Mutation analysis of the 5' untranslated region of the cold shock cspA mRNA of Escherichia coli

Abstract

The mRNA for CspA, a major cold shock protein in Escherichia coli, contains an unusually long (159 bases) 5' untranslated region (5'-UTR), and its stability has been shown to play a major role in cold shock induction of CspA. The 5'-UTR of the cspA mRNA has a negative effect on its expression at 37 degrees C but has a positive effect upon cold shock. In this report, a series of cspA-lacZ fusions having a 26- to 32-base deletion in the 5'-UTR were constructed to examine the roles of specific regions within the 5'-UTR in cspA expression. It was found that none of the deletion mutations had significant effects on the stability of mRNA at both 37 and 15 degrees C. However, two mutations (Delta56-86 and Delta86-117) caused a substantial increase of beta-galactosidase activity at 37 degrees C, indicating that the deleted regions contain a negative cis element(s) for translation. A mutation (Delta2-27) deleting the highly conserved cold box sequence had little effect on cold shock induction of beta-galactosidase. Interestingly, three mutations (Delta28-55, Delta86-117, and Delta118-143) caused poor cold shock induction of beta-galactosidase. In particular, the Delta118-143 mutation reduced the translation efficiency of the cspA mRNA to less than 10% of that of the wild-type construct. The deleted region contains a 13-base sequence named upstream box (bases 123 to 135), which is highly conserved in cspA, cspB, cspG, and cspI, and is located 11 bases upstream of the Shine-Dalgarno (SD) sequence. The upstream box might be another cis element involved in translation efficiency of the cspA mRNA in addition to the SD sequence and the downstream box sequence. The relationship between the mRNA secondary structure and translation efficiency is discussed.

FIG. 1

Cold shock induction of β-galactosidase. (A) Construction of cspA-lacZ fusions. The wild-type cspA is shown on the top. The cspA-lacZ fusion in each expression plasmid is shown from the 5′ end of the cspA promoter upstream region to lacZ. Nucleotide numbers are given starting from the transcription initiation site as +1, as determined by Tanabe et al. (31). The crosshatched, open, dotted, and diagonally striped bars represent the cspA promoter, its 5′-UTR, the cspA coding region, and the lacZ coding region, respectively. The solid boxes indicate the SD sequence. The positions of deleted regions are shown with nucleotide numbers. (B) Induction patterns of various deletion constructs. At mid-log phase, cultures of E. coli AR137 harboring various plasmids were shifted from 37 to 15°C. Samples were taken at 0, 1, 2, 3, 5, 7, and 10 h after the shift, and β-galactosidase activity was measured. The cspA-lacZ fusions were pMM67 (○), pMM022 (●), pMM023 (□), pMM024 (■), pMM025 (▵), and pMM026 (▴).

FIG. 2

Analysis of mRNA stability. (A) Primer extension analysis of the cells harboring the cspA-lacZ fusions. At mid-log phase, cultures of E. coli AR137 harboring various plasmids were shifted from 37 to 15°C. For measurement of the mRNA stability at 37°C, cultures were shifted back to 37°C after 30 min of incubation at 15°C and rifampin was added to the cultures to a final concentration of 200 μg/ml. RNAs were extracted at 0 (lanes 1), 1 (lanes 2), 3 (lanes 3), and 5 (lanes 4) min after the addition of rifampin. For measurement of the mRNA stability at 15°C, rifampin was added 1 h after the temperature downshift, and then RNAs were extracted at 0 (lanes 5), 5 (lanes 6), 10 (lanes 7), and 20 (lanes 8) min after the addition of rifampin. Primer extension was carried out as described previously (21). (B) Graphic presentation of the results shown in panel A for 37 and 15°C, respectively. The radioactivities of transcripts were measured with a phosphorimager and plotted by using the transcript at the zero time point as 100%. ●, pMM022; □, pMM023; ■, pMM024; ▵, pMM025; and ▴, pMM026.

FIG. 3

Analysis of mRNA level and translational efficiency. (A) Primer extension analysis of the cspA-lacZ fusions. At mid-log phase, cultures of E. coli AR137 harboring various plasmids were shifted from 37 to 15°C. RNAs were prepared from the culture at 37°C (0 h; lanes 1) and at 0.5 (lanes 2), 1 (lanes 3), 2 (lanes 4), and 3 (lanes 5) h after the temperature downshift. Primer extension was carried out as described previously (21). (B) Graphic presentation of the relative amounts of mRNA. Relative mRNA amounts (mean values of two experiments) were calculated from the radioactivities of transcripts shown in panel A with the transcript of pMM67 at 37°C as 1. The relative amount is shown on the top of each column. Columns 1, 0 h; columns 2, after 0.5 h; columns 3, after 1 h; columns 4, after 2 h; and columns 5, after 3 h. (C) Relative translational efficiencies of the cspA-lacZ mRNAs. Translational efficiencies at 15°C were calculated by dividing the increment of β-galactosidase activity during the first 2 h after cold shock by the amount of mRNA with the following formula: [(Gal 2 h) × (OD 2 h) − (Gal 0 h) × (OD 0 h)]/(ave mRNA), where (Gal 0 h) and (Gal 2 h) are β-galactosidase activities at 0 and 2 h after temperature downshift, respectively; (OD 0 h) and (OD 2 h) are the optical densities at 600 nm of the cultures at 0 and 2 h after temperature downshift, respectively; and (ave mRNA) is the average of relative mRNA amounts at 0.5, 1, and 2 h after temperature downshift. Relative translational efficiency of each mRNA was calculated by using the efficiency of mRNA of pMM67 as 100%.

FIG. 4

Sequence similarities of cspA, cspB, cspG, and cspI mRNAs around the SD sequence and potential base pairing between cspA mRNA and 16S rRNA. Nucleotide numbers of cspA (31), cspB (7), cspG (23), and cspI (33) mRNA are given starting from the major transcription initiation site as +1. The sequence of 16S rRNA is from the work of Brosius et al. (4). Nucleotides identical in the four csp mRNAs are shown in boldface. The 13-base homologous sequences in cspA, cspB, cspG, and cspI are boxed (the UB). Positions of the SD sequence and the initiation codon are underlined. Potential base pairings between cspA mRNA and 16S rRNA are indicated by vertical lines.

FIG. 5

Role of the 13-base UB sequence in the cspA 5′-UTR in cspA-lacZ expression. (A) Construction of cspA-lacZ fusions. The constructs are drawn in the same manner as shown in Fig. 1A except for a box with wavy lines, which represents the 13-base UB sequence (5′-GCCGAAAGGCACA-3′) located upstream of the SD sequence. pKNJ37 is identical to pMM007 except for the deletion of the 13-base sequence. In both pMM007 and pKNJ37, cspA was translationally fused to lacZ. pKNJ38 is identical to pKM67 (21) except for the addition of the 13-base sequence by replacing the DNA fragment between XbaI and HindIII with synthesized oligonucleotides as described in Materials and Methods. In both pKM67 and pKNJ38, cspA was transcriptionally fused to lacZ. (B) Cold shock induction of β-galactosidase. The cspA-lacZ fusions were pMM007 (○), pKNJ37 (●), pKNJ38 (▵), and pKM67 (▴).

FIG. 6

Comparison of the secondary structures of the 5′-UTRs for the deletion constructs. Secondary structures of the 5′-UTR for each deletion construct were predicted with a nucleotide sequence analysis program (DNASIS-Mac; Hitachi Software Engineering Co. Ltd.) based on the method of Zuker and Stieger (36). Nucleotides are numbered as the position in the cspA mRNA starting from the transcription initiation site as +1. The position of the deletion in each mutant is shown by an arrow with the nucleotide numbers of the deleted region. The highly conserved 13-base sequences upstream of the SD sequence designated the UBs are boxed. The initiation codon and the SD sequence are also boxed.