Subscriber access provided by MT ROYAL COLLEGE
Communication
Affinity-Based Screen for Inhibitors of Bacterial Transglycosylase Wei-Shen Wu, Wei-Chieh Cheng, Ting-Jen R. Cheng, and Chi-Huey Wong J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.7b13205 • Publication Date (Web): 07 Feb 2018 Downloaded from http://pubs.acs.org on February 7, 2018
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 5 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Journal of the American Chemical Society
Affinity-Based Screen for Inhibitors of Bacterial Transglycosylase Wei-Shen Wu,†,‡ Wei-Chieh Cheng,‡ Ting-Jen R. Cheng,‡,* and Chi-Huey Wong†,‡,* †Graduate
Institute of Life Sciences, National Defense Medical Center, 161 Minquan E. Road, Section 6, Neihu, Taipei, 114,
Taiwan ‡Genomics
Research Center, Academia Sinica, 128 Academia Road, Section 2, Nankang, Taipei, 115, Taiwan
Supporting Information Placeholder ient and efficient method to facilitate the isolation of reliable hits from complex mixtures is urgently needed. The objective of this study is to devise a fast and effective separation platform based on the known concept of affinity chromatography for the enrichment of potent natural product inhibitors targeting bacterial TGase (Figure 1). Toward this goal, we immobilized the target enzyme onto nickel chelation beads and used known inhibitors with different dissociation constants to estimate the concentration of binders needed for isolation and characterization. Three Acinetobacter baumannii PBP1b and three Clostridium difficile PBP variants (Figure 2A) were individually expressed in Escherichia coli, and the whole cell extracts were collected and directly incubated with the beads without prior purification. The protein target was successfully immobilized and appeared as a single band on gels (Figure 2B), and the amount of immobilized PBPs was quantitatively determined. It was noted that the
ABSTRACT: The rise of antibiotic resistance has created a mounting crisis across the globe and an unmet medical need for new antibiotics. As part of our efforts to develop new antibiotics to target the uncharted surface bacterial transglycosylase, we report an affinity-based ligand screen method using penicillinbinding proteins immobilized on beads to selectively isolate the binders from complex natural products. In combination with mass spectrometry and assays with moenomycin A and salicylanilide analogues (1−10) as reference inhibitors, we isolated four potent antibacterials confirmed to be benastatin derivatives (11−13) and albofungin (14). Compounds 11 and 14 were effective antibiotics against a broad-spectrum of Gram-positive and Gram-negative bacteria including A. baumannii, C. difficile, S. aureus and drugresistant strains with minimum inhibitory concentrations in the submicromolar to nanomolar range.
Antibiotic resistance has been a serious public health problem that has driven continuous efforts to identify new targets and develop new antibiotics. Bacterial cell wall biosynthesis is essential for growth and division. The major constituent of bacterial cell wall, peptidoglycan (PG), is synthesized by a bi-functional enzyme, penicillin-binding protein (PBP), which consists of two enzymatic domains with transglycosylase (TGase) and transpeptidase (TPase) activities.1 By virtue of its accessible location and critical function, PBP has been an attractive target for the development of new antibacterial agents. In addition, the oligosaccharide backbone of bacterial cell wall is highly conserved and no change has been found. To date, numerous TPase-targeted antibiotics have been successfully developed, whereas only a few compounds have been identified to inhibit TGase but have never been developed into antibiotics for humans.2−5 Accordingly, TGase inhibitors targeting the glycosidic bond formation of bacterial cell-wall PG may be developed as antibiotics to circumvent the emerging resistance of pathogens against available antibacterial drugs. Natural products have historically been recognized as a prolific and dependable source of antibacterial agents.6−8 However, the typical process for isolating compounds with desirable bioactivity from natural product extracts is time-consuming and costly. Despite the values of the broad chemical and biological diversity of natural products, the complexity of the process may interfere with the assay robustness and impede the discovery process. A conven-
Figure 1. Schematic illustration of affinity-based ligand screening. NB: Nickel chelation beads. immobilization efficacy of A. baumannii and C. difficile TGases was about 10-fold higher than that of full-length PBPs (Table S1). Further TGase activity assay demonstrated that both PBP and TGase were enzymatically active after immobilization and PBP was relatively more active than TGase. As expected, the transglycosylation activity (Figure S1) was observed only in the assay 1
ACS Paragon Plus Environment
Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
with full-length PBP and TGase, but not the immobilized TPase (Figure 2C). It was also noted that the immobilized enzymes are much more stable than the free enzymes (Supplementary results and Figure S2) and can be reused. Our studies showed that the bound PBPs exhibited 80% TGase activity after 3 cycles of ligand screening (Figure S3). A transglycosylase inhibitor, moenomycin A (Moe A)9−13 was then utilized to optimize the working conditions and to validate the feasibility of this platform. The enzyme-immobilized beads were incubated with various concentrations of Moe A ranging from 10 to 0.1 µM and then the bound molecule was eluted for matrix-assisted laser desorption-ionization (MALDI) mass spectroscopy analysis. As expected, the bound Moe A can be monitored using full-length PBP-(Figures S4A and S4D) and TGaseimmobilized beads (Figure S4B, and S4E), but not TPaseimmobilized beads (Figure S4C and S4F), and the intensity of Moe A decreased in a concentration-dependent manner. Moreover, the bound Moe A was quantified to calculate the percentage of Moe A captured. As shown in Table S2, greater than 50% of Moe A can be captured by immobilized full-length PBPs. It is noted that the immobilized full-length PBPs can capture more Moe A than immobilized TGases do, and as expected, Moe A is hardly captured by TPase-immobilized beads. In addition, kanamycin was manipulated as a negative control to examine the specificity of TGase-immobilized beads. As shown in Table S3, less than 10% of kanamycin was captured by immobilized enzymes. These results confirmed that the binding of Moe A was specifically toward the TGase domain.
Page 2 of 5
and 4−9) possessed promising TGase inhibitory activities, and the other two compounds (2 and 3) were almost inactive (Table S4). The binding affinity determined by this platform correlated well with the TGase inhibitory activity (Table S4), and the potency of compound mixtures which shared certain structural similarity was distinguishable, providing an indirect and expeditious way to gain insight into the structure-activity relationship. Continuing with our search for new TGase inhibitors, a library of nearly 15,000 crude fermentation broth extracts derived from microbial organisms was preliminarily screened for significant antibacterial activities against A. baumannii, E. coli, methicillinresistant Staphylococcus aureus (MRSA), and other clinically important bacteria. Among these active extracts, two extracts numbered 18894 and 31081 were of particular interests because of their inhibitory activities against purified A. baumannii PBP1b and C. difficile PBP (Figure S6) as well as their activities against bacterial cell wall synthesis (Figure S7). The ethyl acetate extracts of 18894 and 31081 were subjected to affinity-based binder isolation using immobilized A. baumannii TGase, followed by preparative TLC and HPLC analyses of the ligands eluted by methanol. It was shown that the bound ligands from 18894 contained 12 distinguished compounds, and that from 31081 contained 11 compounds (Figure S8). The transglycosylation activity assay was further applied to rule out false positives and to identify three compounds (11−13) from 18894 and one compound (14) from 31081 (Figure 3) as TGase inhibitors. The molecular formulas of 11−14 were established by electrospray ionization mass spectrometry analysis (Figures S9−S12). Their structures were elucidated by 1D and 2D NMR spectral analysis and confirmed to be benastatin derivatives TAN1532B (11), bequinostatin E (12), benastatin B (13) and albofungin (14). The detail assignments of NMR spectra were summarized in the Supporting Information (SI) (Tables S5−S9). O 21
OH
12
HO
OH 15 2
HO
17
7
8
TAN1532B (11) O 21
HO
22
12
16
1
6
OH
Bequinostatin E (12)
17
H
O
10
18
11 19
4
17 18
O 9
OH
5 7
O
6
OH
20
O 28
Benastatin B (13)
4 5 16
27
14
3 2
1
15
OH OH
N
NH2
O
26
21 22
6 7
O
19
8
9 12 13
20
3
14
8
10
Figure 2. Immobilization of His-tagged PBPs onto nickel chelation beads. (A) Three A. baumannii PBP1b variants and three C. difficile PBP variants. (B) After incubation with nickel chelation beads, the immobilized proteins were analyzed by SDS-PAGE. (C) To measure the enzyme activity, the substrate NBD-lipid II was mixed with immobilized enzymes for polymerization, and the product was then digested by N-acetylmuramidase to release peptidoglycan monomer (PGM) which was monitored by ionexchange HPLC. Peaks corresponding to PG monomer were indicated with arrow.
O
OH 15 2
13
11
7
OH
HO
5
9
6
OH
20 19
4
13
10
18 17
14
11
O
16 3
1
12
HO
OH 15 2
HO
O
19
5
9
OH
20
4
8
10
18
3
1 14
13
11
O
16
25
O
24 23
OH
Albofungin (14)
Figure 3. Chemical structures of isolated benastatin derivatives (11−13) and albofungin (14). It was confirmed that the in vitro polymerization of NBD-lipid II was inhibited by 11−14 in a dose-dependent fashion (1.6−800 µM) (Figures S13−S16). As shown in Figure 4, 11 and 13 exhibited better inhibitory activities than 12, indicating that the carbonyl moiety on the C-13 position might weaken the activities of benastatin analogues. In addition, 14 was found to have IC50 in the µM range toward these purified PBPs (Figure 4). It is also noted that A. baumannii PBP1b appeared to be more sensitive to 11 and 14 than other bacterial TGases (Figure 4). Furthermore, the competitive inhibition constants (Ki) of 11 and 14 toward different bacterial TGases were determined to be in the µM range (Table 1) using the Förster resonance energy transfer (FRET)-based TGase activity assay15 (Figures S17 and S18). A binding assay using fluorescent Moe A analogue16 showed that compounds 11 and 14 could not
To further support the applicability of this platform to identify inhibitors from compound pools, a set of salicylanilide analogues (1−9, Table S4 and Figure S5A), a core identified previously as TGase inhibitors from our group14 was exploited. Compounds were pooled into a group under the concentrations ranging from 10 to 0.1 µM. The compound mixtures were subsequently incubated with enzyme immobilized beads and the bound molecules were eluted for MALDI analysis. Seven distinct signals (1 and 4−9) were found when using full-length PBP- and TGaseimmobilized beads (Figure S5B, S5C, S5E, and S5F). The TGase activity analysis further confirmed that these seven compounds (1 2
ACS Paragon Plus Environment
Page 3 of 5
efficiently exchange out the bound fluorescent Moe A analogue (data not shown), indicating that these two compounds are not bound to the donor (the elongating strand) site of TGases.
Compounds 11−14 also showed great antibacterial activities against a broad-spectrum of Gram-positive bacteria, including the drug-resistant strains of enterococci and staphylococci (Tables 1 and S10−13). In agreement with the IC50 results (Figure 4), the minimum inhibitory concentrations (MICs) of 11 and 13 (1−8 µg mL−1) showed better efficacy compared with 12 (16−256 µg mL−1). Again, it was found that the carbonyl moiety on the C-13 position of 12 might weaken its bioactivity. Besides, 11 and 13 were ineffective against most Gram-negative bacteria, but showed moderate activity against A. baumannii (Tables 1 and S11). Notably, the MICs results indicated that 11 and 13 appeared to be inactive against E. coli, while both compounds displayed potent inhibitory activities toward purified E. coli PBP1b, probably due to the outer membrane barrier. Remarkably, 11 and 13 exhibited excellent bactericidal activity (under 4 × MIC treatment) and caused an approximately 3 log10 reduction of viability after 30 minutes against MRSA (Figure S19). 14 exhibited the most potent antibacterial activity with excellent MIC against a wide range of Grampositive bacteria including drug-resistant strains (Tables 1 and S10). However, the MICs of 14 were considerably lower than the Ki value of 14, implying that 14 might interact with additional targets. 14 was also significantly active against Gram-negative bacteria, including A. baumannii, E. coli, Klebsiella pneumonia, and Pseudomonas aeruginosa (Tables 1 and S11). Encouraged by the potency of 11 and 14 against A. baumannii, a panel of ~15 clinically isolated multidrug resistant A. baumannii was tested, and the MICs of 14 displayed no appreciable difference (Table S12), indicating that 14 could be used for the development of antibiotics against multidrug resistant A. baumannii.
B A. baumannii PBP1b 100
IC50 (µM)
80
9.9 11 12 43.4 13 16.0 14 11.3
60 40 20 0 0
1
2
3
Transglycosylase activity (%)
Transglycosylase activity (%)
A
C. difficile PBP 100
IC50 (µM)
80
11 61.9 12 152.1 13 53.3 14 69.8
60 40 20 0 0
Log 10 [Inhibitor] (µM)
1
2
3
Log 10 [Inhibitor] (µM)
D E. coli PBP1b 100
IC50 (µM)
80
11 23.5 12 107.0 13 30.7 14 34.9
60 40 20 0 0
1
2
3
Transglycosylase activity (%)
C Transglycosylase activity (%)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Journal of the American Chemical Society
S. aureus PBP2 100
IC50 (µM)
80
12.1 86.9 31.6 15.3
11 12 13 14
60 40 20 0 0
Log 10 [Inhibitor] (µM)
1
2
3
Log 10 [Inhibitor] (µM)
Figure 4. Dose-response curves for the inhibition of transglycosylase activities by 11−14. The relative activity was determined by HPLC-based TGase activity assay in the presence of A. baumannii PBP1b (90 nM) (A), C. difficile PBP (10 nM) (B), E. coli PBP1b (20 nM) (C), and S. aureus PBP2 (0.75 µM) (D). The results were fitted to a nonlinear regression sigmoidal doseresponse curve to determine the IC50 values.
Table 1. Inhibition Constants (Ki) of 11 and 14 toward Different Bacterial TGases and Antibacterial Activities of 11−14 against Various Pathogenic Microorganisms.
Organisms
Ki (µM) a 11
12 d
MIC (µg mL-1)
13
14
Moe A
n.d.
2.5 ± 0.2
b
11
12
13
14
Moe A
0.0029 ± 0.0001
64
>256
128
2
16
n.d.
n.d.
A. baumannii
1.3 ± 0.2
n.d.
C. difficile
4.5 ± 0.3
n.d.
n.d.
9.2 ± 0.6
0.0007 ± 0.0002
n.d.
n.d.
n.d.
E. coli
7.8 ± 1.2
n.d.
n.d.
6.2 ± 0.5
0.0012 ± 0.0002
>256
>256
>256
1
128
S. aureus
3.7 ± 0.7
n.d.
n.d.
4.6 ± 0.9
n.d.
2
16
4
0.008
0.032
n.d.
n.d.
n.d.
n.d.
n.d.
4
16
4
0.004
0.064
MRSA c a
Shown are the mean values of the data generated from three independent experiments. b The Ki of Moe A are adopted from the literature.15 c MRSA, methicillin-resistant Staphylococcus aureus. dn.d.: Not Determined. In summary, we have developed an effective affinity-based ligcromolar to nanomolar range (Tables 1 and S10−S11), suggesting and screening platform for identification of TGase inhibitors from its potential use as a broad-spectrum antibiotic. Crystallization of fermentation broth extracts. The operative conditions of TGasepurified TGases in complex with 11 and 14 to understand the immobilized beads for affinity-based ligand screening was optidetail mechanism of inhibition and to develop TGase specific mized and the platform was validated using known inhibitors. inhibitors is ongoing. Application of this platform further led to the discovery of benastatin derivatives (11−13) and albofungin (14) with significant ASSOCIATED CONTENT antibacterial and TGase inhibitory activities. Benastatins belong Supporting Information to a class of aromatic polyketides produced by Streptomyces species. Previous report showed that benastatin A and B are active The Supporting Information is available free of charge on the against some Gram-positive bacteria17 but the targets are unACS Publications website at https://pubs.acs.org/. known. Our findings are to our knowledge the first report conSI results, methods, tables, and figures (PDF) cerning the inhibition of bacterial TGases by benastatin derivatives and Gram-negative bacterial TGases by albofungin. AlboAUTHOR INFORMATION fungin (14), a polycyclic xanthone metabolite isolated from ActiCorresponding Author nomyces species, was shown to have antibacterial, antifungal, and antitumor effects.18 We found that 14 was able to inhibit the *
[email protected]. or *
[email protected]. TGase from different strains and the growth of both GramORCID positive and Gram-negative bacteria with MICs in the submi3
ACS Paragon Plus Environment
Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 4 of 5
(10) Lovering, A. L.; de Castro, L. H.; Lim, D.; Strynadka, N. C., Science 2007, 315, 1402-1405. (11) Yuan, Y.; Fuse, S.; Ostash, B.; Sliz, P.; Kahne, D.; Walker, S., ACS Chem. Biol. 2008, 3, 429-436. (12) Sung, M. T.; Lai, Y. T.; Huang, C. Y.; Chou, L. Y.; Shih, H. W.; Cheng, W. C.; Wong, C. H.; Ma, C., Proc. Natl. Acad. Sci. U. S. A. 2009, 106, 8824-8829. (13) Huang, C. Y.; Shih, H. W.; Lin, L. Y.; Tien, Y. W.; Cheng, T. J.; Cheng, W. C.; Wong, C. H.; Ma, C., Proc. Natl. Acad. Sci. U. S. A. 2012, 109, 6496-6501. (14) Cheng, T. J.; Wu, Y. T.; Yang, S. T.; Lo, K. H.; Chen, S. K.; Chen, Y. H.; Huang, W. I.; Yuan, C. H.; Guo, C. W.; Huang, L. Y.; Chen, K. T.; Shih, H. W.; Cheng, Y. S.; Cheng, W. C.; Wong, C. H., Bioorg. Med. Chem. 2010, 18, 8512-8529. (15) Huang, S. H.; Wu, W. S.; Huang, L. Y.; Huang, W. F.; Fu, W. C.; Chen, P. T.; Fang, J. M.; Cheng, W. C.; Cheng, T. J.; Wong, C. H., J. Am. Chem. Soc. 2013, 135, 17078-17089. (16) Cheng, T. J.; Sung, M. T.; Liao, H. Y.; Chang, Y. F.; Chen, C. W.; Huang, C. Y.; Chou, L. Y.; Wu, Y. D.; Chen, Y. H.; Cheng, Y. S.; Wong, C. H.; Ma, C.; Cheng , W. C., Proc. Natl. Acad. Sci. U. S. A. 2008, 105, 431-436. (17) Aoyagi, T.; Aoyama, T.; Kojima, F.; Matsuda, N.; Maruyama, M.; Hamada, M.; Takeuchi, T., J. Antibiot. (Tokyo) 1992, 45, 1385-1390. (18) Fukushima, K.; Ishiwata, K.; Kuroda, S.; Arai, T., J. Antibiot. (Tokyo) 1973, 26, 65-69.
Chi-Huey Wong: 0000-0002-9961-7865
Notes The authors declare no competing financial interests.
ACKNOWLEDGMENT We thank Dr. Shih-Hsien Huang for providing FBLA and Dr. Lin-Ya Huang for providing NBD-lipid II. This work was supported by Academia Sinica and Ministry of Science and Technology [MOST 106-0210-01-15-02, MOST 107-0210-01-19-01].
REFERENCES (1) Sauvage, E.; Kerff, F.; Terrak, M.; Ayala, J. A.; Charlier, P., FEMS Microbiol. Rev. 2008, 32, 234-258. (2) Ostash, B.; Walker, S., Curr. Opin. Chem. Biol. 2005, 9, 459-466. (3) Halliday, J.; McKeveney, D.; Muldoon, C.; Rajaratnam, P.; Meutermans, W., Biochem. Pharmacol. 2006, 71, 957-967. (4) Galley, N. F.; O'Reilly, A. M.; Roper, D. I., Bioorg. Chem. 2014, 55, 16-26. (5) Sauvage, E.; Terrak, M., Antibiotics (Basel) 2016, 5. (6) Koehn, F. E.; Carter, G. T., Nat. Rev. Drug Discov. 2005, 4, 206220. (7) Dias, D. A.; Urban, S.; Roessner, U., Metabolites 2012, 2, 303-336. (8) Newman, D. J.; Cragg, G. M., J. Nat. Prod. 2016, 79, 629-661. (9) Ostash, B.; Walker, S., Nat. Prod. Rep. 2010, 27, 1594-1617.
4
ACS Paragon Plus Environment
Page 5 of 5 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Journal of the American Chemical Society
TOC
5
ACS Paragon Plus Environment
