations using large numbers of patients are warranted to clarify the role of these molecules in the pathogenesis of gastric MALT lymphoma.
- 1Extranodal lymphomas. Edinburgh: Churchill Livingstone, 1994., .
- 2JaffeES, HarrisNL, SteinH, VardimanJW, editors. World Health Organization classification of tumours. Pathology and genetics of tumours of haematopoietic and lymphoid tissues. Lyon: IARC Press, 2001.