VIRULENCE 2021, VOL. 12, NO. 1, 2201–2213 https://doi.org/10.1080/21505594.2021.1950268 RESEARCH PAPER Importance of twitching and surface-associated motility in the virulence of Acinetobacter baumannii Jordi Corral, María Pérez-Varela, Miquel Sánchez-Osuna, Pilar Cortés, Jordi Barbé, and Jesús Aranda Departament De Genètica I Microbiologia, Facultat De Biociènces, Universitat Autònoma De Barcelona, Barcelona, Spain ABSTRACT ARTICLE HISTORY Acinetobacter baumannii is a pathogen of increasing clinical importance worldwide, especially Received 08 October 2020 given its ability to readily acquire resistance determinants. Motile strains of this bacterium can Revised 22 June 2021 move by either or both of two types of motility: (i) twitching, driven by type IV pili, and (ii) surface- Accepted 26 June 2021 associated motility, an appendage-independent form of movement. A. baumannii strain MAR002 KEYWORDS possesses both twitching and surface-associated motility. In this study, we isolated spontaneous A. baumannii; twitching; rifampin-resistant mutants of strain MAR002 in which point mutations in the rpoB gene were surface-associated motility; identified that resulted in an altered motility pattern. Transcriptomic analysis of mutants lacking virulence twitching, surface-associated motility, or both led to the identification of deregulated genes within each motility phenotype, based on their level of expression and their biological function. Investigations of the corresponding knockout mutants revealed several genes involved in the motility of A. baumannii strain MAR002, including two involved in twitching (encoding a minor pilin subunit and an RND [resistance nodulation division] component), one in surface-associated motility (encoding an amino acid permease), and eight in both (encoding RND and ABC compo­ nents, the energy transducer TonB, the porin OprD, the T6SS component TagF, an IclR transcrip­ tional regulator, a PQQ-dependent sugar dehydrogenase, and a putative pectate lyase). Virulence assays showed the reduced pathogenicity of mutants with impairments in both types of motility or in surface-associated motility alone. By contrast, the virulence of twitching-affected mutants was not affected. These results shed light on the key role of surface-associated motility and the limited role of twitching in the pathogenicity of A. baumannii. Introduction extension, attachment, and retraction of type IV pili Acinetobacter baumannii is a gram-negative bacterium [4,5]. It is associated with surface attachment and that causes nosocomial infections in hospitals world­ biofilm formation in a process that includes the pilA, wide, mainly ventilator-associated and bloodstream pilD, and pilT genes [4,6] and the GacS/GacA two- infections in critically ill patients [1]. The pathogenic component regulatory system [7]. Surface-associated success of this bacterium can be attributed to its capa­ motility is an appendage-independent form of move­ city to survive in healthcare environments, its potential ment characteristic of some A. baumannii clinical to acquire and develop antimicrobial resistance, and to isolates and probably driven by the extrusion of extra­ its virulence [2]. In response to the increasing clinical polymeric substances [8,9]. This type of motility is importance of multidrug-resistant bacteria, the World associated with the synthesis of 1,3-diaminopropane Health Organization created a priority list of antibiotic- [8], lipooligosaccharide production [9], proteins resistant bacterial species, including A. baumannii, to related to metabolism, the outer membrane, and nat­ encourage research into effective drugs and the devel­ ural competence [10], and proteins from at least five opment of novel therapeutic strategies[3]. of the six known superfamilies of efflux pumps [11]. While the name Acinetobacter comes from the Regulators of surface-associated motility include Greek word “akineto,” translated as “non-motile,” quorum sensing [12], cyclic diguanylate [13], blue- many strains of A. baumannii are in fact motile, light sensing [14], CheW and CheA chemotactic ana­ through two different mechanisms: twitching motility logs [15,16], and the EnvZ/OmpR two-component and surface-associated motility. Twitching motility is system [17]. The capacity to move by twitching or a coordinated multicellular movement driven by the surface-associated motility depends on the CONTACT Jesús aranda jesus.aranda@uab.cat Supplemental data for this article can be accessed here. © 2021 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. 2202 J. CORRAL ET AL. A. baumannii strain. Thus, strain ATCC 17978 exhi­ revealed the involvement of several of these genes in bits only surface-associated motility, strain AYE only virulence as well. twitching, strain MAR002 is capable of both, and other strains, such as ATCC 19606, are non- motile [18,19]. Materials and methods ]In many different bacterial species, virulence is clo­ Bacterial strains, plasmids and growth conditions sely related to motility, as is the case for A. baumannii [2,20]. This association suggests that bacterial motility The bacterial strains and plasmids used in this study are can be exploited as a specific preventive or therapeutic listed in Table 1. A. baumannii strain MAR002 and antibacterial target, with the aim of either cure or dis­ E. coli strain DH5α were grown at 37°C in Luria- ease prevention [21]. Although the role of twitching Bertani (LB) medium[32] with shaking at 180 rpm. motility in virulence is so far unknown [2], a key role The pCR-BluntII-TOPO (Invitrogen, catalog number for surface-associated motility in the virulence of 451,245) and pBAV1Gm-T5-gfp plasmids were used A. baumannii has been demonstrated in several reports for mutant construction and complementation, respec­ [8,10,11,16,22,23]. tively (Table 1). When needed, rifampin (50 mg/L), Point mutations in the bacterial rpoB gene, encoding the kanamycin (50 mg/L), or gentamicin (20 mg/L for β-subunit of the RNA polymerase, confer rifampin resis­ E. coli and 40 mg/L for A. baumannii) was added to tance and cause substantial changes in the transcriptional the growth medium. A. baumannii growth was mon­ profile of bacterial cells, by affecting the enzyme’s affinity itored using bacterial cultures in LB broth inoculated for several promoters [24–26]. Rifampin inhibits DNA- with an overnight culture at a dilution of 1:100 and dependent RNA synthesis through its interaction with the then incubated at 37°C with shaking at 180 rpm. The RpoB subunit. Mutations in the rpoB gene that produce optical density at 600 nm (OD600) of the cultures was conformational changes in the region of rifampin interac­ measured hourly. To obtain rifampin-resistant (Rifr) tion reduce the affinity of RpoB for this antimicrobial and mutants, saturated cultures of wild-type (WT) thus confer resistance. Moreover, the new conformation of A. baumannii strain MAR002 were plated on rifampin- RpoB modifies the polymerase’s affinity for certain promo­ containing LB plates and incubated at 37°C for 24 h. ters, causing changes in gene expression[27]. In other bac­ terial species, such as Escherichia coli or Pseudomonas Motility assays aeruginosa, mutations in the rpoB gene induce pleiotropic effects that impact processes such as sporulation[28], strin­ The assays were conducted on fresh-agarose plates (0.5% gent response [25,29], motility, and virulence [24,30]. tryptone, 0.25% NaCl, and 0.3% low-agarose; Nzytech) In a previous study conducted in A. baumannii strain prepared on the day of the experiment and using bacterial ATCC 17978, which possesses only surface-associated cultures at the early stationary phase. A Kolle loop was motility, we showed that point mutations in rpoB impair used to inoculate the plates, followed by gently swirling both motility and virulence[23]. A transcriptional analysis the medium until the plastic surface of the plate was of these spontaneous rifampin-resistant mutants led to the covered. Bacterial growth occurred at the agarose-plastic identification of six genes encoding proteins homologous interface. The inoculated plates were incubated at either to the transporters and metabolic enzymes involved in the 30°C or 37°C for 16–24 h. Twitching movement was surface-associated motility of A. baumannii[23]. observed at the agarose-plastic interface as a faint bacterial In the present work, spontaneous rifampin-resistant growth halo, and surface-associated motility on the plate mutants of A. baumannii strain MAR002, a biofilm- surface as a densely branched bacterial halo with irregular hyperproducing strain isolated from a wound sample edges. All assays were carried out at least three times in collected from a patient in Spain[31] and possessing independent experiments, each set up in triplicate. both twitching and surface-associated motility, were Representative images were obtained using a ChemiDoc­ TM isolated. Three different motility-related phenotypes XRS+ system (Bio-Rad). were identified among the mutants: (i) twitching defi­ cient, (ii) surface-associated motility deficient, and (iii) Galleria mellonella killing assays lacking both movement phenotypes. A transcriptional analysis of these mutants allowed us to identify genes The virulence of the A. baumannii strains was determined involved in either or both forms of motility. using the G. mellonella (wax worm) model, as described Experiments with the respective knockout mutants previously [33]. Briefly, ten caterpillars were inoculated VIRULENCE 2203 Table 1. Bacterial strains and plasmids used in this work. Source or Strains or plasmids Relevant characteristics reference E. coli strain DH5α E. coli supE4 ΔlacU169 (ɸ80 ΔlacZ ΔM15) hsdR17, recA1, endA1, gyrA96, thi-1, relA1 Clontech A. baumannii strains [31] MAR002 Wild-type, biofilm-hyperproducing strain isolated from a wound sample T−S− Spontaneous rifampin-resistant MAR002 derivative strain with an impairment in twitching and surface- This work associated motility, Rifr T+S− Spontaneous rifampin-resistant MAR002 derivative strain with an impairment in twitching motility, Rifr This work T−S+ Spontaneous rifampin-resistant MAR002 derivative strain with an impairment in surface-associated motility, This work Rifr LH92_RS00045 MAR002 derivative strain with LH92_RS00045::pCR-BluntII-TOPO disruption, Kmr, Zeor This work LH92_RS01675 MAR002 derivative strain with LH92_RS01675::pCR-BluntII-TOPO disruption, Kmr, Zeor This work LH92_RS02585 MAR002 derivative strain with LH92_RS02585::pCR-BluntII-TOPO disruption, Kmr, Zeor This work LH92_RS04440 MAR002 derivative strain with LH92_RS04440::pCR-BluntII-TOPO disruption, Kmr, Zeor This work LH92_RS04440 LH92_RS04440 strain carrying the pBAV1K-T5-gfp plasmid with the corresponding gene cloned, Kmr, Zeor, This work complemented Gmr LH92_RS05285 MAR002 derivative strain with LH92_RS05285::pCR-BluntII-TOPO disruption, Kmr, Zeor This work LH92_RS05285 LH92_RS05285 strain carrying the pBAV1K-T5-gfp plasmid with the corresponding gene cloned, Kmr, Zeor, This work complemented Gmr LH92_RS05405 MAR002 derivative strain with LH92_RS05405::pCR-BluntII-TOPO disruption, Kmr, Zeor This work LH92_RS05405 LH92_RS05405 strain carrying the pBAV1K-T5-gfp plasmid with the corresponding gene cloned, Kmr, Zeor, This work complemented Gmr LH92_RS06300 MAR002 derivative strain with LH92_RS06300::pCR-BluntII-TOPO disruption, Kmr, Zeor This work LH92_RS06300 LH92_RS06300 strain carrying the pBAV1K-T5-gfp plasmid with the corresponding gene cloned, Kmr, Zeor, This work complemented Gmr LH92_RS06715 MAR002 derivative strain with LH92_RS06715::pCR-BluntII-TOPO disruption, Kmr, Zeor This work LH92_RS06715 LH92_RS06715 strain carrying the pBAV1K-T5-gfp plasmid with the corresponding gene cloned, Kmr, Zeor, This work complemented Gmr LH92_RS07135 MAR002 derivative strain with LH92_RS07135::pCR-BluntII-TOPO disruption, Kmr, Zeor This work LH92_RS08005 MAR002 derivative strain with LH92_RS08005::pCR-BluntII-TOPO disruption, Kmr, Zeor This work LH92_RS12010 MAR002 derivative strain with LH92_RS12010::pCR-BluntII-TOPO disruption, Kmr, Zeor This work LH92_RS13540 MAR002 derivative strain with LH92_RS13540::pCR-BluntII-TOPO disruption, Kmr, Zeor This work LH92_RS13540 LH92_RS13540 strain carrying the pBAV1K-T5-gfp plasmid with the corresponding gene cloned, Kmr, Zeor, This work complemented Gmr LH92_RS13935 MAR002 derivative strain with LH92_RS13935:: pCR-BluntII-TOPO disruption, Kmr, Zeor This work LH92_RS13935 LH92_RS13935 strain carrying the pBAV1K-T5-gfp plasmid with the corresponding gene cloned, Kmr, Zeor, This work complemented Gmr LH92_RS15870 MAR002 derivative strain with LH92_RS15870:: pCR-BluntII-TOPO disruption, Kmr, Zeor This work LH92_RS15870 LH92_RS15870 strain carrying the pBAV1K-T5-gfp plasmid with the corresponding gene cloned, Kmr, Zeor, This work complemented Gmr LH92_RS16065 MAR002 derivative strain with LH92_RS16065::pCR-BluntII-TOPO disruption, Kmr, Zeor This work LH92_RS16065 LH92_RS16065 strain carrying the pBAV1K-T5-gfp plasmid with the corresponding gene cloned, Kmr, Zeor, This work complemented Gmr LH92_RS16130 MAR002 derivative strain with LH92_RS16130::pCR-BluntII-TOPO disruption, Kmr, Zeor This work LH92_RS16130 LH92_RS16130 strain carrying the pBAV1K-T5-gfp plasmid with the corresponding gene cloned, Kmr, Zeor, This work complemented Gmr LH92_RS16880 MAR002 derivative strain with LH92_RS16880::pCR-BluntII-TOPO disruption, Kmr, Zeor This work LH92_RS17175 MAR002 derivative strain with LH92_RS17175::pCR-BluntII-TOPO disruption, Kmr, Zeor This work LH92_RS17175 LH92_RS17175 strain carrying the pBAV1K-T5-gfp plasmid with the corresponding gene cloned, Kmr, Zeor, This work complemented Gmr Plasmids pCR-BluntII-TOPO Cloning vector, Kmr, Zeor Invitrogen pBAV1K-T5-gfp Complementation vector, Kmr Addgene pVRL1 Plasmid carrying a gentamicin cassette, Gmr [38] pBAV1Gm-T5-gfp pBAV1K-T5-gfp derivative vector carrying gentamicin cassette, Gmr This work Rifr, Kmr, Zeor, Gmr stand for resistance to rifampin, kanamycin, zeocin and gentamicin, respectively. via the hemocoel with 10 μL of the corresponding h. All G. mellonella killing experiments were performed at A. baumannii strain (~ 105 CFU). The inoculants were least three times. prepared from exponentially growing cultures (OD600 = 0.1) corresponding to ~ 108 CFU/mL and previously diluted in phosphate-buffered saline (PBS). The concen­ RNA extraction tration of each inoculum was confirmed by colony counts on LB plates. As a negative control, the same number of A High Pure RNA isolation kit (Roche) was used to caterpillars was inoculated with 10 μL of PBS. The control isolate RNA from cultures of A. baumannii strain and treated caterpillars were incubated at 37°C in the dark MAR002 and the corresponding Rifr mutants, accord­ and their survival was checked every 12 h for a total of 96 ing to the manufacturer’s instructions. Prior to RNA 2204 J. CORRAL ET AL. extractions, bacteria growing in LB medium until the described[36]. RT-qPCR was conducted using mid-exponential growth phase (OD600 = 0.4) were pel­ Lightcycler RNA Master SYBR green I (Roche) and leted and then treated with lysozyme (50 mg/mL; resus­ a Lightcycler 480 instrument (LC480; Roche) according pended in Tris-EDTA (TE) buffer) for 10 min at 37°C. to the manufacturer’s instructions. The oligonucleo­ To remove DNA contaminants, all RNA samples were tides listed in Table S1 were used to validate the expres­ incubated with DNase Turbo (Ambion). DNA removal sion of the selected genes identified by RNA-seq. The was confirmed in PCRs using RNA samples. relative mRNA concentrations of the genes of interest were determined according to a standard curve gener­ ated by amplifying an internal fragment of the gyrB RNA-seq gene that was not induced under the tested growth Transcriptomic analysis of A. baumannii strain MAR002 conditions. The expression factor was calculated as WT and the Rifr mutants was performed by Allgenetics (A the ratio of the mRNA concentration of the target Coruña, Spain). Briefly, each strain was grown in LB until gene as expressed by the WT vs. by the studied mutant an OD600 of 0.4–0.6, pelleted, and then stored at −80°C strain. until used in RNA-seq. RNA was extracted as described above and its purity was analyzed using a 2100 Bioanalyzer A. baumannii knockout construction and (Agilent). To prepare each library, the ribosomal RNA was complementation removed using the Ribo-Zero Plus rRNA depletion kit (Illumina). The RNA was then processed using the Gene inactivation was carried out by gene disruption, TruSeq Stranded mRNA library prep kit (Illumina), with as previously described[37]. Briefly, an internal frag­ the fragment size distribution checked using the High ment (from ~ 0.2 to 0.6 kb) of the target gene from Sensitivity DNA kit (Agilent). The libraries were quantified A. baumannii strain MAR002 WT was PCR-amplified using the Qubit dsDNA HS assay kit (Thermo Fisher using the corresponding intF and intR primers listed in Scientific) and pooled in equimolar amounts according to Table S1. The resulting fragment was cloned into the the Qubit results. The resulting pool was sequenced in kanamycin- and zeocin-resistant plasmid pCR-BluntII- a HiSeq 4000 PE100 lane (Illumina). TOPO (Invitrogen), which is unable to replicate in After RNA-seq, the quality of the raw sequencing A. baumannii. The construct was introduced into data was checked, and the trimmed and preprocessed E. coli DH5α by electroporation and selected on kana­ reads were mapped to the A. baumannii MAR002 mycin-containing LB plates. The purified plasmid was reference genome (NZ_JRHB01000001.1 and then introduced into WT A. baumannii by electropora­ NZ_JRHB01000002.1) using STAR 2.6.0a[34]. Read tion and the resulting transformants were selected on counts were quantified, normalized, and filtered follow­ kanamycin-containing plates. Recombinant clones were ing the trimmed mean of M-values method[35]. confirmed by sequencing (Macrogen) of the PCR pro­ Differential gene expression between A. baumannii ducts amplified using the appropriated combination of MAR002 WT and the corresponding Rifr mutants was primers (Table S1), one external to the construction analyzed using the Bioconductor packages NOISeq and (ComFw or ComRv) and the other within the suicide NOISeqBIO. The expression of a gene was considered plasmid (M13FpUC or M13RpUC), according to the to be induced or repressed when the fold change of the insert orientation relative to the vector. The stability of Rifr mutants compared to the parental strain was ≥ 2 or the A. baumannii knockouts was determined in ten ≤ −2, respectively, and the P value of the difference passages without selective pressure, with colony resus­ compared to the control was < 0.05. A Venn diagram pension carried out every 24 h, followed by colony was constructed using the BxToolBox program avail­ counting on LB plates with and without kanamycin at able at bioinforx.com. the last passage. RNAseq data are available at NCBI (National Center For mutant complementation, the target ORF was for Biotechnology Information; https://www.ncbi.nlm. cloned into the pBAV1Gm-T5-gfp plasmid using the nih.gov/) under the accession number PRJNA479736, appropriate primers (Table S1), which included where Rif7, Rif11 and Rif12 correspond to T−S−, T+S− XbaI restriction sequences. The pBAV1Gm-T5-gfp and T−S+ strains, respectively. derivative plasmid was generated from the pBAV1K- T5-gfp vector (a gift from Ichiro Matsumura; Addgene plasmid 26,702), in which the kanamycin Gene expression sequence is substituted by the gentamicin cassette To corroborate the transcriptomic data, gene expres­ from the pVRL1 plasmid[38]. The recombinant plas­ sion was determined by RT-qPCR as previously mid was introduced by electroporation, firstly into VIRULENCE 2205 E. coli strain DH5α and then, once the correct con­ differences in the motility of A. baumannii MAR002 struct had been verified by PCR and sequencing at 30°C vs. 37°C were observed (data not shown). (Macrogen), into the corresponding A. baumannii To evaluate the role of each form of motility in the knockout. The complemented mutants were selected pathogenesis of A. baumannii MAR002, the virulence on kanamycin- and gentamicin-containing plates. To of each of the three types of Rifr mutant with altered assess plasmid stability in the complemented mutants, motility (T−S−, T−S+ and T+S−) was determined in the cells were plated on solid medium with or without G. mellonella killing assays (Figure 1b). Compared to gentamicin. the WT strain, the virulence of the strains with impaired surface-associated motility (T−S− and T+S−) was significantly (P < 0.05) reduced whereas this was Statistical analysis not the case for the Rifr mutant with impaired twitch­ The data were analyzed in a two-tailed, one-way analysis ing motility alone (T−S+). After 96 h of incubation, the of variance (ANOVA) followed by the Tukey test for survival of worms inoculated with mutants lacking sur­ post-hoc multiple-group comparisons. Survival curves face-associated motility was 70% (T−S−) and 50% obtained in the G. mellonella killing assays were plotted (T+S−) while that in the worms inoculated with either using the Kaplan-Meier method and differences in sur­ the T−S+ mutant or the WT strain was 30% and 10%, vival were calculated using the log-rank test. In all cases, respectively (Figure 1b). statistical significance was defined as a P value < 0.05. Identification of A. baumannii strain MAR002 Rifr genes involved in motility Results Global gene expression of the Rifr mutants with altered Twitching and surface-associated motility profiles motility (T−S−, T−S+, and T+S−) was characterized by of Rifr mutants of A. baumannii strain MAR002 numerous significant changes compared to that of WT and the effects on virulence A. baumannii MAR002. Based on two-fold differences As noted above, mutations in the rpoB gene of in the up- or down-regulation of gene expression, 771 A. baumannii ATCC 17,978 change the promoter affi­ genes were differentially expressed in the T−S− strain, of nity of the bacterium’s RNA polymerase, which in turn which 282 were up-regulated and 489 down-regulated affects several phenotypic properties including motility (Table S2). Furthermore, 970 genes were differentially [23]. To analyze whether mutations in the rpoB gene of expressed in the T−S+ strain, 489 of which were up- A. baumannii MAR002, a strain able to move through regulated and 481 down-regulated. Finally, 700 genes twitching and surface-associated motility, affect both were differentially expressed in the T+S− strain, of motility systems, saturated cultures of this strain were which 285 were up-regulated and 415 down-regulated plated in the presence of rifampin. Several spontaneous (Table S2). These results were validated in an RT-qPCR A. baumannii MAR002 Rifr mutants were isolated and analysis comparing selected down-regulated genes from their motility patterns were evaluated. each of the three Rifr mutants with the respective WT Four different movement phenotypes were identified genes from the parental strain (Figure S1). in the mutants: (i) 2.3% impaired in both movement To obtain insights into the involvement of the types (T−S−), (ii) 4.5% impaired in twitching motility deregulated genes in twitching and surface- (T−S+), (iii) 25% impaired in surface-associated motility associated motility, they were grouped according to (T+S−), and (iv) 68.2% without a difference compared their presence or absence in the Rifr mutants as to the WT. A representative from each phenotype lack­ determined from the RNA-seq analyses (Figure 2 ing one or both types of motility was selected (T−S−: and Table S2). The identification of 115 genes shared Arg538His; T−S+: Asp525Val; and T+S−: Ser540Phe) for by strains with impaired twitching motility (T−S+ and further analysis and its rpoB gene was sequenced T−S−; T group) and 202 genes by strains with (Figure 1a). To confirm that the motility impairments impaired surface-associated motility (T+S− and T−S−; were due to point mutations located in the rpoB gene, S group) implied that these groups of genes, along further spontaneous Rifr mutants were isolated in with those exclusive to strain T−S− (227 genes), were which the specific point mutations in the rpoB involved in twitching motility (T group), in surface- sequence were located in the same positions as obtained associated motility (S group), or in both (TS group) previously. These mutants showed that the same muta­ (Table S2). tion in different isolates always implied the same moti­ Eighteen genes (6 per group) from the three groups lity phenotype (data not shown). No significant (T, S, and TS) were selected to study their possible 2206 J. CORRAL ET AL. Figure 1. (a) Diameter (in mm) of the halos formed by the indicated strains during both surface-associated motility and twitching. Each experiment was carried out in triplicate in a minimum of three independent motility assays. A significant decrease in motility (P < 0.05) compared to the parental strain is indicated with an asterisk (*). (b) Representative results of a Galleria mellonella killing assay using the specified strains. Larvae (n = 10 per group) were inoculated with ~ 105 CFU of the indicated A. baumannii strain or PBS as a negative control. A significant decrease in virulence (P < 0.05) compared to the parental strain is indicated with an asterisk (*). Figure 2. Venn diagram of the distribution of deregulated genes in A. baumannii strains T−S−, T+S−, and T−S+. Overlapping areas represent deregulated genes identified in two or all three strains, and the numbers in bold the group of genes hypothetically involved in surface-associated motility, twitching motility, or both in A. baumannii strain MAR002. In the indicated groups, the total number of up-regulated (green) or down-regulated (red) genes is also shown. direct roles in twitching or surface associated motility Construction and motility characterization of in the WT strain of A. baumannii. The selection was A. baumannii knockout mutants of the selected based on the expression levels of these genes, their genes predicted function with respect to either twitching or To assess whether the selected genes (Table 2) played surface-associated motility, and their functional group, a role in motility, knockout mutants constructed by with the latter including cell wall/membrane/envelope gene disruption (Figure S2) were inoculated onto fresh- biogenesis, motility, transport, metabolism, and tran­ agarose motility plates. All mutants had a stability of > scription (Table 2). 90%. The results showed that knockout of the genes VIRULENCE 2207 Table 2. Downregulated genes analyzed in this study. Log2 Fold change − − Gene ID. Protein Description T S T+S− T−S+ TS group LH92_RS01675 WP_001072475.1 HlyD family secretion protein −4.50 LH92_RS02585 WP_000537695.1 Type IV pilus biogenesis/stability protein PilW −1.03 LH92_RS07135 WP_000229914.1 Malonate transporter subunit MadM −2.14 LH92_RS15870 WP_001046417.1 Prepilin-type N-terminal cleavage/methylation domain-containing protein PilE −1.68 LH92_RS16130 WP_021510165.1 Efflux RND transporter periplasmic adaptor −3.90 LH92_RS16880 WP_045900525.1 Type VI secretion system protein VgrG −2.00 S group LH92_RS00045 WP_047479431.1 ABC transporter ATP-binding protein −2.64 −2.13 LH92_RS05285 WP_079267361.1 Amino acid permease −4.66 −2.55 LH92_RS08005 WP_001162376.1 Phosphoglycerate mutase family protein −5.93 −3.99 LH92_RS12010 WP_004840631.1 Multidrug efflux RND transporter outer membrane subunit AdeH −2.69 −1.10 LH92_RS16065 WP_000521916.1 Sulfate ABC transporter substrate-binding protein −3.45 −1.30 LH92_RS17175 WP_161782838.1 PQQ-dependent sugar dehydrogenase −4.10 −1.72 T group LH92_RS04440 WP_000837758.1 Energy transducer TonB −1.81 −6.48 LH92_RS05405 WP_125535337.1 Outer membrane porin family OprD −1.28 −1.22 LH92_RS06300 WP_045899602.1 Type VI secretion system-associated protein TagF −1.07 −1.17 LH92_RS06715 WP_000996083.1 IclR family transcriptional regulator −1.38 −6.70 LH92_RS13540 WP_000669945.1 Hypothetical protein (putative lyase) −1.30 −7.77 LH92_RS13935 WP_001174793.1 Multidrug efflux RND transporter outer membrane subunit AdeK −2.01 −1.73 Figure 3. Diameter (in mm) of the halos formed by the indicated strains during surface-associated motility and twitching. Each experiment was carried out in triplicate in a minimum of three independent motility assays. A significant decrease in motility (P < 0.05) compared to the parental strain is indicated with an asterisk (*). encoding the RND component AdeK and the minor transporter component and a PQQ-dependent sugar prepilin PilE, belonging to the T and TS group, respec­ dehydrogenase; and (iii) the TS group, encoding an tively, produced a loss of twitching motility but had no RND transporter component (Figure 3). effect on surface-associated motility (Figure 3). The remaining knockout mutants (Table 2) showed Conversely, inactivation of the predicted amino acid no alterations in their motility and their behavior was permease encoded by the LH92_RS05285 gene, belong­ similar to that of the WT strain (Figure 3). Since the ing to the S group, abolished surface-associated motility growth curves of all mutants grown in LB broth were without altering twitching motility (Figure 3). comparable to that of the parental strain (data not A complete loss of both twitching and surface- shown), the observed impairments in motility could associated motility characterized the knockouts with not be attributed to reductions in the bacterial growth disruptions in genes belonging to: (i) the T group, rate. including those encoding the energy transducer TonB To demonstrate a direct relation between specific (which interacts with outer membrane receptor pro­ gene disruptions and subsequent movement patterns, teins), the porin OprD (an outer membrane protein), the pBAV1Gm-T5-gfp plasmid, carrying the corre­ the type VI secretion system component TagF, sponding WT gene, was introduced into all mutants a transcriptional regulator, and a putative lyase; (ii) with impaired motility. As expected, complementation the S group, including those encoding an ABC consistently restored the parental phenotype (data not 2208 J. CORRAL ET AL. shown). Plasmid stability was confirmed in all cases in the WT strain, indicating a restoration of virulence through the methodology specified above. (Figure 4). Impaired surface-associated motility but not Discussion twitching motility significantly reduces The motility of A. baumannii involves two different A. baumannii virulence motility systems: twitching and surface-associated The above-described knockout mutants with altered motility. Depending on the strain, one, both, or neither motility were then studied for their pathogenicity, of these movement types is present[18,19]. determined in G. mellonella killing assays. Mutants A. baumannii MAR002, a virulent clinical strain, is lacking genes involved in surface-associated motility able to move via both twitching and surface- showed a significant (P < 0.05) reduction in their viru­ associated motility [31,39]. In previous work, we lence, independent of their twitching behavior showed that spontaneous rifampin-resistant mutants (Figure 4). Specifically, > 75% of the worms inoculated of A. baumannii strain ATCC 17,978 carried point with any of the knockout mutants with impairments in mutations in the β-subunit of the RNA polymerase both twitching and surface-associated motility that altered the transcriptional pattern of genes (Figure 4) and ~ 60% of those inoculated with the involved in surface-associated motility[23]. In this LH92_RS05285 mutant, with impaired surface- study, the isolation of spontaneous Rifr mutants in associated motility only, survived until the end of the A. baumannii strain MAR002 included strains with 96-h experiment (Figure 4). By contrast, virulence was phenotypic changes in surface-associated and/or not significantly affected in knockout mutants with twitching motility (Figure 1a). impaired twitching but not surface-associated motility A connection between motility and virulence has (Figure 4). Likewise, knockout mutants with a motility been demonstrated in several pathogenic bacterial pattern similar to that of the A. baumannii MAR002 species, among them flagellated bacteria [20,21], bac­ WT strain did not show a significant reduction in their teria displaying type IV pili[40], and bacteria with virulence, as < 25% of the inoculated worms survived. appendage-independent motility [8,41,42]. In the Finally, when the worms were inoculated with the A. baumannii MAR002 Rifr mutants lacking twitch­ corresponding complemented strains, survival of the ing motility, virulence was slightly but not signifi­ larvae was reduced in all cases to the level determined cantly reduced compared to mutants deficient in Figure 4. Representative results of a G. mellonella killing assay of the specified strains. Larvae (n = 10 per group) were inoculated with ~ 105 CFU of the indicated A. baumannii strain. All assays were carried out at least three times. For clarity, the results of a representative assay are presented separately and compared with those obtained with the wild-type (WT) strain (dotted line). In strains marked with an asterisk (*), virulence was significantly reduced (P < 0.05) compared with the parental strain (MAR002). In those cases, the survival of worms inoculated with the complemented mutants is also shown (discontinued line). The motility profile of the inoculated strain is indicated by the colored background of each graph. VIRULENCE 2209 surface-associated motility (Figure 1b). In the A. baumannii MAR002 Rifr mutants, all of the Transcriptomic studies have shown an increase in down-regulated genes identified in this work encode the expression of genes involved in the synthesis of putative proteins related to cell envelope biogenesis, type IV pili during the growth of A. baumannii in transporters, and metabolic enzymes, i.e., functions human serum, thus demonstrating the importance of that play important roles in motility and/or virulence. these appendages during bacteremia [2,7,43]. From these down-regulated genes, 18 A. baumannii A transcriptional analysis of the Rifr mutants with knockout mutants were constructed and their motility the three types of altered motility (Table S2) revealed was then analyzed. In two mutants, in which either the three groups of genes putatively involved in either one gene encoding the prepilin PilE or the RND transporter or the other or both motility systems (Figure 2). In component AdeK was inactivated, twitching was P. aeruginosa, a phylogenetic neighbor of impaired whereas in another mutant, encoding an A. baumannii, transcriptomic studies of motile versus amino acid permease, surface-associated motility was non-motile cells revealed an overexpression of genes impaired. Furthermore, the inactivation of eight genes, involved in metabolism and secretion during the devel­ including those encoding RND and ABC components, opment of motility by this species [44,45]. Given the the energy transducer TonB, the porin OprD, the T6SS pleiotropic effects produced by rpoB gene mutations component TagF, an IclR transcriptional regulator, and the complex regulation of motility, the wide a PQQ-dependent sugar dehydrogenase, and range of deregulated genes identified in the transcrip­ a putative pectate lyase, revealed their involvement in tomic analysis was not surprising. The identified genes both twitching and surface-associated motility. included those involved in the above-mentioned pro­ PilE and AdeK were the only proteins with cesses [23,29,46] as well as genes that co-regulate moti­ a demonstrated role in twitching but not in the surface- lity and other biological processes under the control of associated motility of A. baumannii strain MAR002. key bacterial regulators, such as cyclic diguanylate PilE is a prepilin subunit involved in the biogenesis of [13,47], chemotaxis [15,16,48], and two-component type IV pili. These long protein polymers participate in systems [17,49] (Tables S2 and S3). twitching motility, attachment, biofilm formation, and The roles in twitching and/or surface-associated- DNA uptake [50–52]. In P. aeruginosa, PilE and other motility of the down-regulated genes selected from minor pseudopilins comprise the minimal set of com­ the transcriptional analysis of the rpoB mutants were ponents required for pilus assembly[52], which would directly investigated through the construction of also explain why a prepilin subunit is required for the knockouts. Based on the transcriptional data, 18 down- twitching motility of A. baumannii. However, the find­ regulated genes belonging to the three gene groups ing of an involvement of AdeK, the outer membrane were inactivated (Table 2). In 11 of the derivative component of an RND transporter, in twitching was knockout mutants, twitching and/or surface-associated unexpected. The adeI, adeJ, and adeK genes encode, motility were reduced (Figure 3). However, although respectively, the membrane fusion, RND, and outer the deregulated genes could be classified according to membrane components of the efflux pump AdeIJK, their putative involvement in twitching motility (T which participates in the extrusion of several antimi­ group), surface-associated motility (S group), or both crobials in A. baumannii[53]. The absence of AdeK was (TS group), not all knockouts exhibited the expected presumably unrelated to the altered expression of pili phenotype. These results can be explained by the wide whereas the absence of the cytoplasmic transporter range of deregulated genes that lead to a disturbed AdeJ was previously implicated in the impaired surface- motility [25,46], as shown by the transcriptomic data associated motility of A. baumannii ATCC 17,978 and (Table S2), whereas in the knockout mutants only one A. nosocomialis, probably through a disruption in the of these genes was inactivated. secretion of the surfactants required for this type of In the virulence assay, the pathogenicity of the motility [11,54]. Interestingly, it has been reported strain-MAR002-derived knockout mutants was similar that the absence of efflux pumps such as MuxABC- to that of the Rifr mutants and could be linked to the OpmB affects twitching in P. aeruginosa [55,56]. respective motility pattern. Specifically, the pathogeni­ Thus, the extrusion of some molecules may be required city of the mutants with a phenotype similar to that of not only for surface-associated motility but also for the T−S− or T+S− strains but not the T−S+ strain was twitching. For instance, Myxococcus xanthus requires significantly attenuated (Figures 3 and 4), thus demon­ extracellular polysaccharides for attachment and for the strating the greater participation of surface-associated retraction of type IV pili[57]. motility over twitching in the virulence of The amino acid permease encoded by the A. baumannii. LH92_RS05285 gene was the only component detected 2210 J. CORRAL ET AL. in this work that was implicated solely in the surface- siderophores such as acinetobactin[67]. Iron limitation associated motility of A. baumannii. Other genes, in A. baumannii grown under iron-limited conditions encoding transporter-related components, could be results in impaired motility[67], which suggests that the attributed to both twitching and surface-associated lack of siderophores impacts the motility of this bacter­ motility: LH92_RS16065, LH92_RS16130, ium. Finally, inactivation of LH92_RS06300, encoding LH92_RS05405, and LH92_RS04440, encoding an the type VI secretion system-associated protein TagF, ABC transporter component, an RND component, the also impairs the motility of A. baumannii MAR002. outer membrane protein OprD, and TonB, an energy A gene cluster predicted to encode a type VI secretion transducer of transporters, respectively. Protein trans­ system is upregulated in motile populations of porters may be important for motility, through their A. baumannii[22]. Taken together, these findings sug­ extrusion of molecules such as surfactants; and patho­ gest that type VI secretion systems significantly contri­ genesis, by the secretion of virulence factors. In fact, bute to the virulence and motility of A. baumannii, as efflux pumps from at least five of the six different also recently demonstrated in the gram-negative bac­ superfamilies described in A. baumannii were impli­ teria Xanthomonas phaseoli[68]. cated in the bacterium’s surface-associated motility, In summary, this work is the first to identify genes presumably related to their role in the secretion of involved in either twitching, surface-associated motility, extrapolymeric substances [8,9,11,54]. Similarly, in or both in an A. baumannii strain capable of both a recent work by Blaschke et al., outer membrane movement types. The identified genes are mainly proteins were shown to be involved in the surface- involved in cell envelope biogenesis, transport, and associated motility of A. baumannii ATCC 1797810. In metabolism, but they are also clearly associated with P. aeruginosa, a transposon insertion in the tonB gene motility and virulence. Specifically, genes encoding PilE results in a phenotype that includes defective twitching (LH92_RS15870) and the RND efflux pump compo­ motility and the reduced assembly of extracellular pili, nent AdeK (LH92_RS13935) were shown to be asso­ indicating the involvement of TonB in the transport ciated with twitching, and a gene encoding an amino and secretion of pili or in a component required for acid permease (LH92_RS05285) was linked to surface- their formation[58]. associated motility. Other proteins shown in this study to participate in In addition, genes encoding a putative pectate lyase both twitching and surface-associated motility were (LH92_RS13540), a PQQ-dependent sugar dehydro­ a PQQ-dependent sugar dehydrogenase and a putative genase (LH92_RS17175), an IclR transcriptional regu­ pectate lyase, encoded by the genes LH92_RS17175 and lator (LH92_RS06715), an efflux RND component LH92_RS13540, respectively. Metabolic enzymes pro­ (LH92_RS16130), the porin OprD (LH92_RS05405), mote different types of motility in A. baumannii[23] an energy transducer TonB (LH92_RS04440), an ABC and in other bacterial species, such as Pseudomonas spp component (LH92_RS16065), and the type VI secretion [59]., E. coli[60], and Xanthomonas oryzae[61]. PQQ- system component TagF (LH92_RS06300) were found dependent dehydrogenases play a role in the oxidation to be associated with both movement types. The tran­ of monosaccharides. Although their physiological func­ scriptomic profiles described herein constitute tion is uncertain, the results of several studies suggest a starting point for further investigations of the mole­ that they serve as energy-conserving systems [62–64] in cular basis of the pathogenic processes of this multi­ high-energy processes such as motility. By catalyzing drug-resistant bacterium and thus for the development the depolymerization of bacterial capsules, the putative of new therapeutic agents aimed at novel targets. pectate lyase could facilitate bacterial displacement. The pectate lyases encoded by bacteriophage are responsible for the depolymerization of polysaccharides to allow Acknowledgments their specific binding to Acinetobacter bacterial cap­ We thank Joan Ruiz (UAB) and Susana Escribano (UAB) for sules[65]. their excellent technical assistance, the Germán Bou lab Two other detected genes were shown to be involved (Complexo Hospitalario Universitario de A Coruña, Spain) in both twitching and surface-associated motility: for the A. baumannii MAR002 strain, and Paolo Visca LH92_RS06715 and LH92_RS06300. In A. baumannii, (University Roma Tre, Rome, Italy) for kindly providing LH92_RS06715 encodes an IclR family transcriptional plasmid pVRL1. This study was supported by a grant regulator involved in the degradation of p-hydroxy­ (BIO2016-77011-R) from the Ministerio de Economía y Competitividad. The funders had no role in the study benzoate in Acinetobacter spp[66].. In Acinetobacter design, data collection or data interpretation, nor the decision spp., 2,3-dihydroxybenzoate, an iron binding com­ to submit the work for publication. J.A. is a Serra Húnter pound, is also a precursor for more complex Fellow, Generalitat de Catalunya, Barcelona, Spain. VIRULENCE 2211 Disclosure statement [13] Ahmad I, Nygren E, Khalid F, et al. A cyclic-di-GMP signalling network regulates biofilm formation and No potential conflict of interest was reported by the authors. surface associated motility of Acinetobacter baumannii 17978. Sci Rep. 2020;10(1):1991. [14] Mussi MA, Gaddy JA, Cabruja M, et al. The opportu­ Funding nistic human pathogen Acinetobacter baumannii senses This work was supported by the Ministerio de Economía and responds to light. J Bacteriol. 2010;192 y Competitividad [BIO2016-77011-R]. (24):6336–6345. [15] Chen R, Lv R, Xiao L, et al. A1S_2811, a CheA/Y-like hybrid two-component regulator from Acinetobacter baumannii ATCC17978, is involved in surface motility References and biofilm formation in this bacterium. Microbiologyopen. 2017;6(5):e00510. . [1] Wong D, Tb N, Ra B, et al. Clinical and pathophysio­ [16] Corral J, Pérez-Varela M, Barbé J, et al. Direct interac­ logical overview of Acinetobacter infections: a century tion between RecA and a CheW-like protein is of challenges. Clin Microbiol Rev. 2017;30(1):409–447. required for surface-associated motility, chemotaxis [2] Harding CMM, Hennon SWW, Feldman MFF. and the full virulence of Acinetobacter baumannii Uncovering the mechanisms of Acinetobacter bauman­ strain ATCC 17978. Virulence. 2020;11(1):315–326. nii virulence. Nat. Rev. Microbiol. 2018; 16:91–102. [17] Tipton KA, Rather PN. An ompRenvZ two-component [3] Tacconelli E, Carrara E, Savoldi A, et al. Discovery, system ortholog regulates phase variation, osmotic toler­ research, and development of new antibiotics: the ance, motility, and virulence in Acinetobacter baumannii WHO priority list of antibiotic-resistant bacteria and strain AB5075. J Bacteriol. 2017;199(3):e00705–16. tuberculosis. Lancet Infect Dis. 2018;18(3):318–327. [18] Eijkelkamp BA, Stroeher UH, Hassan KA, et al. 4 Harding CM, Tracy EN, Carruthers MD, et al. Adherence and motility characteristics of clinical Acinetobacter baumannii strain M2 produces type IV Acinetobacter baumannii isolates. FEMS Microbiol Pili which play a role in natural transformation and Lett. 2011;323(1):44–51. twitching motility but not surface-associated motility. [19] Skerniškyte J, Krasauskas R, Péchoux C, et al. Surface- MBio. 2013;4(4):e00360–13. related features and virulence among Acinetobacter [5] Chlebek JL, Hughes HQ, Ratkiewicz AS, et al. PilT and baumannii clinical isolates belonging to international PilU are homohexameric ATPases that coordinate to clones I and II. Front Microbiol. 2019;9:3116. retract type IVa pili. PLoS Genet. 2019;15(10): [20] Josenhans C, Suerbaum S. The role of motility as e1008448. a virulence factor in bacteria. Int J Med Microbiol. [6] Dhabaan GN, Abubakar S, Cerqueira GM, et al. 2002;291(8):605–614. Imipenem treatment induces expression of important [21] Erhardt M. Strategies to block bacterial pathogenesis by genes and phenotypes in a resistant Acinetobacter bau­ interference with motility and chemotaxis. Curr Top mannii isolate. Antimicrob Agents Chemother. 2016;60 Microbiol Immunol. 2016;398:185–205. (3):1370–1376. [22] Eijkelkamp BA, Stroeher UH, Hassan KA, et al. H-NS [7] Cerqueira GM, Kostoulias X, Khoo C, et al. A global plays a role in expression of Acinetobacter baumannii virulence regulator in Acinetobacter baumannii and its virulence features. Infect Immun. 2013;81 control of the phenylacetic acid catabolic pathway. (7):2574–2583. J Infect Dis. 2014;210(1):46–55. [23] Pérez-Varela M, Corral J, Vallejo JA, et al. Mutations [8] Skiebe E, De Berardinis V, Morczinek P, et al. Surface- in the β-subunit of the RNA polymerase impair the associated motility, a common trait of clinical isolates surface-associated motility and virulence of of Acinetobacter baumannii, depends on Acinetobacter baumannii. Infect Immun. 2017;85(8): 1,3-diaminopropane. Int J Med Microbiol. 2012;302 e00327–17. (3):117–128. . [24] Alifano P, Palumbo C, Pasanisi D, et al. rpoB poly­ [9] McQueary CN, Kirkup BC, Si Y, et al. Extracellular morphism and RNA polymerase genetic engineering. stress and lipopolysaccharide modulate Acinetobacter J Biotechnol. 2015;202:60–77. baumannii surface-associated motility. J Microbiol. [25] Qi Q, Preston GM, Maclean RC. Linking system-wide 2012;50(3):434–443. impacts of RNA polymerase mutations to the fitness [10] Blaschke U, Skiebe E, Wilharm G. Novel genes cost of rifampin resistance in Pseudomonas aeruginosa. required for surface-associated motility in MBio. 2014;5(6):e01562. Acinetobacter baumannii. Curr Microbiol. 2021;78 [26] Jin DJ, Gross CA. Mapping and sequencing of muta­ (4):1509–1528. tions in the Escherichia coli rpoB gene that lead to [11] Pérez-Varela M, Corral J, Aranda J, et al. Roles of rifampicin resistance. J Mol Biol. 1988;202(1):45–58. efflux pumps from different superfamilies in the [27] Campbell EA, Korzheva N, Mustaev A, et al. Structural surface-associated motility and virulence of mechanism for rifampicin inhibition of bacterial RNA Acinetobacter baumannii ATCC 17978. Antimicrob polymerase. Cell. 2001;104(6):901–912. Agents Chemother. 2019;63(3):e02190–18. [28] Cai XC, Xi H, Liang L, et al. Rifampicin-resistance [12] Clemmer KM, Bonomo RA, PN R. Genetic analysis of mutations in the rpoB gene in Bacillus velezensis surface motility in Acinetobacter baumannii. CC09 have pleiotropic effects. Front Microbiol. Microbiology. 2011;157:2534–2544. 2017;8:178. 2212 J. CORRAL ET AL. [29] Chatnaparat T, Li Z, Korban SS, et al. The stringent [45] Tremblay J, Déziel E. Gene expression in Pseudomonas response mediated by (p)ppGpp is required for virulence aeruginosa swarming motility. BMC Genomics. of Pseudomonas syringae pv. tomato and its survival on 2010;11(1):587. tomato. Mol Plant Microbe Interact. 2015;28(7):776–789. [46] Karthik M, Meenakshi S, Munavar MH. Unveiling the [30] Gliniewicz K, Wildung M, Orfe LH, et al. Potential molecular basis for pleiotropy in selected rif mutants of mechanisms of attenuation for rifampicin-passaged Escherichia coli: possible role for tyrosine in the Rif strains of Flavobacterium psychrophilum. BMC binding pocket and fast movement of RNA Microbiol. 2015;15(1):179. polymerase. Gene. 2019;713:143951. [31] Álvarez-Fraga L, López M, Merino M, et al. Draft [47] Jenal U, Reinders A, Lori C. Cyclic di-GMP: second genome sequence of the biofilm-hyperproducing messenger extraordinaire. Nat Rev Microbiol. 2017;15 Acinetobacter baumannii clinical strain MAR002. (5):271–284. Genome Announc. 2015;3(4):e00824–15. [48] Waite AJ, Frankel NW, Emonet T. Behavioral variabil­ [32] Miller JH Experiments in molecular genetics. Cold ity and phenotypic diversity in bacterial chemotaxis. Spring Harbor Laboratory; 1972. Annu Rev Biophys. 2018;47(1):595–616. [33] Peleg AY, Jara S, Monga D, et al. Galleria mellonella as [49] Prüß BM, Margolin W. Involvement of a Model System To Study Acinetobacter baumannii two-component signaling on bacterial motility and Pathogenesis and Therapeutics. Antimicrob Agents biofilm development. J Bacteriol. 2017;199(18): Chemother. 2009;53(6):2605–2609. e00259–17. [34] Dobin A, Davis CA, Schlesinger F, et al. STAR: ultra­ [50] Lobanovska M, Tang CM, Exley RM. Contribution of fast universal RNA-seq aligner. Bioinformatics. 2013;29 σ70 and σN factors to expression of class II PilE in (1):15–21. Neisseria meningitidis. J Bacteriol. 2019;201(20): [35] Robinson MD, Oshlack A. A scaling normalization e00170–19. method for differential expression analysis of [51] Mattick JS. Type IV pili and twitching motility. Annu RNA-seq data. Genome Biol. 2010;11(3):R25. Rev Microbiol. 2002;56(1):289–314. [36] Aranda J, Garrido ME, Cortés P, et al. Analysis of the [52] Nguyen Y, Sugiman-Marangos S, Harvey H, et al. Protective Capacity of Three Streptococcus suis Proteins Pseudomonas aeruginosa minor pilins prime type IVa Induced under Divalent-Cation-Limited Conditions. pilus assembly and promote surface display of the Infect Immun. 2008;76(4):1590–1598. PilY1 adhesin. J Biol Chem. 2015;290(1):601–611. [37] Aranda J, Poza M, Pardo BG, et al. A rapid and simple [53] Damier-Piolle L, Magnet S, Brémont S, et al. AdeIJK, a method for constructing stable mutants of Acinetobacter resistance-nodulation-cell division pump effluxing baumannii. BMC Microbiol. 2010;10(1):279. multiple antibiotics in Acinetobacter baumannii. [38] Lucidi M, Runci F, Rampioni G, et al. New shuttle Antimicrob Agents Chemother. 2008;52(2):557–562. vectors for gene cloning and expression in [54] Knight DB, Rudin SD, Bonomo RA, et al. Acinetobacter multidrug-resistant Acinetobacter species. Antimicrob nosocomialis: defining the role of efflux pumps in Agents Chemother. 2018;62(4):e02480–17. resistance to antimicrobial therapy, surface motility, [39] Álvarez-Fraga L, Pérez A, Rumbo-Feal S, et al. Analysis and biofilm formation. Front Microbiol. 2018;9:1902. of the role of the LH92_11085 gene of a biofilm [55] Rampioni G, Pillai CR, Longo F, et al. Effect of efflux hyper-producing Acinetobacter baumannii strain on pump inhibition on Pseudomonas aeruginosa tran­ biofilm formation and attachment to eukaryotic cells. scriptome and virulence. Sci Rep. 2017;7(1):11392. Virulence. 2016;7(4):443–455. . [56] Yang L, Chen L, Shen L, et al. Inactivation of [40] Craig L, Pique ME, Tainer JA. Type IV pilus structure MuxABC-OpmB transporter system in Pseudomonas and bacterial pathogenicity. Nat Rev Microbiol. 2004;2 aeruginosa leads to increased ampicillin and carbeni­ (5):363–378. cillin resistance and decreased virulence. J Microbiol. [41] Veith PD, Glew MD, Gorasia DG, et al. Type IX 2011;49(1):107–114. secretion: the generation of bacterial cell surface coat­ [57] Zusman DR, Scott AE, Yang Z, et al. Chemosensory ings involved in virulence, gliding motility and the pathways, motility and development in Myxococcus degradation of complex biopolymers. Mol Microbiol. xanthus. Nature Reviews Microbiology. 2007;5 2017;106(1):35–53. (11):862–872. [42] Pérez-Pascual D, Rochat T, Kerouault B, et al. More [58] Dong Y, Geng J, Liu J, et al. Roles of three TonB than gliding: involvement of GldD and GldG in the systems in the iron utilization and virulence of the virulence of Flavobacterium psychrophilum. Front Aeromonas hydrophila Chinese epidemic strain NJ-35. Microbiol. 2017;8:2168. Appl Microbiol Biotechnol. 2019;3(10):4203–4215. [43] Jacobs AC, Sayood K, Olmsted SB, et al. [59] Kato T, Morohoshi T, Someya N, et al. The ppuI-rsaL- Characterization of the Acinetobacter baumannii ppuR quorum-sensing system regulates cellular moti­ growth phase-dependent and serum responsive lity, pectate lyase activity, and virulence in potato transcriptomes. FEMS Immunol Med Microbiol. opportunistic pathogen Pseudomonas sp. StFLB209. 2012;64(3):403–412. Biosci Biotechnol Biochem. 2015;79(4):673–680. [44] Overhage J, Bains M, Brazas MD, et al. Swarming of [60] Winther T, Xu L, Berg-Sørensen K, et al. Effect of Pseudomonas aeruginosa is a complex adaptation lead­ energy metabolism on protein motility in the bacterial ing to increased production of virulence factors and outer membrane. Biophys J. 2009;97(5):1305–1312. antibiotic resistance. J Bacteriol. 2008;190 [61] Guo W, Zou LF, Cai LL, et al. Glucose-6-phosphate (8):2671–2679. dehydrogenase is required for extracellular VIRULENCE 2213 polysaccharide production, cell motility and the full [65] Oliveira H, Costa AR, Konstantinides N, et al. Ability of virulence of Xanthomonas oryzae pv. oryzicola. phages to infect Acinetobacter calcoaceticus-Acinetobacter Microbial Pathogenesis. 2015;78:87–94. baumannii complex species through acquisition of dif­ [62] Van Schie BJ, Hellingwerf KJ, van Dijken JP, et al. ferent pectate lyase depolymerase domains. Environ Energy transduction by electron transfer via a Microbiol. 2017;19(12):5060–5077. pyrrolo-quinoline quinone-dependent glucose dehy­ [66] Kok RG, D’Argenio DA, Ornston LN. Mutation analy­ drogenase in Escherichia coli, Pseudomonas aeruginosa, sis of PobR and PcaU, closely related transcriptional and Acinetobacter calcoaceticus (var. lwoffi). J Bacteriol. activators in Acinetobacter. J Bacteriol. 1998;180 1985;163(2):493–499. (19):5058–5069. [63] Hommes RW, Van Hell B, Postma PW, et al. The [67] Eijkelkamp BA, Hassan KA, Paulsen IT, et al. functional significance of glucose dehydrogenase in Investigation of the human pathogen Acinetobacter Klebsiella aerogenes. Arch Microbiol. 1985;143 baumannii under iron limiting conditions. BMC (2):163–168. Genomics. 2011;12(1):126. [64] Jam DB, Dokter P, van Schie BJ, et al. Role of quino­ [68] Montenegro Benavides NA, Alvarez BA, Arrieta-Ortiz protein glucose-dehydrogenase in gluconic acid pro­ ML, et al. The type VI secretion system of Xanthomonas duction by Acinetobacter calcoaceticus. Antonie Van phaseoli pv. manihotis is involved in virulence and Leeuwenhoek. 1984;50(1):76–77. in vitro motility. BMC Microbiol. 2021;21(1):14.