Turnip crinkle virus (TCV) is a 4054 b plus-sense RNA virus that belongs to the genus Carmovirus in the Family Tombusviridae. The 3' terminal 200 b of TCV are predicted to fold into 5 hairpins labeled in the 3' to 5' direction as the promoter (Pr), hairpin 5 (H5), hairpin 4b (H4b), hairpin 4a (H4a), and hairpin 4 (H4), using 3' UTR phylogenetic comparisons with other carmoviruses and the RNA structural prediction program, mfold. H5 was found to be a highly-conserved structure containing a large symmetrical loop (LSL) that formed a tertiary interaction between the 3' side of the LSL and the 3' terminal nucleotides using compensatory mutational analysis in vivo. In plants, LSL mutations resulted in a mutation frequency that was increased by as much as 12-fold without inducing error catastrophe. The original mutations frequently reverted and led to second site alterations biased for uridylate to cytidylate and adenylate to guanylate changes. These results suggest that H5 may function as a chaperone to properly fold the RdRp.

 The TCV 5' UTR, which binds 40S ribosomal subunits, contains two short segments exhibiting IRES activity that function synergistically with the 3' terminal region to enhance cap-independent translation <em>in vivo</em>.  In the TCV 3' UTR, H4a, H4b, H5, and flanking sequences, form an internal tRNA-like structure (iTLS) that binds 60S ribosomal subunits and the P-site of salt washed 80S ribosomes.  The iTLS may therefore mediate assembly of 80S ribosomes, which are then transported to the 5' end for translation of virally-encoded proteins.    

 Phylogenetic comparisons of carmovirus 3' UTRs revealed that <em>Cardamine chlorotic fleck virus</em> (CCFV) and <em>Japanese iris necrotic ring virus</em> (JINRV) are capable of forming the 5 elemental features comprising the iTLS.  Ribosome binding and plant cell culture assays showed that only the CCFV iTLS bound 80S ribosomes and could functionally replace the TCV iTLS.  These results suggest that closely-related members of the same viral genus may utilize different strategies for cap-independent translation.