Date Published: April 23, 2019
Publisher: Public Library of Science
Author(s): Santosh K. Srivastava, Kelcey S. King, Nader F. AbuSara, Chelsea J. Malayny, Brandon M. Piercey, Jaime A. Wilson, Kapil Tahlan, Pradeep Kumar.
In Streptomyces clavuligerus, the gene cluster involved in the biosynthesis of the clinically used β-lactamase inhibitor clavulanic acid contains a gene (orf12 or cpe) encoding a protein with a C-terminal class A β-lactamase-like domain. The cpe gene is essential for clavulanic acid production, and the recent crystal structure of its product (Cpe) was shown to also contain an N-terminal isomerase/cyclase-like domain, but the function of the protein remains unknown. In the current study, we show that Cpe is a cytoplasmic protein and that both its N- and C-terminal domains are required for in vivo clavulanic acid production in S. clavuligerus. Our results along with those from previous studies allude towards a biosynthetic role for Cpe during the later stages of clavulanic acid production in S. clavuligerus. Amino acids from Cpe essential for biosynthesis were also identified, including one (Lys89) from the recently described N-terminal isomerase-like domain of unknown function. Homologues of Cpe from other clavulanic acid-producing Streptomyces spp. were shown to be functionally equivalent to the S. clavuligerus protein, whereas those from non-producers containing clavulanic acid-like gene clusters were not. The suggested in vivo involvement of an isomerase-like domain recruited by an ancestral β-lactamase related protein, supports a previous hypothesis that Cpe could be involved in a step requiring the opening and modification of the clavulanic acid core during its biosynthesis from 5S precursors.
The β-lactam class of antibiotics have broad-spectrum activity and include some of the most commonly prescribed agents used for treating bacterial infections [1–3]. They have a long history of use in medicine, but as with other antibiotics, the emergence of resistance is a major problem [3–5]. There are several mechanisms responsible for β-lactam resistance, which include the production of secreted β-lactamases, enzymes that hydrolyze and inactivate certain members of this antibiotic class [6, 7]. Combinations of β-lactamase inhibitors such as clavulanic acid along with β-lactam antibiotics are often used as a strategy for treating some infections caused by β-lactamase-producing antibiotic resistant bacteria [8, 9]. Clavulanic acid belongs to the clavam family of specialized metabolites and it irreversibly inhibits class A β-lactamases, thereby restoring the activity of β-lactam antibiotics against target organisms in such combinations [10, 11]. The activity of clavulanic acid is attributed in part to its 3R,5R stereochemistry, as other naturally occurring clavams have a 5S configuration (collectively referred to as the 5S clavams) and do not inhibit β-lactamases [8, 12]. Commercial production of clavulanic acid is achieved by fermenting Streptomyces clavuligerus, and a cluster of ∼18 genes referred to as the clavulanic acid biosynthetic gene cluster (CA-BGC) encodes components of the core biosynthetic pathway . It has previously been reported that Streptomyces jumonjinensis and Streptomyces katsurahamanus also produce clavulanic acid, but the sequences of their respective CA-BGCs are not available [12, 14]. On the other hand, the genome sequences of organisms such as Streptomyces flavogriseus (ATCC 33331, also known as S. pratensis) and Saccharomonospora viridis (DSM 43017) contain gene clusters closely resembling the S. clavuligerus CA-BGC, but neither has been shown to produce the metabolite to date [13, 15]. In addition, S. clavuligerus is somewhat unique among clavulanic acid producers as it also produces certain 5S clavams as products of a pathway related to clavulanic acid [13, 16]. Clavulanic acid and the 5S clavams have common biosynthetic origins and the pathway involved in their production can be roughly divided into two parts in S. clavuligerus (Fig 1). The “early” steps leading up to the intermediate clavaminic acid are shared during the production of both types of metabolites, with all intermediates possessing 5S configuration . Beyond clavaminic acid (also a 5S clavam) the pathway diverges into specific “late” steps leading to either the 5S clavams or to clavulanic acid (Fig 1) .
In the current study, we examined the function of cpe from the CA-BGC of S. clavuligerus, starting with the significance of the relative arrangement of neighboring genes located in its immediate vicinity. Polycistronic mRNAs often allow for the concerted expression of gene products involved in related biosynthetic pathways [44, 45], and gene knockout studies have implicated cpe as being essential for clavulanic acid production in S. clavuligerus [21, 23]. cpe is transcribed as part of a bicistronic operon along with orf13, the start codon of which also overlaps with the stop codon of cpe (Fig 2A), suggesting potential co-translation [46–48]. In addition, the 3′ ends of orf13 and orf14 overlap (Fig 2A), which is unusual in bacteria . It can be challenging to decipher the precise roles of genes located within operons, particularly in cases where co-translation is involved [50–53]. The disruption of genes located in the 5′ regions of operons can influence the expression of downstream genes and also impact the relative stoichiometry of encoded gene products, thereby leading to polar effects [47, 51, 54]. Therefore, we prepared an in-frame S. clavuligerus cpe deletion mutant for use in the current study, while maintaining its stop codon and context with orf13 to minimize the potential for polar effects. During the process, we also prepared the S. clavuligerus Δcpe::apra deletion mutant in which a disruption cassette was inserted in the opposite orientation to cpe transcription. It was noted that both the in-frame and the insertional mutant could be successfully complemented to restore clavulanic acid production using a plasmid-borne copy of cpe, demonstrating that polar effects were not associated with either of them. Therefore, it seems that despite their relative organization, alternate mechanisms exist to facilitate the translation of Orf13 in the cpe mutants, enabling us to use the in-frame mutant for more detailed in vivo studies.