“…The biological activity of the orthologs of K-Rev observed in exogenous retroviruses, such as H-Rev and HTLV-I Rex, requires the recruitment of these viral regulatory proteins to a discrete primary RNA-binding site that is presented in the context of a larger region of tightly folded RNA termed a "response element" (Hanly et al+, 1989;Malim et al+, 1989Malim et al+, , 1990Zapp & Green, 1989;Bartel et al+, 1991;Heaphy et al+, 1991;Bogerd et al+, 1992)+ Although the primary binding site is clearly essential for biological activity, it is not sufficient and the H-Rev-or Rex-induced export of a target viral RNA therefore requires an almost full-length RNA response element located in cis (Malim et al+, 1989;Huang et al+, 1991)+ Because the biological activity of Rev or Rex response elements is dependent on their folding, these RNA elements can be effectively identified, and their structure predicted, using computer analysis+ Indeed, structures for several retroviral response elements, including both the ;234-nt HIV-1 RRE and the ;254-nt HTLV-I, were first predicted by computer analysis and only subsequently validated by mutational analysis (Hanly et al+, 1989;Malim et al+, 1989;Le et al+, 1990)+ The K-RRE coincides with a predicted unusual RNA folding region Previously, we and others have used functional assays to roughly map the extent of the HERV-K Rev response element (K-RRE) to between residues 8719 and 9152 according to the prototypic HERV-K10 proviral sequence reported by Ono et al+ (1986) (Magin et al+, 1999;Yang et al+, 1999)+ This 434-nt RNA sequence, which is entirely located in the U3 region of the HERV-K LTR, is considerably larger than either the H-RRE or the HTLV-I RxRE, and it therefore appeared likely that the actual extent of the K-RRE would be smaller than this initial estimate+ To gain insight into the true size of the K-RRE, and to determine whether the K-RRE indeed coincides with an unusual folding region (UFR), we used the same computational approach that previously allowed the definition of not only the H-RRE and the HTLV-I RxRE but also the HIV-2 RRE and the Visna-Maedi Virus RRE (V-RRE) (Hanly et al+, 1989;Malim et al+, 1989;Le et al+, 1990;)+ Initially, we focused our attention on sequences extending from 8504 to 9250 within the HERV-K108 proviral clone (Barbulescu et al+, 1999), a sequence that fully includes, but extends significantly beyond, the functionally defined 8719-9152 K-RRE sequence+ Monte Carlo simulations were used to search for any UFR within this HERV-K sequence by calculation of the statistical significance and thermod...…”