A novel nucleoid-associated protein of Mycobacterium tuberculosis is a sequence homolog of GroEL

Abstract

The Mycobacterium tuberculosis genome sequence reveals remarkable absence of many nucleoid-associated proteins (NAPs), such as HNS, Hfq or DPS. In order to characterize the nucleoids of M. tuberculosis, we have attempted to identify NAPs, and report an interesting finding that a chaperonin-homolog, GroEL1, is nucleoid associated. We report that M. tuberculosis GroEL1 binds DNA with low specificity but high affinity, suggesting that it might have naturally evolved to bind DNA. We are able to demonstrate that GroEL1 can effectively function as a DNA-protecting agent against DNase I or hydroxyl-radicals. Moreover, Atomic Force Microscopic studies reveal that GroEL1 can condense a large DNA into a compact structure. We also provide in vivo evidences that include presence of GroEL1 in purified nucleoids, in vivo crosslinking followed by Southern hybridizations and immunofluorescence imaging in M. tuberculosis confirming that GroEL1: DNA interactions occur in natural biological settings. These findings therefore reveal that M. tuberculosis GroEL1 has evolved to be associated with nucleoids.

Figure 1.

Binding of Mtb GroEL1 to DNA. (A) DNA-binding activity of GroELs from different organisms. The proteins (in 1 µM concentration) tested for DNA-binding are lanes, 1: Probe alone; 2: (His)₆-tagged Mtb GroEL1; 3: Mtb GroEL-2; 4: M. leprae Cpn60.2; 5: Human Cpn60; 6: E. coli GroEL; 7: Non-(His)₆ tagged GroEL1. ³²P-labeled inverted repeat DNA was used as the probe for binding assays, as chaperonin expression is known to be controlled by similar inverted repeat sequences. It is clearly seen that only non-(His)₆ tagged GroEL1 binds DNA, while others do not. (B) Determination of the binding constant between Mtb GroEL1 and DNA. The binding constants were determined by performing EMSA reactions with a 10 nM ³²P-labeled CIRCE2FR probe and by varying the concentrations of Mtb GroEL1 between 0–300 nM. These experiments were performed in triplicate and quantified by phosphorimager analysis. The percentage of bound probe was calculated by estimating diminishing intensity of the free probe. Binding was saturated at 300 nM concentration of the protein, which was considered to be 100% binding. The data were fitted with Sigmoidal curve, where 50% saturation corresponded to the K_d value. Error bars indicate standard error.

Figure 2.

The binding of Mtb GroEL1 to different mutants of CIRCE2F/CIRCE2FR element. (A) Binding assays with annealed double-stranded oligonucleotides. The nomenclature used for different oligonucleotide probes is listed in Table 1. Lane1 shows CIRCE2FR probe alone whereas other lanes show binding of Mtb GroEL1 to Lane 2: CIRCE2FR; Lane 3: mutO2; Lane 4: mutO3; Lane 5: mutO4; Lane 6: mutO5 and Lane 7: mutO6, Lane 8: mutO7. The concentration of protein used in this experiment was 1 µM, while that of the radiolabeled oligonucleotide was 100 nM. (B) Binding assays with single-stranded oligonucleotides. The lane labeling is identical as in Figure 2A, except that the oligonucleotides were not annealed with their complementary strands, thus maintaining these in the single-stranded form. (C) Binding of different concentrations of Mtb GroEL1 to 140-bp region upstream of Mtb GroES. Lane 1, probe alone; lanes 2–9, 0.2, 0.4, 0.6, 0.8, 1, 2, 3, 4 µM of Mtb GroEL1. Appearance of more retarded bands indicates non-specific binding of Mtb GroEL1 to DNA.

Figure 3.

Mtb GroEL1 binds preferentially to single-stranded DNA. (A) Sequences of oligonucleotides used for performing EMSA with increasing concentration of Mtb GroEL1 using 5′ end-labeled DNA. (B) The comparison of the binding of dsDNA and ssDNA was studied using radiolabeled dsDNA as probe and 1×, 20×, 100×, 500× Molar excess dsDNA (unlabeled) and ssDNA (unlabeled) as competitor to radiolabeled probe. Lane 1: probe alone, lane 2: probe and 1µM GroEL, lane 3: probe and 1× dsDNA (unlabeled), lane 4: probe and 20× dsDNA (unlabeled), lane 5: probe and 100× dsDNA (unlabeled), lane 6: probe and 500× dsDNA (unlabeled), lane 7: probe and 1× ssDNA (unlabeled), Lane 8: probe and 20× ssDNA (unlabeled), Lane 9: probe and 100× dsDNA (unlabeled), Lane 10: probe and 500× dsDNA (unlabeled).

Figure 4.

Binding of Mtb GroEL1 to γP³² labeled RNA. The binding of different concentration of Mtb GroEL1 to γP³²-labeled RNA sequence (UUCUUGCACUCGGCAUAGGCGAGUGCUA)—lane 1, probe alone, lanes 2–8, 0.25, 0.5, 0.75, 1, 2, 3, 4 µM of Mtb GroEL1. The concentration of γP³²-labeled probe used in all the lanes was 4 nM.

Figure 5.

Mtb GroEL1 interacts with DNA through the minor grove. (A) CIRCE2FR oligonucleotide was annealed with its complementary strand to form dsDNA which was then radiolabeled with γP³² ATP. The probe was incubated with increasing concentration of Actinomycin D, a minor groove binder (lanes 4–8, 0.1 mM–1 mM). The reactions were further incubated with 1 µM GroEL1 and resolved on native PAGE. As controls, labeled DNA was either loaded alone (lane 1), with GroEL1 (lane 2) or with 1 mM Actinomycin D (lane 3). (B) As in panel A, DNA was annealed with its complementary strand to form dsDNA and radiolabeled with γP³² ATP. The probe was incubated with increasing concentration of Methyl Green, a major groove binder (lanes 4–7, 0.1–0.5 mM). The reactions were further incubated with 1 µM GroEL1 and resolved on native PAGE. As controls, labeled DNA was either loaded alone (lane 1), with GroEL1 (lane 2) or with 0.5 mM Methyl Green (lane 3).

Figure 6.

Mtb GroEL1 protects plasmid DNA from DNase I digestion and oxidative radicals, and condenses DNA as evidenced by AFM. (A) Protection of plasmid supercoiling from oxidative damage: 0.8% Agarose gel scan shows two forms of plasmid DNA (pBSK) bands. Lane 1 shows 200 ng plasmid DNA without metal-catalyzed oxidation (MCO) system, lane 2 shows plasmid DNA with MCO system and lane 3 shows plasmid DNA preincubated with 14 µM Mtb GroEL1 with MCO system. (B) Protection of pBSK against DNase I digestion: lane 1 pBSK without DNase I, lane 2: pBSK digested with DNase I, and lane 3: pBSK incubated with GroEL1 prior to digestion with DNase I. 75 ng of pBSK and 7 µM Mtb GroEL1 was used in the assay. (C)–(E) AFM showing Mtb GroEL1 mediated DNA condensation. (C) pBSK DNA, scan size is 3 μm, (D) 1 ng/μl solution of pure Mtb GroEL1, scan size is 1.2 μm and (E) plasmid pBSK and Mtb GroEL1 incubated together for 10 min at room temperature, scan size is 586.5 μm. As is clearly seen, Mtb GroEL1 has profound effect on condensing the plasmid DNA.

Figure 7.

In vivo evidence for the Mtb GroEL1 binding to DNA. (A) Isolation of nucleoid from M. tuberculosis H37Rv strain by sucrose gradient centrifugation. (B) The isolated nucleoids separated on a 10% SDS–PAGE (lane 3) along with Mtb whole cell lysate (lane 2) and purified Mtb GroEL1. (C) The presence of Mtb GroEL1 in nucleoids is confirmed by western blotting (lane 2) along with Mtb whole cell lysate (lane 1) and purified Mtb GroEL1 (lane 3) as control. (D) Southern blot hybridization with restriction enzyme digested genomic DNA of Mtb and ChIP eluted DNA as probe. Lane 1: EcoRI digested Mtb genomic DNA, lane 2: BamHI digested Mtb genomic DNA.

Figure 8.

Mtb GroEL1 is associated with the nucleoid of Mtb. Brightfield image of Mtb H37Rv is shown. Cells were stained with DAPI to visualize the nucleoid (blue) and Alexa fluor 594 conjugated secondary antibodies against Mtb GroEL1 monoclonal antibody (red) to visualize Mtb GroEL1. The superimposition of DAPI stained nucleoid and Alexa fluor 594 stained Mtb GroEL is shown. The pink color clearly indicates the colocalization of GroEL1 with Mtb nucleoids.