Viral infection is the major inducer of transcription of genes encoding various types of interferons. Positive Regulatory Domains (PRDs) have been identified in the IFNb promoter, and further characterization established PRD II as an NFkB binding element, whereas PRD IV is able to interact with ATF-2. The PRD I and PRD III regions serve as high affinity sites for the Interferon Regulatory Factors (IRF) 1 and 2. Interestingly, whereas the NFkB site in the IFNb gene is essential for its induction, no such element can be found in the IFNa promoter regions. However, virus responsive elements that can serve as IRF binding sites have been found in the IFNa promoter region. Initial studies suggested that IRF1 acts as an activator and IRF2 as a suppressor of these genes, however, IRF1-/- and IRF2-/- mice are still able to induce the IFNa and IFNb genes upon viral infection. Interestingly, these studies were able to establish IRF1 as a tumor suppressor whose function is opposed by the actions of IRF2.

Slfns proteins are encoded by a novel family of interferon-induced genes comprised of 10 members, are unique to mammalian organisms, and are not found in C. elegans or drosophila. All pox-viruses carry slfn related sequences in their genome. The slfn proteins bear little to no resemblance to any known proteins, and appear to lack any characterized functional domains. The only structural element we have been able to identify remotely resembles the AAA domain found in some RNA/DNA helicases.

Since this was the only lead we had to the possible function of schlafen proteins, we decided to investigate the nucleic acid binding properties of schlafens. Schlafen 2 (slfn2) is the most strongly and consistently IFNalpha/beta-upregulated family member, and we have therefore focused our attention on slfn2. Over the past year, I have worked on the characterization of the intracellular localization of slfn2, and the interaction of Slfn proteins with single-stranded or double-stranded RNA or DNA. My most recent studies suggested an intrinsic activity of these proteins to indeed bind and modify/process nucleic acids.

In addition, we also tested whether slfn2 would be able to interact in a similar fashion with RNA, and found that slfn2 is indeed able to interact with single-stranded RNA in an EMSA assay in a similar manner as with DNA. The fact that a cytoplasmic protein can interact with RNA and DNA was rather unexpected, and raises even more questions as to the function of the protein. One hypothesis that we are currently testing is based on several recent reports that identified the RIG1 RNA helicase as an intracellular sensor for nucleic acids derived from invading pathogens. Since eukaryotic nucleic acids carry unique posttranslational modifications (e.g. pseudouridine, methyl-cytidine, methyl-adenosine) that are absent in viruses or bacteria, we tested whether slfn2 would still be able to recognize RNA in which these modified nucleotides had been incorporated.

Insertion of modified nucleotides resulting in RNA that resembles more closely eukaryotic RNA causes a significant loss in Slfn RNA binding. This finding supports the notion that slfn proteins may act as sensors for foreign nucleic acids and play a role in the defense against pathogens, likely with oncogenic potential.

We seek to determine how pathological events such as malignant transformation and viral infections interfere with signal transduction pathways, thereby circumventing the antiviral and antiproliferative properties of the IFNs, and how disturbances in these signaling systems contribute to the development of pathological processes.

PUBLICATIONS (resulting from this training)

Vroom K, Severa M (2007). Schlafen: a proapoptotic gene family with an essential function in innate antiviral responses. In preparation.