Identification and Analysis of Cleavage Sites and Subdomains in the Group Specific Antigen (Gag) of the Human Endogenous Retrovirus-K(HML-2)

Interestingly at molecular level humans and chimpanzees are very similar. The average DNA sequence difference between them is ~4% (Varki and Altheide, 2009). Contributing to the difference in humans are human endogenous retroviruses (HERV), which constitute about 8% of our genome. The young age and the similarity to other cancer causing betaretroviruses such as MMTV have led to intense research on the influence of some HERVs on the human genome as well as oncogenic aspects of HERV encoded proteins. Fundamental properties of these ancient retroviruses although have remained undiscovered, due to the low expression of the retroviral proteins. One example is the Gag polyprotein, the major structural protein of all retroviruses with the ability to form virus like particles in the absence of any other retroviral protein. The aim of this project was to decipher the processing sites of the Group specific antigen (Gag) protein of a specific member of the HERV family, HERV-K113, and to identify all cleavage products. Although the provirus of HERV-K113 encodes open reading frames for genes of a complex retrovirus and produces mature virions on low expression levels, it has accumulated mutations which render it uninfectious. To elevate the expression non-synonymous post-integration mutations in the gag-pro-pol region of HERV-K113 had previously been identified by a specific algorithm and reconstituted. The amino acid sequence had then been codon optimized for mammalian cell. Gag was cloned together with the protease and polymerase sequence into the vector pcDNA3.1 to set it under the control of a CMV-promoter. Expression of this new construct, oricoHERV-K113_GagProPol, in mammalian cells led to an enhanced viral protein expression as seen on Western Blot and in Electron microscopic pictures of purified viral pellets. The protease defective construct oricoHERV-K113_GagPro-Pol was established to investigate the maturation process. Comparison of protease defective viral proteins to protease active viral proteins on silver nitrate stained SDS gel analysis revealed the processing of Gag proteins and the maturation of VLPs after the release from cells. Alignments of HERV-K113 Gag with the sequences of two beta retroviruses, MPMV and MMTV, had already shown that the endogenous retroviral Gag harbours the typical domains matrix (MA), capsid (CA) and nucleocapsid (NC) as well as a phosphorylated protein with a late domain between MA and CA. The high expression rate of oricoHERV-K113 enabled the investigation of the exact Gag cleavage sites. Therefore viral proteins were separated by reverse phase high pressure liquid chromatography. Fractions were screened for Gag subdomains on western blot with previously established antibodies and, in some cases, by mass spectrometry. Proteins suspected to contain a subdomain were N-terminally sequenced, to determine the N-terminal cleavage site amino acid sequence. Additional to the common retroviral Gag subdomains MA, CA and NC the subdomain p15 was confirmed between MA and CA. Typical for a late domain encoding protein p15 was shown to be phosphorylated by mass spectrometry. All cleavage sites except the one between MA and p15 belong to type 1 or type 2 cleavage sites. Mutation of the P1 position at the cleavage sites led to changes in cleavage patterns detected on Western Blot. Surprisingly two new short proteins were discovered following NC. Both are rich in glutamine and proline and hence were named QP1 and QP2. In summary the reconstituted protein sequence of oricoHERV-K113_GagProPol facilitated the discovery that the HERV-K113 Gag is cleaved by the protease to release the subdomains MA, p15, CA, NC, QP1 and QP2. The functions of QP1 and QP2 have to be investigated in the future. Since processing is a prerequisite for replication, analysis of the subdomains and their processing could lead to the discovery why no original or improved HERV sequence until now has been shown to replicate