On the problem of establishing the subcellular localization of Dictyostelium retrotransposon TRE5-A proteins by biochemical analysis of nuclear extracts

At first sight a protein that is enriched in extracts prepared from nuclei by means of biochemical methods can be considered to be a nuclear protein in vivo. Although this assumption will hold true for most of the analyzed proteins, it could also lead to false interpretations. We analyzed the subcellular distribution of endogenous and plasmid-borne proteins derived from the retrotransposon TRE5-A of Dictyostelium discoideum. In biochemical fractionation experiments the proteins encoded by TRE5-A open reading frame 1 (ORF1p) and the putative endonuclease encoded in ORF2 (ENp) were found in the detergent-insoluble material containing the nuclei. However, salt extraction of isolated nuclei did not considerably release the TRE5-A proteins. Instead, the TRE5-A proteins were strongly enriched in a fraction that contained the chromosomal DNA after removal of most cytoskeletal and histone proteins. These observations implied that ORF1p and ENp were both attached to chromatin in vivo, but this conclusion was disproved by the expression of genetic fusions of green fluorescent protein with either ORF1p or ENp. We show conclusive evidence that both fusion proteins were located as large aggregates of native protein in the cytoplasm of D. discoideum cells.