Epidemiology and HIV’s Physical Properties

OCT.19 '95 7:33AM 310-446-1664 LAYNE @ UCLA. [phone] P.4
COMMENTARY
sequences in the Human Retrovirus and AIDS Database. At present, there are eight sequence
subtypes for HIV-1 and, given more limited data, there appears to be at least two for HIV-2. Each
subtype has a characteristic phylogenetic map and differing geographic distribution, making it
possible to track the evolution of the epidemic12. The total number of sequences that may fit into a
particular subtype, however, is truly enormous. Assume, for example, that HIV is limited to
utilizing just 2 different amino acids at certain positions within its proteins. Further assume that a
"model" subtype is determined by substitutions in 30 independent positions (or approximately 1%
of the 3000 amino acids in HIV's entire genome). In this case, the model subtype could contain as
many as 230 = 109 different sequences - an insurmountable number even if only a small set of all
possible combinations produced active viruses. Given the difficulty of sequencing thousands of
HIV isolates in their entirety, it seems improbable that current technologies will enable us to
identify genetic sequences that correlate reliably with transmissibility. In other words, certain
subtypes may act as surrogate markers of transmissibility but may not identify the underlying
causes.
Serotyping of IV isolates has been carried out with standardized panels of
inununoglobulins that block the virus from infecting of CD4+ cells. This often used approach has
allowed virologists to categorize isolates according to their patterns of susceptibility and to target
certain HIV serotypes for vaccine development13. Susceptibility to blocking, however, varies
markedly with single amino acid substitutions within the gpl20 envelope glycoprotein. The
enormous combinatorics and limitations on sampling has restricted our ability to identify serotypes
that correlate reliably with transmissibility. We need to develop new measurement-based schemes
that integrate epidemiologic, immunologic and virologic data to understand why more transmissible
HIV isolates are emerging.
Physical properties
Increased transmissibility suggests that greater amounts of active virus pass from infected to
susceptible individuals or that once exposed to the inoculum, there is a greater chance for viral

multiplication and seroconversion. Any one of HIV's physical properties which facilitate these
mechanisms of transmission should be considered as potential contributors (Figs. I and 2).

3