In this study we applied a comparative analysis of changes of the host cell gene expression profile following infection with CPXV, MPXV or VACV, respectively. We addressed the question whether the different characteristics of each virus, especially the distinctive repertoire of host modulating factors encoded by each virus, may in part be reflected by different characteristics of the infected cells. OPV are well known to suppress the antiviral host defence, exploit the host cell machinery for reproduction and to inhibit or delay cell death. However, most of our knowledge about these mechanisms originates from studies using VACV. More recently, after the US outbreak in 2003, MPXV has come into focus [37, 38], but again, our knowledge about the differences in virus–host interactions between VACV and MPXV and even more so between VACV, MPXV and CPXV is still limited. In this study we investigated the host-cell transcriptome of CPXV-, MPXV- or VACV-infected cells to explore the different capabilities of each virus to interact with the host cell.
We found the host cell transcriptome to be mainly unaffected by poxviral infection, despite the major morphological changes induced by infection. Although this holds true for OPV infection in general, the impact of MPXV infection on the transcriptome seems to be uniquely low. This might indicate an even more elaborate “stealth” program than that performed by VACV or CPXV, allowing MPXV to especially well avoid the responses of the innate and adaptive immune systems to the developing infection. It would be interesting to investigate if this is true for MPXV infection in general, or if it might be a characteristic of the highly pathogenic central African MPXV strains in contrast to less pathogenic west African strains .
The data presented are in agreement with the results by Rubins et al. which show no general decline in cellular mRNAs after infection of HeLa cells with MPXV (Zaire strain) or VACV (Western Reserve strain) . In our study we could prove that the same was true for CPXV infection. Other studies describe that the cellular gene expression is generally suppressed in response to infection and that only few genes are specifically induced [19, 23, 26]. In our study, we found that more than two thirds of the few host cell transcripts which exhibited regulation in response to infection were upregulated at 6 h post infection. However, we cannot exclude the possibility of a more pronounced downnregulation of host cell genes at later stages of infection. Furthermore, in our study, only genes were taken into account that exhibited at least 2-fold change in gene expression, to set the focus on genes that exhibit stronger regulation. Therefore, a mild global downregulation of cellular genes which was below the cut-off value might have been neglected (see also Additional file 3). However, while repression of host genes by OPV is thought to be generalised and likely nonspecific, cellular genes which are induced by infection are of particular interest, as they are thought to play key roles in viral replication or host response to infection [19, 22, 25].
We found the infection-induced changes of the gene expression profiles of CPXV-, MPXV- and VACV-infected cells to be largely different. However, we could still identify several cellular transcripts which showed similar modulation after infection with each virus, the most significant one being histone mRNAs. An apparent induction of histone genes has been described previously in other studies [19–21, 26, 27]. However, this is thought to be an experimental artefact caused by de novo polyadenylation of the histone transcripts by the viral poly(A) polymerase, as histone mRNAs are the only mRNAs in eukaryotes that lack a poly(A) tail in general [22, 27]. This is supported by the fact that Yang et al. reported histone mRNAs to be overrepresented at 4 h post VACV infection even after poly(A)-specific mRNA isolation procedures . If this was the case, amplification of histone mRNAs from infected samples would be greatly enhanced by the poly(A)-dependent procedure we used. We therefore decided to exclude histone mRNAs from further analysis.
After exclusion of histone mRNAs, we could identify 87 cellular transcripts which seemed to be commonly modulated by OPV infection in general. Most noticeable was an apparent infection-induced upregulation of genes involved in inactivation of MAPK activity. The upregulated Sprouty (SPRY2/4) and Sprouty-related (SPRED1/2) proteins are known to be induced by growth factor receptor activation-mediated MAPK-ERK activation via Ras as a self-regulatory feedback inhibition mechanism [31–33]. Similarly, the upregulated DUSP5/6 and EGR1/2 genes are commonly induced as early response genes after activation of the MAPK-ERK signalling pathway and act as negative regulators of ERK phosphorylation [34, 35]. Therefore we suggest that the observed upregulation of genes involved in inactivation of MAPK activity may reflect a feedback mechanism towards virus-induced stimulation of MAPK-ERK activity, probably via an enhanced virus-induced signalling through EGF receptors. This theory is supported by the observed overrepresentation of components of the ErbB pathway among the commonly affected transcripts and by the upregulation of EGF family members following infection. Although modulation of the MAPK-ERK pathway seems to be a common effect of OPV infection, its importance differs from virus to virus. Silva et al. showed that inhibition of MEK/ERK signalling resulted in a significant decrease in VACV yield, but had no impact on CPXV replication [36, 39, 40]. The importance of MAPK-ERK signalling in the context of MPXV replication has not been addressed yet.
Despite these few commonly affected gene sets, major differences in the transcriptional response towards infection with CPXV, MPXV or VACV prevailed. Most interesting was an enrichment of genes involved in immunity-associated processes and pathways in CPXV- and MPXV-infected cells, which was strikingly absent following VACV infection. This included especially genes that are involved in leukocyte migration and Toll-like receptor signalling, which seemed to be affected by CPXV as well as by MPXV infection.
Interestingly, we could identify several pathways specific to certain infectious diseases to be affected by CPXV and MPXV infection. As most of these pathways were specific to infection by intracellular pathogens, this result may be explained by universal mechanisms of host modulation and inflammatory host response which could be affected alike by these diseases and by CPXV and MPXV.
A large proportion of the immunity-associated genes which were affected by CPXV and MPXV consisted of pro-inflammatory cytokine genes and genes involved in leukocyte chemotaxis or activation of immune cells. As most of these genes showed pronounced upregulation after CPXV and MPXV infection, this might reflect an inadequate subversion of the hosts’ antiviral response. This may be supported by the fact that an induction of genes implicated in the immune response, e.g. IL6, could also be observed in response to attenuated modified VACV Ankara infection of HeLa cells  but not in response to non-attenuated VACV WR infection . Interestingly, an induction of pro-inflammatory cytokines and chemokines in response to CPXV and MPXV infection was observed by other studies as well. Increased IL8 gene expression following MPXV infection of MK2 cells was reported by Alkhalil et al. . And in vivo, strong secretion of IL-6, IL-8, and G-CSF or of IL-6, IL-8 and CCL-2, respectively, was observed following infection of cynomolgus macaques (Macaca fascicularis) with MPXV or CPXV, respectively [41–43].