The restriction of zoonotic PERV transmission by human APOBEC3G

Abstract

The human APOBEC3G protein is an innate anti-viral factor that can dominantly inhibit the replication of some endogenous and exogenous retroviruses. The prospects of purposefully harnessing such an anti-viral defense are under investigation. Here, long-term co-culture experiments were used to show that porcine endogenous retrovirus (PERV) transmission from pig to human cells is reduced to nearly undetectable levels by expressing human APOBEC3G in virus-producing pig kidney cells. Inhibition occurred by a deamination-independent mechanism, likely after particle production but before the virus could immortalize by integration into human genomic DNA. PERV inhibition did not require the DNA cytosine deaminase activity of APOBEC3G and, correspondingly, APOBEC3G-attributable hypermutations were not detected. In contrast, over-expression of the sole endogenous APOBEC3 protein of pigs failed to interfere significantly with PERV transmission. Together, these data constitute the first proof-of-principle demonstration that APOBEC3 proteins can be used to fortify the innate anti-viral defenses of cells to prevent the zoonotic transmission of an endogenous retrovirus. These studies suggest that human APOBEC3G-transgenic pigs will provide safer, PERV-less xenotransplantation resources and that analogous cross-species APOBEC3-dependent restriction strategies may be useful for thwarting other endogenous as well as exogenous retrovirus infections.

Figure 1. PERV Transmission Assay.

(A) Schematic of the co-culture system. PERV transmitting PK-15 cells are grown on top of the membrane of the insert and human 293T cells on the bottom of the well of the culture dish. Virus particles are depicted diffusing through membrane pores. (B) The zoonotic transmission of PERV from PK-15 cells to 293T cells is shown by the time-dependent accumulation of PERV pol gene DNA in the human cells (solid diamonds and squares). No concomitant transfer of pig genomic DNA occurred through the duration of these experiments (open diamonds and squares). This graph summarizes data for two independently derived PK-15 clones, V1 (squares) and V2 (diamonds). All data points were calculated using results from duplicate Q-PCR reactions of genomic DNA from three parallel (but independent) co-cultures. The error bars indicate the SEM. See the Materials and Methods and Online Figure S1 for additional details, representative raw data and controls.

Figure 2. Human APOBEC3G Inhibits PERV Transmission.

(A) An immunoblot showing PK-15 clones expressing human APOBEC3G (G1 and G2) or a control vector (V1 and V2). PK-15 and 293T cell lysates were used as negative controls. CEM and H9 were used as positive controls for APOBEC3G expression. A non-specific (but pan-species) band is shown as a protein loading control (marked by an asterisk). (B) Q-PCR data using genomic DNA prepared from 293T cells co-cultured with two independently derived APOBEC3G expressing PK-15 clones (G1 and G2, circles and triangles, respectively) or two vector control clones (V1 and V2, diamonds and squares, respectively). The experimental parameters are identical to those used in Figure 1B.

Figure 3. The Sub-cellular Distribution of Human APOBEC3G in Human and Pig Cell Lines.

Sub-cellular distributions of GFP, human APOBEC3G-GFP and pig APOBEC3F-GFP in the indicated live pig and human cell lines. The scale bar in the top left panel indicates 10 µM.

Figure 4. The APOBEC3G-dependent Inhibition of PERV Transmission Is Deamination-Independent.

PERV-specific Q-PCR data using genomic DNA prepared from 293T cells co-cultured with PK-15 clones expressing APOBEC3G (G; triangles), APOBEC3G-E259Q (GE259Q; circles) or empty vector (V; squares). Two datasets, each with an independent PK-15 clone in three replica co-culture wells, were collected in parallel and averaged for each data point. One standard error of the mean is shown. The experimental parameters are identical to those used in Figure 1B. The inset immunoblots show the APOBEC3G and α-tubulin levels of representative PK-15 clones expressing the indicated construct. The human T cell line H9 provided a positive control for APOBEC3G expression.

Figure 5. Genetic Variation in Zoonosed PERV DNA Sequences.

(A) A schematic of the PERV genome showing coding regions (gag, pol and env) and long-terminal repeats (LTRs). The relevant 193 bp pol gene fragment is indicated. (B) An ethidium bromide-stained agarose gel showing that PERV pol gene DNA was amplified readily from 293T cells cultured with PK-15 control clones (None) and pig APOBEC3F over-expressing clones (A3Fo/e) but not with PK-15 clones expressing human APOBEC3G (A3G; top panel). PERV pol gene DNA (top panel) and pig genomic DNA (APOBEC3F locus; middle panel) PCR products were detected in PK-15 genomic DNA, whereas human beta-actin was strongly detected in the 293T cell genomic DNA samples (a much weaker amplification of pig beta-actin occurred because the human primers had partial complementarity to pig sequences, 20/21 and 17/18 nucleotides).