Regions of human immunodeficiency virus type 1 nef required for function in vivo

Abstract

In vivo studies in monkeys and humans have indicated that immunodeficiency viruses with Nef deleted are nonpathogenic in immunocompetent hosts, and this has motivated a search for live attenuated vaccine candidates. However, the mechanisms of action of Nef remain elusive. To define the regions of human immunodeficiency virus type 1 (HIV-1) Nef which mediate in vivo pathogenicity, a series of mutated isogenic viruses were inoculated into human thymic implants in SCID-hu mice. Mutation of several regions, including the myristoylation site at the second glycine and a region encompassing amino acids 41 through 49 of Nef, profoundly affected pathogenicity. Surprisingly, mutations of prolines in either of the two distant PXXP SH3 binding domains did not affect pathogenicity, indicating that these regions are not required for Nef activity in developing T-lineage cells. These data suggest that some functions of Nef described in vitro may not be relevant for in vivo pathogenicity.

FIG. 1

Schematic of the nef gene and isogenic deletion mutant forms. Putative functional regions of Nef are indicated at the top. At the far right is a summary of relative Nef activity, as determined by our studies. ++++, wild-type growth; +, attenuated growth; −, no detectable virus. This interpretation was based on our statistical analyses (see Fig. 2).

FIG. 2

Thymocyte depletion (A) and proviral load (B) in Thy/Liv implants 6 weeks postinfection. Each symbol represents a different implant infected with the corresponding viral strain, as described in Fig. 1. (A) Percentage of CD4-bearing cells (including both CD4⁺ CD8⁺ double-positive and CD4⁺ single-positive subsets), as determined by flow cytometry (1, 2). An implant was considered depleted if it was infected and had fewer than 55% CD4-bearing cells. Nef-minus viruses (Δnef) and strains with deletions 3 or 5 were less capable of depleting CD4-bearing cells than was the wild-type virus (NL). This difference was statistically significant (Fisher’s exact test, two tailed), with P = 0.0002, 0.0004, and 0.00002 for Δnef and deletions 3 and 5, respectively. Deletion 6 was not included in these calculations, because the virus was not infectious (B). Other viruses yielded P > 0.05. (B) Number of HIV-1 genomes in 10⁵ human thymocytes, as determined by quantitative DNA PCR. Symbols on the x axis indicate implants with undetectable HIV DNA. The minimum amount of DNA assayed was the equivalent of 10⁴ thymocytes.

FIG. 3

Schematic of the nef gene and isogenic point mutant forms. As in Fig. 1, a summary of relative pathogenic potential is given at the right. ++++, wild-type growth; +, attenuated growth; ++, somewhat attenuated growth. The statistical data used to make these interpretations are given in the legend to Fig. 4.

FIG. 4

Thymocyte depletion (A) and proviral load (B) in implants 6 weeks postinfection with point mutated viruses. Each symbol represents a different implant infected with the corresponding viral strain, as described in Fig. 3. (A) Percentage of CD4-bearing cells (both CD4⁺ CD8⁺ double-positive and CD4⁺ single-positive subsets), as determined by flow cytometry (Fig. 2). An implant was considered depleted if it was infected and had fewer than 55% CD4-bearing cells. MYR⁻ virus and the strains containing the PXXP4+2 point mutation and the ARGX2 mutations were significantly less capable of depleting CD4-bearing cells (Fisher’s exact test, two tailed), with P = 0.001, 0.0002, and 0.015 for MYR⁻, PXXP4+2, and ARGX2, respectively. Other viruses yielded P > 0.05. (B) Number of HIV-1 genomes in 10⁵ human thymocytes, as determined by quantitative DNA PCR (Fig. 2B).