Chemistry & Biology Article N-Carbamoylation of 2,4-Diaminobutyrate Reroutes the Outcome in Padanamide Biosynthesis Yi-Ling Du,1 Doralyn S. Dalisay,1 Raymond J. Andersen,1,2 and Katherine S. Ryan1,* 1Department of Chemistry 2Department of Earth, Ocean and Atmospheric Sciences University of British Columbia, Vancouver, BC V6T 1Z1, Canada *Correspondence: [email protected] http://dx.doi.org/10.1016/j.chembiol.2013.06.013 SUMMARY literature. It is interesting that a compound identical to padana- mide A, named actinoramide A (Nam et al., 2011), was indepen- Padanamides are linear tetrapeptides notable for the dently reported from Streptomyces sp. CNQ-027. This actinomy- absence of proteinogenic amino acids in their struc- cete strain was isolated from sediment on the opposite side of the tures. In particular, two unusual heterocycles, (S)- Pacific Ocean, near San Diego, CA, suggesting a potentially wide 3-amino-2-oxopyrrolidine-1-carboxamide (S-Aopc) distribution of the padanamides/actinoramides. It is intriguing and (S)-3-aminopiperidine-2,6-dione (S-Apd), are that, whereas padanamide A and actinoramide A are identical, found at the C-termini of padanamides A and B, the minor compounds (actinoramides B and C) co-isolated from Streptomyces sp. CNQ-027 are unique (Figure 1A). Total respectively. Here we identify the padanamide synthesis of padanamides A and B was recently reported (Long biosynthetic gene cluster and carry out systematic et al., 2013), confirming the previous structural elucidations. gene inactivation studies. Our results show that The padanamides attracted our attention for their many padanamides are synthesized by highly dissociated unusual chemical features. In particular, the (S)-3-amino-2-oxo- hybrid nonribosomal peptide synthetase/polyketide pyrrolidine-1-carboxamide (S-Aopc) and (S)-3-aminopiperidine- synthase machinery. We further demonstrate that 2,6-dione (S-Apd) residues at the C termini of padanamides A carbamoyltransferase gene padQ is critical to the for- and B, respectively, are unprecedented in naturally occurring mation of padanamide A but dispensable for biosyn- peptides and lack a clear biosynthetic origin. Other unusual thesis of padanamide B. Biochemical investigations structural features for padanamides include an N-terminal show that PadQ carbamoylates the rare biosynthetic methoxyacetyl group and a 4-amino-3-hydroxy-2-methyl- 5-phenylpentanoic acid (Ahmpp) residue. 3-hydroxyleucine precursor L-2,4-diaminobutyrate, generating L-2- (Hleu) and piperazate (Piz) are also rare, nonproteinogenic amino-4-ureidobutyrate, the presumed precursor to peptidic building blocks. Padanamide B displays cytotoxicity the C-terminal residue of padanamide A. By contrast, against Jurkat T lymphocyte cells, with a half maximal inhibitory the C-terminal residue of padanamide B may derive concentration value of 20 mg/ml, while padanamide A is roughly from glutamine. An unusual thioesterase-catalyzed 3-fold less active in the same assay (Williams et al., 2011). A cyclization is proposed to generate the S-Aopc/S- chemical genomics analysis using yeast deletion mutants sug- Apd heterocycles. gests that padanamide A is involved in the inhibition of cysteine and methionine biosynthesis or that these amino acids are involved in the yeast’s response to the peptide. By contrast, INTRODUCTION padanamide B does not exhibit the same activity at the concen- tration where padanamide A is active (Williams et al., 2011). As Discovery of novel compounds with pharmaceutical potential padanamides A and B differ only in their corresponding C-termi- from marine sources has received increased attention in recent nal residues, these biological data suggest that the unusual years (Gerwick and Moore, 2012). The marine actinomycetes, C-terminal residues of padanamides are quite important to their in particular, have emerged as a promising resource, producing biological activities. Even more intriguing, the positioning of the novel natural products with unusual chemical structures and S-Aopc and S-Apd residues at the C termini of the correspond- various biological activities (Hughes and Fenical, 2010). ing peptides suggests that elucidation of the responsible genes Recently, Streptomyces sp. RJA2928 (Williams et al., 2011), a and enzymes might enable new strategies for combinatorial strain isolated from sediment in the Pacific Ocean, near Padana biosynthesis of linear peptides with new C-terminal modifica- Nahua in Papua New Guinea, has emerged as an exciting source tions (Menzella and Reeves, 2007). for new molecules. The strain produces a series of polyketides, Here, we report the biosynthetic gene cluster of padanamides including the nahuoic acids (Williams et al., 2013), along with from Streptomyces sp. RJA2928. In our work, systematic gene padanamides A and B (Williams et al., 2011), which are highly inactivation mutants are used to delineate the padanamide modified linear tetrapeptides (Figure 1A). The padanamides gene cluster. Our results reveal a hybrid nonribosomal peptide lack any proteinogenic amino acids and possess substructures synthetase/polyketide synthase (NRPS/PKS) system for assem- that have not been previously reported in the natural products bly of the padanamide skeleton, along with a number of enzymes 1002 Chemistry & Biology 20, 1002–1011, August 22, 2013 ª2013 Elsevier Ltd All rights reserved Chemistry & Biology Padanamide Biosynthetic Gene Cluster A O O NH2 CH3 O O N N H OH O H OH O HN N HN N N N H N H N N O H N O H O O O O O O HO HO Padanamide A / Actinoramide A Actinoramide B (MAJOR PRODUCT) H O N O OH H OH O H HN N HN N N N H N H N O H N O O O O O O O HO HO Padanamide B Actinoramide C B C i PadG PadA1 PadA2 PadB PadC PadD iv PadE Padanamide A / * C C A T C A T KS AT KR T C A T TE T T A T E Actinoramide A S O O O S O S S O S O S O ? O H OH O HN N O N O H N H N Me N MeO O 2 HN 2 H H OH Me N O Me O- Me HO HN O O NH O HO O NH2 ii PadG PadG PadG PadE T HO TT O- C A T TE P Padanamide B O O SH PadM HO O O S O S O P Other O O PadK H OH O - O OH enzymes? O O PadF HO MeO HN N OH N N - H Glutamine O Me N O H O O Me O O O O NH2 PadQ H HO H2N H N N OH 2 OH NH2 O NH2 L-2,4-diaminobutyrate (Dab) L-2-amino-4-ureidobutyrate (Aub) C A T TE O O iii H OH O Actinoramide B O O O HN N H N EctB EctA H N H H2N N H OH OH OH Me N O Me O O NH O NH NH 2 2 O NH2 O HO L-aspartate L-2,4-diamino- N-acetyl-L-2,4- O CH3 4-semi-aldehyde butyrate (Dab) diaminobutyrate Figure 1. Padanamide Biosynthetic Pathway (A) Structures of padanamides and actinoramides. (B) Padanamide biosynthetic gene cluster. (C) Proposed biosynthetic route to the padanamides and actinoramides. (i) Hybrid NRPS/PKS assembly of the N-terminal aa of the padanamides. (ii) Biosynthesis of methoxymalonyl-PadG (top) and L-Aub (bottom). (iii) Biosynthesis of L-Dab and N-acetyl-L-2,4-diaminobutyrate from L-aspartate-4-semialdehyde. (iv) For- mation of padanamide A, padanamide B, and actinoramide B using TE-catalyzed cyclization with amide nitrogens as nucleophiles. See also Figure S1 and Table S2. believed to be involved in the formation of unusual building of L-2,4-diaminobutyrate (L-Dab) to L-2-amino-4-ureidobutyrate blocks. Notably, further in vitro biochemical characterization re- (L-Aub), which is a key step in the biosynthesis of the S-Aopc res- veals a carbamoyltransferase that catalyzes the transformation idue of padanamide A. Chemistry & Biology 20, 1002–1011, August 22, 2013 ª2013 Elsevier Ltd All rights reserved 1003 Chemistry & Biology Padanamide Biosynthetic Gene Cluster Table 1. Deduced Functions of ORFs of the Padanamide line from six precursors: (1) 2-methoxyacetate, (2) 3-OH-D- Biosynthetic Gene Cluster leucine, (3) Piz, (4) phenylalanine, (5) methylmalonate, and (6) Size Homolog and Identity/ an unknown molecule that is ultimately converted to the Gene (aa) Similarity (%)a Proposed Function S-Aopc residue. We thus hypothesized that the padanamide orf(À2) 3,363 SBI_09160 (YP_ lantibiotic biosynthesis gene cluster would contain genes for NRPSs and PKSs, along 004967409), 78/82 protein LanM with accessory genes. Accessory genes would be involved in the construction of the 2-methoxyacetyl group and might orf(À1) 171 SBI_09159 (YP_ putative type 2 004967408), 75/83 lantibiotic resemble the starter unit pathway for apoptolidins (Du et al., 2011). Additionally, an ornithine N-monooxygenase would likely padF 390 – unknown generate N5-OH-ornithine, the precursor to Piz, as observed in padG 104 Npun_F3159 (YP_ discrete ACP the biosynthesis of the kutznerides (Fujimori et al., 2007), sangli- 001866560), 36/58 fehrin A (Qu et al., 2011), himastatin (Ma et al., 2011), and other padH 316 QncN (AFJ11257), dehydrogenase complex, Piz-containing natural products. 65/73 E1 component, alpha To find the padanamide gene cluster, we initially explored subunit isolation of adenylation (A) domain and N-oxygenase genes padI 348 QncL (AFJ11255), dehydrogenase complex, from RJA2928 genomic DNA using PCR with degenerate 55/68 E1 component, beta subunit primers. However, we found that Streptomyces sp. RJA2928 genomic DNA contained many distinct A domain and N-oxygen- padJ 409 Tsn2 (ACR50769), dehydrogenase complex, ase genes, making this approach inefficient. We thus turned to 40/53 E2 component genome scanning, assembling 11.3 megabase pairs of nonre- padK 578 PAV_7c00910 (ZP_ FkbH-like protein dundant sequences over 862 contigs with 236-fold read 10866675), 29/44 coverage.