Sumoylation at the Host-Pathogen Interface

Wilson, Van G.

An intriguing but more limited viral effect on Ubc9 has recently been reported for the adenovirus E1A oncoprotein [94] . Interaction of E1A to Ubc9 was first reported in 1996 [25] , and Yousef et al. extended those studies by exploring the interaction in detail [94] . The authors found that the Ubc9 interaction motif on E1A is the sequence EVIDLT in the conserved region 2 (CR2) domain. While this sequence resembles a SIM motif, sumoylation of Ubc9 was not required for the interaction with E1A. Furthermore, unlike the previous examples discussed, E1A interaction with Ubc9 did not appear to affect overall sumoylation in vivo nor did it affect in vitro sumoylation of two test substrates, HDAC4 or E2-25K proteins. Analysis of Ubc9 mutants indicated that E1A binds to Ubc9 in the N-terminal region involving sequences that overlap but are not identical to those involved with SUMO binding by Ubc9. Since SUMO binding by Ubc9 is critical for formation of SUMO chains (poly-sumoylation), E1A was tested for an effect on poly-sumoylation in a yeast assay. Ulp2 deleted yeast cannot grow at 37 °C due to toxic accumulation of poly-sumoylated proteins, but wild type E1A is able to restore efficient growth while a Ubc9 non-binding E1A mutant cannot. The authors concluded from this result that while E1A does not appear to grossly alter the mono-sumoylation pattern that it can inhibit poly-sumoylation which may affect the function of specific host proteins. They also speculated that E1A binding might affect the ability of Ubc9 to be sumoylated at lysine 14. Sumoylation at this residue alters Ubc9 substrate specificity [95] , so this could also have effects on sumoylation of host proteins. How these subtle E1A effects on Ubc9 contribute to viral fitness will require further study.