The ATPase domain of HscC (DnaK homolog) is essential for interfering σ⁷⁰ activity in E. coli

Abstract

HscC, a DnaK homolog in Escherichia coli, consists of adenosine triphosphatase (ATPase), substrate-binding and C-terminal domains. Overexpression of HscC markedly inhibits growth of host cell and reduces the σ⁷⁰-dependent promoter activity presumably by forming a complex with σ⁷⁰. To identify the region(s) of HscC responsible for growth inhibition and complex formation with σ⁷⁰, domain swapping experiments were carried out between DnaK and HscC. Thus the chimeric proteins carrying the ATPase domain of HscC and substrate-binding domains of either HscC or DnaK were found to inhibit the growth of the cell, reduce the σ⁷⁰-dependent promoter activity and form a complex with σ⁷⁰. These results indicate that the ATPase domain of HscC rather than the substrate-binding domain is important for determining its functional specificity.

HscC, DnaK homolog, σ⁷⁰, Adenosine triphosphatase domain

1 Introduction

Members of the Hsp70 family are highly conserved molecular chaperones that play important roles in dealing with cellular proteins that are unfolded or damaged during synthesis or exposure to environmental stress [1,2]. Chaperone proteins that belong to the Hsp70 family have adenosine triphosphatase (ATPase) activity and exhibit typical structural features, such as ATPase, substrate-binding and C-terminal domains [3]. Crystal structure of the ATPase domain of bovine Hsc70 reveals a bi-lobed domain with a deep cleft in which the adenosine triphosphate (ATP) molecule is bound [4]. The substrate-binding domain is necessary and sufficient for substrate binding; however, normal regulation of the substrate-binding activity does not occur in the absence of the ATPase domain [5,,,8]. Communication between ATPase and substrate-binding domains of Hsp70 is fundamental to its function because adenine nucleotides regulate the interaction with substrate proteins [9].

In Escherichia coli three Hsp70 proteins, DnaK, HscA (Hsc66) and HscC (Hsc62) [10,11] and six Hsp40 proteins, DnaJ, CbpA, DjlA, HscB (Hsc20), YbeV and YbeS have been identified [11]. Three Hsp40 proteins, DnaJ, CbpA and DjlA, have been shown to serve as co-chaperones for DnaK [12,,14], whereas HscB was reported to act similarly for HscA [15]. DnaJ interacts with DnaK at either the ATPase domain and/or the substrate-binding domain and these interactions have been shown to be important for physiological functions of DnaK [16,17]. DnaJ accelerates the ATPase activity of DnaK several folds [18] whereas CbpA can compensate for biochemical and physiological function of DnaJ [13,19,20]. Similarly, HscB interacts with HscA and enhances the ATPase activity of HscA in vitro [15].

The HscC protein with a molecular mass of 62 kDa is the smallest among members of the Hsp70 family in E. coli[21]. The ATPase and C-terminal domains of HscC are smaller compared to those of DnaK [11]. Nevertheless, HscC displays significant ATPase activity and binds to gelatin in the same manner as DnaK, consistent with its entry as a member of the Hsp70 family [21]. The ATPase activity of HscC is enhanced by YbeV [22].

We recently reported that HscC physically interacts with σ⁷⁰ responsible for transcription of most of the genes during exponential growth and negatively modulates its activity [23]. As might have been expected, the cloned hscC gene on multicopy plasmid did not complement the temperature sensitivity of ΔdnaK52 mutant. Instead, overexpression of HscC caused severe inhibition of cell growth [23]. In this study, we carried out domain swapping experiments to identify the region(s) of HscC responsible for growth inhibition and for physical interaction with σ⁷⁰ leading to the inhibition of transcription of numerous genes.

2 Materials and methods

2.1 Bacterial strains and plasmid

E. coli MC4100 (F⁻araD139 Δ(argF-lac)U19 rpsL150 relI flbB53O1 deoC1 ptsF25 rbsR) and MARF4 (MC4100/λ pF13-Pbla::lacZ) strains were used in this study. pTrc99A is a derivative of pBR322 [23].

2.2 Media, enzymes and chemicals

Cells were grown in Luria–Bertani (LB) broth. When necessary, different concentrations (w/v) of IPTG (isopropylthio-β-d-galactoside, 0.05, 0.1, 0.3 and 0.5 mM) were added. Ampicillin and chloramphenicol (Wako, Japan) were used at 50 μg ml⁻¹. Restriction enzymes, ligation kit, PCR kit (polymerase chain reaction kit), T4 polynucleotide kinase, alkaline phosphatase (calf intestine) and T4 DNA polymerase were supplied by Takara Shuzo Co., Kyoto, Japan. All chemicals used were supplied by Wako, Japan, except for sodium chloride (Nacalai, Japan), yeast extract, and tryptone peptone (Difco Laboratories).

2.3 β-Galactosidase assay

The bla promoter activity was monitored by assaying β-galactosidase expressed from bla promoter. Cells were grown at 37°C in LB medium to OD₆₀₀∼0.3. 0.3 mM IPTG was added, and samples were taken after 1 h incubation. β-Galactosidase activity was determined according to Miller [24].

2.4 Purification of His-tagged proteins

MC4100 cells producing histidine-tagged versions of DnaK, HscC, putative ATPase and substrate-binding domain-carrying chimera molecules were grown at 37°C in 200 ml of LB cm⁻¹ medium to OD₆₀₀ 0.3. IPTG (0.3 mM) was added to induce proteins, and samples were taken after 2 h. Cells were collected by centrifugation (4000×g, 10 min at 4°C) and resuspended in 20 ml of cold TEG buffer (20 mM Tris–HCl (pH 7.5), 100 mM NaCl, 0.1 mM ethylenediamine tetraacetic acid (EDTA), 20% glycerol). All subsequent manipulations were done at 4°C. Crude extracts were obtained by sonication (6×20 s, level 5, Astrason ultrasonic processor) and centrifugation (11 900×g, 30 min) and were loaded onto a 1 ml nickel (Ni²⁺) column (prepared according to manufacturer's instructions (Probond Resin, Invitrogene) and equilibrated with buffer TEG). Affinity chromatography of extracts was performed at 4°C. Loaded columns were first washed with 20 ml of buffer TEG, eluted with a 0–1000 mM imidazole linear gradient, and fractions of 0.5 ml were collected. Homogeneity of proteins (about 90%) was confirmed by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE).

2.5 SDS–PAGE and Western blotting

Protein samples were mixed with 4×sample buffer (400 mM dithiothreitol (DTT), 40% glycerol, 8% SDS, 0.04% bromophenol blue, 200 mM Tris–HCl (pH 6.8)), boiled for a few minutes and subjected to electrophoresis in 12.5% SDS–polyacrylamide gels. Proteins on the gels were analyzed by Western blotting on polyvinylidene difluoride membranes (PVDF; Immobilon, Millipore). The membranes were first incubated with antibodies against σ⁷⁰, or RGS-His (antibody that recognizes the arginine–glycine–serine–histidine epitope found on proteins encoded by 6×His fragment of pCA24N vectors derived from pQE9, Qiagen) and then with HRP (horseradish peroxidase)-conjugated anti-rabbit IgG antibody (Cappel) or anti-mouse IgG (for RGS-His (Dako)) according to the standard procedures. Immunoreactive bands were visualized by using the enhanced chemiluminescence (ECL) detection kit (Amersham) according to the manufacturer's instructions.

3 Results

3.1 Construction of chimeric hscC/dnaK genes

To perform domain swapping of ATPase, substrate-binding and C-terminal domains of HscC and DnaK, AatII and PmaCI restriction sites were introduced into interdomain regions separating the ATPase and substrate-binding domains, and those separating the substrate-binding and the C-terminal domains, respectively. Introduction of the restriction sites did not affect the overall function of chaperones significantly (data not shown). Besides the amino acid changes caused by creating the restriction sites in hscC (pIRZ20) did not change the toxicity profiles in comparison to the wild-type HscC (Fig. 2). The set of strains carrying chimeric genes shown in Fig. 1 were then constructed by inserting into pTrc99A vector at the cloning site downstream of an IPTG-inducible Ptrc promoter and transforming into MC4100 cells. All chimeric genes carried by a series of pIRZ plasmids (pIRZ10–pIRZ13 and pIRZ19–pIRZ26) could be expressed upon addition of IPTG and the products were obtained mostly in soluble form as visualized by SDS–PAGE followed by staining with Coomassie brilliant blue (data not shown).

2

Overproductions of chimeras carrying the ATPase domain of HscC inhibit cell growth. MC4100 cells containing each of the HscC/DnaK chimeras were treated with different concentrations of IPTG as indicated and tested for growth at 37°C. Growth was assayed by determining the ability of the cells to form colonies on LB agar plates. +, normal growth, and −, no growth.

1

Construction and structure of chimeric dnaK and hscC genes on the multicopy pIRZ plasmids. The domain structures of the DnaK (open) or HscC (hatched) proteins and the positions of domain boundaries are schematically presented. Amino acid changes caused by the introduction of restriction sites are indicated. Two different restriction sites were introduced by sequence-specific mutagenesis. An AatII site was introduced into the boundary separating the ATPase and substrate-binding domains (385/351 amino acid positions for DnaK or HscC, respectively) leading to I351V (isoleucine to valine) changes for HscC, whereas a PmaCI site was created within the interdomain region between the substrate-binding and C-terminal domains (513/452 amino acid positions for DnaK or HscC, respectively) leading to E452H (glutamic acid to histidine) changes for HscC. The set of chimeric or truncated genes shown here were constructed by using these modified prototype genes. Letters on the right side of each construct give the name of the plasmid.

3.2 Chimeras carrying the ATPase domain of HscC show toxicity

It was reported that overexpression of HscC shows severe toxicity [23]. Therefore, to identify domains responsible for this effect, we constructed a series of pIRZ plasmids that produce chimeric HscC and DnaK proteins. MC4100 wild-type strain was transformed with plasmids and plated on LB agar plates containing IPTG. After overnight incubation at 37°C growth inhibition was observed by the overexpression of HscC with or without its C-terminal domain or the chimeras carrying the ATPase domain of HscC (Fig. 2; pIRZ12, pIRZ20, pIRZ21, pIRZ22 and pIRZ23). Subsequently, no growth inhibition was observed by the overexpression of putative ATPase domain of HscC, DnaK with or without its C-terminal domain or chimeras carrying the ATPase domain of DnaK (Fig. 2; pIRZ13, pIRZ19, pIRZ10, pIRZ11, pIRZ24, pIRZ25 and pIRZ26). Furthermore, the overexpression of putative HscC containing substrate-binding and C-terminal domains did not show any growth inhibition (data not shown). This result indicates that the ATPase domain of HscC plays a critical role for the specific growth inhibition effect.

3.3 Chimeras carrying the ATPase domain of HscC specifically inhibit the σ⁷⁰-dependent bla promoter activity

It was previously reported that overexpression of HscC specifically inhibits the σ⁷⁰-dependent promoter activity [23]. To identify the domain(s) that are responsible for this inhibition, we investigated the effect of the above set of chimeric proteins on functioning of the σ⁷⁰-dependent bla promoter in a strain carrying λPbla-lacZ by measuring β-galactosidase activity. The results showed that overexpression of HscC with or without its C-terminal domain or the chimeras carrying the ATPase domain of HscC inhibits the σ⁷⁰-dependent bla promoter activity appreciably (Fig. 3; pIRZ12, pIRZ20, pIRZ21, pIRZ22 and pIRZ23). As shown for the effect on cell growth, the C-terminal domain is not required for this effect. In contrast, overexpression of putative ATPase domain of HscC, DnaK with or without its C-terminal domain or chimeras carrying the ATPase domain of DnaK combined with substrate-binding domain of HscC or DnaK failed to inhibit bla promoter activity (Fig. 3; pIRZ13, pIRZ19, pIRZ10, pIRZ11, pIRZ24, pIRZ25 and pIRZ26). It thus seems evident that the ATPase domain rather than the substrate-binding domain of HscC is primarily responsible for specific inhibition of cell growth and of σ⁷⁰-dependent promoter activity.

3

Overproductions of chimeric proteins carrying the ATPase domain of HscC inhibit the σ⁷⁰-dependent bla promoter activity. Cultures of MARF4 (MC4100/λ pF13-Pbla::lacZ) carrying the indicated plasmids were grown at 37°C to OD₆₀₀ 0.3; aliquots were treated with or without IPTG (0.3 mM) for 1 h, and samples were assayed for β-galactosidase activity. Enzyme activities (average of three individual experiments) are presented in Miller units. Black bars and open bars represent the samples treated with or without IPTG, respectively. Lower panel shows the domain features of each chimeric gene: C or K indicates HscC or DnaK, respectively.

3.4 σ⁷⁰ is co-purified with the chimeric Hsp70 proteins containing the ATPase domain of HscC

Previous observations suggested that overexpression of HscC physically interacts with σ⁷⁰ and this interaction is critical for cell growth [23]. Therefore, to identify the domain(s) of HscC that interacts specifically with σ⁷⁰, we investigated the interaction between chimeric HscC/DnaK proteins and σ⁷⁰ in vivo by employing a co-purification strategy. We constructed plasmids expressing His-tagged chimeric proteins that were functionally active as judged by the inhibitory effects on cell growth (data not shown). The proteins were induced by addition of IPTG to MC4100 cells containing each plasmid and purified from crude cell extracts by Ni²⁺ column chromatography, followed by imidazole elution. Upon immunoblotting analysis with antibody against σ⁷⁰ and His-tag, all chimeric proteins containing the ATPase domain of HscC, as well as intact HscC, were co-eluted with σ⁷⁰ (Fig. 4; pIRZ28, pIRZ29, pIRZ30, pIRZ31 and pIRZ16). In contrast, no significant amount of σ⁷⁰ was co-eluted with the ATPase domain of HscC by itself (pIRZ17) or with the chimeric proteins containing the ATPase domain of DnaK (Fig. 4; pIRZ27, pIRZ32, pIRZ33, pIRZ34, pIRZ14 and pIRZ15). In addition, the putative HscC without ATPase domain of HscC could not interact with σ⁷⁰ (data not shown). These observations strongly suggest that the ATPase domain of HscC is primarily responsible for the specific interaction with σ⁷⁰ which presumably reduces its activity and for the eventual inhibition of host cell growth.

4

σ⁷⁰ co-purified with the chimeric proteins containing the ATPase domain of HscC. The domain structures of the DnaK (open) or HscC (hatched) proteins are schematically presented. Cultures of MC4100 containing different histidine-tagged plasmids were grown to OD₆₀₀ 0.3 and induced by 0.3 mM IPTG for 2 h. Proteins were purified by Ni²⁺ affinity chromatography as described in Section 2. Elution fraction numbers are indicated at the top. The eluted fractions were used for Western blotting analysis with anti-σ⁷⁰ and anti-RGS-His antibody. His-tagged indicates the purified chimeric proteins and σ⁷⁰ indicates the co-purified σ⁷⁰ protein.

4 Discussion

We observed that chimeric proteins that consist of the ATPase domain of HscC and substrate-binding domain of either HscC or DnaK inhibit cell growth presumably by inhibiting transcription from σ⁷⁰-dependent promoters (Figs. 2 and 3). In agreement with these results, all the chimeric proteins carrying the ATPase domain of HscC combined with other domains of either HscC or DnaK were found to form complexes with σ⁷⁰ (Fig. 4). These observations strongly suggested that the ATPase domain is important for specific binding of HscC with σ⁷⁰ that would reduce its activity in initiating transcription from σ⁷⁰-dependent promoters.

Similar domain-exchange experiments between two Hsp70 proteins in yeast SsbI and SsaI were reported, the chimeric proteins carrying the ATPase domain of SsbI (but not SsaI) and two other domains of SsbI or SsaI complement the cold sensitive phenotype of ΔssbI mutant [25]. The ATPase domain is therefore thought to have a crucial role in determining specificity of Hsp70 molecules in their characteristic response to environmental changes. The present results support the idea that the ATPase domain of HscC plays a central role in determining the functional specificity of HscC. The ATPase domain by itself, however, exhibits no inhibitory activities characteristic of HscC, because overexpression of ATPase domain of HscC alone failed to show any inhibition of growth or σ⁷⁰ activity (Figs. 2 and 3). A major question that remains to be resolved is how the chimera Hsp70 molecules carrying the ATPase domain of HscC inhibit transcription from σ⁷⁰-dependent promoter and exhibit cellular toxicity. One possible explanation is that the two functional domains (ATPase and substrate-binding domains) of HscC play important role for the physiological function; the ATPase domain is crucial for the substrate recognition through physical interaction with the substrate (σ⁷⁰) and the substrate-binding domain is important for substrate stabilization. We could not also ignore the other possibility that substrate (σ⁷⁰) initially interacts with a co-chaperone, which might be important for HscC activity, then co-chaperone–substrate complex interacts with the ATPase domain of HscC; a subsequent conformational change of HscC could allow the substrate bound to co-chaperone to be transferred to the substrate-binding domain of HscC. It has been hypothesized that the functional specificity of Hsp70 is determined by physical interaction between one or more Hsp70 domains and Hsp40 proteins [17,25]. The DnaJ protein, a major Hsp40 homolog in E. coli, interacts with DnaK through the ATPase and substrate-binding domains and this interaction is crucial for physiological function of DnaK [17]. Moreover, another Hsp40 homolog in E. coli, HscB, interacts with HscA and enhances the ATPase activity of HscA [15], suggesting that co-chaperones play a crucial role for physiological function of Hsp70 proteins. HscC might also require a co-chaperone for its function that interacts with the ATPase domain. Recently, Yoshimune et al. observed that YbeV, encoding a DnaJ homolog, is closely linked to hscC in E. coli, and stimulates ATPase activity of HscC, thus implicating YbeV to be the co-chaperone for HscC [22].

In this analysis, the ATPase domain was found to be the key factor for determination of the functional specificity of HscC and substrate-binding domains might assist its physiological function of HscC probably through stabilization of binding to the target proteins. This approach should provide further insights into the mechanism of determination of functional specificity among the members of Hsp70 families.

Acknowledgements

This work was supported by a Grant-in-Aid for Scientific Research on Priority Areas from the Ministry of Education, Culture, Sports, Science and Technology of Japan, a grant from CREST, JST (Japan Science and Technology) and in part from NEDO (New Energy and Industrial Technology Development Organization). We thank Takashi Yura for manuscript preparation.