Extracytoplasmic function (ECF) σ factors are dedicated to transcription of a subset of bacterial genes in response to environmental change, usually stress (Sineva et al., 2017; Helmann, 2019). By replacing the primary (or “housekeeping”) σ⁷⁰ subunit of the RNA polymerase (RNAP) holoenzyme with an alternative σ factor, the RNAP is directed to transcribe genes whose expression promotes bacterial fitness in response to the perceived signal. The ECF σ factors constitute Group 4 of the σ⁷⁰ family of σ factors. However, while the primary σ⁷⁰ factors feature four domains (σ₁–σ₄) for which specific functions have been delineated, the ECF σ factors are more diverse, and they contain only the two domains (σ₂ and σ₄) that are required for binding to the core RNAP and to cognate promoters (Feklistov et al., 2014). While this group of proteins was originally named for their involvement in transcription of genes encoding proteins with extracytoplasmic functions (Lonetto et al., 1994), some ECF σ factors have since been described, which control production of proteins with primary functions in the cytoplasm. The number of ECF σ factors encoded within a bacterial genome reflects the likelihood that the bacterium will encounter altered environmental conditions; on average, bacterial genomes encode 6–10 ECF σ factors (Staron et al., 2009; Casas-Pastor et al., 2021).

The classical ECF σ factors are auto-regulatory and co-transcribed with an anti-σ factor, which binds the σ factor directly to sequester it in an inactive state (Staron et al., 2009). The activating signal causes modification or degradation of the anti-σ factor, which results in release of the σ factor to allow its association with the RNAP core enzyme. While anti-σ factors are poorly conserved, many are characterized by a single membrane-spanning region, an N-terminal anti-σ domain located in the cytoplasm, and a variable periplasmic domain (Campbell et al., 2007). Approximately 20% of ECF σ factors appear not to be controlled by an anti-σ factor and to be instead regulated at the level of transcription (Staron et al., 2009).

This review focuses on ECF σ factors, which participate in production of the primary siderophores in pathogenic members of the genus Burkholderia. In response to iron starvation, the bacteria produce and secrete siderophores, which facilitate the cellular uptake of iron. Specifically, pathogenic bacteria face a severe iron limitation in a host environment, and siderophores are therefore important for virulence. A picture is emerging, which suggests that several global transcription factors are involved in regulation of these ecf genes, possibly to optimize their synthesis in a host environment.

The transition metal iron is essential due to its involvement as a cofactor or prosthetic group in a number of key proteins such as hemoglobin, cytochromes, and iron–sulfur cluster-containing proteins (Lawen and Lane, 2013). Yet excess iron is toxic due to its chemical reactivity, and improperly sequestered iron can act as a redox catalyst in the production of damaging reactive oxygen species (ROS) through its participation in the Fenton or Haber–Weiss reactions, demanding tight control on iron homeostasis (Abbasi et al., 2021). Fe(II) is soluble in hypoxic, aqueous solution. However, Fe(II) readily gets oxidized to Fe(III), which is poorly soluble, and much of it is stored in the interior cavity of the nanocage ferritin, while circulating Fe(III) is bound to transport proteins such as transferrin and lactoferrin. For bacterial pathogens, this becomes an issue when they colonize the host environment in which iron is sequestered by these high-affinity proteins, a sequestration that maintains free iron at concentrations much below the 10⁻⁸ to 10⁻⁶ M, which is required to support the growth of the bacterium (Cassat and Skaar, 2013). In addition, host nutritional immune responses to bacterial infection include efforts to limit iron availability further (Weinberg, 1975).

To counter host nutritional immunity, bacterial pathogens employ a range of mechanisms to optimize the scavenging of iron from the host (Caza and Kronstad, 2013). One of these mechanisms is the production and secretion of secondary metabolites named siderophores, which are small (500–1,500 Da) iron-chelators with affinities for iron that may exceed those of host proteins such as transferrin. Secreted siderophores form high-affinity complexes with Fe(III), and the complexes are then transported back into the bacterial cytoplasm by specific transporters (Crosa and Walsh, 2002; Hider and Kong, 2010). This transport process involves TonB-dependent outer membrane receptors recognizing the iron-siderophore complexes specifically and with high affinity, following which the iron-chelate enters the periplasm where it is bound by a periplasmic binding protein. The metal complexes are subsequently transported across the cytoplasmic membrane by ATP-binding cassette (ABC) permeases (Figure 1A; Caza and Kronstad, 2013). TonB-dependent receptors are so named because they make contact with the TonB component of the TonB-ExbB-ExbD complex (Braun and Endriss, 2007; Celia et al., 2019). TonB-ExbB-ExbD is a molecular motor, which utilizes the proton motive force across the cytoplasmic membrane of Gram-negative bacteria to furnish energy to the outer membrane transporters, thereby facilitating uptake of required nutrients, such as iron-siderophore complexes. Once in the cytoplasm, the iron must be released, which may involve hydrolysis of the siderophore by esterases to liberate Fe(III). Alternatively, Fe(III) may be reduced to Fe(II) by ferric reductases, as ferrous iron has much lower affinity for the siderophore (Trindade et al., 2021).

Pathogens typically produce multiple siderophores, which may have redundant functions or play specific roles depending on environment or the route of infection (Takase et al., 2000). Some are produced by the modular non-ribosomal peptide synthetases (NRPS), whereas others are the product of NRPS-independent siderophore synthases (Crosa and Walsh, 2002; Carroll and Moore, 2018). Biosynthetic gene clusters encoding proteins responsible for production and transport of secondary metabolites such as siderophores are often large, and the genes are subject to tight regulation (Thapa and Grove, 2019). Such regulation may involve global transcriptional regulators, which generally regulate a large number of genes with distinct functionalities, and/or it may depend on local cluster-specific regulators, which are dedicated to a certain pathway.

The genus Burkholderia includes a number of serious human pathogens. Phylogenetic analyses have suggested the division of Burkholderia species into distinct clades, one of which comprises plant and human pathogens (Sawana et al., 2014). Within this lineage, members of the Burkholderia cepacia complex (Bcc) include plant pathogens as well as species, which have emerged as dangerous opportunistic pathogens, such as Burkholderia cenocepacia, which can cause life-threatening infections in the lungs of persons living with cystic fibrosis or chronic granulomatous disease (Scoffone et al., 2017). The Burkholderia pseudomallei complex (Bpc) pathogens include B. pseudomallei and its clonal derivative Burkholderia mallei, both of which infect both humans and animals, causing melioidosis in humans and glanders in animals, respectively (Tapia et al., 2019). Burkholderia pseudomallei is a soil saprophyte, while B. mallei is an obligate pathogen. Because both species are highly contagious, inherently resistant to many antibiotics, and associated with a high mortality rate, they are classified as Tier 1 Select Agents. Therefore, many virulence traits are analyzed in the Bpc species Burkholderia thailandensis, which is much less virulent despite sharing significant genetic similarity with the more pathogenic species (Haraga et al., 2008).

All pathogenic Burkholderia species must acquire iron from their host, and they produce several siderophores (Butt and Thomas, 2017). The primary siderophore in Bcc species is thought to be the linear hydroxamate/hydroxycarboxylate ornibactin, which bears some similarity to the pyoverdines produced by pseudomonads, while Bpc species produce the structurally similar malleobactins. Accordingly, the biosynthetic gene clusters are also similar, and disruption of genes within these clusters is associated with reduced virulence (Sokol et al., 2000; Visser et al., 2004; Wand et al., 2011). In addition, secondary siderophores such as pyochelin are produced in many species, and these compounds are thought to have lower affinity for iron compared to ornibactin and malleobactin. The secondary siderophores may be more important in environmental conditions and not in virulence, as exemplified by the genomic deletion of the pyochelin biosynthetic gene cluster in the obligate pathogen B. mallei (Esmaeel et al., 2016).

The biosynthetic gene clusters involved in B. cenocepacia ornibactin and B. pseudomallei malleobactin production have been characterized in detail, and homologous gene clusters have been analyzed in Burkholderia xenovorans and B. thailandensis (Figure 1B; reflecting the annotated mba biosynthetic gene cluster in B. thailandensis; Agnoli et al., 2006; Alice et al., 2006; Franke et al., 2013; Vargas-Straube et al., 2016). These gene clusters are conserved among species and are usually encoded on the larger chromosome. Genes contained within these clusters encode two NRPSs as well as several modifying enzymes (Figure 1B; gray), an ABC transporter implicated in export of siderophores (red), and a set of proteins dedicated to uptake of iron-siderophore complexes. The latter comprise a TonB-dependent outer membrane receptor (green), a periplasmic binding protein (purple), and a cytoplasmic ABC transporter (light blue). The first gene in the clusters encodes an ECF σ factor referred to as MbaS (for malleobactin sigma) in B. pseudomallei and OrbS (for ornibactin sigma) in B. cenocepacia (Agnoli et al., 2006; Alice et al., 2006). Notably, none of the adjacent genes are predicted to encode an anti-σ factor.

MbaS and OrbS belong to a specific subgroup of ECF σ factors. Phylogenetic analyses have been used to classify all bacterial ECF proteins. They were first organized into 43 distinct groups for which specific functions were noted, and this classification was subsequently expanded to comprise 157 phylogenetic groups (Staron et al., 2009; Casas-Pastor et al., 2021). Based on these classifications, ECF σ factors belonging to the originally proposed groups ECF05-ECF09 (merged to generate ECF243 according to the updated classification) are generally involved in iron acquisition. It should be noted that the description of the more recently defined group ECF243 suggests that the genomic context includes an anti-σ factor, which does not accurately reflect the gene clusters under control of OrbS and MbaS; these σ factors were part of the originally described ECF09 family for which the apparent absence of a cognate anti-σ factor was a defining feature. Instead, genes encoding ECF09-family proteins are often adjacent to genes ending MbtH-like proteins (Figure 1B), which generally associate with the adenylation domain of NRPS enzymes (Felnagle et al., 2010).

OrbS was originally identified based on its 40% sequence identity to PvdS (for pyoverdine sigma), which controls biosynthetic genes involved in pyoverdine production in pseudomonads (Cunliffe et al., 1995; Llamas et al., 2014). PvdS has been extensively analyzed in species such as the opportunistic pathogen Pseudomonas aeruginosa, where it controls not only production and transport of pyoverdine but also production of virulence factors such as exotoxin A. Expression of the PvdS regulon requires pyoverdine, which signals to the anti-σ factor FpvR, resulting in activation of PvdS (Lamont et al., 2002; Llamas et al., 2014). Despite sequence similarities, regulation of OrbS and MbaS is clearly different from that of PvdS with no evidence for the involvement of an anti-σ factor or the respective siderophores; instead, production of these ECF σ factors appears to rely on transcriptional regulation of the respective genes. Compared to PvdS, OrbS has a 29-amino acid N-terminal extension, and it appears to control only genes within the ornibactin biosynthetic gene cluster (Agnoli et al., 2019).

Production of malleobactin and ornibactin requires expression of large biosynthetic gene clusters, which demands significant cellular resources. Accordingly, tight transcriptional control is exerted. These siderophores have high affinity for Fe³⁺ and may be of particular importance in the iron-limited host environment. A picture is emerging suggesting that regulation of genes encoding MbaS and OrbS determines availability of these ECF σ factors and not anti-σ factors or the respective siderophores. Notably, the involvement of the global regulator MftR, which has been implicated in controlling genes associated with virulence, suggests that regulatory mechanisms are optimized to ensure maximal expression when two criteria are met—iron limitation and host colonization (Figure 2). However, maximal does not equal optimal. The quorum sensing-dependent repression by ScmR during stationary phase appears to function as an overflow valve, precluding excessive production of high-affinity iron chelators under conditions of high cell density.

AG contributed to conceptualization, writing—original draft, writing—review and editing, visualization, and funding acquisition.

This work was supported by the National Science Foundation (MCB-1714219 to AG).

The author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.