Congenital human cytomegalovirus (HCMV) infection is a major cause of central nervous system structural anomalies and sensory impairments in the newborn. A major goal of our research is to understand at a high-resolution the effects of HCMV infection on the neural lineage specification and maturation of stem and progenitor cells. Elucidation of the genes and cellular processes that are affected will serve as a basis for therapeutic strategies to ameliorate the effects of HCMV infection in newborns. The significance of our studies also extends to the serious problem of HCMV infection in immunocompromised individuals, with recipients of allogeneic transplants having a high risk of severe disease and allograft rejection. This potential problem in stem cell therapy has received little attention thus far. The proposed use of stem cell transplantation in treating neuronal injury and neurodegenerative diseases, as well as transplantation of other organ-specific precursors, makes it imperative to understand how disseminated HCMV infection in immunosuppressed recipients will affect the function and differentiation of the cells.
This past year, we have made significant progress in accomplishing the goals of this project. We used human embryonic stem cell (hESC)-derived primitive pre-rosette neural stem cells (pNSCs) to study host-HCMV interactions in early neural development and found with several different lines of hESC-derived pNSCs that HCMV infection is inefficient and non-progressive. Differentiation of pNSCs into primitive neural progenitor cells (pNPCs) restored some viral early gene expression but not the transactivation of late genes. Impaired viral gene expression in pNSCs was not a result of inefficient viral entry or nuclear import of viral DNA but correlated with deficient nuclear import of the virion-associated protein UL82, which is believed to play a role in removing barriers to viral RNA synthesis. Additionally, we found that viral genomes could persist in pNSCs culture up to a month after infection despite the absence of detectable viral lytic gene expression, although we could also detect expression of viral latency-associated genes, suggesting that the virus becomes latent in pNSCs.
To study in greater depth the molecular basis of the interaction of HCMV with cells of the neural lineage, we have continued high-throughput genomics approaches to analyze HCMV microRNAs, alterations in cellular microRNA and gene expression profiles, and global defects in host alternative splicing in infected and uninfected pNSC-derived NPCs. We found that in infected NPCs, there was specific downregulation of transcripts related to neuron differentiation. These findings demonstrate the capacity of HCMV infection to alter the neural identities of key precursor cells in the developing nervous system. We also analyzed our infected NPC RNA-seq database for differences in host mRNA polyadenylation patterns and found that over a hundred transcripts were significantly altered in terms of their 3′ end cleavage site preference, with the majority of these events resulting in shortened 3′ UTRs.
Our finding that HCMV induces major changes in the transcriptome of NPCs, particularly at the level of neural genes, suggested that the virus might affect these cells functionally. To directly evaluate this, we differentiated pNSCs into midbrain dopaminergic (mDA) neurons and infected these cells at different times of the differentiation process. Seeding pNSCs in differentiation medium for 6 weeks yields a high frequency of mature neurons with long axonal projections. Infection of pNSCs at the start of differentiation greatly reduces generation of beta III-tubulin+ neurons 4 weeks later, and prevents differentiation to mature MAP2+ neurons. Infection after 1 week of differentiation also reduces the number of beta III-tubulin+ neurons and results in massive cell death. Infection at 2 weeks after differentiation start does not reduce the number of βIII-tubulin+ or MAP2+ neurons, but the cells display major anomalies. Since neurons are highly sensitive to oxidative stresses and HCMV infection increases the production of reactive oxygen species in fibroblasts, we investigated whether the same effect occurred in neuronal cultures. When pNSCs were infected at week 4 after differentiation, high levels of ROS were detected. These results suggest that the complex effects of HCMV infection at various stages of neural cell differentiation on both cell survival and maturation may account for the broad range of birth defects.
We expect that the results of these studies will provide an unprecedented resolution of the effects on neurogenesis when HCMV infects a newborn, serve as a foundation for future therapeutic efforts in preventing the birth defects due to HCMV, and provide insight into the serious potential problem of disseminated HCMV in immunosuppressed individuals receiving transplanted allogeneic stem cells.