Background: The most common form of peripheral neuropathy seen in HIV/AIDS is distal sensory polyneuropathy (DSP), which is defined by both spontaneous and evoked pain, paresthesiae, and hypesthesia together with both large and small caliber axonal loss following a 'dying back' pattern of degeneration. To date, little is known about the extent of HIV-1 infection of peripheral nerves or the nature of the infecting virus and its potentially pathogenic effects.  In this study we aim to determine the nature of HIV-1 infection in peripheral nerves of clinically-characterized patients with and without DSP.

Methods: Peroneal nerve biopsies were obtained at autopsy from HIV-1 infected patients with and without DSP.  PCR amplification was performed using primers specific for the C2V3 region of the HIV-1 envelope gene.  HIV-1 co-receptor usage was examined by both sequence analysis of the HIV-1 V3 loop and by the generation of recombinant viruses expressing the C2V3 region.  Pathogenic effects of the recombinant viruses were explored with an in vitro culture method using dorsal root ganglia (DRG) from transgenic rats expressing human CD4 and CCR5.

Results: PCR amplification of DNA and RNA yielded products for sequencing in 7/12 and 1/10 patients respectively.  V3 loop sequence analysis indicated the presence of both X4 and R5 tropic HIV-1 within peripheral nerves.  Similarly, recombinant viruses containing peripheral nerve-derived C2V3 sequences used both CXCR4 and CCR5 for in vitro infection.  DRG cultures consisted of neurons (MAP-2+), Schwann cells (GFAP+) and macrophages (ED-1+). Only macrophages were human CD4+, whilst CCR5 immunopositivity was detected on both macrophages and neurons. Following infection, HIV-1 p24 immunoreactivity was observed only in macrophages with peak viral detection at day 4 post-infection.  Both X4 and R5 recombinant viruses reduced neurite outgrowth from neuronal cell bodies but R5 viruses also induced a reduction in the number of macrophages.

Conclusions: HIV-1 provirus was detected in peripheral nerves of patients both with and without DSP, suggesting that the presence of virus within peripheral nerves per se does not predict the clinical phenotype of sensory neuropathy.  The low level viral RNA detection indicates limited viral replication within peripheral nerves.  The reduction in the number of neurons with processes suggests a pathogenic role for HIV-1 within peripheral nerves independent of co-receptor utilization.

Keywords: Sensory neuropathy; Envelope sequence; HIV-1 co-receptors