“…Initiation of protein synthesis within eukaryotic cells usually occurs by a cap-dependent mechanism in which the 59 terminal cap-structure (m 7 GpppN++) of the cytoplasmic mRNAs is recognized by the translation initiation complex eIF4F (reviewed by Merrick & Hershey, 1996)+ This complex is believed to migrate along the mRNA in association with the small ribosomal subunit until the initiation codon is reached, usually within 50-100 nt+ A second, distinct, cap-independent mechanism of translation initiation is employed by some cellular mRNAs (e+g+, BiP; Macejak & Sarnow, 1991) and the RNAs of picornaviruses (e+g+, poliovirus (PV), encephalomyocarditis virus (EMCV), and foot-and-mouth disease virus (FMDV); reviewed in Belsham & Sonenberg, 1996;Jackson & Kaminski, 1995)+ These viruses have positive-sense RNA genomes that act as mRNAs+ However their RNAs are uncapped and they are translated even when cap-dependent protein synthesis is inhibited following the cleavage of the eIF4G component of the eIF4F cap-binding complex+ A region of about 450 nt, near the 39 end of the long 59 noncoding region (600-1,300 nt in different picornaviruses), is required to achieve internal initiation of protein synthesis and this element is termed an internal ribosome entry site (IRES)+ Within the picornavirus family, there are two major classes of IRES element+ The cardioviruses (e+g+, EMCV) and aphthoviruses (FMDV) share one class of element, whereas the enteroviruses (e+g+, PV) and rhinoviruses share a second type of element+ These classes of IRES element have very different predicted secondary structures and also differ in their biology+ The cardio-/aphthovirus elements function efficiently in rabbit reticulocyte in vitro translation systems whereas the entero-/rhinovirus elements do not+ One common feature of the picornavirus IRES elements is the presence of a polypyrimidine tract near the 39 end of the element+ Kaminski et al+ (1994) also pointed out the presence of two conserved loop sequences within these IRES elements+ In each class, one loop fits the GNRA tetraloop consensus (where N is any nucleotide and R is a purine) whereas the second loop is C rich+ Within the cardio-/aphthoviruses, the GNRA tetraloop is located at the end of a stem-loop, within a hammerhead structure that constitutes part of the largest predicted domain (termed the I domain in the EMCV structure) of the secondary structure+ RNA tetraloops fitting the GNRA consensus are very highly represented within large RNAs with stable tertiary structures (Woese et al+, 1990)+ It is believed that such loops play an important role in RNA tertiary interactions (Pley et al+, 1994;Costa & Michel, 1995, 1997Cate et al+, 1996aCate et al+, , 1996b and also interactions with proteins (Glück et al+, 1992;Legault et al+, 1998)+ A single point mutation in the sequence of the EMCV IRES (nt 380-834 of the EMCV RNA) within this tetraloop (GCGA to GCGC, nt 547-550) severely reduces (.95%) the activity of the IRES element (Robe...…”