Single transmembrane peptide DinQ modulates membrane-dependent activities

Abstract

The functions of several SOS regulated genes in Escherichia coli are still unknown, including dinQ. In this work we characterize dinQ and two small RNAs, agrA and agrB, with antisense complementarity to dinQ. Northern analysis revealed five dinQ transcripts, but only one transcript (+44) is actively translated. The +44 dinQ transcript translates into a toxic single transmembrane peptide localized in the inner membrane. AgrB regulates dinQ RNA by RNA interference to counteract DinQ toxicity. Thus the dinQ-agr locus shows the classical features of a type I TA system and has many similarities to the tisB-istR locus. DinQ overexpression depolarizes the cell membrane and decreases the intracellular ATP concentration, demonstrating that DinQ can modulate membrane-dependent processes. Augmented DinQ strongly inhibits marker transfer by Hfr conjugation, indicating a role in recombination. Furthermore, DinQ affects transformation of nucleoid morphology in response to UV damage. We hypothesize that DinQ is a transmembrane peptide that modulates membrane-dependent activities such as nucleoid compaction and recombination.

Figure 1. Overview of the dinQ/agrAB locus.

(A) Genomic organization. The locus contains three genes originating from separate promoters; two constitutively expressed transcripts, agrA and agrB (red box), and one divergently transcribed gene, dinQ (blue box), which is regulated by LexA repressor binding sites (green box). Transcription initiation sites are indicated by color-coded arrows. Antisense sequences are annotated in orange. Flanking genes are indicated in light yellow. (B) Sequence of the full length arsR-gor intergenic region with hallmarks. Promoter elements, −10 and −35 and terminator sequences are underlined. LexA operator sequences in bold. Transcription start () and stop () for dinQ, agrA and agrB. AgrAB repeat sequences are shadowed. (C) Alignment of agrA/agrB sequences antisense to dinQ. The agrAB repeat of dinQ is antisense to sequences in the agrA and agrB transcripts.

Figure 2. Expression pattern and transcription start of the dinQ/agrAB locus.

(A) Northern analysis of dinQ and agrAB transcripts. A riboprobe specific to either dinQ (upper panel) or agrAB (lower panel) was hybridized to a northern blot with total RNA extracted from strains lacking dinQ (BK4040), agrA (BK4042) or agrB (BK4043) and compared to wt (AB1157) before and after UV exposure. Five dinQ transcripts assigned dinQ-a, -b, -c, -d and -e were detected. (B) A [³²P]-labeled primer specific to dinQ was used in a primer extension assay to determine transcription initiation sites of the various dinQ transcripts. The primer extension assays were performed on total RNA extracted from wt (AB1157) and mutant strains agrA (BK4042) and agrB (BK4043). The base annotated for transcription initiation is indicated with an asterisk (red) for each transcript. Three primer extension products were detected that correlated in size to dinQ-a, -b and -d from 3′ mapping and the northern blot in (A). (C) A [³²P]-labeled primer specific to agrAB was used in a primer extension assay to determine transcription initiation sites of agrA and agrB. Total RNA from mutant strain agrB (BK4043) was used to localize agrA transcription start and total RNA from mutant strain agrA (BK4042) was used to localize agrB transcription initiation, due to the transcripts sequence complementarity. The base annotated for transcription initiation is indicated with an asterisk (red) for each transcript.

Figure 3. DinQ, agrB, and agrA phenotypes.

(A) UV survival of dinQ (BK4040, diamond), agrA (BK4042, cross) and agrB (BK4043, star) single mutants, agrAB double mutant (BK4041, triangle) and the agrAB dinQ triple mutant (BK4044, circle) compared to wt (AB1157, square). The data are presented as mean of three independent experiments with standard deviation. (B) UV survival of wt (AB1157/pKK232-8, square), wt overexpressing dinQ (AB1157/pBK444, diamond) and wt overexpressing dinQ-agrAB (AB1157/pBK440, circle). The data are presented as mean of three independent experiments with standard deviation. (C) Sequence of the full-length dinQ transcript with annotated ORFs. The dinQ transcript contains two ORFs encoding putative peptides of 18 and 49 aa, in which the 49 aa ORF contain four putative start codons (underlined and assigned). DinQ II of 49 aa, DinQ III of 42 aa, DinQ IV of 38 aa and DinQ V of 27 aa which is translated from a GTG. Transcription terminator sequences are underlined. RNA start sites found in the primer extension experiments are assigned with a, b and d. Base mutations are assigned in red. (D) Serially diluted (10⁻¹–10⁻⁵cells ml⁻¹) log phase cultures of wt (ER2566) with expression vector pET28b(+) or vector constructs with putative dinQ encoded peptides (DinQ I–V) were spotted onto LB plates containing no IPTG (left panel) or 0.6 mM IPTG (right panel). Pictures were taken 1 day after incubation at 37°C. (E) Survival of wt (ER2566) overexpressing putative dinQ encoded peptides related to UV exposure. DinQ II (diamond), DinQ V (triangle), dinQ (asterisk) compared to vector controls pET28b(+) (square) and pKK232-8 (cross). Left panel shows survival of IPTG induced DinQ peptides without UV exposure and right panel shows survival of IPTG induced DinQ peptides when exposed to UV (40 J/m²).

Figure 4. DinQ translation and localization.

(A) Translation of putative DinQ peptides I–V expressed from pET28b(+) plasmid constructs (lanes 1–6) and PCR products corresponding to dinQ-a, -b and -d mRNAs (lane 7–10) were analyzed with coupled in vitro transcription/translation kits from Promega. The E. coli T7 S30 extract system for circular DNA was used to analyze DinQ expression from plasmid constructs while S30 extract system for linear templates was used to analyze PCR products (Promega). Labelling was carried out with [¹⁴C]-Leucine. (B) Western analysis of FLAG-tagged endogenous DinQ expression. Protein extracts from UV exposed (20 J/m²) control cells (BK4044) mixed with plasmid pCR2.1-DinQ-3×-FLAG in lane#1 (pos ctrl), wt (BK5300) in lane#2 (neg ctrl), UV exposed (50 J/m²) ΔagrB DinQ-3×FLAG (BK5372) in lane#3, UV exposed (50 J/m²) wt DinQ-3×FLAG (BK5370) in lane#4, unexposed ΔagrB DinQ-3×FLAG (BK5372) in lane#5, unexposed wt DinQ-3×FLAG (BK5370) in lane#6 was resolved by SDS-PAGE and analyzed by western blotting using Monoclonal ANTI-FLAG M2-Alkaline Phosphatase antibody (SIGMA). Gel migration was monitored relative to SeeBlue Plus2 prestained standard (Invitrogen) in kDa. Detection was carried out by NBT/BCIP color development substrates. The intensities of the DinQ bands was analyzed in three independent western blots with the program ImageJ (Rasband,W.S. and ImageJ, U. S. National Institutes of Health, Bethesda, MD, USA) and normalized against a cross reacting higher molecular weight protein band (not shown in Figure 4B). A cross reacting low molecular protein band of unknown origin is present in all lanes. (C) Serially diluted (10⁻¹–10⁻⁵cells ml⁻¹) log phase cultures of wt (BK5300), ΔagrB (BK5342), wt dinQ-K4stop (BK5350), ΔagrB dinQ-K4stop (BK5352), wt dinQ-A108T-C112G-A115T (BK5360) and ΔagrB dinQ-A108T-C112G-A115T (BK5362) were spotted onto LB plates and exposed to UV (0 J/m² and 20 J/m²). Pictures were taken one day after incubation at 37°C. (D) Subcellular localization of DinQ in wt (ER2566) cells carrying plasmid pET28b(+)-3×FLAG-DinQ V. Subcellular fractions was resolved by SDS-PAGE and analyzed by western blotting using antibodies against TolC, Lep and FLAG. Lep and TolC detected inner- and outer membrane proteins, respectively. T: total protein; C: cytoplasmic fraction; IM: inner membrane fraction; OM: outer membrane fraction. (E) DinQ amino acid sequence with predicted secondary structure elements (H = helix, ‘-’ = other) and corresponding reliability index (range 0–9). (F) 3D modelling of DinQ as a regular α-helix embedded in a lipid membrane. Polar patch formed by residues Glu17, Arg20 and Gln24 encircled by dashed ellipse.

Figure 5. Effects of DinQ on membrane potential and intracellular ATP.

(A) Exponentially growing cells were exposed to IPTG for 0, 5 and 20 min followed by measurement of DiBAC₄(3) fluorescence. An increase in fluorescence intensity is observed as a consequence of membrane depolarization. Upper panel displays wt (ER2566) cells carrying expression vector pET28b(+). Lower panel displays wt (ER2566) cells overexpressing DinQ V (pET28b(+)-DinQ V). The graphs are representative of at least three repetitions. (B) The staple diagram shows in vivo levels of ATP before and 20 min after UV exposure of wt (AB1157) and mutant agrB (BK4043) cells. ATP concentrations were determined by a luciferase-based assay (Promega) and the activity is shown as RLU/OD (relative light units related to optical density). The data are presented as mean of three independent experiments with standard deviation.

Figure 6. Genetic interactions between agrB and uvrA/recB and recombination frequency of agrB.

(A) Dilutions of exponentionally growing wild type (AB1157, open square), agrB (BK4148, open diamond), uvrA (BK4180, open triangle), recB (BK4110, black diamond), agrB recB (BK4112, black square), uvrA recB (BK4183, black circle) and agrB uvrA (BK4182, open circle) cells were plated on LB plates and exposed to UV irradiation. Colonies were counted one day after incubation at 37°C. The data are presented as mean of three independent experiments with standard deviation. (B) Recombination frequency of uvrA (BK4180), dinQ (BK4040) and agrB (BK4043/BK4148) determined by Hfr conjugation assays. The data are presented as mean of three independent experiments with standard deviation.

Figure 7. Nucleoid compaction during DNA repair and model of regulation and mechanism of action for DinQ.

(A) Exponentially growing wt (AB1157), dinQ (BK4040) and agrB (BK4043) cells in glucose-CAA medium were exposed to UV (3 J/m²). Samples were harvested at the indicated time points and cells were fixed, stained with Hoechst 33258 and mounted on microscope slides. The panel of pictures shows a mosaic of representative cells. (B) The fraction of cells with compact nucleoids (i.e. cells containing one or two non-extended nucleoids instead of two or four normal nucleoids, respectively) were scored for each time point. The average of three experiments and a total of 250–500 cells were counted for each time point. Error bars indicate the standard deviation. Note that normal nucleoids as well as extended, diffuse nucleoids were scored as “non-compact”. (C) This model summarizes all the data presented in this work, including regulation of transcription and translation, and the proposed mechanism of action at the inner cell membrane. All RNA secondary structures are computer models generated by the mfold web server software.