Previous Article | Next Article ![]()
Antimicrobial Agents and Chemotherapy, August 2007, p. 2965-2968, Vol. 51, No. 8
0066-4804/07/$08.00+0 doi:10.1128/AAC.00219-07
Copyright © 2007, American Society for Microbiology. All Rights Reserved.

Lisa Alleva,1,
Mohammed Alsharifi,2
Aulikki Koskinen,2
Victoria Smythe,2
Arno Müllbacher,2
Jeff Wood,3 and
Ian Clark1*
School of Biochemistry and Molecular Biology,1 John Curtin School of Medical Research,2 Statistical Consulting Unit, Australian National University, Canberra ACT 0200, Australia3
Received 13 February 2007/ Returned for modification 29 March 2007/ Accepted 24 May 2007
|
|
|---|
|
|
|---|
In recent years the threat of H5N1 influenza virus infection has generated increased interest in the mechanisms of influenza disease. As with other influenza virus subtypes (19), a number of papers have argued an important causative role for proinflammatory cytokines in H5N1 disease (5, 10, 29), as well as that caused by a reconstructed version of the strain of influenza virus responsible for massive human mortality in 1918-1919 (13).
Accordingly, we have tested a drug known to reduce this cytokine response for its ability to protect against influenza. Here we report that gemfibrozil, a commonly prescribed drug with long-established lipid-lowering properties (18) but only recently recognized as inhibiting the pathways leading to inflammatory cytokine release (3) (and having associated in vivo activity [4]), increases survival in mouse influenza virus infection. Moreover, mortality was reduced in mice when daily treatment was commenced after illness was evident.
|
|
|---|
Influenza virus. Stocks of influenza virus A/Japan/305/57 (A/Jap, H2N2) were grown in embryonated eggs. Virus-containing allantoic fluid was harvested and stored in aliquots at –70°C. Virus content was determined by hemagglutination using erythrocytes from Gallus domesticus. The stock virus titer used throughout this study was 1.6 x 103 hemagglutination units per ml of allantoic fluid.
Infecting and treating mice. Influenza virus infection was established by inoculating 2,2,2-tribromoethanol (Avertin)-anesthetized mice intranasally with 50 hemagglutination units of virus. Mice were weighed prior to infection and then daily from days 4 to 12 inclusive. Gemfibrozil (Spectrum Chemicals, Gardena, CA) was dissolved in 100% propylene glycol (MP Biomedicals, Solon, OH) and administered to mice at doses between 20 mg/kg and 60 mg/kg by the intraperitoneal (i.p.) route (10 µl/mouse) using BD ultrafine needle insulin syringes, 0.5 ml, 29 gauge (Becton Dickinson, Franklin Lakes, NJ), once daily from days 4 to 10 inclusive after virus exposure. Control mice were given propylene glycol alone (10 µl/mouse) i.p. on the same days. Survival was monitored for 30 days.
Treating mice exposed to LPS. To test the ability of gemfibrozil to affect mortality in a standard model of severe systemic inflammation, 30 mg/kg lipopolysaccharide (LPS; from Escherichia coli serotype O111:B4; Sigma) was administered i.p. to mice. A single 60 mg/kg dose of gemfibrozil dissolved in propylene glycol was administered i.p. to mice 1 h before or 2 h after LPS. Control mice were given propylene glycol alone. Mice were monitored for 14 days.
Statistics. The effect of gemfibrozil on survival time was tested using the log rank test, and the Cox proportional hazards model was used to estimate hazard ratios (9). R (21) was used for these analyses.
|
|
|---|
|
View this table: [in a new window] |
TABLE 1. Weight loss in mice infected with influenza virus (A/Japan/305/57) given gemfibrozil once daily from days 4 to 10 after virus exposure
|
![]() View larger version (14K): [in a new window] |
FIG. 1. Effect of gemfibrozil on survival of BALB/c mice with severe influenza. Mice were infected with influenza virus A/Japan/305/57 and treated with vehicle or 20 mg/kg, 40 mg/kg, or 60 mg/kg gemfibrozil once daily from days 4 to 10 after virus exposure. These results are from a single experiment (vehicle, n = 9; 20 mg/kg, n = 10; 40 mg/kg, n = 9; 60 mg/kg, n = 10).
|
![]() View larger version (9K): [in a new window] |
FIG. 2. Effect of 60 mg/kg gemfibrozil on survival of BALB/c mice with severe influenza. Mice were infected with influenza virus A/Japan/305/57 and treated with vehicle or 60 mg/kg gemfibrozil once daily from days 4 to 10 after virus exposure. These results are pooled from four separate experiments (vehicle, n = 50; gemfibrozil, n = 46). Statistical analysis using the log rank test determined the increase in survival time with gemfibrozil treatment to be significant (P = 0.0026).
|
|
View this table: [in a new window] |
TABLE 2. Weight loss in mice infected with influenza virus (A/Japan/305/57) given 60 mg/kg gemfibrozil once daily from days 4 to 10 after virus exposure
|
![]() View larger version (13K): [in a new window] |
FIG. 3. Effect of 60 mg/kg gemfibrozil on survival of BALB/c mice with severe systemic inflammation. Mice were given 30 mg/kg LPS (E. coli serotype O111:B4) and treated with a single dose of vehicle or 60 mg/kg gemfibrozil either 1 h before or 2 h after LPS injection. These results are from a single experiment; n = 20 for each group. Statistical analysis using the log rank test determined that a single dose of 60 mg/kg gemfibrozil 1 h before LPS significantly increased survival (P < 0.001).
|
|
|
|---|
agonists (2).
These agents have recently been shown to have an additional activity. They reduce inflammation at its source, decreasing serum levels of tumor necrosis factor (TNF) and gamma interferon (17), as well as interleukin-6 (IL-6) (24) in patients with atherosclerosis, with more recent studies demonstrating that lower serum TNF is a reflection of reduced secretion of TNF from peripheral blood mononuclear cells (28) and adipocytes (27). They also promote production of the anti-inflammatory cytokine IL-4 and protect against experimental autoimmune encephalitis, the animal model of multiple sclerosis (16). In the atherosclerosis of experimental diabetes, fibrates reduced superoxide production and expression of nuclear factor
B (NF-
B), which regulates inflammation (4). Further, LPS-stimulated astrocyte production of TNF, IL-1ß, IL-6, and nitric oxide was inhibited by fibrates (25).
In recent years, the threat of H5N1 influenza virus infection has generated increased interest in the mechanism of illness and fatality in influenza disease. Nearly 20 years ago (7) it was proposed, because of its clinical similarities to malaria, that influenza disease arose from excessive systemic release of proinflammatory cytokines. Gradually the idea took root for various influenza virus subtypes (19), particularly for more pathogenic strains. For example, H5N1/97 viruses induce much higher transcription of the TNF gene than do H3N2 or H1N1 viruses (5), TNF-related apoptosis-inducing ligand and TNF mRNA are up-regulated in human monocyte-derived macrophages infected with H5N1/97 virus (29), and high levels of inflammatory cytokines and chemokines are associated with a fatal outcome (10). Moreover, a reconstructed version of the strain of influenza virus responsible for massive human mortality in 1918-1919 has recently been reported to induce a strong proinflammatory cytokine response, including TNF, during the fatal infections it causes in mice (13) and monkeys (14).
Although we used an influenza virus subtype that is no longer circulating, the H2N2 influenza virus subtype caused a pandemic in 1957 with clinical manifestations similar to those of the 1968 (H3N2) pandemic. Further, human H2N2 infections are clinically indistinguishable in the early phase of infection from those caused by the prevailing human influenza virus subtypes H3N2 and H1N1 (26). In this sense, the clinical similarity between H2N2 and H3N2 infections makes our model relevant to the current circulating H3N2 influenza virus subtypes and thus relevant to the treatment of severe seasonal influenza. With the threat of an H5N1 influenza pandemic, finding novel ways to reduce mortality in pathogenic avian influenza is a priority. While our model might also be relevant to avian influenza, since the early phase of human H2N2 infections is also clinically indistinguishable from early-phase H5N1 infection (26), we have plans to extend the findings of this study to determine if gemfibrozil also decreases mortality in H5N1 influenza virus infections in mice.
Although gemfibrozil is as yet untried in human influenza, it could have two major advantages. First, it is already approved for daily human use, albeit by a different route, to lower plasma lipids and cholesterol (18, 23). Second, from our data, enhanced survival did not depend on giving gemfibrozil before the onset of illness, since treatment begun 4 days after exposure to virus, when mice were already sick and had lost weight, was effective. This implies that gemfibrozil has the potential to be a treatment rather than a preventative in human disease, allowing limited stocks of the drug to be focused where required in a pandemic. The idea that severe systemic inflammatory disease arises through overproduction of proinflammatory cytokines in influenza (6, 8) now also has general acceptance. This made it attractive to test the effects of gemfibrozil on influenza, a disease acknowledged to operate through these cytokines. No fibrate appears to have been previously tested against an infectious disease.
Using anti-inflammatory agents against influenza is a recent suggestion, with Fedson proposing statins as a prophylaxis and treatment for an influenza pandemic (11, 12). While the concept has been gaining favor (20), no direct data are as yet available. Since the aim of this study was to find an agent useful in animals already sick from influenza, our initial trials included simvastatin at the human daily maintenance dose for lowering blood lipids. Unlike gemfibrozil, simvastatin had a negligible effect on sick animals under these conditions. Nevertheless, the use of statins in influenza, including as prophylactic agents, warrants closer examination, since the epidemiological data that support the protective effects of statins arose from sepsis patients who were already taking these drugs at the time they became ill (1).
We also found that gemfibrozil gave some protection, in survival terms, against the severe systemic inflammatory illness that results from administering LPS (Fig. 3). This suggests that it could also protect against other similar inflammatory conditions as well as influenza. In addition, activity against LPS implies that at least a major effect of gemfibrozil against influenza is to inhibit inflammatory cytokines, not the virus. Nevertheless, the mechanisms by which gemfibrozil exerts the effects we have observed are yet to be elucidated. Our next priorities are to examine whether inflammatory cytokine production is inhibited and to determine if gemfibrozil has any direct antiviral action.
Although the published mouse 50% lethal dose of gemfibrozil is 3,162 mg/kg for a single orally administered dose, and 300 mg/kg/day of gemfibrozil, given for 3 or 12 months, is well tolerated in rats (15), we are wary of extrapolating this to pathogen-infected mice without further study. Also, it is yet to be ascertained how outcomes of oral and i.p. dosages compare. As noted elsewhere (3), gemfibrozil reduces the production of inflammatory cytokines, molecules that are involved in the host response against pathogens as well as disease pathophysiology (6). The literature on the effects of anti-TNF agents in rheumatoid arthritis provides an example of this useful and harmful duality of inflammatory cytokines (22). This possibility in influenza will be examined in several ways, including further parallel studies using gemfibrozil against influenza and LPS toxicity. The LPS model is useful because it involves no cytokine-susceptible infectious agent and causes pathology only through excessive cytokine production.
In summary, if these results translate to human infections, gemfibrozil may prove to be a readily testable and useful treatment for influenza, both in high-risk individuals with the currently circulating influenza virus strain and in any pandemic resulting from an antigenic shift in a subtype, including the current avian H5N1 influenza virus.
Published ahead of print on 11 June 2007. ![]()
A.B. and L.A. contributed equally to this study. ![]()
|
|
|---|
but not by PPAR
activators. Nature 393:790-793.[CrossRef][Medline]This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2010 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»