Differential inhibition of long interspersed element 1 by APOBEC3 does not correlate with high-molecular-mass-complex formation or P-body association

Abstract

The human cytidine deaminase APOBEC3G (A3G) and other APOBEC3 proteins exhibit differential inhibitory activities against diverse endogenous retroelements and retroviruses, including Vif-deficient human immunodeficiency virus type 1. The potential inhibitory activity of human APOBEC proteins against long interspersed element 1 (LINE-1) has not been fully evaluated. Here, we demonstrate inhibition of LINE-1 by multiple human APOBEC3 cytidine deaminases, including previously unreported activity for A3DE and A3G. More ancient members of APOBEC, cytidine deaminases AID and APOBEC2, had no detectable activity against LINE-1. A3A, which did not form high-molecular-mass (HMM) complexes and interacted poorly with P bodies, was the most potent inhibitor of LINE-1. A3A specifically recognizes LINE-1 RNA but not the other cellular RNAs tested. However, in the presence of LINE-1, A3A became associated with HMM complexes containing LINE-1 RNA. The ability of A3A to recognize LINE-1 RNA required its catalytic domain and was important for its LINE-1 suppression. Although the mechanism of LINE-1 restriction did not seem to involve DNA editing, A3A inhibited the accumulation of nascent LINE-1 DNA, suggesting interference with LINE-1 reverse transcription and/or integration or intracellular movement of LINE-1 ribonucleoprotein. Thus, association with P bodies or cellular HMM complexes could not predict the potency of APOBEC3 anti-LINE-1 activities. The catalytic domain of APOBEC3 proteins may be important for proper folding and target factors such as RNA or protein interaction in addition to cytidine deamination.

FIG. 1.

Effect of human APOBEC proteins on LINE-1 retrotransposition. (A) 293T cells were transfected with the negative control pLINE-1_RPJM111-EGFP (JM111) or pLINE-1_RP-EGFP in the absence (with pcDNA3.1+ empty vector) or presence of APOBEC expression vectors. Relative retrotransposition was measured by detecting EGFP expression in cells using fluorescence-activated cell sorting analysis 5 days after transfection and calculated by setting the value for pLINE-1_RP-EGFP cotransfected with an empty vector at 100%. (B) Immunoblot showing similar expression levels of APOBEC proteins in 293T cells cotransfected with various plasmids expressing HA-tagged APOBEC proteins and pLINE-1_RP-EGFP. APOBEC proteins were detected with anti-HA, and ribosomal p19 was detected as a loading control. The expression of AID-V5 from transfected 293T cells was detected with an anti-V5 antibody (data not shown). (A) 293T cells were transfected with the negative control pLINE-1_RPJM111-EGFP (JM111) or pLINE-1_RP-EGFP in the absence (with pcDNA3.1+ empty vector) or presence of APOBEC3 expression vectors. Relative retrotransposition was measured by detecting EGFP expression in cells using fluorescence-activated cell sorting analysis 5 days after transfection and calculated by setting the value for pLINE-1_RP-EGFP cotransfected with an empty vector at 100%. (D) Immunoblot showing similar expression levels of APOBEC3 proteins in 293T cells cotransfected with various plasmids expressing HA-tagged APOBEC proteins and pLINE-1_RP-EGFP. APOBEC proteins were detected with anti-HA, and ribosomal p19 was detected as a loading control.

FIG. 2.

Formation of intracellular HMM complexes containing APOBEC3 proteins and selective interaction of APOBEC3 proteins with cellular mRNAs. (A) Velocity sucrose gradient fractionation of APOBEC3-containing complexes. 293T cells were transfected with APOBEC3 expression vectors. Three days later, the cells were lysed and subjected to 10-to-30% continuous sucrose gradient fractionation. Samples were collected and analyzed by immunoblotting with an anti-HA monoclonal antibody. (B) Velocity sucrose gradient fractionation of A3A was not affected by RNase treatment. Cell lysates expressing A3A-HA were treated in the absence or presence of RNase and then analyzed by sucrose gradient fractionation. (C and D) Coimmunoprecipitation of cellular mRNAs with APOBEC3 proteins. HA-tagged APOBEC3 proteins were expressed in transfected 293T cells. Cell lysates from transfected 293T cells were immunoprecipitated with anti-HA antibody conjugated to agarose beads. RNAs coprecipitated with various APOBEC3 proteins were analyzed by qRT-PCR using primers specific for beta-actin (C) or GAPDH (D). The nonspecific RNA (containing the control vector) binding to the assay system was set at 1.

FIG. 3.

Localization of APOBEC3 proteins and P bodies. (A) 293T cells were transfected with various APOBEC3 expression vectors. APOBEC3 proteins were stained with anti-HA (red). P bodies were visualized by anti-DDX6 (RCK/p54), a P-body marker (green). Some A3DE and A3G were strongly associated with P bodies, while A3A and A3C seemed to have a weaker association. Nuclei were counterstained by DAPI (4′,6′-diamidino-2-phenylindole). (B) Anti-HA antibodies staining A3A and the mutant A3AE72K. A3A appears to have a fairly homogeneous localization in both the cytoplasm and nucleus. The A3AE72K mutant shows aberrant localization, forming aggregate-like spots in the cytoplasm and nucleus. Nuclei were counterstained with DAPI.

FIG. 4.

Association of A3A with LINE-1 RNA in HMM complexes. (A) Anti-HA immunoblot of sucrose velocity gradient fractions of lysed 293T cells cotransfected with HA-tagged A3A plus an empty vector (pcDNA3.1) (top), pLINE-1_RP-EGFP (middle), or pLINE-1_RP-EGFP with RNase treatment prior to sucrose velocity gradient centrifugation (bottom). A3A was shifted to fractions 6 to 9 in the presence of pLINE-1_RP-EGFP. (B) Immunoprecipitation of A3A-HA from fractions 6 to 9 in the presence of pLINE-1_RP-EGFP. (C) Coprecipitation of LINE-1 RNA with A3A-HA from fractions 6 to 9 in the presence of pLINE-1_RP-EGFP. LINE-1 RNA was not detected in fractions 6 to 9 of the sucrose gradient samples containing LINE-1_RP-EGFP in the absence of A3A-HA.

FIG. 5.

An intact catalytic site in A3A is necessary for LINE-1 RNA binding and LINE-1 inhibition. (A) 293T cells were transfected with the negative control pLINE-1_RP(JM111)-EGFP or pLINE-1_RP-EGFP in the absence or the presence of A3A or A3Am. Retrotransposition was detected by fluorescence-activated cell sorting analysis of EGFP expression in 293T cells 5 days after transfection and was calculated by setting pLINE-1_RP-EGFP cotransfected with an empty vector (pcDNA3.1) at 100%. (B) Anti-HA immunoblot showing similar levels of expression of A3A and A3Am in 293T cells cotransfected with pLINE-1_RP-EGFP. (C) Coprecipitation of LINE-1 RNA with A3A-HA but not A3AE72K-HA. Ethidium bromide-stained agarose gel showing reverse-transcribed and PCR-amplified EGFP precipitated with A3A-HA but not with A3AE72K-HA. RT-PCR detection was within the detection sensitivity of the assay, as indicated by the dilution standard (lanes 5 to 8). (D) Immunoblot showing that both A3A-HA and A3Am-HA were immunoprecipitated. (E) A3A but not A3AE72K inhibits LINE-1 DNA accumulation. Ethidium bromide-stained agarose gel showing accumulation of LINE-1 DNA in 293T cells cotransfected with A3A or A3Am and pLINE-1_RP-EGFP. LINE-1 DNA was amplified by nested PCR primers in EGFP.