Introduction

Antibiotics are the drugs of choice to treat bacterial infections. Because of the emergence of dangerous multi-drug resistant strains of bacteria, currently approved antibiotics are losing base against bacterial diseases1,2. Recent studies suggest that by 2050 a continuous rise in antibacterial resistance might result in the death of 10 million people per year3. Among pathogenic bacteria, ‘ESKAPE pathogens’ are of particular concern. Within this class of organisms, ‘KAPE’ represent gram-negative bacteria with increased susceptibility to attain multi-drug resistance4. Gram-negative bacteria are more threatening to humans due to the presence of the outer polysaccharide layer and multi-drug efflux transporters. This renders many small molecules including antibiotics eventually ineffective. This phenomenon has led to the creation of novel antibacterial agents distinct from the existing classes of molecules and their targets as well. A present interest is focused towards water soluble metal-based small molecular systems with organic moieties, which can cross the cell wall barrier and specifically target crucial bacterial molecules. Attention has been paid to coordination complexes because of their antimicrobial5,6, antifungal7,8,9 and anticancer activities10,11 applied in industry, ecology, or medicine. These complexes exhibited versatile electrochemical properties with tunable coordination chemistry which subsequently offer an enormous scope for the design of new entities. Cis-platin and its derivatives are well known examples which interact with DNA and human superoxide dismutase accounting for leads as both an antitumor drug and to fight amyotrophic lateral sclerosis12,13. Cell toxicity is well known for copper itself however its complexes often exhibit several fold better responses10,14,15,16. We previously showed the anticancer activity of copper(II) complexes exhibiting up to 100 fold better responses compared to Cu2+10,14.

One of the important functions of protein-folding helper proteins is to interact with aggregation-prone protein folding intermediates by recognizing non-natively exposed hydrophobic patches and shuffling disulfide bond (disulfide isomerases) or to accelerate peptidyl–prolyl cis–trans isomerization17,18. SlyD harbours both a peptidyl–prolyl isomerase and chaperone activity19,20 and is involved in several bacterial metabolic pathways including twin-arginine translocation (Tat) transport, biosynthesis of [NiFe] hydrogenases, and metal storage/release21. SlyD and its homologous proteins are present in all prokaryotes and archaea21,22. SlyD had first been characterized because of its PPIase and molecular chaperone activities23, which are located on two separate domains but mechanistically coupled21,24,25,26,27. Additionally, deletion of the slyD gene resulted in a marked reduction of the hydrogenase activity in cell extracts prepared from anaerobic Escherichia coli cultures28 and the slyD gene was classified as ‘Death Gene’ because it is involved in a pathway that leads to death of a majority of cells29. Moreover, being unique to prokaryotes and archaea; and essential for organism survival, these molecules could become interesting targets for developing small molecules which obstruct their activity inside the cells.

For several coordination compounds, antiprolifirative, antimicrobial, and antifungal activity had been identified by co-incubating them with cell cultures. However, their molecular mechanisms30,31,32 of cell death often remain elusive33. In structure based drug-design, to a large extent the proposed complexes hardly cross the cell wall/membrane barrier to reach their targets in the complex cellular environment, although they are quite effective binding-partners under in vitro conditions33. In our present studies, a coordination complex of Cu2+ and anthracenyl terpyridine (Fig. 1a) (hereafter refer as Cu2+ complex)10 has been found to inhibit bacterial growth by blocking the active side of SlyD (present in prokaryotes and archaea). We observed by in cell NMR that this promising antibacterial activity concomitantly occurs with Cu2+ complex binding to the SlyD. In vitro NMR spectroscopy showed that Cu2+ complex interacts with most of the residues forming the PPIase active side of the FKBP domain. These interactions inhibit both chaperone and PPIase activity of SlyD. We propose that these missing functions of SlyD led to a negative control of major metabolic pathways and hence a retardation of cell growth of E. coli, S. aureus (gram-positive) and P. aeruginosa (gram-negative) was observed.

Figure 1: Molecular systems and their effect on bacterial growth.
figure 1

(a) Small molecules used in the present study: (i) Cu2+ complex, (ii) Zn2+ complex, (iii) Cu2+ and (iv) Zn2+. (b) Growth curve analysis showing the role of SlyD in survival of bacteria against Cu2+ complex. Time dependent optical density of E. coli BL21(DE3) transformed with EcSlyD. The green arrow indicates the time point of IPTG induction (■ and curve) or addition of Cu2+ complex ( curve) and grey arrow indicated the addition of the Cu2+ complex ( and curve) or IPTG ( curve). (c) IC50 value determination for E. coli strain BL21(DE3): ■–E. coli strain BL21(DE3), –cells transformed with EcSlyD (non-induced), and –cells transformed with EcSlyD and induced with IPTG. In (b) and (c), error bars represent mean ± s.d. for triplicate experiments.

Full size image

Results

The Cu2+ complex inhibits bacterial growth

Metal ions cannot easily pass the cell wall of gram-negative bacteria such as E. coli34. Despite their complex composition (lipopolysaccharides, lipoproteins and peptidoglycans), however, the Cu2+ complex is able to enter inside of the E. coli (Figure S1). To evaluate the antibacterial property, growth curves were analyzed up to the stationary phase for the E. coli strain BL21(DE3). Complete bacterial growth inhibition was observed by incubation above a concentration of 2 μM of the Cu2+ complex. Lower concentrations (1.0 μM and 0.5 μM) inhibited only about 70–80% while beyond these concentrations cell growth was not effective (Figure S2).

Role of SlyD during cell growth inhibition

To evaluate the role of the molecular chaperone SlyD in the growth and maintenance of bacterial cells after treatment with the Cu2+ complex, growth curves were analyzed with cells over expressing E. coli SlyD (1–165) (EcSlyD). At an optical density of 0.5–0.6 of exponentially growing EcSlyD gene transformed E. coli BL21(DE3) cells ( and ■ symbols in Fig. 1b), SlyD over expression was induced by IPTG (green arrow in Fig. 1b). After 30 minutes of IPTG induction, cultures ( and ) were supplemented with 4 μM of the Cu2+ complex (grey arrow in Fig. 1b). After 2 h, no further increase in absorbance was observed in the non-induced bacteria ( in Fig. 1b). However, the IPTG-induced bacterial cells showed a significant survival against the Cu2+ complex induced toxicity ( in Fig. 1b). The latter cells could not grow as normal as the non-treated control due to the inhibitory role of the Cu2+ complex at 4 μM which could not be compensated by the presence of a high numbers of EcSlyD molecules. To show further the involvement of SlyD in growth and inhibition, the EcSlyD gene transformed E. coli BL21(DE3) cells were initially grown up to optical densities of 0.5–0.6 and then 4 μM of Cu complex was added (, indicated by green arrow) and further grown for 30 minutes. Thereafter the cells were induced with IPTG (indicated by the grey arrow). These cells showed better survival after Cu2+ complex treatment. Note that non-induced bacteria showed a similar growth curve as E. coli strain BL21(DE3) without EcSlyD transformation (Figure S3).

The in vitro antimicrobial activity of the Cu2+ complex in terms of half-maximal inhibitory concentration (IC50)31 was found to be 0.44 ± 0.02 μM and 0.55 ± 0.01 μM for E. coli strain BL21(DE3) and E. coli strain BL21(DE3) containing the EcSlyD plasmid but non-induced, respectively (Fig. 1c). These studies suggested the potential of the Cu2+ complex in inhibiting the bacterial growth. For the EcSlyD over expressing E. coli strain BL21(DE3) an IC50 value of 2.89 ± 0.09 μM was measured (Fig. 1c). This ~5 fold increased IC50 is in line with the increase in the optical density observed in the respective growth curves (Fig. 1b). Note, that free Cu2+ did not show significant bacterial inhibition at the concentrations studied herein (Figure S4). As a second control, we employed 9-anthracene methanol in the absence of Cu2+, which showed only 10–20% of inhibition at maximum concentration. Third, we replaced Cu2+ with Zn2+ in the coordination complex10,35. This molecule showed growth inhibition (10–20%) similar to 9-anthracene methanol (Figure S4). Together, these control experiments confirm that a Cu-metal centre in the 9-anthracene-terpyridine complex is required for efficient growth inhibitor.

Interactions of the Cu2+ complex with EcSlyD studied by in cell NMR

In cell NMR spectroscopy is a well suited method to investigate the conformation, stability and dynamics of proteins inside living cells36,37,38,39. The high complexity of the intracellular environment makes it difficult to analyze the interactions of small molecules with the protein of interest under physiological conditions due to low protein concentrations and a high background for NMR experiments. To investigate the interaction of EcSlyD with the Cu2+ complex inside E. coli BL21(DE3) cells, we recorded 2D 1H-15N HSQC spectra after IPTG induction. The induction level is high enough to detect the well-dispersed resonances of EcSlyD (black contours in Fig. 2a). Between 7.5 ppm and 8.5 ppm, backbone amide protons are overlapped by cellular background signals (Figure S5a). After incubation with the Cu2+ complex, several resonances either changed their chemical shifts or showed decreased intensities. These spectral changes confirm that the Cu2+ complex indeed can pass the cell wall of E. coli and interact with the molecular chaperone EcSlyD. To exclude that EcSlyD from outside the cells contributes to the spectra e.g. due to cell leakage, we recorded a 2D 1H-15N HSQC of the supernatant indicating no detectable signals (Figure S5b).

Figure 2
figure 2

(a) In cell NMR revealing the binding of the Cu2+ complex to EcSlyD inside BL21(DE3) cells. Overlaid sections of 2D 1H-15N HSQC spectra are shown before (black) and after addition of the Cu2+ complex. Enlarged boxes indicate examples of cross peaks showing decreased intensity, changes in chemical shifts and missing cross peaks upon Cu2+ complex interaction. (b) Inhibition of molecular chaperone activity of EcSlyD by Cu2+ complex. Symbols indicated the insulin aggregation. Conditions: ■–control experiment without EcSlyD, –in the presence of 40 μM EcSlyD; , and –inhibition of chaperone activity by Cu2+ complex at 40, 70 and 100 μM, respectively. (c) Inhibition of the PPIase activity of EcSlyD. PPIase activity was carried out at 25 μM () and 50 μM () of Cu2+ complex taking no inhibitor as a control (●). The data points were fitted according to the Michaelis-Menten equation.

Full size image

Inhibition of the molecular chaperone activity of EcSlyD

Next, we examined, whether the in cell NMR detected binding of the Cu2+ complex to EcSlyD relates to its chaperone and PPIase activity. One established assay for molecular chaperone activity is the suppression of aggregation of the insulin B chain after reduction by DTT (black and green symbols in Fig. 2b)26,40. This chaperone function could be inhibited by stoichiometric amounts of EcSlyD in a concentration-dependent fashion shown in Fig. 2b. This observation clearly suggests that the chaperone activity of SlyD can be controlled by the Cu2+ complex. Note, Cu2+ perchlorate, 9-anthracene methanol or the corresponding Zn2+ complex showed a negligible effect on inhibition of insulin aggregation under these conditions (Figure S6).

Inhibition of cis/trans peptide prolyl isomerase (PPIase) activity

The residues responsible for PPIase activity in SlyD reside in the FKBP domain26,41. The catalytic PPIase activity of EcSlyD was examined with a standard assay, which is the catalysis of refolding of the reduced and carboxymethylated form of the S54G/P55N variant of ribonuclease T1 (RCM-T1), which is limited by a slow trans to cis isomerization reaction of Pro3922. EcSlyD showed a Km value of 1.4 ± 0.4 μM (black symbols in Fig. 2c) which is in good agreement with the literature value22. In the presence of 25 μM and 50 μM of the Cu2+ complex this PPIase assay suggests non-competitive inhibition of EcSlyD as judged by the corresponding Michaelis–Menten plots (Fig. 2c). Data analysis yields a Ki value of 23.6 μM.

Binding of the Cu2+ complex to EcSlyD by fluorescence spectroscopy

To gain insights into the molecular details of the EcSlyD inhibition, we performed a fluorescence titration experiment with the Cu2+ complex (Fig. 3a). The intrinsic EcSlyD fluorescence intensity dominated by Tyr 34 and Tyr 68 in the FKBP domain and Tyr 71 in the linker between FKBP and IF domain. Intensity dropped as a function of the Cu2+ complex concentration, resulting in a dissociation constant (KD) of 45 μM. We found a 2:1 stoichiometry and assumed the same affinity for two binding sites at EcSlyD. In contrast, Cu2+ or 9-anthracene methanol did not result in a significant decrease in the fluorescence intensity (Fig. 3a). Additionally, we studied ∆IF EcSlyD, which showed a KD of 63.65 ± 4.5 μM for the Cu2+ complex, indicating that the IF domain only marginally modifies the affinity to the inhibitor. Neither the Cu2+ complex nor the latter two compounds changed the far UV-CD spectrum of EcSlyD indicating that the structural integrity of the protein remains intact (Fig. 3b).

Figure 3: Binding studies of the Cu2+ complex with EcSlyD.
figure 3

(a) Relative fluorescence intensity plots of EcSlyD analysed at 335 nm during titration with ■–Cu2+ complex, –Cu2+. –9-anthracene methanol and –∆IF SlyD vs Cu2+ complex. (b) CD spectrum of EcSlyD in the presence and absence of Cu2+ complex (10 μM to 50 μM). The black spectrum corresponds to free EcSlyD. High absorption beyond 215 nm did not allow useful spectra to be obtained upon addition of the Cu2+ complex.

Full size image

Residue specific interaction of the Cu2+ complex with EcSlyD

In vitro NMR spectroscopy42,43 can be applied to probe the binding of the Cu2+ complex to EcSlyD at residue-level resolution. The applied 2D 1H−15N-HSQC experiment solely detects nitrogen-bound protons of 15N-EcSlyD. At a stoichiometry of 1:2 of EcSlyD versus the Cu2+ complex we analysed differences in the chemical shifts and resonance intensities of individual residues. Several resonances were broadened beyond the detection limits, while other residues changed their chemical shift positions (Fig. 4a). Mapping these properties onto the structure shows that significant changes are at the FKBP domain and few at the IF domain (Fig. 4b). EcSlyD residues Y13, V23, D24, L32, Y34, S40, L41, I42, L45, Y68, L130, and F132 are important for the PPIase activity of the FKBP domain26. Interestingly, the intensity of all of these residues decreased upon addition of the Cu2+ complex (Fig. 4c) with only one exception (L130). This confirms the observed inhibition of PPIase activity by the Cu2+ complex, because it binds to the active site of EcSlyD. Similarly, the residues of the IF domain were analysed. Upon Cu2+ complex addition the intensities of Leu75, Val76, Thr100, Gln102, Gly103, Val105, and Gly121 were strongly affected, while those of Ala98 and Val119 were moderately affected (Fig. 4b). These results are in good agreement with the inhibition of the chaperone activity by the Cu2+ complex.

Figure 4: In vitro NMR spectroscopic analysis of the Cu2+ complex binding to EcSlyD.
figure 4

(a) 2D 1H-15N HSQC overlaid spectra of 15N EcSlyD in the free (black) and bound state (red). Red and blue labels are the residues involved in the PPIase activity. (b) Mapping of cross peak intensities on EcSlyD structure. Residues with missing NMR resonances are marked in red, 20–50% loss in intensities are indicated in blue and no significant change in grey. (c) Close up view of the PPIase active sites. Key residues are labeled by their side chains using the same color coding as in (b).

Full size image

In order to study the binding affinity of the Cu2+ complex to the individual EcSlyD domains, the NMR intensities of titrating residues were analysed (Figure S9). Residues of the FKBP domain yielded an averaged KD value of 87 ± 30 μM and for the IF domain we found 105 ± 31 μM. The values match well with the affinity obtained with fluorescence spectroscopy. Cu2+ perchlorate and 9-anthracene methanol did not show interactions with these residues of the FKBP and IF domain corresponding to the respective PPIase and chaperone function (Figure S8).

Toxicity of the Cu2+ complex for pathogenic bacterium

We tested the toxicity of the Cu2+ complex against the pathogenic bacteria Staphylococcus aureus (gram-positive) and Pseudomonas aeruginosa (gram-negative) as chosen from the ‘ESKAPE pathogens’ category4. The IC50 was found to be 0.90 ± 0.1 μM and 0.81 ± 0.07 μM for Staphylococcus aureus and Pseudomonas aeruginosa, respectively (Fig. 5). The results indicate that the Cu2+ complex is effective even for pathogenic bacteria. The obtained IC50 values are close to the findings with E. coli. Note again, that Cu2+ perchlorate, 9-anthracene methanol, and the Zn2+ complex have a negligible effect on bacterial growth (Figure S10).

Figure 5: Growth analysis of pathogenic bacteria in presence of the Cu2+ complex.
figure 5

IC50 value were determined for Staphylococcus aureus () and Pseudomonas aeruginosa (■). Error bars represent mean ± s.d. for triplicate experiments.

Full size image

Discussion

Molecular chaperones are key molecules for cell growth and maintenance44. SlyD is a metallo-chaperone21 in prokaryotes and archaea additionally harbouring a PPIase domain22 and a C-terminal Ni2+ binding domain to deliver these metal ions during [NiFe]-hydrogenase maturation45. Furthermore, SlyD guides proteins to the Tat-dependent translocation of folded proteins across the cytosolic membrane of bacteria46. Targeting such crucial functions to inhibit bacterial growth might be a new strategy to counteract the emergence of multi-drug resistance. Coordination complexes are gaining interest in this respect for medical applications11,47. Although known coordination complexes exhibit promising antibacterial activities, their mechanisms of action are less explored. Recent examples include the complexes of [CuX2(INH)2]·nH2O (X = Cl and n = 1; X = NCS and n = 5; X = NCO and n = 4; INH = isoniazid), which have been optimized for in vitro antibacterial activity48. The compound N-(2-hydroxybenzylidene)-1-ethyl-1, 4-dihydro-7-methyl-4-oxo-1, and 8 naphthyridine-3-carbohydrazide and their Cu(II), Co(II) and Zn(II) complexes were synthesized and characterized and their antibacterial activities have been tested under in vitro conditions49. Cu(II) complexes with the chemotherapeutic antibacterial norfloxacin and N-donor mixed-ligands have been recently reported50.

For the Cu2+ complex studied here, we found at low micromolar concentrations antibacterial activity after readily crossing the bacterial cell wall probably because of the hydrophobic character of the anthracenyl terpyidine moiety. In vivo NMR revealed that EcSlyD is involved in this inhibition (Fig. 1). Overexpression of EcSlyD cannot fully overcome the inhibitory function of the Cu2+ complex indicating that this complex might influence additional cellular processes (Fig. 1 and Figures S2 and S3).

Few molecular chaperones such as heat shock proteins (HSPs) have been proposed as drug targets in cancer therapy51. More recently, Kondoh and Osada performed high throughput screens to identify molecular chaperones for the same purpose52. Geldanamycin and radicicol have been identified as inhibitors of HSP9053. Other chaperones such as HSP60 or HSP70 have received much less attention as a potential drug target51. To the best of our knowledge bacterial chaperones have not been reported in this context so far. It is well known that antibacterial small molecules can interfere with cell wall synthesis (e.g., penicllins), cell membrane function (e.g., polymixin B), protein synthesis (e.g., aminoglycosides) and nucleic acid synthesis (e.g., quinolones)54.

The here presented Cu2+ complex readily crosses the bacterial cell wall which might be of advantage in terms of drug resistance e.g. by efflux-mediated extrusion. Nano-particles have recently emerged in this area55,56,57 because of their slightly larger diameters compared to the drug efflux pumps58,59. Vancomycin-capped gold nano-particles enhanced the in vitro antibacterial activities of vancomycin-resistant Enterococcus and E. coli strains60 as well as S. aureus61 without developing resistance62. Graphene-based nano-particles have been developed which kills gram-positive and negative bacteria upon near-infrared laser irradiation via the cross linking property63.

SlyD is a metallo-chaperone and the full-length protein can bind up to 7 Ni2+ ions with a histidine rich C-terminal extension of the PPIase and IF domain pair64. The EcSlyD variant in the present study contains one Ni2+ binding site (His149, His151 and His153), which has been characterized by NMR spectroscopy45. The paramagnetic metal ion enhances the relaxation rates of NMR active nuclei in close proximity (paramagnetic relaxation enhancement (PRE) effect) resulting in a line broadening often below the detection limit. In the EcSlyD/Ni2+ complex studied earlier, residues mainly following E141 at the C-terminus and residues close to Y34 in the PPIase domain show strong PRE45. An almost identical pattern was observed for the EcSlyD complex with Cu2+ as the metal ion (Fig. S8c). The PRE pattern caused by the Cu2+ complex strongly differs from the pattern of single metal ions (Fig. 4a). Many more resonances are affected, which correspond to residues of the active sites of the PPIase and IF domains. As 9-anthracene methanol causes no significant changes in the NMR spectra of EcSlyD (Figure S8a), we can conclude that only the Cu2+ coordination complex can effectively block the PPIase and chaperone function of EcSlyD. We expect minor conformational changes of EcSlyD to happen upon inhibitor binding, because none of the detectable resonances (Fig. 4a) show significant chemical shift changes. This is in agreement with the CD analysis of free and bound EcSlyD (Figs 3a and S7).

In vitro studies of EcSlyD substantiated that PPIase and chaperone function is located at two different domains and an interplay of both domains is required for full PPIase activity19,20. The isolated IF domain of Thermos thermophilus SlyD (TtSlyD) still has chaperone activity whereas the ΔIF variant lacks this function65. Excision of the IF domain from TtSlyD led to a 100-fold decrease in PPIase activity towards protein substrates65 and a complete loss of activity for EcSlyD66. The Cu2+ complex studied here binds EcSlyD with a stoichiometry of 2:1 and inhibits both the chaperone and the PPIase activities (Fig. 2). NMR chemical shift mapping of complexed EcSlyD (Fig. 4) revealed binding to both the active site of the PPIase domain26 and the IF domain41,67 with about the same affinity. The majority of mapped contacts are hydrophobic residues distant from the metal binding site at the C-terminus. Therefore, we propose that the anthracenyl terpyridine moiety of the Cu2+ complex facilitates binding of the metal not to the C-terminus of EcSlyD but to both active sites by hydrophobic interactions to block its function.

In summary, it is expected that inhibition of cellular chaperone and PPIase activities will affect proper protein folding and functioning of their various substrates and thus bacterial proteostasis, which causes retardation of cell growth. Besides its chaperone and PPIase function, SlyD is known to be involved in other biochemical reactions, which are further cellular functions affected by SlyD inhibition. SlyD is present in both gram-positive and -negative bacteria. We demonstrated that both representatives of both families of pathogenic bacteria (S. aureus (gram-positive) and P. aeruginosa (gram-negative)) exhibited similar growth inhibition (IC50) as E. coli in the presence of the Cu2+ complex. Although we observed well documented binding to EcSlyD and its functional consequences, we cannot rule out that the Cu2+ complex interacts with additional cellular proteins or DNA68 which cannot be completely revealed by the presented in-cell NMR experiment and need to be identified in future experiments. Targeting a molecular chaperone is a feasible way towards the generation of new antibacterial molecules, which is supported by our findings at a proof-of-principle stage.

Methods

Growth curve analysis

E. coli BL21(DE3) cells were grown in the double YT (dYT) media in a 50 ml of culture volume. For the growth curve analysis, a starting OD600nm of 0.05 was taken by adding aliquots of the freshly grown culture at the log phase. Growth inhibition was analyzed by adding the 0–6 μM of the Cu2+ complex. These culture flasks were incubates at 37 °C with rotation at 300 rpm. Aliquots of growing culture were taken at different time point. Their growths were monitored by OD 600nm. Aliquots were taken out until the OD 600nm of the control experiment without Cu2+ complex reached to the stationary phase.

In order to understand the role of SlyD in bacterial growth survival against the Cu2+ complex, E. coli BL21(DE3) cells were transformed with the EcSlyD plasmid corresponding to the amino acid sequence (1–165). Growth curves in case of transformed bacteria were analyzed in the similar way as stated above. The EcSlyD over production was initiated by inducing the culture with 1 mM of IPTG. The transformed E. coli BL21(DE3) cells were grown in presence of 30 μg/ml kanamycin.

In our experiment we have chosen SlyD (residues 1–165) of E. coli, a truncated version of full-length SlyD lacking the unstructured C terminal tail. Removing this tail does not affect the chaperone and PPIase functions of SlyD21,26. The Cu2+/Zn2+ complex or 9-anthracene methanol was dissolved in dimethyl sulfoxide (DMSO) and further diluted for the respective experiments. The final concentration of DMSO was <1% in any experiment.

IC50 value determination

For the IC50 value determination, freshly grown E. coli BL21(DE3) cells or EcSlyD transformed E. coli BL21(DE3) cells (non-induced with IPTG) at their log phase were used. The 10 ml of culture media was taken and the Cu2+ complex with concentrations ranging from 0–6 μM were added in a separate tube. A starting OD600nm of 0.05 was used for this experiment. In a second experiment, EcSlyD transformed cells were grown until OD600nm of 0.6–0.7. This culture was then induced by adding 1 mM of IPTG. The culture was further grown for 3 h so that the cells had a sufficient number of EcSlyD copies. These bacteria were used for a IC50 value determination as described for non-induced E. coli BL21(DE3) cells. All these cultures were incubated for 10 h at 37 °C with rotation at 300 rpm. OD600nm of the bacteria without the Cu2+ complex was considered as 100% bacterial growth. The transformed E. coli BL21(DE3) cells were grown in presence of 30 μg/ml kanamycin. The experiments were performed in triplicate and repeated at least two times.

Pathogenic bacteria, Staphylococcus aureus and Pseudomonas aeruginosa were grown in nutrient broth. Their IC50 value was determined by co-incubating them with Cu2+ complex concentrations ranging from 0–8 μM. The growth of bacteria was monitored at OD600nm as described above.

In cell NMR

The in cell NMR experiment was performed as described earlier69. In brief EcSlyD transformed E. coli BL21(DE3) cells were grown in M9 minimal media containing 30 μg/ml kanamycin until an OD600nm of 0.6–0.7 was reached. These cells were induced with the 1 mM of IPTG and further incubated for 3 h at 37 °C with 300 rpm rotation. These cells were gently harvested by centrifugation at 5000 rpm at 4 °C. The cells were then washed 3 times with 50 mM Na2HPO4, 100 mM NaCl, pH 7.5. A 25% cells slurry with 10% D2O in a Shigemi tube was used in order to minimize the effect of settling of the bacteria. This culture was considered the control experiment. A 2D 1H-15N HSQC spectrum was recorded for 2 h at 25 °C. Thereafter, 100 μM of Cu2+ complex was added to this culture and incubated for 1 h at room temperature. After a short mixing period this slurry was transferred to the NMR tubes and a 2D 1H-15N HSQC spectrum was recorded for 2 h. In order to see if cell lysis might have occurred during the course of time and various treatments, the slurry was taken out of the NMR tube and centrifuged at 10000 rpm at 4 °C for 10 minutes. The supernatant was taken out and a 2D 1H-15N HSQC spectrum was recorded for 2 h.

EcSlyD expression and purification

For the expression and purification, the plasmid containing the gene corresponding to SlyD (1–165) of E. coli was transformed into the E. coli BL21(DE3) cells. The protein expression was carried out at 37 °C with IPTG induction. Purification of EcSlyD was performed under denaturing conditions (GdmCl) as described earlier22.

Molecular chaperone activity (insulin aggregation assay)

The insulin aggregation assay was performed as previously described22,26. In brief, insulin was dissolved in 6 M HCl and the pH was adjusted to 2.6 with NaOH. The aggregation assay was started at a concentration of 30 μM insulin (in 50 mM Na2HPO4, 100 mM NaCl, pH 7.5) with the addition of 10 mM DTT. Aggregation was retarded by addition of 40 μM EcSlyD and the Cu2+ complex at concentrations of 0, 40, 70 and 100 μM. The reductive cleavage of the insulin chain B aggregation was monitored by increase in scattered light at 400 nm. The reaction was followed in a Jasco FP-6500 fluorescence spectrometer at 25 °C.

PPIase activity

PPIase activity was analyzed as described earlier22. In brief, the catalytic efficiency of EcSlyD was examined as a catalyst of the refolding of the reduced and carboxymethylated form of the S54G/P55N double mutant of ribonuclease T1 (RCM-T1). This reaction is limited by the slow trans to cis isomerization of Pro39. The reaction was carried out at 0–17 μM of RCM-T1 and a fixed concentration of 20 nM EcSlyD. Enzyme inhibition was carried out at 0, 25, and 50 μM of Cu2+ complex. The experiments were carried in 50 mM Na2HPO4, 100 mM NaCl, pH 7.5 at 25 °C. The data was analysed by the Michaelis–Menten formalism with the enzyme kinetic module of Sigma Plot 12.

Fluorescence binding

For the fluorescence studies, a solution of 5 μM of EcSlyD was used for binding studies. In a typical titration experiment, 4 μM of Cu2+ complex was added at each step and the spectrum was recorded after a short mixing period. Titration of Cu2+ perchlorate or 9-anthracene methanol was also carried out in the same way. A blank titration (without protein) was recorded, where the Cu2+ complex or Cu2+ perchlorate or 9-anthracene methanol was added. This blank titration was subtracted from the main experiment before the analysis. The titration experiment was carried in 50 mM Na2HPO4, 100 mM NaCl, pH 7.5. The whole solution was excited at 280 nm on a Jasco FP-6500 fluorescence spectrometer at a temperature of 25 °C.

Fluorescence data were analyzed according to the following equation:

with P = (P0/(V0 + L))V0 and L1 = (L0/(V0 + L))L and Q–fluorescence intensity, P–protein concentration, L–ligand concentration, V–volume, n–binding sites and 0–indicates start point.

CD spectroscopy

For the CD spectroscopic studies a solution of 10 μM of EcSlyD was used. In typical titration experiments 10 μM of the Cu2+ complex was added. After a short mixing period, a spectrum was recorded after each titration step. Titration of Cu2+ perchlorate or 9-anthracene methanol was carried out at the same protein concentration. Cu2+ complex or Cu2+ perchlorate or 9-anthracene methanol titrated against buffer was used as blank. This blank titration was subtracted from the main experiment before the analysis. The titration experiment was carried in 50 mM Na2HPO4, 100 mM NaCl, pH 7.5. The spectrum was recorded on Jasco J815 spectrometer at a temperature of 25 °C.

In vitro NMR spectroscopy

The spectra were recorded at a concentration of 60 μM of EcSlyD. NMR titration experiments were carried out by the addition of 1 μl of 10 mM Cu2+ complex until there was no further change in NMR intensity observed. The spectrum corresponding to the 1:2 complex formations was used for the analysis. The NMR titration with Cu2+ perchlorate or 9-anthracene methanol was carried out in the same way. The spectra were recorded in 50 mM Na2HPO4, 100 mM NaCl, pH 7.5, with a Bruker 800 MHz Avance III spectrometer equipped with a CP–TCI cryoprobe at 25 °C. Spectra were processed using the programs NMRPipe and NMR Draw. The previously determined NMR assignments of the EcSlyD resonances determined previously were used for analysis26.

Additional Information

How to cite this article: Kumar, A. and Balbach, J. Targeting the molecular chaperone SlyD to inhibit bacterial growth with a small molecule. Sci. Rep. 7, 42141; doi: 10.1038/srep42141 (2017).

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