Acinetobacter baumannii is one of the most common nosocomial pathogens responsible for a wide range of concerning diseases, such as pneumonia, bacteremia or secondary meningitis (Wong et al., 2017). The rise of healthcare-associated infections caused by multidrug resistant strains of A. baumannii, together with the scarce development of new antimicrobials in the last decades, represents an important health threat (De Oliveira et al., 2020). In fact, A. baumannii, is one of the Gram-negative ESKAPE pathogens identified by the World Health Organization (WHO) as critical priority for antibiotic discovery (World Health Organization [WHO], 2017).

Although the main characteristic of this pathogen is its ability to acquire new antimicrobial resistance, it shows several mechanisms involved in virulence, persistence and stress adaptation that enhance its pathogenicity (Harding et al., 2018). Within this context, many researchers have focused their efforts in developing alternatives to conventional antibiotics, such as antivirulence agents, that can work alone or together with antibiotics to overcome A. baumannii infections (Dickey et al., 2017).

Iron is an essential micronutrient for bacteria to infect and multiply in tissues and body fluids of the host, playing an important role in pathogenesis. The mechanisms of bacterial iron acquisition include: (i) expression of transporters involved in the uptake of ferrous iron, such as the Feo system; (ii) extraction of heme-iron from hemoproteins; (iii) capture of iron from transferrin and lactoferrin; or (iv) synthesis of siderophores (Sheldon et al., 2016). Siderophores are high-affinity iron-chelating molecules synthetized by microorganisms to scavenge extracellular ferric iron from the environment. Baumanoferrin, fimsbactin and preacinetobactin-acinetobactin (referred to as acinetobactin) are the most common siderophore systems detected in A. baumannii (Yamamoto et al., 1994; Proschak et al., 2013; Penwell et al., 2015). The most extensively studied is acinetobactin which is considered the major siderophore of A. baumannii and it is highly conserved among all A. baumannii strains (Antunes et al., 2011). All the genes required for the synthesis (basA-J), efflux (barA/B) and uptake (bauA-F) of acinetobactin are located in a 26.5-kb chromosomal region (Supplementary Figure 1A) (Mihara et al., 2004), with the exception of the entA homolog gene, found elsewhere in the chromosome (Penwell et al., 2012). Its biosynthesis follows the logic of a non-ribosomal peptide synthetase (NRPS) assembly system, where three precursors are bound in equimolar quantities into the preacinetobactin molecule: N-hydroxyhistamine, L-threonine and 2,3-dihydroxybenzoic acid (DHBA) (Figure 1A) (Yamamoto et al., 1994; Song and Kim, 2020). Once the preacinetobactin synthesis is completed, the siderophore is secreted to the extracellular space (Figure 1B) where two reactions can occur: (i) preacinetobactin stabilization by chelation of ferric iron or (ii) non-enzymatically and irreversibly isomerization to acinetobactin at pH > 7 (Shapiro and Wencewicz, 2016; Moynié et al., 2018). The fimsbactins A-F siderophores are present in a small fraction of the A. baumannii isolates (Antunes et al., 2011; Proschak et al., 2013). These siderophores, also derived from a NRPS assembly system, are structurally related to acinetobactin by the presence of catecholate, phenolate oxazoline, and hydroxamate metal-binding motifs (Proschak et al., 2013). Moreover, the cluster fbsA-Q, coding for the fimsbactins, consists of 18 genes with high functional similarity to those present in the acinetobactin cluster indicating redundancy between both siderophores pathways (Supplementary Figure 1B) (Dorsey et al., 2004; Mihara et al., 2004; Proschak et al., 2013). Although the expression of fimsbactins was shown to be enough to support the growth of A. baumannii in serum, these siderophores are not required for survival during bacteremia (Sheldon and Skaar, 2020).

Indeed, acinetobactin has been shown to be essential for the virulence of A. baumannii during greater wax moth (Galleria mellonella) and murine bacteremia and pneumonia infections (Gaddy et al., 2012; Penwell et al., 2012; Martínez-Guitián et al., 2020; Sheldon and Skaar, 2020). However, no studies have explored the contribution of the entire collection of acinetobactin genes to the infectious process.

Herein, we have performed an in-depth analysis of the role of the acinetobactin cluster in the virulence of A. baumannii which allowed us to identify potential targets for the design of new antimicrobials against this pathogen.

In the last decades, the emergence of A. baumannii multidrug-resistant strains has become a worldwide concerning problem derived from the scarcity of effective therapeutic options against this bacterium. Hence, the World Health Organization (WHO) included A. baumannii as a critical priority pathogen claiming an urgent need of efficient alternatives to the known antibiotics (World Health Organization [WHO], 2017). Within this context, the search and identification of new therapeutic targets in A. baumannii have become a priority.

The pathogenesis success of A. baumannii is partially linked to the synthesis of active siderophores that supply the iron needed for its essential role in crucial metabolic events. Among all the siderophore systems identified, acinetobactin is considered the major siderophore of this bacterium (Yamamoto et al., 1994). Since then, several studies have focused on unraveling its regulation, chelating mechanisms and its role in the virulence of the bacterium (Mihara et al., 2004; Gaddy et al., 2012; Shapiro and Wencewicz, 2016; Sheldon and Skaar, 2020). After confirming that acinetobactin-related metabolism is a crucial virulence factor and that it is highly conserved among A. baumannii strains, the iron(III) uptake system mediated by acinetobactin has been proposed as a potential therapeutic target to combat this multidrug-resistant pathogen (Antunes et al., 2011; Gaddy et al., 2012; Sheldon and Skaar, 2020). However, the contribution of each individual gene involved in the acinetobactin metabolism in the infectious process is still unclear. Thus, we have performed an in-depth analysis of the acinetobactin gene cluster by conducting different phenotypical assays with the well-known strain A. baumannii ATCC 17978 and 19 isogenic mutant derivative strains lacking genes involved in the biosynthesis and transport of acinetobactin.

Three different siderophore-mediated iron uptake systems (acinetobactin, baumanoferrin and fimsbactin) were identified from the reference strain A. baumannii ATCC 17978, which corresponding gene clusters were found upregulated under in vitro iron-limiting conditions and in vivo infection (Eijkelkamp et al., 2011; Murray et al., 2017; Martínez-Guitián et al., 2020). Acinetobactin and fimsbactins are synthesized through non-ribosomal peptide synthetase (NRPS) assembly systems, sharing the DHBA and L-threonine precursors (Proschak et al., 2013; Song and Kim, 2020). This could explain the high level of genetic redundancy between both clusters where most of the genes involved in the biosynthesis of the acinetobactin have a potential redundant gene in the fimsbactin cluster. Hence, although fimsbactins are only present in a small percentage of A. baumannii strains, fimsbactin genes could complement the inactivation of some acinetobactin genes, suggesting a redundancy in both pathways. Within this context, A. baumannii ATCC 17978 confers the perfect background for the present study.

Bioinformatic analysis of the acinetobactin gene cluster showed that basG, basC and entA genes do not have a potential redundant gene. The lack of redundant basG and basC genes in the fimsbactin cluster is easily explained since these genes are involved in the biosynthesis of the precursor of acinetobactin N-hydroxyhistamine (Shapiro and Wencewicz, 2016) and this moiety is not present in the fimsbactins. On the other hand, the entA gene is always located outside of the acinetobactin cluster, varying its location between strains. In A. baumannii ATCC 17978, acinetobactin and fimsbactins share the gene entA, which is located in the fimsbactin cluster (Penwell et al., 2012). The individual deletion of these three genes resulted in a drastic reduction of the virulence when compared to the wild-type strain. Our results agreed with a recent study published by Sheldon and Skaar where they demonstrated that the deletion of gene basG impairs growth on human serum, transferrin or lactoferrin as sole iron sources, and severely attenuates survival of A. baumannii ATCC 17978 in a murine bacteremia model (Sheldon and Skaar, 2020).

Although Dorsey et al. predicted that basC gene had an essential function in the biosynthesis of the acinetobactin on the basis of its involvement in the synthesis of N-hydroxyhistamine (Dorsey et al., 2004), this hypothesis was never investigated until now. Siderophore-content analysis of the cell free supernatants of both ΔbasG and ΔbasC using our SPE-HLB/HPLC-MS methodology showed that the deletion of these genes resulted in the complete inhibition of acinetobactin production. The low virulent phenotype of these two mutants could be related to the lack of the siderophore. As the previous case, it would be expected that the deletion of the entA gene will inhibit the biosynthesis of both acinetobactin and fimsbactin siderophores.

Deletion of basB and basD also led to a significant decrease in virulence characterized by an impaired fitness under iron-limiting conditions and increased mice survival. We have previously reported that basB is an essential gene for the virulence of A. baumannii during pneumonia in mice and for bacteria growth under iron-limiting conditions (Martínez-Guitián et al., 2020). In addition, Gaddy et al. demonstrated that basD gene is essential for the biosynthesis of acinetobactin and for the bacterial growth under iron-depleted conditions in A. baumannii ATCC 19606 (Gaddy et al., 2012). Both genes code for proteins involved in the last steps of the biosynthetic pathway of acinetobactin, where DHBA and L-Threonine precursors are linked (BasD) and the resulting intermediate is bonded to N-hydroxyhistamine (BasB) to give preacinetobactin (Hasan et al., 2015; Shapiro and Wencewicz, 2016; Song and Kim, 2020). Both ΔbasB and ΔbasD mutants were unable to synthetize acinetobactin.

A closer analysis of the basJ gene, showed that its deletion did not have any effect in the biosynthesis of acinetobactin. However, the deletion of both acinetobactin (basJ) and fimsbactin (fbsB) redundant genes resulted in the loss of acinetobactin and fimsbactins production. This fact demonstrates that in absence of the basJ gene, A. baumannii ATCC 17978 can use fbsB to synthesize acinetobactin. This is a clear example of molecular redundancy whereby two genes have the same function or when an alternative pathway fulfills the mission role of an inactivated gene. Pathogens used it to adapt to a continuous changing environment, avoiding the antimicrobial defenses of their hosts (Ghosh and O’Connor, 2017). Based on our results, we predict that the redundancy of ΔbasF,ΔbasI,ΔbasH,ΔbasA, and ΔbasE strains possibly lead to the unchanged ability to synthesize acinetobactin.

Among the genes involved in the transport of acinetobactin, only the gen bauA, coding for the outer-membrane receptor, was found to be essential for the virulence during a murine sepsis model. Previous studies have shown that the gene bauA is essential for the virulence of A. baumannii ATCC 19606 since its deletion led to a decrease in the ability of the bacteria to infect, divide inside body fluids of mice and in fitness under iron-limiting conditions (Gaddy et al., 2012). In fact, BauA was proposed as a good vaccine candidate since mice injected with recombinant BauA were able to produce antibodies against this protein. In addition, passive immunization using serum anti-BauA protected mice from infection (Esmaeilkhani et al., 2016). Our data slightly differ with this study since we have not observed any reduction in the fitness of A. baumannii ATCC 17978 when bauA was deleted. This discrepancy could be explained by the higher susceptibility of the strain ATCC 19606 to chelate iron (III) compared to the ATCC 17978 strain, possibly due to the lack of fimsbactins and baumanoferrin production in the ATCC 19606 strain (Antunes et al., 2011; Proschak et al., 2013; Penwell et al., 2015; Ramirez et al., 2019).

Both BarA and BarB proteins belong to the acinetobactin secretion system (efflux) of the ABC superfamily. Mutant strains lacking the genes involved in the synthesis of these proteins, ΔbarA and ΔbarB, showed a significant decrease in fitness under iron-limiting conditions. Notwithstanding, no statistical differences between the percentage of survival of the mice infected with these mutant strains and the mice infected with the parental strain were observed. SPE-HLB/HPLC-MS analysis of ΔbarB mutant strain cultures showed the presence of acinetobactin in its cell-free supernatant, which indicates that this single mutant strain did not prevent the efflux of acinetobactin outside the bacteria. A previous study carried out by Penwell et al. in A. baumannii ATCC 19606, showed a growth defect under iron-limiting conditions and a 60% decrease of acinetobactin effluxed in the ΔbarA/ΔbarB cell-free supernatant compared with the parental strain (Penwel, 2013). The reduction in the efflux of acinetobactin matches with the partial loss of virulence in the ΔbarA and ΔbarB mutant strains. It is known that A. baumannii possesses a wide variety of transport mechanisms. It is possible that under stress conditions, the bacteria could use non-specific transporter systems to secrete and uptake acinetobactin and do not lose the iron-battle against the host (Iacono et al., 2008; Coyne et al., 2010, 2011; Fernando and Kumar, 2012).

Several researchers have focused their efforts on the development of new inhibitors based on acinetobactin metabolism. Inhibitors of BasE, an enzyme involved in biosynthesis of the acinetobactin, have shown a powerful inhibitory activity (Neres et al., 2013). Analogous of acinetobactin have also shown bacteriostatic activity as they were able to block the transport of the iron-acinetobactin complex inside the bacteria (Bohac et al., 2017; Shapiro and Wencewicz, 2017). In the last years, siderophore conjugates using the “Trojan Horse” antibiotic drug delivery strategy has become more popular for combating this microorganism. In fact, cefiderocol (fetcroja) was the first cathecol-substituted siderophore cephalosporin approved by the FDA and EMA (Ji et al., 2012; Wencewicz and Miller, 2013; Ghosh et al., 2017; Parsels et al., 2021).

In summary, we performed an in-depth analysis of the role of each individual gene of the acinetobactin metabolism in the virulence of A. baumannii ATCC 17978, allowing us to identify six potential targets for the design of new antimicrobials against this microorganism: five of them involved in its biosynthesis (entA, basG, basC, basD, and basB) and one related to its transport (bauA). Due to the similar function and potentially similar structure of the enzymes involved in the biosynthesis of acinetobactin and fimsbactin, inhibitors against the remaining biosynthetic steps could also have the potential to be effective by inactivating both redundant proteins.

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

The animal study was reviewed and approved by Hospital Universitario A Coruña, Spain, project code P102.

KC-P, SR-F, NT-T, and LÁ-F performed mutant construction. MM-G, JV-U and KC-P performed phenotypic experiments and animal models. LA performed the analysis of the siderophore-content. AB, MP, CJ, and LÁ-F designed and supervised the experiments and wrote the manuscript. GB and JR revised the manuscript. All authors read and approved the final manuscript.

The authors declare 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.