Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Cold adaptation in the environmental bacterium Shewanella oneidensis is controlled by a J-domain co-chaperone protein network


DnaK (Hsp70) is a major ATP-dependent chaperone that functions with two co-chaperones, a J-domain protein (JDP) and a nucleotide exchange factor to maintain proteostasis in most organisms. Here, we show that the environmental bacterium Shewanella oneidensis possesses a previously uncharacterized short JDP, AtcJ, dedicated to cold adaptation and composed of a functional J-domain and a C-terminal extension of 21 amino acids. We showed that atcJ is the first gene of an operon encoding also AtcA, AtcB and AtcC, three proteins of unknown functions. Interestingly, we found that the absence of AtcJ, AtcB or AtcC leads to a dramatically reduced growth at low temperature. In addition, we demonstrated that AtcJ interacts via its C-terminal extension with AtcC, and that AtcC binds to AtcB. Therefore, we identified a previously uncharacterized protein network that involves the DnaK system with a dedicated JDP to allow bacteria to survive to cold environment.


Every organism possesses a complex network of molecular chaperones to control protein homeostasis, also called proteostasis1,2. In stress conditions, the roles of the chaperones become preponderant to prevent the formation of toxic misfolded and aggregated proteins. DnaK in bacteria or Hsp70 in eukaryotes is a major ATP-dependent molecular chaperone playing a key role in the proteostasis network3,4,5. It participates in the folding/unfolding, trafficking, disaggregation, and degradation of a multitude of client proteins, and therefore DnaK is involved in almost all cellular functions, from DNA replication to cell division and metabolism.

DnaK is a two-domain protein containing a nucleotide-binding domain that is connected by a flexible linker to a substrate-binding domain1,5,6. The substrate-binding domain is composed of a β-sheet subdomain and an α-helical lid. Allosteric communication between the nucleotide-binding domain and the substrate-binding domain is essential for the chaperone cycle of DnaK7,8,9. When ATP is bound to the nucleotide-binding domain of DnaK, the substrate-binding domain is in an open conformation with the α-helical lid contacting the nucleotide-binding domain1,8,10. In this conformation, substrate proteins interact with the substrate-binding domain with low affinity and high exchange rate. In contrast, in the ADP-bound conformation of DnaK, the α-helical lid closes over the substrate in the β-sheet subdomain resulting in high-affinity binding of the substrate and low exchange rate8,11.

DnaK conformational changes are regulated by two co-chaperones: a J-domain protein (JDP) and a nucleotide exchange factor. The JDP targets DnaK to substrate polypeptides, stimulates DnaK ATP hydrolysis leading to closure of the substrate-binding domain and trapping of the substrate. The nucleotide exchange factor allows ADP to ATP exchange from the ADP-bound DnaK-substrate complex, triggering substrate release1,8,12.

The JDPs were named from the first extensively studied member of this family, DnaJ from E. coli, and they all possess a J-domain of ~70 amino acids that is essential for the co-chaperone interaction with DnaK4,5,12,13. The J-domain is composed of four helices forming a hairpin, and a flexible loop that connects helices II and III which bears the HPD motif, a signature of the JDP proteins. Interestingly, there is a high diversity of JDPs that allows DnaK to possess such a wide variety of functions in the cells, and several JDPs are usually found in a given organism4,12,13. The JDPs have been classified into three categories, depending on their domain organization besides the J-domain12,13. Class A JDPs possess the structural features of E. coli DnaJ: a J-domain followed by a G/F rich region, two similar β-barrel domains containing a Zinc finger, and a C-terminal domain allowing dimerization. Class B is similar to class A, except that the Zinc finger region is missing. Class C represents the most diverse category of JDPs since these proteins possess a conserved J-domain that is associated with various domains different from that of class A and B. In some examples, these additional domains are responsible of targeting specific substrates to DnaK, they can allow the JDPs to localize at a specific position in the cell therefore helping the recruitment of DnaK at this position, and many are still of unknown functions12,13. Therefore, although many JDPs have been studied, much more work is required to understand the functional diversity and molecular mechanisms of many class C JDPs, in particular for bacteria which contain a great reservoir of uncharacterized JDPs14.

Bacteria from the Shewanella genus are ubiquitous aquatic γ-proteobacteria found in marine and fresh waters, and in sediments15. Given their natural ecological niches, these bacteria possess the ability to adapt to many stress conditions, including variations in a vast range of temperatures (from 4 °C to more than 42 °C), in hydrostatic pressure, or in osmotic conditions. Remarkably, Shewanella oneidensis appears like an exquisite bacterial model for chaperone study. Indeed, we have recently shown that Hsp90, another major chaperone, plays a crucial role for the growth of S. oneidensis at high temperature16,17, although Hsp90 is expendable in Escherichia coli18. By exploring the genome of S. oneidensis, we found that it encodes a functional DnaK system composed of the DnaK chaperone, the nucleotide exchange factor GrpE, and several JDPs, including DnaJ. Indeed, we identified four putative uncharacterized class C JDPs. Interestingly, S. oneidensis seems to have a strong requirement for the DnaK system since dnaK and dnaJ genes cannot be deleted19.

In this paper, we show that the DnaK system driven by a JDP, AtcJ, is required for cold adaptation in S. oneidensis. In addition, we found that AtcJ interacts with another protein of unknown function, AtcC, encoded in the same operon and that this protein is also essential during cold stress. Finally, we found that AtcC forms a complex with AtcB showing that a network of interaction allows the bacteria to cope with cold environments.


AtcJ is a functional J-domain protein

In addition to the canonical class A JDP DnaJ and the two class C JDPs DjlA and HscB4, we found that the genome of S. oneidensis encodes four putative uncharacterized class C JDPs. Here, we focus on one of them, SO_1850, that we called AtcJ (Adaptation to cold protein J). AtcJ is a 94-amino acid protein with a molecular mass of 11 kDa. As for all proteins of the Jdp family, it possesses a conserved J-domain that is composed of four helices and contains the classical tripeptide 31HPD33 located on an exposed loop connecting helices II and III (Fig. 1a, b)4,5,12. The J-domain of AtcJ shares 63% of sequence similarity with that of DnaJ from E. coli. Besides the J-domain, AtcJ contains a 21 amino acid C-terminal extension of unknown function only found in the proteins homologous to AtcJ (Supplementary Fig. 1A).

Fig. 1
figure 1

AtcJ is a functional J-domain protein. a Structure of the J-domain of E. coli DnaJ (PDB: 1XBL). Helices are shown in green, loops are in gray, and the conserved HPD tripeptide is in red. b Sequence alignment of the J-domains of E. coli DnaJ (DnaJEc), S. oneidensis DnaJ (DnaJSo), and AtcJ using the ClustalW/Omega method and Boxshade (ExPaSy). Black boxes indicate that the residues are strictly conserved and gray boxes that they have conserved substitutions. Green lines indicate the helices of DnaJEc shown in a. The HPD motif is framed in red. c Schematic of the chimera. The J-domain of DnaJEc is shown in green, and the glycine–phenylalanine-rich domain (G/F), the zinc-binding domain (Zn), and the C-terminal domain (CTD) are shown in light orange. The J-domain (J) of AtcJ is green hatched, and the C-terminal extremity of 21 residues is colored blue. The AtcJ chimera (AtcJchim) possesses the J-domain of AtcJ and the G/F, Zn and CTD domains of DnaJEc. The AtcJchimH31Q chimera has the H31Q point mutation in the J-domain of AtcJ. d Production of the wild-type chimera (AtcJchim) suppresses the growth phenotype of the E. coli Δ3 strain (dnaJ cbpA djlA) at high temperature. E. coli Δ3 strain containing the pBad33 vector (p) or plasmids as indicated were grown at 28 °C. Ten time serial dilutions were spotted on LB-agar plates containing 2% L-arabinose (w/v). Plates were incubated overnight at 37 °C and 43 °C. e Production of the wild-type chimera (AtcJchim) partially suppresses the motility phenotype of the E. coli Δ3 strain (dnaJ cbpA djlA). Same strains as in d were spotted on LB soft agar (0.3% (w/v)), and motility was observed by the circumference halo around the initial spot after 1 day at 28 °C. In d and e, plates are representative of at least three experiments. f Stimulation of the DnaK ATPase activity by AtcJ. ATPase activities were measured at 37 °C with 10 µM DnaK, 50 µM AtcJ or AtcJH31Q, or 2 µM DnaJ where indicated. The data from three replicates are shown as mean ± SD. t Test analysis shows that the difference measured is significant (**p < 0.01)

To test if the J-domain of AtcJ was functional, we constructed a chimera between E. coli DnaJ (DnaJEc) and AtcJ by substituting the J-domain of DnaJEc by that of AtcJ (Fig. 1c), and we took advantage of the known phenotypes (high-temperature sensitivity and motility defects) of the Δ3 E. coli strain devoid of three DnaK co-chaperones (DnaJ, CbpA, and DjlA)20,21,22. As expected, the growth of the Δ3 strain containing an empty vector was strongly reduced at high temperature, 43 °C, compared with 37 °C and production of DnaJEc from a plasmid rescued this phenotype (Fig. 1d)21. Interestingly, production of the AtcJ chimera allowed growth of the Δ3 strain at 43 °C, indicating that the J-domain of AtcJ is functional and can replace that of DnaJEc (Fig. 1d). Similarly, production of the AtcJ chimera restored the absence of motility of the Δ3 strain, although at a lower level than DnaJEc (Fig. 1e).

It is known that a single H to Q mutation in the conserved HPD motif of the J-domain proteins prevents physical and functional interactions with DnaK23,24. The chimera was therefore mutated accordingly (H31Q) (Fig. 1c), and as expected, production of this mutated chimera did not restore the growth defect at high temperature and the loss of motility of the Δ3 strain (Fig. 1d, e). As a control, we checked that the mutated chimera was produced at the same level as the wild-type chimera (Supplementary Fig. 1B). These experiments indicate that the J-domain of AtcJ can functionally replace that of DnaJ, therefore allowing association with DnaK.

Binding of JDPs to DnaK is known to trigger allosteric changes in DnaK that result in ATPase activity stimulation7,12, and as a control, we found that DnaJ from S. oneidensis stimulated the ATPase activity of DnaK (Fig. 1f; Supplementary Fig. 1C). To clearly demonstrate that AtcJ is a functional JDP, we measured DnaK ATPase activity in the presence of full-length AtcJ. We observed that, although at a lower level than with DnaJ, the ATPase activity of DnaK was stimulated 3.3 times by wild-type AtcJ, whereas very low stimulation was measured in the presence of AtcJH31Q (Fig. 1f). In the presence of the nucleotide exchange factor GrpE from S. oneidensis, the ATPase activity of DnaK was increased, and a 2.7-fold stimulation was observed when AtcJ was added (Supplementary Fig. 1C). Here again, AtcJH31Q did not almost stimulate the ATPase activity of DnaK in the presence of GrpE (Supplementary Fig. 1C).

Altogether, these experiments indicate that AtcJ possesses a bona fide J-domain that allows physical and functional binding with the DnaK chaperone.

AtcJ is essential for cold resistance

To understand the role of AtcJ in S. oneidensis, we compared the growth of an atcJ deletion strain with a wild-type strain. The two strains were grown at 28 °C, a permissive temperature for S. oneidensis, and dilutions were spotted on plates that were subsequently incubated at several temperatures. Strikingly, we found that the growth of the atcJ deletion strain was dramatically reduced at low temperature, 7 °C, compared with the wild-type strain that grew well after a few days (Fig. 2a). Similar phenotypes were also observed at temperatures ranging between 5 °C and 10 °C. At 15 °C and 28 °C, no relevant growth difference was observed between the two strains, as well as at 35 °C, a heat-stress temperature for S. oneidensis (Fig. 2a)16.

Fig. 2
figure 2

AtcJ allows cold adaptation. a Growth of S. oneidensis wild-type and mutant (ΔatcJ) strains at several temperatures on LB-agar plates. After initial growth at 28 °C, strains were diluted to OD600 = 1, and 10-time serial dilutions were spotted on LB-agar plates. Plates were incubated 10 days at 7 °C, 5 days at 15 °C, and 1 day at 28 °C and 35 °C. b Growth of S. oneidensis wild-type and mutant (ΔatcJ) strains at 7 °C and 28 °C in liquid media. After initial growth at 28 °C, strains as in A were diluted to OD600 = 0.05, and were incubated at 7 °C or 28 °C with shaking. Absorbance was measured with time. The data from three replicates are shown as mean ± SD. c Complementation of the growth phenotype. Wild-type or ΔatcJ strains containing as indicated the pBad33 vector (p) or the plasmids producing wild-type AtcJ, AtcJH31Q or AtcJΔC that lacks the last 21 residues were treated as in a, and dilutions were spotted on agar plates containing 0.2% arabinose. Plates were incubated at 7 °C (10 days) or at 28 °C (1 day). In a and c, plates are representative of at least three experiments. d The ΔatcJ strains containing the same plasmids as in c were grown overnight at 28 °C with 0.2% arabinose. Protein extracts were analyzed by western blot using anti-AtcJ antibody. The upper arrow indicates the bands corresponding to AtcJ-WT or H31Q mutant, and the lower arrow AtcJΔC. The HtpG protein (lower panel) was detected by a specific antibody to show that the same amount of the different extracts was loaded on the gel

We also compared the growth of the wild-type and the atcJ deletion strains in liquid cultures. Interestingly, although the wild-type strain grew well at 7 °C, growth of the atcJ deletion strain was strongly impaired at this temperature (Fig. 2b). At 28 °C, the two strains grew similarly as expected (Fig. 2b).

To confirm that AtcJ is essential at low temperature for S. oneidensis growth, atcJ was cloned into an inducible vector, and the resulting plasmid was introduced in the atcJ deletion strain. We observed that, at 7 °C, the production of AtcJ from a plasmid rescued the growth defect of the atcJ deletion strain to a level similar to that of the wild-type strain containing an empty vector (Fig. 2c). Interestingly, we found that the production of the AtcJH31Q mutant that is affected for DnaK stimulation (Fig. 1f) did not allow growth of the atcJ deletion strain at low temperature (Fig. 2c). We checked by western blot that AtcJH31Q was produced in the same amount as the wild-type AtcJ (Fig. 2d). These results strongly suggest that the essential function of AtcJ to support S. oneidensis growth at low temperature requires a functional J-domain as expected for a DnaK co-chaperone.

As described above, AtcJ is a very short Jdp that only contains 21 amino acids in addition to its J-domain. Using the growth phenotype of the atcJ deletion strain, we addressed the importance of its C-terminal region by producing a truncated mutant protein, AtcJΔC that lacks the last 21 amino acids. We found that AtcJΔC did not rescue the slight growth of the atcJ deletion strain at low temperature, indicating that the C-terminal extension of AtcJ is also essential for its function (Fig. 2c, d).

atcJ is encoded in an operon involved in cold resistance

By analyzing the genome of S. oneidensis, we noticed that a predicted three-gene operon of unknown function (SO_1849, SO_1848, and SO_1846) that we called atcA, atcB, and atcC, respectively, was encoded 120 -bp downstream of atcJ. Comparison of 80 bacterial genomes from the Shewanella, Moritella, and Aeromonas genus indicates that the synteny between atcJ and the three other genes is conserved, and that, in the majority of bacterial genomes (74/80), only a few nucleotides (8–24 bp) separated the end of atcJ from the beginning of the following gene of the putative operon. This observation suggests that, in S. oneidensis, atcJ could be encoded in the same operon as the three other genes. This hypothesis was explored by RT-PCR using converging primers located in each of the four genes (Fig. 3a). After RNA retro-transcription, PCR amplifications were observed between atcJ and atcA, atcA and atcB, and atcB and atcC (Fig. 3b), meaning that a same transcript contains several genes. atcJ was amplified from cDNA as a positive control (Fig. 3b, lane 1), whereas no amplification was observed using mRNA as matrix with atcJ primers (Fig. 3b, lane 5). These experiments strongly suggest that atcJ, atcA, atcB, and atcC are part of the same operon.

Fig. 3
figure 3

atcJ is encoded in an operon crucial for cold resistance. a Schematic of the atc operon. Specific oligonucleotides indicated by little arrows were used to amplify DNA fragments from the cDNA matrix. The fragments correspond to atcJ (1), atcJ-atcA (2), atcA-atcB (3), and atcB-atcC (4). b cDNA amplification. The total RNA from wild-type S. oneidensis strain grown at 7 °C to exponential phase was extracted and retro-transcribed in cDNA. PCR were performed using cDNA as a matrix and the oligonucleotides indicated in a. No amplification (lane 5) was observed when the primers 1 (see panel a) were used in the PCR with RNA as a matrix, indicating that the RNA preparation was not contaminated by the chromosome. c Growth of the ΔatcA, ΔatcB, and ΔatcC mutant strains. After initial growth at 28 °C, S. oneidensis strains were diluted to OD600 = 1, and 10-time serial dilutions were spotted on LB-agar plates. Plates were incubated 10 days at 7 °C and 1 day at 28 °C. d Complementation of the growth phenotype. Wild-type, ΔatcB, or ΔatcC strains containing as indicated the pBad33 vector (p) or plasmids with the atcB or atcC gene were treated as in c, and dilutions were spotted on 0.2% arabinose LB-agar plates that were incubated at 7 °C for 10 days. In c and d, plates are representative of at least three experiments

The proteins AtcA (20 kDa), AtcB (29 kDa), and AtcC (34 kDa) from S. oneidensis as well as their homologs in other organisms have not yet been characterized. Classical bioinformatics analysis did not identify any known domains, protein sequence motifs, signal peptide or transmembrane segment. Since we showed that AtcJ is important for cold resistance, we wondered whether the other proteins encoded in the same operon could also play a role in this resistance mechanism. Wild-type and atcA, atcB, or atcC deletion strains were grown overnight at permissive temperature, diluted and spotted on plates that were incubated at 7 °C or 28 °C. We found that although the atcA deletion strain grew like the wild-type, the atcB, and atcC deletion strains had very poor growth at 7 °C and behave similarly to the atcJ deletion strain (Figs. 2a, 3c). At 28 °C, all the strains grew as the wild-type. Growth defects at low temperature of the atcB and atcC deletion strains were complemented by the production from a plasmid of AtcB and AtcC proteins, respectively, in the presence of the arabinose inducer (Fig. 3d). These results support the hypothesis that AtcJ, AtcB, and AtcC together work in the same pathway involved in cold resistance. However, growth defects of the atcJ, atcB, or atcC strains at low temperature were not rescued by the production of the other Atc proteins suggesting a specific role for these proteins (Supplementary Fig. 2).

Given the importance of AtcJ, AtcB, and AtcC at low temperatures, we tested if the expression of the atcJABC operon is induced at low temperature. Wild-type S. oneidensis cells were grown at 7 °C, 28 °C, and 37 °C, RNA was isolated, reverse transcribed in cDNA, and tested by quantitative PCR. We found that expression of the atcJABC operon did not almost vary with temperatures, although a very slight increase of expression was measured at low temperature (Supplementary Fig. 3A). In contrast, the expression of dnaK was strongly increased at 37 °C as expected25,26, and interestingly we observed that it was reduced at 7 °C (Supplementary Fig. 3B).

AtcJ interacts with AtcC

To better understand the role of AtcJ, we wondered if it could interact with a protein from the atc operon. We first used a bacterial two-hybrid assay based on the reconstitution of the T18 and T25 catalytic domains of the adenylate cyclase of Bordetella pertussis27,28. When AtcJ fused to the T18 domain and AtcC fused to the T25 domain were produced together in an adenylate cyclase deleted E. coli strain, we measured a high level of β-galactosidase activity which reflects the interaction between AtcJ and AtcC (Fig. 4a). This interaction is relevant since production of T18-AtcJ with the T25 domain alone, or T25-AtcC with the T18 domain alone led to background levels of β-galactosidase activity. In contrast, no interaction was detected between AtcJ and AtcA or AtcB, indicating that AtcC could be the only partner of AtcJ in the atc operon (Fig. 4a).

Fig. 4
figure 4

AtcJ interacts with AtcC. a Bacterial two-hybrid assay showing that AtcJ and AtcC interact. E. coli Bth101 strain co-transformed as indicated with the T18 and T25 plasmids coding for protein fusion between T18 and AtcJ, AtcJΔC, or the last 21 amino acids of AtcJ (AtcJ-C), and protein fusion between T25 and AtcA, AtcB, or AtcC were grown overnight at 28 °C. β-galactosidase activity was measured as explained in the Methods section. The data from three replicates are shown as mean ± SD. bd ITC experiments. Binding assays were performed at 25 °C with 36 µM AtcC and 285 µM AtcJ (b), 36 µM AtcC and 285 µM AtcJΔC (c), or 36 µM AtcC and 285 µM of the pep21 peptide corresponding to the last 21 amino acids of AtcJ (d). Top panels show heat exchange upon ligand titration, and bottom panels show the integrated data with binding isotherms fitted to a single-site binding model. The data shown are representative of two independent experiments. e Thermal Shift Assay experiments were performed as indicated in Methods with 10 µM AtcC, 40 µM AtcJ, or 40 µM pep21. Proteins were mixed as indicated in the figure, SYPRO Orange was added, and the temperature was increased from 10 °C to 90 °C with a 0.5 °C step size. Fluorescence was measured at each temperature. Dot lines indicate the temperature melting point of AtcC alone (purple, 31 °C ± 0.5), with AtcJ (blue, 44 °C ± 0.5), or with pep21 (light blue, 44.5 °C ± 0.5). The data shown are representative of four independent experiments

Since we found that AtcC as well as the C-terminal extension of AtcJ are both essential to support cold resistance (Figs. 2c, 3c), we wondered if this region could be involved in AtcC binding. To answer this point, we constructed two mutants of AtcJ fused to the T18 domain, one lacking the last 21 amino acids, AtcJΔC, and the other producing only the last 21 amino acids, AtcJ-C. Strikingly, we observed that binding with AtcC was abolished with the mutant AtcJΔC, whereas an interaction was measured with AtcJ-C (Fig. 4a). These results support the idea that the C-terminal extension of AtcJ is essential for AtcC binding, and that the J-domain is not involved in this interaction.

To confirm the interaction between AtcJ and AtcC and to have more insight into the affinity and thermodynamic parameters, we performed isothermal titration calorimetry using purified AtcC with wild-type AtcJ, AtcJΔC, or a synthetic peptide corresponding to the last 21 amino acids of AtcJ that we called pep21. We found that wild-type AtcJ strongly interacted with AtcC with a Kd of 7 nM ± 2 and a stoichiometry ratio of 1 (Fig. 4b), whereas no binding was detected with the AtcJΔC mutant (Fig. 4c). Moreover, the synthetic peptide pep21 interacted with AtcC with a Kd of 80 nM ± 12, confirming that the C-terminal region of 21 amino acids of AtcJ is the main site of interaction with AtcC (Fig. 4d). As a control, we found that a scrambled peptide did not interact with AtcC showing that AtcC interacts specifically with the last 21 amino acids of AtcJ (Supplementary Fig. 4).

We then developed a thermal shift assay to monitor the stability of AtcC in the presence or the absence of the peptide corresponding to the last 21 residues of AtcJ. In this experiment, purified AtcC was analyzed in a thermocycler with increasing temperature in step-by-step mode to induce thermal unfolding of the protein. A specific fluorophore, the Sypro orange, is quenched in aqueous solution and becomes unquenched when it interacts with exposed hydrophobic regions of the protein during unfolding. Therefore, the increase in fluorescence reflects the unfolding of AtcC. We found that AtcC alone had a melting temperature (Tm) of 31 °C ± 0.5 (Fig. 4e). Strikingly, when AtcC was incubated with the peptide, the Tm shifted to 44.5 °C ± 0.5 (Fig. 4e). Similar results were also observed when full-length AtcJ was incubated with AtcC, the Tm shifted to 44 °C ± 0.5 (Fig. 4e). We were not able to accurately determine a Tm value for AtcJ alone because its unfolding gave a very weak fluorescent signal for unknown reasons (Fig. 4e).

Together, these results indicate that binding of the last 21 amino acids of AtcJ stabilizes AtcC, either by directly contacting a region prone to unfolding, or by triggering a conformational change in AtcC that becomes more resistant to thermal unfolding.

AtcC binds to AtcB

Our data show that AtcJ, AtcB, and AtcC, that are encoded in the same operon, are functionally related to cold resistance in S. oneidensis, and that AtcJ interacts with AtcC. We next wondered if AtcA, AtcB, and AtcC are physically connected. Using a two-hybrid assay, we found that AtcB and AtcC interacted (Fig. 5a). However, no interaction was observed between AtcA and the two other proteins (Fig. 5a).

Fig. 5
figure 5

AtcB interacts with AtcC. a Bacterial two-hybrid assay. E. coli Bth101 strain co-transformed as indicated with the T18 and T25 plasmids coding for protein fusion between T18 and AtcC or AtcB, and protein fusion between T25 and AtcA or AtcB were grown overnight at 28 °C. β-galactosidase activity was measured as explained in the Methods section. The data from three replicates are shown as mean ± SD. b Co-purification assay. AtcB with a CBP tag was produced with or without AtcC with a 6-His tag in the E. coli MG1655 strain. CBP-AtcB was purified using CBP affinity resin and bound proteins were analyzed by western blot with anti-CBP and anti-His antibodies as indicated. c In vitro co-purification assay with purified proteins. Purified AtcB (16 µg) or HtpG (16 µg) with a Strep-tag, and purified AtcC (100 µg) with a 6His-tag were mixed as indicated (final volume 200 µL). Proteins with a Strep-tag were pulled down on Strep-Tactin beads, washed several times, and proteins were separated on SDS-PAGE. The results shown in b and c are representative of three independent experiments

The interaction between AtcB and AtcC was also confirmed by a co-purification experiment. Plasmids allowing the production of AtcC with a 6His-tag and AtcB with a CBP tag, as well as control plasmids, were co-transformed in an E. coli strain and the proteins were produced by adding the arabinose inducer. We purified AtcB on calmodulin beads and observed on a western blot with a 6His antibody that AtcC was co-purified with AtcB (Fig. 5b).

To make sure that the interaction between AtcB and AtcC was direct, in vitro co-purification experiments using purified proteins were performed. AtcB with a Strep-tag and AtcC with a 6His-tag were incubated together and AtcB was pulled down on Strep-Tactin beads. Interestingly, we observed that AtcC co-purified with AtcB (Fig. 5c), confirming that the binding is direct. As negative controls, we showed that AtcC neither interacted with the Strep-Tactin beads nor interacted with another protein with a Strep-tag, HtpG (Fig. 5c).

Altogether, our results indicate that there is a network of interactions between proteins encoded in the atc operon. We found that AtcC interacts with AtcJ and with AtcB, and that together these proteins support cold resistance.


In this paper, we identified a previously uncharacterized JDP, AtcJ, which is essential for the growth of S. oneidensis at low temperature. atcJ is the first gene of a 4-gene operon (atcJABC) that is involved in cold adaptation since the absence of atcB or atcC resulted in the same cold-sensitive phenotype as the ΔatcJ strain. Interestingly, a network of interaction was identified between proteins encoded from this operon: AtcJ interacts with AtcC, and AtcC interacts with AtcB (Fig. 6). In addition, discrete regions of AtcJ were identified for DnaK binding (i.e., the J-domain), and for interaction with AtcC (i.e., the C-terminal extremity of 21 amino acids) (Fig. 6).

Fig. 6
figure 6

Model of the role of the Atc proteins in cold resistance. The AtcJ, AtcA, AtcB, and AtcC proteins are encoded from the atcJABC operon. AtcJ binds DnaK with its J-domain, and binds AtcC with its last 21 amino acids (C-term). In addition, AtcC interacts with AtcB. This network of interaction could allow the transfer of AtcC, AtcB, or another protein to DnaK, or could target the DnaK system to a specific location in the cell. Altogether, these proteins play a key role to support bacterial growth at low temperature. Black solid arrows indicate physical or functional interactions demonstrated in this study; gray dashed arrows indicate putative interactions

Our experiments strongly suggest that AtcJ works as a co-chaperone of the DnaK system to support cold resistance. Indeed, AtcJ possesses a bona fide J-domain that can replace that of DnaJ in complementation assays of the characterized phenotypes of an E. coli strain devoid of JDPs (Fig. 1d, e). In addition, AtcJ stimulated the ATPase activity of DnaK (Fig. 1f), whereas AtcJH31Q that has the well-characterized single point mutation in the highly conserved HPD motif that prevents DnaK binding did not activate the ATPase (Fig. 1f). Finally, the production of AtcJH31Q did not support growth of the ΔatcJ strain at low temperature (Fig. 2c). However, it is important to mention that the stimulation of the ATPase activity of DnaK was much higher with DnaJ than with AtcJ (Fig. 1f; Supplementary Fig. 1C). This result was expected because in addition to the J-domain, other domains of DnaJ (absent in AtcJ) including the G/F region participate in the stimulation23,24,29.

Based on our data, we propose that AtcJ, by contacting both the DnaK chaperone by its J-domain and AtcC by its C-terminal extremity could connect DnaK to AtcC or to the complex formed by AtcC and AtcB (Fig. 6). AtcC or AtcB might then be transferred to DnaK as substrates, although the quite high binding affinity (7 nM) we determined for AtcJ and AtcC (Fig. 4b) seems not to be in agreement with a transient interaction expected between a co-chaperone and its substrate30,31. In other hypotheses, the AtcJ-AtcC-AtcB proteins could help in the recruitment of DnaK substrates, or target DnaK to a specific location in the cell. Altogether, the Atc proteins by a still unknown mechanism play a key role in supporting cold adaptation.

We found that the atcJABC operon is constitutively expressed from 7 °C to 37 °C (Supplementary Fig. 3A), although we showed that AtcJ, AtcB, and AtcC are crucial for bacterial growth only at low temperature (Figs. 2a, 3c). One possibility is that these proteins have to be already synthesized to allow cell survival to a sudden cold shock. On the other hand, these proteins might play an alternative role in addition to cold adaptation.

The operonic organization of AtcJ suggests that AtcJ might have evolved to be dedicated to work with AtcC or AtcB or to a protein targeted by the AtcB–AtcC complex, in contrast to DnaJ that interacts with many substrate proteins. This is consistent with other class C JDP that have selective client binding12,13. For example, the class C JDP HscB (Jac1 in mitochondria) that is conserved from bacteria to eukaryotes interacts with IscU (Isu), a scaffold protein involved in Fe–S cluster assembly32. Interestingly, HscB is encoded in the same operon as IscU in bacteria. HscB transfers IscU to a DnaK homolog, HscA (Ssq1), that facilitates the loading of Fe–S clusters into the recipient apo-proteins. Another example is given by eukaryotic scSwa2 (hsDNAJC6) that specifically binds clathrin proteins coated on vesicles, and with Hsp70 dissociate clathrin from vesicles, a process important to control membrane traffic pathways33.

Some class C JDPs are also known to recruit Hsp70 at specific locations in the cell like for instance Zuo1 in S. cerevisiae at the ribosome or Pam18 at mitochondrial translocons12,13. Therefore, an alternative model would be that the complex AtcB-AtcC might localize AtcJ, and therefore DnaK at some particular location where it might be needed to cope with cold environments. This hypothesis is in agreement with the reduced expression of dnaK observed at low temperature (Supplementary Fig. 3B) that could potentially lead to DnaK scarcity. Therefore, AtcJ could locally concentrate DnaK where it is required.

In bacteria, low temperature is a stress known to affect several cellular components and pathways, including a decrease in membrane fluidity, stabilization of RNA secondary structures leading to reduced transcription and translation, a diminution in protein export, and a slowdown in protein folding34,35,36,37. Bacterial cells have therefore developed ways to counteract these adverse effects. For example, they modify their membrane composition to maintain viscosity34, and some Shewanella species produce a specific polyunsaturated fatty acid called eicosapentaenoic acid38. In addition, bacteria adapt their RNA metabolism by producing a set of cold-shock proteins (Csp) and other proteins such as RNA helicase and exoribonucleases34,39,40. Chaperone proteins are also important players to cope with cold conditions and favor the folding of substrates. For example, the level of the trigger factor (TF), a chaperone that interacts with the ribosome, increases in E. coli and in some psychrophilic bacteria when temperature decreases41,42. Moreover, the absence of DnaK or its three co-chaperones together20,43, or of the export chaperone SecB44 lead to cold-sensitive phenotypes in E. coli, indicating that these chaperones are involved in protein homeostasis at low temperature. Recently, it has been shown that cold conditions could induce prophage excision in S. oneidensis leading to inhibition of trans-translation, a mechanism important for protein synthesis quality control, and consequently promoting biofilm formation at low temperatures45.

Future work will aim at understanding if the Atc proteins we identified participate in known mechanisms of cold adaptation, such as membrane composition modification, protein secretion, protein folding or assembly, or if they are involved in any other unanticipated pathways. Therefore, this work paves the way for new insight about stress adaptation mediated by a co-chaperone of the DnaK system and dedicated proteins.


Growth conditions and strains

The wild-type S. oneidensis strain used in this study is MR1-R46. Strains were grown aerobically in rich LB medium at the temperatures indicated in the figures. When necessary, LB-rich medium was supplemented with chloramphenicol (25 µg/mL), kanamycin (25 µg/mL), ampicillin (50 µg/mL), or streptomycin (100 µg/mL). The ΔatcJ, ΔatcA, ΔatcB, and ΔatcC strains were constructed by homologous recombination as described before47. atcJ (SO_1850) and atcC (SO_1846) coding sequences were deleted entirely. However, since the coding sequences of atcA (SO_1849) and atcB (SO_1848) overlap, and since the Shine-Dalgarno sequence of atcC is located at the 3′ extremity of atcB, atcA was deleted from base pair 1 to 520 (over 540 bp) and atcB from base pair 2 to 747 (over 759 bp). The E. coli Δ3 strain (W3110 background) is deleted of the genes coding for the three JDPs dnaJ, cbpA, and djlA21. The E. coli Bth101 was used for bacterial two-hybrid experiments27,28. Transformations in E. coli and conjugations in S. oneidensis were done as described before47,48.

Plasmid constructions

Plasmids used to obtain the ΔatcJ, ΔatcA, ΔatcB, and ΔatcC strains were constructed as follows. A 1 kb sequence composed of two overlapping PCR fragments of 500 bp upstream and downstream of the genes to be deleted was PCR-amplified from the chromosome of S. oneidensis MR-1 with specific primers (supplementary Table 1) and cloned into the suicide vector pKNG10147,49 at the ApaI and SpeI restriction enzyme sites for atcB and atcC, or BamHI and SpeI for atcA and atcJ.

To construct plasmids pET-DnaKSo, pET-DnaJSo, pET-GrpESo, pET-AtcJ, and pET-AtcC allowing the overproduction of DnaK-6His, DnaJ-6His, 6His-GrpE, AtcJ-6His, and AtcC-6His, respectively, genes were PCR-amplified from the chromosome of S. oneidensis MR-1 using forward oligonucleotides with the NdeI restriction site, and reverse nucleotides with the SacI restriction site. To construct plasmid pET-AtcB allowing the overproduction of AtcB-6His, atcB was PCR-amplified using forward oligonucleotides with the NdeI restriction site, and reverse nucleotides with the EcoRI restriction site. A sequence coding for a 6-His tag (6xCAC) was added on the forward primer for pET-GrpESo, and on the reverse primers for pET-DnaKSo, pET-DnaJSo, pET-AtcJ, pET-AtcB, and pET-AtcC (Supplementary Table 1). After digestion, the DNA fragments were inserted into the pET24b vector (Novagen) digested with the corresponding restriction enzymes. To construct pET52-AtcB and pET52-HtpGSo allowing production of Strep-AtcB and Strep-HtpG, the same strategy as above was used except that the PCR products and the pET52 vector (Novagen) were digested with the KpnI and SalI enzymes for atcB, and KpnI and SacI enzymes for htpG.

For two-hybrid experiments, plasmids pT18-AtcJ, pT18-AtcB, and pT18-AtcC coding for AtcJ, AtcB, and AtcC fused to the C-terminal extremity of the T18 domain of the adenylate cyclase were constructed as described above using the restriction enzymes EcoRI and XhoI and the pUT18-C-linker vector27. The pT18-AtcJ-C plasmid was constructed by cloning a synthetic sequence corresponding to the last 21 amino acids of AtcJ in pUT18-C-linker using the restriction sites EcoRI and XhoI. To construct pT25-AtcA, pT25-AtcB, and pT25-AtcC coding for fusion of AtcA, AtcB, and AtcC to the C-terminal extremity of the T25 domain of the adenylate cyclase, the restriction sites used were PstI and XbaI for atcA, and EcoRI and XhoI for atcB and atcC. The DNA fragments were subsequently inserted into pKT25-linker27.

To construct plasmid pBad33-AtcJ, atcJ was amplified by PCR with a forward oligonucleotide containing the XmaI restriction site and a Shine-Dalgarno sequence, and a reverse oligonucleotide containing the XbaI restriction site. After digestion, the PCR product was inserted into the pBad33 vector50. The same strategy was used to clone DnaJ from E. coli into pBad33 by doing a PCR from E. coli MG1655 chromosome using primers containing XmaI and XbaI restriction sites (supplementary Table 1). Plasmids pBad33-AtcA, pBad33-AtcB, and pBad33-AtcC were constructed by digesting the plasmids pET-AtcB and pET-AtcC with XbaI and SalI enzymes, and the digestion products were cloned into the pBad33 vector previously digested with the same enzymes. Therefore, AtcA, AtcB, and AtcC produced from these plasmids contain a 6-His tag at their C-terminal extremity.

To construct pBad33-AtcJchim coding for the chimera in which the J-domain of DnaJ is replaced by that of AtcJ, a 2-step PCR reaction was performed. First, the sequences coding for the J-domain of AtcJ (amino acids 1 to 73) was amplified by PCR using a reverse primer that overlaps with the forward primer allowing the amplification of the sequence coding for DnaJEc without its J-domain (from amino acid 71–376) (Supplementary Table 1). Second, the two overlapping PCR products were mixed and used as matrix for a PCR amplification with external primers containing the XmaI and XbaI restriction enzyme sites, before insertion into the pBad33 vector50.

For co-purification experiments, pBad24-CBP-AtcB was constructed by digesting pT25-AtcB with XbaI and XhoI enzymes, and the digestion product was cloned into the pBad24-CBP vector51 previously digested with the same enzymes.

To add the H31Q mutation in the J-domain of AtcJ, plasmids pET-AtcJ, pBad33-AtcJ, and pBad33-AtcJchim were mutated using the QuickChange kit (Agilent) according to the manufacturer instructions, resulting in plasmids pET-AtcJH31Q, pBad33-AtcJH31Q, and pBad33-AtcJchimH31Q. The same strategy was used to obtain the pET-AtcJΔC, pT18-AtcJΔC, and pBad33-AtcJΔC plasmids by adding the stop codon TAG instead of AAC coding for N74 by QuickChange reaction. All constructions were verified by sequencing.

E. coli Δ3 complementation assays

For complementation of the E. coli phenotypes by the chimera, plasmids pBad33-AtcJchim, pBad33-AtcJchimH31Q, or pBad33-DnaJEc were introduced in the E. coli Δ3 strain (W3110 background) deleted of dnaJ, cbpA, and djlA21. Strains were grown at 28 °C until OD600 = 3, diluted to OD600 = 1, and 10-time serial dilutions were spotted on LB-agar plates supplemented with 2% (w/v) L-arabinose. Plates were then incubated at 37 °C or 43 °C for 24 h. In addition, 2 µL of cells diluted to OD600 = 1 were spotted on LB soft agar 0.3% (w/v) supplemented or not with L-arabinose and incubated at 28 °C for 24 h.

Bacterial growth

To follow S. oneidensis growth in liquid cultures, cells were grown at 7 °C or 28 °C as indicated after overnight precultures at 28 °C, and OD600 was measured with time. To follow S. oneidensis growth on plates, cells in late exponential growth phase grown at 28 °C were diluted to OD600 = 1. Then, 2 µL of 10-time serial dilutions were spotted on LB-agar plates supplemented or not with 0.2% (w/v) L-arabinose. Plates were then incubated at 7 °C for 10 days, 15 °C for 5 days, and 28 °C or 35 °C for 24 h.

Control of the amount of the AtcJ or AtcJchim mutants

To check that the same amount of wild-type and mutant proteins were produced, the E. coli Δ3 strains containing the plasmids coding for AtcJchim wild-type or mutant, and the S. oneidensis strains containing the plasmids coding for AtcJ wild-type or mutants were grown at 28 °C with 0.2% L-arabinose. The total protein extracts from the same amount of cells were heat-denatured, loaded on SDS-PAGE, transferred by western blot and revealed with anti-DnaJ antibody for AtcJchim wild-type and mutant, anti-AtcJ antibody for AtcJ wild-type and mutants, or anti-HtpG to detect the HtpG protein as a loading control.

Bacterial Two-hybrid assay

Bacterial two-hybrid assays were performed as described earlier27,28 with some modifications. Bth101 strain lacking the adenylate cyclase gene27,28 was co-transformed with a first plasmid coding for the T18 domain fused to AtcJ wild-type, AtcJΔC, AtcJ-C, AtcB, or AtcC, and a second plasmid coding for the T25 domain fused to AtcA, AtcB, or AtcC. For negative controls, pUT18-C-linker or pKT25-linker vectors were used27. After 2 days at 28 °C, cells were grown overnight in liquid LB medium supplemented with 25 µg/mL kanamycin, 50 µg/mL ampicillin, and 1 mM IPTG. Cells were lysed by adding PopCulture Reagent solution (Agilent) and 1 mg/mL lysozyme during 15 min, before addition of buffer Z (100 mM phosphate buffer pH = 7, 10 mM KCl, 1 mM MgSO4, 50 mM β-mercaptoethanol). Then, 2.2 mM ortho-nitrophényl-β-galactoside (ONPG) was added, and β-galactosidase activity was measured using a modified Miller assay adapted for use in a Tecan Spark microplate reader as described earlier52.

Protein purification

BL21(DE3) strains (Novagen) containing pET-DnaKSo, pET-GrpESo, pET-AtcJ wild-type and mutants, pET-DnaJSo, pET52-AtcB, or pET52-HtpGSo were cultivated in LB medium at 37 °C until OD600 = 0.8. Protein production was induced by adding 1 mM iso-propyl-β-D-thiogalactopyranoside (IPTG), and after 2 h of culture at 28 °C, cells were collected by centrifugation. To produce AtcC, BL21(DE3) strain containing pET-AtcC was grown at 37 °C until OD600 = 0.8, 0.1 mM IPTG was added, the cells were incubated at 16 °C for 20 h and collected.

Cells containing His-tag proteins were resuspended in buffer A (40 mM phosphate buffer pH = 7.4, 500 mM NaCl, 10% glycerol), and lysed by French Press. Lysates were centrifuged 10 min at 10,000 rpm, and the supernatants were centrifuged 45 min at 45,000 rpm. The resulting supernatants were then loaded on a His-Trap HP/FF column (GE Healthcare) using an AKTA Pure system (GE Healthcare), the column was washed successively with buffer A containing 5, 20, and 50 mM imidazole. Proteins were eluted with 100, 200, and 500 mM imidazole in buffer A.

Cells containing Strep-AtcB and Strep-HtpG proteins were resuspended in buffer C (100 mM Tris-HCl pH = 8, 150 mM NaCl, 1 mM EDTA) and lysed by French Press. Lysates were centrifuged 10 min at 10,000 rpm, and the supernatants were centrifuged 45 min at 45,000 rpm. The resulting supernatants were then loaded on a Strep-Trap HP column (GE Healthcare) using an AKTA Pure system (GE Healthcare). The column was washed with buffer C, and proteins were eluted in the same buffer containing 2.5 mM desthiobiotin (Sigma-Aldrich).

For 6His- and Strep-tag protein purification, fractions containing the proteins of interest were pooled and concentrated using Amicon Centrifugal Filters (Millipore). DnaK, AtcJ wild-type and mutants, AtcC, and AtcB were purified further using a gel filtration column HiLoad 16/600 Superdex 200 pg (GE Healthcare) in buffer B (50 mM Tris-HCl, 100 mM KCl, 10% glycerol, and 1 mM DTT). GrpE, DnaJ, and HtpG were loaded on a PD-10 column (GE Healthcare) in buffer B. Protein concentrations were measured by Bradford assays. Proteins were stored at −80 °C. The pep21 synthetic peptide corresponding to the last 21 amino acids of AtcC (NNVIVTDPNSAMRELWDQFYP) was obtained from Smart Bioscience.

ATPase activity assay

DnaK ATPase activity assays were performed at 37 °C with DnaK (10 µM), AtcJ wild-type or mutant (50 µM), DnaJ (2 µM), and GrpE (2 µM) when indicated, in 25 mM Hepes pH = 7.5, 150 mM KCl, 5 mM MgCl2, 2 mM ATP, using a pyruvate kinase/lactate dehydrogenase enzyme-coupled assay as described before53.

Co-purification assays

Co-purification experiments on cellular extracts were performed as described earlier16. Briefly, E. coli MG1655 strains containing pBad24-CBP or pBad24-CBP-AtcB with pBad33-AtcC-6His or pBad33 as indicated were grown at 37 °C until OD600 = 0.8. Gene expression was induced by adding 0.05% L-arabinose, and cells were collected after 1 h. Cells were then resupended in buffer D (10 mM Tris-Hcl pH = 8, 150 mM NaCl, 1 mM Mg acetate, 1 mM imidazole, 2 mM CaCl2, 0.1% Triton X-100, 0.1% Tween-20, 20 mM β-mercaptoethanol), and lysed by French Press. After centrifugation, supernatants were incubated for 1 h at 4 °C with calmodulin beads (Agilent). Beads were then washed five times with buffer D and resuspended with loading buffer. After denaturation, proteins were loaded on SDS-PAGE, transferred by western blot, and visualized using anti-CBP (Merck) or anti-His (Thermo) antibodies.

For co-purification experiments using purified proteins, Strep-AtcB (16 µg), Strep-HtpG (16 µg), or AtcC 6-His (100 µg) were mixed as indicated in the figure in buffer E (100 mM Tris-HCl pH = 8, 150 mM NaCl, 1 mM EDTA, 0.05% Triton X-100) before the addition of 50 µL Strep-Tactin beads (IBA Lifesciences). After a 1-h incubation at 4 °C, beads were washed four times with 1 mL of buffer E, and eluted with 20 µL of buffer E supplemented with 5 mM desthiobiotin (Sigma-Aldrich). Proteins were heat denatured and separated by SDS-PAGE subsequently stained with Coomassie Blue.

Isothermal titration calorimetry

The AtcJ, AtcJΔC, and AtcC proteins were dialyzed overnight at 4 °C in buffer B to avoid discrepancies from buffer mismatch. The pep21 peptide was solubilized in buffer B. ITC experiments were performed at 25 °C using the MicroCal PEAQ-ITC (Malvern). In total, 285 µM of ligands were titrated using 19 injections of 2 µl against 36 µM of AtcC in the ITC cell with a constant stirring speed of 750 rpm, with the exception of pep21, for which only 8 injections were needed to extrapolate the data. Duplicates were performed for each measurement. To account for the dilution heat of the titrated ligands, AtcJ-WT, AtcJΔC, and pep21 were titrated, respectively, into buffer and subtracted from the measured data with AtcC. The data were fitted using a “One Set of Sites” model in the PEAQ-ITC Analysis Software.

Thermal shift assays

AtcC (10 µM), AtcJ (40 µM), or pep21 (40 µM) were mixed together as indicated in the figure in buffer B (final volume: 20 µL). SYPRO Orange (Sigma-Aldrich) was added to a final concentration of 10X according to the manufacturer, the temperature was increased of 0.5 °C every 30 s from 10 to 90 °C, and fluorescence was measured with time with a BioRad CFX96 Touch Real‐Time PCR instrument. Determination of protein melting point was done using BioRad CFX Manager 3.1 software54.

RNA preparation, PCR from cDNA, and qRT-PCR

RNA from S. oneidensis wild-type grown at 37 °C, 28 °C, or 7 °C to exponential growth phase (OD600 = 2) was extracted and reverse transcribed into cDNA as described earlier for other bacteria55. PCR was performed using specific primers (Supplementary Table 1) and cDNA as a matrix. For the negative control, RNA was used as a matrix to make sure that the RNA preparation was not contaminated by the chromosomal DNA. Amplification was performed using GoTaq® DNA Polymerase (Promega), and the following cycling parameters: 2′ at 95 °C, followed by 29 cycles of 30″ at 95 °C, 30″ at 55 °C, 2′30″ at 72 °C, and a final elongation of 5′ at 72 °C. Samples were loaded on 1% agarose gel prepared in TBE (Tris-Borate-EDTA) buffer and visualized with ethidium bromide staining.

Quantitative real-time PCR (qPCR) analyses were performed on a CFX96 Real-Time System (BioRad) as described before55. The results were analyzed using BioRad CFX Manager software, version 3.1 (BioRad, France). The 16s RNA gene was used as a reference for normalization. For each point, technical and biological duplicates were performed.

Statistics and reproducibility

Independent experiments were repeated three times to generate replicates used to calculate mean and standard deviation. Where indicated, t tests were performed with GraphPad Prism Software version 6.05 to compare two sets of unpaired data using a parametric test with a Welch’s correction.

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.

Data availability

All data generated or analyzed during this study are included in this published article (and its supplementary information files). Full-length gels and blots are showed in the Supplementary Fig. 5. All source data underlying the graphs and charts presented in the main figures is made available as Supplementary Data 1.


  1. 1.

    Balchin, D., Hayer-Hartl, M. & Hartl, F. U. In vivo aspects of protein folding and quality control. Science 353, aac4354 (2016).

    Article  Google Scholar 

  2. 2.

    Finka, A., Mattoo, R. U. H. & Goloubinoff, P. Experimental milestones in the discovery of molecular chaperones as polypeptide unfolding enzymes. Annu. Rev. Biochem. 85, 715–742 (2016).

    CAS  Article  Google Scholar 

  3. 3.

    Finka, A., Sharma, S. K. & Goloubinoff, P. Multi-layered molecular mechanisms of polypeptide holding, unfolding and disaggregation by HSP70/HSP110 chaperones. Front. Mol. Biosci. 2, 29 (2015).

    Article  Google Scholar 

  4. 4.

    Genevaux, P., Georgopoulos, C. & Kelley, W. L. The Hsp70 chaperone machines of Escherichia coli: a paradigm for the repartition of chaperone functions. Mol. Microbiol. 66, 840–857 (2007).

    CAS  Article  Google Scholar 

  5. 5.

    Mayer, M. P. & Gierasch, L. M. Recent advances in the structural and mechanistic aspects of Hsp70 molecular chaperones. J. Biol. Chem. (2018).

    Article  Google Scholar 

  6. 6.

    Bukau, B., Weissman, J. & Horwich, A. Molecular chaperones and protein quality control. Cell 125, 443–451 (2006).

    CAS  Article  Google Scholar 

  7. 7.

    Mayer, M. P. Intra-molecular pathways of allosteric control in Hsp70s. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 373, (2018).

    Article  Google Scholar 

  8. 8.

    Mayer, M. P. & Kityk, R. Insights into the molecular mechanism of allostery in Hsp70s. Front. Mol. Biosci. 2, 58 (2015).

    Article  Google Scholar 

  9. 9.

    Zhuravleva, A., Clerico, E. M. & Gierasch, L. M. An interdomain energetic tug-of-war creates the allosterically active state in Hsp70 molecular chaperones. Cell 151, 1296–1307 (2012).

    CAS  Article  Google Scholar 

  10. 10.

    Kityk, R., Kopp, J., Sinning, I. & Mayer, M. P. Structure and dynamics of the ATP-bound open conformation of Hsp70 chaperones. Mol. Cell 48, 863–874 (2012).

    CAS  Article  Google Scholar 

  11. 11.

    Bertelsen, E. B., Chang, L., Gestwicki, J. E. & Zuiderweg, E. R. P. Solution conformation of wild-type E. coli Hsp70 (DnaK) chaperone complexed with ADP and substrate. Proc. Natl Acad. Sci. USA 106, 8471–8476 (2009).

    CAS  Article  Google Scholar 

  12. 12.

    Kampinga, H. H. & Craig, E. A. The HSP70 chaperone machinery: J proteins as drivers of functional specificity. Nat. Rev. Mol. Cell Biol. 11, 579–592 (2010).

    CAS  Article  Google Scholar 

  13. 13.

    Craig, E. A. & Marszalek, J. How do J-proteins get Hsp70 to do so many different things? Trends Biochem. Sci. 42, 355–368 (2017).

    CAS  Article  Google Scholar 

  14. 14.

    Kampinga, H. H. et al. Function, evolution, and structure of J-domain proteins. Cell Stress Chaperon. (2018).

    Article  Google Scholar 

  15. 15.

    Hau, H. H. & Gralnick, J. A. Ecology and biotechnology of the genus Shewanella. Annu. Rev. Microbiol. 61, 237–258 (2007).

    CAS  Article  Google Scholar 

  16. 16.

    Honoré, F. A., Méjean, V. & Genest, O. Hsp90 is essential under heat stress in the bacterium Shewanella oneidensis. Cell Rep. 19, 680–687 (2017).

    Article  Google Scholar 

  17. 17.

    Honoré, F. A., Maillot, N. J., Méjean, V. & Genest, O. Interplay between the Hsp90 chaperone and the HslVU protease to regulate the level of an essential protein in Shewanella oneidensis. mBio 10, e00269-19 (2019).

  18. 18.

    Bardwell, J. C. & Craig, E. A. Ancient heat shock gene is dispensable. J. Bacteriol. 170, 2977–2983 (1988).

    CAS  Article  Google Scholar 

  19. 19.

    Yang, H. et al. Genome-scale metabolic network validation of Shewanella oneidensis using transposon insertion frequency analysis. PLoS Comput. Biol. 10, e1003848 (2014).

    Article  Google Scholar 

  20. 20.

    Genevaux, P., Schwager, F., Georgopoulos, C. & Kelley, W. L. The djlA gene acts synergistically with dnaJ in promoting Escherichia coli growth. J. Bacteriol. 183, 5747–5750 (2001).

    CAS  Article  Google Scholar 

  21. 21.

    Perrody, E. et al. A bacteriophage-encoded J-domain protein interacts with the DnaK/Hsp70 chaperone and stabilizes the heat-shock factor σ32 of Escherichia coli. PLoS Genet. 8, e1003037 (2012).

    CAS  Article  Google Scholar 

  22. 22.

    Shi, W., Zhou, Y., Wild, J., Adler, J. & Gross, C. A. DnaK, DnaJ, and GrpE are required for flagellum synthesis in Escherichia coli. J. Bacteriol. 174, 6256–6263 (1992).

    CAS  Article  Google Scholar 

  23. 23.

    Mayer, M. P., Laufen, T., Paal, K., McCarty, J. S. & Bukau, B. Investigation of the interaction between DnaK and DnaJ by surface plasmon resonance spectroscopy. J. Mol. Biol. 289, 1131–1144 (1999).

    CAS  Article  Google Scholar 

  24. 24.

    Wall, D., Zylicz, M. & Georgopoulos, C. The NH2-terminal 108 amino acids of the Escherichia coli DnaJ protein stimulate the ATPase activity of DnaK and are sufficient for lambda replication. J. Biol. Chem. 269, 5446–5451 (1994).

    CAS  PubMed  Google Scholar 

  25. 25.

    Gao, H. et al. Global transcriptome analysis of the heat shock response of Shewanella oneidensis. J. Bacteriol. 186, 7796–7803 (2004).

    CAS  Article  Google Scholar 

  26. 26.

    Neidhardt, F. C. & VanBogelen, R. A. Positive regulatory gene for temperature-controlled proteins in Escherichia coli. Biochem. Biophys. Res. Commun. 100, 894–900 (1981).

    CAS  Article  Google Scholar 

  27. 27.

    Battesti, A. & Bouveret, E. The bacterial two-hybrid system based on adenylate cyclase reconstitution in Escherichia coli. Methods San. Diego Calif. 58, 325–334 (2012).

    CAS  Article  Google Scholar 

  28. 28.

    Karimova, G., Pidoux, J., Ullmann, A. & Ladant, D. A bacterial two-hybrid system based on a reconstituted signal transduction pathway. Proc. Natl Acad. Sci. USA 95, 5752–5756 (1998).

    CAS  Article  Google Scholar 

  29. 29.

    Kityk, R., Kopp, J. & Mayer, M. P. Molecular mechanism of J-domain-triggered ATP hydrolysis by Hsp70 chaperones. Mol. Cell 69, 227–237.e4 (2018).

    CAS  Article  Google Scholar 

  30. 30.

    Feifel, B., Schönfeld, H. J. & Christen, P. D-peptide ligands for the co-chaperone DnaJ. J. Biol. Chem. 273, 11999–12002 (1998).

    CAS  Article  Google Scholar 

  31. 31.

    Perales-Calvo, J., Muga, A. & Moro, F. Role of DnaJ G/F-rich domain in conformational recognition and binding of protein substrates. J. Biol. Chem. 285, 34231–34239 (2010).

    CAS  Article  Google Scholar 

  32. 32.

    Vickery, L. E. & Cupp-Vickery, J. R. Molecular chaperones HscA/Ssq1 and HscB/Jac1 and their roles in iron-sulfur protein maturation. Crit. Rev. Biochem. Mol. Biol. 42, 95–111 (2007).

    CAS  Article  Google Scholar 

  33. 33.

    Fotin, A. et al. Structure of an auxilin-bound clathrin coat and its implications for the mechanism of uncoating. Nature 432, 649–653 (2004).

    CAS  Article  Google Scholar 

  34. 34.

    Barria, C., Malecki, M. & Arraiano, C. M. Bacterial adaptation to cold. Microbiol. Read. Engl. 159, 2437–2443 (2013).

    CAS  Article  Google Scholar 

  35. 35.

    Phadtare, S. Recent developments in bacterial cold-shock response. Curr. Issues Mol. Biol. 6, 125–136 (2004).

    CAS  PubMed  Google Scholar 

  36. 36.

    Feller, G. Protein folding at extreme temperatures: current issues. Semin. Cell Dev. Biol. 84, 129–137 (2018).

    CAS  Article  Google Scholar 

  37. 37.

    Piette, F., Struvay, C. & Feller, G. The protein folding challenge in psychrophiles: facts and current issues. Environ. Microbiol 13, 1924–1933 (2011).

    CAS  Article  Google Scholar 

  38. 38.

    Kawamoto, J. et al. Eicosapentaenoic acid plays a beneficial role in membrane organization and cell division of a cold-adapted bacterium, Shewanella livingstonensis Ac10. J. Bacteriol. 191, 632–640 (2009).

    CAS  Article  Google Scholar 

  39. 39.

    Gao, H., Yang, Z. K., Wu, L., Thompson, D. K. & Zhou, J. Global transcriptome analysis of the cold shock response of Shewanella oneidensis MR-1 and mutational analysis of its classical cold shock proteins. J. Bacteriol. 188, 4560–4569 (2006).

    CAS  Article  Google Scholar 

  40. 40.

    Keto-Timonen, R. et al. Cold shock proteins: a minireview with special emphasis on Csp-family of enteropathogenic Yersinia. Front. Microbiol. 7, 1151 (2016).

    Article  Google Scholar 

  41. 41.

    Kandror, O. & Goldberg, A. L. Trigger factor is induced upon cold shock and enhances viability of Escherichia coli at low temperatures. Proc. Natl Acad. Sci. USA 94, 4978–4981 (1997).

    CAS  Article  Google Scholar 

  42. 42.

    Piette, F. et al. Proteomics of life at low temperatures: trigger factor is the primary chaperone in the Antarctic bacterium Pseudoalteromonas haloplanktis TAC125. Mol. Microbiol. 76, 120–132 (2010).

    CAS  Article  Google Scholar 

  43. 43.

    Bukau, B. & Walker, G. C. Cellular defects caused by deletion of the Escherichia coli dnaK gene indicate roles for heat shock protein in normal metabolism. J. Bacteriol. 171, 2337–2346 (1989).

    CAS  Article  Google Scholar 

  44. 44.

    Ullers, R. S., Ang, D., Schwager, F., Georgopoulos, C. & Genevaux, P. Trigger Factor can antagonize both SecB and DnaK/DnaJ chaperone functions in Escherichia coli. Proc. Natl Acad. Sci. USA 104, 3101–3106 (2007).

    CAS  Article  Google Scholar 

  45. 45.

    Zeng, Z. et al. Cold adaptation regulated by cryptic prophage excision in Shewanella oneidensis. ISME J. 10, 2787–2800 (2016).

    Article  Google Scholar 

  46. 46.

    Bordi, C., Iobbi-Nivol, C., Méjean, V. & Patte, J.-C. Effects of ISSo2 insertions in structural and regulatory genes of the trimethylamine oxide reductase of Shewanella oneidensis. J. Bacteriol. 185, 2042–2045 (2003).

    CAS  Article  Google Scholar 

  47. 47.

    Baraquet, C., Théraulaz, L., Iobbi-Nivol, C., Méjean, V. & Jourlin-Castelli, C. Unexpected chemoreceptors mediate energy taxis towards electron acceptors in Shewanella oneidensis. Mol. Microbiol 73, 278–290 (2009).

    CAS  Article  Google Scholar 

  48. 48.

    Miller, J. H. Experiments in Molecular Genetics (Cold Spring Harbor Laboratory Press, New York, 1972).

  49. 49.

    Sarker, M. R. & Cornelis, G. R. An improved version of suicide vector pKNG101 for gene replacement in gram-negative bacteria. Mol. Microbiol. 23, 410–411 (1997).

    CAS  Article  Google Scholar 

  50. 50.

    Guzman, L. M., Belin, D., Carson, M. J. & Beckwith, J. Tight regulation, modulation, and high-level expression by vectors containing the arabinose PBAD promoter. J. Bacteriol. 177, 4121–4130 (1995).

    CAS  Article  Google Scholar 

  51. 51.

    Battesti, A. & Bouveret, E. Improvement of bacterial two-hybrid vectors for detection of fusion proteins and transfer to pBAD-tandem affinity purification, calmodulin binding peptide, or 6-histidine tag vectors. Proteomics 8, 4768–4771 (2008).

    CAS  Article  Google Scholar 

  52. 52.

    Baaziz, H. et al. ChrASO, the chromate efflux pump of Shewanella oneidensis, improves chromate survival and reduction. PloS One 12, e0188516 (2017).

    Article  Google Scholar 

  53. 53.

    Genest, O. et al. Uncovering a region of heat shock protein 90 important for client binding in E. coli and chaperone function in yeast. Mol. Cell 49, 464–473 (2013).

    CAS  Article  Google Scholar 

  54. 54.

    Gambari, C., Boyeldieu, A., Armitano, J., Méjean, V. & Jourlin-Castelli, C. Control of pellicle biogenesis involves the diguanylate cyclases PdgA and PdgB, the c-di-GMP binding protein MxdA and the chemotaxis response regulator CheY3 in Shewanella oneidensis. Environ. Microbiol. (2018).

    Article  Google Scholar 

  55. 55.

    Schoeffler, M. et al. Growth of an anaerobic sulfate-reducing bacterium sustained by oxygen respiratory energy conservation after O2-driven experimental evolution. Environ. Microbiol. (2018).

    Article  Google Scholar 

Download references


We thank Pierre Genevaux, Marianne Ilbert, Sue Wickner, Aurélia Battesti, Chantal Iobbi-Nivol, Cécile Jourlin-Castelli, Michel Fons and Olivier Lemaire for help and fruitful discussions, Yann Denis for assistance with cDNA preparation and qPCR. Pierre Genevaux, Marie-Pierre Castanié-Cornet, and Latifa El-Antak are thanked for kindly providing strains, antibodies, and peptides. This work was supported by the Centre National de la Recherche Scientifique and Aix Marseille Université (AMU). O.G. was supported by a grant from the Agence Nationale de la Recherche (ANR-16-CE11–0002–01).

Author information




N.J.M., F.A.H., D.B., V.M., and O.G. designed the research, interpreted the data, and wrote the paper. N.J.M., F.A.H., D.B., and O.G. performed experiments.

Corresponding author

Correspondence to Olivier Genest.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Maillot, N.J., Honoré, F.A., Byrne, D. et al. Cold adaptation in the environmental bacterium Shewanella oneidensis is controlled by a J-domain co-chaperone protein network. Commun Biol 2, 323 (2019).

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI:

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing