Conditional DnaB Protein Splicing Is Reversibly Inhibited by Zinc in Mycobacteria

Abstract

Inteins, as posttranslational regulatory elements, can tune protein function to environmental changes by conditional protein splicing (CPS). Translated as subdomains interrupting host proteins, inteins splice to scarlessly join flanking sequences (exteins). We used DnaB-intein1 (DnaBi1) from a replicative helicase of Mycobacterium smegmatis to build a kanamycin intein splicing reporter (KISR) that links splicing of DnaBi1 to kanamycin resistance. Using expression in heterologous Escherichia coli, we observed phenotypic classes of various levels of splicing-dependent resistance (SDR) and related these to the insertion position of DnaBi1 within the kanamycin resistance protein (KanR). The KanR-DnaBi1 construct demonstrating the most stringent SDR was used to probe for CPS of DnaB in the native host environment, M. smegmatis We show here that zinc, important during mycobacterial pathogenesis, inhibits DnaB splicing in M. smegmatis Using an in vitro reporter system, we demonstrated that zinc potently and reversibly inhibited DnaBi1 splicing, as well as splicing of a comparable intein from Mycobacterium leprae Finally, in a 1.95 Å crystal structure, we show that zinc inhibits splicing through binding to the very cysteine that initiates the splicing reaction. Together, our results provide compelling support for a model whereby mycobacterial DnaB protein splicing, and thus DNA replication, is responsive to environmental zinc.IMPORTANCE Inteins are present in a large fraction of prokaryotes and localize within conserved proteins, including the mycobacterial replicative helicase DnaB. In addition to their extensive protein engineering applications, inteins have emerged as environmentally responsive posttranslational regulators of the genes that encode them. While several studies have shown compelling evidence of conditional protein splicing (CPS), examination of splicing in the native host of the intein has proven to be challenging. Here, we demonstrated through a number of measures, including the use of a splicing-dependent sensor capable of monitoring intein activity in the native host, that zinc is a potent and reversible inhibitor of mycobacterial DnaB splicing. This work also expands our knowledge of site selection for intein insertion within nonnative proteins, demonstrating that splicing-dependent host protein activation correlates with proximity to the active site. Additionally, we surmise that splicing regulation by zinc has mycobacteriocidal and CPS application potential.

Keywords: DNA helicase; conditional protein splicing; intein; mycobacteria.

FIG 1

Kanamycin-DnaBi1 (KD) constructs sort into three phenotypic classes. (A) KD schematic demonstrating splicing-dependent resistance (SDR). N-exteins (blue) and C-exteins (green) were interrupted with DnaBi1 (red) from M. smegmatis at native serine residues (+1S [purple]) of KanR protein. The WT nucleophilic cysteine at position 118 (C118 [yellow]) attacks the amino terminal alanine residue (gray) to initiate splicing. Mutation of C118 to alanine renders a splicing-inactive mutant. (B) Three phenotypes emerge from KD constructs. M. smegmatis DnaBi1 was cloned into the KanR protein at one residue before each of 16 native serine residues of KanR protein that serve as +1S. A nonsplicing C118A variant of each KD construct served as a control for splicing. Growth of E. coli-containing KD constructs in the presence of increasing kanamycin concentrations resulted in splicing-dependent or splicing-independent resistance (SDR [KD154] or SIR [KD189], respectively) and in no-resistance (NR [KD200]) phenotypes. Representative titers corresponding to kanamycin concentrations (0, 12.5, 100, and 300 μg/μl) are shown here. E. coli cells expressing uninterrupted KanR and empty vector (KanS) served as positive and negative controls for kanamycin resistance, respectively.

FIG 2

DnaBi1 insertion site selection and characteristics. (A) Graded resistance to increasing levels of kanamycin is affected by the position of insertion of DnaBi1. Data represent quantitative spot titers of CFU (right) for each of 16 KD constructs (top) with splicing-active (yellow) and splicing-inactive (gray) mutants challenged with increasing kanamycin concentrations (left; 12.5 to 2,000 μg/ml). CFU data are graphed adjacent to the uninterrupted positive control (KanR) and the empty vector negative control (KanS). Values are based on results from three biological replicates. Raw CFU data for each construct (WT and C118A) at each kanamycin concentration can be found in Table S3. (B) Serine residues for DnaBi1 insertion. The positions of the +1S residues of the SDR (red), NR (dark blue), and SIR (gray-blue) phenotypes are highlighted in the structure of KanR protein (PDB ID: 4FEW). (C) Fourteen of 16 (87.5%) of KanR +1S are more than 30% surface exposed. The relative solvent-accessible surface area (rSASA) of each serine residue (+1S) used for DnaBi1 insertion as a function of its position within the primary amino acid sequence of the KanR protein is illustrated. The exposure threshold of 30% is indicated with a horizontal dashed line. (D) Relative distances from +1S to dimer interface (DI) of KanR. The distance of each +1S residue from the DI of KanR (measured in angstroms [Å]) is plotted against the distance of each +1S from the DI along the amino acid (AA) sequence of KanR. (E) Proximity of +1S to KanR active site (AS). The distance from each +1S residue to the KanR AS (measured in angstroms [Å]) is plotted against the distance of each +1S from the KanR AS along the amino acid (AA) sequence of KanR.

FIG 3

DnaBi1 splicing is inhibited by zinc in the native M. smegmatis. Two-fold dilutions of cells, starting at OD₆₀₀ of 3 to 5, were spotted onto media in the presence of kanamycin and/or zinc as indicated. Survival of M. smegmatis expressing KD154 was selectively reduced >100-fold only in the presence of kanamycin and zinc compared to M. smegmatis expressing uninterrupted KanR. The concentration of kanamycin was 300 μg/ml and that of zinc was 100 μM where present. Three biological replicates were performed with an average reduction in survival of 170-fold (± standard deviation of 74-fold), for KD154 in the presence of 300 μg/ml kanamycin and 100 μM zinc compared to KanR lacking the intein.

FIG 4

Zinc potently inhibits DnaBi1 from M. smegmatis and M. leprae in vitro. (A) MIG schematic. The in-gel fluorescent reporter construct with maltose-binding protein, MBP-intein-GFP (MIG), was used to monitor M. smegmatis and M. leprae DnaBi1 splicing. Fluorescent product sizes indicate precursor (P), ligated extein (LE), or off-pathway N- and C-terminal cleavage reactions. (B) Zinc reversibly inhibits M. smegmatis DnaBi1 splicing. The gel of MIG splicing shows an accumulation of precursor and a concomitant reduction of ligated exteins in the presence of zinc compared to untreated MIG (left). This inhibition was reversed in the presence of the zinc chelator EDTA (right). Results of quantitation of MIG reporter under conditions of increasing zinc levels in the presence or absence of EDTA are shown below the representative gel (stack plots) where the ratio of splice products is plotted. (C) Splicing of M. leprae intein is also inhibited by zinc. Similarly to the M. smegmatis results, zinc-treated MIG DnaBi1 from M. leprae showed an accumulation of precursor and a reduction in the levels of ligated exteins compared to the control, in the low micromolar range (top left). Data are representative of results from three biological replicates, and where error bars are present, values are expressed as averages ± standard deviations.

FIG 5

Zinc binds the catalytic center of DnaBi1. (A) Crystal structure of Mycobacterium smegmatis DnaBi1 (red) bound to zinc (gray) at 1.95 Å (PDB ID: 6OWN). A single zinc ion is coordinated by C118, V119, and Y128 (residues shown as sticks). (B) A magnification of the zinc ion and coordinating residues. C118, V119, and Y128 are colored by atom as carbon (green), nitrogen (blue), oxygen (red), and sulfur (yellow).