Background:  HIV-1 invades the central nervous system (CNS) early after infection and evolves independently within the CNS compartment. Previous studies comparing HIV isolates from plasma and cerebrospinal fluid (CSF) have demonstrated variable compartmentalization of CSF isolates. However, few studies have characterized dynamic virologic interactions between blood and CSF. Pleocytosis, which is believed to represent inflammatory cell trafficking into CSF from blood, is common in HIV infection. We hypothesized that during the development of pleocytosis, trafficking of HIV-infected lymphocytes into CSF from extra-neural sources would result in equilibration of sequences derived from CSF and blood.

Methods:  A chronically infected HIV-1 patient who interrupted his ART was intensively investigated. Viral load changes in paired blood and CSF samples were documented along with the development of CSF pleocytosis. Viral sequences encoding the C2V3 loop of env were amplified by nested real-time polymerase chain reaction (RT-PCR) from paired blood and CSF samples. Replicate PCR products were cloned and sequenced. Phylogenetic analysis was performed using fDNAML on clonal sequences (Phylip 3.5 and G. Olsen). Assessment of degree of inter-compartment segregation was performed by testing for panmixis using gene phylogenies (Slatkin Maddison test) as implemented in MacClade (Sinauer X).

Results:  The C2V3 regions of 3 paired plasma and CSF samples were cloned, sequenced and analyzed. Viruses in blood and CSF were well compartmentalized before CSF pleocytosis developed. When CSF pleocytosis developed, circulating blood viral isolates were found in CSF.  

Conclusions:  Phylogenetic studies of HIV-1 viral isolates in paired blood and CSF revealed evidence of viral trafficking during CSF pleocytosis. These findings are consistent with the view that CSF is a virologically dynamic site that reflects changing interactions between extra- and intra-neural compartments. This is linked, in part, to changes in therapy and may explain why studies of CSF viral genetics do not always find evidence of compartmentalization. Further characterization of blood–CSF interactions may assist in evaluating the potential utility of CSF as a marker of virologic events in brain parenchyma, where ongoing viral replication may trigger neural injury.