BI-2 destabilizes HIV-1 cores during infection and Prevents Binding of CPSF6 to the HIV-1 Capsid

Background The recently discovered small-molecule BI-2 potently blocks HIV-1 infection. BI-2 binds to the N-terminal domain of HIV-1 capsid. BI-2 utilizes the same capsid pocket used by the small molecule PF74. Although both drugs bind to the same pocket, it has been proposed that BI-2 uses a different mechanism to block HIV-1 infection when compared to PF74. Findings This work demonstrates that BI-2 destabilizes the HIV-1 core during infection, and prevents the binding of the cellular factor CPSF6 to the HIV-1 core. Conclusions Overall this short-form paper suggests that BI-2 is using a similar mechanism to the one used by PF74 to block HIV-1 infection.


Findings
The ability of the novel HIV-1 inhibitor BI-2 to potently block HIV-1 infection has been correlated with stabilization of in vitro assembled HIV-1 CA-NC complexes [1][2][3]. Crystal structure of the drug with the N-terminal domain of capsid (CA NTD ) revealed that BI-2 binds in the site 2 pocket [1], as it has been shown for the small-molecule inhibitor PF74 [1,4,5]. Using a novel capsid stability assay, we have demonstrated that BI-2 and PF74 stabilize in vitro assembled HIV-1 capsidnucleocapsid (CA-NC) complexes [2]. Counter intuitively, PF74 destabilizes the HIV-1 core during infection of cells [5]. In addition, several reports have demonstrated that PF74 prevents the binding of the cellular factor cleavage and polyadenylation specific factor 6 (CPSF6) to the viral capsid [2,6]. Previous observations have shown that BI-2 stabilizes in vitro assembled HIV-1 CA-NC complexes by using two different assays [1,2]. Because BI-2 has been suggested to inhibit HIV-1 infection, at least in part, by stabilizing the viral capsid [1,2], we investigated the effects of BI-2 in infection by analyzing 1) HIV-1 DNA metabolism, 2) the fate of the HIV-1 capsid, 3) binding of CPSF6 to HIV-1 capsid, and 4) the ability of BI-2 to block infection by other retroviruses.

BI-2 blocks infection of HIV-1 after reverse transcription but prior to nuclear import
We initially studied the ability of BI-2 to block HIV-1-GFP infection in canine Cf2Th cells at the indicated concentrations ( Figure 1A). As a control, we performed similar experiments using the small-molecule PF74 [1,2,4,5]. Our experiments showed that 50 μM of BI-2 is equivalent to 5 μM of PF74 when comparing inhibition of HIV-1-GFP infection ( Figure 1A). These drugs did not exhibit cellular toxicity at the used concentrations, as determined by propidium iodide exclusion [7]. Next we challenged dog Cf2Th cells with similar amounts of HIV-1-GFP in the presence of BI-2. Infections were harvested at 7, 24 and 48 hours post-infection to analyze late reverse transcripts (LRT) (B), formation of 2-LTR circles (C) and infectivity (D), respectively. As a control, we performed similar infections in the presence of DMSO. To control for a block in reverse transcription, we used the inhibitor nevirapine [8], which completely blocks HIV-1-GFP reverse transcription ( Figure 1B). BI-2 did not affect the occurrence of reverse transcription when compared to the effect of nevirapine ( Figure 1B); this result is reminiscent of the effect of the related small molecule BI-1 to reverse transcription [1]. However, BI-2 potently blocked the formation of 2-LTR circles ( Figure 1C). These results indicated that BI-2 blocks HIV-1-GFP infection after reverse transcription but prior to nuclear import, as demonstrated for BI-1 [1]. PF74 had a greater effect on the occurrence of reverse transcription when compared to BI-2, and potently blocked the formation of 2-LTR circles ( Figure 1B-C), as previously shown [4,5]. Inhibition of HIV-1-GFP infection by BI-2 was comparable to PF74 at the indicated concentrations ( Figure 1D). Previous observations showed that BI-1, a similar molecule to BI-2, did not affected the occurrence of reverse transcription [1]. Next we measured occurrence of HIV-1 reverse transcription in the presence of different concentrations of BI-2. To this end, we challenged dog Cf2Th cells with similar amounts of HIV-1-GFP in the presence of the indicated concentrations of BI-2, and measured the occurrence of reverse transcription and infection at 7 and 48 hours post-infection, respectively ( Figure 1E). In agreement with previous findings using BI-1 [1], these experiments showed that BI-2 does not affect the occurrence of reverse transcription. As a control, we performed similar infections in the presence of nevirapine ( Figure 1E), an inhibitor of reverse transcription. In addition, we monitored HIV-1 and HIV-1-T107N LRTs at 7, 24, and 48 hours post-infection in the presence of BI-2 or PF-74 ( Figure 1F). Similarly, we found that BI-2 did not affect the formation of HIV-1 LRTs. In addition, BI-2 did not affect the formation of LRTs by HIV-1-T107N.

BI-2 destabilizes the HIV-1 core during infection
We investigated the fate of the HIV-1 capsid in the presence of BI-2. For this purpose, we challenged Cf2Th cells with HIV-1-GFP in the presence of 50 μM BI-2 and performed the fate of the capsid 12 hours post-infection, as previously described [9][10][11]. As shown in Figure 2A, the use of BI-2 destabilized the HIV-1 core during infection when compared with the DMSO control. As a control, we used 5 μM PF74 that destabilized the HIV-1 core (Figure 2A), as previously shown [5]. Our results suggested that BI-2, like PF74, destabilizes the HIV-1 core during infection. To show that the destabilization of the HIV-1 core observed in the presence of BI-2 is specific to capsid, we performed the fate of the capsid assay using an HIV-1-GFP virus bearing the mutation T107N, which confers HIV-1 resistance to BI-2 and PF74 [1,4]. As shown in Figure 2B, BI-2 and PF74 did not affect the stability of the HIV-1 core bearing the change T107N. These results suggested that the ability were measured as described [2]. Input and Pellet fractions were analyzed by Western blotting using antibodies against HIV-1 CA p24. As control, stability of in vitro assembled HIV-1 CA-NC complexes in stabilization buffer was measured. Similar results were obtained in three independent experiments. (D) Stability of wild type (upper panel) or T107N mutant (lower panel) in vitro assembled HIV-1 CA-NC complexes in stabilization buffer containing cell extracts at increasing concentrations of BI-2 was measured, as described [2]. Similar results were obtained in three independent experiments. of these drugs to destabilize the HIV-1 core is specific to capsid. As a control, we showed that TRIM5α rh destabilizes the HIV-1 core ( Figure 2B), as previously shown [12,13]. Next, we tested the ability of BI-2 to stabilize in vitro assembled HIV-1 CA-NC complexes using our previously published assay [2]. As we have previously shown, BI-2 as well as PF74 stabilize HIV-1 CA-NC complexes ( Figure 2C) [2]. These results showed that BI-2, like PF74, stabilizes in vitro assembled HIV-1 CA-NC complexes, which is in agreement with previous reports [1,2]. Contrary to in vitro assembled HIV-1 CA-NC complexes that are mainly composed of capsid hexamers [14], the HIV-1 core is composed of capsid pentamers and hexamers [15,16]. The mature fullerene core is an assembly of capsid subunits displaying multiple quasi-equivalent conformations, which arise in part The ability of CPSF6 to bind in vitro assembled HIV-1 CA-NC complexes in the presence of BI-2 was analyzed as previously described [22]. Briefly, extracts of 293 T cells transiently transfected with a CPSF6-FLAG construct (Input) were incubated with in vitro assembled HIV-1 or SIV mac CA-NC complexes in the presence BI-2 for 1 h. Subsequently, extracts were applied onto 70% sucrose cushion and centrifuged, and the pelleted fraction was collected (Pellet). Input and Pellet fractions were analyzed using anti-FLAG and anti-p24 antibodies. As control, the binding of CPSF6 to in vitro assembled HIV-1 CA-NC complexes was studied in the presence of PF74. Similar results were obtained in three independent experiments and a representative experiments is shown. To control for the bona fide origin of in vitro assembled SIV mac CA-NC complexes, we tested the ability of TRIM5α protein from Tamarin monkeys tagged with an HA epitope (TRIM5α Tamarin - from the flexibility between the N-terminal and C-terminal domains of capsid. These multiple quasi-equivalent conformations result in the formation of hexamers and pentamers, which allow the formation of a curved capsid lattice. One possibility is that BI-2 and PF-74 limits the flexibility of the capsid to a range compatible only with the formation of hexamers; this might be the reason that BI-2 and PF74 stabilize in vitro assembled HIV-1 CA-NC complexes but destabilize the HIV-1 core during infection. A second possibility is that BI-2 requires the presence of cellular factors in order to destabilize in vitro assembled HIV-1 CA-NC complexes. To rule out that the ability of BI-2 to destabilize capsid complexes depends upon the presence of cellular factors, we tested the ability of BI-2 to destabilize in vitro assembled HIV-1 CA-NC complexes in the presence of cellular extracts. As shown in Figure 2D, the presence of cellular extracts did not alter the ability of BI-2 to destabilize capsid. As a control, similar experiments were performed using the capsid mutant T107N, which is resistant to BI-2 ( Figure 2D). Future structural studies will shed light on this discrepancy.

BI-2 prevents the binding of CPSF6 to in vitro assembled HIV-1 CA-NC complexes
Expression of CPSF6 is required for the HIV-1 infection phenotype observed in human TNPO3-depleted cells [17][18][19]. We and others have previously demonstrated that the small-molecule HIV-1-inhibitor PF74 prevents the binding of CPSF6 to HIV-1 capsid [6,17]. Because of the similar phenotypes observed for HIV-1-GFP infection when using BI-2 and PF74, we tested the ability CPSF6 to bind in vitro assembled HIV-1 CA-NC complexes in the presence of BI-2. As shown in Figure 3A, BI-2 prevents the ability of CPSF6 to bind in vitro assembled HIV-1 CA-NC complexes. As previously shown, PF74 also prevented the binding of CPSF6 to in vitro assembled HIV-1 CA-NC complexes [17]. Interestingly, BI-2 and PF74 also inhibited the binding of CPSF6 to in vitro assembled simian immunodeficiency virus (SIV mac ) CA-NC complexes ( Figure 3A). As a control to show the bona fide origin of the SIV mac capsid, we showed that TRIM5α from tamarin monkeys binds to in vitro assembled SIV mac CA-NC complexes ( Figure 3A) [20]. These results suggested that BI-2 prevents the binding of CPSF6 to the HIV-1 and SIV mac cores. Next, we performed a dose response curve to better understand the ability of BI-2 to prevent the binding of CPSF6 to in vitro assembled HIV-1 CA-NC complexes. As shown in Figure 3B, we observed that using BI-2 at 50 μM completely inhibit the binding of CPSF6 to in vitro assembled HIV-1 CA-NC complexes. For comparison, we showed a dose response curve to understand the ability of PF74 to interfere with the binding of CPSF6 to in vitro assembled HIV-1 CA-NC complexes. As we have previously shown using PF74 at 5 μM inhibited the ability of CPSF6 to bind in vitro assembled HIV-1 CA-NC complexes [17]. Altogether these results showed that BI-2 prevents the ability of CPSF6 to interact with the HIV-1 core. Because CPSF6 binds to nucleic acids, and HIV-1 CA-NC complexes are assembled in the presence of nucleic acids [21], we performed a control to demonstrate that the interaction of CPSF6 with in vitro assembled HIV-1 CA-NC complexes is only dependent upon the capsid protein.
To this end, we tested the ability of CPSF6 to bind in vitro assembled HIV-1 CA-NC complexes bearing the change N74D, which confers HIV-1 resistance to the overexpression of cytosolic CPSF6 [17,18]. As shown in Figure 3C, CPSF6 did not bind to vitro assembled HIV-1 CA-NC complexes bearing the change N74D. These results indicated that CPSF6 is specifically binding to capsid, as shown [17].

Ability of BI-2 to block infection by different retroviruses
Next we explored the ability of BI-2 to block infection by different retroviruses. For this purpose, we challenged Cf2Th cells with increasing amounts of different retroviruses expressing GFP as reporter of infection (Figure 4 Viruses expressing GFP as a reporter were prepared as previously described [23]. Interestingly, BI-2 potently blocked HIV-1 and SIV mac but not HIV-2 ROD , BIV, FIV, EIAV, N-MLV, B-MLV and Mo-MLV. As a control, we performed similar infections in the presence of PF74 ( Figure 4). As previously shown PF74 blocks HIV-1-GFP and SIV mac -GFP infection [5,6,17,24]. Interestingly, we found a parallel between the ability of BI-2 to inhibit infection by HIV-GFP and SIV mac -GFP with the ability of BI-2 to prevent the binding of CPSF6 with the HIV-1 and SIVmac cores.
This short-form article thoroughly examined and compared the effects of BI-2 and PF74 on HIV-1 infection. Our novel findings demonstrate that BI-2, similar to PF74, destabilizes the HIV-1 core during infection and prevents the binding of CPSF6 to the HIV-1 core.