ENO1 in HIV-1 target cells influences viral infectivity
We previously reported that GAPDH and ENO1 are incorporated into HIV-1 particles, and virion-incorporated GAPDH and ENO1 inhibit viral reverse transcription independently of each other [24, 26]. Therefore, in this study, we examined whether GAPDH and ENO1 expression levels were changed by HIV-1 infection. Cell lysates derived from a T-cell line, CEM cells, and a chronically HIV-1-infected T-cell line, CEM/LAV-1 cells, were subjected to western immunoblotting. As a result, we detected no significant difference in the expression level of GAPDH between CEM and CEM/LAV-1 cells (Fig. 1A). In addition, GAPDH knockdown in TZM-bl cells maintained comparable levels of HIV-1 infectivity (Figs. 1B and C). On the other hand, the expression level of ENO1 was lower in CEM/LAV-1 cells than in CEM cells (Fig. 1D), and ENO1 knockdown in TZM-bl cells increased HIV-1 infectivity (Figs. 1E and F) without affecting cell viability (Fig. 1G). These findings suggest that ENO1 may have higher inhibitory activity against HIV-1 infection than GAPDH in target cells, and ENO1 in viral target cells may inhibit HIV-1 replication via a mechanism similar to or different from that underlying the inhibitory activity of virion-packaged ENO1.
ENO1 knockdown in HIV-1 target cells increases viral integration efficiency
To gain further insight into the function of ENO1 in HIV-1 target cells, we examined postentry steps in ENO1 knockdown cells. As shown in Fig. 2A, flow cytometry showed that treatment of TZM-bl cells with ENO1-specific siRNA had no effect on surface expression levels of the HIV-1 receptors CD4, CCR5 and CXCR4. In addition, measurement of cytosolic p24 isolated from HIV-1-infected TZM-bl cells by a previously described method  demonstrated that HIV-1 was able to penetrate into target cells regardless of the ENO1 expression level in the cells (Fig. 2B). We next investigated whether ENO1 knockdown in TZM-bl cells enhances the viral reverse transcription because low-level-ENO1-packaging virus, which was prepared by transfection of CEM/LAV-1 cells with an ENO1-specific siRNA, showed an increased reverse transcription efficiency . Unexpectedly, ENO1 knockdown has no effect on the abundance of late R/gag products of reverse transcription (Fig. 2C). Furthermore, the number of copies of two-long-terminal-repeat (2-LTR) circular DNA products, which are generally used as a marker of viral cDNA nuclear import, also showed no significant difference between ENO1-knockdown cells and control cells (Fig. 2D). However, when we performed nested Alu-gag PCR analysis, which is generally used for calculation of integrated viral cDNA, we found that ENO1 knockdown increased the integration efficiency, which correlates with an enhanced HIV-1 infection (Fig. 2E). These findings indicate that unlike the inhibitory effect of virion-packaged ENO1, ENO1 in viral target cells inhibits HIV replication by preventing HIV-1 integration.
ENO1 overexpression in HIV-1 target cells decreases viral integration efficiency
Because ENO1 knockdown in the viral target cells resulted in increased HIV-1 infectivity, we next investigated the effect of ENO1 overexpression. To overexpress ENO1 in viral target cells, TZM-bl cells were transfected with an ENO1-V5 expression vector (Fig. 3A). As a result, ENO1 overexpression in HIV-1 target cell decreased viral integration efficiency (Fig. 3B). This result indicated that ENO1 impaired HIV-1 infection in viral target cells. To further examine whether ENO1 overexpression affects CD4, CXCR4 and CCR5 expression levels, we performed flow cytometry using anti-CD4, anti-CCR5 and anti-CXCR4 antibodies. The results showed that ENO1 overexpression had no effect on the expression levels of HIV-1 receptors, indicating that the viral entry step was unaffected by ENO1 overexpression (Fig. 3C). In addition, quantitative real-time PCR showed that ENO1 overexpression also had no effect on the number of copies of late R/gag (Fig. 3D) and 2-LTR circular DNA products (Fig. 3E). However, as expected, nested Alu-gag PCR analysis showed that ENO1 overexpression decreased viral integration efficiency compared with control vector treatment (Fig. 3F). Previously, we reported that high-level-ENO1-packaging virus, which was prepared by cotransfection of HEK293 cells with pNL-CH and ENO1-V5 expression vector, showed decreased number of copies of viral reverse transcription products . Therefore, on the basis of these findings, we hypothesized that ENO1 is a bifunctional inhibitory protein that inhibits reverse transcription and integration processes.
Nuclear ENO1 prevents HIV-1 integration
To prevent HIV-1 integration, ENO1 should be in the viral target cell nucleus. Therefore, we confirmed the subcellular localization of ENO1. Endogenous ENO1 in TZM-bl cells was stained with specific antibodies and detected by fluorescence microscopy. We observed a weak nuclear ENO1-specific signal (Fig. 4A, upper panel). Interestingly, a clearer signal from V5-tagged ENO1, which was expressed by transfection in TZM-bl cells, indicated that ENO1 was present in the nucleus (Fig. 4A, lower panel). To clarify the ENO1 localization in more detail, we next fractionated the cells and detected endogenous ENO1, ENO1-V5, lactate dehydrogenase (LDH) and histone deacetylase1 (HDAC1) by western immunoblotting. LDH and HDAC1 were detected as a cytosolic marker and a nuclear marker, respectively. As shown in Fig. 4B, large amounts of ENO1 and ENO1-V5 were detected in the cytosolic fraction. Interestingly, small amounts of ENO1 and ENO1-V5 were also detected in the nuclear fraction, suggesting that the suppression of HIV infection by ENO1 overexpression (Fig. 3F) depended on V5-tagged ENO1 translocated into the nucleus. To eliminate the effects of the V5 tag, treatment of the untagged ENO1 expression vector increased the amount of ENO1 in the nucleus (Additional file 1: Figure S1A) and enhanced the inhibitory effect of ENO1 (Additional file 1:Figure S1B). Next, since TZM-bl cells are derived from HeLa cells, not immune cells, we also determined whether a small amount of ENO1 is present in the nucleus of immune cells. CD4+ T-cell line CEM cells were fractionated into the cytosol and nucleus fractions using the same method as that for TZM-bl fractionation. As shown in Fig. 4C, a small amount of ENO1 located in the nucleus (Fig. 4C). To further confirm whether acute HIV-1 infection affects ENO1 nuclear translocation, TZM-bl cells were infected with HIV-1 and fractionated. As a result, the same amount of ENO1 was detected in the nuclear fraction from noninfected or infected cells, indicating that ENO1 nuclear localization was unaffected by acute HIV-1 infection. These results suggest that the larger the amount of ENO1 present in the nucleus, the more HIV-1 integration is inhibited.
ENO1 has bifunctional inhibitory activities on HIV-1 infection
Finally, we clarified the bifunctional inhibitory activities of ENO1 in more detail. First, we prepared a high-level-ENO1-packaging virus by cotransfection of HEK293 cells with pNL-CH and ENO1-V5 expression vector. As shown in Fig. 5A, lane 3, the high-level-ENO1-packaging virus showed significant decreases in its infectivity in normal TZM-bl cells. This finding is consistent with our previous findings . Furthermore, as shown in Fig. 3B, the control virus showed significant decreases in its infectivity in the ENO1-overexpressing TZM-bl cells (Fig. 5A, lane 2). As expected, the high-level-ENO1-packaging virus showed a greater reduction in its infectivity in ENO1-overexpressing TZM-bl cells (Fig. 5A, lane 4). Second, we prepared a low-level-ENO1-packaging virus from culture supernatants of ENO1-specific-siRNA-treated CEM/LAV-1 cells. As shown in Fig. 5B, lanes 1 and 2, the control WT virus showed about 60% reduction in its infectivity in ENO1-overexpressing TZM-bl cells. In contrast, the low-level-ENO1-packaging virus also showed about 60% reduction in infectivity in the ENO1-overexpressing TZM-bl cells (Fig. 5, lanes 3 and 4). These findings indicate that ENO1 in viral producer and target cells has bifunctional inhibitory activities on HIV-1 replication.