Modulation of Quorum Sensing and Biofilms in Less Investigated Gram-Negative ESKAPE Pathogens

Pathogenic bacteria have the ability to sense their versatile environment and adapt by behavioral changes both to the external reservoirs and the infected host, which, in response to microbial colonization, mobilizes equally sophisticated anti-infectious strategies. One of the most important adaptive processes is the ability of pathogenic bacteria to turn from the free, floating, or planktonic state to the adherent one and to develop biofilms on alive and inert substrata; this social lifestyle, based on very complex communication networks, namely, the quorum sensing (QS) and response system, confers them an increased phenotypic or behavioral resistance to different stress factors, including host defense mechanisms and antibiotics. As a consequence, biofilm infections can be difficult to diagnose and treat, requiring complex multidrug therapeutic regimens, which often fail to resolve the infection. One of the most promising avenues for discovering novel and efficient antibiofilm strategies is targeting individual cells and their QS mechanisms. A huge amount of data related to the inhibition of QS and biofilm formation in pathogenic bacteria have been obtained using the well-established gram-positive Staphylococcus aureus and gram-negative Pseudomonas aeruginosa models. The purpose of this paper was to revise the progress on the development of antibiofilm and anti-QS strategies in the less investigated gram-negative ESKAPE pathogens Klebsiella pneumoniae, Acinetobacter baumannii, and Enterobacter sp. and identify promising leads for the therapeutic management of these clinically significant and highly resistant opportunistic pathogens.


INTRODUCTION
Many international organizations have declared antibiotic resistance (AR) to be a global public health concern, requiring concerted action plans to tackle this problem and to decrease its huge social, medical, and economic burden (Michael et al., 2014;Spellberg et al., 2016).
The emergence of AR strains is favored by microbiostatic substances, which only inhibit microbial multiplication, but also by the improper administration of microbicidal drugs with respect to dose interval and active concentration (Vrancianu et al., 2020a,c). Some of the multiple negative consequences of AR are the unprecedented increase in infectious disease frequency, illness by gram-negative bacteria in intraspecies communication are the N-acyl homoserine lactones (AHLs) (Watson et al., 2002), while gram-positive bacteria use autoinducing peptides (AIPs) (Jayaraman and Wood, 2008).
Among many important advantages offered by QS to bacterial pathogens, there is the ability to colonize and/or invade the host, as well as to develop biofilms on natural tissues (skin, mucosa, endothelial epithelia, and teeth) or medical devices (central venous catheters, peritoneal, urinary catheters, dental materials, cardiac valves, intrauterine contraceptive devices, contact lenses, and other implants) (Lazãr and Chifiriuc, 2010) and, thus, to persist in the host.
The disruption of bacterial QS by QS inhibitors (QSIs) represent a promising approach for fighting BAIs (Davies et al., 1998;Chandra Kalia, 2013;Fong et al., 2018). As most of the described QS signaling systems include twocomponent systems (TCS), namely, the AI (QS molecule) and the receptor, also known as response regulator (RR), which impacts on the transcription of target genes (Papenfort and Bassler, 2016), the QS modulation strategies follow one of two directions: (i) interference with signal generation and (ii) signal reception . Both directions cluster various approaches and are summarized in Table 1. However, in some situations, the AI (i.e., AI-3) can bind to a sensor kinase (SK), instead of an RR (Kim et al., 2020).
Numerous in vitro and in vivo experimental data (Brackman et al., 2011) on biofilm formation and antibiofilm unconventional strategies were reported (Roy et al., 2018), but their efficiency and safety need to be validated in clinical studies. The local delivery of QSIs in biofilms seems to be a promising lead, allowing a quick assessment of therapeutic efficiency. Therefore, developing appropriate local delivery systems and ways would be of most importance in future research. Despite the huge amount of data, only a few of the available QSIs are reaching the stage of clinical studies and, eventually, the bedside, and sometimes they have been approved for other biological activities, such as antimicrobial (e.g., azithromycin, which inhibits the alginate synthesis; vegetal extracts; natural compounds, which can also act as QSIs in subinhibitory concentrations) or antitumoral agents (Saurav et al., 2016;Rémy et al., 2018). There are also few patents using lactonase or acylase QSIs, proposed mainly as antibiofouling agents . From the most challenging resistant species, known as ESKAPE (Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, and Enterobacter species), and then changed to ESCAPE (E. faecium, S. aureus, Clostridium difficile, A. baumannii, P. aeruginosa, and Enterobacteriaceae), the gram-negative bacilli are the most problematic because of the lack of novel classes of antimicrobial agents efficient against these multiple-(MDR), extended (XDR), and pan-drug resistant (PDR) strains. From February 2017, these emerging MDR bacteria are also listed as critical in the WHO (World Health Organization) priority pathogen list for the research and development of new antimicrobials 1 .
The most known investigated biofilm regulators and their described mechanisms for Klebsiella sp., Acinetobacter sp., and Enterobacter sp., which are less investigated MDR ESKAPE agents, are presented in Table 2.
Biofilm-associated infections often involve ESKAPE pathogens as etiological agents. Therefore, the purpose of this paper is to reveal and discuss the progress on the development of antibiofilm and anti-QS strategies in the less investigated gram-negative ESKAPE pathogens, such as A. baumannii, K. pneumoniae, and Enterobacter sp., and their potential contribution to the personalized control of infections produced by these emerging opportunistic pathogens.
QS SIGNALING AND MICROBIAL BIOFILMS IN Acinetobacter sp., Klebsiella sp., AND Enterobacter sp.
Among the most dangerous threats concerning infection control gathered under the acronym ESKAPE, some of them are less investigated than the well-known QS experimental models, such as P. aeruginosa and S. aureus. However, the increasing incidence in the etiology of hospital-acquired infections and BAIs as well as the multiple intrinsic and acquired resistance mechanisms of the gram-negative species from the ESKAPE group, Acinetobacter sp., Klebsiella sp., and Enterobacter sp., justify the urgent need for the development of novel and effective antimicrobial strategies to target them. Table 3 summarizes some of the recent approaches investigated for BAI management in A. baumannii, Klebsiella sp., and Enterobacter sp.
Acinetobacter spp. is one of the hospital "superbugs, " considered today the most important nosocomial pathogen and the first priority on the WHO pathogen list requiring novel antibiotics, mainly due to its tolerance to desiccation, MDR mechanisms, and ability to develop medical device BAIs. Biofilm-forming ability seems to be much higher in clinical than in environmental isolates. The ability of A. baumannii clinical strains to form biofilms on abiotic substrata and epithelial cells increases their genetic resistance. Thus, at least 92% of the biofilm-forming nosocomial isolates seem to be MDR (Babapour et al., 2016), while an increased detection rate and expression of the blaPER-1 gene encoding for betalactam resistance were recorded in biofilm-forming isolates (Lee et al., 2008;Gaddy and Actis, 2009). It was reported that CarO and OmpA outer membrane proteins are interacting physically with the OXA-23 carbapenemase, leading to an enhanced carbapenem resistance by cumulating non-enzymatic and enzymatic resistance mechanisms (Chopra et al., 2013).
-protein extracellular fibers involved in host cell adhesion and invasion, control the formation and architecture of E. cloacae biofilms, modulate adherence to abiotic and biotic surfaces, -regulate biofilm formation in Enterobacter sp. Mezzatesta et al., 2012;Soria-Bustos et al., 2020Soares et al., 2016 Recent genomic studies highlight the presence of much greater virulence determinants in A. baumannii than previously thought. The virulence genes, including those involved in biofilm formation and the current progress in developing antibiofilm agents in A. baumannii-derived infections, are summarized by Eze et al. (2018). The success of A. baumannii in host colonization mainly depends on its adherence capacity, but other virulence factors are also incriminated. These include: K1 capsular polysaccharides, surface antigen protein 1, outer membrane porins (which are involved in adhesion, biofilm formation, and drug resistance), Bap (biofilm-associated protein), inflammatory cytokine induction molecules (Eze et al., 2018;Harding et al., 2018), iron transport systems and siderophores (such as acinetobactin), poly-(1-6)-N-acetylglucosamine (PNAG, which is one of the most important structures for biofilm formation correlated with higher resistance), activation of phosphomannomutase/phosphoglucomutase (algC) gene [encoding for alginate and lipopolysaccharide (LPS) during biofilm development and correlated with the MDR level], type I chaperone-usher pilus system (Csu pili) regulated by QS (which is critical for the adherence to inert substrata), LHp2_11085 factor involved in adherence to inert and cellular substrata, and A1S_0114 regulatory gene of surface proteins and pili-assembly system expression (Shin et al., 2009;Xiang et al., 2012;Zarrilli, 2016;Álvarez-Fraga et al., 2017). The enormous adaptability of resistant strains, supported by the acquisition and dissemination of resistance and virulence markers, renders it a dangerous opportunistic pathogen, particularly in the case of immunosuppressed patients from the intensive care units (Vrancianu et al., 2020b,c).
The QS system in Acinetobacter sp. has been described as homologous to the LuxR receptor (AbaR) and LuxI synthase (AbaI) system from V. fischeri. However, phylogenetic studies indicate that its QS genes (abaI and abaR) were acquired horizontally from Halothiobacillus neapolitanus (Bhargava et al., 2010(Bhargava et al., , 2015a. More than 63% of the Acinetobacter spp. analyzed strains produced more than one AHL (≥C10), including N-(3-hydroxydodecanoyl)-L-homoserine lactone (OH-dDHL). The QS mechanism plays an important role in Acinetobacter spp. motility, expression of multidrug efflux pumps, and biofilm development (Figure 1). However, little is known about the cascade of genes associated with various mechanisms controlled by the QS system in A. baumannii (López et al., 2017). Iron limitation seems to regulate the expression of virulence and QS factors in A. baumannii clinical strains, including the biofilm development capacity (Kim H. W. et al., 2013;Modarresi et al., 2015). In their turn, the QS signaling molecules could chelate iron, inducing the occurrence of the stress response (Minandri et al., 2016). This could explain the persistence of A. baumannii biofilms in iron-depleted environments. Siderophores can chelate iron, zinc, copper, and other metals, interfering thus with the activity of antibiotics and host molecules while modulating the oxidative stress (Figure 1). Studies related to the inhibition of QS in Acinetobacter are limited. The lack of QSIs for this clinically significant pathogen is a mounting concern, especially with the increasing frequency of MDR strains (Subhadra et al., 2016).
Klebsiella pneumoniae is a versatile opportunistic pathogen, exhibiting many virulence features, allowing it to colonize different inert substrata, including the urinary catheters, such as fimbriae (of type 1 and 3), capsular polysaccharides, factors involved in aggregative adhesion, and siderophores (Stahlhut et al., 2012;Vuotto et al., 2014;Paczosa and Mecsas, 2016). If in the preantibiotic era K. pneumoniae was considered an important etiological agent of community-acquired (CA) infections (such as severe pneumonia in debilitated patients), presently, because of its high resistance to last-resort antibiotics, such as carbapenems and colistin, the spectrum of K. pneumoniae infections has broaden, including CA and healthcare-associated life-threatening infections (Lederman and Crum, 2005). K. pneumoniae is involved in 5-7% of all healthcare-associated infections (Clegg and Murphy, 2016). K. pneumoniae uses QS TCSs to control the host-pathogen interactions and to coordinate the virulence and the AR mechanisms, including the rapid development of biofilms on abiotic surfaces (Balestrino et al., 2005;Srinivasan et al., 2012;Tiwari et al., 2017). The QS system in K. pneumoniae is mediated by the AI-2 AI encoded by a homolog of luxS from V. harveyi (Figure 2), and also by N-octanoyl homoserine lactone (C8-HSL) and N-3-dodecanoyl-L-homoserine lactone (C12-HSL) (Balestrino et al., 2005;Yin et al., 2012). Mutations in luxS are correlated with an increased expression of two LPS synthesis-related genes, wbbM and wzm, also involved in biofilm formation (Sun et al., 2016). Recent data confirm the involvement of the AI-2 QS system in the expression of LPS and PNAG biosynthesis, as well as biofilm development of an extensively drug-resistant K. pneumoniae clinical isolate (Chen et al., 2020; Figure 2).
Enterobacter genus, especially through its two most prominent species, Enterobacter aerogenes and Enterobacter cloacae, is a versatile nosocomial pathogen with serious implications in respiratory and urinary tract infections (Sanders and Sanders, 1997). Unfortunately, little is known about quorum control and pathogenesis in this group of bacteria. Most of the available data come from food-associated studies. It uses C4 and C6-HSLs as QS signaling molecules (Yin et al., 2012;Lau et al., 2013), encoded by a LuxR homolog, which has been found to negatively regulate bacterial adhesion and biofilm formation (Shankar et al., 2013). AI-2-mediated QS has also been suggested to play a role in intercellular communication within Enterobacter spp. (Figure 3), as Lsr-type receptors have been found in strains of Enterobacter cancerogenus, E. cloacae, and Enterobacter mori (Rezzonico et al., 2012;Tay and Yew, 2013). The AI-3 activity, initially reported in enterohemorrhagic E. coli O157:H7, was also detected in E. cloacae isolated from normal microbiota (Reading and Sperandio, 2006).
A recent study documented the cloning and characterization of a transcriptional regulator, luxR homolog from Enterobacter asburiae, as well as the functionality and specificity of EasR protein in response to different AHL signaling molecules to activate gene transcription from QS target promoters (Figure 3). However, further genome-wide comparative transcriptomics are needed to elucidate the possible roles of QS, especially in the pathogenicity of different Enterobacter spp. (Lau et al., 2020; Figure 3).

BIOFILM AND QS MODULATORS IN Acinetobacter, Klebsiella, AND Enterobacter
As BAIs produced by ESKAPE pathogens are currently very difficult to treat, the modulation of key molecular mechanisms  of biofilm development, including the QS signaling, represents a very promising alternative in handling such infections (Lazar and Chifiriuc, 2011).
Below, we present the results of different studies that have reported the use of natural and synthetic compounds or other mechanical, physical, or biological strategies to modulate QS and inhibit biofilm development in Acinetobacter, Klebsiella, and Enterobacter species.

Antibiotics and Antiseptics
When used in specific amounts, antibiotics could act as intermicrobial signaling agents, impacting on the biofilm homeostasis, motility, and type three secretion system (TTSS) (Linares et al., 2006). Subinhibitory concentrations of antibiotics modulate the QS mechanism, in contrast to bactericidal effects observed at high concentrations (Rémy et al., 2018). Some antibiotics (i.e., streptomycin, gentamicin, and myomycin), utilized in subinhibitory concentrations, have been found to inhibit QS signaling in A. baumannii. Streptomycin can act as an antagonist of AbaR and inhibits QS in A. baumannii by downregulating the abaI gene, encoding for the AI synthase, resulting in the corresponding decrease in 3-oxo-C12-HSL production (Saroj and Rather, 2013). It has been shown that subinhibitory concentrations of trimethoprim-sulfamethoxazole completely inhibit the pilin expression in A. baumannii, disrupting the biofilms formed on inert substrata and promoting a planktonic lifestyle, with bacterial cells more susceptible to antimicrobial agents (Harding et al., 2018).
In associated wound infections, drug-resistant A. baumannii forms biofilms, which are very recalcitrant to topical antibacterial agents. However, when associated with ambroxol, a respiratory mucus secretolytic agent, the topical antibacterial agents proved to be effective against wound-associated BAIs (Huang et al., 2012).
A number of Food and Drug Administration (FDA)approved drugs with different pharmacological activities, including erythromycin (antibiotic), chloroquine (antimalarial), levamisole (antiparasitic), and propranolol (adrenergic blocker), were recently proven to interfere with QS and virulence in MDR A. baumannii clinical isolates. These drugs repressed the expression of abaI gene in vitro and showed significant virulence repression in A. baumannii, both in vitro and in vivo, expressed by improved mice survival rates. In addition, molecular docking studies against AbaI and AbaR proteins of QS system in A. baumannii revealed the potential inhibition of QS by these drugs (Seleem et al., 2020).

Natural QS Modulators
Quorum sensing inhibition was first observed in natural habitats (Pietschke et al., 2017). The QSIs were synthesized either by other organisms to protect from pathogenic bacteria or by bacteria to gain a survival advantage (Saurav et al., 2017). The concept of QS modulators includes numerous types of natural or synthetic molecules, but also phages and cells, like quorum quenching (QQ) bacteria.
The natural QSIs are used as a backbone to obtain synthetic QSIs. As antipathogenic drugs, the QSIs could be used either to treat or prevent infections and synergize with current antibiotics and anti-infectious immune effectors (Romero et al., 2012;Tang and Zhang, 2014;Pietschke et al., 2017).
Microbial metabolites produced by drinking water bacteria proved to inhibit Acinetobacter calcoaceticus biofilms (Simões et al., 2008). The 5-episinuleptolide, a natural compound isolated from Sinularia leptoclados inhibited the A. baumannii biofilm development as well as the MDR A. baumannii strains by decreasing the poly-PNAG expression (Tseng et al., 2016). There are reports on utilizing several natural compounds as QSIs that interfere with AHL receptors in A. baumannii, including patulin, clavacin, vanillin, and alliin (Cady et al., 2012). Linalool, a major compound Coriandrum sativum essential oil, flavonoids from Glycyrrhiza glabra, and Salvia glutinosa essential oil have been shown to exhibit QSI as well as anti-A. baumannii virulence and biofilm activity (Bhargava et al., 2015b;Alves et al., 2016;Tutar, 2016;Shaaban et al., 2019). Vegetal QSIs have also been proposed to be used as natural food preservatives to prevent opportunistic food-borne infections. Petunidin, a dark-red or purple water-soluble pigment found in many red berries (an O-methylated anthocyanidin of the 3-hydroxy type) at sub-MIC values, drastically reduced the EPS production in K. pneumoniae, the antibiofilm effect being much enhanced when acting synergistically with conventional antibiotics. Molecular modeling studies predicted that petunidin induces changes in the 3D structure of the LasR receptor protein, suggesting that it acts as an effective competitive inhibitor of QS signaling through the LasR receptor pathway (Gopu et al., 2015). The essential oil from cumin seeds reduced biofilm formation in K. pneumoniae, but without any direct connection with QS pathway inhibition.
Lactonases are natural enzymes able to degrade AHL-type AIs. Two clusters of AHL lactonases were described in prokaryotes: AiiA and AttA. The AiiA lactonase has been shown to decrease the number of E. cloacae cells during early biofilm formation in continuous biofilm models, while flagellin and outer membrane protein expressions were downregulated (dos Reis Ponce et al., 2012;Kim I. H. et al., 2013). AttA cluster was described in K. pneumoniae and regulates the fermentative metabolism and virulence in this bacterium (Sun et al., 2016).
Engineered and natural lactonases (e.g., a thermostable engineered mutant of phosphotriesterase-like lactonase from Geobacillus kaustophilus, with enhanced catalytic activity on different AHLs ranging from 6 to 12 carbons) induced a significant decrease in A. baumannii-associated biofilm development (Chow et al., 2010(Chow et al., , 2013(Chow et al., , 2014. The large AHL spectrum of these enzymes proves their promising potential to fight infections associated with various gram-negative bacteria using AHL-mediated QS signaling. It has also been reported by Zhang et al. (2017) that the recombinant enzyme, MomL, is also able to degrade QS molecules, and thus reducing the biofilm formation and increasing the in vitro susceptibility of biofilm cells to different antibiotics of some Acinetobacter sp. strains and of P. aeruginosa PAO1. However, the results were strain dependent, and when this enzyme was tested against polymicrobial biofilms and wound-associated biofilm infections, it was ineffective, probably due to the fact that the in vivo conditions could affect the stability of the enzyme and its penetration through the biofilm matrix (Zhang et al., 2017).
Antimicrobial peptides (AMPs) are considered promising candidates for developing antimicrobial and anti-inflammatory agents and an example of how the natural antimicrobial strategies from the living world could be exploited or mimicked to create effective antimicrobial drugs (Aminov, 2010). It was demonstrated that the AMP LL-37 disrupted the structure of A. baumannii biofilms at low concentrations of 2.5 µg/ml (Shin et al., 2009). Also, a natural AMP complex (defensin, cecropin, diptericin, and proline-rich peptide families that are produced during bacterial infections) from the blow fly maggot Calliphora vicina has been proven to be active both on the cellular and matrix components of A. baumannii biofilms, at the same time lacking toxicity toward human immune cells (Gordya et al., 2017). Magainin 2 AMP has been proven to be an effective treatment for A. baumannii infections (Kim et al., 2018). Many other AMPs, such as CAMEL (a hybrid AMP consisting of cecropin from Hyalophora cecropia and melittin from Apis mellifera), pexiganan, cecropins identified in Musca domestica, and myxinidin isolated from Myxine glutinosa, revealed antibiofilm activity against resistant A. baumannii strains (Han et al., 2017). Natural AMPs can be a starting point for the biosynthesis of AMPs with similar functions, being an attractive therapeutic option for preventing and controlling A. baumannii BAIs (Vrancianu et al., 2020a).

Synthetic Modulators
Antagonists of diguanylate cyclase enzyme that synthesize c-di-GMP, a second messenger signal essential for biofilm formation, were proven to inhibit QS and biofilm formation in A. baumannii (Sambanthamoorthy et al., 2014). Recently Subhadra et al. (2016) reported the efficacy of virstatin, a small organic molecule, as an inhibitor of biofilm formation and motility in A. baumannii and Acinetobacter nosocomialis, acting by inhibiting the anoR/I signaling pathway (Subhadra et al., 2016). Various non-native AbaR ligands inhibited AHL-mediated QS and, subsequently, A. baumannii surface motility and biofilm formation (Stacy et al., 2012). One of the strongest AbaR antagonists (with very low IC50 values less than 20 µM) largely contained aromatic acyl groups, whereas the AbaR agonists closely resembled OH-dDHL (Stacy et al., 2012). A 2-aminoimidazole-based antibiofilm agent proved to effectively decrease biofilm development on indwelling medical devices (Peng et al., 2011). The dihydrofolate reductase inhibitor N2, N4-disubstituted quinazoline-2,4-diamines, has been shown to decrease by 90% the number of biofilm-embedded cells at concentrations similar to MIC, being more effective than tigecycline (Fleeman et al., 2017).

QSI-Antibiotic Synergic Combinations
QSIs often exhibit a synergic antibiofilm activity with antibiotics (Algburi et al., 2017). Scientists suggest using furanone in combination with antibiotics; this approach being more acceptable by patients (Grandclément et al., 2015). The antivirulence compounds affecting the cell wall composition may render bacteria more susceptible to antibiotics; therefore, the association of antivirulence agents with current antibiotics could be anticipated as efficient against biofilms (Escaich, 2010). The QS-controlled bacterial adherence and colonization could be inhibited using novel inhibitors of pili synthesis, represented by sortases or specific inhibitors of TTSS. Synergic combinations of antibiotics and QSIs are expected to be soon evaluated for QS modulation in A. baumannii, Klebsiella sp., and Enterobacter sp.

Nanomaterials
Nanotechnology offers promising leads for fighting BAIs by developing nanoantimicrobials and antibiofilm materials and by improving the drug loading and the controlled release of antimicrobial agents into biofilms. Numerous nanostructured materials have been developed to target biofilm pathogens, including the less investigated ESKAPE gram-negative species discussed in this study. Liposomes of different compositions proved to be efficient carriers for ciprofloxacin and meropenem against K. pneumoniae biofilms (Gubernator et al., 2007) and polymyxin B/clarithromycin against A. baumannii and Acinetobacter lwoffii biofilms (Khan et al., 2016;Halbus et al., 2017).
Gallium nitrate is a potent inhibitor of A. baumannii biofilm formation and a disruptor of mature biofilms developed in human serum, probably also due to iron depletion in the multicellular communities formed by A. baumannii (Runci et al., 2016).
Metallic nanoparticles (NPs) have a great potential for antimicrobial applications, exhibiting multiple mechanisms of action, such as membrane lesions induced by direct contact or indirectly, by the release of free metal ions, protein inactivation, nucleic acid damages, and release of reactive oxygen species (ROS) (Grumezescu et al., 2015;Samrot et al., 2020;Tripathy et al., 2020). They could also exhibit synergic action with the host immune effectors. Metallic NPs can also be associated with the current antibiotics or other pharmaceutically active compounds to overcome the resistance threat, particularly in hospital settings (Tudose et al., 2015a(Tudose et al., ,b, 2016. The small size and tailored properties of NPs seem to represent an advantage for penetrating more efficiently the biofilm matrix (Holban et al., 2016;Abdelghany et al., 2020). Silver NPs alone and associated with biocides or antibiotics (imipenem) proved very active on A. baumannii, both planktonic and biofilm growth (Hendiani et al., 2015). Zinc oxide NPs were also reported to impact the biofilm formation of different gram-positive and gramnegative pathogens, being considered future nanoantibiotics (Visinescu et al., 2015(Visinescu et al., , 2018Muzammil et al., 2018). It has also been shown that K. pneumoniae uropathogenic strains isolated from complicated urinary tract infections have shown decreased adherence to silver-treated silicone or latex catheters (Gabriel et al., 1995). Titanium NPs also show promising antibacterial and antiadhesive properties against A. baumannii and K. pneumoniae strains (Ibrahem et al., 2014).
The medical device-associated infections could be prevented by developing antiadherent materials or coatings. For example, "biospecific polymers" coated with antiadhesive molecules or doped with inhibitory biofilm-associated gene expression could represent an alternative (Pascual, 2002). Urinary catheters coated with nitrofurazone proved to have an enhanced resistance to biofilm development by ESKAPE urinary pathogens (Johnson et al., 2012;Zhu et al., 2019).

Biofilm Dispersal Diffusible Signal Factors
Bacteria can induce dispersal to escape from the biofilm macrostructure in response to a broad range of input signals. The biofilm dispersal process is the starting point of systemic infections since it triggers the release of bacteria into the host (Marks et al., 2013;Guilhen et al., 2017). Therefore, knowledge of the regulation of dispersal factors would help control the development of BAIs and the systemic spread of bacteria. It is known that biofilm dispersal could be triggered by (i) environmental factors [i.e., availability of iron (Musk et al., 2005;Glick et al., 2010), carbon source (Uppuluri et al., 2010;Bonnichsen et al., 2015), presence of heavy metals (Petrova et al., 2015), temperature (Nguyen et al., 2015), pH (Uppuluri et al., 2010), and oxygen limitation ] and (ii) bacteria and host-produced signaling molecules [i.e., AHLs, AIPs, diffusible signal factors (DSFs) (Ueda and Wood, 2009;Tao et al., 2010;Periasamy et al., 2012), human intestine epithelial cell signals (Sanchez et al., 2016), and nitric oxide (Li et al., 2013)].
Diffusible signal factors was originally found in Xanthomonas campestris and is a new family of widely conserved QS fatty acid signals in gram-negative bacteria, which regulate biofilm formation, motility, virulence, and AR (Deng et al., 2014). Studies showed that DSFs act as interspecies biofilm regulators since, for example, DSF produced by P. aeruginosa can disperse biofilms of other gram-negative (i.e., E. coli, K. pneumoniae, and Proteus mirabilis), gram-positive (i.e., Streptococcus pyogenes, Bacillus subtilis, S. aureus), but also yeast (Candida albicans) strains (Davies and Marques, 2009). However, little is known regarding DSF regulation in the development and dispersal of K. pneumoniae, A. baumannii, and Enterobacter sp. biofilms. Recent studies showed that DSFs and other fatty acids inhibit key virulence mechanisms, such as planktonic growth, capsule production, and cell adhesion and induce biofilm dispersal in K. pneumoniae (Rahmani-Badi et al., 2014;Gupta et al., 2020;Kumar et al., 2020). Chowdhury et al. (2021) reported a cis-2hexadecenoic acid (c2-HAD) DSF homolog encoding gene (rpfF) in Enterobacter sp. This DSF was proven to control the intestinal invasion of Salmonella sp. and control the main virulence regulon in this microorganism (Chowdhury et al., 2021). c2-HAD is supposed to interfere also with the virulence of other intestinal gram-negative bacteria, including Enterobacter species. Several mono-unsaturated chain fatty acids that could act as DSFs were demonstrated to affect QS communication and inhibit motility and biofilm formation of A. baumannii clinical isolates. These fatty acids decreased the expression of the regulator abaR from the LuxIR-type QS system. Consequently, they reduced the AHL production with a direct impact on biofilm dispersal (Nicol et al., 2018).
All the strategies proposed above are based on the molecular regulation of key phenotypes of virulence and resistance. These approaches are preferred in recent years since researchers believe that signaling modulation could prevent the development of the infectious process and the selection of resistant mutants compared to classical antibiotics. QS and biofilm control by such molecules could not interfere with population fitness but with some social behavior of microorganisms, which are key for their pathogenesis.

CRISPR (Clustered Regularly Interspaced Short Palindromic Repeat) System
One of the most attractive leads to the fight against bacterial resistance is the CRISPR-Cas system, first described by Ishino et al. (1987). CRISPR-Cas is considered a bacterial immune defense system that could specifically recognize and degrade foreign nucleic acids. The CRISPR platform has been used to achieve rapid genomic editing by deletions, insertions, and point mutations, to investigate the oxidative stress (OxyR) mechanisms, as well as the abaI gene role in biofilm formation of A. baumannii Vrancianu et al., 2020a).
Studies suggested that certain bacteria employ the Cas proteins of CRISPR-Cas3 systems to target their own genes, which also alters the virulence during host invasion. It seems that numerous gram-negative bacteria use QS signaling to control adaptive immunity through the regulation of multiple CRISPR-Cas systems. Such interference has been revealed in Serratia sp., Pseudomonas sp. (Høyland-Kroghsbo et al., 2017;Noirot-Gros et al., 2019), and Salmonella sp. (Cui et al., 2020), where the tool is currently investigated to understand QS molecular signaling and biofilm formation. Cas3 deletion upregulated the luxS regulated operon related to QS and downregulated biofilm-forming-related genes in Salmonella interfering with pathogenicity island 1 genes related to the TTSS (Cui et al., 2020). Figure 4 shows the main phenotypes that impact on biofilm development and could be targeted for the management of ESKAPE infections.

IN VIVO MODELS OF INFECTION
Animal models are crucial to developing new therapeutics and vaccines and play critical roles in the assessment of understanding infection. Over the years, several animal models were developed for A. baumannii, K. pneumoniae, and Enterobacter sp.
FIGURE 4 | Overview of the modulated phenotypes in biofilms produced by Acinetobacter, Klebsiella, and Enterobacter, which could be modulated by diverse QS modulators, NPs, and antibiotics. Nanoparticles can inhibit the attachment and early biofilm development, being also able to deliver antibiotics, QSIs, and other bioactive molecules in the biofilm. Phenotypes such as motility and attachment are modulated by such molecules, and this controls biofilm initiation. Maturation is influenced by EPS inhibition and QS interference, which also impacts the metabolic changes in the biofilm such as stress response, tolerance, and persistence. Finally, dispersion of mature biofilms could be also controlled by some molecular inhibitors controlling cell detachment.
Pneumonia mouse models have been most widely developed in A. baumannii. However, most of the tested strains do not infect immunocompetent mice or induce only self-limiting pneumonia with no or very limited local and systemic dissemination. Other studies use immunocompromised (i.e., neutropenic) mice or treat mice with mucin or agar to increase host susceptibility to A. baumannii and bacterial virulence (van Faassen et al., 2007;Eveillard et al., 2010). Despite their limitation, these mouse models are useful for investigating bacteria virulence and the development of the infectious process. Recently, a mouse model of A. baumannii-associated pneumonia using a clinically isolated hypervirulent strain showed reliable reproduction of the most relevant features of human acute pulmonary infection and pathology (Harris et al., 2013). A similar model (Luna et al., 2019) was utilized to analyze the efficiency of antibiotic combinations in MDR A. baumannii pneumonia. The study demonstrated the synergistic effects of the combination of colistin with fosfomycin and minocycline, respectively, as therapeutic options in carbapenem-resistant A. baumannii mouse infection (Ku et al., 2019). Mouse models are also investigated to reveal A. baumannii strains and clones, which are associated with increased risk fatality and are circulating in the human population (Nutman et al., 2020).
Caenorhabditis elegans is another model developed for investigating A. baumannii virulence and biofilm development in vivo. The nematode is currently the preferred model for screening the infection, resistance, and virulence correlations in most clinically relevant Acinetobacter species (Cosgaya et al., 2019).
Few in vivo models were developed to study K. pneumoniae infection. Wax moth Galleria mellonella has been utilized to study key virulence mechanisms in Klebsiella sp., such as cell death associated with bacterial replication, avoidance of phagocytosis by phagocytes, the attenuation of host defense responses, and the production of antimicrobial factors. Numerous studies support the utility of G. mellonella as a surrogate host for assessing infections with K. pneumoniae (Insua et al., 2013). However, some research reports the better utility of murine models to investigate K. pneumoniae infection and host interaction. Along with their proven utility in the elucidation of pneumonia mechanisms, mouse models were recently used to evaluate K. pneumoniae gastrointestinal (GI) colonization and host-tohost transmission. Using an oral route of inoculation and fecal shedding as a marker for GI colonization, authors showed that K. pneumoniae can asymptomatically colonize the GI tract in immunocompetent mice and modifies the host GI microbiota. A hypervirulent K. pneumoniae isolate evaluated in that study was able to translocate from the GI tract and cause a hepatic infection that mimicked the route of human infection. Authors claim expression of the capsule is required for colonization. Also, treatment with antibiotics of infected mice led to changes in the host microbiota and the development of a transient supershedder phenotype, which enhanced transmission efficiency. Therefore, mouse model can be used to determine the contribution of host and bacterial factors toward K. pneumoniae dissemination (Young et al., 2020).
In Enterobacter sp., in vivo models of infection are scarce. G. mellonella larvae were recently studied to determine the antibacterial efficacy of various drugs and proved its utility also for the investigation of host-pathogen interactions in E. cloacae. The study concluded that G. mellonella killing significantly depends on the number of E. cloacae cells injected in a dose-dependent manner. Moreover, survival can be reduced by increasing the postinoculation temperature. Also, treatment of lethal E. cloacae infection with antibiotics with proven in vitro activity significantly prolonged the survival of larvae, as compared with antibiotics to which the bacteria were resistant.
The therapeutic benefit arising from the administration of antibiotics was also correlated with a reduced burden of E. cloacae cells in the hemolymph (Yang et al., 2017).
Our study highlights that G. mellonella larvae proved to be the most investigated in vivo infection model for A. baumannii (Jeon et al., 2019), K. pneumoniae, and Enterobacter sp. (Cieślik et al., 2021). However, more models currently applied for other gram-negative bacteria (i.e., Drosophila melanogaster, zebrafish, mouse, rat) are expected to emerge in the near future in order to enhance knowledge regarding biofilm infections determined by less investigated emerging ESKAPE pathogens.

CONCLUSION
To the best of our knowledge, this is the first paper focusing on the current progress in developing antibiofilm and anti-QS strategies for fighting the less investigated gramnegative ESKAPE pathogens: K. pneumoniae, A. baumannii, and Enterobacter sp.
The surveyed literature reveals some promising leads for the development of efficient strategies against these problematic superbugs, such as combinations of QSIs and/or antibiotics administered locally or with improved and controlled targeted delivery by using nanocarriers. Researchers are currently exploiting the great perspectives offered by CRISPR-Cas in the research of BAIs. It will probably soon be applied in the investigation of less analyzed ESKAPE pathogens.
A promising priority lead is represented by natural QSIs that could provide an ecological approach, with great therapeutical and preventive value, and can be used as the backbone to obtain synthetic, non-pollutant QSIs. This approach will foster the development of social microbiology, which will exploit the antagonistic biological relationships for finding attractive and intelligent anti-infectious strategies.
The development of QS modulation strategies for clinically significant biofilm-producing pathogens such as K. pneumoniae, A. baumannii, and Enterobacter sp. could be of mounting importance for effectively controlling the nosocomial and CA BAIs, especially with the continuing evolution of MDR, XDR, and PDR strains. QS modulators are also less likely to select for resistance and eventually would have fewer side effects and ecotoxicity.
In vivo models are very useful to decipher molecular mechanisms during infection and also the utility of newly developed agents aiming to control virulence, biofilm modulation, and resistance of less investigated ESKAPE bacteria. More and specific infection models are expected to emerge in the next few years.