The Gram-positive strain Amycolatopsis mediterranei S699 is a member of the phylum Actinomycetota and capable of producing rifamycin B which belongs to a group of ansamycin antibiotic polyketides built by modular polyketide synthases type-I (PKS-I) (Kim et al., 1996; August et al., 1998; Floss and Yu, 1999; Floss et al., 2011; Kisil et al., 2021). Rifamycin provides the structural foundation for semisynthetic processes that lead to clinically important drugs, in particular rifampicin, which is used as a part of the therapy regime for curing tuberculosis, one of the world’s most infectious diseases caused by Mycobacterium tuberculosis (Global Tuberculosis Report 2023). Rifamycin binds to the β-subunit of the DNA-dependent RNA-Polymerase and acts bactericidal by suppressing mRNA synthesis (Wehrli et al., 1968). These derivatives have a high bioavailability in comparison to rifamycin B (Maggi et al., 2009). Rifamycin B itself has a low bioavailability and can only be used for traveler’s diarrhea (Di Stefano et al., 2011). Its complex structure allows limited structural changes which subsequently led to six clinically effective semisynthetic derivatives. Further modifications did not result in clinically usable structures (Sood et al., 2024). Semi-synthetic modifications of rifamycin B are primarily done at position -3 of the naphthoquinone ring and result in more lipophilic structures with increased bioavailability and wide-ranging indications (Cannata and Tamagnone, 1985; Maggi et al., 2009). In addition to their use against Mycobacterium tuberculosis these derivatives have also been used against leprosy caused by M. leprae as well in AIDS-related mycobacterial infection and treatment of traveler’s disease caused by Escherichia coli (Rees et al., 1970; Waters et al., 1978; Alvisi et al., 1987). Rifampicin was introduced into the market in 1968 and became a first-line drug within the therapeutic scheme against M. tuberculosis infections (Sensi, 1983; Pozniak et al., 1999). As with many antibiotics, multi-drug resistance strains of M. tuberculosis emerged in TB patients due to inadequate medical supervision, incorrect use, and inadequate compliance (Global Tuberculosis Report 2023). Since semi-synthetical changes of rifamycin B only led to a few clinically used rifamycin derivatives, new approaches were developed to gain additional rifamycin derivatives and overcome the problem posed by multidrug-resistant strains (MDR) of mycobacteria.

One approach is to directly manipulate the rifamycin polyketide biosynthetic gene cluster of the rifamycin B-producing strain Amycolatopsis mediterranei S699 resulting in new structural derivatives of rifamycin B, which then can also be converted into further semi-synthetical structures (Tang et al., 2012; Nigam et al., 2014). In practice, this could be done by combinatorial biosynthesis of the modular rifamycin PKS-I using homologous recombination (Lal et al., 2000). Rifamycin B is built by a modular PKS-I which consists of a loading module using 3-amino-5-hydroxy benzoic acid (AHBA) as a starter unit followed by the addition of two acetates and eight propionates by the ten extension modules (Floss and Yu, 1999; Kato et al., 2002; Floss et al., 2011). These modules are responsible for the successive polyketide chain assembly by either methyl malonyl- or malonyl-CoA. The modules differ in their composition of domains resulting in different functional groups for each of these C₂-extender units. Responsible for the chain extension is the keto synthase (KS), acyltransferase (AT), and acyl-carrier protein (ACP) domain of each module. Depending on the composition of the module they are additionally accompanied by a dehydrogenase (DH) or keto reductase (KS) (Khosla et al., 1999). The use of combinatorial biosynthesis of ery PKS (erythromycin polyketide biosynthetic gene cluster) that leads to new erythromycin analogues has been amply demonstrated in Saccharopolyspora erythreaea (Staunton, 1998; Lü et al., 2020). In ery PKS the 6-deoxyerythronolide b synthase (DEBS) is equal to the rifamycin PKS-I and confirmed that DEBS domains can be swapped with each other, modified, or exchanged to domains of different PKS-I biosynthesis clusters (McDaniel et al., 1999). The rifamycin B biosynthetic gene cluster of Amycolatopsis mediterranei S699 was shown to be more rigid to combinatorial approaches (Floss, 2006). It could finally be manipulated resulting in the 24-desmethyl rifamycin B-producing strain DCO36 ( Figure 1 ) by exchange of rifAT6 with rapAT2 (Nigam et al., 2014; Sood et al., 2024). In Lal and his coworker’s approach malonyl-CoA (rapAT2) was used as an extender unit instead of the original incorporation of methyl malonyl-CoA (rifAT6). Besides 24-desmethyl rifamycin B DCO36 still produces rifamycin B due to the assumingly broad selectivity of the new AT domain ( Figure 2B ). However, by using an antibacterial assay against rifampicin-sensitive and resistant strains of M. tuberculosis they were able to show that the novel 24-desmethyl rifamycin B when converted into 24-desmethyl rifampicin or 24-desmethyl rifamycin S has strong antibacterial activity. This result showed promising potential for the commercial use of these compounds. The creation of such a new derivative comes at the expense of a lower yield compared to the wild-type Amycolatopsis mediterranei S699. While industrial strains like N. mediterranei N813 can produce up to 24 g/L, these newly developed mutant strains have an antibiotic yield of 2–20 mg/L ( Figure 1 ) (Lal et al., 1995; Jin et al., 2002; Nigam et al., 2014; Sood et al., 2024).

The biosynthetic gene cluster of rifamycin also contains genes encoding regulatory proteins. The deduced amino acid sequence of rifO is putatively involved in the regulation of the production of rifamycin by the production of the B-factor (3’-(1-butyl phosphoryl) adenosine) ( Supplementary Figure 4 ) (Kawaguchi et al., 1984, 1988; Azuma et al., 1990). RifO shows similarity to 2’,3’-cyclic nucleotide 2”-phosphodiesterases. The B-factor occurs naturally and has been isolated from yeast extracts. It was shown that an external addition of the B-factor to the media can stimulate rifamycin B synthesis. This effect was already noticeable at an extremely low concentration of 10 ng/mL (Kawaguchi et al., 1984). It was mentioned by Azuma et al. that rifamycin biosynthesis was inhibited in the absence of rifO. Therefore, we decided to investigate whether RifO is important for rifamycin biosynthesis in Amycolatopsis mediterranei S699 and whether rifO can be used in DCO36 to increase 24-desmethyl rifamycin B production.

The protein encoded by rifQ is part of the feedback regulatory system of the rifamycin biosynthetic cluster ( Figure 2 ) (Lei et al., 2018; Singhvi et al., 2021). The expression of rifQ leads to RifQ which is present as a homodimer (46.9 kDa) within the cell. RifQ is part of the TetR family and its overall structure can be broken down into two DNA-binding domains on each monomer and a regulatory core, which is responsible for the rifamycin recognition and dimerization by making hydrophobic contacts within the regulatory core (Cuthbertson and Nodwell, 2013). RifQ acts in the early growth phase as a repressor for the expression of rifP encoding a transmembrane transporter of the major facilitator superfamily (MFS) with 14 transmembrane domains (53 kDa) (Absalón et al., 2007; Drew et al., 2021; Singhvi et al., 2021). The hydrogen/drug antiporter (H⁺/Rifamycin) functions as an efflux pump via an electrochemical proton gradient. Rifamycin can remove repression in a later growth phase by binding to the RifQ-homodimer. A certain threshold of intracellular rifamycin concentration leads to the binding of two rifamycin molecules to the RifQ-homodimer resulting in a conformational change of RifQ. RifQ in the form of [Rifamycin-RifQ] complex loses its function as a repressor of rifP expression. As a consequence, rifP is expressed already during the early growth phase which leads subsequently to a higher overall yield of rifamycin.

Previous studies showed that the removal of rifQ in Amycolatopsis mediterranei leads to a higher antibiotic yield in comparison to the wild-type (Lei et al., 2018). Also, it was suggested that this approach may lead to a positive effect in mutant strains (Singhvi et al., 2021). More specifically, in silico experiments displayed that 24-desmethyl rifamycin B has a 14 times weaker binding affinity to RifQ than rifamycin B. As a consequence, expression of rifP is inhibited more intensely in the 24-desmethyl rifamycin B-producing mutant resulting in low production.

Here we report, besides our studies on rifO (see above) the generation and characterization of a rifQ mutant in DCO36.

This study demonstrates the adaptation of the modified PCR-targeted Streptomyces gene replacement method to remove or exchange genes in Amycolatopsis mediterranei S699 and its mutant strains like the 24-desmethyl rifamycin B-producing strain DCO36 ( Figure 3 ) (Gust et al., 2003). For Streptomyces species a wide range of suitable cloning vectors and transformation methods are well established (Kieser et al., 2000). Nonetheless, the adaptation of these methods and plasmids to Amycolatopsis mediterranei S699 is an extremely difficult process that does not always necessarily lead to the desired success (Lal et al., 1991). The development of suitable plasmids is also a very time-consuming procedure (Tuteja et al., 2000; Dhingra et al., 2003). The combinative use of the suicidal vector pKGLP2 and the replicative and consecutive PermE promoter-containing plasmids pUWL-H (respectively pUWL-HA), shown in this study, enables a reliably quick way to switch off, exchange, or specifically overproduce genes (Petzke, 2010; Myronovskyi et al., 2011). Furthermore, the results indicate the potential that the successful implementation of the modified PCR-targeted Streptomyces gene replacement method and use of the mentioned plasmids will have on previous time-consuming and problematic approaches to manipulate rifamycin polyketide synthase gene cluster or the modification of regulatory elements of Amycolatopsis mediterranei strains. The adaptation of the method and plasmids led subsequently to the presented results in this study.

This study investigated whether, for the 24-desmethyl rifamycin B derivative-producing strain DCO36, the removal of rifQ leads to increased production of rifamycin B derivative. After the removal of rifQ, the respective strain can produce and integrate the transmembrane transporter RifP into the membrane already at the early growth phase. This can lead to a higher yield of 24-desmethyl rifamycin B as the intracellular concentration does not need to reach the threshold concentration to inactivate RifQ.

This study also investigated the role of RifO and its involvement in the regulation of rifamycin B production by its production of the B-Factor. With the removal of rifO in Amycolatopsis mediterranei S699, the production of rifamycin B should be decreased when RifO plays a crucial role in the production of rifamycin B and its derivatives.

This study investigated the potential of an internal increase of B-Factor as it was shown by Kawaguchi et al., that the external addition of B-Factor to the media stimulates the rifamycin B synthesis of cells. This led to the approach to enable cells to constitutively overexpress rifO followed by an intrinsic rise of B-factor, which itself can increase rifamycin production. This potential positive effect of constitutive overexpression of rifO on the rifamycin or its derivatives production was investigated for DCO36.

This study illustrates the effective adaptation of a modified PCR-targeted Streptomyces gene replacement technique for gene manipulation in Amycolatopsis mediterranei S699 and its mutant strains, including DCO36, known for producing 24-desmethyl rifamycin B. While Streptomyces species benefit from a diverse array of established cloning vectors and transformation methods, applying these techniques to Amycolatopsis mediterranei S699 proves notably challenging, with success not always guaranteed (Lal et al., 1991; Kieser et al., 2000). The development of suitable plasmids for this purpose is a time-intensive process (Tuteja et al., 2000; Dhingra et al., 2003). However, by utilizing the suicidal vector pKGLP2 in conjunction with replicative and consecutive PermE promoter-containing plasmids like pUWL-H (and pUWL-HA), this study offers a novelty reliable and efficient method for gene manipulation, facilitating the quick switching off, exchanging, or targeted overproduction of genes in Amycolatopsis mediterranei strains. These findings hint at the potential impact of this adapted method and plasmid use in streamlining earlier laborious approaches to modifying the rifamycin polyketide synthase gene cluster or altering regulatory elements in Amycolatopsis mediterranei strains. The successful adaptation of this method and plasmids paved the way for the results presented in this study.

As the removal of rifQ and the overexpression of rifO lead to an increase in Amycolatopsis mediterranei S699 mutant strains of rifamycin B derivatives, these genetic modifications can, in general, be seen as new useful tools to increase production like it was previously only shown for the wild type by the removal of rifQ (Lei et al., 2018). Based on the impact of these modifications a combination of the removal of rifQ as well as a simultaneous overproduction of rifO could further increase 24-desmethyl rifamycin B production. The approaches might also be usable in upcoming rifamycin B derivative-producing mutant strains. A developing understanding and further studies of factors and genes involved in the induction or increase of the biosynthetic cluster of rifamycin and its derivatives can lead to further genetic modifications to increase antibiotic yield.

To quantify the antibiotic production rate of rifamycin B derivative-producing strains, these mutants should be grown in the fermenter to gain a better quantitative understanding. Since the growth of these strains depends on multiple factors like dissolved oxygen concentration, glucose concentration, and pH during the growth or building of foam which secondly influences the production of rifamycin B and its derivatives, the growth in the fermenter is an interesting aspect for these optimized mutant strains. This allows further quantitative statements to be made, such as the antibiotic production performance in g/L, which can further expand the results presented in this work of the genetic changes carried out in mutant strains.

The removal of rifQ and the overexpression of rifO resulting in higher antibiotic yields in mutant strains can potentially close the gap to commercial use of their rifamycin B derivatives like it is aimed for 24-desmethyl rifamycin B of DCO36 (Sood et al., 2024).

The original contributions presented in the study are included in the article/ Supplementary Material , further inquiries can be directed to the corresponding author/s.

MM: Conceptualization, Investigation, Methodology, Writing – original draft. EB: Methodology, Writing – review & editing. IS: Investigation, Methodology, Writing – original draft. US: Writing – review & editing. RL: Conceptualization, Writing – review & editing. AB: Conceptualization, Funding acquisition, Methodology, Project administration, Resources, Supervision, Writing – original draft, Writing – review & editing.