Phospholipase A2 in Experimental Allergic Bronchitis: A Lesson from Mouse and Rat Models

Background Phospholipases A2 (PLA2) hydrolyzes phospholipids, initiating the production of inflammatory lipid mediators. We have previously shown that in rats, sPLA2 and cPLA2 play opposing roles in the pathophysiology of ovalbumin (OVA)-induced experimental allergic bronchitis (OVA-EAB), an asthma model: Upon disease induction sPLA2 expression and production of the broncho-constricting CysLTs are elevated, whereas cPLA2 expression and the broncho-dilating PGE2 production are suppressed. These were reversed upon disease amelioration by treatment with an sPLA2 inhibitor. However, studies in mice reported the involvement of both sPLA2 and cPLA2 in EAB induction. Objectives To examine the relevance of mouse and rat models to understanding asthma pathophysiology. Methods OVA-EAB was induced in mice using the same methodology applied in rats. Disease and biochemical markers in mice were compared with those in rats. Results As in rats, EAB in mice was associated with increased mRNA of sPLA2, specifically sPLA2gX, in the lungs, and production of the broncho-constricting eicosanoids CysLTs, PGD2 and TBX2 in bronchoalveolar lavage (BAL). In contrast, EAB in mice was associated also with elevated cPLA2 mRNA and PGE2 production. Yet, treatment with an sPLA2 inhibitor ameliorated the EAB concomitantly with reverting the expression of both cPLA2 and sPLA2, and eicosanoid production. Conclusions In both mice and rats sPLA2 is pivotal in OVA-induced EAB. Yet, amelioration of asthma markers in mouse models, and human tissues, was observed also upon cPLA2 inhibition. It is plausible that airway conditions, involving multiple cell types and organs, require the combined action of more than one, essential, PLA2s.

Accordingly, the control of PLA 2 activities has been proposed for treating respiratory inflammatory/allergic diseases. Cellular PLA 2 s are generally classified into the intra-cellular cytosolic and the Ca 2+ -independent PLA 2 s (cPLA 2 s and iPLA 2 s, respectively), and the secretory PLA 2 s (sPLA 2 s). Previous studies have assigned a role for secretory and cytosolic PLA 2 s in inflammatory/allergic processes, while the iPLA 2 does not seem to be significantly involved in airway pathology [6][7][8][9][10]. However, these studies have not produced an unequivocal conclusion.
In a previous study, we investigated the involvement of PLA 2 s and eicosanoids in asthma pathophysiology using a rat model of ovalbumin (OVA)-induced experimental allergic bronchitis (EAB) [4,11], as expressed by broncho-constriction, airway remodeling, the levels of the broncho-dilator PGE 2 and the bronchoconstrictor Cysteinyl-LTs (CysLTs) in bronchoalveolar lavage (BAL). Upon induction of EAB these indices were up-regulated, except for PGE 2 which was markedly reduced. Concomitantly, sPLA 2 expression in lung tissue was enhanced, while cPLA 2 expression was markedly decreased. All these parameters were reversed upon amelioration of the disease by treatment with an sPLA 2 inhibitor, resulting in elevation of cPLA 2 and PGE 2 along with suppression of sPLA 2 and Cys-LTs [4,11]. PGE 2 , generally considered a pro-inflammatory mediator, is a potent broncho-dilator and inhibits smooth muscle cell proliferation [11][12][13][14][15]. It has thus been postulated that, unlike other organs, the lung is unique in benefiting from the action by PGE 2 [11]. Therefore, the results obtained with the rat EAB model, seemed to make a clear physiological sense, suggesting that sPLA 2 plays an important role in the onset and progression of asthma, while cPLA 2 is involved in the disease abatement.
However, differing results were presented in studies with mouse models, mostly using PLA 2 genetic manipulations. Henderson et al. [16,17] assigned a key role to sPLA 2 gX, showing that physiological and biochemical markers of OVA-induced asthma were reduced in sPLA 2 gX -deficient mice [16]. These markers were enhanced when the mouse sPLA 2 gX was replaced with human sPLA 2 gX, or inhibited by treatment with a specific sPLA 2 gX inhibitor [17]. Munoz et al. [18] reported that cell migration and airway hyperresponsiveness were attenuated in OVA-sensitized PLA 2 gVdefficient mice, as well as by treatment of mice (WT) with sPLA 2 gV antibody. Similarly, Giannattasio et al. [19] showed that Dermatophagoide farina-induced lung inflammation was attenuated in sPLA 2 gV-deficient mice.
On the other hand, Uozomi et al. [20] showed that in cpla2deficient mice OVA-induced anaphylactic response and bronchial reactivity to methacholine were significantly reduced. Similarly, Bickford et al. [21] showed that mice sensitized/stimulated with Aspergillus fumigatus exhibited marked elevation of cPLA 2 c mRNA expression. These discrepancies might be due to differences in methodologies and/or genetic manipulations, or might reflect the involvement of more than one PLA 2 type. To explore these possibilities, in the present study we examined the role of PLA 2 s in OVA-induced EAB in mice, without genetic manipulation of PLA 2 , using the same methodology and procedures applied to rats in our previous study [4,11]. It was found that, similar to our findings with rats [4,11], OVA-induced EAB in mice was associated with enhanced sPLA 2 expression and production of broncho-constricting eicosanoids. However, in contrast to EAB in rats, cPLA 2 mRNA expression and PGE 2 production were elevated in the mouse model. Yet, in both models, the disease was markedly ameliorated by treatment with a cell-impermeable sPLA 2 inhibitor.

Ethic statement
This study includes experiments with mice, all conducted according to the instruction and permit of the Hebrew University Ethical Committee

Induction of experimental allergic bronchitis (EAB) in mice
As in our previous study in rats [4,11], in the present study EAB was induced in BALB/c female mice by a weekly IP injection of 0.3 ml PBS containing 100 mg OVA, and 2 mg of the adjuvant Al(OH) 3 for three weeks, followed with four weeks of challenge by three weekly intranasal (IN) OVA administration (100 mg in 50 mL PBS). EAB development was assessed by two common tests: 1. Pulmonary function. by airway response to allergen or methacholine, using two non-invasive methods: a. Enhanced pause (Penh): Unrestrained conscious mice were placed in a whole-body plethysmograph (Buxco Electronics, Troy, NY), measuring flow-derived pulmonary function (Penh), as previously described [4,11,17,22]. b. Airway resistance using the occlusion technique (Roccl): Nonsedated mice, with closed mouth, were breathing through a nose-mask connected to a pneumotach (flow-meter) with a mouth pressure port. The pneumotach was attached to 2 differential pressure transducers, connected through preamplifiers (Hans Rudolph, Shawnee, KS, USA) producing analog signals of flow and mouth pressure, digitized by a data acquisition program (LabView National Instruments, Austin, TX). Peak pressure was measured while the mouse was breathing against an occluded pneumotach for 3-5 breaths.
The pressures generated at the beginning and at the end of occlusion (inspiratory and expiratory, respectively) were divided by the respective adjacent peak flow immediately before and after the occlusion. Resistance (Rocclud) was calculated as peak pressure divided by the adjacent peak flow. Airway resistance is expressed as the percent change compared to baseline (level before treatment). Airway reactivity was assessed before challenge (baseline) and 5 minutes after IN challenge with either OVA or increasing methacholine dose (0, 40, 80, 320, 640, and 1280 mg in 20 mL PBS).
2. Gene expression of arginase-I and mammalian acidic chitinase in lung tissue. both enhanced in asthma. Arginase-I is involved in L-arginine metabolism and the subsequent inhibition of NO production, typical of type-2 responses [23,24]. Although chitin does not exist in mammals, chitinases and chitinase-like proteins have been observed in mice and human subjects [25]. The prototypic acidic mammalian chitinase is induced during T H 2 inflammation through an IL-13-dependent mechanism, and plays an important role in the pathogenesis of T H 2 inflammation and IL-13 effector pathway activation [25][26][27]. The respective primers are depicted in Table 1.

Histological Analysis by Hematoxylin and Eosin Staining
Lungs preserved in 4% formaldehyde were dehydrated, sliced longitudinally, and embedded in paraffin. Histological sections of 4 mm thick were cut on a microtome, placed on glass slides, deparaffinized and stained sequentially with hematoxylin (for nuclear material) and eosin (for cytoplasmic material). PLA 2 mRNA expression in lung was determined by RT-PCR, using conventional methods [28]. Total RNA was purified from lung tissues (SV Total RNA isolation kit, containing DNase I Promega Corporation, Madison, WI) to remove possible genomic DNA contamination. RNA integrity was tested by 1% agarose gel electrophoresis. cDNA was prepared from total RNA (2 mg/ml) using MuLV reverse transcriptase (Applied Biosystems). Primers were designed using the Primer Express program (Applied Biosystems). Target mRNA was calculated in reference to the endogenous 18S ribosomal RNA, while the naive group was used as a calibrating factor. The respective primers are depicted in Table 1.

Treatment with cell-impermeable sPLA 2 inhibitor
As in the previous study of EAB in rats [4,11], we have tested the effect of a cell-impermeable sPLA 2 inhibitor, composed of PLA 2 -inhibiting lipid (specifically derivatized phosphatidyl ethanolamine), conjugated to hyaluronic acid (HyPE), which prevents the inhibitor's internalization, thereby designed to confine the inhibitory action to the cell membrane. This inhibitor has been shown to suppress the action of exogenous sPLA 2 s and diverse related inflammatory conditions in a number of studies [4,11,30]. The mice were treated during the challenge, one hour before each OVA challenge, with IN administration of HyPE (200 mg in 50 ml at the first two challenges, followed by 40 mg in 40 ml, until one day before sacrifice).
Statistical analysis was done using one-way ANOVA, followed by Tukey multiple comparison. Conventionally, P#0.05 was considered significant.

Induction of OVA-induced EAB in mice
showing that methacholine challenge exerted airway resistance to air flow in a dose-dependent manner (Fig. 1), concomitantly with enhanced expression of arginase-I and chitinase mRNA (Fig. 2). Similar to our findings with rats [4,11], the elevation of these physiological and biochemical markers was inhibited by treatment with the sPLA 2 inhibitor.

Airway response to OVA challenge
Mice with OVA-induced EAB responded to OVA challenge with markedly enhanced airway resistance, as expressed both by Penh (Fig. 3A) and resistance (Fig. 3B). Similarly, EAB induction was associated with peribronchial infiltration of inflammatory cells, as shown in the histology micrographs (Fig. 4A) and in the respective morphometric measurement (Fig. 4B). These figures also show that pre-treatment with the sPLA 2 inhibitor completely prevented the disease development, reverting both the airway response (Figs. 3A &3B) and the inflammatory cell infiltration (Figs. 4A & 4B), to their level in naïve mice.

PLA 2 s expression in lungs
As noted above, in the rat model the disease induction was associated with suppression of cPLA 2 expression [4,11], while studies with mice suggested that the disease induction involved elevated expression of PLA 2 gIVC (cPLA 2 c) [21] RNA expression, as well as sPLA 2 gV [18,19] and sPLA 2 gX [16,17]. In the present study we have found that while sPLA 2 gV, sPLA 2 g1B and sPLA 2 gIII were not affected by the disease induction, the expression of both sPLA 2 gX and cPLA 2 c was markedly increased, and both were suppressed by treatment with the sPLA 2 inhibitor (Fig. 5). The elevated sPLA 2 expression is in agreement with our findings in the rat model [4,11], and with other studies in mice [16][17][18]. However, the elevated cPLA 2 expression, while in agreement with others' studies with mice [20,21], is in contrast to our findings in rats, where cPLA 2 was suppressed in the disease state, and resumed upon treatment with the sPLA 2 inhibitor. Fig. 6 shows that, in parallel to the PLA 2 expression, EAB induction was associated with enhanced production of both the broncho-constricting PGD 2 , TXB 2 and CysLTs, and the broncho-dilating PGE 2 , which is in agreement with previous studies with mice [4,11].The elevation of the broncho-constricting eicosanoids in the disease state is in accordance with our findings in the rat model [4,11]. However, the elevated PGE 2 production observed here is in contrast to our findings in the rat model, where the disease induction was associated with suppression of PGE 2 .

Expression of 5-lipoxygenase
In recent years, airway inflammation has been shown to undergo temporal changes from the inflammatory phase, where 5lipoxygenase (5-LO) produces the broncho-constricting LTs, to a resolution phase, in which 15-LO produces anti-inflammatory lipid mediators, such as protectins and resolvins [31][32][33][34][35]. The EAB model applied in the present study does not reach the resolution phase. Accordingly, as shown in Fig. 7, the EAB induction was associated with elevation of 5-LO protein expression, which was suppressed by treatment with the sPLA 2 inhibitor, whereas 15-LO expression was not affected by the disease or its treatment (not shown).

PLA 2 expression
As discussed in the Introduction, previous studies with mouse models of asthma have produced differing results, showing that the disease was associated with increased expression of sPLA 2 gX [16,17], sPLA 2 gV [18,19], or cPLA 2 [20,21], and was ameliorated by treatment with specific inhibitors or genetic manipulations of these enzymes. Our previous study with the rat EAB model [4,11] conforms to the studies with mice, pointing to sPLA 2 s as a key player in asthma pathophysiology [16,17]. However, in contrast to the previous studies with mice that associated the disease with elevated cPLA 2 expression [20,21], in rats we have found that the disease was associated with suppression of cPLA 2 expression. To determine whether these discrepancies reflect differences between species or methodologies (e.g., genetic manipulation, stimulants, selection of PLA 2 isoforms studied), in the present study we applied to mice, with no genetic manipulation, the same protocol of OVAinduced EAB used in the rat study [4,11]. The results of both models, summarized in Table 2 show that, similar to the findings with rats, OVA-induced EAB in mice was associated with increased sPLA 2 gX, conforming to the finding of Henderson et al. [16,17], while sPLA 2 gV was not affected. However, contrary to our findings with rats, where cPLA 2 expression was suppressed in the disease state, OVA-induced EAB in mice was associated with elevated expression of cPLA 2 c and cPLA 2 a, which agrees with previous mouse studies assigning a role for cPLA 2 in asthma pathophysiology [20,21]. In addition, Giannattasio et al. [36]   reported that IgG-stimulated human lung mast cells are a source for several sPLA 2 s that contribute to LTC 4 production, known to facilitate asthma development. Subsequently, in the present study we also examined mRNA expression of some of the reported sPLA 2 isoforms, specifically sPLA 2 gXIIA, sPLA 2 gXIIB, sPLA 2 -gIB, sPLA 2 gIII, and sPLA 2 gVI in the mice lung, and found that none of them was affected in the OVA-induced EAB (not shown). It therefore seems that the results would differ between animal models, depending on the species and methodologies used.
Another limitation of the OVA-induced EAB in mice, and possibly of the other models discussed above, is indicated by the finding that EAB is associated with elevation of 5-LO (Fig. 7), known to be involved in the disease induction, whereas 15-LO, which involved in the disease resolution [33], was not affected (data not shown). This might suggest that these animal models reflect different phases of the course of the disease. It is not unlikely that PLA 2 expression varies at different phases and this contributes to the discrepancies between the expressions of PLA 2 isoforms observed in the various studies with animal models.

Lipid mediators
As shown in Table 2 the induction of EAB in rats was associated with suppressed production of PGE 2 , concomitantly with enhanced production of Cys-LT, and both were reversed upon disease amelioration [4,11]. This is physiologically sound, since PGE 2 is a broncho-dilator, and Cys-LTs is a bronchoconstrictor [37]. However, OVA-induced EAB in mice is associated with elevation of both the broncho-dilator PGE 2 and the bronco-constricting eicosanoids, Cys-LTs [4], TBX 2 and PGD 2 . This is in agreement with the above-discussed studies reporting that in the mouse asthma model the disease state is characterized by elevated production of both types of eicosanoids [16,20], and these were inhibited, along with the other disease indices, by inhibition of either sPLA 2 or cPLA 2 .
Notably, in the study with a mixed human lung cell population, cPLA 2 inhibition decreased the ionomycin-induced production of PGD 2 , LTB 4 and TXA 2 , but not that of PGE 2 [37]. Since PGE 2 is a broncho-dilator [4], the authors considered that as a positive outcome of the treatment. In line with that, in the present study, the treatment with an sPLA 2 inhibitor strongly suppressed, practically to the basal (naïve) level, the elevated production of CysLTs, TXB 2 and PGD 2 , while PGE 2 level was only partially reduced (Fig. 6), thereby turning their balance toward the broncho-dilating PGE 2 . This supports the notion that airway pathophysiology is ultimately determined by the balance between the dilating and constricting lipid mediators.   It should be noted that the research on inflammatory lipid mediators in airway conditions has addressed predominantly the eicosanoids. However, PLA 2 activity is also responsible for the production of lyso-phospholipids, some of which are known to be potent inflammatory/allergic mediators; e.g. lyso-phosphatidylserine activates mast cells to secret histamine, lyso-phosphatidic  acid induces muscle cell proliferation, and lyso-phosphatidylcholine is the precursor of PAF [4,38] and more [39]. Therefore, the focus on eicosanoids might provide only part of the picture, as it ignores the potentially major role of lyso-phospholipids and the respective PLA 2 activities in airway pathophysiology.

Use of PLA 2 inhibitors
An intriguing phenomenon presented by the present and previous studies on PLA 2 in asthma-related pathophysiology in mouse and rat models, is that the disease was successfully treated by specific inhibitors or genetic manipulations of specific PLA 2 s, including sPLA 2 gX, sPLA 2 gV and cPLA 2 c [16][17][18][19][20][21].
Similarly, the study of Hewson et al. [37] showed that a specific inhibitor of cPLA 2 a inhibited the contarctility of AMP-stimulated isolated human tarcheal rings, as well as eicosanoids production by mixed human lung cells and IgE-stimulated mast cells. On the other hand, in a recent study (Mruwat et al., unpublished), we have found that the production of inflammatory/allergic cytokines (IL-5, IL-13, IL-17 and INF-c) by cultured human nasal polyps stimulated with super antigen, was associated with increased expression of sPLA 2 gX, and suppression of cPLA 2 a expression. Yet, both cytokine production and PLA 2 expression were reversed by treatment with the sPLA 2 inhibitor used in the present study.
Taken together, the studies with animal models and human tissues discussed above, appear to suggest that since several tissues and cell types take part in the pathophysiology of asthma and related airway conditions [9,36], it is plausible that the disease development requires a combined (likely sequential) action of more than one essential PLA 2 -from different cell types -and blocking one of them would significantly attenuate the disease. This hypothesis conforms to the model of Murakami et al. [40], proposing various modes of cross-talk between sPLA 2 and cPLA 2 in the induction of airway diseases.
In conclusion, the findings and considerations summarized above demonstrate that animal models can provide only limited insight into the role of PLA 2 isoenzymes in the pathophysiology of human airway diseases. As these conditions involve multicellular/ multi-organ processes, it is plausible to conclude that human asthma and related conditions require the combined action of more than one essential PLA 2 isoform. By changing the ratio between the pro-and anti-inflammatory lipid mediators -eicosanoids and lyso-phospholipids -PLA 2 inhibition would determine the disease resolution. Which PLA 2 isoform(s) should be the target for pharmacological inhibition is yet to be explored and will ultimately be decided based on comprehensive clinical studies.