DNA AND CELL BIOLOGY Volume 19, Number 1, 2000 Mary Ann Liebert, Inc. Pp. 39–46 Downstream Sequence Adjacent to AUG Affects Translation of Chloramphenicol Acetyl Transferase in Eukaryotic Cells MARINA CIULLO, GIOVANNA DEL POZZO, MONICA AUTIERO, and JOHN GUARDIOLA ABSTRACT The CAT gene is widely used as a reporter in eukaryotic systems because of the efficient translation of its mRNA. We report here that a sequence occurring in the CAT mRNA at 115 nucleotides from CAT AUG is essential for translation. This sequence includes a stem-loop structure, which, however, exhibits a calculated stability significantly lower than that required for a hairpin to act as an enhancer of translation in vitro. Re- placement of this region with the corresponding sequence from mRNAs that are normally translated in eu- karyotic systems drastically reduced translation of CAT in COS cells, although the consensus sequence around the AUG, known to be required for high-level translation initiation, was conserved. These observations may be relevant for the exploitation of the CAT reporter system for analysis of the mechanisms of translation ini- tiation by means of fusion constructs. 39 INTRODUCTION T HE SCANNING MODEL for the initiation of translation in eu- karyotes postulates the binding of a 40S subunit factor com- plex to the 59 end CAP site of the mRNA followed by a linear migration up to the first AUG codon (Kozak, 1989a). The nu- cleotide sequence surrounding the AUG site is important to en- sure fidelity and efficiency of translation. This sequence corre- sponds to a consensus motif that is conserved in almost all vertebrate mRNAs. Positions 2 3 (G or A) and 1 4 (G) are par- ticularly critical in determining the strongest effect on transla- tion (Kozak, 1991, 1995, 1997). Although secondary structures located upstream of the AUG codon may show an inhibitory effect on translation (Ganoza and Louis, 1994; Hess and Dun- can, 1996; Joyce and Perry, 1990; Kozak, 1986, 1989b), sec- ondary structures introduced downstream of an AUG placed in a suboptimal context may increase the rate of this process (Kozak, 1990, 1991; Ganoza and Louis, 1994; Frolov and Schlesinger, 1996). This effect appears to be particularly im- portant for initiation from non-AUG codons (Joyce and Perry, 1990, Hann, 1994). In some experiments (Kozak, 1990), a cas- sette containing an alternative weak AUG initiation site and a sequence forming a secondary structure exhibiting a D G 52 19 Kcal/mol, placed at different distances from the AUG, was in- troduced upstream of the natural start site of the CAT reporter gene. In vitro translation in a rabbit reticulocyte system showed that the alternative weak AUG was preferred to the natural CAT AUG, when the secondary structure was placed at a distance of 14 nucleotides. It was also found that the stability of the sec- ondary structure, rather than its primary sequence, was impor- tant to elicit the observed behavior (Kozak, 1990). The inter- pretation given for this effect was that the presence of this structure slows the progression of the translational complex along the mRNA, thereby allowing more time for recognition of the AUG codon, with the facilitation being maximal when the ribosome stalls with its AUG-recognition center directly over the AUG. The role of stem-loop structures in vivo is less characterized. Hypothetical stem-loops can be predicted from sequences lying downstream of the noncanonical start codons of translation of viral and cellular mRNAs (Kozak, 1990). Whereas the CAT gene is widely used as a reporter for the analysis of promoter and enhancer activities in transfected eu- karyotic cells, it has seldom been employed as a tool for the study of the effect of mRNA sequences preceding the AUG on the control of protein synthesis (Strugnell and Browder, 1997; Mizokami and Chang, 1994). In our laboratory, we have intro- duced the use of CAT as a reporter to study the translation of eukaryotic mRNAs coding for proteins such as the MHC class II HLA-DRA subunit (Ciullo et al ., in preparation). In the pre- sent work, we found that the 59 terminal region of CAT cod- ing from nt 1 1 to 1 27 is essential for the in vivo synthesis of the protein. This observation may be instrumental for the com- International Institute of Genetics and Biophysics, CNR, Naples, Italy.