Virchows Arch (2010) 457:659–667 DOI 10.1007/s00428-010-0992-7

ORIGINAL ARTICLE

The expression and distribution of group IIA phospholipase A2 in human colorectal tumours Tuulia Avoranta & Jari Sundström & Eija Korkeila & Kari Syrjänen & Seppo Pyrhönen & Jukka Laine

Received: 9 June 2010 / Revised: 1 September 2010 / Accepted: 29 September 2010 / Published online: 12 October 2010 # Springer-Verlag 2010

Abstract Secretory phospholipase A2 group IIA (IIA PLA2) is a protein shown to be increased in various human malignancies. The expression profile of this protein, however, is controversial in colorectal carcinoma. The aim of this study was to examine the distribution and expression of IIA PLA2 protein in benign, premalignant and malignant colorectal tumours as well as in peritumoural mucosa. Seven hyperplastic polyps, 24 adenomas and 83 colorectal carcinomas were stained with immunohistochemistry (IHC) for IIA PLA2. Four hyperplastic polyps, 12 adenomas and nine carcinomas were also evaluated for the sites of IIA PLA2 expression using mRNA in situ hybridisation (ISH). There was no immunoreactivity for IIA PLA2 in hyperplastic polyps. A total of 79% of adenomas and 31% of carcinomas showed IIA PLA2-immunopositive tumour cells in IHC, and the expression was localised to epithelial cells with ISH. In carcinomas, IIA PLA2-immunopositive apoptotic cells and necrosis were also found. The epithelial cells in the peritumoural mucosa showed immunopositivity for IIA PLA2 in 96% of cases, with considerably stronger intensity adjacent to carcinoma than in the more distal mucosa. Moreover, IIA PLA2-immunopositive malignant epithelial cells were found in 44% of cases in the invasive front of carcinomas. Our results suggest that the IIA PLA2 T. Avoranta (*) : E. Korkeila : K. Syrjänen : S. Pyrhönen Department of Oncology and Radiotherapy, University of Turku and Turku University Hospital, Savitehtaankatu 1 PB 52, 20521 Turku, Finland e-mail: [email protected] J. Sundström : J. Laine Department of Pathology, University of Turku and Turku University Hospital, Kiinamyllynkatu 10, 20520 Turku, Finland

protein content is dramatically decreased in malignant colorectal tumours as compared with adenomas. The protein is also found in the apoptotic cells, necrosis, peritumoural mucosa and in the invasive front of carcinomas. Keywords Group IIA phospholipase A2 . Localisation . Colorectal adenoma . Colorectal carcinoma

Introduction Phospholipase A2 (PLA2) family consists of several enzymes hydrolysing the sn-2 ester bond of membrane phospholipids, thus creating a lysophospholipid and releasing fatty acids—predominantly arachidonic acid (AA). Cyclooxygenases (COX) change free AA into eicosanoids, such as prostaglandin E2 (PGE2) [1]. Most colorectal cancers arise from pre-existing adenomas in a process called adenoma-to-carcinoma sequence. Because it takes several years for a normal cryptal cell to undergo neoplastic molecular changes and develop into clinically recognised neoplasm [2], it is important to find biological markers that are altered early in colorectal carcinogenesis. Because deregulation of the COX-2/PGE2 pathway is demonstrated to influence colorectal carcinogenesis via a number of distinct mechanisms [3], functional defects of PLA2 proteins have been regarded as potential predictive molecular markers in colorectal cancer [4–6]. Secretory phospholipase A2 group IIA (IIA PLA2) is a 14-kDa enzyme of PLA2-family found in a number of tissues and body fluids [7]. The concentration of group IIA PLA2 enzyme in blood plasma is known to increase dramatically in several diseases involving generalised inflammation [8, 9], and the content of IIA PLA2 has been demonstrated to be upregulated in a range of human

660

malignancies, such as prostate [10–12], gastric [13] and intestinal cancer [14, 15]. In the gastrointestinal tract, IIA PLA2 is particularly found in the Paneth cells of the small intestine and in the goblet cells of the large intestine [16, 17]. In contrast to normal colorectal mucosa, where little or no IIA PLA2 protein is present [15, 16, 18], increased expression of IIA PLA2 detected at protein and messenger RNA (mRNA) level has been reported in colorectal adenomas of familial adenomatous polyposis (FAP) patients [19]. Increased levels of IIA PLA2 protein have also been demonstrated in chronic inflammatory bowel disease [20–23] as well as in normal colorectal mucosa adjacent to adenocarcinoma [15, 18]. The expression profile of IIA PLA2 in colorectal carcinoma is more controversial; there are studies indicating that both the amount of IIA PLA2 content is increased [14, 15] and it is attenuated [18] in colorectal carcinomas. Furthermore, little data are available on the exact location of IIA PLA2 protein synthesis in colorectal tumours at mRNA level. In this study, we examined the distribution and sites of synthesis of IIA PLA2 in benign, premalignant and malignant human colorectal tumours as well as in peritumoural normal mucosa, with special emphasis on rectal carcinomas.

Materials and methods Samples Clinicopathological features of the study population are described in Table 1. The present series included two separate sets of series; the first consisted of paraffinembedded tissue samples of seven hyperplastic polyps (five from the colon, two from the rectum), 14 colorectal adenomas (nine from the colon, four from the rectum, one localisation missing) and 11 carcinomas (eight from the colon, three from the rectum) collected between 1994 and 1999. The second set consisted of paraffin-embedded samples from 72 rectal carcinomas collected between 2000 and 2009. The biopsies and resection samples in both sets had been taken during colonoscopies or operations performed at Turku University Hospital. Representative samples were taken following the standard criteria of colorectal sampling to establish the histological diagnosis. For this study, the tissue samples were obtained from the files of the Department of Pathology, Turku University Hospital. Among the 72 rectal carcinomas in the second set, none had distant metastases at the time of operation, and none of the patients had received any preoperative treatment prior to surgery. To make the study population as homogenous as

Virchows Arch (2010) 457:659–667

possible, only tumours located in the medial or distal rectum were included. In addition, adenomas with different degrees of dysplasia were found in ten resection samples of carcinomas. These ten adenomas were included in the analyses of adenomas. Immunohistochemistry for IIA PLA2 protein IIA PLA2 protein distribution was analysed using immunohistochemical staining (IHC) in formalin-fixed, paraffinembedded 5 μm-thick sections. The first set was stained using a G-protein purified in-house polyclonal rabbit IIA PLA2 IgG with a dilution of 1.04 μg/ml of IgG as described in detail earlier [24]. The cross-reactivity of the antibody has been tested against group IB, IIA, IID, IIE, IIF, III, V, X, XIIA and XIIB PLA2s. The antibody was shown to be specific for IIA PLA2 without significant cross-reactivity with other secretory small molecular weight PLA2s [25, 26]. For the larger second set (72 rectal carcinomas) Power Vision Kit (Immuno Vision Technologies, Co, USA) and LabVision Autostainer (Thermo Fisher Scientific, CA, USA) was used with the same IIA PLA2 antibody and dilution as mentioned above. For Autostainer the antigen was retrieved 2 × 7 min in pH 6.0 with DakoTarget Retrieval Solution (Dako, Glostrup, Denmark), and the slides were washed twice with Tris-buffered saline. In controls, the primary antibody was replaced by preimmune rabbit serum, with a dilution of 1:3,000. To further confirm the specifity of the in-house IIA PLA2 antibody, a commercial human monoclonal IIA PLA2 antibody (dilution of 1:3,000, Cayman Chemicals, Ann Arbor, Michigan, USA, catalog number 160500) was used for a group of ten rectal carcinomas with the Power Vision Kit and Lab Vision Autostainer methods as mentioned above. Assessment of IHC Assessment of IHC was performed by two authors (TA and JL) on light microscopy blinded to the clinicopathological data. The immunopositivity for IIA PLA2 in hyperplastic polyps, adenomas and carcinomas was determined by estimating the percentage of epithelial cells showing positive staining for IIA PLA2. For statistical purpose, the content of IIA PLA2 protein was determined to be zero when the percentage of immunopositive epithelial cells in the tumour was below 1%. In carcinomas, IIA PLA2-immunopositive apoptotic cancer cells, necrotic areas (within the tumour), and cells in the invasive front were defined to be present or absent, respectively. The apoptotic cells were identified by their typical morphology (condensed, fragmented chromatin and a clear halo around the apoptotic cell). The invasive fronts were identified histologically by the presence of “tumour

Virchows Arch (2010) 457:659–667 Table 1 Clinicopathological features of the study population

661 Variable

Hyperplastic polyp n=7

Sex Male Female Mean age (SD) Tumour localisationa Colon Rectum Postoperative tumour statusb T1 T2 T3 T4 Postoperative nodal statusb N0 N1

a

Missing data of localisation of one adenoma b Missing data of the T, N and stage of 11 carcinomas

Adenoma n=24

Carcinoma n=83

3 (43%) 4 (57%) 61.4 (±7.9)

14 (58%) 10 (42%) 73.3 (±8.8)

37 (45%) 46 (55%) 74.0 (±10.8)

5 (71%) 2 (29%)

9 (39%) 14 (61%)

8 (10%) 75 (90%) 5 27 37 3

45 (62%) 15 (21%)

N2 Postoperative stage (TNM)b Stage I Stage II Stage III Tumour grade Gr I Gr II Gr III

budding” and an infiltrating growth pattern in the tumour– host interface. The peritumoural mucosa was divided in the mucosa adjacent to carcinoma (proximal) and in the mucosa slightly further from carcinoma (distal). The presence or absence of IIA PLA2 protein in the peritumoural mucosa was determined, and the intensity was estimated and divided into four classes: 0 “no staining”, 1 “weak staining”, 2 “moderate staining” and 3 “intense staining”. For statistical purpose, classes 0 and 1 as well as classes 2 and 3 were combined in the case of proximal epithelium. The staining intensities of distal epithelium were as well divided into two groups; “no staining” and “weak/moderate/intense staining”. In situ hybridisation for IIA PLA2 mRNA Four hyperplastic polyps, 12 adenomas and nine carcinomas were evaluated for the expression of IIA PLA2 mRNA using the method described in detail earlier [27]. Briefly, in situ hybridisation (ISH) was performed on formalin-fixed, paraffin-embedded tissue sections by probing with human group II PLA2 anti-sense (test) and sense (control) singlestranded RNA riboprobes. A 0.45-kb cDNA sequence

(7%) (38%) (51%) (4%)

12 (17%) 26 (36%) 19 (26%) 27 (38%) 9 (38%) 14 (58%) 1 (4%)

16 (19%) 56 (68%) 11 (13%)

covering the protein coding area of human group II PLA2 in pUC18 plasmid [28] was inserted into the HindIII and BamHI sites of pGEM-3Z transcription vector (Promega, Madison, WI, USA). Digoxigenin-labelled anti-sense and sense RNA probes were synthesised by in vitro transcription with T7 and SP6 RNA polymerases, respectively, by using a DIG RNA Labelling Kit (Boehringer Mannheim, Mannheim, Germany), and the yields were estimated by using a DIG Nucleic Acid Detection Kit (Boehringer Mannheim). The probes were purified with Quick Spin 25 columns (Boehringer Mannheim). The digoxigenin label was detected with alkaline phosphatase-labelled anti-digoxigenin Fab fragments by using disodium-3-(4-metoxyspiro{1,2-dioxyetane3,2′-(5′-chloro)tricyclo[3, 3, 1, 13, 7]decan}-4-yl) phenylphosphate (CSPD®, Boehringer Mannheim) as a substrate. The percentage of epithelial cells positive for IIA PLA2 ISH was estimated by a pathologist (JL) on light microscopy. Statistical analysis All statistical analyses were run using PASW Statistics® 18.01 for Windows (SPSS, Inc., Chicago, USA) software package. Frequency tables were analysed using the χ2 test, with the likelihood ratio or Fisher’s exact test for

662

Virchows Arch (2010) 457:659–667

categorical variables; 2×2 tables were used to calculate odds ratio (OR) and 95% confidence interval (CI) using the exact method. Non-parametric tests were used to compare the means of continuous variables because the mean IIA PLA2 -positivity in tumours was not normally distributed. The differences in the immunopositivity between tumours were analysed with Mann–Whitney or Kruskal–Wallis test for two- and multiple independent samples, respectively. In the ten cases with both adenoma and carcinoma present in the same specimen, the concordance of IIA PLA2immunopositivity between adenomas and carcinomas was calculated using non-parametric paired-samples test (Wilcoxon-signed ranks test). Correlation between IIA PLA2 protein content in IHC and mRNA expression in ISH was calculated using bivariate correlation test (Spearman’s rho) for non-parametric variables. All statistical tests were performed as two-sided and declared significant at p value <0.05.

Results The results for IHC and ISH for IIA PLA2 are summarised in Tables 2 and 3, and the examples of the localisation of IIA PLA2 protein and mRNA in colorectal tumours and adjacent mucosa are presented in Figs. 1 and 2. Protein content and mRNA expression of IIA PLA2 in hyperplastic polyps and adenomas There was no immunoreactive IIA PLA2 protein in hyperplastic polyps. A total of 19 (79%) of 24 adenomas showed positive immunoreaction for IIA PLA2 protein in Table 2 The immunoreactivity of group IIA PLA2 protein in colorectal tumours

Protein content and mRNA expression of IIA PLA2 in carcinomas A total of 57 (69%) of all carcinomas were totally devoid of immunoreactive IIA PLA2 protein (Fig. 2a). In the 26 carcinomas (31%) positive for IIA PLA2 protein, the percentage of immunopositive cancer cells varied from 1% to 20% (mean 0.8%, 95% CI 0.2–1.4). There was no difference in the mean percentage of IIA PLA2immunopositive cells between colonic and rectal carcinomas (p=0.63). ISH was performed in nine carcinomas and showed expression of IIA PLA2 mRNA in three of them. The percentage of positive malignant epithelial cells varied from 0% to 10% (mean 1.3%). The IIA PLA2 protein content in IHC and mRNA expression in ISH were intimately correlated (r = 0.93, p < 0.001) in hyperplastic polyps, adenomas and carcinomas.

Tumour type and localisation

Number of cases (%)

Number of IIA PLA2 -positive cases (%)

Mean IIA PLA2positivity in percents (95% CI)b

Significance (p value)c

Hyperplastic polyp

7

0

0

=0.001 vs. adenoma

Colon Rectum Adenomaa a Localisation data of one adenoma missing

IHC (Fig. 1a–b). The percentage of epithelial cells positive for IIA PLA2 protein in 24 adenomas varied from 0% to 60% (mean 14.8%, 95% CI 6.9–23). There was no significant difference in the mean percentage of IIA PLA2-immunopositive epithelial cells between colonic and rectal adenomas (p=0.30). In the adenoma presented in Fig. 1, there was an old haemorrhage in the stroma containing hemosiderin-phagocyted macrophages. Hemosiderin is seen both in the IHC and ISH (Fig. 1a–d). ISH was performed in four hyperplastic polyps and 12 adenomas. There were no IIA PLA2 mRNA expression in hyperplastic polyps whereas 67% of adenomas showed IIA PLA2 mRNA expression varying from 1% to 30% (mean 5.0%, Fig. 1c–d).

5 (71) 2 (29) 24

Colon

19 (79) 9 (39)

6 (67)

14.8 (6.9–23) 10.8 (0–24)

b

Mean% of IIA PLA2 immunopositive epithelial cells in tumours. p value=0.018 for differences between groups (hyperplastic polyps, adenomas and carcinomas) in colonic tumours and <0.001 in rectal tumours as analysed with Kruskal–Wallis test c

Mann–Whitney U test

Rectum Carcinoma Colon Rectum

14 (61)

12 (86)

17.1 (5.1–29)

8 (10) 75 (90)

26 (31) 2 (25) 24 (32)

0.8 (0.2–1.4) 0.3 (0–0.6) 0.9 (0.2–1.5)

83

<0.001 vs. carcinoma 0.058 vs. colon carcinoma 0.30 vs. rectum adenoma <0.001 vs. rectum carcinoma 0.63 vs. rectum carcinoma

Virchows Arch (2010) 457:659–667 Table 3 The expression of group IIA PLA2 mRNA in colorectal tumours with ISH

663 Tumour type

Number of cases

Number of IIA PLA2 mRNA-positive cases (%)

Mean IIA PLA2 mRNA content in percents (min, max)

Hyperplastic polyp Adenoma

4 12

0 8 (67)

0 5.0 (min 0%, max 30%)

Carcinoma

9

3 (33)

1.3 (min 0%, max 10%)

In the 72 rectal carcinomas of the second set, there was a tendency to an increase of IIA PLA2-immunopositivity with tumour grade, but the difference did not reach statistical significance (p=0.094). In five carcinomas with an excessive mucinotic component, there were significantly (p=0.025) more epithelial cells positive for IIA PLA2 protein (mean 5.4%, min 0, max 20, Fig. 2b) than in carcinomas with no excessive mucin production (n=67, mean 0.54%, min 0, max 10). In the 15 rectal carcinomas (21%) with IIA PLA2immunopositive apoptotic cells (Fig. 2c), there also were more immunoreactive vital malignant epithelial cells than in the carcinomas with no detectable IIA PLA2 immunopositive apoptotic cells (p<0.001). The same difference was seen in the 15 carcinomas (21%) with IIA PLA2immunopositive necrotic areas (p=0.001, Fig. 2c). Apoptotic cells and necrotic areas showed concordant immunoreaction for IIA PLA2 in 58 cases (81%, OR=8.2, p= 0.001) and discrepant reaction in 14 cases (19%).

Fig. 1 Localisation of IIA PLA2 protein and mRNA in colonic adenoma. a Colonic adenoma showing IIA PLA2immunopositive epithelial cells in IHC. The immunoreactive material was usually localised in the apical cytoplasm of epithelial cells. Occasional IIA PLA2-immunopositive metaplastic Paneth cells were seen. Bar 200 μm. b Same adenoma as in a stained with rabbit 0-serum (negative control for IHC). c Positive reaction with mRNA ISH in the same epithelial cells of the adenoma that showed positive immunoreaction for IIA PLA2 protein in a. Anti-sense (test) probe. d An adjacent section of the colonic adenoma shown in c. Sense (control) probe. An area of hemosiderin-containing macrophages is marked with an asterisk

Comparison of immunopositivity for IIA PLA2 protein between hyperplastic polyps, adenomas and carcinomas The differences between the mean percentages of IIA PLA2 protein detected with IHC in hyperplastic polyps, adenomas and carcinomas were statistically significant both in the colon and rectum (p=0.018 and p<0.001 in Kruskal–Wallis test, respectively). Adenomas always showed the highest content of IIA PLA2 protein. When the IIA PLA2-immunopositivity in only adenomas and carcinomas was compared according to tumour location, the differences between the mean percentages of IIA PLA2 protein were highly significant in the rectum (p<0.001) but did not reach the statistical significance in the colon (p=0.058). In the ten cases with an adenoma present in the same specimen with carcinoma, adenoma always showed IIA PLA2 protein in IHC (mean 20.8%, 95% CI 4.9–37), whereas carcinoma showed immunopositivity for IIA PLA2 protein in three specimens only (mean 0.4%, min 0, max 2, p=0.001, see Fig. 2a).

664

Virchows Arch (2010) 457:659–667

Fig. 2 Localisation of IIA PLA2 protein in colorectal tumours and peritumoural mucosa. a Rectal adenoma (arrow) and proximal peritumoural mucosa (arrowhead) showing IIA PLA2immunopositive epithelial cells in IHC. Epithelial cells in carcinoma (asterisk) are completely devoid of IIA PLA2 protein. When present, the immunoreactive material was usually localised in the cytoplasm of epithelial cells either in apical or granular pattern. Bar 200 μm. b Carcinoma with an excessive mucinotic component showing IIA PLA2-immunopositive epithelial cells (compare to carcinoma in a). c Apoptotic cell (arrow) and necrotic area (asterisk) showing positive immunoreaction for IIA PLA2 protein. In eight cases (11%), there were both IIA PLA2 immunopositive apoptotic cells and necrotic areas in the same specimen, and in 50 cases (69%) both apoptotic cells and

necrotic areas were devoid of IIA PLA2 immunoreaction. d Distal peritumoural mucosa showing a weak positive immunoreaction for IIA PLA2 protein (compare to proximal peritumoural mucosa in a). e Invasive front showing immunopositive cells in the tumour–host interface. Usually, it was impossible to determine whether immunoreactive cells were stroma cells or apoptotic or necrotic cancer cells, but the ISH showed positive reaction only in malignant epithelial cells. Outside the invasive front, there usually were some stroma cells positive for IIA PLA2 with IHC and ISH. Furthermore, immunopositivity for IIA PLA2 was present in the elastic lamina of blood vessel walls. f Higher magnification of another invasive front with numerous IIA PLA2-positive cells. Bar 100 μm

IIA PLA2 in the peritumoural normal mucosa and in the invasive front

ably stronger than the staining intensity in the more distal mucosa (Fig. 2d; 16% showing moderate staining and none showing intense staining; p=0.012, OR 8.7). The peritumoural mucosa also showed the expression of IIA PLA2 with ISH. In 32 out of 72 carcinomas (44%), IIA PLA2-positive cells were found in the invasive front (i.e. in the tumour–

The proximal peritumoural mucosa showed immunoreaction for IIA PLA2 protein in 96% of cases. The staining intensity in the proximal peritumoural mucosa (Fig. 2a; 67% showing moderate or intense staining) was consider-

Virchows Arch (2010) 457:659–667

host interface, Fig. 2e–f). These 32 carcinomas showed a tendency to have more often IIA PLA2 protein detected with IHC in their apoptotic and necrotic malignant cells when compared to carcinomas with no IIA PLA2 immunopositive cells in their invasive front (p=0.08 for apoptosis and p<0.001 for necrosis). It is important to note that hemosiderin pigment seen in the stromal macrophages (as in Fig. 1) was usually absent in the stroma of the invasive front and did not interfere in the interpretation of IIA PLA2 immunohistochemistry.

Discussion Secretory phospholipase A2 group IIA is an enzyme involved in the mediation of inflammation [8, 9]. The expression of IIA PLA2 has been demonstrated to be increased in various human malignancies [1], including colorectal cancer [14, 15], and it has been suggested to play an important role in tumour development and progression [29–31]. In mice, the IIA PLA2 gene has been implicated as a candidate gene for Modifier of Min-1 locus [29, 32], which modifies tumour number in mice carrying a dominant germline mutation in their Apc-gene (the Minmice) [33]. This raised the question of the human counterpart as a potential tumour suppressor gene. Although the gene for human IIA PLA2 (PLA2G2A) has been mapped to chromosome 1p35-36 [34], a region often lost in colorectal cancer [35], sequence analysis of the complete coding region has not revealed any somatic mutations of the PLA2G2A gene in colorectal cancer, suggesting that this gene is not likely to be a tumour-suppressor gene in humans [36]. In the present study, we examined the distribution and sites of synthesis of IIA PLA2 with IHC and ISH in hyperplastic polyps, adenomas and carcinomas, as well as in non-neoplastic peritumoural mucosa. None of the hyperplastic polyps showed IIA PLA2 protein or mRNA. Altogether 79% of adenomas contained IIA PLA2 protein, whereas only 31% of the colorectal carcinomas showed immunopositivity for IIA PLA2 protein. The expression of IIA PLA2 in adenomas and carcinomas was localised in epithelial cells with ISH, indicating that IIA PLA2 protein and mRNA are present in the same cells. Interestingly, in the ten cases with both adenoma and carcinoma adjacent to each other, epithelial cells of all adenomas were immunopositive for IIA PLA2 protein, whereas carcinoma showed IIA PLA2 protein in 33% of cases only. This may suggest that the malignant cells lose their ability to express IIA PLA2 when invasive carcinoma develops in the adenoma. Our results are in line with the findings of Kennedy et al. [19] who reported elevated gene and protein expression of IIA PLA2 in colorectal adenomas from FAP patients, and those of Edhemovic et al. [18] who found almost no IIA

665

PLA2 activity in the colorectal carcinoma. In addition, Praml et al. [37] reported that lack of IIA PLA2 expression is very common among colorectal cancer cell lines in contrast to normal mucosa. In contrast, both the studies of Wendum et al. [14] and Buhmeida et al. [15] showed positive staining for IIA PLA2 in over 50% of the colorectal carcinomas. In the latter study, the patients with IIA PLA2-negative tumours had a better prognosis than the patients with positive tumours [15]. At least some of the discrepancies between these observations could be due to different methods, antibodies and their dilutions used in these different studies. The ISH method we have used is based on RNAase-free conditions [27]. When IHC and ISH are positive in the same cells (i.e. epithelial cells), the interpretation is usually fairly easy, but in the context of apoptotic and necrotic cells, it may be difficult. In the present study, we found IIA PLA2 protein and mRNA usually in the same epithelial cells, and the presence of IIA PLA2 protein and mRNA in colonic epithelium correlated intimately in adenomas and carcinomas for individual samples. Moreover, the antibody used in this study has earlier been used in IHC and ISH in the samples from Chron’s disease and colitis ulcerosa patients with good results (high signal, low background and concordance between IHC and ISH in Paneth cells of the small intestine and colon) [20, 21, 38]. We found five rectal carcinomas with excessive mucinotic component containing considerably more IIA PLA2positive cancer cells than the non-mucinotic carcinomas. Increased content of IIA PLA2 in mucinotic carcinomas has been reported previously by Edhemovic et al. [18]. As mucinotic rectal carcinomas are commonly considered high-grade carcinomas, it is possible that poorly differentiated tumours express more IIA PLA2 than more differentiated tumours. In our series, we noticed a tendency to an increase of IIA PLA2 immunoreaction with tumour grade, but the difference did not reach statistical significance. We also analysed IIA PLA2 immunopositivity in apoptotic and necrotic cells. Interestingly, there were more IIA PLA2-positive apoptotic and necrotic cells in carcinomas with more immunopositive vital cancer cells. In most cases, apoptotic cells and necrotic areas showed a concordant pattern in immunoreaction for IIA PLA2. These findings are noteworthy, as IIA PLA2 has been suggested to promote apoptosis of cancer cells [39]. One possible explanation is that the typically lost membrane asymmetry in cancer cells enables increased IIA PLA2 activity [39]. Furthermore, AA, the product of IIA PLA2’s enzymatic activity, can stimulate an important apoptotic pathway via the activation of sphingomyelinases and the subsequent release of ceramide [40, 41]. In the light of our study, it seems unlikely that phospholipase itself would cause cell death because relatively few cancer cells expressed IIA

666

PLA2 protein. The more likely explanation is that the apoptotic or necrotic cancer cells are somehow engaged to express IIA PLA2, which could assist the host in removing apoptotic cells. In the present series, the peritumoural non-neoplastic epithelium showed positive immunoreaction for IIA PLA2 in 96% of the cases. This is consonant with the findings of Edhemovic et al. [18] and Buhmeida et al. [15] who reported increased content of IIA PLA2 in normal mucosa adjacent to carcinoma. IIA PLA2 is inducible by proinflammatory cytokines, such as tumour necrosis factor α and interleukin 1β [42], both being abundantly present in the vicinity of carcinoma [43]. Thus, this increased IIA PLA2 protein content in normal mucosa adjacent to carcinoma may well result from an increased secretion of pro-inflammatory cytokines by the cancer cells as suggested earlier, too [18]. Our observation of the weak or absent intensity of immunoreaction in normal colonic mucosa more distally to cancer supports this view. In the studies where the normal colorectal epithelium has been reported to be devoid of IIA PLA2 protein [15, 16, 18], the specimens have been taken from the mucosa more distant to cancer site. Frequently, the invasive front of rectal carcinomas showed immunopositive cells even in the absence of IIA PLA2 immunoreactivity in vital cancer cells in the same sample. This is an interesting finding because the cells in the invasive front are thought to represent a dynamic interface between pro- and anti-tumour factors [44]. The glands in the invasive front often showed disintegrated structures suggesting that these cells have possibly lost their symmetric membrane structure, which in turn may result in increased IIA PLA2 activity. In addition, abundant inflammatory cell infiltrate almost invariably present in the invasive front may also trigger IIA PLA2 production in some cells. Our observations could explain the previous results of Tribler et al. [45] who found that the peripheral areas of colon carcinoma contained significantly higher levels of IIA PLA2 protein than the samples taken from the central tumour regions. In our study, it was not always possible to determine in IHC whether immunoreactive cells in the invasive front were stromal cells or apoptotic or necrotic cancer cells. However, ISH showed IIA PLA2 mRNA only in malignant epithelial cells suggesting that at least the majority of IIA PLA2 protein present in the invasive front originates from the tumour cells rather than from the stromal cells. We conclude that the majority of colorectal adenomas show IIA PLA2 expression, whereas the expression is disturbed and IIA PLA2 protein content is dramatically reduced in malignant colorectal tumours as compared to adenomas. In addition, peritumoural mucosa shows in-

Virchows Arch (2010) 457:659–667

creased expression and content of IIA PLA2. It is possible that the down-regulation of IIA PLA2 has a role in colorectal carcinogenesis, but it remains to be established how and when the malignant cells lose their capability to produce IIA PLA2. In addition, the questions whether IIA PLA2 is needed in apoptotic cell death in colorectal carcinomas and whether radiation therapy affects IIA PLA2 expression via cancer cell death remain unanswered. Furthermore, our results confirm that IIA PLA2 in the invasive front is derived from the malignant epithelial cells. Acknowledgments We are grateful to Sinikka Kollanus for her skilful and expeditious help in laboratory work, Jaakko Liippo for aid with the digital pictures and Prof. Timo Nevalainen for the polyclonal rabbit IIA PLA2 antibody used in the present study. This research work has been supported by the grants from The Special Government Funding (EVO) allocated to Turku University Hospital. The authors declare that they have no conflict of interest.

References 1. Laye JP, Gill JH (2003) Phospholipase A2 expression in tumours: a target for therapeutic intervention? Drug Discov Today 8:710–716 2. Tanaka T (2009) Colorectal carcinogenesis: review of human and experimental animal studies. J Carcinogen 8:5 3. Greenhough A, Smartt HJ, Moore AE, Roberts HR, Williams AC, Paraskeva C, Kaidi A (2009) The COX-2/PGE2 pathway: key roles in the hallmarks of cancer and adaptation to the tumour microenvironment. Carcinogenesis 30:377–386 4. Yuan CJ, Mandal AK, Zhang Z, Mukherjee AB (2000) Transcriptional regulation of cyclooxygenase-2 gene expression: novel effects of nonsteroidal anti-inflammatory drugs. Cancer Res 60:1084–1091 5. Wendum D, Comperat E, Boëlle PY, Parc R, Masliah J, Trugnan G, Fléjou JF (2005) Cytoplasmic phospholipase A2 alpha overexpression in stromal cells is correlated with angiogenesis in human colorectal cancer. Mod Pathol 18:212–220 6. Panel V, Boëlle PY, Ayala-Sanmartin J, Jouniaux AM, Hamelin R, Masliah J, Trugnan G, Fléjou JF, Wendum D (2006) Cytoplasmic phospholipase A2 expression in human colon adenocarcinoma is correlated with cyclooxygenase-2 expression and contributes to prostaglandin E2 production. Cancer Lett 243:255–263 7. Nevalainen TJ, Haapanen TJ (1993) Distribution of pancreatic (group I) and synovial-type (group II) phospholipases A2 in human tissues. Inflammation 17:453–464 8. Nevalainen TJ, Kortesuo PT, Rintala E, Märki F (1992) Immunochemical detection of group I and group II phospholipases A2 in human serum. Clin Chem 38:1824–1829 9. Rintala EM, Nevalainen TJ (1993) Group II phospholipase A2 in sera of febrile patients with microbiologically or clinically documented infections. Clin Infect Dis 17:864–870 10. Graff JR, Konicek BW, Deddens JA, Chedid M, Hurst BM, Colligan B, Neubauer BL, Carter HW, Carter JH (2001) Expression of group IIa secretory phospholipase A2 increases with prostate tumor grade. Clin Cancer Res 7:3857–3861 11. Jiang J, Neubauer BL, Graff JR, Chedid M, Thomas JE, Roehm NW, Zhang S, Eckert GJ, Koch MO, Eble JN, Cheng L (2002) Expression of group IIA secretory phospholipase A2 is elevated in

Virchows Arch (2010) 457:659–667

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

prostatic intraepithelial neoplasia and adenocarcinoma. Am J Pathol 160:667–671 Mirtti T, Laine VJ, Hiekkanen H, Hurme S, Rowe O, Nevalainen TJ, Kallajoki M, Alanen K (2009) Group IIA phospholipase A as a prognostic marker in prostate cancer: relevance to clinicopathological variables and disease-specific mortality. APMIS 117:151–161 Leung SY, Chen X, Chu KM, Yuen ST, Mathy J, Ji J, Chan AS, Li R, Law S, Troyanskaya OG, Tu IP, Wong J, So S, Bostein D, Brown PO (2002) Phospholipase A2 group IIA expression in gastric adenocarcinoma is associated with prolonged survival and less frequent metastasis. Proc Natl Acad Sci USA 99:16203–16208 Wendum D, Svreck M, Rigau V, Boëlle PY, Sebbagh N, Parc R, Masliah J, Trugnan G, Fléjou JF (2003) COX-2, inflammatory secreted PLA2, and cytoplasmic PLA2 protein expression in small bowel adenocarcinomas compared with colorectal adenocarcinomas. Mod Pathol 16:130–136 Buhmeida A, Bendardaf R, Hilska M, Laine J, Collan Y, Laato M, Syrjänen K, Pyrhönen S (2009) PLA2 (group IIA phospholipase A2) as a prognostic determinant in stage II colorectal carcinoma. Ann Oncol 20:1230–1235 Nevalainen TJ, Grönroos JM, Kallajoki M (1995) Expression of group II phospholipase A2 in the human gastrointestinal tract. Lab Invest 72:201–208 Hong KH, Bonventre JC, O’Leary E, Bonventre JV, Lander ES (2001) Deletion of cytosolic phospholipase A(2) suppresses Apc (Min)-induced tumorigenesis. Proc Natl Acad Sci USA 98:3935– 3939 Edhemovic I, Snoj M, Kljun A, Golouh R (2001) Immunohistochemical localization of group II phospholipase A2 in the tumours and mucosa of the colon and rectum. Eur J Surg Oncol 27:545–548 Kennedy BP, Soravia C, Moffat J, Xia L, Hiruki T, Collins S (1998) Overexpression of the non-pancreatic secretory group II PLA2 messenger RNA and protein in colorectal adenomas from familial adenomatous polyposis patients. Cancer Res 58:500–503 Haapamäki MM, Grönroos JM, Nurmi H, Alanen K, Kallajoki M, Nevalainen TJ (1997) Gene expression of group II phospholipase A2 in intestine in ulcerative colitis. Gut 40:95–101 Haapamäki MM, Grönroos JM, Nurmi H, Alanen K, Nevalainen TJ (1999) Gene expression of group II phospholipase A2 in intestine in Crohn’s disease. Am J Gastroenterol 94:713–720 Minami T, Tojo H, Shinomura Y, Matsuzawa Y, Okamoto M (1994) Increased group II phospholipase A2 in colonic mucosa of patients with Crohn’s disease and ulcerative colitis. Gut 35:1593–1598 Minami T, Shinomura Y, Miyagawa J, Tojo H, Okamoto M, Matsuzawa Y (1997) Immunohistochemical localization of group II phospholipase A2 in colonic mucosa of patients with inflammatory bowel disease. Am J Gastroenterol 92:289–292 Nevalainen TJ, Märki F, Kortesuo PT, Grütter MG, Di Marco S, Schmitz A (1993) Synovial type (group II) phospholipase in cartilage. J Rheumatol 20:325–330 Grönroos JO, Laine VJO, Nevalainen TJ (2002) Bactericidal group IIA phospholipase A2 in serum of patients with bacterial infections. J Infect Dis 185:1767–1772 Nevalainen TJ, Eerola LI, Rintala E, Laine VJ, Lambeau G, Gelb MH (2005) Time-resolved fluoroimmunoassays of the complete set of secreted phospholipases A2 in human serum. Biochim Biophys Acta 1733:210–223 Laine VJ, Grass DS, Nevalainen TJ (1999) Protection by group II phospholipase A2 against Staphylococcus aureus. J Immunol 162:7402–7408 Di Marco S, Märki F, Hofstetter H, Schmitz A, van Oostrum J, Grütter MG (1992) Purification, analysis, and enzymatic activity of recombinant human synovial fluid phospholipase A2 and N-terminal variants. J Biochem 112:350–354

667 29. Cormier RT, Hong KH, Halberg RB, Hawkins TL, Richardson P, Mulherkar R, Dove WF, Lander ES (1997) Secretory phospholipase Pla2g2a confers resistance to intestinal tumorigenesis. Nat Genet 17:88–91 30. Cormier RT, Bilger A, Lillich AJ, Halberg RB, Hong KH, Gould KA, Borenstein N, Lander ES, Dove WF (2000) The Mom1AKR intestinal tumor resistance region consists of Pla2g2a and a locus distal to D4Mit64. Oncogene 19:3182–3192 31. Belinsky GS, Rajan TV, Saria EA, Giardina C, Rosenberg DW (2007) Expression of secretory phospholipase A2 in colon tumor cells potentiates tumor growth. Mol Carcinog 46:106– 116 32. MacPhee M, Chepenik KP, Liddell RA, Nelson KK, Siracusa LD, Buchberg AM (1995) The secretory phospholipase A2 gene is a candidate for the Mom1 locus, a major modifier of ApcMininduced intestinal neoplasia. Cell 81:957–966 33. Dietrich WF, Lander ES, Smith JS, Moser AR, Gould KA, Luongo C, Borenstein N, Dove W (1993) Genetic identification of Mom-1, a major modifier locus affecting Min-induced intestinal neoplasia in the mouse. Cell 75:631–639 34. Johnson LK, Frank S, Vades P, Pruzanski W, Lusis AJ, Seilhamer JJ (1990) Localization and evolution of two human phospholipase A2 genes and two related genetic elements. Adv Exp Med Biol 275:17–34 35. Leister I, Weith A, Brüderlein S, Cziepluch C, Kangwanpong D, Schlag P, Schwab M (1990) Human colorectal cancer: high frequency of deletions at chromosme 1p35. Cancer Res 50:7232– 7235 36. Riggins GJ, Markowitz S, Wilson JK, Vogelstein B, Kinzler KW (1995) Absence of secretory phospholipase A2 gene alterations in human colorectal cancer. Cancer Res 55:5184–5186 37. Praml C, Amler LC, Dihlmann S, Finke LH, Schlag P, Schwab M (1998) Secretory type II phospholipase A2 (PLA2G2A) expression status in colorectal carcinoma derived cell lines and in normal colonic mucosa. Oncogene 17:2009–2012 38. Nevalainen TJ, Haapanen TJ, Rajala P, Ekfors T (1998) Expression of group IIA phospholipase A2 in Paneth cells of an adenoma of the rectum. A case report. APMIS 106:780– 784 39. Fijneman RJ, Cormier RT (2008) The roles of sPLA2-IIA (Pla2g2a) in cancer of the small and large intestine. Front Biosci 13:4144–4174 40. Cao Y, Pearman AT, Zimmerman GA, McIntyre TM, Prescott SM (2000) Intracellular unesterified arachidonic acid signals apoptosis. Proc Natl Acad Sci USA 97:11280–11285 41. Ilsley JN, Nakanishi M, Flynn C, Belisnky GS, De Guise S, Adib JN, Dobrowsky RT, Bonventre JV, Rosenberg DW (2005) Cytoplasmic phospholipase A2 deletion enhances colon tumorigenesis. Cancer Res 65:2636–2643 42. Pruzanski W, Stefanski E, Vadas P, Kennedy BP, van den Bosch H (1998) Regulation of the cellular expression of secretory and cytosolic phospholipases A2, and cyclooxygenase-2 by peptide growth factors. Biochim Biophys Acta 1403:47–56 43. Szkaradkiewicz A, Marciniak R, Chudzicka-Strugala I, Wasilewska A, Drews M, Majewski P, Karpinski T, Zwozdziak B (2009) Proinflammatory cytokines and IL-10 in inflammatory bowel disease and colorectal cancer patients. Arch Immunol Ther Exp 57:291–294 44. Zlobec I, Lugli A (2009) Invasive front of colorectal cancer: dynamic interface of pro-/anti-tumor factors. World J Gastroenterol 15:5898–5906 45. Tribler L, Jensen LT, Jorgensen K, Brünner N, Gelb MH, Nielsen HJ, Jensen SS (2007) Increased expression of group IIA and X secretory phospholipase A2 in peritumoural versus central colon carcinoma tissue. Anticancer Res 27:3179–3185