Proteins are dynamic in nature and exist in a set of equilibrium conformations on various timescale motions. The flexibility of proteins governs various biological functions, and therefore elucidation of such functional dynamics is essential. In this context, we have studied the structure–dynamics–stability–activity relationship of bacteriophage T7 lysozyme/endolysin (T7L) native-state ensemble in the pH range of 6–8. Our studies established that T7L native state is conformationally heterogeneous, as several residues of its C-terminal half are present in two conformations (major and minor) in the slow exchange time scale of nuclear magnetic resonance (NMR). Structural and dynamic studies suggested that the residues belonging to minor conformations do exhibit native-like structural and dynamic features. Furthermore, the NMR relaxation experiments unraveled that the native state is highly dynamic and the dynamic behavior is regulated by the pH, as the pH 6 conformation exhibited enhanced dynamics compared with pH 7 and 8. The stability measurements and cell-based activity studies on T7L indicated that the native protein at pH 6 is ∼2 kcal less stable and is ∼50% less active than those of pH 7 and 8. A comprehensive analysis of the T7L active site, unfolding initiation sites and the residues with altered dynamics outlined that the attenuation of stability and activity is a resultant of its enhanced dynamic properties, which, in turn, can be attributed to the protonation/deprotonation of its partially buried His residues. Our study on T7L structure–dynamics–activity paradigm could assist in engineering novel amidase-based endolysins with enhanced activity and stability over a broad pH range.
To accomplish biological functions, proteins must conserve their native state with a considerable degree of structural flexibility. The stability of native state is governed by the surrounding environmental factors, such as pH, temperature, solvent and ionic strength, due to their effect on inter- and intramolecular interactions that are crucial for the stability and integrity of proteins [1–5]. Failure of protein to survive in such environmental conditions produces biological inactivation and/or misfolding of proteins that can be associated with various fatal diseases [6,7]. The native state occupies the lowest energy level down the folding funnel and comprises an ensemble of conformations existing in dynamic equilibrium under various physiological conditions [8–10]. The ensemble also involves a small population of fluctuating low-lying energy conformations that are very close to the free energy minima [11–13]. These multiple or alternative conformations have an effect on protein function, flexibility, allosteric regulations and folding [14–17]. Therefore, it is of great interest to identify that how these multiple conformations lead to diverse functions [14,16].
Nuclear magnetic resonance (NMR) spectroscopy is a unique technique to characterize low energy states/alternative conformations of proteins at an atomic resolution [18–21]. Furthermore, NMR is very much sensitive to analyze slow conformational exchange and interconversion of multiple conformations that are more rigorously inferred using NMR spin relaxations [22,23]. Identification and characterization of such low energy states under mild external perturbation, such as pH, urea and temperature, unravel the native-state heterogeneity [24–27]. Furthermore, proteins are inherently dynamic which is responsible for the existence of time-dependent structures, rather than occupying a single conformation [28,29]. Elucidating the dynamic properties of a protein is important to understand the structural basis of its functional behavior . Indeed, proteins are dynamic over a wide range of timescales [28,31,32], but determining the number of distinct dynamic processes and identifying functionally relevant conformations are still challenging.
Classically, bacteriophage endolysins are the peptidoglycan degrading enzymes that lyse bacterial cell wall at the end of bacteriophage replication to release the progeny of virons [33–35]. Recently, these endolysins were recognized as ‘enzybiotics’ due to their antimicrobial characteristics and have found various applications in the fields of agriculture, food safety, medicine and biotechnology [36–40]. The bacteriophage endolysins are broadly classified into four categories: (a) N-acetylmuramidase; (b) transglycosylase; (c) endopeptidase and (d) N-acetylmuramoyl-l-alanine . Among these, T7 endolysin or T7 lysozyme (T7L) encoded by the class-II gp3.5 gene of T7 bacteriophage is an N-acetylmuramoyl-l-alanine amidase, which cleaves an amide bond between the N-acetylmuramic acid (NAM) and l-alanine of stem peptide of Gram-negative peptidoglycan layer [42,43]. T7L is a 17 kDa (151 amino acids) single-domain α/β-fold structured protein . T7L and other members of the amidase family show significant pH-dependent lysis activity characteristics [45–48]. This class of endolysins is highly active in the pH range of 6.5–8.5 and their activity drastically decreases below pH 6. In recent work, our group established that below pH 6, T7 endolysin exhibits a structural transition, and it forms partially collapsed conformations in the range of pH 3–5. Such a reversible pH-dependent structural transition is responsible for the loss of activity of amidases at low pH values .
The current study deals with the identification of T7L native-state heterogeneity and elucidation of dynamic regulations using NMR chemical shifts, temperature coefficients and relaxation dynamics. Our results establish that T7L native state exhibits significant conformational dynamics in response to a mild change in pH without changing its secondary and tertiary structures. Furthermore, biophysical and amidase activity assays suggested that the lysis activity and structural stability of T7L are highly regulated by its dynamics. Such a detailed characterization of the multiple conformers of T7L native state is important for understanding its structure–function relationship. Furthermore, this work could be of great interest to evaluate the structural and dynamic properties of native sate ensemble of other single-domain amidase proteins of bacteriophage origin.
Materials and methods
Protein expression and purification
Circular dichroism spectroscopy
Size exclusion chromatography
T7L protein samples of 70 µM at pH 6, 7 and 8 were loaded separately onto a column pre-equilibrated with buffers of respective pH values (20 mM phosphate, 0.1 M NaCl, 5 mM EDTA and 1% glycerol). Size exclusion chromatography (SEC) experiments were performed using HiLoad 16/60 Superdex 75 prep grade column (GE Healthcare) on an AKTA prime FPLC system (GE Healthcare) with a flow rate of 1 ml/min. The total volume (Vt) and excluded volume (Vo) of the column were 120 and 45 ml, respectively.
In vitro enzymatic assay
The amidase activity of T7L native states at pH 6, 7 and 8 was measured using an absorbance-based turbidimetric assay and a green fluorescence protein (GFP)-based fluorimetric assay.
For turbidimetric assay, the Escherichia coli BL21 (DE3) cells were grown at 37°C and 220 rpm to an OD600 nm of 0.6–0.7 and sensitized with 0.1 M EDTA for 5 min. The cells were centrifuged at 6700×g for 1 min and pellets were resuspended in 10 mM potassium phosphate buffer (pH 6–8) containing 10 µM ZnSO4. After adding 50 µg of untagged T7L, decrease in the optical density (at 600 nm) of sensitized E. coli cells was monitored to measure the lysis efficiency of the protein at different pH values. The suspension of untreated E. coli cells was used as a control. All the measurements were performed in triplicate at 25°C for a period of 0–12 min, with 2 min time interval.
Fluorimetric assay was performed using E. coli BL21 (DE3) cells containing an inducible GFP plasmid. GFP was expressed by growing cells at 37°C until it reached its exponential phase (OD600 nm of ∼0.6), followed by IPTG (0.5 mM) induction and incubation at 37°C for 2 h. The cells were sensitized as described above and incubated them with T7L at 25°C for 10 min. Supernatants of lysed cells were recovered for fluorescence spectroscopic analysis, and the extent of lysis of E. coli cells was measured by monitoring the intensity changes of GFP fluorescence emission in the wavelength range of 490–650 nm using a Hitachi F-4600 fluorescence spectrophotometer. The experiments were performed in triplicate and the untreated GFP expressing cell suspension was used to normalize the emission profiles.
2D 1H-15N heteronuclear single quantum coherence (HSQC) experiments were recorded at pH 6, 7 and 8 at 25°C using 15N labeled ∼250–500 µM T7L. The spectral widths in the 1H and 15N dimensions were set as 16 and 32 ppm, respectively. All the triple-resonance, relaxation and temperature-dependent experiments of native state T7L were performed in 20 mM phosphate buffer containing 0.1 M NaCl, 5 mM EDTA and 10% D2O.
For urea-based T7L 1H-15N HSQC experiments, 10.5 M stock solutions of urea were prepared at pH 6 and 7 and then added to 250 µM lyophilized T7L. The lyophilized protein samples were diluted with respective buffers using urea stock solution to give final urea concentrations from 0 to 9.5 M. All the HSQC experiments were recorded on Bruker 500 MHz instrument, and the NMR data obtained were processed using Bruker software TOPSPIN version 3.2.
Backbone resonance assignment
For the backbone resonance assignment of multiple conformations of T7L native state, three-dimensional NMR experiments were carried out on Bruker Avance III 800 MHz instrument equipped with a TXI cryoprobe, pulse shaping and pulse field gradient capabilities. A set of triple-resonance experiments, such as HNCO, HNCA, HNCACB and CBCA(CO)NH, were recorded at 25°C using 1 mM T7L at pH 7. The 13C carrier frequency was set at 54 ppm for HNCA, 42 ppm for HNCACB and CBCA(CO)NH and 174 ppm for HNCO . The assignment was achieved by analyzing the resultant spectra using CARA . Secondary chemical shifts (Δδ) of the Cα carbons were calculated based on the full spectral assignments at pH 7, 25°C and the random coil database .
The 15N transverse relaxation experiments (T2) and steady-state 1H-15N heteronuclear NOE (het-NOE) measurements of ∼250 µM T7L were performed at pH 6, 7 and 8 using the pulse sequences described by Peng and Wagner, and Farrow et al. [55,56]. For T2 (R2) measurements, the following relaxation delays were used: 18.56, 37.12, 55.68, 74.24, 92.8, 111.36 and 129.92 ms. Het-NOE experiments were performed using a proton saturation time of 3.0 s and a relaxation delay of 3.0 s. Steady-state 1H-15N NOEs were calculated as a ratio of intensities of the peaks with and without proton saturation. The errors in the NOEs were obtained as described by Farrow et al. . Relaxation data (R2 and NOE) analysis was performed using Sigma Plot 12.0 software.
Amide proton temperature coefficient
The temperature coefficients of amide protons (NH) of T7L (∼250 µM) at pH 7 were calculated based on temperature-dependent 1H-15N HSQC spectra. The data were collected in the temperature range from 283 to 318 K at regular intervals of 5 K. All the temperature measurements were limited to 318 K to avoid the formation of precipitate/aggregation. Amide proton temperature coefficients (ΔδHN/ΔT) were calculated by fitting the proton chemical shift changes to a linear equation using Sigma Plot 12.0 software [57,58].
Existence of multiple conformations in native-state ensemble of T7L
The backbone resonance assignment of T7L native state at pH 7 was reported earlier by our research group . After assigning most of the available sequential connectivities in the HSQC spectrum of T7L native state at pH 7, there are still several resonances present which do not correspond to any new stretch or missing/unassigned residues. Such resonances draw a serious attention towards the assignment/sequence-specific location of those residues. During the course of NMR analysis, we noted that a substantial set of low intensity peaks (∼25 peaks) closer to the major amide resonances are present in the 1H-15N HSQC spectrum (Figure 1A). Indeed, it is evident from the HSQC spectrum that the intensities of the NH resonances are not uniform indicating an inherent conformationally dynamic nature. The backbone resonance assignment enabled us to establish the sequential connectivities for these minor cross-peaks. It was found that majority of these peaks belong to some of the already assigned residues and corresponded to a major and a minor conformation. The resonance assignments of these alternative stretches of T7L clearly establish that some of the protein segments are accessing the multiple conformations in the slow exchange regime of the NMR timescale (Figure 1A). The alternative sequential connectivity of residues, G117, A118 and V119 which are present in two conformations, is shown in Figure 1B using HNCACB strip plots for the stretch of E116–L120. A total of 19 peaks/resonances were assigned to be the minor conformation peaks at pH 7. The assignment summary and the structural localization of multiple conformation accessing residues are shown in Figure 1C,D. The chemical shifts of the minor and major conformations for all the assigned 19 residues are listed in Supplementary Table S1. Residues exhibiting the multiple conformations belong to four different regions of the polypeptide chain (Figure 1C,D). Seven residues are located in α2-helix, two residues in β5-sheet and three residues are located in α3-helix region, followed with five residues in C-terminal loop. The data clearly suggest that C-terminal segments contribute extensively to the structural heterogeneity of T7L in its native conformation.
Backbone resonance assignment of T7 endolysin minor conformations at pH 7 and 25°C.
Native-like structural preferences of minor conformations
It is essential to depict the structural characteristics of these assigned minor confirmations as they can display native or non-native secondary structural features. To elucidate the structural preferences of T7L minor conformations, we have calculated the secondary chemical shifts assessed from Cα secondary chemical shifts and compared with those of corresponding major conformations (Figure 2A). The positive values of 13Cα indicate population of α-helical conformations, while the negative values represent β-sheet structural propensity [59,60]. The minor conformations of some residues, such as Q100, M101, Q102, S103, L108, T110, K114, K137, R138 and W139, have propensities of α-helix; G117, A118 and V119 have the propensity of β-sheet. Rest of all the minor conformers displayed random coil-like propensities. The secondary structural preferences of these minor conformations suggest that they do exhibit the native-state-like conformations as the native state consists of α2-helix from residues A99 to E116, α3-helix from residues L136 to W140 and β5-sheet from A118 to A122 residues.
Structural analysis of T7L multiple conformations.
To further substantiate the native-like features of the minor conformations, temperature-dependent amide proton chemical shift analysis was performed (Figure 2B,C). The amide proton chemical shifts vary linearly over the small range of temperature changes and the value of temperature coefficient more positive than −4.5 ppb/K is engaged with strong H-bonding [57,58]. It was observed that amide protons of minor confirmations experience similar linear temperature-dependent profile as shown by major confirmations. All the measured temperature coefficients are shown in Figure 2D and listed in Supplementary Table S2. Most of the minor conformation residues belonging to the α2-helix show NH temperature coefficients below −4.5 ppb/K, thus establishing that these residues are involved in H-bonding and are present in the structural elements. The observed patterns are consistent with the structural view offered by the secondary chemical shifts. Henceforth, these combined results suggest that these minor conformations have structural characteristics that are similar to those of the native conformation.
Concentration-dependent population analysis of multiple conformations
NMR studies, as a function of temperature/concentration over a wide range, have been used to characterize the multiple conformations of individual residues . Therefore, to know the behavior of T7L minor states, we have performed concentration-dependent quantification of the relative populations of the major and minor conformations using 250 µM and 1 mM protein samples, respectively. The intensity plot of the major and minor conformations indicated that changing of the protein concentration modulates the ratio of their populations (Figure 2E). It has been observed that the ratio of the major to minor conformations at 1 mM is ∼2–3 fold higher than those at 250 µM. It suggests that increasing the concentration is associated with a decrease in minor conformer population. Altogether, these results suggest that the major and minor conformers are under conformational equilibrium.
pH-dependent conformational heterogeneity of T7L
To assess the heterogeneity of T7L in native ensemble, biophysical and NMR-based experiments were performed at different pH (6–8). SEC experiments, secondary and tertiary CD features of T7L suggested that the protein exhibits similar structural features in the chosen pH values (Supplementary Figure S1A–C). As 1H and 15N chemical shifts are sensitive probes for the environment of a given residue in a protein structure, we have recorded the 1H-15N HSQC in the pH range of 6–8 to further confirm the similar structural features. It is evident from the HSQC spectra of both pH 6 and 8 that T7L shows similar spectral features as that of pH 7 in terms of chemical shift dispersion along both 1H and 15N dimensions (Supplementary Figure S2).
As the 1H-15N HSQC spectra at these two pH values are very similar to pH 7, the transfer of assignment is readily possible. The assignments at pH 6 and 8 were obtained by direct transfer of assignments using pH 7 assignment. The number of peaks present in the HSQC spectra at pH 6, 7 and 8 is 120, 140 and 140, respectively. The pH 6 state experiences the broadening of several residues due to enhanced dynamics (discussed in the next section). The broadened residues at pH 6 are marked in circles in Figure 3A and Supplementary Figure S2.
Comparison of multiple conformations in the native-state ensemble.
Furthermore, identification of the residues having conformational heterogeneity in the native states at pH 6, 7 and 8 provided a conscious view of the landscape of the native-state ensemble. The HSQC spectra revealed that all the major and minor conformation resonances at pH 7 are also present at pH 8 (Figure 3A,B and Supplementary Figure S2). But, two situations were observed in the case of pH 6; (a) resonances corresponding to both the major and minor conformations completely disappeared, which includes T97, L145 and V146 (black circles) and, (b) residues are intact in their major conformations, while the minor conformations were vanished; these residues include V119, K137, E144 and D149 (sky blue circles) (Figure 3A). The locations of these residues are marked on the sequence of T7L (Figure 3B).
Conformational dynamics of heterogeneous T7L native state
NMR relaxation experiments are widely used to assess the dynamics of proteins over a magnitude of timescales. To determine the backbone dynamics conformationally heterogeneous T7L native state, we have performed transverse relaxation (R2) and het-NOE experiments in the pH range of 6–8 (Figure 4). Transverse relaxation rates (R2) are sensitive to ms–µs exchange processes and provide information about the residue-specific conformational exchange. These exchange processes contribute to the conspicuous enhancement of R2 values. But, het-NOE values represent the ps–ns timescale motions that correspond to fast dynamics.
NMR based backbone dynamics of T7L native-state ensemble.
The transverse relaxation rates that indicate the occurrence of slow conformational transitions for multiple conformations accessing residues provide information about the distinct dynamic characteristics of the major and minor conformations. The R2 values for minor conformations in the structural elements α2-helix and β5-sheet are quite similar to those for the major species of the T7L, indicating that both conformations exhibit a similar magnitude of conformational dynamics in the ms–µs timeframe at all measured pH values, due to their similar native-like structural nature as of major conformations (Figure 4A–C). However, the R2 values for few of the C-terminal minor conformation residues are lower than those of the major counterparts indicating the flexibility of the terminal residues. Furthermore, het-NOE experiments suggest that the heterogeneous residues in both the major and minor conformations exhibit a similar type of faster timescale motional preferences (Figure 4D–F).
To characterize the degree of protein flexibility and details of the motional dynamics in native-state ensemble, detailed analysis of relaxation experiments was carried out at pH 6, 7 and 8 by comparing the resonances of the major conformations across the polypeptide chain (Figure 4). The average R2 values obtained are 24.1 ± 1.5, 24.3 ± 1.0 and 24.3 ± 1.2 s−1 at pH 6, 7 and 8, respectively, signify the similar overall molecular tumbling of T7L. However, significant differences were observed in the transverse relaxation rates at residue level upon comparing their counterparts at pH values 6–8. At pH 8, a group of residues S25, H37−G41, K71, D87, T147 and G151 had displayed large R2 values (Figure 4A). Such enhanced R2 values indicate the presence of conformational exchange at these residues. Further at pH 7, apart from the residues observed at pH 8, more number of residues showed enhanced R2 values (Figure 4B). They include G72, V83−D87, L107, K142 and V146. These residues are essentially in the vicinity or neighborhood extension of the previously observed residues at pH 8, indicating that the conformational exchange has enhanced at pH 7. Further lowering of the pH to 6 resulted in pronounced conformational exchange as several of the residues showing higher R2 values at pH 7 and 8 have broadened out and an extended set of new residues (Q8-F16, F49, I50, T56, V57, E63-A65, G85, I86, D88, A113, S148 and D149) popped up with enhanced R2 values (Figure 4C).
Furthermore, to throw light on the differential nature of the faster timescale motions, het-NOE experiments were analyzed (Figure 4D–F). The average values were observed to be 0.81 ± 0.04, 0.82 ± 0.03 and 0.83 ± 0.02 at pH 6, 7 and 8, respectively, suggesting the characteristic folded structural conformation of native states. Upon decreasing pH, the N-terminus showed decreased NOE values indicating its greater flexibility, which is also evident from the R2 data (Figure 4D–F). Overall, no major differences in the het-NOE patterns are observed between pH 7 and 8 native state. However, in the case of pH 6, some of the residues, such as I15, E58, D62, G79, I86, D89 and G117, showed significantly less NOE values compared with pH 7 and 8 states, indicating the enhanced dynamic flexibility at these sites on faster timescales. Summarizing the results on protein dynamics obtained in the present study, it can be concluded that these native states have conformational dynamic heterogeneity. Overall, the R2 and NOE analyses suggested the native-state ensemble at pH 6 is conformationally more dynamic as compared with pH 7 and 8.
Assessing the stability features of T7L native-state conformations
The structural stability of native fold has regulatory consequences on the functions of proteins. To assess the stability of T7L, chemical and thermal denaturation studies were performed in the pH range of 6–8. Temperature-dependent far-UV CD measurements were performed to analyze the thermal stability and reversibility of T7L native conformations. Recently, we have reported that at pH 7, T7L is thermally stable up to 60°C and thereafter it forms irreversible soluble and insoluble aggregates . Consistent with our earlier report, pH 8 conformation also showed similar thermal characteristics to those of pH 7 (Supplementary Figure S3A,B). However, a distinct thermal behavior has been noted at pH 6. Rather than being stable up to 60°C, the protein started partially melting above 40°C suggesting that the conformation at pH 6 is comparatively unstable to that of pH 7 and 8 (Supplementary Figure S3C).
To delineate the chemical stability, urea-induced denaturation profiles of T7L native conformations were also obtained using far-UV CD spectroscopy (Figure 5A–C). It is evident from CD profiles that at 0 M urea T7L exhibits native conformations at all these three pH values. Increasing the urea concentration from 0 to 10 M associated with the gradual melting of secondary structural elements indicated by the decrease in ellipticity at 222 nm. Furthermore, to determine the free energy change and the transition midpoints of the urea unfolding process, the transition curves are fitted to a two-state model (Figure 5D). Although the mid-point of transitions is nearly similar for all the three native conformations, the free energy values clearly indicated that pH 6 state is comparatively less stable compared with pH 7 and 8 native-state ensemble (Table 1). Moreover, the unfolding transition of pH 6 showed a continuous unfolding in the range of 0–3 M urea in contrast with pH 7 and 8 unfolding transitions, which possessed a flat baseline in this range of denaturant concentration.
Chemical denaturation of T7L.
|8||−6.1 ± 0.3||5.6|
|7||−5.9 ± 0.3||5.6|
|6||−4.1 ± 0.2||5.3|
|8||−6.1 ± 0.3||5.6|
|7||−5.9 ± 0.3||5.6|
|6||−4.1 ± 0.2||5.3|
To validate the differences in the unfolding transitions, we have recorded the 15N–1H HSQC spectra of T7L at various urea concentrations, ranging from 0 to 9.5 M, at both pH 7 and 6 (Figure 5E,F). Consistent with the far-UV CD results, NMR-based HSQC studies revealed that at pH 7, the native population of T7L is predominated up to 2 M urea concentration, as shown by wide chemical shift dispersion of peaks in both the 1H and 15N dimensions. At 4–6 M urea, pH 7 exhibited unfolding transition zone and then achieved complete denaturation at 8 M urea (Figure 5E). In the case of pH 6, T7L has signatures of folded species only at 0 M urea. A significant reduction in N–H resonances and chemical shift dispersion in the HSQC spectra indicated that between 2 and 6 M urea, the protein had its melting transition, and complete denaturation was achieved at ∼8 M urea (Figure 5F). These results are consistent with the CD data of pH 6, which evidenced the melting of T7L at low concentrations of urea. All these results establish that T7L possesses differential stability characteristics in its native-state ensemble, and moreover, the pH 6 T7L conformation is ∼2 kcal/mol less stable than those of pH 7 and 8.
Lysis activity characteristics of T7L
In the pH range of 6–8, although T7L native states have similar structural preferences, their stabilities have differed. The altered stabilities can potentially modulate the amidase activity of the T7L protein. Hence, to assess its amidase activity features, in vitro cell-based enzymatic assays based on absorbance and fluorescence were performed. The absorbance-based turbidimetric assay was performed using sensitized E. coli cells as a control, and the lysis characteristics of T7L were measured in the pH range of 6–8 (Figure 6A). The lysis activity profiles suggested that T7L shows similar and maximum decrease in optical density (OD600 nm) of E. coli cells at pH 7 and 8 suggesting for the maximum activity at these pH values as compared with pH 6, where only 50–55% of activity is noted at the end of 10 min (Figure 6A). Further to validate the turbidimetric assay results, the fluorimetric assay was performed for the quantification of amidase activity of T7L native states (Figure 6B). The fluorescence emission of total GFP content released from lysed cells in the supernatants was monitored by using fluorescence spectroscopy. The intensity of the GFP fluorescence is directly proportional to the extent of cell lysis at a given pH value. The emission profiles demonstrated that pH 7 and 8 supernatants show the maximum intensity of GFP fluorescence, as compared with that of pH 6, which showed ∼55% of pH 7 and 8 intensity, thus indicating its reduced ability to lyse the cells. These amidase assay results demonstrate that the pH 6 native state is compromised not only in its structural stability but also in its enzymatic activity as compared with pH 7 and 8 native-state populations.
Amidase activity of T7L native-state conformations at pH 6–8.
Mechanistic insights into heterogeneity, differential stability and activity features of T7L native-state ensemble
Deciphering the conformational heterogeneity, stability and dynamic properties of proteins in their native ensemble provides mechanistic insights into the shallowness of native state, and its role on the modulation of functional characteristics upon mild environmental perturbations such as pH/temperature . The current study elucidated that the native state of T7L is heterogeneous as we could able to identify two interconverting conformations of the protein on a slow NMR timescale. The identified minor conformation exhibits native-like structural and dynamic features and is in dynamic equilibrium with the major conformation.
It has been reported in the literature that the minor conformations of proteins can display native-like or unfolded characteristics . For example, studies on human intestinal fatty acid binding protein (hIFABP) revealed the coexistence of four conformational states (one major state and three minor intermediate states) in the absence of ligands that undergo slow conformational exchanges on the millisecond timescale . It is interesting to note that, unlike T7L studied here; these minor conformations show differential structural and functional preferences in the same protein. The authors also reported that the native-like minor conformations are irrelevant to fatty acid transport whereas unfolded minor conformations facilitate the fatty acid entry process . The study on Kunitz-type domain of human-type VI collagen α3 (VI) chain (domain C5) evidenced its multiple conformations slowly exchanging on the NMR timescale . The NMR study demonstrates that domain C5 is a highly dynamic at a temperature between 10 and 30°C and exists in at least two or three conformations. The ratios of the major to minor conformations at temperatures 10 and 30°C indicate that the amount of the minor conformation for the residues located in the β-sheet decreased with increasing temperature. The authors concluded that multiple conformational exchange processes must be responsible for structure fluctuation in this part of domain C5 . For FK506-binding protein FKBP12, Mustafi et al.  reported that the residues exhibiting resonance doubling not only extend the majority of the active site but also span the rest of the protein structure. Many of the residues that exhibited resonance doubling also participated in conformational line-broadening transition. Indeed, all the well-resolved minor conformations do show similar values of transverse relaxation rates (R2) to those of the major conformations, which indeed is the case observed for T7L .
The environmental perturbation, such as pH/temperature, is very handy in modulating the conformational heterogeneity and dynamic behavior of the native-state ensemble, henceforth to elucidate the origin of these fascinating biophysical features of the native ensemble. Indeed, providing rationales using the pH perturbation are of relative ease, as it is directly associated with the protonation/deprotonation of specific amino acids in the chosen pH range. Considering this in mind, T7L has been subjected to a small pH perturbation (pH 6–8), where the protein remained in its native state. The dynamic studies suggested that although the protein is completely folded, it exhibits enhanced dynamics at pH 6 compared with pH 7 and 8. The major dynamic changes are associated around the enzymatic active site of the protein (Figure 7). Several studies in the literature reported that the enhanced dynamics of the protein can potentially alter its conformational stability and also its activity or functional features [30,62–65]. In line with several of the literature reports, such enhanced dynamics has resulted in partial loss of its stability and enzymatic activity for T7L. A close examination of the distances of the residues involved in enhanced dynamics (such as S25, H37-G41, G72, K71, G85, I86, D88, V146, T147 and R150) indicated that they are present in the close vicinity of catalytic site (Figure 7A,B). Thus, these residues may have a significant role in maintaining the conformation of active site and can influence the arrangement/interactions of active site residues during peptidoglycan recognition and cleavage. Furthermore, our recent NMR experiments on partially folded (PF) T7L conformations established the α1-helix (V30–Q40) and the loop-34 (R61–S77) are involved in the unfolding initiation mechanism of this protein . Combining these dynamics results with our previous reports on T7L-PF conformations suggest that there is a strong overlap of residues that dictate the activity and stability of T7L.
Structure-dynamics-stability-activity relationship of T7L.
Moreover, interestingly several of these broadened/enhanced dynamic residues are around the stretches of His residues that can have a strong influence due to their protonation/deprotonation mechanism. Hence, the differential sensitivity of conformational dynamics in the chosen pH range of 6–8 can be attributed to the partial protonation of non-catalytic His side chains; H37 (pKa 5.57), H75 (pKa 6.22) and H124 (pKa 5.07) as calculated computationally [49,66]. It is worth noting that T7L is a zinc amidase, where two His residues (His 18—pKa 4.2; and His 123—pKa 4.9) are involved in coordination with the metal ion . The pKa values of these two His residues in the range of 4–5 indicate that they are not completely exposed and are partially/fully buried. Furthermore, the pKa values of the H18 and H123 also suggest that there will be ∼80–90% of their population in the deprotonated and metal co-ordinated form at pH 6. Indeed, Zhou et al. suggested that the loss of metal (zinc) coordination due to His protonation occurs below pH 6, and the authors did not observe drastic effects of protonation in the pH range of 6–11 . In the current scenario, a certain extent of the activity loss can also be attributed to the altered effects of Zn2+ coordination, in addition to the observed conformational dynamics. Indeed, the altered dynamics and catalytic attenuation are offshoots of the combined protonation effects of non-catalytic and buried catalytic His residues that are present across the polypeptide chain. Henceforth, our comprehensive experimental data establish that the stability and activity of T7L are coupled, which, in turn, is essentially modulated by its native-state dynamics.
The current work deciphered that the native state of T7L exhibits conformational heterogeneity with the presence of multiple conformations, and the minor conformations do possess native-like structural and dynamic characteristics. The native conformations present at pH 6 and 7/8 have differential structural stability, enzymatic activity and dynamic behaviors. As decoding the dynamics–stability–function paradigm of a native ensemble of proteins is important for understanding the relationship between molecular flexibility and function, the biophysical and NMR-based atomic insights provided in the present study can be of great use to decipher such correlated relationships of other single-domain endolysins that share common structural fold like T7L. Furthermore, as engineering of endolysins with broad specificity and enhanced activity is currently one of the fascinating research areas to combat the bacterial species that develop resistance over traditional antibiotics, our mechanistic insights into the pH-dependent dynamics and stability of T7L could be handy in engineering novel single-domain amidases with enhanced stability and activity over a broad pH range.
K.M.P. designed the project. M.S., N.J. and D.K. performed the experiments. M.S., N.J. and K.M.P. analyzed the data. M.S. and K.M.P. wrote the manuscript. All the authors reviewed and accepted the final version of the manuscript.
This work is supported by the Science and Engineering Research Board—SB/YS/LS-380/2013, Department of Biotechnology-Innovative Young Biotechnologist Award fellowship—BT/07/IYBA/2013-19, GKC-01/2016-17/212/NMCG-Research from National Mission for Clean Ganga-Ministry of Water Resources and startup aid from Ministry of Human Resource Development-Indian Institute of Technology Roorkee, to K.M.P. and SERB—EMR/2016/001756 to D.K.
M.S. acknowledges the receipt of JRF/SRF fellowship from UGC-NET. We acknowledge the support of biophysical and 500 MHz NMR instrumentation facilities at Institute instrumentation centre, IIT-Roorkee and 800 MHz NMR facility at CMBR, Lucknow.
The Authors declare that there are no competing interests associated with the manuscript.