SUMOylation of viral proteins is widespread and serves to modify or regulate the properties of those proteins. Papillomaviruses are a large group of small DNA viruses that infect the skin, leading to benign lesions (warts) that in some cases can progress to malignancy. The papillomavirus life cycle is intimately connected with the differentiation process of stratified epithelium, and several viral early proteins function to modulate the host cell environment. One of the critical early proteins is the E2 protein, which functions in both viral replication and transcription. In the present paper, we demonstrate that E2 proteins are SUMOylated and that overexpression of SUMOylation results in a dramatic increase in intracellular levels of the E2 protein. We have shown previously that there is increased SUMOylation during keratinocyte differentiation, suggesting that the levels of E2 protein may be tied to changes in the cellular SUMOylation state during differentiation. In addition to itself being regulated by SUMOylation, E2 appears to influence the SUMOylation state of one of its binding partners, the cellular transcription factor, C/EBP (CCAAT/enhancer-binding protein). Overall, these observations indicate a complex interplay between this viral protein and the host SUMOylation system.

SUMOylation

SUMOylation is a protein modification system related to the ubiquitin family that utilizes a small protein, known as SUMO (small ubiquitin-related modifier), as the modifying group [1]. There are three well-characterized SUMOs, SUMO1, SUMO2 and SUMO3; SUMO1 shares ∼48% identity to SUMO2/3 at the primary sequence level, and SUMO2 and SUMO3 share ∼97% identity to each other in the mature forms. Each of the SUMOs is covalently attached to substrate proteins through a series of enzymatic steps utilizing a single heterodimeric activating enzyme [SAE1/SAE2 (SUMO1-activating enzyme subunits 1 and 2); the E1 enzyme], a single monomeric conjugating enzyme [Ubc9 (ubiquitin-conjugating enzyme 9); the E2 enzyme] and one of several SUMO ligases {e.g. PIAS [protein inhibitor of activated STAT (signal transducer and activator of transcription)] proteins, Pc2 (polycomb 2 protein) and RanBP2 (Ran-binding protein 2); the E3 enzymes}. Recent proteomics approaches indicate that numerous cellular proteins, primarily in the nucleus, are SUMOylated [2,3]. Additionally, there are distinct differences in the populations of proteins modified by SUMO1 versus SUMO2/3, implying that the functional and biological roles of these two families are likely to be somewhat different. Nonetheless, most of the known SUMOylation targets are transcription factors or other proteins involved in chromatin structure, regulation and expression [4], supporting a fundamental role for this modification system in the cellular life cycle.

Viruses and SUMOylation

Consistent with the nuclear predominance of SUMOylation, most of the nuclear DNA viruses (parvovirus, adenovirus, papovavirus and herpesvirus) have viral proteins that are SUMOylated and/or interact with the SUMOylation machinery, as does HIV and several other RNA viruses [5]. While the functional significance of the viral–SUMOylation interactions is not always precisely defined, presumably these viruses have evolved to utilize or subvert the SUMOylation system to their own replicative advantage. Our own studies have focused on the papillomavirus system as these small DNA viruses are particularly dependent on the host cell machinery for their maintenance and reproduction.

Papillomavirus and SUMOylation

HPVs (human papillomaviruses) are small non-enveloped viruses with a genome of supercoiled, double-stranded DNA (approx. 8000 base-pairs); there are more than 100 HPV types identified and characterized based on their DNA sequences similarity. These viruses infect both mucosal and cutaneous epithelial tissues, and the HPV life cycle is strongly influenced by the process of keratinocyte differentiation [6]. HPV is responsible for skin warts, laryngeal papillomas and cervical carcinoma, and is also correlated with certain skin cancers as well as some head and neck sarcomas. Sexually transmitted HPVs infect the genital track and are classified as ‘low risk’ or ‘high risk’ based on their capability to cause cervical carcinoma.

Establishment of infection with HPV requires initial entry into the basal cells of the epithelium, but the productive life cycle of HPV is directly linked to keratinocyte differentiation and stratification [7]. In basal cells, where the non-productive stage of the viral life cycle takes place, viral gene expression is very limited. Minimal viral replication occurs in synchrony with basal cell proliferation, resulting in a persistent low copy number of viral genomes that segregate into daughter cells at division. When normal basal cells leave the replicative layer, they detach from the basal membrane, move toward the terminal differentiated layers of the epithelium and progressively exit the cell cycle to finally be shed at the skin surface. However, when HPV-infected basal cells leave the basal layer and move towards the surface of the skin, they re-enter the S-phase and keep their proliferating properties, perturbing the normal process of differentiation. Concomitant with this process, productive viral replication ensues, resulting in enhanced numbers of viral genome copies and production of new virions.

The viral genome contains eight ORFs (open reading frames) producing six non-structural regulatory proteins (E1, E2, E4, E5, E6 and E7) and two structural viral capsid proteins (L1 and L2). E1 and E2 are the viral replication proteins and will be the focus of the present review. E1 is the viral origin recognition protein and eventually assembles into a hexameric helicase that both catalyses viral DNA strand separation and also recruits host cell DNA replication proteins [8]. Previous work from our laboratory has demonstrated that E1 is SUMOylated, that PIAS proteins can serve as SUMO ligases for this reaction and that SUMOylation of E1 is necessary for proper nuclear localization [9,10]. The mechanism by which SUMOylation affects E1 nuclear localization is still unclear as SUMOylation does not appear to affect nuclear import [11] or export (G. Rosas-Acosta and V.G. Wilson, unpublished work) of E1 directly.

E2 is a multifunctional protein (Figure 1) that (i) acts as a loading factor to assist assembly of E1 on the viral origin, (ii) facilitates viral genome segregation by tethering the viral genome to host chromosomes, and (iii) has transcriptional regulatory activity affecting both viral and cellular genes. E2 binds directly to viral DNA sequences to control viral gene expression, but can also affect transcription through protein–protein interactions. For example, E2 binds the host transcriptional factor, C/EBP (CCAAT/enhancer-binding protein), and decreases C/EBP transcriptional activity through a molecular mechanism that has not been determined [12]. Ongoing studies in our laboratory have begun to examine the role of SUMOylation in these E2 functions.

Schematic diagram of the HPV E2 domain structure

Figure 1
Schematic diagram of the HPV E2 domain structure

The N-terminal region possesses the transcriptional transactivating function and also the mitotic tethering function through binding to the host chromatin protein, Brd4. The C-terminal domain possesses site-specific DNA binding and dimerization activities. Between the N- and C-terminal functional domains is a hinge domain whose length varies with different HPV types. There is a single high-probability SUMOylation site that is conserved among high-risk HPV E2s and maps to the DNA-binding domain as shown.

Figure 1
Schematic diagram of the HPV E2 domain structure

The N-terminal region possesses the transcriptional transactivating function and also the mitotic tethering function through binding to the host chromatin protein, Brd4. The C-terminal domain possesses site-specific DNA binding and dimerization activities. Between the N- and C-terminal functional domains is a hinge domain whose length varies with different HPV types. There is a single high-probability SUMOylation site that is conserved among high-risk HPV E2s and maps to the DNA-binding domain as shown.

High-risk HPV E2 proteins are readily SUMOylated in vitro and in Escherichia coli using SUMO1. In contrast, E2 shows a strong preference for SUMO2/3 in vivo, with little detectable SUMOylation by SUMO1. Additionally, exogenous expression of SUMO2 or SUMO3 with Ubc9 dramatically stabilizes intracellular E2 levels, while SUMO1 shows no effect on E2 stability. However, this effect on stability does not appear to be directly related to SUMOylation of E2 itself, as mutation of the predominant SUMOylation site, lysine, of E2 does not render the mutant unresponsive to the effect of SUMO2/3 plus Ubc9. Intriguingly, we have recently demonstrated that there is an up-regulation of Ubc9 and SUMO2/3 at both the transcript and protein levels during keratinocyte differentiation, while SUMO1 is relatively unchanged [13]. These combined results suggest that in papillomavirus-infected skin, E2 protein levels may be regulated by changes in SUMOylation state coupled with the differentiation process (Figure 2). Thus, in basal cells where SUMO2/3 is low, E2 would be unstable and would be present at levels sufficient only for low copy number viral genome persistence. As infected cells leave the basal level and begin differentiation, increased SUMO2/3 SUMOylation could lead to E2 stabilization and higher protein concentrations consistent with productive viral genome amplification.

Model of SUMO2/3 regulation of E2 during keratinocyte differentiation

Figure 2
Model of SUMO2/3 regulation of E2 during keratinocyte differentiation

The image shows HPV-infected, stratified, human keratinocytes with the brackets indicating the basal cell layer. Details of the model are described in the text.

Figure 2
Model of SUMO2/3 regulation of E2 during keratinocyte differentiation

The image shows HPV-infected, stratified, human keratinocytes with the brackets indicating the basal cell layer. Details of the model are described in the text.

In addition to being a SUMO substrate, we have also demonstrated that E2 can inhibit the SUMOylation of a binding partner, the cellular transcription factor, C/EBP β1. As SUMOylation of C/EBP β1 is known to reduce its transcriptional activity [14], this E2-mediated reduction in C/EBP β1 SUMOylation level may account for the known activation of C/EBP β1 by E2. Interestingly, a similar effect has also been shown for the papillomavirus E7 protein, which reduces SUMOylation of its binding partner, pRb (retinoblastoma protein) [15]. These E2 and E7 results further illustrate that the interplay between viral proteins and the SUMOylation system is complex and extends beyond their simply being substrates for SUMO modification.

Conclusions

Viruses are well adapted to targeting host cell systems for their own benefit, and the SUMOylation system is no exception. It is clear from the work of numerous groups that many viral proteins are modified by SUMOylation. Our work with papillomaviruses has shown that two early proteins, E1 and E2, are both substrates for SUMOylation and that this modification has functional implications for these viral proteins. In addition to the direct effects of SUMO modification on E1 and E2 protein function, the evidence emerging with HPV E2 and E7 proteins is that the viral proteins can also affect SUMOylation of their host cell binding protein partners. Defining additional viral effects on host SUMOylation and their functional consequences will be exciting areas for future study.

Regulation of Protein Function by SUMO Modification: A Biochemical Society Focussed Meeting held at Manchester Conference Centre, Manchester, U.K., 25–27 June 2007. Organized and Edited by R. Hay (Dundee, U.K.) and A. Sharrocks (Manchester, U.K.).

Abbreviations

     
  • C/EBP

    CCAAT/enhancer-binding protein

  •  
  • HPV

    human papillomavirus

  •  
  • STAT

    signal transducer and activator of transcription

  •  
  • PIAS

    protein inhibitor of activated STAT

  •  
  • SUMO

    small ubiquitin-related modifier

  •  
  • Ubc9

    ubiquitin-conjugating enzyme 9

References

References
1
Wilson
V.G.
SUMOylation: Molecular Biology and Biochemistry
2004
Norfolk
Horizon Biosciences
2
Rosas-Acosta
G.
Russell
W.K.
Deyrieux
A.
Russell
D.H.
Wilson
V.G.
Mol. Cell. Proteomics
2005
, vol. 
4
 (pg. 
56
-
72
)
3
Vertegaal
A.C.O.
Andersen
J.S.
Ogg
S.C.
Hay
R.T.
Mann
M.
Lamond
A.I.
Mol. Cell. Proteomics
2006
, vol. 
5
 (pg. 
2298
-
2310
)
4
Heun
P.
Curr. Opin. Cell Biol.
2007
, vol. 
19
 (pg. 
350
-
355
)
5
Boggio
R.
Chiocca
S.
Curr. Opin. Microbiol.
2006
, vol. 
9
 (pg. 
430
-
436
)
6
Hebner
C.M.
Laimins
L.A.
Rev. Med. Virol.
2006
, vol. 
16
 (pg. 
83
-
97
)
7
Longworth
M.S.
Laimins
L.A.
Microbiol. Mol. Biol. Rev.
2004
, vol. 
68
 (pg. 
362
-
373
)
8
Wilson
V.G.
West
M.
Woytek
K.
Rangasamy
D.
Virus Genes
2002
, vol. 
24
 (pg. 
275
-
290
)
9
Rangasamy
D.
Woytek
K.
Khan
S.A.
Wilson
V.G.
J. Biol. Chem.
2000
, vol. 
275
 (pg. 
37999
-
38004
)
10
Rosas-Acosta
G.
Langereis
M.A.
Deyrieux
A.
Wilson
V.G.
Virology
2005
, vol. 
331
 (pg. 
190
-
203
)
11
Bian
X.-L.
Rosas-Acosta
G.
Wu
Y.-C.
Wilson
V.G.
J. Virol.
2007
, vol. 
81
 (pg. 
2899
-
2908
)
12
Hadaschik
D.
Hinterkeuser
K.
Oldak
M.
Pfister
H.J.
Smola-Hess
S.
J. Virol.
2003
, vol. 
77
 (pg. 
5253
-
5265
)
13
Deyrieux
A.F.
Rosas-Acosta
G.
Ozbun
M.A.
Wilson
V.G.
J. Cell Sci.
2007
, vol. 
120
 (pg. 
125
-
136
)
14
Subramanian
L.
Benson
M.D.
Iniguez-Lluhi
J.A.
J. Biol. Chem.
2003
, vol. 
278
 (pg. 
9134
-
9141
)
15
Ledl
A.
Schmidt
D.
Muller
S.
Oncogene
2005
, vol. 
24
 (pg. 
3810
-
3818
)