In this work we report the modification of the gene expression profile induced by the expression of HPV16 E2.
We used the C-33A cell line to study the changes in the expression level of 10,000 human transcripts when the HPV16 E2 is expressed. In C-33A cells there is not evidence of HPV infection and they represent a convenient model to study the effect of E2 on cellular gene expression, without the involvement of another viral gene. Traditionally it has been considered that the effects observed in the regulation of cellular genes when the protein E2 is expressed in cervical carcinoma derived cell lines is due to the repression of the expression of the viral oncogenes E6 and E7 [2, 26–28]; however, in this work we showed that HPV16 E2 induces changes in the expression of cellular genes, independently of the regulation of the viral oncoproteins E6 and E7.
The present study showed that HPV16 E2 importantly alters the expression profile of cellular genes, preferentially in a negative way, although a large number of genes were up-regulated.
It is well known that E2 protein suppress the activity of papillomavirus promoters by binding to low-affinity binding sites, leading to the displacement of cellular binding factors [8, 29–31]. A similar scenario has been proposed for several cellular promoters, since in cultured primary keratinocytes it has been observed that HPV8 E2 represses the transcriptional activity of the β4 integrin promoter, due in part to its binding to a specific E2 binding site on the promoter and leading to displacement of at least one cellular DNA binding factor. However, growing evidence indicates that protein-protein interactions could be even more significant for E2-mediated transcriptional regulation of cellular genes since has been shown that E2 protein from several papillomavirus physically and functionally interact with a variety of cellular regulatory transcription factors, including Sp1, C/EBP, CBP/p300 and p53 [10, 11, 16, 32, 33]. The interaction with Sp1 is apparently one of the most relevant for transcriptional regulation of cellular genes by E2 since it is involved in the down-regulation of the hTERT promoter by HPV18 E2, but also the interaction with this transcription factor plays an important role in the transcriptional activation of several cellular promoters, including p21 by HPV18 E2  or SF2/ASF by HPV16 E2 proteins .
Even when E2 shows cooperative activation with a variety of sequence-specific DNA binding factors such as AP1, USF, TEF-1, NF1/CTF, and C/EBP [11, 34–37], a direct interaction between E2 and these cooperation partners has only been shown for HPV18 E2 with C/EBP, in the transactivation of the involucrin promoter , suggesting the transcriptional cooperation may also occur without a direct binding of these cellular proteins with E2.
The analysis of our data indicated that HPV16 E2 negatively regulates a higher number of genes (1048 genes) than those positively regulated (581 genes) (Additional file 1, Table S1 and Additional file 2, Table S2). In agreement with results previously reported [10, 11], we found that HPV16 E2 up-regulate the expression of involucrin and cyclin-dependent kinase inhibitor 1A (p21) genes. However, we observed these genes up-regulated at a level below the established cutoffs for our analysis, indicating the relevance of the data set provided by our study for understanding the role of HPV E2 as a regulator of cellular gene expression.
Although we do not rule out the possibility that several genes are regulated by a direct interaction of E2 with specific sequences in the particular promoters, the global effect observed suggest that it could be the consequence of the interaction of E2 with several cellular transcription factors such as Sp1 (apparently the most important). E2 protein could destabilize protein-protein interactions between Sp1 and co-activators resulting in the negative regulation of the transactivation function of Sp1 itself, or E2 bound on the promoter via Sp1 may promote the recruitment of transcriptional co-activators such as p300/CBP and pCAF, leading to the transactivation of cellular promoters [38, 39].
On the other hand, some HPV E2 proteins have been shown to interact with TBP and a number of components of the basal transcription machinery [18, 40–45], regulating the recruitment of the pre-initiation complex and affecting both viral and cellular gene expression. Previous works in our group have demonstrated that E2 protein interacts and cooperates with TAF1 in the activation of E2-dependent viral promoters [18, 40]. An analysis of our results indicates that 55 genes regulated by E2 have a natural regulation for TAF1 (data not shown).
Transregulation of specific cellular promoters could be also dependent on levels of the E2 protein in cells, since high levels of HPV16 E2 are known to result in inhibition of cell growth and promotion of apoptosis probably because with higher E2 levels, cellular metabolism may be compromised, leading to a reduced ability of cellular factors to control expression of several cellular genes. However, even we observed an abundant expression of E2, cell viability and different metabolic aspects of the cells (such as cell death) were not affected in a period of 72 hours post-infection with the AdenoE2 virus (data not shown).
This allows us to assume that the observed modulation of cellular gene expression by E2 is not the consequence of induced quiescence or apoptosis, thus the mechanisms of gene expression regulation by E2 only implicate its transcriptional regulatory properties, strongly influenced by its interaction with cellular proteins.
As expected, the results of the microarray analysis showed that HPV16 E2 affect a variety of cellular pathways (Tables 3 and 4), some of them altered in early stages of cervical cancer development, when E2 is still expressed before the integration of the viral genome into cell chromosome.
Interestingly we observed that the expression of a high number of genes on the WNT-pathway is modified for the expression of E2. In the last few years it has been reported that WNT-pathway is activated by the expression of E6 and E7 viral oncogenes [46–48]. However, our observations suggest that E2 is also targeting this pathway probably with different consequences than the induced by the viral oncogenes, since a tight regulation and controlled coordination of the WNT signaling cascade is required to maintain the balance between proliferation and differentiation. Recently it has been proposed that essentially all cellular information - i.e. from other signaling pathways, nutrient levels, etc. - is funneled down into a choice of coactivators usage, either CBP or p300, by their interacting partner beta-catenin (or catenin-like molecules in the absence of beta-catenin) to make the critical decision to either remain quiescent, or once entering cycle to proliferate without differentiation or to initiate the differentiation process . Since CBP and p300 are also interactors for E2, the function of the WNT-pathway could be deeply modified by the low availability of the coactivators when the viral protein is present.
The control of this pathways in the viral cycle could have biological consequences as in the case of the regulation of cell proliferation, since the induction of Cyclin A expression by E2, orchestrated with a negative regulation of RhoA, known inhibitor of the cell proliferation, would allow the entry into the S phase of cell cycle [50–52]. Similarly, E2 expression could be also involved in apoptosis regulation since it negatively regulate genes involved in this process, such as caspase 9 (CASP9) , whose product is an effector of cell death. In the same way, EGR2 [54, 55] is negatively regulated bringing as a consequence the inhibition of cytochrome c releasing it from the mitochondria. Interestingly, several genes mainly expressed in keratinocytes from the basal layers of stratified epithelia, such as type I keratins (keratin 14, 24 and 34) [56–58], were down-regulated in cells expressing E2 suggesting that the process of cell differentiation could be also regulated by this viral product.