MLN-273 had been extensively tested in mice and showed a good safety profile, except for weight lose at high doses. To choose an appropriate dose for injection, we tested the toxicities of MLN-273 at different doses in HBV transgenic mice. Using 8-week-old HBV transgenic mice, MLN-273 was tested at doses of 0.5, 0.8, 1.0, and 1.2 mg/kg of body weight. The treated mice were weighed daily for 2 weeks. At the highest dose of 1.2 mg/kg, 1 of 5 mice died and the rest lost over 20% of their body weight. At the dose of 1.0 mg/kg, the mice lost about 10% of their body weight. At the doses of 0.5 and 0.8 mg/kg, the mice lost 5 to 10% of their body weight. To test for the toxicity of MLN-273 in these mice at these doses, blood counts and chemistry were examined. There were no significant differences in mice with or without treatment, indicating that MLN-273 is not toxic at the doses used, which is consistent with the animal toxicology data of this drug from the manufacturer. Therefore, MLN-273 at a dose of 0.8 mg/kg was selected for further experiments.
By using a recombinant adenovirus expressing HBV (ad-HBV) and a recombinant baculovirus expressing WHV (bv-WHV), we have previously demonstrated that proteasome inhibitors substantially enhance viral replication of the HBVX−
virus but not that of the wild-type HBV, suggesting an important role of proteasome pathways in HBX-dependent viral replication (30
). To study this effect in vivo
, the proteasome inhibitor MLN-273 was then tested in HBV transgenic mice expressing wild-type or X-negative HBV.
Eight-week-old HBV transgenic mice were injected with 0.8 mg/kg of MLN-273 i.v. on days 0 and 3. Serum samples harvested at weeks 0, 1, and 4 postinjection were analyzed for HBV DNA, HBeAg, and HBsAg levels. In the transgenic mice expressing wild-type HBV (HBVWT), the serum HBV DNA and HBeAg levels were slightly increased after MLN-273 injection, and the serum HBsAg level did not change much with treatment (Fig. ). Consistent with previous reports, transgenic mice with the HBX-defective HBV genome (HBVX−
) showed much lower HBV DNA levels than the HBVWT mice. In HBVX−
mice, both HBV DNA and HBeAg levels increased significantly (>100-fold for HBV DNA, ~10-fold for HBeAg) at week 1 postinjection. As expected, this effect disappeared at week 4. In contrast, serum antigen HBsAg levels did not show any change in the course of the experiment (Fig. ), consistent with our previous in vitro
) showing that HBsAg levels do not correlate with HBV DNA levels. The increase in serum HBeAg level likely reflected the increase in serum HBV DNA titer, because viral transcription was not affected by the proteasome inhibitor treatment (see Fig. ). Studies on other transgenic mouse models have also shown that HBcAg, which is a component of the virion, can be degraded in vivo
to protein species with HBeAg reactivities (20
). Our previous study concerning the effect of proteasome inhibitors on HBV replication in vitro
showed that HBeAg production was minimally affected by proteasome inhibitor treatment, whereas viral replication was markedly affected (30
). This observation is consistent with the fact that serum HBeAg has been used as a marker of viremia in vivo
prior to the development of sensitive HBV quantification assays (17
FIG. 2. Effect of MLN-273 on serum HBV markers in transgenic mice. HBV transgenic mice at the age of 8 weeks were treated with MLN-273 at a dose of 0.8 mg/kg or with PBS, and serum and liver samples were harvested at different time points as described. In HBVWT (more ...)
FIG. 3. Effect of MLN-273 on hepatic HBV viral replication, transcription, and protein expression in transgenic mice. Livers from the HBV transgenic mice were harvested for DNA, RNA, and protein analyses. (A) Effect on HBV replication. HBV core-associated DNA (more ...)
To further test the effect of MLN-273 on HBV replication, transcription, and protein expression, livers were harvested from the transgenic mice at week 1 after injection. Cytoplasmic viral core particles were isolated, and core-associated viral DNA was extracted for the analysis of replicative intermediates. MLN-273 had little effect on viral replication in the liver of the HBVWT mice. Similar to serum HBV DNA levels, viral replication and core protein levels in the HBVX− livers were much lower than those in HBVWT livers. MLN-273 treatment significantly increased and restored the HBV replication and core protein levels to almost the same levels as those of HBVWT mice (Fig. ). However, MLN-273 treatment had little effect on the viral transcripts, including the 3.5-, 2.4-, and 2.1-kb mRNAs (Fig. ). This observation supports the notion that MLN-273 acts on viral replication without affecting viral transcription in an HBX-dependent manner.
To confirm that the proteasome activities were indeed inhibited by MLN-273 in the HBV transgenic mice, these mice were treated with MLN-273 at doses of 0.4 and 0.8 mg/kg. At 24 h postinjection of MLN-273, the livers were harvested and proteasomes were purified for proteasome activity determination. Without MLN-273 treatment, the proteasome activity of the HBVWT mice was somewhat lower than that of the HBVX−
mutant mice (0.584 ± 0.15 versus 0.748 ± 0.21 pmol/s; P
< 0.01) (Fig. ). This minor effect could possibly be due to inhibition of the proteasome activities by the X protein from the wild-type HBV genome (14
). In both HBVWT and HBVX−
mice, MLN-273 treatment reduced the proteasome activities significantly, but the magnitude of reduction appeared to be greater in the HBVX−
mice (Fig. ). These data indicate that MLN-273 was indeed active in inhibiting the proteasome activities in vivo
FIG. 4. Effect of MLN-273 on hepatic proteasome activities of HBV transgenic mice. HBV transgenic mice at the age of 8 weeks were injected i.v. with MLN-273 at doses of 0.4 and 0.8 mg/kg. At 24 h postinjection of MLN-273, the livers were harvested and the proteasomes (more ...)
The HBV transgenic mouse experiments demonstrated that proteasome inhibitor MLN-273 enhanced viral replication markedly in the HBVX−
mice but only slightly in the HBVWT transgenic mice. To further confirm this finding, we applied another HBV mouse model, in which a recombinant adenovirus containing a replication-competent HBV genome can be inoculated into C57BL/6 mice and initiate HBV replication in the mouse liver (25
). Eight-week-old C57BL/6 mice were inoculated with 1 × 108
PFU/ml of ad-HBV i.v., followed by two injections of MLN-273 i.p. at a dose of 0.8 mg/kg on days 0 and 1. Serum samples were harvested on days 2, 4, and 6 and analyzed for HBV DNA, HBeAg, and HBsAg levels. In mice injected with ad-HBV, there was no significant difference with or without MLN-273 treatment (Fig. ). In mice infected with ad-HBVX−
, the levels of serum HBV DNA and HBeAg were much lower (around 35%) than those in mice infected with ad-HBV, but HBsAg levels were comparable. However, MLN-273 treatment significantly enhanced both serum HBV DNA and HBeAg levels in ad-HBV-infected mice by 10- to 20-fold as early as 1 day postinjection but had little effect on HBsAg levels (Fig. ). To ensure that measurement of serum HBV DNA represents authentic replicating HBV and not injected recombinant ad-HBV, we showed that by adenovirus-specific PCR, adenoviral DNA (detection limit of 103
copies/ml) could no longer be detected in serum 2 days after injection. Therefore, the detected HBV DNA indeed represents authentic replicating HBV DNA.
FIG. 5. Effects of MLN-273 on HBV replication in C57BL/6 mice infected with HBV adenoviruses. C57BL/6 mice at the age of 8 weeks were infected with ad-HBVWT (A) and ad-HBVX− (B) and treated with proteasome inhibitor MLN-273 at a dose of 0.8 mg/kg or with (more ...)
Our study demonstrates that cellular proteasome plays an important role in HBV replication in vivo
and that this effect is mediated by the function of HBX. This result is consistent with our previous findings demonstrating the structural and functional interaction of HBX with the proteasome complex in cell culture models. In these in vivo
experiments, we showed that the effect of proteasome inhibition on HBV replication resides in a posttranscriptional step. It is possible that the assembled core particles containing the replicative intermediates are targets of cellular proteasome degradation. HBX, functioning like a proteasome inhibitor (14
), blocks this crucial step during productive HBV infection in vivo
. Further studies are necessary to confirm this possibility and to characterize the mechanism of these functional interactions. Proteasome has also been linked to interferon-mediated suppression of HBV replication in vitro
and in vivo
), suggesting that proteasome may play an important role in the innate antiviral response in addition to its role in antigen presentation and activation of adaptive immunity. Cellular proteasome has already been shown to play a direct antiviral role in other viral infections (18
). Therefore, targeting the interaction between HBX and proteasome may provide a novel therapeutic strategy for anti-HBV development.