A preclinical assessment to repurpose drugs to target type 1 diabetes-associated type B coxsackieviruses

A.-B. Sioofy-Khojine1, A. Honkimaa1 and H. Hyöty1,2

What’s new?

• Type B coxsackieviruses are associated with type 1 diabetes.
• Although many drugs have shown anti-enterovirus properties, no antiviral drug has currently been licensed specifically for the treatment of enterovirus infections.
• Repurposing antiviral drugs represent an attractive opportunity in type 1 diabetes prevention trials.
• Dose and spectrum tests against type B coxsackieviruses showed that Hizentra, enviroxime, ribavirin, pleconaril and favipiravir would be the best candidates for this strategy.
• Ribavirin and itraconazole were less effective but were able to impair virus replication.
• Our results can be used to design clinical trials aimed at eradication from the pancreas of persistent infection and preventing/treating type 1 diabetes.
• Using antiviral drugs to treat children with type 1 diabetes will help to elucidate the involvement of enteroviruses, including coxsackieviruses, over the course of the condition.

Abstract

Aim To screen several antiviral drugs systematically for their efficacy against type B coxsackieviruses.
Methods Ten drugs with different antiviral mechanisms were analysed for their efficacy against prototype strains of type B coxsackieviruses in A549 cells. Cell viability was quantified in fixed cells using a colorimetric assay. Median effective dose was interpolated from the triplicated experiments and the dose–response curves were generated for each drug–virus combination. Drug cytotoxicity was similarly quantified and selectivity indices calculated.
Results Hizentra, pleconaril, fluoxetine, norfluoxetine, ribavirin, favipiravir and guanidine hydrochloride were able to abrogate infection by all tested viruses, with the exception of complete inefficacy of pleconaril against coxsackievirus B3 and favipiravir against coxsackievirus B2. The effective doses for Hizentra, enviroxime, ribavirin and pleconaril were clearly below their therapeutic serum concentrations, while the effective concentrations of fluoxetin, norfluoxetine and itraconazole exceeded their therapeutic serum concentrations. Lovastatin and azithromycin did not efficiently block type B coxsackieviruses.
Conclusion Hizentra, enviroxime, pleconaril, ribavirin and favipiravir are effective against type B coxsackieviruses in vitro in their therapeutic serum concentrations. These antiviral drugs are therefore attractive candidates for type 1 diabetes prevention/treatment trials. They can also be used in other clinical conditions caused by type B coxsackieviruses.

Introduction

Type B coxsackieviruses (CVBs) belong to the species enterovirus B within the family Picornaviridae, icosahedral non-enveloped RNA viruses with mRNA polarity. CVBs include six members (CVB1–6) causing, for example, the common cold, aseptic meningitis, herpangina, myocarditis, pericarditis, and multi-organ life-threatening infections in young infants. They constitute ~24% of all reported enteroviruses by the National Enterovirus Surveillance System 1970–2018 [1]. They have also been linked to chronic dilated cardiomyopathy and chronic fatigue syndrome [2–4].
Type 1 diabetes has also been linked to CVBs [5–7]. It has been suggested that a persistent infection in insulin-producing pancreatic β cells could contribute to the development of type 1 diabetes [8,9]; therefore, those antiviral drugs that are effective against CVBs could offer an option for preventing/treating type 1 diabetes by eradicating such infections. In fact, to assess this opportunity, people with newly diagnosed type 1 diabetes are currently being recruited into the DiViDIntervention trial in Norway (EudraCT No. 2015 003350-1), a randomized controlled trial using antiviral drugs
Currently, no antiviral drug has been licensed for the treatment of enterovirus infections, as pleconaril did not receive approval for treatment of the common cold. Nevertheless, pleconaril does reduce mortality in severe enterovirus infections in young infants [11]. In addition, certain drugs that are prescribed for other indications have shown anti-enterovirus activity in vitro, but the effect of these compounds against different CVB members has not been systematically studied. The aim of the present study, therefore, was to evaluate the possibility of repurposing clinically used drugs, which have shown activity against at least one enterovirus strain in vitro, to block the infection of all prototype strains of CVB in cell culture. The aim was to identify drugs with broad anti-CVB activity, to facilitate their possible use in the treatment of severe and/or persistent CVB infections and in clinical trials evaluating the possible role of CVB in type 1 diabetes.
We selected drugs targeting different steps in the viral replication cycle, including pleconaril (a pocket factor replacer and uncoating blocker of enteroviruses), enviroxime (enterovirus 3A blocker), ribavirin (a nucleoside analogue), Hizentra (an immunoglobulin that contains neutralizing CVB antibodies), favipiravir (an inhibitor of RNA-dependent RNA polymerases of RNA viruses), fluoxetine (commercially known as Prozac) and its active metabolite norfluoxetine (both potential enterovirus 2C blockers), itraconazole (enterovirus 3A blocker through targeting cellular oxysterolbinding protein), azithromycin (interferon and interferon-stimulated gene upregulator), and lovastatin (possible viral receptor downregulator or disruptor of the membranes of viral replication organelles). Guanidine hydrochloride (an enterovirus 2C ligand) was used as a positive control.

Methods

Antiviral activity assay

Mycoplasma-free A549 (human lung carcinoma epithelial, ATCC) cells were used. Cells were maintained in Ham′s F12 media (Sigma-Aldrich, St Louis, MO, USA) with 10% fetal calf serum (Gibco, Waltham, MA, USA) and penicillin-streptomycin (Gibco) in a humidified chamber with 5% CO2 at 37°C. Cell monolayers were infected in 96 well plates (Nunc, Roskilde, Denmark) with 95% confluency in the absence of antibiotics with individual CVB (ATCC prototype strains, Wesel, Germany; Table S1) with different concentrations of drugs and 10 000 tissue culture infectivity dose 50% of each virus (quantified using the Reed and Munch method). Viable plastic-bound cells were fixed-stained 48 h after infection using formaldehyde (0.5%)-crystal violate (0.05%; Electron Microscopy Sciences, Oy FF-Chemicals Ab, Haukipudas, Finland) for 20 min at room temperature. The intensity of the colour was identified by solubilizing the stained cells using 1% sodium dodecyl sulphate in phosphate-buffered saline for 10 min at room temperature and measuring the absorbance of each well at 595 nm using an ELISA plate reader (Perkin Elmer Victor2V Multilabel Counter 1420-040). The cell viability (%) was calculated comparing the absorbance values of the test wells to those of mock-infected wells representing 100% cell viability.
Since Hizentra (human immunoglobulin) and pleconaril interact directly with the virus to block the infectivity, they were incubated with the virus at 37°C for 1 h prior to the infection and throughout the post infection period. As recommended (see Table S2), azithromycin and lovastatin were added to the cells 24 h prior to infection and the other drugs were added on the cells 1 h before virus infection (Table S1). Control wells included no drug or virus. Drug cytotoxicity was also monitored by analysing cells exposed to the drug without any virus for 48 h.

Identification and quantification of antiviral effect

The median effective dose (ED50) that prevents 50% of cell death due to virus infection was calculated from the triplicated experiments for each drug and individual viruses. ED50 directly demonstrates the efficiency of the drug against individual viruses. The graphs were generated using the viability data and concentrations of the drugs used in the study. The built-in log(agonist) vs response statistics package, based on a non-linear regression method in the GRAPHPAD PRISM program (version 5.02), was used to generate the graphs. In some cases, the ED50 could not be calculated since some drugs did not reach 50% blockage of the virus infection. The cytotoxic concentration 50% (CC50; the concentration of the drug that kills half of the cells in uninfected samples) of each drug was also calculated in the absence of the viruses, using a similar approach and different concentrations of each drug. The selectivity index (SI [SI = CC50/ED50]), representing the safety profile of the drug related to antiviral dose, was calculated for each drug (Table 1). The program was not able to calculate the CC50 for lovastatin; therefore, the value of the highest lovastatin concentration used in the assay was used in place of a CC50 value. For all drugs a conservative ‘minimum SI’ is presented using the highest concentration of the ED50 against CVBs. All background information and relevant references for virus strains and the drugs used in the study are reported in Table S1 and S2.

Results

Individual ED50 values for each drug–virus combination are reported in Table 1 and the dose– response curves are reported in Fig. 1. Briefly, Hizentra, pleconaril, fluoxetine, norfluoxetine and guanidine hydrochloride were able to abrogate infection by all tested viruses, with the known exception that pleconaril was not able to block the ATCC strain of CVB3 (Table 1). Ribavirin blocked all viruses but with less than 100% efficacy and was more cytotoxic than any other drug. Lovastatin blocked only CVB2 with lower efficacy than other drugs. Azithromycin blocked only CVB3.
The ED50 values (horizontal broken lines crossing the dose response curves in Fig. 1) for Hizentra, enviroxime, ribavirin and pleconaril were clearly below the serum concentrations reached in clinical treatments (vertical lines in Fig. 1), suggesting that the antiviral effect could be reached by the doses recommended for their routine clinical indications. By contrast, for fluoxetine/norfluoxetine and itraconazole ED50 values were higher than the serum concentrations reached by recommended doses (up to 4.5 and two times higher, respectively). Lovastatin and azithromycin performed poorly and did not even reach the ED50 for many CVB members. By contrast, Hizentra blocked the infection of all CVBs, even at low concentrations, therefore the ED50 value was not calculated as it was far below the recommended doses used in the clinic. Since Hizentra is a pooled antiserum, its antiviral activity is reached mainly by neutralizing antibodies; therefore, lower reactivity to CVB6 reflects a generally lower titre of neutralizing antibodies to this virus, that could be attributable to a low prevalence of CVB6 in the human population.

Discussion

This study characterized the anti-CVB activity of drugs that had already been reported to block some members of enteroviruses but not exclusively tested for activity against infection by all CVB members (proposed to be associated with type 1 diabetes). Hizentra, enviroxime, pleconaril, ribavirin and, in many cases, favipiravir, were highly effective against CVBs in vitro at the concentrations reached in serum at their recommended therapeutic doses (identified by other studies shown in Table 1). It is possible, therefore, that these drugs could provide clinical benefits in acute or persistent CVB infections. Further studies, particularly carefully designed clinical trials, would be needed to explore this possibility among people with type 1 diabetes. This has already started with the DIVIDintervention trial in people newly diagnosed with type 1 diabetes [10]. That unique trial uses a combination of pleconaril and ribavirin, which were both found to be effective against CVBs in the present study.
The advantage of the antiviral drugs tested in the present study is that their safety profile is already characterized. For example, as well as potentially causing anaemia, ribavirin is contraindicated during pregnancy and in the male partners of women planning pregnancy because of its teratogenic potential [12]. Fluoxetine is not licensed in children below the age of 7 years because of a greater risk of suicide during the first few months of treatment and it may also disturb glucose metabolism in people with diabetes requiring adjustments to insulin dosage [12]. Fluoxetine may also induce QT prolongation when used together with azithromycin or itraconazole [12]. Lovastatin has been contraindicated for use with itraconazole because of the substantial increased risk of developing myopathy, rhabdomyolysis and acute renal failure [12]. Immunoglobulin (Hizentra) concentrates, which have been used off-label for measles prophylaxis and for the prevention of poliomyelitis before the polio vaccine era, is known to be associated with thromboembolism [12].
In conclusion, the present study identified several drugs (Hizentra, enviroxime, pleconaril, ribavirin and favipiravir) that could be beneficial in the treatment of CVB infections and which could be useful in type 1 diabetes intervention trials.

References

[1] CDC-USA (May 10, 2018) The National Enterovirus Surveillance System (NESS), USA. Available at https://www.cdc.gov/surveillance/ness/pubs.html. Last accessed 6 June 2019.
[2] Bowles NE, Bayston TA, Zhang HY, Doyle D, Lane RJ, Cunningham L et al. Persistence of enterovirus RNA in muscle biopsy samples suggests that some cases of chronic fatigue syndrome result from a previous, inflammatory viral myopathy. J Med 1993; 24:145–160.
[3] Klingel K, Hohenadl C, Canu A, Albrecht M, Seemann M, Mall G et al. Ongoing enterovirusinduced myocarditis is associated with persistent heart muscle infection: quantitative analysis of virus replication, tissue damage, and inflammation. Proc Natl Acad Sci U S A 1992; 89:314–318.
[4] Cunningham L, Bowles NE, Lane RJ, Dubowitz V, Archard LC. Persistence of enteroviral RNA in chronic fatigue syndrome is associated with the abnormal production of equal amounts of positive and negative strands of enteroviral RNA. J Gen Virol 1990; 71 ( Pt 6):1399–1402.
[5] Sioofy-Khojine AB, Lehtonen J, Nurminen N, Laitinen OH, Oikarinen S, Huhtala H et al. Coxsackievirus B1 infections are associated with the initiation of insulin-driven autoimmunity that progresses to Favipiravir type 1 diabetes. Diabetologia 2018; 61:1193–1202.
[6] Oikarinen S, Tauriainen S, Hober D, Lucas B, Vazeou A, Sioofy-Khojine A et al. Virus antibody survey in different European populations indicates risk association between coxsackievirus B1 and type 1 diabetes. Diabetes 2014; 63: 655–662.
[7] Laitinen OH, Honkanen H, Pakkanen O, Oikarinen S, Hankaniemi MM, Huhtala H et al. Coxsackievirus B1 is associated with induction of beta-cell autoimmunity that portends type 1 diabetes. Diabetes 2014; 63:446–455.
[8] Krogvold L, Edwin B, Buanes T, Frisk G, Skog O, Anagandula M et al. Detection of a low-grade enteroviral infection in the islets of langerhans of living patients newly diagnosed with type 1 diabetes. Diabetes 2015; 64:1682–1687.
[9] Richardson SJ, Rodriguez-Calvo T, Gerling IC, Mathews CE, Kaddis JS, Russell MA et al. Islet cell hyperexpression of HLA class I antigens: a defining feature in type 1 diabetes. Diabetologia 2016; 59: 2448–2458.
[10] EU Clinical Trials Register (EudraCT): European Medicines Agency © 1995-2018 | 30 Churchill Place, Canary Wharf, London E14 5EU. EudraCT Number: 2015-003350-41, The Diabetes Virus Detection and Intervention Trial (DiViDIntervention). Available at
[11] Rotbart HA, Webster AD; Pleconaril Treatment Registry Group. Treatment of potentially lifethreatening enterovirus infections with pleconaril. Clin Infect Dis 2001; 32:228–235.
[12] ©2018 PDR L (2018) Drug Summary. Available at https://www.pdr.net/browse-by-drug-name, Last accessed 11 November 2018.