Platelet-activating factor (1-O-alkyl-2-acetyl-sn-glycero-3-phosphocholine, PAF), a potent proinflammatory phospholipid mediator, has remarkably diverse biological effects in diseases ^[1], including sepsis, which arises through body's inflammation response to infection and is a leading cause of death and disability for patients in an intensive care unit ^[2]. PAF synthesis is up-regulated in response to bacterial endotoxins both in vivo and in vitro^[3], ^[4]. Although it was recently reported that PAF may protect mice against lipopolysaccharide (LPS)-mediated sepsis ^[5], many studies indicated that increased concentrations of PAF may contribute to the deleterious effects of systemic inflammation in the pathogenesis of severe sepsis ^[2].

The inactivation of PAF is mediated by PAF acetylhydrolase (PAF-AH), a calcium-independent phospholipases A2 with specificity for hydrolysis of this lipid mediator ^[6]. The plasma form of PAF-AH (pPAF-AH) is a secreted protein in the blood that serves to inactivate PAF and PAF-like phospholipids ^[6]. This enzyme accounts for all of the PAF-inhibitory activity found in human serum, limiting the normal serum half-life of PAF to only a few minutes ^[1], ^[7]. Except for the sepsis caused by severe leptospirosis ^[8], the activity of pPAF-AH is diminished in human sepsis ^[7], ^[9], ^[10], ^[11] as a consequence of endotoxin and cytokine-induced reduction of the pPAF-AH encoding gene transcription and possible inactivation by oxidant injury ^[11], ^[12].

A potential therapeutic strategy for sepsis is to facilitate the inactivation of PAF with the supplement of pPAF-AH. The results of the clinical trials of recombinant human pPAF-AH in patients with severe sepsis were controversial (Table 1). In 2006, Gomes et al. reported that the administration of exogenous recombinant pPAF-AH reduced mortality and inflammatory injury relevant to the clinical syndrome (Table 1), based on the depressed pPAF-AH activity in C57BL/6 and Swiss mouse models induced by LPS or cecal ligation and puncture (CLP) (Table 2). However, this result is partly in contradictory to the previous studies in rodents challenged with LPS, which showed an increase of pPAF-AH activity in Syrian hamsters, rats, and C57BL/6 mice (Table 2). Therefore, it would be important to clarify the response of pPAF-AH against LPS-treatment in mice, and/or explore alternative animals suitable for simulating the role of pPAF-AH in human sepsis.

Previously, we found that Mongolian gerbil has its normal pPAF-AH level similar to that of human, and the patterns of the change of PAF-AH level in serum during the course of severe leptospirosis in gerbil model are similar to that of severe leptospirosis patients, including the levels of elevation ^[8]. These findings were consistent with the fact that LPS of Leptospira interrogans is much less virulent than that of Escherichia coli, and has little effect on the sepsis caused by leptospirosis ^[13]. Therefore, among experimental rodents and rabbits, gerbil is likely to be a good candidate to develop an animal model to mimic the role of pPAF-AH in human diseases, particularly, the LPS-induced sepsis ^[8]. In this study, we examined the in vivo effects of LPS on pPAF-AH activity in C57BL/6 mice and gerbils.

Compared to the control group (0 mg/kg body weight, saline only), the C57BL/6 mice and gerbils with LPS-treatment (3 or 5 mg/kg body weight) appeared acutely ill and displayed signs of lethargy, and then they were euthanized while the animals appeared moribund after LPS-treatment (Table 3). Most of the animals died after LPS injection, and autopsy showed the volume increase of spleen in all the animals of both C57BL/6 mice and gerbils (Table 3). Therefore, the clinical manifestations of LPS-induced gerbil model were apparently similar to that of C57BL/6 mouse sepsis model ^[9].

Blood was collected by cardiac puncture and pPAF-AH activity was measured. We found that both 3 and 5 mg/kg body weight LPS induced the elevation of the pPAF-AH activity in C57BL/6 mice, and the elevated levels were similar in these two dose group (Figure 1). In contrast, after LPS-treatment (3 or 5 mg/kg body weight), the pPAF-AH activity of gerbils decreased compared to that of the control, and the decreased levels were similar in the doses of 3 and 5 mg/kg body weight (Figure 1).

Our result showed that LPS caused the elevation of pPAF-AH activity in C57BL/6 mice (Figure 1), which was the same with that reported by Memon et al. ^[14] (Table 2), but different from the study of Gomes et al. ^[9] (Table 2). Although the reason for the different effects of LPS in C57BL/6 mice was yet to be elucidated, the present study, together with the studies in Syrian hamsters, rats, and C57BL/6 mice challenged by LPS ^[14], ^[15], ^[16], ^[17] (Table 2), showed that, among rodent species, only the gerbil demonstrated the decrease of pPAF-AH activity under the exposure of LPS (Figure 1), which was similar to the response of pPAF-AH measured in sepsis patients ^[7], ^[9], ^[10], ^[11]. Therefore, the LPS-induced sepsis in gerbil could be used to study the pharmacological effect of recombinant pPAF-AH in sepsis, and may provide pre-clinical evaluation of pPAF-AH in sepsis for the additional clinical trials in the future.

LPS is a relatively pure compound that can be stably stored in lyophilized form. Therefore, the LPS model is much easier to be established than the surgical CLP model, which is the “gold standard” in sepsis research ^[18]. However, the LPS model is known to be different from the sepsis in human and CLP model with respect to the profile of cytokine release ^[18], ^[19]. PAF is a phospholipid cytokine implicated in a wide range of biological and pathologic responses ^[1], and thus, the different responses of pPAF-AH, the regulator of serum PAF ^[6], between sepsis patients and LPS models of Syrian hamster, rat and mouse (Table 2 and Figure 1) might be accounted by the differences in the profile of cytokine release. The similar responses of pPAF-AH between sepsis patients and gerbil LPS model (Figure 1), as well as between the severe leptospirosis patients and the gerbil leptospirosis model ^[8], may implicate the similar response of cytokine release between human and gerbil. This possible underline mechanism should be further tested in order to fully characterize this novel model, which is potentially advantageous in mimicking the cytokine response in sepsis.

Competing Interests: The authors have declared that no competing interests exist.

Funding: This work was supported in part by grants from the National Natural Science Foundation of China (No. 30830002, 30970077, 30900051, 30970125, 30770111 and 30770820), the National Key Program for Infectious Diseases of China (No. 2008ZX10004-002, 2008ZX10004-009 and 2009ZX10004-712), and Shanghai Leading Academic Discipline Project (T0206), Shanghai Natural Science Foundation of China (No. 062R14056), Shanghai Municipal Education Commission and Shanghai Education Development Foundation (“Chen Guang” project, 2007CGB06). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.