Ubiquitination of MHC class I heavy chains is essential for dislocation by human cytomegalovirus-encoded US2 but not US11.

The human cytomegalovirus-encoded glycoproteins US2 and US11 target newly synthesized major histocompatibility complex class I heavy chains for degradation by mediating their dislocation from the endoplasmic reticulum back into the cytosol, where they are degraded by proteasomes. A functional ubiquitin system is required for US2- and US11-dependent dislocation of the class I heavy chains. It has been assumed that the class I heavy chain itself is ubiquitinated during the dislocation reaction. To test this hypothesis, all lysines within the class I heavy chain were substituted. The lysine-less class I molecules could no longer be dislocated by US2 despite the fact that the interaction between the two proteins was maintained. Interestingly, US11 was still capable of dislocating the lysine-less heavy chains into the cytosol. Ubiquitination does not necessarily require lysine residues but can also occur at the N terminus of a protein. To investigate the potential role of N-terminal ubiquitination in heavy chain dislocation, a lysine-less ubiquitin moiety was fused to the N terminus of the class I molecule. This lysine-less fusion protein was still dislocated in the presence of US11. Ubiquitination could not be detected in vitro, either for the lysine-less heavy chains or for the lysine-less ubiquitin-heavy chain fusion protein. Our data show that although dislocation of the lysineless class I heavy chains requires a functional ubiquitin system, the heavy chain itself does not serve as the ubiquitin acceptor. This finding sheds new light on the role of the ubiquitin system in the dislocation process.

Human cytomegalovirus-encoded US2 and US11 target unassembled MHC class I heavy chains for degradation.

Surface MHC class I molecules serve important immune functions as ligands for both T and NK cell receptors for the elimination of infected and malignant cells. In order to reach the cell surface, MHC class I molecules have to fold properly and form trimers consisting of a heavy chain (HC), a beta2-microglobulin light chain and an 8-10-mer peptide. A panel of ER chaperones facilitates the folding and assembly process. Incorrectly assembled or folded MHC class I HCs are detected by the ER quality-control system and transported to the cytosol for degradation by proteasomes. In human cytomegalovirus-infected cells, two viral proteins are synthesized, US2 and US11, which target MHC class I HCs for proteasomal degradation. It is unknown at which stage of MHC class I folding and complex formation US2 and US11 come into play. In addition, it is unclear if the disposal takes place via the same pathway through which proteins are removed that fail to pass ER quality control. In this study, we show with a beta2m-deficient cell line that US2 and US11 both target unassembled HCs for degradation. This suggests that US2 and US11 both act at an early stage of MHC class I complex formation. In addition, our data indicate that US11-mediated degradation involves mechanisms that are similar to those normally used to remove terminally misfolded HCs.

Subtle sequence variation among MHC class I locus products greatly influences sensitivity to HCMV US2- and US11-mediated degradation.

Human cytomegalovirus (HCMV) interferes with cellular immune responses by modulating surface expression of MHC class I molecules. Here, we focused on HCMV-encoded unique short (US) 2 and US11, which bind newly synthesized MHC class I heavy chains (HCs) and support their dislocation into the cytosol for subsequent degradation by proteasomes. Not all MHC class I locus products are equally sensitive to this down-modulation. The aim of this study was to identify which domains, and ultimately which residues, are responsible for the resistance or sensitivity of MHC class I molecules to US2- and US11-mediated down-regulation. We show that, besides endoplasmic reticulum-lumenal regions, the C-terminus of class I molecules represents an important determinant for allele specificity in US11-mediated degradation. HLA-E becomes sensitive to US11-mediated down-regulation when its cytoplasmic tail is extended. Interestingly, this only requires two additional residues, lysine and valine, at its C-terminus. For US2, the MHC class I allele specificity is largely determined by a small region at the junction of the alpha2/alpha3 domain of the HC. It is quite remarkable that minor changes, in only four residues, can completely revert the sensitivity of naturally US2-resistant HLA-E molecules. With this study we provide better insights into the features underlying the selectivity in MHC class I down-regulation by US2 and US11.

Human cytomegalovirus-encoded US2 differentially affects surface expression of MHC class I locus products and targets membrane-bound, but not soluble HLA-G1 for degradation.

Human CMV (HCMV) can elude CTL as well as NK cells by modulating surface expression of MHC class I molecules. This strategy would be most efficient if the virus would selectively down-regulate viral Ag-presenting alleles, while at the same time preserving other alleles to act as inhibitors of NK cell activation. We focused on the HCMV unique short (US) region encoded protein US2, which binds to newly synthesized MHC class I H chains and supports their dislocation to the cytosol for subsequent degradation by proteasomes. We studied the effect of US2 on surface expression of individual class I locus products using flow cytometry. Our results were combined with crystal structure data of complexed US2/HLA-A2/beta(2)-microglobulin and alignments of 948 HLA class I database sequences of the endoplasmic reticulum lumenal region inplicated in US2 binding. This study suggests that surface expression of all HLA-A and -G and most HLA-B alleles will be affected by US2. Several HLA-B alleles and all HLA-C and -E alleles are likely to be insensitive to US2-mediated degradation. We also found that the MHC class I endoplasmic reticulum-lumenal domain alone is not sufficient for degradation by US2, as illustrated by the stability of soluble HLA-G1 in the presence of US2. Furthermore, we showed that the membrane-bound HLA-G1 isoform, but also tailless HLA-A2, are targeted for degradation. This indicates that the cytoplasmic tail of the MHC class I H chain is not required for its dislocation to the cytosol by US2.

Amino acid composition of alpha1/alpha2 domains and cytoplasmic tail of MHC class I molecules determine their susceptibility to human cytomegalovirus US11-mediated down-regulation.

During co-evolution with its host, human cytomegalovirus has acquired multiple defense mechanisms to escape from immune recognition. In this study, we focused on US11, which binds to MHC class I heavy chains and mediates their dislocation to the cytosol and subsequent degradation by proteasomes. To examine which domains of class I heavy chains are involved in this process, we constructed chimeric HLA molecules of US11-sensitive and -insensitive class I molecules (HLA-A2 and HLA-G, respectively). Pulse-chase experiments were performed to evaluate protein stability and interactions between class I heavy chains and US11. Flow cytometry was employed to assess the effect of US11 on surface expression of the different chimeras. Our results indicate that the alpha1 and alpha2 domains of HLA molecules are important for the affinity of US11 association. However, the degradation efficiency seems to rely mostly on cytosolic tail residues. We found that the nonclassical HLA-G molecule is insensitive to US11-mediated degradation solely because it lacks essential tail residues. A deletion of the last two tail residues in full-length MHC class I molecules already caused a severe reduction in degradation efficiency. Altogether, our data provide new insights into the mechanism by which US11 down-regulates MHC class I molecules.