attI1-Located Small Open Reading Frames ORF-17 and ORF-11 in a Class 1 Integron Affect Expression of a Gene Cassette Possessing a Canonical Shine-Dalgarno Sequence

TEXT

A search of the Integrall integron database (see integrall.bio.ua.pt) (1) and GenBank (see www.ncbi.nlm.nih.gov/GenBank/) revealed that approximately 20% of gene cassettes of class 1 integrons lack a translation initiation region (TIR). Hanau-Berçot et al. showed previously that the presence of a small DNA segment of 36 nucleotides (nts) (open reading frame 11 [ORF-11]) upstream of the start codon of an aacA4 gene cassette can compensate for the absence of TIR (2); however, the underlying mechanism remains unclear. We observed that ORF-11 was also present upstream in about 40% of gene cassettes possessing a canonical TIR. ORF-11 overlaps the inner boundary of the 5′ conserved segment (CS) of class 1 integrons, with the stop codon of the ORF located in the 7-bp core site (G↓TTRRRY) (3). ORF-11 can be created during recombination, provided that the inserted cassette sequence introduces an in-frame stop codon. Interestingly, a second ORF (ORF-17), comprising 51 nts and overlapping with ORF-11, was observed within the attI1 site in virtually all class 1 integrons (Fig. 1; see also representative integrons in Table S1 in the supplemental material). Unlike ORF-11, formation of ORF-17 does not depend on the sequence of the gene cassette inserted downstream of attI1. These two ORFs possess canonical TIRs (Fig. 1).

• Open in new tab
• Download powerpoint

FIG 1

Schematic presentation of part of the cloned sequence (307 nts; plasmid pHP329) from intI1 to the 5′ end of bla_GES-1. The nucleotides are numbered as in GenBank accession number EU598463. Diamond-bounded line indicates the Pc (hybrid 1) promoter, and the −35 and −10 regions are underlined. The attI1 site is indicated by gray shading. Sequences related to the 7-bp core site are indicated by thin black arrows, and the recombination crossover point is marked by a vertical arrow. ORF-17 and ORF-11 are shown with arrows (with vertical and horizontal lines, respectively). The SD sequences are presented within rectangles. Initiation codons are in bold letters, and the nucleotides of the sequence provided by the inserted cassette are in lowercase italic letters.

In this study, we found indications that ORF-17 and ORF-11 contribute to the expression of TIR-containing gene cassettes.

A set of six plasmid variants of the previously described GES-1-encoding pHP329 was used (Table 1). pHP329 was a pACYC177 derivative containing a bla_GES-1 gene cassette preceded by a hybrid 1 Pc promoter; the cloned sequence had been obtained from In58-G1, a Pseudomonas aeruginosa class 1 integron with a variable region possessing bla_GES-1 as a single gene cassette (4). Shine-Dalgarno (SD) sequences and/or initiation codons of ORF-17, ORF-11, and bla_GES-1 were subjected to site-specific mutagenesis using 12 mutagenic primers (see Table S2 in the supplemental material) and a QuikChange site-directed mutagenesis kit (Stratagene). Also, pCTRL, a promoterless variant of pHP329 constructed by reverting the 29-bp Pc sequence and deleting five of the 17 bp between the −35 and −10 hexamers, was used for control purposes (4). Plasmid sequences were verified on the two strands using an ABI 377 sequencer (Applied Biosystems). Escherichia coli DH5α was utilized as a plasmid host.

MICs of various β-lactams were determined by Etest (bioMérieux). β-Lactamase-containing extracts were obtained by ultrasonic treatment of cells suspended in phosphate buffer (0.1 M [pH 7.0]) and clarified by ultracentrifugation. Production of GES-1 was verified by isoelectric focusing in polyacrylamide gels containing ampholytes (pH, 3.5 to 9.5) (APBiotech) and staining with nitrocefin (Oxoid Ltd) (5). β-Lactamase activity was quantified by spectrophotometry using nitrocefin as a reporter substrate (Δε_Μ⁴⁸² = 17,400 Μ⁻¹ cm⁻¹ [6]). Results are expressed as units of activity. One unit was the amount of enzyme hydrolyzing 1 nmol of substrate min⁻¹ mg⁻¹ of protein at 30°C and pH 7.0.

Quantification of the bla_GES-1 transcripts was carried out by quantitative real-time reverse transcription-PCR (RT-PCR). RNA extraction and RT-PCR were performed as described previously using aphA1 as the reference gene (4). Results are expressed relative to the amount (percentage) of bla_GES-1 transcript from pHP329.

The results are summarized in Table 1. The highest β-lactamase activity was observed in extracts of the pHP329-harboring strain. Mutagenesis of TIR-17 (pHP393) caused a slight (15%) reduction in hydrolytic activity. A similar reduction (19%) in GES-1 production was observed when TIR-11 was mutagenized (pHP392). Notably, simultaneous modifications of TIR-17 and TIR-11 (pHP206) resulted in a “synergistic” effect, with the apparent expression efficiency reduced by 69%. On the other hand, the presence of intact TIR-17 and TIR-11 did not compensate for the loss of hydrolytic activity (89% reduction) seen after mutagenesis of the SD sequence of bla_GES-1 (pHP396). MICs of β-lactams tested against the strains were in line with production levels of GES-1 (Table 1). To study further the role of ORF-17, plasmid pHP392 was used to construct two additional variants, pHP207 and pHP210. Determination of β-lactamase activities mediated by these two plasmids clearly showed that the effect of ORF-17 on gene cassette expression was primarily due to the presence of a functional SD sequence (Table 1); therefore, in trans activity is unlikely. It was also observed that hydrolytic activities conferred by the constructs pHP207 (containing a stop codon instead of an ATG) and pHP210 (lacking a functional SD sequence) matched those of the strains carrying pHP392 and pHP206, respectively. Finally, quantification of the bla_GES-1 transcripts showed that the relative amounts of mRNAs varied significantly, and at the same time, their differences were in good agreement with those observed in their respective hydrolytic activities. These findings, along with the fact that transcription in all constructs was driven by the same promoter, indicates that both ORF-17 and ORF-11 were required to achieve the maximum expression efficiency of GES-1.

As was hypothesized by Depardieu et al., small ORFs located in the vicinity of cassettes possessing canonical SD sequences may increase expression of the genes (7). This hypothesis seems to have been validated here. Indeed, the newly described ORF-17, which is common among class 1 integrons (see Table S1 in the supplemental material, and ORF-11 significantly contributed to expression of the adjacent bla_GES-1 cassette. Regarding the responsible mechanism, the most plausible hypothesis that we can offer is that the ribosomes accumulated on the SD sequences located in the 5′ untranslated region may act as cis mRNA stabilizers, increasing the half-life of the transcripts.

This study had certain limitations: (i) The constructs used lacked part of the characteristic 3′ CS of class 1 integrons (8). Although unlikely, the possibility that this feature affected gene cassette expression cannot be excluded. (ii) The role of ORF-11 was not examined in detail, as was done for ORF-17. (iii) From the available data, it is not possible to predict the effects, if any, of the two ORFs on expression of genes in class 1 integrons with multicassette variable regions. (iv) The use of a system carrying a promoter of moderate strength (4, 9) might exaggerate the impact of ORF-17 and ORF-11. Despite these limitations, and our assumptions regarding the mechanisms involved, the experimental data presented here suggest that the role of the attI1 sequence in the regulation of expression of TIR-possessing gene cassettes merits further investigation.