Acinetobacter baumannii: Emergence of a Superbug, Past, Present, and Future

Abstract

By exhibiting resistance to most known antibiotics or quickly acquiring resistance to antimicrobials it was once susceptible to, Acinetobacter baumannii has attracted increasing attention worldwide since the last decades of the previous century. The resistance abilities of the bacterium were soon shown to be so advanced that it was even able to resist antibiotics that had only just been discovered and used for first time. Utilizing complex mechanisms of resistance, combining different modalities, and achieving new resistant traits established A. baumannii as one of the most clinically important and challenging pathogens of the new century, being categorized by the World Health Organization as a critical priority bacterium for which new antibiotics are urgently needed. After even last-resort, broad-spectrum antibiotics were rendered useless, the fight against this superbug began to be led by the reintroduction of once abandoned antimicrobials, new combination therapies and novel modalities of treatment. In this chapter, we will look at the history and background of Acinetobacter species and then specifically focus on A. baumannii, explaining its clinical importance in detail, reviewing the most recent findings regarding its mechanisms of resistance, latest modalities of treatment and newest areas of research towards opening new frontiers in the management of infections caused by multi-resistant strains of this bacterium.

1. Introduction

The story of the Acinetobacter genus dates back to the early 20th century when in 1911, Dutch microbiologist Martinus Willem Beijerinck named a newly discovered organism isolated from soil using minimal media enriched with calcium acetate: Micrococcus calco-aceticus [1]. Since he did not describe the microorganism but only named it, his report was greatly overlooked [2]. Hence, in the meantime, the same microorganism was described under various names over several decades, including: Diplococcus mucosus, Micrococcus calcoaceticus, Alcaligenes haemolysans, Mima polymorpha, Bacterium anitratum, Herellea vaginicola, Moraxella lwoffi, Achromobacter anitratus, Moraxellalwoffi var. glucidolytica, Achromobacter mucosus, and Neisseria winogradskyi [1]. Due to this, the original strain has been lost and the name proposed by Beijernik was used by few authors [2].

In 1954 Brisou and Prévot coined the name Acinetobacter from the Greek “ακινɛτοσ” [akinetos], meaning non-motile, to separate non-motile from motile microorganisms within the genus Achromobacter [3]. Years later in 1968 Paul Baumann, through a comprehensive survey, suggested that all of the above-listed species belonged to a single genus and could not be further sub-classified into different species based on the phenotypical characteristics, for which the name Acinetobacter was proposed [4]. In 1971, based on the results of Baumann’s publication subcommittee on nomenclature of Moraxella and allied bacteria, the genus Acinetobacter was officially acknowledged [5]. In 1974 the genus Acinetobacter was listed in Bergey’s manual of systematic bacteriology, described as a single species named Acinetobacter calcoaceticus [1]. Despite that, in the “Approved List of Bacterial Names” of 1980, based on the observation that some Acinetobacter species were able to acidify glucose, the species were listed as A. calcoaceticus (which was further subdivided into the two biovars A. calcoaceticus bv. anitratus and A. calcoaceticus bv. lwoffii) and Acinetobacter lwoffii [6]. These nomenclatures, however, were never approved by taxonomists [1].

Acinetobacter are defined as gram negative (however, since Acinetobacter species are often difficult to de-stain through the process of Gram staining, they are often incorrectly identified as gram positive), strictly aerobic, catalase-positive, oxidase-negative, non-motile, non-fermenting, non-fastidious bacteria with a DNA G + C content of 39–47% [7]. Acinetobacter has been classified in the family Moraxellaceae, within the order Pseudomonadales, and the class Gammaproteobacteria since 1991 [8].

By 2015 only 33 species of Acinetobacter had been identified. Rapid technological advances in recent years have caused faster and more precise identification of novel strains. Continued reports of such novel strains have doubled the number of described species in the last 6 years. Within the 4 years from 2017 to 2020 alone, 22 new Acinetobacter species were identified [9]. By May 2021 67 Acinetobacter species had already been validly named, and 20 additional species were under tentative species designation. Most identified species, however, are non-pathogenic, environmental microorganisms [9].

A. baumannii is a ubiquitous opportunistic coccobacillus, widely distributed in the environment, having reservoirs in nearly all environmental niches [8]. Despite isolation of A. baumannii from soil, water, crude oil, sewage, solid surfaces, human skin, raw meat, milk, milk products, vegetables, livestock, fish, shrimp, plants, and so on, natural habitats of this bacterium are still poorly understood as it is almost exclusively isolated from close contact communities and hospital environments [8].

Smith et al., performed and published the first whole genome sequencing of A. baumannii (strain ATCC 17978) in 2007 [10]. However, identification of A. baumannii in routine diagnostic laboratories has been found to be quite complicated due to bacterium’s phylogenetical closeness to several other Acinetobacter species, collectively called A. calcoaceticus-A. baumannii (Acb) complex [8].

Traditionally four species of Acinetobacter A. calcoaceticus (formerly known as genomic species 1), A. baumannii, Acinetobacter pittii (formerly known as genomic species 3), and Acinetobacter nosocomialis (formerly known as genomic species 13TU), were members of Acb complex [11]. In recent years, other genomic species have been proposed for inclusion, including: Acinetobacter seifertii (formerly known as Acinetobacter genomic species “close to 13TU”), A. lactucae (formerly known as Acinetobacter NB14 and synonym of A. dijkshoorniae), and the Acinetobacter genomic species “between 1 and 3” [12]. As mentioned earlier, all species of the complex are phenotypically indistinguishable and genetically are closely related to the extent that molecular methods are required for their identification [12]. Acb complex members (except for A. calcoaceticus) are most frequently associated with human pathogenicity and health association infections (HAI). While the pathogenicity of A. calcoaceticus is still unknown, other members of the complex are frequently identified as major human pathogens, with A. baumannii being the most prominent one, having the highest associated HAI, followed by A. pittii and A. nosocomialis [12].

Although until 1975 A. baumannii was still treated by antibiotic monotherapy with relatively good susceptibility, the rate of antibiotic resistance (ABR) of the bacterium has been rising ever since [13]. In the early years of the 21st century, the Infectious Disease Society of America highlighted a group of bacteria including Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, A. baumannii, Pseudomonas aeruginosa, and Enterobacter spp. as “the ESKAPE pathogens” for their ability to “escape” antibiotics and resist antibiotic treatment, mutually representing a new paradigm in pathogenesis, resistance, and transmission [14]. Multidrug resistance (MDR) species of A. baumannii with resistance to almost all known antibiotics (including last-resort broad-spectrum antibiotics like carbapenems) are being isolated from hospital environments worldwide, posing a great burden on the health care system [13]. Resistance to carbapenem (a broad-spectrum β-lactam antibiotic) is considered to be a marker of extensive antimicrobial resistance since it involves a large range of co-resistance to other unrelated antibiotic classes [8]. In 2017, the World Health Organization (WHO) listed carbapenem-resistant (CR) A. baumannii as a critical priority bacterium for which new antibiotics are urgently needed [8]. The rate of clinical outbreaks of CR A. baumannii in Europe ranges between 1% to over 30%, being most intensive in eastern and southeastern Europe [8].

The range of A. baumannii clinical isolates among all gram-negative aerobes differs from 0.7% in North America to 4.6% in the Middle East [8]. Historically the Middle East is one of the regions most known for MDR outbreaks of A. baumannii, giving it the memorable “Iraqibacter” title [15]. Several MDR outbreaks of the bacterium were reported among US military hospitals in Iraq, Afghanistan, and Kuwait during the Iraq war [15, 16], making A. baumannii a global concern. Several research studies have been conducted ever since to understand the antibiotic resistance mechanism of the bacterium. However, so far, our understanding of the pathology, epidemiology, and MDR mechanism of A. baumannii is limited. Worrisomely, restricted options exist for the treatment of MDR A. baumannii infections, where reportedly, Colistin is the only antibiotic still having therapeutic effects on such strains, making the bacterium extremely difficult to treat [8]. Isolation of colistin-resistant A. baumannii reports have been rising in recent years [17, 18, 19], emphasizing the emergent need for new antimicrobials, treatment strategies, stricter control, and better management of colonization of patients by a rational antibiotic stewardship programme.

In this chapter, we describe the latest findings regarding the pathogenicity and mechanisms of antibiotic resistance of A. baumannii, then review the available routine and novel treatment strategies for this new-age superbug.

2. Clinical relevance and mechanisms of the pathogenicity of A. baumannii

The survival of A. baumannii on solid surfaces for extended periods of time is linked with the ability of the bacterium to cause nosocomial outbreaks. A majority of isolates of the bacterium are collected from the hospital environment and are closely connected with HAI, specifically in patients hospitalized in intensive care units (ICU) or in immunocompromised patients [8]. The bacterium can colonize curtains, linen fomites, bed rails, sinks, tables, medical equipment, etc., and hence can be transmitted through the vicinity of infected or colonized patients [20]. Contamination of respiratory support devices or intravenous access devices such as central venous catheters (CVC) is a major source of infections in critically ill patients [20].

The most important and common manifestations of A. baumannii include ventilator-associated pneumonia (VAP); CVC-associated bloodstream infections (BSI) (both VAP and BSI having the highest mortality rate); urinary tract infections (UTI) (although not very frequent, it causes up to 1.6% of urinary tract-related infections in ICU patients associated with prolonged catheter-related complications); and central nervous system infections such as meningitis, skin, soft tissue, and wound infections [21]. Infections caused by A. baumannii are associated with extended hospitalization, older age, and male gender [21].

A. baumannii has also gained the ability to infect the general population (however, to a lesser extent in comparison to HAI), causing community-acquired pneumonia (some reports stating that community-acquired pneumonia of A. baumannii is more fulminant than nosocomial pneumonia caused by the same bacterium, causing death within 8 days on average from diagnosis [22]), bacteraemia, skin and soft tissue infections, endocarditis, secondary meningitis, and ocular infections connected to contact lens use [11, 23, 24]. These infections are associated more with male gender and are more common in patients with comorbidities such as diabetes melitus, chronic obstructive pulmonary disease (COPD), renal disfunctions, and unhealthy lifestyle, including alcoholism or heavy smoking [21]. Origin of infection in 20–70% of patients infected with A. baumannii remains unknown [1].

The mortality rate of A. baumannii largely varies depending on strain type, type of infection, degree of immunosuppression and comorbidities of infected patients. The mortality rate of the bacterium can vary from 26% in a general hospital setting up to 43% in ICUs [25]. A. baumannii is the principal agent of VAP, accounting for 15% of HAI, which accounts for approximately 50% of total use of antibiotics and the highest mortality rate in ICUs [13].

A. Baumannii also remains an important threat to neurosurgery patients, causing up to 4% of all meningitis and shunt-related infections, with a mortality rate as high as 70% [26].

The virulence of A. baumannii, as mentioned earlier, is strain dependant and involves complicated mechanisms including adhesion, cell invasion, cytotoxicity, and serum persistence [8].

Adhesion and invasion: Outer membrane proteins (OMPs) are by far the most well studied structures of the bacterium, contributing to cell adhesion, invasion, and cytotoxicity [27]. OmpA (also known as Omp38), a conserved, abundant porin with a molecular weight of 38 kDa, is one of the adhesins which plays an essential role in the bacterium’s adhesion and cell invasion [27]. After attachment of A. baumannii to epithelial cells (especially with high affinity for epithelial cells of the respiratory tract), the bacterium secretes OmpA into the cell, leading to bacterial uptake through actin rearrangement and membrane reorganization of the attacked cell, a process known as the zipper-like mechanism [27]. After uptake, internalized bacteria are located in membrane-bound vacuoles [27]. Following that, internalized OmpA adhesins are translocated to the nucleus and mitochondria of the invaded cell, causing the release of a group of proapoptotic molecules, promoting cell apoptosis and hence acting as cytotoxins [28]. OmpA plays an important role in the enhancement of biofilms on plastic and is a crucial component in its formation [27]. Omp33–36 (also known as Omp34) is another adhesin which contributes to cell adhesion and cytotoxicity [8]. Studies show that mutation in the Omp33 gene results in significant reduction in the bacterium’s ability in adhesion, invasion, and apoptosis [29]. Besides enabling the bacteria to attach to host cells, Omp33 induces cell apoptosis by activating caspase and modulating autophagy, leading to the accumulation of sequestosome 1 and LC3B-11 autophagosome [29]. Some of the other cell surface proteins contributing to the bacteria’s cell adhesion and consequently to its virulence include: Biofilm-associated protein (Bap), by increasing the membrane hydrophobicity [30], Acinetobacter trimeric autotransporter (Ata), by mediating adherence through binding to type IV collagen in the extracellular matrix of attacked cells [31], and the FhaB/FhaC, CdiA/CdiB type Vb secretion system [32]. Ata has been shown to cause secretion of pro-inflammatory cytokines (such as interleukin (IL) 6, IL8), leading to caspase-dependent cellular apoptosis [33]. Hydrophobic abilities allow the bacterium to attach to foreign materials, the same characteristic that helps bacteria to attach to materials such as plastics used in intravascular access devices. It is has been shown that surface hydrophobicity is highly expressed in bacteria isolated from patients in comparison to the normal flora of the skin [7]. Bap is directly involved in the formation of a mature biofilm in medically relevant surfaces such as titanium, polystyrene, and polypropylene [7].

Recently, the response regulator BfmR and the sensor kinase BfmS, in the two-component regulatory system BfmRS, have been shown to play a role, as non-OMP adhesins, in cell attachment of the bacterium [34]. The two-component regulatory system BfmRS can sense extracellular signals important in the formation of biofilm, causing expressions of the chaperone-usher assembly system accountable for the formation of pili (the needed component for biofilm formation on polystyrene surfaces). It has been suggested that pili may be relevantly attributed to bacterial attachment in biotic and abiotic surfaces [8]. Some of those indicators include the ability of A. baumannii to produce different adhesins and pili in response to different environmental conditions (B1sA mediated light-regulated type 1 pilus assembly system PrpABCD [35]), and the association between twitching motility caused by type IV pili and cell adhesion [36]. An association between motility and adhesion of A. baumannii was shown as well, in bacteria with disrupted expression of a homolog of the histone-like nucleoid structuring protein (H-NS) known as a global transcriptional repressor [37].

Phospholipase enzymes play a significant role in the bacteria’s invasion ability [8]. Both phospholipase C (PLC) and phospholipase D (PLD) inactivation lead to impaired cell invasion [38]. Phospholipases are essential enzymes necessary for the metabolism of phosphatidylcholine and the release of phosphorylated head groups and polar head groups from PLD trough PLC cleavage, imposing change in the constancy of the membrane of host cells [38] and so facilitating invasion. OMPs containing phosphorylcholine contribute to A. baumannii cell adhesion and invasion by binding to the platelet-activating factor receptor on human lung epithelial cells, causing an activation signaling cascade involving G protein, clathrin, β-arrestins, vacuolar movement proteins and intercellular calcium, leading to the invasion of cells [39].

To date, our understanding of host cell receptors’ contribution in bacteria’s adhesion and invasion is still very limited.

2.1 Cytotoxins

Other than the already-mentioned cytotoxic abilities of OmpA, Omp33 and Ata, the lipopolysaccharide (LPS) envelope of A. Baumannii is another main contributor to the bacterium’s cytotoxicity and immunogenicity towards the host cell [40]. LPS, as the major component of the outer leaflet of the outer membrane of the bacterium, consisting of an O-antigen and a lipid A moiety (acting as a chemotactic agent), stimulates the secretion of tumor necrotic factors (TNFs) and IL-8 from macrophages [41]. Absence or disturbance in the production of any of the first three enzymes in the lipid A biosynthetic pathway (namely acyltransferase (LpxA), deacetylase (LpxC), and N-acyltransferase (LpxD)) reduces serum survival of the bacterium and lead to less lethal infections, highlighting the immunogenicity of LPS [42].

γ-glutamyl transferase enzyme (GGT), a secretion from the type II secretion system, is another protein that contributes to the bacteria’s cytotoxicity via caspase activation leading to ATP depletion followed by cell-cycle arrest and apoptosis [43]. It has been shown that strains of A. baumannii with higher levels of extracellular GGT activity cause more severe tissue damage, through increased inflammation and oxidative activities with elevated phenoloxidase, lysozyme, lactate dehydrogenase, and lipid peroxidation [43]. Higher levels of GGT in serum was also associated with oxidative stress and exacerbation of COPD [44]. The type IV secretion system, based on its identification within pathogenicity islands after whole genome sequencing, as well as the type VI secretion system, by killing competing bacteria, have been suggested to contribute to the pathogenicity of the bacteria; however, the complete mechanism and molecular interactions of these systems are not yet fully understood [8, 10, 45].

2.2 Serum persistence

In addition to the mentioned mechanisms, serum persistence indirectly affects the pathogenicity of the bacterium [8]. Capsular exopolysaccharide (CPS) is an important component mediating in persistence and hence, better survival of many invasive bacteria under unfavorable and harsh conditions, protecting them from phagocytosis of humoral immune attacks and inhibiting activation of alternative complement pathway [46] (an ability reported from OmpA proteins as well [47]). Although many isolates of A. baumannii from patients express the K locus gene cluster, which is needed for biosynthesis and export of CPS [46], having CPS alone should not be considered as a pathogenic determinant as many non-pathogenic microorganisms possess this envelope component as well [8]. Protective characteristics of CPS have been suggested through a series of studies, showing that mutation in its associated genes significantly decreases or in some cases impairs the survival of the bacterium in tissue. For example, two capsule-associated genes (namely ptk and epsA), believed to encode a putative protein tyrosine kinase and putative polysaccharide export OMP mutation, caused a significant decrease in survival of the bacterium in soft tissue [48].

The ability of the bacterium to degrade H₂O₂, utilized by catalases KatE and KatG, which could decrease production of reactive oxygen species by the innate immune system, is another mechanism used by A. baumannii to increase its serum resistance [8]. Universal stress protein A also showed the capacity to protect the bacterium from low PH, H₂O₂, and 2,4-DNP [49].

Mutation of LpsB glycosyltransferase involved in LPS synthesis caused a reduction in the resistance of A. baumannii to human serum, suggesting the role of LPS in the bacteria’s persistence [50].

Finally, metal acquisition systems, the BfmRS system through biofilm formation, phospholipases, and the novel plasminogen protein CipA all were shown to have an effect on the bacteria’s ability to survive longer in vivo [49, 51]. CipA links to the active form of plasmin and plasminogen, leading to a breakdown of fibrinogen, facilitating the spread of bacteria [49]. A. baumannii utilizes a complex set of metal acquisition systems, including iron (BasD, BauA, NfuA), zinc (ZnuABC, ZigA), and manganese (MumC, MumT) acquisition systems [8]. As an example, an iron acquisition system consisting of three siderophores (namely baumannoferrin, fimsbactin, and acinetobactin as highly converged iron chelators) is used by the bacteria to thrive under iron-reduced circumstances [52].

3. Mechanisms of resistance

Currently there are reports of isolates of A. baumannii which are resistant to all known antibiotics [7]. Misuse and overuse of antibiotics, lack of a proper antibiotic stewardship programme and the innate capability of the bacterium to adapt to new environmental challenges and acquire new resistant capacities are all part of the bigger picture which has made A. baumannii into the superbug it is today, for which we barely have any proper treatment. The genetic setup of the bacterium is capable of rapid developments towards antimicrobial resistance, making A. baumannii a natural transformant [7]. This ability is most likely due to the bacteria’s lack of mismatch repair system gene (called mutS), causing increased mutation rates. It is unknown if the bacterium is naturally capable of transformation or if this change in pathogenicity and antibiotic resistance is due to environmental conditions [53]. Until the 1970s, ampicillin, gentamycin, and nalidixic acid were still used in A. baumannii treatment, as single or combination therapy. By 1975, high rates of resistance to mentioned antibiotics had already been reported [7]. To date, antibiotics such as penicillins, cephalosporins, tetracyclines, chloramphenicol, and most aminoglycosides have lost their efficacy against the bacterium [54].

A. baumannii uses a complex set of mechanisms for its antibiotic resistance [54]. Enzymatic hydrolysis of antibiotics, genetic modifications, and efflux pumps are adopted mechanisms used by the bacterium to escape the effect of antimicrobials.

3.1 Resistance to beta-lactams

In A. baumannii, resistance to beta-lactam antibiotics is achieved by the utilization of beta-lactamases, causing degradation of the antibiotic, modification in penicillin-binding proteins, and decreased permeability to antibiotics through change in outer membrane porins or expulsion of the antibiotic from the cell with the help of efflux pumps [55].

3.1.1 Beta-lactamases

Beta-lactamases are generally classified by two main schemes; the first one is the widely accepted Ambler Molecular classification scheme, which classifies enzymes based on their amino acid sequence, and the second one is the Bush-Jacoby classification scheme, which is based on biochemical characteristics of the enzymes [56, 57]. Ambler classification classifies beta-lactams into four major sub-classes A, B, C, and D based on their distinguishing amino acid motifs. Class A, C, and D are serin carbapenemases, achieving their effect through hydrolysing their substrate by the active site, forming acyl groups for which they need serin as a cofactor. Class B are collectively called bacterial metallo beta-lactamases (MBLs), using at least one zinc ion in their active site to hydrolyse beta-lactam antibiotics. MBLs are further subdivided into the three subclasses of B1, B2, and B3 [57]. Bush-Jacoby classification classifies these enzymes into the three subgroups of 1 (cephalosporinases, equivalent of class C), group 2 (serin beta-lactamases, equivalent of class A and D), and group 3 (MBLs, equivalent of class B) [56].

AmpC cephalosporinase, from the molecular class C beta lactamase, is the most prevalent enzyme utilized by A. baumannii against beta-lactam antibiotics [53]. This enzyme is encoded by the bla gene, making the bacteria achieve resistance against penicillins as well as narrow- and extended-spectrum cephalosporins. Moreover, ampC enzymes can also cause resistance to a combination of these antibiotics along with beta-lactam inhibitors [53]. Other than ampC, A. baumannii strains may contain a wide range of different classes of beta lactamases, namely PER-1, VEb-1, CTX-M, TEM, and SHV from class A, MBLs; IMP, SIM, VIM from class B and OXA from class D [58]. Having OXA-type carbapenemases is the mainstay for resistance against these last-resort antibiotics [58]. Class D carbapenemases are capable of hydrolysing isoxazolyl penicillin drugs such as oxacillin. Although more than 574 variants of the enzyme have been identified so far, few of them have carbapenemase activity. Therefore, class D beta lactamases are subdivided into 12 main Oxacillinase groups, taking OXA abbreviation. Among them, OXA23, OXA24, and OXA58 are plasmids encoding carbapenemases and are the main variants used by the bacterium for carbapenem degradation. OXA25, OXA26 and OXA40 are other enzymes utilized by A. baumannii from class D, which are chromosome mediated, unlike the first three mentioned enzymes. Coexistence of OXA23 and a class B MBL NDM-1 has been reported in a pan-resistant isolate from India [59].

3.1.2 Target alteration of the antibiotics and modification in membrane permeability

Reduced expression of OMPs and alteration in penicillin-binding proteins (PBPs) are two mechanisms used by A. baumannii to decrease uptake and permeability to beta-lactam antibiotics. Since hydrophilic antibiotics can cross the bacterial outer cell membrane only through OMPs, a decrease in their expression can decrease the permeability of antibiotics. OprD a 43 kDa protein and CarO a 29 kDa protein are among the most studied OMPs the bacterium uses to downregulate permeability [60]. Loss of CarO, secondary to the porin’s gene disruption by a distinct insertion element, was shown to be associated with imipenem and meropenem resistance [61]. Since no specific imipenem binding site was found in CarO, it is believed that this porin forms nonspecific channels [62]. Heat-modifiable protein HMP-AB (homologous to OmpA of Enterobacteriaceae and OmpF of P. aeruginosa) and OmpW are some of the other notable OMPs identified to have an effect on the beta-lactam resistance [63, 64]. Proteomic studies showed difference in CarO expression and structural changes in isoforms of OmpW in multidrug-resistant strains of A. baumannii in comparison to control strains; interestingly, there was no difference in expression of OprD porins among the two strains [65].

PBPs play a crucial role in the synthesis of peptidoglycan used in the bacterial cell wall. They also catalyze transglycosylation and the cross-linking by transpeptidation of peptidoglycan. A. baumannii can develop beta-lactam resistance by decreasing the affinity of PBPs to beta-lactams, overproducing a critical PBP, or by producing new or altered PBPs [66].

3.1.3 Increased efflux pumps

So far, six families of efflux pumps have been identified: the resistance nodulation cell division family, the small multidrug resistance superfamily, the ATP-binding cassette (ABC) family, the major facilitator superfamily, the multidrug toxic compound extrusion family, and the recently identified proteobacterial antimicrobial compound efflux family [67]. The presence of efflux pumps provides resistance to multiple classes of antibiotics. The AdeABC efflux pump, a member of the resistance nodulation cell division family, is one of most studied and well-defined pumps in A. baumannii, conferring resistance to a group of multiclass antibiotics including aminoglycosides, fluoroquinolones, chloramphenicol, erythromycin, cefotaxime, tetracyclines, trimethoprim and ethidium bromide [68]. Once this efflux pump is overexpressed, it can confer resistance to carbapenems as well [68]. AdeABC contains a three-component structure: AdeB forms the transmembrane component, AdeA forms the inner membrane fusion protein, and AdeC forms the OMP [1, 68]. This efflux pump is encoded chromosomally and is regulated by a two-component system: a sensor kinase (AdeS) and its associated response regulator (AdeR) [69]. Point mutation within this regulatory system is associated with pump overexpression and hence, carbapenem resistance. However, mutation is not the only mechanism necessary for its overexpression [69]. Several other efflux pumps identified in A. baumannii so far include: AbeS, from a small multidrug resistance efflux pump group, AdeABC, AdeIJK, and AdeFGH of the ABC group, as well as CraA, AmvA/AedF, and Tet(B) [7].

3.2 Resistance to aminoglycosides

Most frequently, A. baumannii modifies amino or hydroxyl group by means of aminoglycoside-modifying enzymes, namely: adenylases, acetylases, methyltransferases, and phosphotransferases, to confer resistance to aminoglycosides [53]. Presence of genes coding for aminoglycoside-modifying enzymes within class 1 integrins is commonly seen in antibiotic-resistant strains of A. baumannii [70].

The bacterium uses alteration of the target ribosomal protein as the other mechanism for resistance against this class of antibiotics [71]. 16S rRNA methylation has been described for A. baumannii (armA) strains impairing aminoglycoside binding to its target site, causing high levels of resistance to all clinically important aminoglycosides, including amikacin, gentamycin, and tobramycin [72, 73]. armA is a plasmid born gene within a transposon (Tn1548) which has very similar characteristics to several genes across gram-negative organisms [73]. Besides the mentioned mechanisms, A. baumannii utilizes the AdeABC efflux pump and the AbeM pump (member of the multidrug and toxic compound extrusion (MATE) family) to increase the efflux of aminoglycosides from the cell [74]. AdeABC pumps, however, are somewhat less effective against amikacin and kanamycin due to their more hydrophilic characteristics [74].

3.3 Resistance to quinolones

Mutations in the gyrA and parC genes, resulting in phenotypic changes in DNA gyrase and topoisomerase IV causing reduced drug affinity by interference with target site binding, are the main mechanisms used by A. baumannii to resist quinolones [75]. Decreased permeability by reduced expression of OMPs and increased efflux through efflux pumps such as AdeABC and MATE pump AdeM (as mentioned for other classes of antibiotics) is used here as well, for resistance against quinolones [76]. Inhibition of binding of DNA gyrase and/or topoisomerase by protection of DNA through plasmid-encoded quinolone-resistant determinants gnrA, gnrB, and gnrS is another notable mechanism utilized by the bacterium to resist these antibiotics [77].

3.4 Resistance to colistin

Chromosomal DNA genes of the bacterium encode resistance to colistin via two mechanisms, target modification and remodeling of the outer membrane. The first mechanism is due to mutation in lipid A encoding genes (namely lpxA, lpxC, and lpxD) resulting in loss of LPS, which is an initial target for colistin [78]. The two-component system pmrAB (a response regulator and sensor kinase), in response to environmental conditions such as change in pH, Mg²⁺, and Fe³⁺, regulates the expression of genes causing synthesis of lipid A. Point mutation (as the second mechanism) in pmrA and pmrB causes upregulation of their gene expression and hence, remodeling of the outer membrane [79].

3.5 Resistance to Tetracyclines

Efflux and ribosomal protection are two mechanisms used for resistance against tetracyclines by A. baumannii. Tetracycline-specific efflux pumps encoded by the tet(A) and tet(B) determinants have been identified in the bacterium [80]. Tet(A) genes are located within a transposon similar to Tn1721, in association with an IS element [80]. Apart from these class-specific efflux pumps, tetracyclines are susceptible to the effect of multi-efflux systems such as AdeABC as well [81]. The effect on microsomal protection is conferred by tet(O) and tet(M) determinants, the later being described rarely for A. baumannii [82].

3.6 Resistance to trimethoprim-sulfamethoxazole and chloramphenicol

Resistant gene-containing integrons are very commonly seen among multidrug-resistant strains of the bacterium. The 3′-conserved region of an integron containing a qac gene fused to a sul gene has been shown to provide resistance to sulphonamides [83]. Similarly, the trimethoprim-resistant gene (dhfr) and chloramphenicol resistant gene (cat) are located on integron structures [70]. Efflux pumps also contribute to conferring resistance to these antibiotics [53].

4. Antibiotic treatment options for multi-resistant strains of A. baumannii and new non-antibiotic modalities under development

4.1 Antibiotic treatment

4.1.1 Carbapenems

For years carbapenems were a mainstay of antimicrobial therapy against A. baumannii. Since 1990 these antibiotics have been used as last-resort, broad-spectrum antimicrobials for treatment of complicated life-threatening infections caused by this bacterium. Interestingly, Acinetobacter’s first resistance to carbapenems was reported in 1985, the year imipenem was discovered, indicating that the resistance mechanism had existed even before the antibiotic’s first use [13]. Rate of resistance to carbapenems is highly geographical and is connected to regional distribution of strain types and percentage of antibiotic prescription and misuse. Currently on average, the general percentage of resistance of A. baumannii to carbapenems is 74–92% globally [84]. The percentage of resistance rises from 15% in Europe and North America to 40% in Latin America and up to 85–100% in south and east Asia [85, 86, 87]. This high resistance level renders use of carbapenems and consequently other beta-lactams ineffective in multi-resistant A. baumannii strains. Therefore, choosing treatment with carbapenems is rational in areas with lower resistance to this antibiotic.

4.1.2 Polymyxins Colistin (polymyxin E)

Polymyxins are bactericidal antibiotics that function through cell membrane disruption. Polymyxins’ positively charged cationic region bind to the negatively charged hydrophilic part of LPS, causing loss of integrity of the cell membrane [88]. The emergence of pan-resistant strains of A. baumannii led to the reintroduction of these historically discarded antibiotics (due to colistin’s cell toxicity, especially nephrotoxicity ranging between 11% up to 76% in different reports, use of colistin was discontinued shortly after its introduction in the 1950s). Colistin nephrotoxicity is suggested to be due to accumulation of the drug in proximal renal tubules causing oxidative damage [89]. Other than nephrotoxicity, colistin may induce neurotoxicity (although infrequently) when administered via the respiratory tract, causing bronchoconstriction and cough, and may result in chemical meningitis [90]. Several other medication side effects, such as ataxia, apnoea, delirium, visual disturbances, seizures, etc., have been reported in connection to use of this antibiotic [91]. Monotherapy of colistin raises two distinct complications resulting in resistance against it: on the one hand, due to administration of colistin as a pro-drug in the form of colistin methane sulfonate, achievement of therapeutic serum concentration is mostly complicated and suboptimal, causing emergence of regrowth. On the other hand, selective resistance of bacterial subpopulation, called heteroresistance, renders the antibiotic ineffective [92]. Thus, combination therapy with other active agents to achieve synergy has been proposed. Although there are still debates regarding the advantages of monotherapy vs. combination therapy, extensive review of literature has suggested combination therapy as a superior option both in terms of microbiological clearance and clinical cure [93, 94]. The combination of colistin with carbapenem or colistin with rifampicin are the most studied and well-established choices [95]. Several other combinations including colistin and aminoglycosides, daptomycin, co-trimoxazole, fosfomycin, sulbactam, etc. exist in literature [96]. Despite all of this, resistance to this last-resort antibiotic has been reported (the first report of A. baumannii being resistant to colistin was published in Czechia in 1999) and is rising worldwide [97]. Resistance to colistin in the US and Europe is documented at around 11%, with the highest resistance so far reported from India with 53%, Iran 48% and Spain 40% [13, 98].

4.1.3 Tetracyclines minocycline: tigecycline

Tetracyclines function by entering the outer membrane of gram-negative bacteria via protein channels, causing conformational changes in the RNA of the bacteria by binding to its 30S ribosomal unit [99], blocking entry of aminoacyl transferase RNA into the site A of the ribosome and consequently stopping protein production [99]. Since minocycline is a more lipophilic type (along with doxycycline) among tetracyclines, this antibiotic has better tissue penetration and antimicrobial activity [99]. As mentioned in the mechanisms of resistance section of this chapter, the main mechanism of resistance against tetracyclines is efflux of the drug by A. baumannii’s efflux pumps, namely tet(A) and tet(B). It has been shown (in vitro) that tet(A) pumps are not effective against minocycline and strains lacking tet(B) pumps are susceptible to this antibiotic [100]. Although tigecycline was synthetically designed to overcome the effect of efflux pumps (commonly seen in other tetracyclines), development of other resistance mechanisms (such as overexpression of various efflux pumps, including AdeABC), has made this once promising antibiotic far less effective [101]. Interestingly, so far, the effect of overexpression of RND pumps and increase of resistance has not been reported on minocycline and therefore, tigecycline-resistant A. baumannii can still display susceptibility to this antibiotic [99]. Due to the mentioned properties, minocycline has attracted a lot of attention during the last decade or so and has become more clinically used. Still, superiority in use of this antibiotic as monotherapy or combination therapy is not clear and more large clinical studies are needed to provide a better understanding of the efficacy of this drug. To date, published case reports (with small cohorts of patients) have shown good relative efficacy and tolerance of this antibiotic in clinical practice [102, 103, 104].

4.2 Bacteriophages

In 2010 the first report of isolation and characterization of phage against A. baumannii was published, showing lytic behavior of AB1 and AB2 phages against the bacterium [105]. Since then, many lytic phages have been identified; however, the first use of phage in a human in Western countries was reported no sooner than 2017. Being a successful attempt, this experiment paved the way for the initiation of a phage therapy programme just a year later at the University of California. They recently published a case report of the first 10 patients being treated by this modality [106]. Currently, only five clinical case reports exist on phage therapy against A. baumannii [107]. All these reports share a similar trend, using phage cocktails combined with antibiotics in more than one dose, administered intravenously or in the case of respiratory infection, nebulized. Although general positive impact was seen in all of these reports, phage therapy was not always correlated with complete recovery in all cases [107].

4.3 Monoclonal antibodies

Neutralizing virulence factors of the bacterium by the binding of monoclonal antibodies is the mainstay mechanism of action of these molecules against A. baumannii. Many studies in recent years have been dedicated to the development and safety of monoclonal antibodies in laboratory animals. So far, MAb-based treatment (a monoclonal antibody) is one modality showing promising effectivity in the management of infection in animals without driving resistance to antibiotics [108]. Historically, the clinical efficacy of MAbs and other passive immunization therapies was based on enhancing microbial clearance induced by opsonophagocytosis and activating complement system [108]. However, it has been suggested that MAbs may confer their effect not only through microbial clearance but also by altering the inflammatory settings of the host regardless of microbial count [108]. This hypothesis has been further supported by a recent publication which demonstrated that despite a decrease of bacterial burden through the effect of antibodies, the antibacterial effect of MAbs was insufficient to mediate clinical improvement, as measured by survival of the host [108]. It was suggested that efficacy of MAb therapy depends more on “normalizing the inflammatory response to infection than on reducing bacterial burden” [108]. Due to their complicated mechanism of action, safety concerns, and very expensive price in terms of mass production, so far, monoclonal antibodies are still only in animal research stages, but are providing a promising modality for the future of antimicrobial treatment.

4.4 Antimicrobial peptides

Antimicrobial peptides (AMPs) are products of eukaryotic and prokaryotic organisms as part of their innate host immune response [13]. Since they have bactericidal capacities and are broad spectrum, with low immunogenicity and low resistance, interest in them has been increasingly rising in recent years as alternatives to antibiotic treatment. As an example, efficacy of a hybrid of cecropin A and melittin has been reported in management of peritoneal sepsis caused by pan-resistant A. baumannii in an animal model [109]. Several reports have already been published showing the antimicrobial effect of Brevinin 2, alyteserin 2, catonic α-helical peptide, A3-APO, and D-RR4 against multidrug-resistant strains of the bacterium in animals [109, 110]. However, cytotoxicity, enzymatic degradation, and high production costs still need to be evaluated before we may expect to see them in clinical use for humans.

4.5 Vaccination

Since A. baumannii has an intracellular lifestyle, vaccine development strategies against it should include the induction of acquired cellular immunity with long-term memory, and be targeted for a population cohort of patients prone to be infected with the bacterium [13]. The first attempts to produce a vaccine against A. baumannii were initiated during the last decade, testing almost all modalities of vaccines, including whole-cell vaccines using attenuated or killed bacteria in mice, subunit vaccines using OmpA, Ata, and Omp 33–36, conjugate vaccines, multicomponent vaccines, and nucleic acid vaccines [13]. However, none of them have entered clinical trials so far, indicating our lack of knowledge in understanding the complex bacterium-host relationship. To date, most of the vaccine candidates developed against A. baumannii have been focusing on protein-based technologies (including the mentioned opsonophagocytic antibody-mediated killing and/or antibody-mediated toxin inhibition) [13]. Since it has been shown that polarized (T helper) Th2 will not be sufficient for long-term protection, new efforts have been shifted to achieve a mixed Th1/Th2 or Th1/Th17 response. Thanks to the technology acceleration associated with the COVID vaccine, the horizon has been expanded on nucleic acid vaccines against A. baumannii and other pathogens, and interesting developments in this area in the future are expected.

4.6 Other modalities

Some of the other modalities that have been explored in the clamor caused by the urgent need for a new treatment against multi-resistant A. baumannii include the use of clustered, regularly intercepted short palindromic repeat (Cas) systems to eliminate the resistant gene, producing susceptible species; utilization of metal chelators, essential in the expression of bacterial virulence factors, as targets for designing new antimicrobials; and toxin-carrying liposomes [111, 112].

5. Conclusion

A. baumannii is one of the most complicated and challenging to treat pathogens of the modern era of medicine. The bacterium’s complex mechanisms of resistance, its genetic set-up, and its ability to survive on different surfaces and tolerate harsh environmental conditions make A. baumannii an advanced nosocomial pathogen causing clinical outbreaks all around the world. Since most antibiotics are ineffective against it, there is an urgent need for new antibiotics or other modalities of treatment (phage therapy, vaccination, etc.) in the near future. Meanwhile, available options for the management of A. baumannii include a rational antimicrobial stewardship programme to minimize selective pressure in support of growth of multi-resistant strains, decreased misuse of antibiotics, early diagnosis and proper isolation of infected and colonized patients, and proper choice in antibiotic therapy (carbapenems, minocycline or tigecycline in case of susceptibility to these antimicrobials or combination therapy with colistin + another antibiotic agent (i.e. carbapenems) in case of resistance) once infection with such strains has been established.

6. Conflict of interest

The authors declare no conflict of interest.