oa MiSeq-Next Generation Sequencing Approach in Investigating the Evolutionary Dynamics of Viral Infection in Children with Type I Diabetes

Background

There is no longer a question as to whether viruses contribute to the pathogenesis of type 1 diabetes (T1D), as we recently reviewed, but rather how they contribute and, in particular, the role of viral diversity and evolution in the disease process. The recent finding of enterovirus (EV) capsid protein VP1 in pancreatic autopsy samples from the JDRF Network for Pancreatic Organ Donors with Diabetes (nPOD) supports earlier case series in which EVs (Coxsackievirus B, CVB)were isolated from pancreatic tissue and inoculated into human islets, causing functional impairment and β-cell death. The most interesting observation from the nPOD data is the patchy distribution of insulitis, with MHC class II hyper-expression on β-cells, which was co-located with viral protein. Indeed, it is well established that specific EV strains demonstrate β-cell tropism; we and others have shown that EVs infect and replicate in β-cells (Fig. 1), inducing inflammation, cytokine production and functional damage. There is also substantial epidemiological evidence that EVs have more than an occasional role in the disease; in our meta-analysis of >4000 cases, the odds ratio (OR) was ∼10 for EV infection at T1D onset vs controls, and OR∼4 for EV infection and islet autoimmunity (IA). While the genetic and immunological components of disease are not in question, the capacity for EVs to evolve is completely unexplored in the pathogenesis of human T1D. This information is critical for development of EV vaccines to prevent T1D, which is currently underway.

HYPOTHESIS-1: Variation in the capsid and non-structural regions of the EV genome determine β-cell tropism

AIM-1: To characterize human EV isolates in cases of IA and T1D using NGS – to identify regions in the EV genome associated with β-cell tropism

HYPOTHESIS-2: Increased genetic diversity of EVs at the full genome level is associated with seroconversion to IA and T1D

AIM-2: To examine the evolutionary dynamics and genetic diversity of EVs at the full genome level from children with IA and T1D, and to quantify the extent of intra-host evolution of EVs within an infection and the kinetics of intra-host virus evolution between infections.

Research Plan and Methods

EV prototype strains and clinical isolates from children with AI and T1D that infect and replicate in β-cells. Cohorts: Viruses in Genetically at Risk (VIGR), Environmental Determinants of Islet Autoimmunity (ENDIA), and children at onset of T1D (EET1DPP2). Samples collected at the study visit or at diagnosis of T1D. RNA extracted with QIAamp viral RNA, quantitative RT-PCR were performed on the Roche LC-480 platform. NGS: Full-length EV genomes were amplified as a single 7.4 kb fragment by RT-PCR, and NGS performed using the Illumina MiSeq sequencer. Phylogenetic analysis of the full-length viral consensus sequences performed using the neighbour joining and maximum likelihood method. Statistical analysis with R software. Trees were constructed from alignment of complete genome sequences by using best-fit models and visualised using FigTree (Figure2). Comparisons were performed with Viral Epidemiology Signature Pattern Analysis (VESPA).

Results and Discussion

Two of the EVs from IA+ cases had an N to S amino acid (AA) substitution within the 2C protein, which became dominant after 10 days passage in the islets. The 2C protein encodes for the viral helicase and lies just upstream of the viral region that shares significant homology with human GAD65. EV isolate from another IA+ case has 5 AA differences within the capsid protein VP4 at residues 3, 16, 18, 50 and 61 (Figure3). VP4 is an internal capsid protein linked to the genome. VP4 has been shown in vitro to be a target of human antibodies that enhance CVB induced synthesis of interferon α (IFN-α). CVB-induced IFN-α plays a role in the initiation and/or maintenance of chronic CVB infection in human islets. Antibodies directed towards the region 11–30 of the VP4 capsid enhance infection of peripheral blood cells with CVB4 in vitro. Therefore, our preliminary data suggest VP4 may be a determinant of ‘diabetogenicity’. Our novel NGS data will contribute to vaccine development from a global perspective. Our ultimate goal is to reduce the future burden of T1D.