Nef from pathogenic simian immunodeficiency virus is a negative factor for vaccinia virus

Abstract

The nef gene of human and simian immunodeficiency viruses (HIV and SIV) is important for pathogenicity and maintenance of high virus loads. We previously reported that recombinant vaccinia viruses (rVVs) expressing nef from attenuated SIVmac1A11 (vNef1A11) produced typical plaques on thymidine kinase-deficient 143B cells, whereas rVVs expressing nef derived from the pathogenic SIVmac239 (vNef157) formed plaques with altered morphology. Here, we show that vNef157 is attenuated in normal and nude mice, whereas the pathogenicity of vNef1A11 is similar to that of a control virus. Thus, Nef157 is an attenuating factor in the vaccinia virus (VV) system, contrasting sharply with its function in lentiviruses. We also show that Nef157 inhibits VV cell-to-cell spread, causing formation of atypical plaques regardless of thymidine kinase deficiency, neoplasticity, and species of the infected cell line. We hypothesized that Nef157 interferes with VV spread by association with actin, but no direct colocalization of Nef and the cytoskeletal actin network was detected. Instead, higher levels of Nef157 protein were observed, although mRNAs for both nef genes were produced at comparable levels. Thus, the mechanism behind such Nef157 protein accumulation and Nef157-mediated VV attenuation could be related to the process that causes an opposite effect in its native SIV system, making SIVmac239 more pathogenic than SIVmac1A11.

Fig. 1.

vNef157 consistently produced atypical, smaller plaques than vNef1A11 and v50 in various cell lines. (A) Human TK^- 143B, African green monkey (BS-C-1, BS-C-40, and Vero), murine L929, and canine MDCK cells were infected with rVV, and, 36 h later, photographs of representative plaques were taken. The sizes of vNef1A11 and v50 plaques were very similar in all cell lines, but vNef157 consistently produced atypical, smaller plaques. (B) Significant plaque size differences (t test, P < 0.0001) were observed in v50 or vNef1A11 vs. vNef157 comparisons in all cell lines, but only Vero cells showed a plaque size difference between v50 and vNef1A11. Error bars indicate the SEM. The number of days postinfection (dpi) is indicated in parentheses after the cell line name.

Fig. 2.

vNef157 consistently produced lower virus yields than vNef1A11 and v50 in various cell lines at low MOI but equivalent virus yields at high MOI. Monolayers of multiple cell lines were infected with rVVs at a low MOI of 0.01 (A and C) or high MOI of 5 (B). At 0, 6, 12, 24, and 48 h postinfection (A) or at 24 (B) or 48 (C) h postinfection, both intracellular (Upper) and extracellular (Lower) virus fractions were collected and titered in BS-C-1 cells. The data shown represent the mean viral yields from triplicate samples assayed in duplicate. Error bars indicate the SEM.

Fig. 3.

vNef157 is less pathogenic than vNef1A11 and v50 in immunodeficient and immunocompetent mice. (A) Groups of 10 athymic BALB/cBy nude mice were inoculated i.p. with 10⁷ PFU of rVV. All nude mice injected with v50 and vNef1A11 succumbed by 40 days postinoculation, whereas mice infected with vNef157 exhibited a much greater survival rate (60% by day 40). The Kruskal-Wallis test for the three groups yielded a P value of 0.001 (significant difference of at least one group). Pairwise comparisons yielded a P value >0.9 for v50 vs. vNef1A11 (no significant difference) and values of 0.003 for v50 vs. vNef157 and 0.001 for vNef1A11 vs. vNef157 (highly significant differences). (B) Groups of 10 CB6F₁ mice were inoculated intranasally with 5 × 10⁶ PFU of rVV. The mean group weight (±SEM) was expressed as the percentage of the mean weight of that group of animals immediately after infection. Significant differences in weight loss were observed in v50- and vNef157-inoculated mice (ANOVA followed by Tukey's test) from days 1 through 14 postinfection (P <0.05). vNef1A11-inoculated mice also displayed weight loss similar to v50-inoculated mice but recovered much faster, showing a statistically significant difference when compared with v50-inoculated mice from days 8 through 12 postinfection (P < 0.05). vNef1A11-inoculated mice differed statistically from vNef157-inoculated mice at day 5 (P < 0.01) and from days 6 through 8 (P <0.001). Error bars indicate the SEM.

Fig. 4.

SIV Nef did not colocalize with cellular actin. TK^- 143B (A and B) and BS-C-1 (C and D) cells were infected with vNef1A11 (A and C) and vNef157 (B and D), and, at 48 h postinfection, cells were fixed and processed for immunofluorescent staining. Nef1A11 and Nef157 were detected with an anti-Nef antibody and a secondary antibody conjugated with Alexa Fluor 568 (red), cellular actin was detected with phalloidin conjugated to Alexa Fluor 488 (green), and the cellular nucleus was located with DAPI stain (blue). Neither Nef1A11 nor Nef157 appeared to colocalize with actin. Instead, the Nef proteins and actin exhibit different distribution patterns, and the intensity of Nef staining was, on average, stronger in vNef157-infected cells than vNef1A11-infected cells.

Fig. 5.

Nef157 was expressed at higher levels than Nef1A11 in rVV-infected cells. TK^- 143B cells were infected by vNef157 and vNef1A11 at a MOI of 1. At 18 h postinfection, cells were collected for anti-Nef immunoblotting analysis. Quadruplicates of cell lysates from independent infections showed that the Nef protein (≈30 kDa) in vNef157-infected cell lysates could be detected by immunoblotting after a 1:27 sample dilution and that Nef from vNef1A11-infected cells was undetectable after a 1:9 dilution. Data are representative of quadruplicates of immunoblots. Baculovirus-expressed Nef served as a positive control. MW, molecular weight standards.

Fig. 6.

Nef157 and Nef1A11 mRNA levels in VV-infected cells were similar. BS-C-1 and TK^- 143B cells in six-well plates were infected with vNef1A11 or vNef157 at a MOI of 1. At 100 min (1.7 h) and at 6 and 12 h, cells were harvested, and the nef mRNA levels were measured by real-time TaqMan PCR. Quantification of nef mRNA is reported as SIV nef transcripts per culture, which is equivalent to one well of infected cells in a six-well plate. The data shown represent the mean values from triplicate samples. Error bars indicate the SD.