Arsinothricin, an Arsenic-Containing Non-Proteinogenic Amino Acid Analog of Glutamate, Is a Broad-Spectrum Antibiotic
Total Page:16
File Type:pdf, Size:1020Kb
ARTICLE https://doi.org/10.1038/s42003-019-0365-y OPEN Arsinothricin, an arsenic-containing non-proteinogenic amino acid analog of glutamate, is a broad-spectrum antibiotic Venkadesh Sarkarai Nadar1,7, Jian Chen1,7, Dharmendra S. Dheeman 1,6,7, Adriana Emilce Galván1,2, 1234567890():,; Kunie Yoshinaga-Sakurai1, Palani Kandavelu3, Banumathi Sankaran4, Masato Kuramata5, Satoru Ishikawa5, Barry P. Rosen1 & Masafumi Yoshinaga1 The emergence and spread of antimicrobial resistance highlights the urgent need for new antibiotics. Organoarsenicals have been used as antimicrobials since Paul Ehrlich’s salvarsan. Recently a soil bacterium was shown to produce the organoarsenical arsinothricin. We demonstrate that arsinothricin, a non-proteinogenic analog of glutamate that inhibits gluta- mine synthetase, is an effective broad-spectrum antibiotic against both Gram-positive and Gram-negative bacteria, suggesting that bacteria have evolved the ability to utilize the per- vasive environmental toxic metalloid arsenic to produce a potent antimicrobial. With every new antibiotic, resistance inevitably arises. The arsN1 gene, widely distributed in bacterial arsenic resistance (ars) operons, selectively confers resistance to arsinothricin by acetylation of the α-amino group. Crystal structures of ArsN1 N-acetyltransferase, with or without arsinothricin, shed light on the mechanism of its substrate selectivity. These findings have the potential for development of a new class of organoarsenical antimicrobials and ArsN1 inhibitors. 1 Department of Cellular Biology and Pharmacology, Florida International University, Herbert Wertheim College of Medicine, Miami, FL 33199, USA. 2 Planta Piloto de Procesos Industriales Microbiológicos (PROIMI-CONICET), Tucumán T4001MVB, Argentina. 3 SER-CAT and Department of Biochemistry and Molecular Biology, University of Georgia, Athens, GA 30602, USA. 4 Berkeley Center for Structural Biology, Lawrence Berkeley Laboratory, Berkeley, CA 94720, USA. 5 Division of Hazardous Chemicals, National Institute for Agro-Environmental Sciences, NARO, Tsukuba, Ibaraki 305-8604, Japan. 6Present address: Manchester Institute of Biotechnology, School of Chemistry, University of Manchester, 131 Princess Street, Manchester M1 7DN, UK. 7These authors contributed equally: Venkadesh Sarkarai Nadar, Jian Chen, Dharmendra S. Dheeman. Correspondence and requests for materials should be addressed to B.P.R. (email: brosen@fiu.edu) or to M.Y. (email: myoshina@fiu.edu) COMMUNICATIONS BIOLOGY | (2019) 2:131 | https://doi.org/10.1038/s42003-019-0365-y | www.nature.com/commsbio 1 ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-019-0365-y rsenic is the most pervasive environmental toxic element1. highly toxic pentavalent organoarsenical. It is chemically unre- AHere we describe how bacteria harness arsenic to create a lated to other organoarsenicals and has the potential to be the potent broad-spectrum antibiotic. New antibiotics are progenitor of a new class of organoarsenical antibiotics. With urgently needed because the emergence of resistance has rendered every new antibiotic, resistance inevitably arises. The enzyme PPT nearly every clinically used antibiotic ineffectual. Human tuber- N-acetyltransferase (PAT) confers resistance to PPT by acetylat- culosis, the top global infectious disease killer, which is caused by ing its α-amino group. A curious observation has been that many Mycobacterium tuberculosis, has become even more difficult to arsenic resistance (ars) operons have an arsN1 gene that encodes treat due to the drug resistance2. The World Health Organization a pat ortholog. Why an enzyme for PPT resistance should be in declares multidrug-resistant tuberculosis a global public health an ars operon was a mystery. The identification of AST as a crisis, calling for a pressing need for development of new and natural product suggested that the biological function of ArsN innovative antibiotics3. In addition to M. tuberculosis, the World could be to act as an AST resistance. Here we show that ArsN1 Health Organization recently issued a global priority pathogen list acetylates both AST and PPT but with higher affinity for AST, of antibiotic-resistant bacteria that pose the greatest threat to indicating that ArsN1 is an AST-selective N-acetyltransferase. human health to guide and promote research and development of We crystallized ArsN1 and solved the apo and substrate-bound new antibiotics4. structures. This knowledge can be utilized to design new and The use of arsenicals as antimicrobial and anticancer agents is novel drugs that evade or inhibit resistance mechanisms. well-established5,6. The first synthetic antimicrobial agents were the organoarsenicals atoxyl (p-aminophenylarsenate, also known Results as p-arsanilic acid) and salvarsan (arsphenamine). While sal- AST is a broad-spectrum antibiotic. To determine whether AST varsan is no longer in clinical use, the organoarsenical melarso- has antibiotic activity, we examined its ability to inhibit growth of prol, developed in 1949, is still recommended by the World bacteria using environmental isolates. AST was equally effective Health Organization for treatment of second-stage Trypanosoma 7 against both Gram-negative and Gram-positive bacteria (Fig. 2a). brucei sleeping sickness . Atoxyl and the related synthetic aro- Each species was inhibited to the same degree by 25 μM AST and matic arsenicals roxarsone (4-hydroxy-3-nitrophenylarsenate) 400 μM L-PPT, except for B. gladioli GSRB05 and Pseudomonas and nitarsone (4-nitrophenylarsenate) are antimicrobials used for 10 8 putida KT2440. B. gladioli GSRB05 is the producer of AST ,soit the prevention of Coccidia and Histomonas infections in poultry . is not unexpected that this strain might be resistant to the anti- Although no longer in wide use in the United States, roxarsone is biotic it produces. As discussed below, the arsN1 gene confers still produced and utilized worldwide. Finally, arsenic trioxide resistance in P. putida KT2440. Our results demonstrate that AST is currently the treatment of choice in humans for all-trans 9 is a broad-spectrum antibiotic effective against both Gram- retinoic acid unresponsive acute promyelocytic anemia . negative and Gram-positive bacteria. In Escherichia coli, AST is Here we demonstrate that a recently discovered arsenic- considerably more inhibitory than inorganic As(III) and is similar containing natural product, arsinothricin (2-amino-4-(hydro- to that of highly toxic trivalent methylarsenite (MAs(III)) xymethylarsinoyl)butanoate, AST) (Fig. 1a), produced by the rice 10 (Fig. 2b). Given that, in general, pentavalent arsenicals are rela- rhizosphere microbe Burkholderia gladioli GSRB05 , has broad- tively benign and much less toxic compared to trivalent species11, spectrum antibiotic activity. Biosynthetic AST is a mimetic of the this is a striking result. To our best knowledge, except thiolated Streptomyces antibiotic L-phosphinothricin (2-amino-4-(hydro- species6, AST is the only known pentavalent arsenic species that xymethylphosphinyl)butanoate or L-PPT) with an arsenic in exhibits high toxicity. place of the phosphorus of L-PPT (Fig. 1b). L-AST and L-PPT are non-proteinogenic amino acid analogs of L-glutamate (Fig. 1c) and act through inhibition of glutamine synthetase. Most toxic AST inhibits glutamine synthetase. The mechanism of action of arsenicals contain trivalent As(III). AST is unusual in being a L-PPT is irreversible inhibition of bacterial glutamine synthe- tase12. L-PPT also inhibits plant glutamine synthetase, which is the basis for its use as the broad-spectrum systemic herbicide O O Glufosinate13. Because of the structural similarity with PPT a As (Fig. 1b), it was reasonable to propose that the target of AST is AST O– – O also bacterial glutamine synthetase. We compared the effect of + fi NH3 AST and L-PPT on puri ed E. coli glutamine synthetase activity. The Km of glutamine synthetase was found to be 2.7 ± 0.6 mM for O O L-glutamate, consistent with the previous determination12. The μ b P observed Ki values for AST and L-PPT are 0.3 ± 0.05 M and PPT O– – O 0.4 ± 0.15 μM, respectively, indicating that AST is as effective an + NH3 inhibitor of glutamine synthetase as is L-PPT. O O AST is an effective antibiotic with pathogenic bacteria. Inhi- c bition of glutamine synthetase has been proposed to be a Glutamate – O O – potential therapeutic strategy against tuberculosis14. Pathogenic + NH3 mycobacteria, including M. tuberculosis, secrete large amounts of an extracellular glutamine synthetase that is involved in synthesis α O O of the poly- -L-glutamine layer, a cell wall component that is found exclusively in pathogenic strains and considered essential d S MSO O – to their virulence15. In fact, L-methionine S-sulfoximine HN + (L-MSO) (Fig. 1d), the first glutamine synthetase inhibitor NH3 described16, effectively inhibits M. tuberculosis growth both Fig. 1 Chemical structure of glutamate and analogs. a Arsinothricin in vitro and in vivo15,17. To examine the potential of AST as a (AST); b phosphinothricin (PPT); c glutamate; d methionine sulfoximine drug for tuberculosis, we analyzed the effect of AST on a related (MSO) pathogenic strain, M. bovis BCG, and compared it with L-PPT 2 COMMUNICATIONS BIOLOGY | (2019) 2:131 | https://doi.org/10.1038/s42003-019-0365-y | www.nature.com/commsbio COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-019-0365-y ARTICLE ab 1.4 1.4 Gram-negative Gram-positive ) 1.2 1.2 600nm 1.0 ) 1.0 0.8 600nm 0.8 0.6 0.6 0.4 Growth (A 0.4 Growth (normalized A 0.2 0.2 0.0 0.0 01020304050607080 W3110 [Arsenic] μM KT2440 GSRB05 E. coli Bacillus cereus P. putida Bacillus megaterium Sinorhizobium meliloti Shewanella putrefaciens Burkholderia gladioli Corynebacterium glutamicum cd5 Control 1.4 10 μM L-PT ) 1.2 4 100 μM L-PT 10 μM AST 600nm ) 1.0 3 600nm 20 μM AST 0.8 0.6 2 10 μM L-MSO Growth (A 0.4 50, 100 μM L-MSO 1 100 μM AST μ 20 M L-MSO Growth (normalized A 0.2 50 μM AST 0 0.0 Control AST L-PPT L-MSO 01234 Week e 1.2 ) 1.0 570nm 0.8 0.6 0.4 0.2 Growth (normalized A 0.0 0 20 40 60 80 100 [Arsenic] μM Fig.