Viral Vectors Take On HIV Infection.

The lack of robust, classical vaccines against virulent viruses has prompted the search for alternative approaches to the prevention and treatment of infection. A key target in these efforts has been the human immunodeficiency virus (HIV), in which the identification of broadly neutralizing antibodies (bNAbs) in patients led to attempts at passive prophylaxis through the administration of protein. Although this practice appears to be safe in humans, logistical and compliance issues make the maintenance of therapeutic levels of antibodies highly challenging. The use of gene-transfer therapy to mediate gene expression is a tactic that enables long-term expression of bNAbs or other molecules (including antigens, immune regulatory proteins and signals, and proteins or RNAs) that directly target either the products of the viral genome or the host proteins required for viral entry or replication. The use of a tissue (e.g., muscle or liver) as a depot for the expression of a bNAb against the pathogen essentially bypasses the need to stimulate an endogenous adaptive immune response. For example, a recombinant adeno-associated vector (rAAV) delivered into the skeletal muscle of monkeys expressed a bNAb against envelope glycoprotein 120 (gp120) that protected the recipients against repeated exposure to the simian immunodeficiency virus (SIV).1 This approach also provided protection against HIV infection and eliminated preexisting HIV infection in a study of mice.2 Such favorable outcomes of studies in animals have led to a phase 1 trial involving healthy men on the safety and immunogenicity of rAAV1-PG9DP against HIV infection (ClinicalTrials.gov number, NCT01937455), the results of which are expected by early 2016. However, among the great number of variant gp120 isolates, there are many HIV-1 strains that are resistant or only partially neutralized, even with substantial serum levels of bNAb. This issue is particularly important in the case of the quasispecies that evolve in patients with chronic infections, which suggests that multiple, high-titer antibodies will probably be required for generalized use. The most conserved regions of the gp120 envelope are the epitopes that bind to the primary receptor (CD4) and a coreceptor, usually CCR5 or CXCR4. The soluble, immunoadhesin form of the CD4 receptor, CD4-Ig, can block HIV infection in culture but has not been successful in the clinic.1,2 Another approach, suggested by the amino acid sequences of some bNAbs, involves the use of peptide mimetics in sulfotyrosinecontaining coreceptor domains (i.e., the domains of CCR5 or CXCR4), which interact with gp120 in most if not all HIV-1 clades, albeit with a low affinity in the absence of the CD4 receptor.3 Thus, engineering a protein that binds gp120 to block its binding to both the CD4 receptor and one of its coreceptors is an attractive approach.4 Recently, Gardner et al.5 achieved striking results by using an optimized bifunctional protein that is delivered by means of gene transfer. They used a fusion protein consisting of CD4-Ig and a CCR5 sulfotyrosine-containing mimetic, referred to as eCD4-Ig (Fig. 1), and showed that it had a stronger neutralizing effect on viral activity than that of the best bNAbs (at an order of magnitude of 1 to 2 logs) as measured by viral neutralization against a wide range of HIV-1, SIV, and even HIV-2 clades in cultured cells. The investigators designed two rAAV vectors, one expressing eCD4-Ig and the second expressing an enzyme, tyrosine-protein sulfotransferase 2, to ensure that eCD4-Ig was properly sulfated. After dual rAAV–vector gene transfer into the muscle, the investigators challenged humanized mice and macaque monkeys with multiple exposures to HIV and SIV, respectively. The rAAVtreated animals were protected against infection,