Enterovirus strain and type-specific differences in growth kinetics and virus-induced cell destruction in human pancreatic duct epithelial HPDE cells.

Enterovirus infections have been suspected to be involved in the development of type 1 diabetes. However, the pathogenetic mechanism of enterovirus-induced type 1 diabetes is not known. Pancreatic ductal cells are closely associated with pancreatic islets. Therefore, enterovirus infections in ductal cells may also affect beta-cells and be involved in the induction of type 1 diabetes. The aim of this study was to assess the ability of different enterovirus strains to infect, replicate and produce cytopathic effect in human pancreatic ductal cells. Furthermore, the viral factors that affect these capabilities were studied. The pancreatic ductal cells were highly susceptible to enterovirus infections. Both viral growth and cytolysis were detected for several enterovirus serotypes. However, the viral growth and capability to induce cytopathic effect (cpe) did not correlate completely. Some of the virus strains replicated in ductal cells without apparent cpe. Furthermore, there were strain-specific differences in the growth kinetics and the ability to cause cpe within some serotypes. Viral adaptation experiments were carried out to study the potential genetic determinants behind these phenotypic differences. The blind-passage of non-lytic CV-B6-Schmitt strain in HPDE-cells resulted in lytic phenotype and increased progeny production. This was associated with the substitution of a single amino acid (K257E) in the virus capsid protein VP1 and the viral ability to use decay accelerating factor (DAF) as a receptor. This study demonstrates considerable plasticity in the cell tropism, receptor usage and cytolytic properties of enteroviruses and underlines the strong effect of single or few amino acid substitutions in cell tropism and lytic capabilities of a given enterovirus. Since ductal cells are anatomically close to pancreatic islets, the capability of enteroviruses to infect and destroy pancreatic ductal cells may also implicate in respect to enterovirus induced type 1 diabetes. In addition, the capability for rapid adaptation to different cell types suggests that, on occasion, enterovirus strains with different pathogenetic properties may arise from less pathogenic ancestors.

Isolation of saffold virus type 2 in green monkey kidney cells.

Saffold viruses (SAFV) have been discovered recently and they are classified into Theilovirus species in genus Cardiovirus in the Picornaviridae family. SAFV, especially those belonging to the genotype 2, have been difficult to propagate in laboratory cell lines. This study describes the successful isolation of an efficiently growing SAFV-2 strain directly from a stool specimen by standard virological methods. The availability of SAFV isolates that can be propagated to high titers is crucial to the future studies on pathogenesis and epidemiology of these novel human viruses.

Cellular tropism of human enterovirus D species serotypes EV-94, EV-70, and EV-68 in vitro: implications for pathogenesis.

Enterovirus 94 (EV-94) is an enterovirus serotype described recently which, together with EV-68 and EV-70, forms human enterovirus D species. This study investigates the seroprevalences of these three serotypes and their abilities to infect, replicate, and damage cell types considered to be essential for enterovirus-induced diseases. The cell types studied included human leukocyte cell lines, primary endothelial cells, and pancreatic islets. High prevalence of neutralizing antibodies against EV-68 and EV-94 was found in the Finnish population. The virus strains studied had wide leukocyte tropism. EV-94 and EV-68 were able to produce infectious progeny in leukocyte cell lines with monocytic, granulocytic, T-cell, or B-cell characteristics. EV-94 and EV-70 were capable of infecting primary human umbilical vein endothelial cells, whereas EV-68 had only marginal progeny production and did not induce cytopathic effects in these cells. Intriguingly, EV-94 was able to damage pancreatic islet ß-cells, to infect, replicate, and cause necrosis in human pancreatic islets, and to induce proinflammatory and chemoattractive cytokine expression in endothelial cells. These results suggest that HEV-D viruses may be more prevalent than has been thought previously, and they provide in vitro evidence that EV-94 may be a potent pathogen and should be considered a potentially diabetogenic enterovirus type.

Vitronectin receptors, alpha v integrins, are recognized by several non-RGD-containing echoviruses in a continuous laboratory cell line and also in primary human Langerhans' islets and endothelial cells.

Previously published data suggest that the RGD-recognizing integrin, alphavbeta3, known as the vitronectin receptor, acts as a cellular receptor for RGD-containing enteroviruses, coxsackievirus A9 (CAV-9) and echovirus 9 (E-9), in several continuous cell lines as well as in primary human Langerhans' islets. As this receptor is also capable of binding the ligands by a non-RGD-dependent mechanism, we investigated whether vitronectin receptors, alpha v integrins, might act as receptors for other echoviruses that do not have the RGD motif. Blocking experiments with polyclonal anti-alphavbeta3 antibody showed that both primary human islets and a continuous laboratory cell line of green monkey kidney origin (GMK) are protected similarly from the adverse effects of several non-RGD-containing echovirus (E-7, -11, -25, -30, -32) infections. In contrast, corresponding studies on primary human endothelial cells showed that the receptor works only for E-25, E-30, E-32 and CAV-9. The inhibitory effect of the antibody was not restricted to prototype strains of echoviruses, as GMK cells infected with several field isolates of the corresponding serotypes were also protected from virus-induced cytopathic effects. Co-localization of virus particles with the receptor molecules in both GMK and primary human endothelial cells was demonstrated by live-cell stainings and confocal microscopy. Remarkably, in spite of similar virus-receptor co-localization and a comparable protective effect of the alphavbeta3 antibody, the entry pathways of the studied virus strains seemed to be divergent.

Amino acids of Coxsackie B5 virus are critical for infection of the murine insulinoma cell line, MIN-6.

It was shown recently that 15 successive passages of a laboratory strain of the Coxsackie B virus 5 in a mouse pancreas (CBV-5-MPP) resulted in apparent changes in the virus phenotype, which led to the capacity to induce a diabetes-like syndrome in mice. For further characterization of islet cell interactions with a passaged virus strain, a murine insulinoma cell line, MIN-6, was selected as an experimental model. The CBV-5-MPP virus strain was not able to replicate in MIN-6 cells in vitro but required adaptation over a few days for progeny production and the generation of cytopathic effects. In order to determine the genetic characteristics required for virus growth in MIN-6 cells, the whole genome of the MIN-6-adapted virus variant was sequenced, and critical amino acids were identified by comparing the sequence with that of a virus strain passaged repeatedly in the mouse pancreas. The results of site-directed mutagenesis demonstrated that only one residue, amino acid 94 of VP1, is a major determinant for virus adaptation to MIN-6 cells.

An enterovirus strain isolated from diabetic child belongs to a genetic subcluster of echovirus 11, but is also neutralised with monotypic antisera to coxsackievirus A9.

An enterovirus strain (designated D207) isolated from a Slovakian diabetic child and originally serotyped as coxsackievirus A9 (CAV-9) was found to cause rapid cytolysis coinciding with severe functional damage of the surviving cells in primary cultures of human pancreatic islets. This finding prompted us to clone the isolate for full-length genome sequencing and molecular characterization as the prototype strain of CAV-9 is known to cause only minimal damage to insulin-producing beta-cells. Based on capsid-coding sequence comparisons, the isolate turned out to be echovirus 11 (E-11). Phylogenetic analyses demonstrated that E-11/D207 was closely related to a specific subgroup B of E-11 strains known to cause uveitis. To study further antigenic properties of isolate E-11/D207 and uveitis-causing E-11 strains, neutralization experiments were carried out with CAV-9- and E-11-specific antisera. Unlike the prototype strains, the isolate E-11/D207 and uveitis-causing E-11 strains were well neutralized with both CAV-9- and E-11-specific antisera. Attempts to identify recombination of the capsid coding sequences as a reason for double-reactivity using the Simplot analysis failed to reveal major transferred motifs. However, peptide scanning technique was able to identify antigenic regions of capsid proteins of E-11/D207 as well as regions cross-reacting with an antiserum raised to CAV-9. Thus, double specificity of E-11/D207 seems to be a real characteristic shared by the phylogenetically closely related virus strains in the genetic subgroup B of E-11.

In vitro toxicity of cereulide on porcine pancreatic Langerhans islets.

Cereulide is a K(+) ionophore cytotoxic and mitochondriotoxic to primary cells and cell lines of human and other mammalian origins. It is a heat-stable, highly lipophilic (logK(ow) 5.96) peptide (1152 g mol(-1)) produced by certain strains of Bacillus cereus, a bacterium connected to emetic food poisonings. In this study the pancreatic toxicity of purified cereulide, and cereulide-containing bacterial extracts, was studied using fetal porcine Langerhans islets in culture. Exposure to 1ngml(-1) of purified cereulide caused necrotic cell death of the islet cells impairing their insulin content within 2 days. Cell extracts of cereulide-positive B. cereus strains connected to food poisoning or isolated from foodstuffs were toxic, corresponding to their measured cereulide content. Extracts of B. cereus strains producing or not producing the B. cereus diarrheal toxin, but no cereulide, were tolerated by the porcine islet cultures up to concentrations 1000-fold higher compared to extracts from strains containing cereulide, and up to exposure times of 7d. Cereulide thus was identified as the B. cereus-produced substance toxic towards porcine fetal Langerhans islets and beta cells.

Diabetogenic effects of the most prevalent enteroviruses in Finnish sewage.

Common enterovirus infections appear to initiate or facilitate the pathogenetic processes leading to type 1 diabetes (T1D) and also sometimes precipitate the clinical disease. We have recently demonstrated that (1) enterovirus-positive islet cells were seen on postmortem pancreatic specimens of several T1D patients but not in the corresponding samples of nondiabetic controls, and (2) several different enteroviruses can be associated with T1D. Enterovirus infections are transmitted from person to person by fecal-oral or respiratory routes, which means that infections usually start from the respiratory or gastrointestinal mucosa. Regardless of the clinical symptoms of the disease, viral replication continues in the submucosal lymphatic tissue for several weeks, up to a couple of months, and during that time the virus is excreted into the feces and translocated to the environment. Monitoring of sewage samples for enteroviruses can be used as a tool in epidemiologic studies of enterovirus. Finland has successfully used environmental control data in poliovirus surveillance for decades. About 24 samples have been collected annually from the Helsinki region, which covers about 20% of the population. In the present study, we have reanalyzed the sewage samples of the years 1993-2004 for nonpolio enteroviruses by inoculating them into five different continuous cell lines known to cover a wide range of serotypes. Isolated strains were identified by RT-PCR and VP1 sequencing. The most commonly detected serotypes were coxsackie B viruses (CBV1-5) and echoviruses (E6, 7, 11, 25, 30). Diabetogenic effects of the most prevalent enterovirus serotypes were studied in primary human beta cells.

A whole cell immunization-derived monoclonal antibody that protects cells from coxsackievirus A9 infection binds to both cell surface and virions.

Coxsackievirus A9 (CAV-9) infects human rhabdomyosarcoma (RD) cells using an unidentified RGD-independent receptor. Monoclonal antibodies were prepared by immunizing mice with intact RD cells and by selecting cells from the cytopathic effect of CAV-9 for protection. Here we describe a monoclonal antibody that binds to host cell plasma membrane and protects cells from virus infection. In addition, binding of the virus to cell monolayers was more efficient in the presence of the antibody, suggesting that the antibody is also capable of recognizing virus particles. Immunoprecipitation and electron microscopy studies with highly purified virus preparations verified binding of the monoclonal antibody to the virus particles. The antibody also recognized coxsackievirus A21 and all three serotypes of poliovirus, but without affecting their infectivity. The amino acid sequence of CAV-9 recognized by the monoclonal antibody was identified by peptide mapping and by producing escape mutants in the presence of the antibody.

Phenotypic and genetic changes in coxsackievirus B5 following repeated passage in mouse pancreas in vivo.

Common enterovirus infections appear to initiate or facilitate the pathogenetic processes leading to type 1 diabetes, and also sometimes precipitate the clinical disease. In experimental infection of mice, coxsackieviruses have shown to have a strong affinity for the exocrine tissue, while even in lethal cases, the islets remain unaffected. The virus strain most intensively studied in this respect is the diabetogenic variant E2 of coxsackievirus B4. In addition, it is known that all six serotypes of coxsackie B viruses can be made diabetogenic by repeated passages in either mouse pancreas in vivo or in cultured mouse beta-cells in vitro. However, the genetic determinants of the phenomenon have not been determined. In the present study, a laboratory strain of coxsackievirus B5 was passaged repeatedly in mouse pancreas in vivo. After 15 passages, the virus phenotype was clearly changed and infection of the variant resulted in a diabetes-like syndrome in mice characterized by chronic pancreatic inflammation together with dysregulation in glucose metabolism, loss of pancreatic acinar tissue, and mild insulitis. In order to characterize the genetic determinants involved in mouse pancreas adaptation, the passaged virus variant together with the parental virus strain was cloned for molecular characterization. The whole genome sequencing of both virus strains revealed only limited differences. Altogether, eight nucleotides were changed resulting in five amino acid substitutions, of which three were located in the capsid proteins.

Mechanisms of coxsackievirus B5 mediated beta-cell death depend on the multiplicity of infection.

Coxsackievirus infections may trigger and accelerate pancreatic beta-cell death, leading to type I diabetes. Unrestricted coxsackievirus B5 replication in cultured beta-cells inoculated with high multiplicity leads to rapid lytic cell death. Evidence from other virus-host cell systems indicates that host cell responses to infection may depend on the multiplicity of infection (MOI). Thus, the aim of this study was to compare the mechanisms of beta-cell death during high versus low multiplicity of coxsackievirus B5 infection. Cultures of highly differentiated mouse insulinoma cells and primary adult human islets were infected with coxsackievirus B5 at multiplicities of >1,000 or <0.5 TCID50 per cell. The results of nuclear morphology and viability stainings, TUNEL staining and electrophoretic DNA fragmentation analysis showed high multiplicity infection to predominantly induce necrosis and transient apoptosis. In low multiplicity culture, however, necrosis was only moderately induced and apoptosis increased steadily with time. This was best demonstrated by a tenfold higher apoptosis/necrosis ratio than after high multiplicity inoculation. Expression of gamma-glutamyl cysteine synthetase increased in both infective cultures but the level of intracellular glutathione permanently depleted only at high multiplicity and recovered fully at low multiplicity. Thus, apoptosis represents an important mechanism of beta-cell death after low multiplicity of coxsackievirus B5 infection. This process is associated with maintenance of a physiological intracellular glutathione profile differing dramatically from the high multiplicity infection during which necrosis dominates and intracellular thiol balance deteriorates. These data suggest that the pattern and mechanisms of coxsackievirus B5 infection induced beta-cell death depend on the MOI.

Mechanisms of beta cell death during restricted and unrestricted enterovirus infection.

Coxsackie B virus (CVB-5) infections potentially trigger and accelerate pancreatic beta cell damage leading to type 1 diabetes. In vivo, all viruses face natural resistance mediated by various host factors which restrict the progression of infection. Thus, the aims of this study were to generate a tissue culture model of restricted coxsackie B virus infection in primary islet cells by preventing the production of viral progeny with a selective inhibitor of viral RNA replication and to investigate the mechanisms of virus-induced islet cell death during productive and restricted infective conditions. Cultured foetal porcine islet cells were infected effectively with the prototype strain of coxsackievirus B5 (CVB-5). Nuclear viability stainings and electron microscopy showed productive infection to result in dominantly necrotic cell death with additional slight induction of apoptosis during the 7 days of follow-up. The restricted conditions were created by addition of guanidine-hydrochloride (G-HCl) into culture medium. At 1 mM concentration, it significantly protected the infected cells from necrosis and thus maintained high viability. This was associated with increased significantly apoptosis. In perifusion analysis, the cellular ability to release insulin was reduced, although the metabolic integrity was preserved as shown by MTT-analysis and cellular ATP levels. These data show that restriction of CVB-5 replication with G-HCl protects islet cells against virus-induced necrosis. However, restriction of viral replication shifts the mechanism of cell death from necrosis toward apoptosis. A slowly progressing subclinical infection of islets could thus lead to increased beta-cell apoptosis.

Enterovirus infection and activation of human umbilical vein endothelial cells.

Gastrointestinal tract associated lymphoid tissue is considered to be the main replication site for enteroviruses. In order to invade tissues to reach pancreatic islets, cardiac muscles, and other secondary replication sites, the virus has to survive circulation in the blood and find a way to get through endothelial cells. In the present study, the susceptibility of human endothelial cells to infections caused by human parechovirus 1 and several prototype strains of enteroviruses, representing different species (human poliovirus, human enterovirus B and C), and acting through different receptor families was examined. Primary endothelial cells isolated from human umbilical vein by collagenase perfusion and also an established human endothelial cell line, HUVEC, were used. Primary endothelial cells were highly susceptible to several serotypes of enteroviruses (coxsackievirus A13, echoviruses 6, 7, 11, 30, and poliovirus 1). However, coxsackievirus A 9 and echovirus 1 infected only a few individual cells while human parechovirus 1 and coxsackie B viruses did not show evidence of replication in primary endothelial cells. In general, primary endothelial cells were more sensitive to infection-induced cytolytic effect than HUVEC. Activation of endothelial cells by interleukin-1beta did not change the pattern of enterovirus infection. Immunofluorescence stainings of infected primary endothelial cells showed that expression of activation markers, E-selectin, and intercellular adhesion molecule-1, was clearly increased by several virus infections and the former molecule also by exposing cells to UV-light inactivated coxsackieviruses. In contrast, human leukocyte antigen-DR expression was not increased by virus infection.

Molecular and biological analysis of echovirus 9 strain isolated from a diabetic child.

The full-length infectious cDNA clone was constructed and sequenced from the strain DM of echovirus 9, which was recently isolated from a 6-week-old child at the clinical onset of type 1 diabetes. Parallel with the isolate DM, the full-length infectious cDNA clone of the prototype strain echovirus 9 Barty (Barty-INF), was constructed and sequenced. Genetic relationships of the sequenced echo 9 viruses to the other members of the human enterovirus type B species were studied by phylogenetic analyses. Comparison of capsid protein sequences showed that the isolate DM was closely related to both prototype strains: Hill and Barty-INF. The only exception was the inner capsid protein VP4 where serotype specificity was not evident and the isolate DM clustered with the strain Hill and the strain Barty-INF with echovirus 30 Bastianni. Likewise, the nonstructural protein coding region, P2P3, of isolate DM was more similar to strain Hill than to strain Barty-INF. However, like echovirus 9 Barty, the isolate DM contained the RGD-motif in the carboxy terminus of capsid protein VP1. By blocking experiments using an RGD-containing peptide and a polyclonal rabbit antiserum to the alpha(v)beta(3)-integrin, it was shown that this molecule works as a cellular receptor for isolate DM. By using primary human islets, it was shown that the isolate DM is capable of infecting insulin-producing beta-cells like the corresponding prototype strains did. However, only isolate DM was clearly cytolytic for beta-cells. The infectious clones that were made allow further investigations of the molecular features responsible for the diabetogenicity of the isolate DM.

Screening enteroviruses for beta-cell tropism using foetal porcine beta-cells.

Primary adult human insulin-producing beta-cells are susceptible to infection by prototype strains of coxsackieviruses (CV) and infection may result in impaired beta-cell function and/or cell death, as shown for coxsackie B virus (CVB) types 4 and 5, or have no apparent immediate adverse effects, as shown for CVA-9. Because of the limited availability of human pancreatic beta-cells, the aim of this study was to find out if foetal porcine pancreatic islets could be used as a substitute in enterovirus (EV) screening. These cells resemble human beta-cells in several biological properties. CVB infection resulted in a rapid progressive decline of insulin content and reponsiveness to insulin release. The amount of virus inoculum sufficient for this destruction was small, corresponding to only 55 infectious units per pancreas. In contrast to CVBs, CVA-9 replicated poorly, and sometimes not at all, in foetal porcine beta-cells. The first signs of functional impairment and cell destruction, if present at all, were seen only after 1-3 weeks of incubation. Furthermore, CVA-16, several strains of echoviruses and human parechovirus type 1 were unable to replicate in foetal porcine pancreatic beta-cells. Based on these results, foetal porcine islets are somewhat more sensitive to CVB infection than adult human islets, whereas many other human EV strains do not infect porcine beta-cells. Therefore, foetal porcine beta-cells cannot be used for systematic screening of human EV strains and isolates for beta-cell tropism, but they might provide a useful model for detailed studies on the interaction of CVBs with beta-cells.