Differential Effects of the Putative GBF1 Inhibitors Golgicide A and AG1478 on Enterovirus Replication

The genus Enterovirus, belonging to the family Picornaviridae, includes well-known pathogens, such as poliovirus, coxsackievirus, and rhinovirus. Brefeldin A (BFA) impedes replication of several enteroviruses through inhibition of Golgi-specific BFA resistance factor 1 (GBF1), a regulator of secretory pathway integrity and transport. GBF1 mediates the GTP exchange of Arf1, which in activated form recruits coatomer protein complex I (COP-I) to Golgi vesicles, a process important in transport between the endoplasmic reticulum and Golgi vesicles. Recently, the drugs AG1478 and Golgicide A (GCA) were put forward as new inhibitors of GBF1. In this study, we investigated the effects of these putative GBF1 inhibitors on secretory pathway function and enterovirus replication. We show that both drugs induced fragmentation of the Golgi vesicles and caused dissociation of Arf1 and COP-I from Golgi membranes, yet they differed in their effect on GBF1 localization. The effects of AG1478, but not those of GCA, could be countered by overexpression of Arf1, indicating a difference in their molecular mechanism of action. Consistent with this idea, we observed that GCA drastically reduced replication of coxsackievirus B3 (CVB3) and other human enterovirus species, whereas AG1478 had no effect at all on enterovirus replication. Time-of-addition studies and analysis of RNA replication using a subgenomic replicon both showed that GCA suppresses RNA replication of CVB3, which could be countered by overexpression of GBF1. These results indicate that, in contrast to AG1478, GCA inhibits CVB3 RNA replication by targeting GBF1. AG1478 and GCA may be valuable tools to further dissect enterovirus replication.The Enterovirus genus, belonging to the Picornaviridae family, includes many important human pathogens, such as poliovirus, human rhinovirus, echovirus, and coxsackievirus. These are small, nonenveloped viruses that contain a single-stranded RNA genome of positive polarity. The genome is approximately 7.5 kb in length and encodes a single large polyprotein, which is processed into capsid proteins, encoded in the P1 region, and the nonstructural P2 and P3 region proteins that mediate viral RNA replication.Brefeldin A (BFA), a fungal metabolite, is a well-known inhibitor of enteroviruses. BFA blocks transport of cargo from the endoplasmic reticulum (ER) to the Golgi vesicles by disrupting the Golgi vesicles and ER-Golgi intermediate compartment (ERGIC) integrity through inhibition of several guanine nucleotide exchange factors (GEFs), including Golgi-specific BFA resistance factor 1 (GBF1), BFA-inhibited GEF 1 (BIG1), and BIG2 (3, 18). These GEFs regulate the activity of GTPase ADP-ribosylation factor 1 (Arf1) by stimulating GTP exchange. Upon activation, Arf1-GTP binds to Golgi membranes where it induces formation of secretory vesicles via recruitment of coatomer protein complex I (COP-I), a coatomer protein involved in the transport between the Golgi vesicles and the ER. The inhibitory effect of BFA on enterovirus replication is attributed to the inhibition of GBF1 and does not seem to involve BIG1 or BIG2 (2, 11). Besides enteroviruses, other plus-strand RNA viruses, such as mouse hepatitis virus and hepatitis C virus, also seem to rely on GBF1 for efficient replication (2, 8, 11, 21). The viral protein 3A of the enteroviruses poliovirus and coxsackievirus B3 (CVB3) has been shown to interact directly with GBF1 (22, 22a, 23), but the exact function of this interaction remains to be established.Recently, two compounds, AG1478 and Golgicide A (GCA), have been proposed to specifically inhibit GBF1. AG1478 was identified by screening a library of compounds for their ability to induce Golgi complex disassembly (13). AG1478, known as an inhibitor of the epidermal growth factor receptor (EGFR), had effects on the Golgi membranes highly similar to those of BFA through a mechanism not involving the inhibition of EGFR. Arf1-GTP pulldown assays showed that AG1478 inhibited Arf1 activation. Furthermore, overexpression of GBF1 was shown to counter the effect of AG1478 on COP-I localization. Based on these results, AG1478 was proposed to be a GBF1 inhibitor.GCA was identified in a high-throughput screen for small molecules that protected Vero cells from the effects of Shiga toxin (15). Similar to AG1478 and BFA, GCA was reported to fragment the Golgi vesicles and to inhibit Arf1 activation. Furthermore, overexpression of either wild-type GBF1 or the BFA-resistant mutant GBF1-M832L relieved the effects of GCA. In addition, the authors constructed a structural model of the catalytic Sec7 domain of GBF1 in complex with GCA, showing that GCA binds GBF1 at the same site as BFA. Collectively, their results provided convincing lines of evidence that GCA specifically inhibits GBF1 in a manner similar to BFA and does not act on BIG1 and BIG2.BFA has been instrumental in elucidating the membrane requirements for enterovirus replication. Therefore, we investigated the effects of AG1478 and GCA on enterovirus replication after first characterizing the effects of these drugs on BGM cells, the cell line that we routinely use in our studies on coxsackievirus B3 replication. Treatment with other AG1478 or GCA fragmented the Golgi vesicles and caused dissociation of Arf1 and COP-I from Golgi membranes, yet these drugs had different effects on GBF1 localization. Interestingly, the effects of AG1478, but not those of GCA, could be countered by overexpression of Arf1. Next, GCA was found to abrogate enterovirus replication, whereas surprisingly AG1478 did not affect replication at all. Together these results indicate that AG1478 on one hand and GCA and BFA on the other hand have different mechanisms of action, leading to a disparate effect on enterovirus replication.