Study of the Role of Cytosolic Phospholipase A2 Alpha in Eicosanoid Generation and Thymocyte Maturation in the Thymus

The thymus is a primary lymphoid organ, home of maturation and selection of thymocytes for generation of functional T-cells. Multiple factors are involved throughout the different stages of the maturation process to tightly regulate T-cell production. The metabolism of arachidonic acid by cyclooxygenases, lipoxygenases and specific isomerases generates eicosanoids, lipid mediators capable of triggering cellular responses. In this study, we determined the profile of expression of the eicosanoids present in the mouse thymus at different stages of thymocyte development. As the group IVA cytosolic phospholipase A2 (cPLA2α) catalyzes the hydrolysis of phospholipids, thereby generating arachidonic acid, we further verified its contribution by including cPLA2α deficient mice to our investigations. We found that a vast array of eicosanoids is expressed in the thymus, which expression is substantially modulated through thymocyte development. The cPLA2α was dispensable in the generation of most eicosanoids in the thymus and consistently, the ablation of the cPLA2α gene in mouse thymus and the culture of thymuses from human newborns in presence of the cPLA2α inhibitor pyrrophenone did not impact thymocyte maturation. This study provides information on the eicosanoid repertoire present during thymocyte development and suggests that thymocyte maturation can occur independently of cPLA2α.


Introduction
The thymus has a central role in the immune system as it supports the development, the differentiation and the selection of T-cells [1][2][3]. Thymic development of the T-cell precursors is finely regulated. Firstly, the T-cell precursors originating from the bone marrow enter in the thymus through the cortex. These immature T-cells, called thymocytes, differently express the T-cell receptor (TCR) co-receptors CD4 and CD8 at their surface, an indication of the T-cell maturation state. Owing to the lack of expression of CD4 and CD8 immediately after their mice and it was not reversed by the exogenous addition of PGE 2 (up to 10μM) [35]. Thus, whether prostanoids actually participate in thymocyte development remains unclear.
AA, which is mainly esterified at the sn-2 position of phospholipids, has to be released from the membrane phospholipids to be metabolized into eicosanoids. The availability of AA is thus a rate-limiting step for the production of eicosanoids [36]. Phospholipases A 2 (PLA 2 ) catalyze the hydrolysis of phospholipids in sn-2, generating free fatty acids and lysophospholipids [37]. So far, more than 20 mammalians PLA 2 s have been described. The PLA 2 repertoire includes; 1) the secreted PLA 2 s (dependent of calcium), 2) the intracellular PLA 2 s of group VI independent of calcium, 3) the intracellular PLA 2 s of group IV dependent of calcium (with the exception of the group IVC (cPLA 2 gamma), which does not rely on calcium for its activity), 4) the lysosomal PLA 2 , 5) the adipose-specific PLA 2 and 6) the platelet-activating factor acetylhydrolases. The most studied and best-described PLA 2 is the cytosolic PLA 2 of group IVA, also called cPLA 2 α. This enzyme is ubiquitously expressed in mammalian cells, and cPLA 2 α gene ablation in mice showed its critical role in fertility, particularly in fetus implantation and labor [38,39]. Importantly, the exogenous injection of PGE 2 and of a stable analog of PGI 2 (carbaprostacyclin) restored normal implantation in cPLA 2 α deficient mice [40], further supporting the notion of functional coupling between cPLA 2 α and prostaglandins. In concurrent studies, the function of cPLA 2 α in eicosanoid production in a context of inflammation is also exemplified, as cPLA 2 α deficient mice were resistant to experimental asthma, and the macrophages isolated from these mice failed to produce PGE 2 , platelet activating factor, leukotriene B4 and leukotriene C4 [38,39]. A series of subsequent studies confirmed a dominant role of cPLA 2 α in eicosanoid production in several processes, including immunity, reproduction, inflammation and cancer [7,[37][38][39][40][41][42][43][44][45]. While cPLA 2 α is expressed in thymocytes [46], whether it plays a role in eicosanoid generation and thymocyte maturation is unknown.
For this study, we portrayed the eicosanoids produced in the thymus at different stages of thymocyte maturation and considered the potential role of cPLA 2 α in this process. As the role of eicosanoids in the thymus has been invoked, we further hypothesized that cPLA 2 α might contribute to thymocyte maturation. We found that the production of eicosanoids is modulated accordingly to the maturation of thymocytes, and that the production of eicosanoids and thymocytes can proceed independently of cPLA 2 α.

Ethic statement
This study was reviewed and approved by our institutional review board (Comité Éthique de la Recherche du CHU de Québec) before the study began.
Human thymuses from newborns and young children were obtained under an approved institutional review board protocol (Comité Éthique de la Recherche du CHU de Québec) following written consent of the parents after a cardiac surgery (CHU de Quebec). This consent procedure was approved by the Comité Éthique de la Recherche du CHU de Québec.
In this study, Guidelines of the Canadian Council on Animal Care were followed in a protocol approved by the Animal Welfare committee at Laval University (Quebec City, Canada) and all efforts were made to minimize suffering. Fetal thymus harvesting was performed after euthanasia of fetuses on ice. Adult thymuses were obtained after an isoflurane anesthesia followed by euthanasia with CO 2 . deficient mice was maintained by crossing heterozygous males and females, and the littermate cPLA 2 α wild type (WT) and cPLA 2 α KO were used for our experiments. The identification of cPLA 2 α genotypes was performed using DNA isolated from mouse tail. The tails were digested with DirectPCR Lysis Reagent (Tail) (Viagen Biotech) and Proteinase K (Invitrogen) according to the manufacturer protocol and PCR amplification was performed using HotStarTaq DNA Polymerase (Qiagen) and the following primers: cPLA 2 α forward (5 0 -TTCTCTGGTGTGAT GAAGGC-3 0 ), cPLA 2 α reverse 5 0 -AAACTGACTGTAGCATCACAC-3 0 ), NeoForward (cPLA 2 α KO) (5 0 -ATCGCCTTCTTGACGAGTTC-3 0 ). The following PCR steps were used: 15 minutes at 95°c, 35 cycles of 45 seconds at 94°c, 60 seconds at 65°c and 60 seconds at 72°c and the final step is 10 minutes at 72°c. The PCR products were then separated on 1.5% agarose gel containing ethidium bromide. The WT and KO products were distinguished by visualization of bands at 224 and 570 bp, respectively.

Human thymus
Small sections of human thymuses ( 2mm 3 ) were cultured as already described for mouse FTOCs. The human FTOCs-like were fed daily by complete medium replacement with solvent control (DMSO) or pyrrophenone (Cayman Chemical) for 5 days at 37°C with 5% CO 2 .

Flow cytometry analysis
Thymuses were mechanically dissociated into single cell suspensions in PBS. The absolute cell number present in each thymus was determined by cell counting and labeling with fluorochrome-conjugated antibodies was performed according to the manufacturer protocols. The following antibodies were used: PE-Cy7 Hamster Anti-Mouse CD3e (145-2C11), PE Rat Anti-Mouse CD4 (RM4-5), APC Rat Anti-mouse CD8a (53-6.7), PE-Cy7 mouse Anti-Human CD3 (clone SK7), PE mouse Anti-human CD4 (RPA-T4) and APC mouse Anti-Human CD8 (RPA-T8). All antibodies and their related isotype controls were purchased from BD Biosciences. Flow cytometry analysis was performed on a BD FACSCanto II Flow cytometer (BD Biosciences, San Jose, California, USA) and analyzed using FlowJo software (Ashland, Oregon USA).

Mass spectrometry analysis of eicosanoids
Eicosanoids from 1ml FTOC supernatants and crushed mouse adult thymuses were analyzed by combined liquid chromatography/tandem mass spectrometry, as already described [47]. The FTOC supernatants were collected daily and conserved at -80°C before analysis. Thymuses from cPLA 2 α WT and KO adult mice were crushed in 1ml PBS 1X and conserved at -80°C before analysis. Culture media (in absence of FTOC) was used as negative control for our analyses. Deuterium standards purchased from Cayman Chemical were used to detect the set of eicosanoids listed in the Table 1.

Statistical analyses
All data are presented as mean ± SEM. Statistical significance between 2 groups was determined using unpaired Student t tests. All the statistical analyses were performed using Prism software 4.00 (GraphPad Software, CA, USA).

Eicosanoid profiling during thymocyte maturation
To determine the eicosanoids produced by the thymus through different stages of thymocyte maturation, we compared the lipid profile generated in FTOC supernatants (E15.5) after 1, 3 and 5 days of culture. The full-set of eicosanoids that was evaluated is presented in Table 1. LTC 4 , LTD 4 , LTE 4 , 8-HETE, Tetranor-12-HETE, Resolvin D2, Resolvin E1, 11α-PGF 2 α, 2,3-Dinor-11β-PGF 2 α, 11-dehydro TXB 2 and 11,12-DHET were undetectable in FTOC supernatants and adult mouse thymuses. Furthermore, we found profound changes in the eicosanoid expression profile during the course of thymocyte maturation, with LTB 4 and LXA 4 representing the majority (>50%) of the eicosanoids expressed through the first 3 days of culture ( Fig 1A and 1B, left and middle panel). At day 5 of culture, 14,15-DHET was the second most abundant lipid mediator produced by FTOCs after LTB 4 , while LXA 4 appeared essentially absent (Fig 1A and 1B, right panel). Next, we wished to verify the expression of eicosanoids present in the thymus of adult mice (6-8 weeks). In this case, we found that LTB 4 remains among the most abundant lipid mediator present in the thymus, followed by LXA 4 and 5-HETE ( Fig 1C). The production of eicosanoids by FTOCs, adult thymus and the modulation of their production during the course of thymocyte development, prompted our examination of the role of cPLA 2 α. Using FTOCs and adult thymuses from cPLA 2 α deficient mice, we observed that the majority of the most abundant eicosanoids could be produced independently of the expression of cPLA 2 α (Fig 1B and 1C). The ablation of the gene coding for cPLA 2 α led to the absence of 5,15-DiHETE, 5,14-LXB 4 and TXB 2 at day 1, of 2,3-Dinor TXB 2 , 2,3-Dinor-6-Keto PGF 1 α and 5,15-DiHETE at day 3 and of 14,15-DHET and 5-OxoETE at day 5 of culture in FTOCs, suggesting that cPLA 2 α is implicated in the generation of these lipids (Fig 1B). Furthermore, 14,15-DHET at day 1, 14,15-DHET and 11-HETE at day 3, 8,9-DHET, 5,6-LXA 4 and 5,6-DiHETE at day 5 were only detected in cPLA 2 α KO FTOC supernatants (Fig 1B) while significantly more Resolvin D1 was observed in absence of cPLA 2 α in mouse adult thymuses ( Fig  1C), suggesting that cPLA 2 α expression can also negatively regulate the production of some eicosanoids. Taken together, these results demonstrate that the production of eicosanoids is modulated accordingly to the development stages of thymocytes, and that the majority of the eicosanoids detected in mouse fetal and adult thymuses are produced independently of cPLA 2 α. These observations also point to a contribution of cPLA 2 α in expression of a subset of less abundant eicosanoids in the thymus.
The disruption of the cPLA 2 α gene does not affect the maturation of thymocytes in FTOC Although cPLA 2 α appeared dispensable for the biosynthesis of most eicosanoids, subtle changes in the lipid expression profile in the thymus were observed in absence of cPLA 2 α. Furthermore, cPLA 2 α might be implicated in the generation of eicosanoids in discrete cellular lineages in the thymus, which might not be possible to estimate when measuring the complete pool of eicosanoids produced by the entire organ. We thus wished to verify whether the cPLA 2 α is implicated in thymocyte maturation, and we firstly used a genetic approach in FTOCs [33,34]. The thymocytes present in the cultured thymus from cPLA 2 α WT and KO littermate mice were examined cytofluorometrically, and no differences in their maturation were observed. Indeed, the four subpopulations studied, the DN (CD3 + /CD4 -/CD8 -), the DP (CD3 + /CD4 + /CD8 + ) and the SP (CD3 + /CD4 + /CD8and CD3 + /CD4 -/CD8 + ) thymocytes showed the same repartition in the cPLA 2 α WT and KO FTOCs (Fig 2A and 2B). In light of these results, cPLA 2 α is dispensable for the maturation of thymocytes in mice.
Evaluation of the impact of the cPLA 2 α inhibitor pyrrophenone on thymocyte maturation We next used a pharmacological approach to confirm our observations made in genetically engineered mice. The cPLA 2 α inhibitor pyrrophenone (PP) [48] suppresses AA release from an activated monocytic cell line and PGE 2 release by renal mesangial cells with an IC 50 of 24nM and 8nM, respectively [48]. WT FTOCs were cultured during 5 days in absence or in presence of 10 and 100 nm of PP and the repartitioned thymocyte subpopulations were analyzed by flow cytometry. As for the genetic approach, cPLA 2 α appeared dispensable, as no differences were observed in the maturation of thymocytes in presence of PP compared to those left untreated (Fig 3A and 3B). Rocca [21,35]. This effect of NS-398 was considered unspecific, as it was recapitulated in COX-2 deficient FTOCs and it was not reversed by the addition of PGE 2 [35]. Using high concentrations (1μM) of the cPLA 2 α inhibitor, we observed an increase and a decrease of DN and DP thymocyte populations, respectively (Fig 4A and 4B). Furthermore, we observed an increase of the two SP populations (Fig 4A and 4B). Thus, high dose of PP affects the maturation of mouse thymocytes. We next wished to confirm the specificity of the inhibitor, here using cPLA 2 α KO FTOCs. We observed that PP, used at 1 μM, impedes the maturation of thymocytes deficient in cPLA 2 α (Fig 5A and 5B). Furthermore, the exogenous addition of AA and of PGE 2 , which was reported involved in thymocyte maturation [21], to PP-treated FTOCs did not restore normal thymocyte maturation (Fig 6A and 6B). Taken together, these results demonstrate that high doses of PP inhibit thymocyte maturation through the inhibition of another target than cPLA 2 α, most likely irrelevant to AA and prostaglandin release.

cPLA 2 α gene disruption does not impact thymocyte maturation in the adult mouse
Prior studies evaluated the role of prostaglandins in thymocyte maturation in the adult [21]. Having confirmed that cPLA 2 α is dispensable in thymocyte maturation at a fetal development   thymocyte subpopulations were observed between WT and KO thymuses (Fig 7A and 7B). Thus, the cPLA 2 α is dispensable for normal thymocyte maturation in adult mice.

Pharmacological inhibition of cPLA 2 α does not impact human thymocyte maturation
Having demonstrated that the maturation of fetal and adult mouse thymocytes could proceed independently of cPLA 2 α, we wished to confirm our observations using human thymocytes. For this, we used thymuses from human newborns and young children suffering of cardiac malformation and undergoing thymectomies. To determine the role of the cPLA 2 α in human thymocyte maturation, small sections of human thymuses were cultured in absence or in presence of different concentrations of PP and then the thymocyte subpopulations were determined cytofluorometrically. We observed no differences in the percentage of different thymocyte subsets (DN, DP, SP CD4 + and SP CD8 + ) when thymuses were treated with PP up to 1μM (Fig 8A and 8B). Thus, cPLA 2 α appears dispensable for the maturation of human thymocytes.

Discussion
In this study, we reveal for the first time the elaborated set of eicosanoids produced by the thymus through different stages of thymocyte development. LTB 4 and LXA 4 were the most abundant eicosanoids found in thymus. LTB 4 is a recognized pro-inflammatory mediator involved in phagocyte chemotaxis, [49] while LXA 4 displays anti-inflammatory activities and mediates clearance of apoptotic cells [50,51]. LTB 4 and LXA 4 might play roles in the thymus, such as the recruitment of phagocytes and the stimulation of apoptotic cell clearance. The actual role of these lipids in the thymus is worth investigating, especially when it is considered that 98% of the thymocytes die by apoptosis in the thymus [52,53]. Furthermore, we showed that the eicosanoid expression profile is modulated through the differen t thymocyte maturation stages, pointing to tight regulation of enzymes implicated in eicosanoid generation in the thymus. Future studies are thus necessary to verify the role of eicosanoids in thymus and the regulation mechanisms behind their production.
Through its important role in eicosanoid production, cPLA 2 α plays major roles in several physiological and pathophysiological processes, including immunity, reproduction, cancer and inflammation [7,[37][38][39][40][41][42][43][44][45]. Prostaglandins and their receptors are expressed in the thymus, and prior studies suggested that they are necessary for proper thymocyte maturation. Furthermore, it was demonstrated that the thymus is the organ with the highest concentration of thromboxane receptor, which is mostly expressed on DP thymocytes and appears implicated in the induction of thymocyte apoptosis [22] [25]. Herein, we surmised that cPLA 2 α might participate in eicosanoid generation and thymocyte maturation. To our surprise, we observed that production of most abundant eicosanoids and thymocyte maturation in the thymus occur independently of cPLA 2 α.
While cPLA 2 α is expressed in thymocytes [46] and its mRNA expression is modulated throughout development (S1 Fig), the exact role of cPLA 2 α in the thymus thus remains obscure. We investigated the impact of cPLA 2 α on the major populations of thymocytes based on surface expression of CD3, CD4 and CD8 receptors. However, cPLA 2 α and its products might have more subtle roles, and might regulate the development of other T-cell subpopulations such as T regs and γδ T-cells. Indeed, we showed that absence of cPLA 2 α has an impact on some less abundant eicosanoids. Whether these eicosanoids, and thus the cPLA 2 α, are involved in the function or development of scarce cellular populations is unknown. Furthermore, cPLA 2 α might be implicated in the production of eicosanoids that both positively and negatively regulate maturation of thymocytes. Thus, the overall effects of cPLA 2 α deficiency on thymocyte phenotype would be imperceptible. Finally, lysophosphatidic acid is involved in lymphocyte transmigration from the high endothelial venules of lymph nodes [54]. Whether cPLA 2 α and its products are also implicated in processes such as thymocyte entry or egress is unknown. As cPLA 2 α and AA metabolites are expressed in the thymus, the delineation of their exact role in the establishment of T-cell repertoire remains of great interest.
The prior demonstration of a role of prostaglandins in thymocyte maturation [21] was our impetus for our investigation of cPLA 2 α in the thymus. However, the actual role of prostaglandins in thymocyte maturation is currently debated. Indeed, two distinct studies reported divergent results. Whereas a first study suggested that COX-1 and COX-2-derived PGE 2 participate in thymocyte maturation [21], a second one described that mice lacking expression of COX-1 and COX-2, EP-1 and EP-2 display normal thymocyte maturation [35]. What explains the discrepancies between these two studies is unclear, but we speculate that specific housing animal facility environment or background genetic drift might have contributed. Our results cannot settle the debate. Indeed, cPLA 2 α is not the only PLA 2 enzyme expressed in the thymus [55][56][57] and other enzymes might participate to prostaglandin production in its absence. Hence, PGE 2 levels are not altered by the absence of cPLA 2 α in the thymus (Fig 1B and 1C). Furthermore, sPLA 2 X, which is highly efficiently at releasing AA from the cellular outer leaflet, is also expressed in the thymus [55,56,58,59]. As we also excluded sPLA 2 X in thymocyte maturation (S2A and S2B Fig), other PLA 2 and or lipases expressed in thymus [56,57] might thus compensate the absence of the cPLA 2 α and sPLA 2 X for the production of prostaglandins, We further observed that high concentrations of the cPLA 2 α inhibitor PP impair thymocyte maturation in mice, but not in humans. Similarly to the observations made by Xu et al. using high concentrations of the COX-2 inhibitor NS-398, [35] we demonstrated that the effect of PP at high concentrations (around 125 time higher than the IC 50 ) is independent of its ability to inhibit its specific target. It seems unplausible that the unspecific target(s) of NS-398 and PP are the same. Indeed, the two compounds are structurally highly different and the unspecific effects observed on thymocytes are also distinct. Given that PP has no impact on human thymocyte development, we suggest that its unspecific target expressed in mice has no human ortholog, or that the human ortholog has a much lower affinity for the inhibitor. An unspecific effect of PP has recently been reported in a distinct study [60]. The authors demonstrated that the release of AA and lactate dehydrogenase from cPLA 2 α KO fibroblasts was efficiently inhibited by PP through the prevention of mitochondrial calcium uptake. The inhibition of this process in FTOCs could explain the reduction in thymocyte maturation but remains to be established.
In sum, our study provides novel information concerning the broad repertoire of eicosanoids present in the thymus and on the role of cPLA 2 α in thymocyte development. As a plethora of molecules drive T-cell functions in lymphoid organs and in the periphery, our study adds to the comprehension of mechanisms that are key in immunity. Representative thymocyte subpopulation distribution in WT and KO sPLA 2 X FTOC after 5 days of culture. Thymocytes were identified by flow cytometry using fluorochrome-conjugated antibodies directed against CD3, CD4 and CD8. B. WT and KO sPLA 2 X fetal thymuses were cultured during 5 days as FTOCs. After mechanical dissociation of fetal thymuses, thymocytes were labeled with fluorochrome-conjugated antibodies directed against CD3, CD4, and CD8 and analyzed by flow cytometry. Data are mean ± SEM of 4 independent experiments and the number of fetal thymuses for each genotype is: sPLA 2 X +/+ (n = 7); sPLA 2 X -/-(n = 13). NS (non significant). (TIF)