Articles Cite This: ACS Chem. Biol. XXXX, XXX, XXX−XXX
Identification of Noncatalytic Lysine Residues from Allosteric Circuits via Covalent Probes Jens Bongard,† Marian Lorenz,† Ingrid R. Vetter,‡ Patricia Stege,‡ Arthur T. Porfetye,‡ Anna Laura Schmitz,§ Farnusch Kaschani,§ Alex Wolf,∥ Uwe Koch,∥ Peter Nussbaumer,∥ Bert Klebl,∥ Markus Kaiser,*,§ and Michael Ehrmann*,† †
Microbiology, Faculty of Biology, Zentrum für Medizinische Biotechnologie (ZMB), Universität Duisburg-Essen, Universitätsstr. 2, 45117 Essen, Germany ‡ Max-Planck-Institute of Molecular Physiology, Otto-Hahn-Str. 11, 44227 Dortmund, Germany § Chemical Biology, Faculty of Biology, Zentrum für Medizinische Biotechnologie (ZMB), Universität Duisburg-Essen, Universitätsstr. 2, 45117 Essen, Germany ∥ Lead Discovery Center GmbH, Otto-Hahn-Str. 15, 44227 Dortmund, Germany S Supporting Information *
ABSTRACT: Covalent modifications of nonactive site lysine residues by small molecule probes has recently evolved into an important strategy for interrogating biological systems. Here, we report the discovery of a class of bioreactive compounds that covalently modify lysine residues in DegS, the rate limiting protease of the essential bacterial outer membrane stress response pathway. These modifications lead to an allosteric activation and allow the identification of novel residues involved in the allosteric activation circuit. These findings were validated by structural analyses via X-ray crystallography and cell-based reporter systems. We anticipate that our findings are not only relevant for a deeper understanding of the structural basis of allosteric activation in DegS and other HtrA serine proteases but also pinpoint an alternative use of covalent small molecules for probing essential biochemical mechanisms.
B
out to be covalent modifiers of noncatalytic lysine residues, thereby enabling the identification of novel residues involved in an allosteric activation circuit.24
esides their well-established application as highly effective drugs,1 covalent probes also represent useful chemical tools. For example, proteome-wide profiling approaches such as activity-based protein profiling (ABPP) or global reactivity profiling are based on the use of covalent probes.2−5 Moreover, covalent probes may serve as starting points for the development of noncovalent small molecule modulators.6,7 Most covalent probes thereby target active site residues of enzymes or cysteine residues (catalytic or noncatalytic) that display a particularly pronounced chemical reactivity.8−10 More challenging is however the development of chemical probes that target other noncatalytic residues with less chemical reactivity under physiological conditions, and first success stories on targeting noncatalytic tyrosine, serine, or glutamic acid moieties have been reported.11−15 Recently, noncatalytic lysine residues have come into focus for covalent chemical probe discovery.16−21 In fact, a recent global profiling study has revealed that many noncatalytic lysine residues harbor enough chemical reactivity for covalent modification even at cellular pH levels, pinpointing that selective targeting of lysines may evolve into a similar generic strategy as covalent cysteine modification.22,23 Here, we report the serendipity-based finding of a class of small molecule activators of DegS, the rate-limiting enzyme of the essential bacterial outer membrane stress signaling pathway, that turned © XXXX American Chemical Society
■
RESULTS AND DISCUSSION Quinazolin-4(3H)-one Hydroxamic Esters As DegS Activators. Our discovery evolved from a small molecule screen for inhibitors of the human HTRA1 serine protease. This screen comprising 185 000 compounds yielded [2-(3,5dimethoxyphenyl)-4-oxoquinazolin-3-yl] 5-chloro-2-methoxy benzoate as an HTRA1 inhibitor (1, Figure 1A). Subsequent biochemical validation via SDS-PAGE analysis however revealed that protease inhibition was associated with concomitant HTRA1 degradation, suggesting compound-triggered HTRA1 autoproteolysis as the underlying mechanism of inhibition (Figure S1). To evaluate the specificity of this effect, we tested this compound versus other serine proteases. While compound 1 caused a slight inhibition of the classic serine proteases chymotrypsin, trypsin, and elastase (Figure S2), a dose-dependent activation of the bacterial HTRA1 homologue DegS was observed, resulting in enhanced cleavage of its native Received: January 30, 2018 Accepted: April 10, 2018
A
DOI: 10.1021/acschembio.8b00101 ACS Chem. Biol. XXXX, XXX, XXX−XXX
Articles
ACS Chemical Biology
Figure 1. Quinazolin-4(3H)-one hydroxamic esters are allosteric activators of DegS. (A) Chemical structure of [2-(3,5-dimethoxyphenyl)-4oxoquinazolin-3-yl] 5-chloro-2-methoxybenzoate, the initially found allosteric activator. (B) Proteolysis of RseA by DegS. The substrate RseA was incubated with DegS in the presence of the allosteric activator peptide DNRLGLVYQF and various concentrations of 1. Samples were taken at the time points indicated and subjected to SDS-PAGE followed by staining with Coomassie blue. (C) Allosteric activation mechanism of DegS. Upon binding of stress signals (e.g., peptides corresponding to the C-termini of outer membrane proteins) to the PDZ domain of DegS, loop 3 of the protease domain is reorienting its position, thereby triggering a series of additional conformational changes that are required for reaching the active conformation. (D) Chemical structures of derivatives of 1 (left panel) that trigger differential levels of DegS activation (right panel). Statistical t test against DMSO control; *p < 0.05, **p < 0.01.
Figure 2. Identification of covalently targeted residues involved in allosteric activation. (A) Proposed mechanism of covalent modification of DegS. Nu corresponds to the O of a serine/threonine side chain or the N of a lysine side chain. (B) Modified residues in DegS identified by LC-MS. Top and bottom views of DegS protease domain (PDB entry 3LGI) with active site Ser residue (red). Modified Lys residues are highlighted in blue, modified Thr in magenta, and modified Ser in green. Below, representational peptides including the modified residue and scores from MASCOT analyses. (C) DegS activity. The substrate RseA was incubated with DegS in the presence of the allosteric activator peptides Boc-FFF−OH and DNRLGLVYQF (YQF). Activities of DegS derivatives S70A, K95A, K243A, and K247A are shown. The mean values are derived from at least two independent experiments. (D) RseA digests as described in C using additional DegS mutants.
substrate RseA (Figure 1B). DegS activation by 1 was interesting, as the scaffold of this compound, i.e., a quinazolin-4(3H)-one hydroxamic ester, can be considered a type of “activated ester” that may thus display bioreactivity, e.g.,
against active site Ser but also catalytic or noncatalytic Lys residues.22,25 Note that in all experiments, DegS constructs missing the N-terminal transmembrane segment (41 amino acids) were used. B
DOI: 10.1021/acschembio.8b00101 ACS Chem. Biol. XXXX, XXX, XXX−XXX
Articles
ACS Chemical Biology
Figure 3. Validation of the relevance of K243 and K247 for allosteric activation. (A) Cartoon of reporter assay for determining DegS activity in vivo. The regulatory protease DegS is allosterically activated by binding the C-terminus of MBP-′OmpC to its PDZ domain. Activated DegS initiates proteolytic degradation of the antisigma factor RseA that is further degraded by the proteases RseP, ClpXP, and Lon. Degradation of RseA activates SigmaE to interact with its target promoters such as P3rpoH. As the lacZ gene is under control of P3rpoH, the specific activity of β-galactosidase of the reporter strain serves as a measure of DegS activity. (B) β-Galactosidase activity in ΔdegS background with plasmid-encoded DegS variants indicated with (+AHT) and without (-AHT) induction of MBP-′OmpC. (C) Difference density (Fo-Fc, green) and 2Fo-Fc density (blue) of the DNRLGLVYQF peptide in monomer A of the DegS structure (the structure was refined without the peptide). The C-terminal residues LVYQF are visible in the electron density; the other residues are disordered. (D) The activated structure (gray, this work) superimposed onto the inactive conformation of DegS (green, PDB ID 4RQY). Salt bridges between the K243 and K247 side chains and D320/E342 of the PDZ domain (top, with the bound peptide in cyan) are indicated by dashed green and yellow lines. Charge−charge interactions with R186 in the active conformation are shown as dashed blue lines, and side chains are shown as sticks. The large shift of loop L3 and of the side chain of R178 are indicated by the arrow.
self-destruction upon compound treatment, although its structural basis is much less explored.31,32 In contrast, the pronounced specificity of DegS prevents autoproteolysis allowing straightforward characterization. In addition, compound 1 represents the first nonpeptidic DegS activator. We therefore decided to focus our subsequent investigations on DegS. Structure−activity relationships (SAR) were investigated using four derivatives of the quinazolin-4(3H)-one hydroxamic ester scaffold (Figure 1D). Derivatives 2 and 3 featured the same substitution at the 2-phenyl residue of the quinazolin4(3H)-one moiety as 1 but different substitutions at the benzoic acid moiety. In contrast, derivatives 4 and 5 had the same substitution at the benzoic acid residue as 1 but different
Activation of DegS′ proteolytic activity is triggered by an allosteric activation mechanism involving a spatial rearrangement of loop L3 of the protease domain. This results in a classic disorder−order transition of the activation domain comprising the loops L1, L2, and LD, leading to an activation of the neighboring protease domain of the DegS trimer which is the biologically active unit.26 This rearrangement is caused by peptide or protein allosteric activators that bind with their Ctermini into the peptide binding pocket of the PDZ domain where the penultimate (−1) residue interacts with loop L3 (loop nomenclature according to Perona and Craik)27 of the protease domain (Figure 1C).28−30 Of note, a similar loop L3dependent allosteric activation mechanism has also been proposed for HTRA1, which may also explain the observed C
DOI: 10.1021/acschembio.8b00101 ACS Chem. Biol. XXXX, XXX, XXX−XXX
Articles
ACS Chemical Biology
alone (DegSΔPDZ) by compound 1 increases the activity (Figure S4). K95A had only minor if any effects on DegS activation. In contrast, K243A caused an about 2- to 3-fold increase in activity in the presence of activating peptides, while K247A caused an about 2-fold reduction of DegS activity. Of note, K243A has been previously reported to cause an about 2fold activation in the presence of an activating peptide.30 Both mutants were inactive in the absence of activating peptides, which suggests that chemical modification identified K243 and K247 as residues potentially implicated in DegS regulation. To obtain additional information, we changed K243 and K247 to a negatively charged (Glu), hydrophobic (Ile), and bulky hydrophobic (Trp) residue and measured the activity of the purified mutants in the presence of the two activating peptides (Figure 2D). K243E activated DegS more than 6-fold compared to wt DegS, while K243I, K247I, and K247W were strongly inhibitory. In addition, K247E did not affect DegS activity in the presence of Boc-FFF−OH but slightly activated DegS in the presence of DNRLGLVYQF, while K243W had no significant effects. Together, these data indicate that residues K243 and K247 are implicated in the regulation of DegS activity. They also suggest that chemical modification by small molecules can serve as a tool to study the regulation of enzymatic activity. Impact of Identified Sites for Allosteric Activation. To validate these results, we measured the effects of selected mutations in living E. coli cells using a reporter assay for DegS activity. Here, DegS activity causes activation of transcription factor RpoE, which specifically activates promoter 3 (P3) of the rpoH gene encoding the cytoplasmic heat shock sigma factor. Therefore, E. coli cells harboring a P3:rpoH-lacZ fusion can be used to determine DegS activity via detection of increased βgalactosidase (LacZ) activity (Figure 3A).36 For artificial induction of DegS activity in this strain, a gene fusion was constructed between the malE gene encoding maltose binding protein (MBP) and a fragment of the ompC gene encoding the 50 C-terminal residues of the outer membrane porin C (Figure S5). The C-terminus of OmpC was chosen because its four Cterminal residues VYQF act as an allosteric activator of DegS. Expression of the artificial inducer MBP-′OmpC thus caused an about 5-fold increase of LacZ activity in the presence of the inducer anhydrotetracycline (AHT). As expected, in cells where RpoE activity is uncoupled from DegS activity, i.e., in ΔdegS or ΔrseA mutants, induction of LacZ activity was not observed. To measure the effects of the constructed degS alleles, plasmids encoding wt degS and degS K243E, K243I, and K247I were transformed into the ΔdegS derivative of the reporter strain harboring an expression vector for the production of the DegS activator MBP-′OmpC. Under conditions that activated production of MBP-′OmpC, the presence of wt degS increased β-galactosidase (LacZ) activity about 4-fold and degS K243E about 2-fold above wt degS, while LacZ activity was reduced in the presence of degS K243I and degS K247I by about 4-fold and 5-fold, respectively (Figure 3B). Note that the protein levels of the DegS mutants were very similar (Figure S6). Therefore, these in vivo data confirm the biochemical data thus implicating K243 and K247 in the regulation of DegS activity. The reporter assay is, however, not suitable to study the effects of the chemical probe in vivo as the probe is probably too unspecific and precipitates in the growth medium at higher concentrations. As the PDZ domains were only poorly resolved in previous crystal structures, we aimed at obtaining higher resolution
substitutions at the 2-phenyl moiety of the quinazolin-4(3H)one hydroxamic scaffold. All compounds activated DegS with slightly different activities (Figure 1D and Figure S3). Derivatives 2, 4, and 5 were moderately more active than 1, while derivative 3 was slightly less active. Mapping of Covalent Modification Sites. To elucidate the molecular basis of DegS activation and the underlying SAR, we performed mass spectrometric mapping of modified residues as we reckoned that the quinazolin-4(3H)-one hydroxamic ester may display bioreactivity (Figure 2A). DegS was incubated with 50 μM 5, and subsequent mass spectrometric mapping for the corresponding covalent 2methoxy-5-chloro-benzoic acid modification identified three modified Lys (K95, K243, and K247), two Ser (S46 and S70), and one Thr (T68) residue (Figure 2B). The serines and the threonine are located in loops, but the side chain of K95 is mostly buried, except for the amino group. Furthermore, no residues in the PDZ domains were modified, although the two lysines (K310 and K340) and most of the serine residues (two out of four) and threonine residues (three out of four) within this domain are solvent-accessible. This already indicates that selection of the modification sites is not only based on accessibility but probably involves more or less selective binding by the modifier compound. While K243 and K247 are located at the interface between the protease and PDZ domain and thus at the “classic” allosteric activation interface,28 all other modified residues are located within the protease domain, far away from the PDZ domain (Figure 2B). These data suggested that allosteric activation may be caused by covalent chemical modification. To characterize the implications of these observations, we tested whether this activation mechanism can substitute “standard” allosteric activation via activating peptides that bind to the PDZ domain. Accordingly, we measured proteolytic activity of DegS and a DegS variant lacking its PDZ domain (DegSΔPDZ corresponding to amino acids 42 to 251 of DegS) in the absence of the allosteric activator and compound 1. While DegS was inactive, slight constitutive activity of DegSΔPDZ was detected, as observed earlier (Figure S4).29 However, the addition of 1 led to an increase in activity of DegS and DegSΔPDZ, indicating that activation by chemical modification can be independent of allosteric peptides. These data suggest that chemical modification has the potential to identify residues in DegS that are implicated in the regulation of catalytic activity. To initially address this hypothesis, modified residues S70, K95, K243, and K247 were individually mutated to Ala (S46 and T68 were not investigated as these residues are located at surface exposed loops and are not evolutionarily conserved). The activities of the purified mutants were measured in the presence of two established activating peptides, Boc-FFF−OH and a peptide that resembles the C-terminus of the OmpC protein, DNRLGLVYQF (Figure 2C).28,33−35 Mutant S70A displayed a slight, i.e., about 30%, increase in DegS activity. S70 is located in the LA loop of DegS that has not been implicated in the regulation of DegS so far. In the related protein DegP, the (here much longer) LA loop mediates hexamerization of the two catalytically active DegP trimers. However, in both proteins, the surface exposed loop LA connects two beta strands that lead directly to the active site. A conformational change at the tip of the connecting loop could thus influence the conformation of the active site residues. This effect might offer one explanation why modification of the protease domain D
DOI: 10.1021/acschembio.8b00101 ACS Chem. Biol. XXXX, XXX, XXX−XXX
Articles
ACS Chemical Biology structures of DegS in the presence of the activating peptide DNRLGLVYQF. Our crystals grew in a new space group compared to the known DegS structures in the protein database (P2 with cell dimensions 71.6 Å, 53.7 Å, 132.5 Å, 90°, 101.8°, 90°) and diffracted at a resolution of up to 2.2 Å (Table S1). The “conventional” resolution at an I/σ cutoff of 3 would be 2.65 Å. Monomer A of the trimer in the asymmetric unit displayed a very well resolved PDZ domain, whereas the PDZ domains of monomers B and C are apparently more mobile and partially disordered, similar to other DegS crystal structures. The activator peptide of monomer A displays excellent electron density (Figure 3C), and the side chains in the relevant interface region are also well-defined (Figure 3D). Compared to the structure with the clearest density of the activator peptide so far (PDB ID 3GDV with a YQF peptide), our peptide DNRLGLVYQF shows well-defined density for residues LVYQF (Figure 3C). A salt bridge/charge network between K243, K247 (of the protease domain), E320, D324 (of the PDZ domain), and R178 of loop L3 of the protease domain seems to fix this loop in the inactive position (Figure 3D).28 Upon binding of the activator peptide to the PDZ domain, loop L3 rearranges and moves toward the bound peptide where both NH2 groups of R186 interact with the backbone carbonyl of the Gln of the activator peptide. Moreover, binding of the allosteric peptide promotes new contacts between PDZ and the protease domain, leading ultimately to a rearrangement of loop L3 including R178. Consequently, the repositioned arginine can interact with the backbone carbonyls of L164, G165, and Q166 of loop LD of a neighboring monomer. This interaction stabilizes the activation domain and thus the active site. Accordingly, the charge-reversal mutation K243E should activate protease function by inducing rearrangement of loop L3. The K243− E324 salt-bridge is located at the center of the protease−PDZ domain interface in the inactive state (Figure 3D). Abrogating this contact is thus expected to stabilize the active over the inactive conformation. In contrast, the effects of K247E are lower because K247 is located at the periphery of the interdomain interface, making it presumably less potent in influencing the switch in conformation and activity. The mutations K243I and K247W/I/A might affect the position of the PDZ domain and thus the stability of the domain interface. The effects of the K243W mutation seem to be less deleterious than K247W, probably because of the less tightly packed surroundings of this residue.
Figures S1−S6, Table S1, and supplementary methods (PDF) Accession Codes
Atomic coordinates have been deposited in the PDB under the accession code 6EW9.
■
*Phone: + 49 (0) 201 183 4980. E-mail:
[email protected]. *Phone: +49 (0) 201 183 2949. E-mail:
[email protected]. ORCID
Markus Kaiser: 0000-0002-6540-8520 Michael Ehrmann: 0000-0002-1927-260X Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work was funded by Deutsche Forschungsgemeinschaft (to M.E. and M.K.), by Ministry of Innovation, Science and Research (MIWF) of North Rhine Westphalia (to Lead Discovery Center GmbH, FKZ 314-400 012 10 and 323-400 003 12), and by CRC1093 (to M.E., M.K., I.R.V., and A.T.P.). We thank the beamline staff of P11 at the PETRA synchrotron, Hamburg, DE, for support.
■
REFERENCES
(1) Baillie, T. A. (2016) Targeted covalent inhibitors for drug design. Angew. Chem., Int. Ed. 55, 13408−13421. (2) Cravatt, B. F., Wright, A. T., and Kozarich, J. W. (2008) Activitybased protein profiling: from enzyme chemistry to proteomic chemistry. Annu. Rev. Biochem. 77, 383−414. (3) Niphakis, M. J., and Cravatt, B. F. (2014) Enzyme inhibitor discovery by activity-based protein profiling. Annu. Rev. Biochem. 83, 341−377. (4) Weerapana, E., Wang, C., Simon, G. M., Richter, F., Khare, S., Dillon, M. B., Bachovchin, D. A., Mowen, K., Baker, D., and Cravatt, B. F. (2010) Quantitative reactivity profiling predicts functional cysteines in proteomes. Nature 468, 790−795. (5) Serafimova, I. M., Pufall, M. A., Krishnan, S., Duda, K., Cohen, M. S., Maglathlin, R. L., McFarland, J. M., Miller, R. M., Frodin, M., and Taunton, J. (2012) Reversible targeting of noncatalytic cysteines with chemically tuned electrophiles. Nat. Chem. Biol. 8, 471−476. (6) Erlanson, D. A., Wells, J. A., and Braisted, A. C. (2004) Tethering: fragment-based drug discovery. Annu. Rev. Biophys. Biomol. Struct. 33, 199−223. (7) Schmidt, M. F., and Rademann, J. (2009) Dynamic templateassisted strategies in fragment-based drug discovery. Trends Biotechnol. 27, 512−521. (8) Lagoutte, R., Patouret, R., and Winssinger, N. (2017) Covalent inhibitors: an opportunity for rational target selectivity. Curr. Opin. Chem. Biol. 39, 54−63. (9) Chaikuad, A., Koch, P., Laufer, S., and Knapp, S. (2018) Targeting the protein kinases cysteinome. Angew. Chem., Int. Ed. 57, 4372. (10) Backus, K. M., Correia, B. E., Lum, K. M., Forli, S., Horning, B. D., Gonzalez-Paez, G. E., Chatterjee, S., Lanning, B. R., Teijaro, J. R., Olson, A. J., Wolan, D. W., and Cravatt, B. F. (2016) Proteome-wide covalent ligand discovery in native biological systems. Nature 534, 570−574. (11) Shannon, D. A., and Weerapana, E. (2015) Covalent protein modification: the current landscape of residue-specific electrophiles. Curr. Opin. Chem. Biol. 24, 18−26.
■
CONCLUSION Our studies have demonstrated the potential of bioreactive small molecules targeting noncatalytic residues such as lysines in identifying essential residues in allosteric circuits. We anticipate that our study will pave the way for a broader use of covalent probes in similar biochemical research questions.
■ ■
AUTHOR INFORMATION
Corresponding Authors
MATERIALS AND METHODS
All materials and experimental procedures are included in the Supporting Information.
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acschembio.8b00101. E
DOI: 10.1021/acschembio.8b00101 ACS Chem. Biol. XXXX, XXX, XXX−XXX
Articles
ACS Chemical Biology (12) Fadeyi, O. O., Hoth, L. R., Choi, C., Feng, X. D., Gopalsamy, A., Hett, E. C., Kyne, R. E., Robinson, R. P., and Jones, L. H. (2017) Covalent enzyme inhibition through fluorosulfate modification of a noncatalytic serine residue. ACS Chem. Biol. 12, 2015−2020. (13) Martin-Gago, P., Fansa, E. K., Winzker, M., Murarka, S., Janning, P., Schultz-Fademrecht, C., Baumann, M., Wittinghofer, A., and Waldmann, H. (2017) Covalent protein labeling at glutamic acids. Cell Chem. Biol. 24, 589−597. (14) Gu, C., Shannon, D. A., Colby, T., Wang, Z. M., Shabab, M., Kumari, S., Villamor, J. G., McLaughlin, C. J., Weerapana, E., Kaiser, M., Cravatt, B. F., and van der Hoorn, R. A. L. (2013) Chemical proteomics with sulfonyl fluoride probes reveals selective labeling of functional tyrosines in glutathione transferases. Chem. Biol. 20, 541− 548. (15) Crawford, L. A., and Weerapana, E. (2016) A tyrosine-reactive irreversible inhibitor for glutathione S-transferase Pi (GSTP1). Mol. BioSyst. 12, 1768−1771. (16) Akcay, G., Belmonte, M. A., Aquila, B., Chuaqui, C., Hird, A. W., Lamb, M. L., Rawlins, P. B., Su, N., Tentarelli, S., Grimster, N. P., and Su, Q. B. (2016) Inhibition of Mcl-1 through covalent modification of a noncatalytic lysine side chain. Nat. Chem. Biol. 12, 931−936. (17) Choi, S., Connelly, S., Reixach, N., Wilson, I. A., and Kelly, J. W. (2010) Chemoselective small molecules that covalently modify one Lys in a non-enzyme protein in plasma. Nat. Chem. Biol. 6, 133−139. (18) Pettinger, J., Le Bihan, Y. V., Widya, M., van Montfort, R. L. M., Jones, K., and Cheeseman, M. D. (2017) An irreversible inhibitor of HSP72 that unexpectedly targets lysine-56. Angew. Chem., Int. Ed. 56, 3536−3540. (19) Zhao, Q., Ouyang, X. H., Wan, X. B., Gajiwala, K. S., Kath, J. C., Jones, L. H., Burlingame, A. L., and Taunton, J. (2017) Broadspectrum kinase profiling in live cells with lysine-targeted sulfonyl fluoride probes. J. Am. Chem. Soc. 139, 680−685. (20) Pettinger, J., Jones, K., and Cheeseman, M. D. (2017) Lysinetargeting covalent inhibitors. Angew. Chem., Int. Ed. 56, 15200−15209. (21) Dalton, S. E., Dittus, L., Thomas, D. A., Convery, M. A., Nunes, J., Bush, J. T., Evans, J. P., Werner, T., Bantscheff, M., Murphy, J. A., and Campos, S. (2018) Selectively targeting the kinome-conserved lysine of PI3Kδ as a general approach to covalent kinase inhibition. J. Am. Chem. Soc. 140, 932−939. (22) Hacker, S. M., Backus, K. M., Lazear, M. R., Forli, S., Correia, B. E., and Cravatt, B. F. (2017) Global profiling of lysine reactivity and ligandability in the human proteome. Nat. Chem. 9, 1181−1190. (23) Ward, C. C., Kleinman, J. I., and Nomura, D. K. (2017) NHSEsters as versatile reactivity-based probes for mapping proteome-wide ligandable hotspots. ACS Chem. Biol. 12, 1478−1483. (24) Merdanovic, M., Clausen, T., Kaiser, M., Huber, R., and Ehrmann, M. (2011) Protein quality control in the bacterial periplasm. Annu. Rev. Microbiol. 65, 149−168. (25) Cognetta, A. B., 3rd, Niphakis, M. J., Lee, H. C., Martini, M. L., Hulce, J. J., and Cravatt, B. F. (2015) Selective N-hydroxyhydantoin carbamate inhibitors of mammalian serine hydrolases. Chem. Biol. 22, 928−937. (26) Clausen, T., Kaiser, M., Huber, R., and Ehrmann, M. (2011) HTRA proteases: regulated proteolysis in protein quality control. Nat. Rev. Mol. Cell Biol. 12, 152−162. (27) Perona, J. J., and Craik, C. S. (1995) Structural basis of substrate specificity in the serine proteases. Protein Sci. 4, 337−360. (28) Wilken, C., Kitzing, K., Kurzbauer, R., Ehrmann, M., and Clausen, T. (2004) Crystal structure of the DegS stress sensor: How a PDZ domain recognizes misfolded protein and activates a protease. Cell 117, 483−494. (29) Hasselblatt, H., Kurzbauer, R., Wilken, C., Krojer, T., Sawa, J., Kurt, J., Kirk, R., Hasenbein, S., Ehrmann, M., and Clausen, T. (2007) Regulation of the σE stress response by DegS: how the PDZ domain keeps the protease inactive in the resting state and allows integration of different OMP-derived stress signals upon folding stress. Genes Dev. 21, 2659−2670. (30) Sohn, J., Grant, R. A., and Sauer, R. T. (2007) Allosteric activation of DegS, a stress sensor PDZ protease. Cell 131, 572−583.
(31) Truebestein, L., Tennstaedt, A., Mönig, T., Krojer, T., Canellas, F., Kaiser, M., Clausen, T., and Ehrmann, M. (2011) Substrate-induced remodeling of the active site regulates human HTRA1 activity. Nat. Struct. Mol. Biol. 18, 386−388. (32) Krojer, T., Sawa, J., Huber, R., and Clausen, T. (2010) HtrA proteases have a conserved activation mechanism that can be triggered by distinct molecular cues. Nat. Struct. Mol. Biol. 17, 844−852. (33) Hauske, P., Mamant, N., Hasenbein, S., Nickel, S., Ottmann, C., Clausen, T., Ehrmann, M., and Kaiser, M. (2009) Peptidic small molecule activators of the stress sensor DegS. Mol. BioSyst. 5, 980− 985. (34) Hauske, P., Ottmann, C., Meltzer, M., Ehrmann, M., and Kaiser, M. (2008) Allosteric regulation of proteases. ChemBioChem 9, 2920− 2928. (35) Merdanovic, M., Mönig, T., Ehrmann, M., and Kaiser, M. (2013) Diversity of allosteric regulation in proteases. ACS Chem. Biol. 8, 19−26. (36) Mecsas, J., Rouviere, P. E., Erickson, J. W., Donohue, T. J., and Gross, C. A. (1993) Escherichia coli heat shock gene mutants are defective in proteolysis. Genes Dev. 7, 2618−2628.
F
DOI: 10.1021/acschembio.8b00101 ACS Chem. Biol. XXXX, XXX, XXX−XXX
