Possible Involvement of Intracellular Calcium-Independent Phospholipase A2 in the Release of Secretory Phospholipases from Mast Cells—Increased Expression in Ileal Mast Cells of Crohn’s Disease

Increased activity of secretory phospholipases A2 (sPLA2) type-II was previously observed in ileum of Crohn’s disease (CD). Our aims were to explore the involvement of calcium-independent (i)PLA2β in the release of sPLA2s from the human mast cell (MC) line (HMC-1) and investigate expressions of cytosolic (c)PLA2α, iPLA2β, sPLA2-IIA and sPLA2-V in MCs of CD ileum. The release of sPLA2 was investigated in HMC-1 by immunocytochemistry and ELISA. The expression intensities of PLA2s in mucosal MCs, and the proportion of PLA2-positive MCs, were investigated in normal ileum and in ileum from patients with CD by immunohistochemistry. The calcium ionophore-stimulated release of sPLA2-IIA and sPLA2-V from HMC-1 was reduced by the iPLA2-inhibitor bromoenol lactone. All four PLA2s were detectable in mucosal MCs, both in normal ileum and in CD, but the proportion of iPLA2β-containing mucosal MCs and the expression intensity of sPLA2-IIA was increased in CD. Results indicate that iPLA2β is involved in the secretion of sPLA2s from HMC-1, and suggest that iPLA2β-mediated release of sPLA2 from intestinal MCs may contribute to CD pathophysiology. Ex vivo studies on isolated mucosal mast cells are however needed to clarify the precise role of MC PLA2s in the inflammatory processes of CD.


Introduction
Mediators released from activated intestinal mast cells (MCs) have shown to be of pathophysiological significance in Crohn's disease (CD) [1,2], for instance, by promoting intestinal fibrosis or by decreasing the mucosal barrier against immune-activating antigens [3][4][5]. However, MC mediators do not necessarily have only detrimental effects in CD, since intestinal MCs also are thought to have a role in host defense against bacterial, viral and parasitic agents [5]. MC mediators of potential relevance for inflammatory conditions include, for instance, eicosanoids [6] and other lipid mediators (i.e., platelet-activating factor and lysophospholipids) generated upon activation of one or several isoforms of the phospholipase A 2 superfamily (PLA 2 ) [7]. The expression of different PLA 2 isoenzymes in MCs of the human intestinal mucosa is still unknown, both in the normal intestine and in CD. exclusion assay or by a MTT toxicology assay and was not affected during the experimental conditions used in this study.
To investigate if the expressions of iPLA 2 β and cPLA 2 α could be further increased upon activation of the MCs, 5 × 10 5 HMC-1 were incubated for 48h in 1 mL culture medium with or without (controls) 25 ng/mL of TNFα. TNFα is of fundamental importance in inflammatory conditions such as CD [30], and may influence PLA 2 expression and activity [31,32]. The expressions of cPLA 2 α, iPLA 2 β, sPLA 2 -IIA and sPLA 2 -V were after incubation analyzed by Reverse Transcriptase-PCR and immunocytochemical staining.

Reverse Transcriptase-PCR of PLA 2 s
Total RNA was extracted from HMC-1 using Ultraspec™-II RNA Isolation System (Nordic Biosite, Täby, Sweden). One µg of total RNA was converted into cDNA using Omniscript ® Reverse Transcription RT Kit (Qiagen, Solna, Sweden) according to the manufacturer's instructions, and amplified using PuRe Taq RTG PCR beads (GE Healthcare, Buckinghamshire, UK) and primers (Life Technology Ltd., Paisley, UK). Due to a high expression, the cDNA for sPLA 2 -IIA had to be diluted 10× before subjected to conventional Reverse Transcriptase-PCR. Primers and running schedules used in PCR are summarized in Table 1. The final PCR products were loaded on 1.5% agarose gels, and identified as previously described [33]. C, 20 s; 72 • C, 45 s a The first cycle was preceded by an initial denaturation step at 94 • C for 5 min, and the last cycle was followed by an elongation step at 72 • C for 5 or 7 (cPLA 2 ) min.

Immunocytochemical Staining of PLA 2 s
HMC-1 were smeared on poly-L-lysine coated glass (Sigma) as previously described [33]. The samples were fixed in ice-cold acetone for 5 min at -20 • C and then blocked with 50% of serum in PBS for 1h at room temperature (RT). The samples were incubated with either 1:50 mouse monoclonal FITC-conjugated anti-human sPLA 2 -V antibody (Santa Cruz, Dallas, Texas, USA) or 1:200 mouse monoclonal anti-human sPLA 2 -IIA (Cayman Chemical Co, Ann Arbor, MI, USA) for 16h at 4 • C. Biotin-conjugated 1:250 secondary rabbit anti-mouse (DakoCytomation, Glostrup, Denmark) was applied to samples with sPLA 2 -IIA antibody for 1h at RT and then 1:100 FITC-conjugated streptavidin (DakoCytomation) for 30 min at RT. In addition, samples were incubated with either 1:100 Alexa-488 conjugated mouse monoclonal anti-human cPLA 2 α (Santa Cruz) or 1:250 rabbit polyclonal anti-human iPLA 2 β (Cayman) for 16h at 4 • C. FITC-conjugated secondary antibody goat anti-rabbit (Jackson ImmunoResearch Laboratories Inc, West Grove, PA, USA) was applicated at a dilution of 1:400. The slides were mounted with Vectashield ® mounting medium with propidium iodide (Vector Laboratories Inc, Burlingame, CA, USA). Negative controls without primary antibodies or with a FITC-conjugated isotype matched irrelevant antibody (Santa Cruz) were included in all experiments.

Release of Fatty Acids
To further explore the involvement of cPLA 2 in AA-mobilization in activated HMC-1, 14 C-AA labelled cells were stimulated with the frequently used MC activator calcium ionophore A23187 [10,[34][35][36][37][38], in the presence and absence of known enzyme inhibitors. Cells were suspended in 25 mL supplemented medium with 0.1% fatty-acid free bovine serum albumin (Sigma) and labelled for 16 h with 0.1 µCi [1-14 C]AA (New England Nuclear, Perkin Elmer, Wellesley, MA, USA) per 5 × 10 5 cells, before washed two times with PBS supplemented with 0.1% fatty-acid free bovine serum albumin [37]. Labelled cells (5 × 10 5 cells in a final volume of 2.7 mL) were then treated for 4h with 2 µM of the calcium ionophore A23187 (Sigma) only, or in combination with 200 nM of the protein kinase C activator phorbol myristate acetate (PMA) (Sigma). The combination of A23187 and PMA has previously shown to induce a synergistic release of AA in other cell systems, an effect attributed to an increased activation of cPLA 2 [37,39,40].
As an attempt to investigate the relative contribution of cPLA 2 and iPLA 2 in the A23187-stimulated AA release, cells were pre-incubated with the combined cPLA 2 and iPLA 2 inhibitor methyl arachidonyl fluoro-phosphonate (MAFP) (Sigma) [41], or the specific iPLA 2 inhibitor bromoenol lactone (BEL) (Sigma) [41]. Cells were pre-treated for 30 min with MAFP (0 µM, 10 µM or 25 µM) or BEL (0 µM, 10 µM or 25 µM) prior to incubation with A23187 (2 µM) for an additional 4 h. All treatments with stimulators and inhibitors were performed in the absence of serum but in the presence of 0.1% fatty acid-free bovine serum albumin. The amount of 14 C-AA released into the culture medium was analyzed by liquid scintillation counting. The inhibitors were added 30 min prior to adding the stimulators. To evaluate the AA specificity of the involved PLA 2 , a comparable stimulation of 14 C-oleic acid (OA) (Perkin Elmer) labelled cells was performed.
To investigate if iPLA 2 is involved in the ionophore-stimulated sPLA 2 secretion in HMC-1, 25 µM of the inhibitor BEL was added 30 min before A23187, when appropriate. All treatments with stimulators and inhibitors were performed in the absence of serum. Cells were centrifuged and the medium was collected. The β-hexosaminidase activity was determined as previously described [43], and the amount of sPLA 2 -IIA was determined by sandwich-ELISA according to the manufacturer's instructions (Cayman).
The amounts of remaining sPLA 2 -IIA and sPLA 2 -V in stimulated cells were investigated by immunocytochemical staining as described above. Due to its low basal expression, sPLA 2 -V had to be upregulated by 25 ng/mL TNFα (Sigma) for 48h prior to stimulation with A23187 in this set of experiments.

Patients
Specimens from ileum were achieved during surgery at Linköping University Hospital from 5 patients with ileal CD and 5 patients with colonic cancer, as non-IBD controls. The CD patients constituted of 3 men and 2 women with a median age of 53 years (range 43-65) and disease duration of 15 years (range 9-28). According to the Montreal classification, all patients had an active disease, however, tissue obtained for analyzes were dissected from mild-inflamed ileum. The non-IBD control group constituted of microscopically normal ileal specimens from 3 men and 2 women with a median age of 71 years (range 62-87). None of the patients within the non-IBD control group had received pre-operative chemo-or radiotherapy or had signs of generalized disease. The study was approved by the Committee of Human Ethics, Linköping (ethical number 02-154, 09/04/2002) and all included subjects gave their informed written consent before the study was initiated.

Preparation of Ileal Tissues
Surgical ileal specimens from patients with CD and non-IBD controls were immediately after division of the ileocolic artery, put in ice-cold oxygenated Krebs buffer and specimens were stripped of external muscle and myenteric plexus, as previously described [44]. Segments of ileal mucosa were fixed in 4% buffered formaldehyde in PBS for 24h in 4 • C, embedded in paraffin and sectioned to a thickness of 5 µm.

Immunohistochemical Staining of PLA 2 s
Slides with sections were hydrated according to standard procedures followed by incubation for 10 min with background sniper (Histolab, Gothenburg, Sweden). After washed in PBS, slides were incubated for 16h at 4 • C with 1:200 mouse monoclonal-anti-human MC tryptase antibody (Santa Cruz) in combination with either 1:50 rabbit polyclonal-anti-human sPLA 2 -IIA (Novus Biologicals, Bio-Techne, Abingdon, UK), 1:50 rabbit polyclonal-anti-human sPLA 2 -V (Bio-Techne), 1:50 goat polyclonal cPLA 2 β antibody (Santa Cruz), or 1:50 rabbit polyclonal-anti-human iPLA 2 β (Santa Cruz). Slides were rinsed and incubated with secondary antibodies (MC: 1:4 ready to use Alexa Fluor 594-conjugated-goat-anti-mouse (Invitrogen, Oregon, USA); cPLA 2 β: 1:200 Alexa Fluor 488-conjugated donkey-anti-goat (Life technologies); iPLA 2 β, sPLA 2 -IIA, sPLA 2 -V: 1:200 Alexa Fluor 488-conjugated donkey-anti-rabbit (Life technologies) for 1h at RT. After repeated rinsing, slides were mounted with Prolong ® Gold Antifade with DAPI (Life Technologies) and evaluated in a Nikon E800 fluorescence microscope connected to software NIS elements (Nikon Instruments Inc. Tokyo, Japan) in a blinded fashion by two independent researchers. Three sections per individual were stained for each double-staining, and negative controls with primary antibodies excluded were included in all experiments. The total number of MCs co-localizing with the different PLA 2 s were manually quantified at 600× magnification. The intensities of the different PLA 2 -stainings were measured by Image J Fiji software (National Institutes of Health, Bethesda, MD, USA). Approximately 6-8 area-units per section were counted. All area-units were of the same size and only area-units that were fully covered by tissue were used.

Statistical Analysis
Data were analyzed using the GraphPad Prism Software (GraphPad Software Inc., CA, USA). Parametric data are expressed as mean ± SEM and depending on the experimental layout, statistical analyses were undertaken with one-way ANOVA, repeated measures ANOVA, and Bonferroni post-test. Non-parametric data are given as median (25th-75th interquartile range) and comparisons between groups were done with Kruskal-Wallis and Mann-Whitney U tests.

iPLA 2 is the Predominating High-Molecular-Weight PLA 2 Expressed by HMC-1
HMC-1 was found to have a basal expression of both iPLA 2 β mRNA ( Figure 1A) and iPLA 2 β protein ( Figure 1B). In contrast, cPLA 2 α revealed no basal mRNA expression ( Figure 1A), and the protein expression was very low ( Figure 1B). Treatment with 25 ng/mL TNFα for 48 h did neither affect the iPLA 2 β mRNA expression ( Figure 1A) nor the iPLA 2 β protein expression ( Figure 1B). On the contrary, TNFα stimulation had an inconsistent effect on the cPLA 2 α expression, increasing the mRNA stimulation had an inconsistent effect on the cPLA 2 α expression, increasing the mRNA expression ( Figure 1A) without affecting the protein expression ( Figure 1B).

Secretory PLA2-IIA and V are Expressed by HMC-1
Immunostaining revealed a basal expression of sPLA2-IIA mRNA ( Figure 2A) and sPLA2-IIA protein ( Figure 2B) in HMC-1. Neither the mRNA nor the protein expression was affected in cells stimulated with 25 ng/mL TNFα for 48h (Figure 2A,B). HMC-1 were also found to have a basal expression of sPLA2-V mRNA ( Figure 2A) and sPLA2-V protein ( Figure 2B), although, less pronounced as compared to corresponding expressions of sPLA2-IIA (Figure 2A,B). However, in contrast to sPLA2-IIA, the expressions of sPLA2-V mRNA and proteins were increased in TNFαstimulated cells (Figure 2A,B). Reverse Transcriptase-PCR analysis; the PCR products were identified as sPLA2-IIA (238 bp), sPLA2-V (559 bp) or 18S rRNA (531 bp). Note that the cDNA for sPLA2-IIA was diluted ten times compared to the cDNA for sPLA2-V. Samples are two representatives out of seven independent runs. (B) Immunocytochemical analysis. Green staining is for sPLA2-IIA or sPLA2-V and red staining is for visualization of cell nuclei (magnification × 600). Results are representative for three independent experiments.

Secretory PLA2-IIA and V are Expressed by HMC-1
Immunostaining revealed a basal expression of sPLA2-IIA mRNA ( Figure 2A) and sPLA2-IIA protein ( Figure 2B) in HMC-1. Neither the mRNA nor the protein expression was affected in cells stimulated with 25 ng/mL TNFα for 48h (Figure 2A,B). HMC-1 were also found to have a basal expression of sPLA2-V mRNA ( Figure 2A) and sPLA2-V protein ( Figure 2B), although, less pronounced as compared to corresponding expressions of sPLA2-IIA (Figure 2A,B). However, in contrast to sPLA2-IIA, the expressions of sPLA2-V mRNA and proteins were increased in TNFαstimulated cells (Figure 2A,B). Reverse Transcriptase-PCR analysis; the PCR products were identified as sPLA2-IIA (238 bp), sPLA2-V (559 bp) or 18S rRNA (531 bp). Note that the cDNA for sPLA2-IIA was diluted ten times compared to the cDNA for sPLA2-V. Samples are two representatives out of seven independent runs. (B) Immunocytochemical analysis. Green staining is for sPLA2-IIA or sPLA2-V and red staining is for visualization of cell nuclei (magnification × 600). Results are representative for three independent experiments.  Stimulation with calcium ionophore A23187 caused an obvious time-dependent increase in the release of radioactivity from 14 C-AA-labelled cells ( Figure 3A). The increase was discernible after 1 h but not significant until 4 h of treatment compared to controls at each time point ( Figure 3A). cPLA 2 α is generally regarded as the main regulator of cellular AA mobilization [8], however, a comparable release of radioactivity also from A23187-stimulated 14 C-OA-labeled HMC-1 clearly demonstrated that the ionophore-stimulated PLA 2 activity was not AA-specific ( Figure 3B). Stimulation with the combination of A23187 and the protein kinase C activator PMA showed that PMA had no further impact on the A23187-stimulated AA release, neither at 30 min (data not shown) nor at 4 h ( Figure 3C). release of radioactivity from 14 C-AA-labelled cells ( Figure 3A). The increase was discernible after 1 h but not significant until 4 h of treatment compared to controls at each time point ( Figure 3A). cPLA2α is generally regarded as the main regulator of cellular AA mobilization [8], however, a comparable release of radioactivity also from A23187-stimulated 14 C-OA-labeled HMC-1 clearly demonstrated that the ionophore-stimulated PLA2 activity was not AA-specific ( Figure 3B). Stimulation with the combination of A23187 and the protein kinase C activator PMA showed that PMA had no further impact on the A23187-stimulated AA release, neither at 30 min (data not shown) nor at 4 h ( Figure  3C).

cPLA2α is not Involved in Calcium
The PLA2-inhibitors MAFP (general) and BEL (iPLA2-specific) were found to reduce the A23187stimulated AA release in a dose-dependent manner and at a comparable extent ( Figure 4A,B).

iPLA2 is involved in the A23187-stimulated release of sPLA2-IIA and sPLA2-V from HMC-1
Stimulation with A23187 induced degranulation of the HMC-1 cells in a dose-dependent manner, demonstrated as an increased β-hexosaminidase release ( Figure 5A). Simultaneously, The PLA 2 -inhibitors MAFP (general) and BEL (iPLA 2 -specific) were found to reduce the A23187-stimulated AA release in a dose-dependent manner and at a comparable extent ( Figure 4A,B). Stimulation with calcium ionophore A23187 caused an obvious time-dependent increase in the release of radioactivity from 14 C-AA-labelled cells ( Figure 3A). The increase was discernible after 1 h but not significant until 4 h of treatment compared to controls at each time point ( Figure 3A). cPLA2α is generally regarded as the main regulator of cellular AA mobilization [8], however, a comparable release of radioactivity also from A23187-stimulated 14 C-OA-labeled HMC-1 clearly demonstrated that the ionophore-stimulated PLA2 activity was not AA-specific ( Figure 3B). Stimulation with the combination of A23187 and the protein kinase C activator PMA showed that PMA had no further impact on the A23187-stimulated AA release, neither at 30 min (data not shown) nor at 4 h ( Figure  3C).
The PLA2-inhibitors MAFP (general) and BEL (iPLA2-specific) were found to reduce the A23187stimulated AA release in a dose-dependent manner and at a comparable extent ( Figure 4A,B).

iPLA2 is involved in the A23187-stimulated release of sPLA2-IIA and sPLA2-V from HMC-1
Stimulation with A23187 induced degranulation of the HMC-1 cells in a dose-dependent manner, demonstrated as an increased β-hexosaminidase release ( Figure 5A). Simultaneously,  Stimulation with A23187 induced degranulation of the HMC-1 cells in a dose-dependent manner, demonstrated as an increased β-hexosaminidase release ( Figure 5A). Simultaneously, A23187 caused a dose-dependent release of sPLA 2 -IIA, as detected by ELISA ( Figure 5B) and further confirmed by immunocytochemical visualization (Figure 5C). In addition, A23187 caused a dose-dependent release of sPLA 2 -V, as visualized by immunocytochemistry ( Figure 5C). Due to the low basal expression of sPLA 2 -V, the immunocytochemistry was performed after up-regulation of sPLA 2 -V with TNFα, as illustrated in Figure 2B.
Pre-incubation with the iPLA 2 -specific inhibitor BEL prior to A23187 stimulation, diminished both the degranulation of HMC-1 ( Figure 6A) and the release of sPLA 2 -IIA and sPLA 2 -V ( Figure 6B,C). A23187 caused a dose-dependent release of sPLA2-IIA, as detected by ELISA ( Figure 5B) and further confirmed by immunocytochemical visualization ( Figure 5C). In addition, A23187 caused a dosedependent release of sPLA2-V, as visualized by immunocytochemistry ( Figure 5C). Due to the low basal expression of sPLA2-V, the immunocytochemistry was performed after up-regulation of sPLA2-V with TNFα, as illustrated in Figure 2B. Green staining is for sPLA2-IIA or sPLA2-V and red staining is for visualization of cell nuclei (magnification × 600). Note that the expression of sPLA2-V had to be upregulated by TNFα, as described in Figure 2A and B. **p < 0.01, ***p < 0.001 vs. controls. Data from three independent experiments.
Pre-incubation with the iPLA2-specific inhibitor BEL prior to A23187 stimulation, diminished both the degranulation of HMC-1 ( Figure 6A) and the release of sPLA2-IIA and sPLA2-V ( Figure  6B,C). Green staining is for sPLA2-IIA or sPLA2-V and red staining is for visualization of cell nuclei (magnification × 600). Note that the expression of sPLA2-V had to be upregulated by TNFα, as described in Figure 2A and B, to be illustrated. ** p < 0.01, *** p < 0.001 vs. controls. Data from three independent experiments.

Mucosal MCs express all four PLA2 isoforms investigated
Cells positively stained with the MC tryptase antibody were found in both control and CD ileal mucosa. MCs from controls and CD patients were found to express all four PLA2 isoforms investigated, i.e., the two intracellular high molecular isoforms, cPLAα and iPLA2β, and the two secretory isoforms, sPLA2-IIA and sPLA2-V ( Figure 7A-D). Both intracellular and secretory PLA2s Green staining is for sPLA 2 -IIA or sPLA 2 -V and red staining is for visualization of cell nuclei (magnification × 600). Note that the expression of sPLA 2 -V had to be upregulated by TNFα, as described in Figure 2A and B, to be illustrated. ** p < 0.01, *** p < 0.001 vs. controls. Data from three independent experiments.

Mucosal MCs Express all four PLA 2 Isoforms Investigated
Cells positively stained with the MC tryptase antibody were found in both control and CD ileal mucosa. MCs from controls and CD patients were found to express all four PLA 2 isoforms investigated, i.e., the two intracellular high molecular isoforms, cPLAα and iPLA 2 β, and the two secretory isoforms, sPLA 2 -IIA and sPLA 2 -V ( Figure 7A-D). Both intracellular and secretory PLA 2 s were also found on cells not positive for MC tryptase, and in addition, there were MCs present not expressing any PLA 2 . MC and PLA2 expressions were quantified manually at 600× magnification and results are given as median (25th-75th percentile). Red = MCs, Green = PLA2, Blue = DAPI, nuclei staining. *p < 0.05 vs. controls.

Increased Proportion of iPLA2β-Containing Mucosal MCs of CD Ileum
For the intracellular forms there was a higher percentage of MCs expressing iPLA2β in CD compared to controls, p < 0.05 ( Figure 7A), but no significant difference in expressions of cPLA2α, p = 0.11 ( Figure 7B). Measurements of intensity (Median (25th-75th percentile)) showed no difference between the groups (iPLA2β: CD 13.

Discussion
The present study demonstrates, for the first time, that human ileal MCs of normal and CD mucosa contain the sPLA2 isoforms sPLA2-IIA and sPLA2-V, as well as the intracellular high molecular isoforms cPLA2α and iPLA2β. In addition, studies on the human MC cell line HMC-1 demonstrated that iPLA2β might have a role in the release of sPLA2-IIA and sPLA2-V. Thus, our MC and PLA 2 expressions were quantified manually at 600× magnification and results are given as median (25th-75th percentile). Red = MCs, Green = PLA 2 , Blue = DAPI, nuclei staining. *p < 0.05 vs. controls.

Increased Proportion of iPLA 2 β-Containing Mucosal MCs of CD Ileum
For the intracellular forms there was a higher percentage of MCs expressing iPLA 2 β in CD compared to controls, p < 0.05 ( Figure 7A), but no significant difference in expressions of cPLA 2 α, p = 0.11 ( Figure 7B). Measurements of intensity (Median (25th-75th percentile)) showed no difference between the groups (iPLA 2 β:   0-50.1)). In contrast, intensity measurements showed a significantly higher expression intensity of sPLA 2 -IIA in MCs of CD patients compared to controls, p < 0.05 ( Figure 7C), but no difference between groups in the expression of sPLA 2 -V ( Figure 7D).

Discussion
The present study demonstrates, for the first time, that human ileal MCs of normal and CD mucosa contain the sPLA 2 isoforms sPLA 2 -IIA and sPLA 2 -V, as well as the intracellular high molecular isoforms cPLA 2 α and iPLA 2 β. In addition, studies on the human MC cell line HMC-1 demonstrated that iPLA 2 β might have a role in the release of sPLA 2 -IIA and sPLA 2 -V. Thus, our results point to a possible role of iPLA 2 β in the release of sPLA 2 s from MCs of the human ileal mucosa. sPLA 2 -IIA and V are frequently associated with inflammatory conditions [14,23]. Even though sPLA 2 -II is known to be present in the CD intestine [45,46], including submucosal MCs [18], no studies on sPLA 2 -V expressions in CD intestine, or sPLA 2 -II expressions in intestinal mucosal MCs, have been reported. We previously demonstrated [46] that the distal ileal mucosa is rich in PLA 2 -II mRNA and that the expression of this mRNA and the corresponding enzyme activity accompanies recurrent new ileal inflammation after ileocolonic resection for CD. However, the cells responsible for this increased expression and activity have previously not been identified. In the present study we demonstrated that the expression of sPLA 2 -IIA was higher in MCs from ileal CD mucosa compared to MCs from control patients. Further, we found that the proportion of iPLA 2 β-expressing mucosal MCs was increased in CD ileum compared with controls; i.e., among all MCs present, more MCs expressed iPLA 2 β in ileum from CD patients. These findings suggest that MCs may contribute to the increased sPLA 2 -II expression and activity in CD ileum [46].
Although iPLA 2 β is generally thought to be involved in various cellular and pathological conditions [9], its expression and role in the human intestine has never been investigated. However, our results on HMC-1 support previous findings demonstrating a possible role for iPLA 2 β in MC exocytosis [10], and one might speculate that the increased proportion of iPLA 2 β-expressing MCs found in CD may reflect a greater release of various MC mediators in the CD intestine. Intestinal barrier dysfunction, leading to increased transfer of luminal bacteria to the lamina propria is thought to be a factor of importance in the pathogenesis of CD [47]. Considering the proposed protective role of iPLA 2 β in the intestine [11,12], it is tempting to speculate that iPLA 2 β might have a role in releasing bactericidal sPLA 2 s from MCs in the intestinal mucosa. Indeed, several sPLA 2 s, in particular sPLA 2 IIA, are known to have antibacterial activities [7,13,21].
Considering the proposed species differences with regard to both MC characteristics [48] and PLA 2 expression [14] a human experimental MC cell model was used for the studies on sPLA 2 release. Although various aspects of PLA 2 s have been extensively studied in rodent MCs [10,[15][16][17]38,[49][50][51], not much is known about the expression and regulation of these enzymes in MCs of human origin. We chose to work with the human MC cell line HMC-1 [29] because it has been frequently used for studies on various aspects of MC biology, and this cell line has been reported to produce several different eicosanoids upon stimulation with calcium ionophore [35,36]. However, the PLA 2 s responsible for generating the required free AA is not known, and studies concerning the expression and activity of PLA 2 -enzymes of HMC-1 are still lacking. It was necessary thus to confirm the presence of intracellular and secretory PLA 2 s in this cell line before using it for studies on sPLA 2 release. Interestingly the HMC-1 was found to have a basal expression of iPLA 2 β protein, whereas the expression of cPLA 2 α was very low. Neither the protein nor the mRNA expression of iPLA 2 β was apparently affected by TNFα. This lack of effect of TNFα suggests that iPLA 2 β is not regulated by inflammatory agents in HMC-1, a finding well in line with the proposed role of iPLA 2 β as a homeostatic enzyme in cellular phospholipid metabolism [9]. In contrast, TNFα increased the mRNA but not the protein expression of cPLA 2 α. Thus, increasing the level of cPLA 2 α mRNA in HMC-1 seems not to per se induce translation into cPLA 2 α protein, but additional stimulators of translation seem to be needed. These findings are in line with a previous study [52], showing that transforming growth factor β-1 stimulates cPLA 2 gene expression in human intestinal MCs without affecting the level of cPLA 2 protein. The translation of gene expressions to protein levels is a multistep process and Schwanhausser et al. [53] has concluded that translational rate constants were the dominant factors in controlling protein levels, and that half-life of the proteins are highly involved in the translation as well. In addition to the findings of increased iPLA 2 β, the HMC-1 were found to have a basal expression of both sPLA 2 -V and sPLA 2 -IIA, which is in line with previous reports on rodent MCs [16] and human lung MCs [20]. However, when HMC-1 were stimulated with TNFα, both mRNA and protein expressions of sPLA 2 -V were increased, whereas the mRNA and protein levels of sPLA 2 -IIA was unaltered. Our results on HMC-1 are in line with previous studies showing that despite close similarities between group IIA and V [14], their expression and regulation may differ [54].
The expression of cPLA 2 α protein appears to be very low in HMC-1. Therefore, to clarify if cPLA 2 α activity is present in HMC1, the release of AA and OA was compared in A23187-stimulated cells. The calcium ionophore A23187 caused a marked elevation of fatty acid release from the HMC-1. This fatty acid release was not restricted to AA, and about equally reduced by the specific iPLA 2 inhibitor BEL [41] and the combined iPLA 2 and cPLA 2 inhibitor MAFP [41]. Also, the A23187-stimulated AA release was not augmented by the attempt to increase the cPLA 2 α activity by combined stimulation with PMA [37,39,40]. Taken together, these findings strongly suggest that one or several PLA 2 s, different from the AA specific cPLA 2 α, is accountable for the A23187-stimulated AA release in HMC-1. One possible candidate is iPLA 2 β, since the AA release was reduced by BEL and iPLA 2 is known to release AA in other cell systems [9,10]. However, BEL and MAFP reduced about 50% of the AA release induced by A23187, indicating contribution of one or several MAFP/BEL-insensitive PLA 2 s, for instance sPLA 2 s [14]. It was out of the scope of the present study to investigate in detail which particular PLA 2 s are involved in the AA release from HMC-1. However, our results may suggest a role for iPLA 2 and clearly indicate that the cPLA 2 α activity of HMC1 is very low and in line with the low cPLA 2 α protein levels found.
Whereas several studies have implicated a role for cPLA 2 and sPLA 2 in the release of AA from rodent MCs [6,15,17,50,51,55], only one study, so far, has reported involvement of iPLA 2 [10]. Indeed, A23187 was found to release radiolabeled AA from mouse bone marrow-derived MCs (BMMCs) and rat basophilic leukemia MCs (RBL 2H3) by an iPLA 2 -dependent mechanism [10], a finding in line with our results in HMC-1.
The mechanism of MC degranulation involves cellular events leading to an increased cytosolic Ca 2+ -concentration [42]. Evidently, we found that A23187 stimulates degranulation (i.e., stimulated the release of β-hexosaminidase) of HMC-1 and release of sPLA 2 . This is in line with a previous study on ionophore-stimulated BMMCs [34]. The A23187-stimulated release of sPLA 2 -IIA and sPLA 2 -V was reduced by the iPLA 2 inhibitor BEL, suggesting a role for iPLA 2 in the A23187-stimulated sPLA 2 release from HMC-1. Although BEL is known to inhibit degranulation of BMMCs and RBL 2H3 cells [10], and also to inhibit exocytosis in other cell types [25,28], this is, as far as we know, the first study suggesting a role for iPLA 2 in the regulation of sPLA 2 release. Indeed, our finding that BEL inhibited not only the A23187-stimulated release of sPLA 2 , but also the release of β-hexosaminidase, may indicate a role of iPLA 2 in MC degranulation and release of MC mediators in general.
Although the results of the present study suggest that iPLA 2 β is involved in the release of sPLA 2 s from A23187-stimulated cells, the precise mechanism by which iPLA 2 β is activated by A23187 has to be evaluated. However, one possible mechanism might be that depletion of calcium stores by A23187 results in displacement of inhibitory calmodulin from iPLA 2 [49].
Both iPLA 2 β [10, [25][26][27][28], and cPLA 2 α [8, 56,57] have been implicated in vesicle trafficking and exocytosis. However, due to the low (perhaps absent) cPLA 2 α activity of the HMC-1, it is not likely that cPLA 2 α is involved in the release of sPLA 2 s. Our finding that cPLA 2 α is expressed in human intestinal MCs may suggest, however, that also this intracellular PLA 2 might be involved in MC exocytosis in the human intestine. Clearly, further studies on MCs isolated directly from the human intestine are needed to evaluate the precise roles of iPLA 2 β and cPLA 2 α in the release of sPLA 2 s from MCs in the normal and inflamed human intestine.
Although our results suggest that iPLA 2 β is involved in the degranulation and release of sPLA 2 in HMC-1, this is not necessarily true for other experimental MC models or during other experimental settings. For example, a study on BMMCs [38] demonstrated, in contrast with a previous report [10], that iPLA 2 β is not involved in the release of β-hexosaminidase from these MCs. It is also worth mentioning that species differences among MCs may influences their behavior [48], and that it is unknown to what extent the role and regulation of a particular PLA 2 in rodent MCs correspond to its role and regulation in human MCs.
BEL is a widely used inhibitor of iPLA 2 , with limited effect on cPLA 2 and sPLA 2 [9,41]. Indeed, BEL is to date the only irreversible specific inhibitor of iPLA 2 available, however, BEL may have other unspecific side effects as well, resulting in cytotoxic effects [58]. In the present study, the viability of HMC-1 was routinely evaluated and no detrimental effect of BEL was found. Thus, it seems likely that iPLA 2 was the target of BEL in HMC-1. However, to verify this, further studies using gene silencing techniques are needed.

Conclusions
In conclusion, this study suggests that iPLA 2 β might be involved in the secretion of sPLA 2 s from HMC-1, suggesting that an iPLA 2 β-mediated release of sPLA 2 from intestinal MCs may contribute to increased sPLA 2 -II activity. Further, cPLA 2 α, iPLA 2 β, sPLA 2 -IIA and sPLA 2 -V are all present in mucosal MCs of both normal ileum and in the mild-inflamed ileum of CD. However, CD ileum possessed an increased proportion of iPLA 2 β-containing MCs. Taken together, results may suggest that iPLA 2 β may have a previously unrecognized role in human MCs, i.e., regulation of sPLA 2 secretion. However, further ex vivo studies are needed to confirm this and to evaluate the precise role of iPLA 2 β in the release of sPLA 2 s from isolated ileal MCs and its importance in the pathophysiology of CD.