|Home | About | Journals | Submit | Contact Us | Français|
The aminoglycoside, geneticin (G418), was recently shown to have antiviral activity against bovine viral diarrhea virus (BVDV). Since BVDV, dengue virus (DENV) and yellow fever virus (YFV) all belong to the Flaviviridae family, it seemed possible that a common step in their life cycle might be affected by this aminoglycoside. Here it is shown that geneticin prevented the cytopathic effect (CPE) resulting from DENV-2 infection of BHK cells, in a dose-dependent manner with an EC50 value of 3±0.4 μg/ml. Geneticin had no detectable effect on CPE caused byYFV in BHK cells. Geneticin also inhibited DENV-2 viral yield with an EC50 value of 2±0.1 μg/ml and an EC90 value of 20±2 μg/ml. With a CC50 value of 165±5 μg/ml, the selectivity indexof anti-DENV activity of geneticin in BHK cells was established to be 66. Furthermore, 25 μg/ml of geneticin nearly completely blocked plaque formation induced by DENV-2, but not YFV. In addition, geneticin, inhibited DENV-2 viral RNA replication and viral translation. Gentamicin, kanamycin, and the guanidinylated geneticin showed no anti-DENV activity. Neomycin and Paromomycin demonstrated weak antiviral activity at high concentrations. Finally, aminoglycoside-3′-phosphotransferase activity of neomycin-resistant gene abolished antiviral activity of geneticin.
Dengue infection is caused by one of four serotypes of dengue virus (DENV), which is a member of the Flaviviridae family. It occurs in many tropical and subtropical regions and has expanded over the last 30 years to include more than 100 countries (1). There are about 50 million cases of DENV infection annually (Anonymous, 2000).
The epidemiological evidence indicates that immunity to one serotype of DENV increases the chance of a more severe disease upon infection with a second serotype by about ten-fold (a process known as antibody-dependent enhancement of infection (ADE) (Kurane et al., 1994). Although a direct link between ADE and severity of the disease is yet to be established, the concerns that ADE will occur among vaccines, making vaccinated individuals more susceptible to severe disease, have hampered the development of monovalent dengue vaccines. Presently, tetravalent dengue vaccines, targeting all four serotypes, are being developed (Raviprakash et al., 2008), which should potentially abolish ADE, unless virus will continue mutating into new serotypes.
Some aminoglycosides are considered to have antiviral activities. Hygromycin B was shown to inhibit replication of herpes simplex virus (Lacal et al., 1983), mouse hepatitis virus (Macintyre et al., 1991a,b), HIV type 1 (Gatti et al., 1998), influenza virus (Ghendon et al., 1981), and both encephalomyocarditis virus and Semliki forest virus (Lacal et al., 1980). Neomycin and recently developed neomycin analogs were also demonstrated to inhibit HIV replication and viral entry (Zapp et al., 1993; Herold & Spear, 1994; Herold et al., 1994; Hung et al., 2002; Litovchick et al., 2000, 2001; Langeland et al., 1986, 1987).
We previously demonstrated that the aminoglycoside geneticin, although structurally distinct from hygromycin B, inhibited bovine viral diarrhea virus (BVDV), which belongs to the Flaviviridae family (Collett et al., 1988). These results allowed us to hypothesize that geneticin might also have antiviral activity against other members of the Flaviviridae family, such as dengue virus and yellow fever virus.
In this study, we demonstrated that geneticin, a neomycin analog widely used to select for transfected eukaryotic cells (Santerre et al., 1984; Danielson et al., 1989), inhibited DENV-2 induced cytopathology, viral titers, and viral RNA replication and translation. However, surprisingly, geneticin had no effect on cytopathology and proliferation of YFV, demonstrating selectivity of this aminoglycoside to DENV. Close structural analogs of geneticin, such as gentamicin, kanamycin, and guanidilated geneticin at Rings I and II (gG418), had no effect on DENV proliferation, suggesting that the structural specificity of geneticin’s antiviral activity is due to Ring I and II of this aminoglycoside.
The morphology of BHK cells infected with DENV-2 (MOI of 1) was similar to mock-infected control cells up to 48 hrs post-infection (hpi). However, by 72 and 96 hpi, a large number of infected cells were found to shrink and lose their morphological characteristics. By 96 hrs post-infection, about 90% of monolayer was destroyed.
To measure cell viability at 96 hrs in mock-infected and DENV-2-infected BHK cells treated with or without geneticin, we used Almar Blue assay (see Methods). We determined that geneticin reduced viability of BHK cells with a CC50 value of 165±5 μg/ml (Fig. 1, bottom panel). Geneticin prevented DENV-induced CPE (MOI of 1) and increased cell viability of infected cells with an EC50 value of about 3±0.4 μg/ml (Fig. 1, top panel). However, geneticin had no protective activity against CPE induced by YFV (Fig. 1, top panel). Morphologically, cells treated with EC90–100 concentration of geneticin (25 μg/ml) were shown to be indistinguishable from mock-infected controls (data not shown). Interestingly, other aminoglycosides structurally distinct from geneticin, such as hygromycin B, butirosin A, streptozocin, netilmicin, sisomicin, amikacin, tobramycin, ribostamycin, and apramycin had no protective activity against DENV-2 -induced CPE.
To further characterize the antiviral property of geneticin, the effect of the drug on virus yield was determined. We demonstrated that geneticin inhibited the yield of viral titers of DENV-2 even after 72 hpi with an EC50 value of 2±0.1 μg/ml and an EC90 value of about 20±2 μg/ml (Fig. 2A), which was consistent with antiviral activity of geneticin in the CPE assay (Fig. 1). We then used the EC90–100 concentration of geneticin (25 μg/ml) to determine the effect of the drug on DENV-induced plaque formation in BHK cells. Thus, we demonstrated (Fig. 2B) that geneticin inhibited the number of plaques produced by the virus, and the sizes the remaining plaques were shown to be significantly reduced, suggesting that the drug inhibited proliferation and spread of the virus. However, in agreement with CPE studies, geneticin had no inhibitory effect on YFV-induced plaque formation (data not shown). This further indicates the difference in antiviral activity of geneticin between these related viruses.
To understand the mechanism underlying geneticin inhibition of DENV-induced CPE and virus yield reduction, we investigated the effects of geneticin on viral RNA synthesis and protein production. RT-qPCR was employed to measure intracellular accumulation of viral RNA 3, 6, 12, 24, and 48 hpi with DENV-2 (m.o.i. of 1) in the presence or absence of 25 μg/ml geneticin. In the first 6 hpi, viral RNA accumulation did not display significant difference between treatment groups, suggesting that the drug did not block the early events of DENV infection, such as viral entry. Our data show that geneticin inhibited viral RNA synthesis by 40% at 12 hpi, and by 24 and 48 hpi; the addition of geneticin resulted in nearly 90% decrease in viral RNA, compared to infected control without geneticin ( Fig. 3A).
DENV, like other flaviviruses, expresses a polyprotein, which is consequently processed to individual viral proteins. Therefore, to assess the effect of geneticin on translation of viral proteins, we used viral E protein as a marker of DENV-2 translation at 6, 12, 24, and 48 hpi at an MOI of 1. Interestingly, no E protein band was observed until 48 hpi (Fig 3B). Not surprisingly, geneticin treatment (25 μg/ml) inhibited formation of E protein by about 80% (Fig. 3B and C). Moreover, the absence of additional E protein bands in geneticin-treated sample further suggest that geneticin does not affect processing of E protein of DENV-2.
We also tested the antiviral activity of structurally similar geneticin analogs (Fig 4A), such as gentamicin and kanamycin A. Ring II is structurally similar in all of these aminoglycosides. Geneticin and gentamicin also share structural identity of Ring III. However, among all tested aminoglycosides, only geneticin significantly improved cell protection against DENV-2 (Fig. 1, top and Fig. 4B). Interestingly, neomycin and paromomycin, with high structural homology with geneticin in Rings I and II, showed modest anti-DENV activity at concentrations between 300 and 1000 μg/ml. However, antiviral activity of those aminoglycosides even at 1000 μg/ml never reached full antiviral activity observed with 12 μg/ml of geneticin.
This modest activity of neomycin and paramomycin suggest that Rings I and II might be responsible for antiviral activity of geneticin. Therefore, we modified geneticin by guanidilating primary amines of Rings I and II of geneticin to test the role of primary amines in the anti-DENV activity of geneticin. Our results showed that guanidinylated geneticin (gG418) (Fig 5A, structure) does not exert any antiviral activity against DENV-2 (Fig. 5A). Likewise, it appears that YFV and BVDV were not blocked by gG418 either (data not shown).
It has been well established that cells containing neomycin-resistant gene (neor), encoding aminoglycoside 3′-phosphotransferase, are insensitive to the toxic concentrations of geneticin, due to the 3′-phosphorylation of geneticin (Thompson et al., 1999). To test if 3′-phosphorylated geneticin has antiviral activity, we created neor-containing BHK cells. The neor-containing BHK cells were viable in the presence of geneticin up to 1000 μg/ml (Fig. 5B, bottom). However, the antiviral activity against DENV-2 of geneticin was completely abolished (Fig. 5B, top).
In the present study, we demonstrated that geneticin selectively inhibits DENV-2 proliferation by 1) protecting BHK cells against the cytopathic effect of DENV-2, with an EC50 value of about 3±0.1 μg/ml (Fig. 1); 2) reducing the viral yield with an EC50 value of 2±0.1 μg/ml and EC90 value of about 20±2 μg/ml (Fig. 2A ), consistent with the antiviral activity of geneticin in CPE assay; 3) inhibiting DENV-2 plaque formation in both the number and the size of the plaques at its EC90 concentration of 25 μg/ml (Fig. 2B); 4) blocking DENV-2 RNA and protein synthesis (Fig. 3). The selectivity index of anti-DENV activity of geneticin in BHK cells was established to be 66. However, YFV was not inhibited by geneticin in the same cell line, where drug showed clear anti-DENV-2 activity.
The molecular mechanism of antiviral activity of geneticin remains unclear. Geneticin has shown antiviral activity against different RNA viruses in a variety of cell lines, such as MDBK (Birk et al., 2008), BHK, bottle neck dolphin skin cells (data not shown), and rabbit kidney cell (manuscript in preparation), suggesting that geneticin-mediated antiviral mechanism is cell type-independent. Therefore, the antiviral activity of geneticin is more likely dependent on the commonly shared cellular antiviral mechanisms. Alternatively, the drug can selectively affect different viral structures and functions. The other possibility is that geneticin may directly inhibit DENV through specifically interacting with the viral structural or nonstructural proteins or binding to viral RNA. Thus, further exploration of virus selectivity of geneticin and comparison of antiviral mechanisms for different viruses will be required.
Here, we show evidence that geneticin inhibits DENV-2 life cycle at the stages different from that of BVDV (Birk et al., 2008). DENV-2 RNA accumulation was not inhibited by 6 hpi hours of infection (Fig. 3A), suggesting that DENV-2 entry into the cells was not blocked by geneticin, which is similar to the effect of the drug on BVDV infection of MBDK cells. Geneticin inhibited BVDV assembly and release and protected MDBK cells against BVDV-induced CPE and reduced viral yield, without inhibiting BVDV protein and RNA synthesis. Distinctly different from its anti-BVDV activity, geneticin impeded the production of DENV-2 RNA already at 12 hpi by about 40% and prevented the synthesis of viral E protein and viral RNA at 48 hpi by 80% and 90%, respectively (Fig. 3). These data could suggest a possibility that geneticin, being an inhibitor of protein synthesis (Cabanas et al., 1978; Eustice & Wilhelm, 1984), inhibits viral translation, resulting in a decrease of viral RNA synthesis. However, this hypothesis is not supported by the fact that other translational inhibitors, such as kanamycin and gentamycin, had no effect of DENV infection. Moreover, although it was recently demonstrated that both geneticin and gentamicin, and other related aminoglycosides could suppress stop codons and induce frame-shifting (Lai et al., 2004; Konno et al., 2004; Borovinskaya et al., 2007a,b), only geneticin was shown to inhibit DENV-2, suggesting that the antiviral activity of geneticin is different from inhibition or deregulation of translation. Alternatively, it is also possible that geneticin inhibited viral machinery or RNA folding, preventing accumulation of viral RNA, and then resulting in overall decrease in viral proteins.
Structure and function analysis of geneticin and its close structural analogs, including Gentamicin, kanamycin, neomycin and paromomycin, revealed that DENV-2 was responsive to geneticin at an EC50 and an EC90 approximately 2–3 and 20 μg/ml, respectively. Neomycin and paromomycin showed modest anti-DENV-2 activities at 300 through 1000 μg/ml, which did not result in the antiviral effect observed after treatment with geneticin at a 100-fold lower concentration. Both gentamicin and kanamycin did not show antiviral properties (Fig. 4).
As shown in Fig. 4, panel A, Ring II is structurally similar in geneticin, gentamicin and kanamycin, paromomycin and neomycin. Geneticin and gentamicin also share structural identity of Ring III. Furthermore, Ring I of paromomycin and neomycin share the same structure with geneticin, with an exception of functional group at C6, where geneticin has a secondary alcohol, and neomycin and paromomycin have primary amine and primary alcohol, respectively. Based on the fact that only geneticin, neomycin and paromomycin showed anti-DENV activity, we concluded that Ring I is the essential component in geneticin as an anti-DENV-2 agent, and Ring II also play a modest, but important, role in antiviral activity. This is somewhat different from the ability of those aminoglycosides to inhibit protein synthesis via binding to the A site of 30S ribosomes (Benveniste & Davies, 1973) where the ability of geneticin binding to RNA is assigned to its Ring II (Vicens & Westhof, 2003). To confirm the role of geneticin Ring I and II in anti-DENV-2, a novel molecule was generated by guanidinylating the primary amines of geneticin Ring I and II (Fig. 4, panel C). The modified geneticin, gG418, indeed had no antiviral activity, suggesting chemical selectivity of geneticin against DENV-2. Furthermore, geneticin administration to the neor-containing BHK cells showed that phosphorylation of geneticin diminished its antiviral activity, implying that the hydroxyl group in the 3′ –OH of Ring I of geneticin contributed significantly to the anti-DENV activity.
Aminoglycosides are known to be active against HIV (Zapp et al., 1993; Tassew & Thompson, 2003), HCMV (Lobert et al., 1996), HSV (Langeland et al., 1986, 1987; Garcin et al., 1990), and HDV (Rogers et al., 1996; Chia et al., 1997). Thus, the natural and synthesized aminoglycoside-based antibiotics can be a potent tool to explore and develop new specific antiviral drugs against pathogenic RNA viruses. In addition, because geneticin and Gentamicin are the most efficacious therapeutic agents for the treatment of 15% of severe inherited genetic diseases (Yang et al., 2007) are associated with nonsense mutation, it might be foreseeable to develop geneticin or its analog as anti-DENV therapeutics to inhibit viremia and DENV-induced clinical complications. Overall, our data suggest that geneticin represents a novel lead compound for broad-spectrum virus-selective antivirals.
This work was supported by NIH/NIDA grant K01 DAO18262 and NIH 5U01AI061441.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.