Information Technology Reference
In-Depth Information
adaptation to the cell-mediated and humoral branches of the immune system. We
also lack detailed understanding of the dynamics of immune responses to HIV.
Here, I describe a simulation approach to understanding and predicting the adap-
tation of HIV to immune surveillance at the molecular genetic level. Published esti-
mates of fundamental population genetic parameters for HIV-1 allow the realistic
simulation of HIV intrapatient evolution in the absence of selection. Selection is
more difficult to simulate realistically at the molecular level, however, because it
requires knowledge of the effects of all genetic changes on replication rate within a
given environment. For example, to simulate selection at the protein level would
require knowledge of the fitness effect of each amino acid at each site of a protein, or
protein region, and the effects on fitness of interactions among amino acids at differ-
ent sites. This information is not available for any protein region. However, in the
case of selection by host cell receptors, fitness effects can be modeled easily for
HIV-1 because the mean site-specific frequency of an amino acid is positively corre-
lated with its effect on fitness (da Silva 2006a). For the special case of immune se-
lection, in which the environment coevolves antagonistically with the virus, knowl-
edge is required of the dynamics of the immune response as well as the targeted
protein region (epitope).
The evolution of a viral population infecting a single patient is examined by fo-
cusing on the intensely studied third variable region (V3) of the HIV-1 exterior enve-
lope glycoprotein (gp120). I begin by providing background information on the HIV
replication cycle and the special role of V3 in interactions with cell-surface receptors
and neutralizing antibodies. Then I briefly review what little is known about the
dynamics of the humoral response to HIV and the evolutionary response of the virus.
This is followed by a description of the model, focusing on the methods used to
estimate viral fitness. Finally, results are presented from simulations that explore the
effects of coreceptor selection, antibody selection, and the interaction of these on
viral adaptation.
9.2 The HIV Replication Cycle
HIV virus particles (virions) contain two single-stranded copies of their RNA ge-
nome, which is approximately 9.5 kilobases long and includes nine protein-coding
genes. In order to replicate its genome a virion must infect a cell, reverse transcribe
its genome into DNA, and integrate the DNA copy into the host's genome. The first
step, infection, requires that gp120, which is on the virion's surface, interact with
protein receptors on the cell surface. Typically, gp120 binds to a CD4 receptor mole-
cule, which causes conformational changes to gp120 that allow it to then bind to
either of two chemokine receptors: CCR5 or CXCR
(Coffin 1999). As a result of
these interactions, the primary targets of infection are CD4
+
T cells expressing either
CCR5 or CXCR4. The second major step in the replication cycle is the reverse tran-
scription of the viral genome by the viral enzyme reverse transcriptase. Reverse
transcription of the viral genome is error prone and lacks proofreading, resulting in
the high mutation rate characteristic of retroviruses (~10
-5
mutations per nucleotide
per replication cycle
(Mansky and Temin 1995)). Reverse transcriptase also jumps
