Genomic Comparative Analysis of Two Multi-Drug Resistance (MDR) Acinetobacter baumannii Clinical Strains Assigned to International Clonal Lineage II Recovered Pre- and Post-COVID-19 Pandemic

Simple Summary Acinetobacter baumannii is a problematic bacterium that causes hard-to-treat hospital infections worldwide. Multiple cases of A. baumannii/SARS-CoV-2 co-infection were reported during the pandemic. This fact raised the question of whether the strains in those co-infections had or acquired unique genetic traits. This study is a comparative analysis of two strains from the same clonal group, but one was isolated before the pandemic, and the other was isolated from a patient with COVID-19. Their genomes had a high similarity, indicating that they may have derived from a unique background. However, each genome had numerous unique genes that were involved in virulence and resistance to antimicrobials. These differences could result from adaptative evolution to the human body infected with SARS-CoV-2. Abstract Background: After the emergence of COVID-19, numerous cases of A. baumannii/SARS-CoV-2 co-infection were reported. Whether the co-infecting A. baumannii strains have distinctive characteristics remains unknown. Methods and Results: A. baumannii AMA_NO was isolated in 2021 from a patient with COVID-19. AMA166 was isolated from a mini-BAL used on a patient with pneumonia in 2016. Both genomes were similar, but they possessed 337 (AMA_NO) and 93 (AMA166) unique genes that were associated with biofilm formation, flagellar assembly, antibiotic resistance, secretion systems, and other functions. The antibiotic resistance genes were found within mobile genetic elements. While both strains harbored the carbapenemase-coding gene blaOXA-23, only the strain AMA_NO carried blaNDM-1. Representative functions coded for by virulence genes are the synthesis of the outer core of lipooligosaccharide (OCL5), biosynthesis and export of the capsular polysaccharide (KL2 cluster), high-efficiency iron uptake systems (acinetobactin and baumannoferrin), adherence, and quorum sensing. A comparative phylogenetic analysis including 239 additional sequence type (ST) 2 representative genomes showed high similarity to A. baumannii ABBL141. Since the degree of similarity that was observed between A. baumannii AMA_NO and AMA166 is higher than that found among other ST2 strains, we propose that they derive from a unique background based on core-genome phylogeny and comparative genome analysis. Conclusions: Acquisition or shedding of specific genes could increase the ability of A. baumannii to infect patients with COVID-19.


Introduction
Infections that are caused by Acinetobacter are associated with mortality rates as high as 60% [1][2][3][4]. Multi-drug resistance (MDR) isolates are becoming more common, and many of them include carbapenems as the antibiotics to which they are immune [5,6]. Consequently, the World Health Organization (WHO) and the Centers for Disease Control and Prevention (CDC) have designated A. baumannii as a high-priority pathogen for antibiotic research and development.
A. baumannii causes a variety of infections, such as ventilator-associated pneumonia, bloodstream, and catheter-associated urinary tract infections [7,8]. This bacterium is often part of polymicrobial infections, which are common in hospitalized patients [9,10]. Gram-negative organisms can take advantage of being a component of polymicrobial infections [11][12][13][14]. Co-infections of Acinetobacter with other bacterial, fungi, or viral pathogens complicate the design of successful therapies [15][16][17][18]. Our recent results showed that A. baumannii and Staphylococcus aureus can coexist at the site of infection, producing a more aggressive process [19]. A. baumannii can sense and undergo phenotypic changes in response to molecules that are secreted by S. aureus [20]. Other kinds of co-infections are those involving viruses and bacteria, which raise significant medical concerns. Coinfections that are caused by different bacteria and the SARS-CoV-2 virus were widely reported [21][22][23][24][25][26]. A study found that OXA-23-producing carbapenem-resistant sequence type (ST) 2 A. baumannii (ST2-CRAB) isolates were the etiologic agent of outbreaks among COVID-19 patients at the ICU in Tehran, Iran [27]. A. baumannii/SARS-CoV-2 co-infection has not been reported yet in South America. The present study aimed to analyze and compare two ST2 A. baumannii clinical strains from Argentina. One of them was isolated from a patient before the pandemic (AMA166), and the other from a COVID-19 patient (AMA_NO). Extensive genomic analysis suggested that the two strains derive from a common ancestor. The studies also identified genetic determinants that shaped the strains' antimicrobial resistance profile. The differences in resistance profiles between both strains could result from their adaptative evolution.

Bacterial Isolates
A. baumannii AMA166 was isolated in 2016 (Argentina) from a 64-year-old patient with type 2 diabetes and arterial hypertension. The A. baumannii AMA_NO strain was isolated in 2021 (Argentina) from a 58-year-old patient with COVID-19. Both patients were hospitalized and received colistin as monotherapy. Both strains were cultured in LB medium and were identified using MALDI-TOF MS [28].

Genome Sequencing
Genomic DNA was extracted using the DNeasy Blood and Tissue kit (Qiagen Germantown, MD, USA) following the manufacturer's instructions. Whole genome sequencing was carried out on the NextSeq 550 Illumina (MiGS sequencing service). Quality control of sequencing was performed using the FASTQ software. De novo assembly and quality assessment were done with the SPAdes and the QUAST software, respectively [29,30]. The Whole Genome Shotgun project has been deposited at GenBank with accession numbers JANKJZ000000000 and JANKKA000000000 for AMA166 and AMA_NO, respectively.

Comparative Genome Analysis
Genome annotation of both strains was performed using the PROKKA software [31]. The ortholog functional assignment was done using EggNOG v2.0 (default parameter) [32]. The taxonomy assignment was performed by ANI% with the seven reference genomes from A. baumannii-calcoaceticus complex using the JSpeciesWS software using the default parameter [33].
To determine the core-genome, the initial datasets of the 2938 ST2 genomes were downloaded from NCBI GenBank (January 2022). The core-genome genes were obtained using the Roary software using the default parameters (Table S1) [42]. A total of 722 core genes were concatenated and aligned using the MAFFT software (parameter: -auto) [43]. The core-genome sequences were aligned and clustered by nucleotide identity (nucleotide identity > 98%) using the "Decipher" R package [44]. One reference genome of each cluster was selected from a total of 239 clusters (Table S1). All the selected genomes were included in the study to perform the core-genome phylogeny analysis.
The core genome phylogeny analysis was performed applying the maximum likelihood method using the RAxML software using the default parameters [45]. The substitution genetic model was done by JModelTest2 using the default parameters [46]. SNPs were extracted using the snp-sites software using the default parameters [47]. The genes that were unique to each genome were extracted from "gene_presence_absence.csv" of Roary output.
The co-linearity analysis was performed using the progressive mauve algorithm with default parameters [48].

RNA Extraction and qRT-PCR Analysis
The A. baumannii AMA_NO strain was cultured in LB broth for 24 h at 37 • C with shaking. RNA extraction was performed using the Direct-zol RNA miniprep Kit (ZYMO research, Irvine, CA, USA). Quantification of RNA was performed using a DeNovix DS-11+ Spectrophotometer. RNA quality was assessed on a 1.5% agarose gel via gel electrophoresis. DNase treatment was performed following the manufacturer's instructions (Thermo Fisher Scientific, Waltham, MA, USA) and quantified as previously described [49,50]. The absence of DNA was confirmed by PCR amplification of the 16S rDNA gene. Reverse transcription was carried out using the iScript Reverse Transcription Supermix for qRT-PCR (BioRad, Hercules, CA, USA) according to the manufacturer's instructions. qRT-PCR was performed using iQ™SYBR Green Supermix (BioRad, Hercules, CA, USA) per the manufacturer's recommendations. Specific oligonucleotides to amplify bla OXA-23 and bla NDM-1 were used. Transcriptional levels of each sample were normalized to the transcriptional level of rpoB. The relative quantification of gene expression was performed using the comparative threshold method 2 −∆Ct [51]. The reactions were performed in four technical and three biological replicates, respectively. The statistical analysis (t-test) was performed using GraphPad Prism (GraphPad software, San Diego, CA, USA). A p-value < 0.05 was considered significant.

Antibiotic Susceptibility Assays
The antimicrobial resistance profile was determined by disk diffusion (10 µg ampicillin/sulbactam, 30 µg amikacin, 30 µg cefepime, 10 µg ceftazidime 5 µg ciprofloxacin, 10 µg imipenem, 10 µg gentamicin, 10 µg meropenem, 15 µg tigecycline, 30 µg minocycline, or 10 µg colistin) and minimum inhibition concentration determination according to the Clinical and Laboratory Standards Institute (CLSI) recommendations [52]. The experi- ments were repeated at least three times for each strain. The results were interpreted with CLSI guidelines, except for colistin and tigecycline, in which cases European Committee on Antimicrobial Susceptibility Testing (EUCAST) and Food and Drug Administration (FDA) recommendations were used, respectively. The CLSI, EUCAST, and FDA publish guidelines for antimicrobial susceptibility testing (AST) that provide recommendations for testing and interpreting the susceptibility of microorganisms to antimicrobial agents. These guidelines include recommendations for standardized methods, quality control procedures, and interpretive criteria for AST [52,53].

Sequencing of AMA166 and AMA_NO. Genomic and Phylogenomic Comparative Analyses
Strains AMA_NO and AMA166, isolated post-and pre-COVID pandemic, were used to study changes at the genomic level to identify gene acquisition that could indicate patterns of adaptative genomic evolution [54,55]. The source of the AMA_NO strain was a patient with COVID-19. The whole genome sequences of both strains were of good quality, and depth was greater than 50X coverage ( Table 1). The assembly quality was evaluated using the QUAST software and produced 79 (AMA_NO) and 38 (AMA166) contigs [30]. Both genomes had similar sizes with a difference of 254.427 bp between them. The N50 of AMA_NO and AMA166 were 125.130 and 220.435, respectively. The numbers of tRNAs identified were 61 (AMA_NO) and 60 (AMA 166) ( Table 1). Mobile genetic elements, such as insertion sequences (ISs), prophages, and plasmids, play an important role in genome evolution and the dissemination of antimicrobial resistance genes in A. baumannii. ISs were identified using the BLAST algorithm and the ISFinder database [39]. A total of three ISs were common in both bacteria, but strain AMA_NO possessed nine more (Table S2). Also, the disrupted transposases of ISs (as pseudogene) were found in both genomes. There were twelve and two disrupted IS transposases that were found in AMA_NO and AMA166 genomes, respectively (Table S2).
Analysis using the PHASTER online software found four prophages that were common to both strains (Table S2). The PHASTER prediction classification detected one intact category prophage, one incomplete category prophage, and two questionable category prophages in both genomes. Plasmids were not found when searching for known rep and mob genes [41].
A total of 41 and 38 non-coding RNA (ncRNA) regulatory elements were identified in strains AMA_NO and AMA166, respectively ( Table 1). The three ncRNA that were present only in strain AMA_NO were ALILL, group-II-D1D4-2, and Intron_gpII (Table S3). ALILL pseudoknot is an RNA element that induces frameshifting. This element was identified through comparative analysis of a class of transposable elements belonging to the IS3 family and is conserved across bacterial species, such as Lactobacillus lactis, Escherichia coli, Acinetobacter species, etc. [56][57][58]. Group II introns (group-II-D1D4-2 and Intron_gpII) are a large class of self-catalytic ribozymes and mobile genetic elements that are found within the genes of all three domains of life. Remarkably, AMA_NO contains seven disrupted IS3 family transposase genes (Table S2). Considering these results, we hypothesize that the Group II introns might have had regulatory functions of the IS3-family insertion sequences activity in ancestors of the AMA_NO strain. The presence of incomplete IS3 family sequences could be due to a reduction of the genome and a possible path towards a greater specialization to occupy one or more specific ecological niches. The MLST profile was determined using the Pasteur and Oxford scheme. AMA_NO and AMA166 belong to the ST2 (Pasteur)/ST208 (Oxford) clone (Clonal Complex 2). With the Oxford scheme, duplication of gdhB was identified, corresponding to alleles 3 and 189 (paralogous genes). Considering Gaiarsa et al.'s report, the allele 3 was considered, assigned the ST208 as the MLST profile [59]. A core-genome phylogenetic analysis of ST2 (Pasteur) clone showed two main phylogenetic clusters (A and B) ( Figure 1). Both strains clustered together in the phylogenetic cluster A with high similarity with the A. baumannii ABBL141 strain isolated in the USA (Figure 1). In the A. baumannii ABLL141, identified as ST208 in the scheme Oxford, the same duplication of gdhB alleles that were found in our strains (AMA_NO and AMA166) is present (  A comparison of gene content showed that 3546 genes were shared by both genomes, while 269 and 14 were unique to AMA_NO and AMA166, respectively (Table S3). In addition, co-linearity analysis was performed to evaluate the genome structural variation between AMA_NO and AMA166. There were six local collinear blocks (LCBs), which indicate genomic regions with the highest homology, that were determined between both genomes. LCB 1 was observed inverted and translocated in AMA166, while LCB 4 and 5 were translocated in AMA166 ( Figure S1).
To assess the possible biological activity or function of unique genes, functional annotation was performed using the EggNog Mapper v 2.0 software [32]. At least one category was assigned to 237 out of 269 and 10 out of 14 unique genes of AMA_NO and AMA166, respectively (Table S4). The potential functions of the unique genes in AMA166 were transposases, hydrolases, and monoxidases. Some possible functions of the unique genes in AMA_NO were related to virulence, such as folate biosynthesis (folKP, sul1, sul2), pyruvate metabolism (adh, frmA, adhP, pdhD), biofilm formation (impC), and tyrosine metabolism (adh, frmA, adhP) (Table S4). These results indicate that post-pandemic A. baumannii ST2 may have acquired virulence-associated genes, which suggests an increase in pathogenicity and resistance to treatment (Table S4). Similar observations were found in other opportunistic pathogens, such as Klebsiella pneumoniae and Aspergillus [60,61].

Antibiotic Resistance, Virulence, and Its Association with Horizontal Genetic Transfer (HGT) Elements
A total of 18 and 10 ARG were identified in strains AMA_NO and AMA166, respectively ( Figure 2). The intrinsic β-lactamases bla OXA-66 and bla ADC-25 were present in both genomes ( Figure 2); according to their genetic environment, they do not seem to be associated with a transposable element. Both strains include a copy of Tn2008, a transposon that contains the bla OXA-23 gene. Identical TnAbaR multidrug resistance genomic islands harboring tetB (tetracycline resistance), strA, and strB (streptomycin resistance) were found in strains AMA_NO and AMA166 genomes (Figure 3). This TnAbaR possesses an identical resistance gene array to those in 907 out of the 6702 A. baumannii genomes deposited in GenBank at the time of this study. When considering only the A. baumannii ST2 clone genomes, the TnAbaR present in strains AMA166 and AMA_NO was found in 202 out of 2938 genomes currently in GenBank (Figure 3). There were three efflux pumps, AdeABC, AdeFGH, AdeIJK, and their regulator systems that were found in both genomes. These efflux pumps are associated with fluorquinolone, carbapenem, cephalosporin, phenicol, macrolide, tetracycline, rifampicin, glycylcycline, and lincosamide resistance [62][63][64]. Also, the ade efflux pumps are associated with biofilm formation, fitness, and pathogenesis [62]. Both genomes contain the class 1 integron with the same gene cassette array, aacC1-orfP-orfQ-aadA1, in the variable region. This gene cassette array is present in 167 of the 2938 ST2 A. baumannii genomes in GenBank.
The human body has specific and non-specific defenses against infection. Early ies on the latter led to the discovery of the "hypoferremic response," which consist reduction of free iron levels in blood and fluids in response to a bacterial invasion The consequence of the hypoferremic response is the iron starvation of invading bac Later, it was found that iron is not the only essential element whose availability is red to interfere with bacterial growth. This understanding originated the concept of "n tional immunity," the group of non-specific defense strategies based on deprivation metal ions [77,78]. Bacteria must circumvent these nutritional limitations to establ the body and cause disease. A common mechanism to overcome the lack of available is the biosynthesis of a siderophore that is secreted to the environment and competes iron with high-affinity iron-binding proteins. Then, the iron-siderophore complexe recognized by specific bacterial receptors to uptake iron [79,80].
A. baumannii iron uptake systems compete with high-affinity iron-binding hos teins to capture essential iron for the survival and progress of the infection. Antunes identified six iron-uptake systems in A. baumannii. They are coded for by the acinetob locus, baumannoferrin locus, fimsbactin locus, Heme uptake cluster 2, Heme uptake ter 3, and feoABC genes [81]. The acinetobactin locus, baumannoferrin locus, Heme cl 2, and feoABC genes were found in AMA_NO and AMA166 (Table S6). However, th netobactin and baumannoferrin loci were incomplete in both strains and are most p bly nonfunctional. In the acinetobactin locus, the genes coding for basE (ACICU_0 and a hypothetical protein (ACICU_02575) were missing. In the case of the baumann rin locus, there was a missing gene that encodes a siderophore synthetase (ACICU_01 The four iron-uptake systems that were identified in A. baumannii AMA_NO and AM were also found in 545 out of 2938 ST2 genomes (Table S7). The complete acinetob and baumannoferrin loci were present in 2273 out of 2938 (77.36%) and 2839 out of ST2 genomes (96.63%), respectively. The complete sets of other systems were fou different percentages of ST2 genomes (Table S7). While every ST2 genome contains at one iron uptake system (Table S6), there is heterogeneity in the combinations of sys that are present in each strain. A previous study that included 111 A. baumannii ST nomes found that they all harbor incomplete baumannoferrin and acinetobactin loc This homogeneity can be explained by the small ST2 genomes sample, which migh have been sufficient to represent the iron-uptake cluster variability of the ST2 clone rately.
It must be noted that only the acinetobactin iron-uptake system has been linke high virulence phenotype [83]. The absence of acinetobactin in 665 ST2 strains cou the reason behind the low pathogenicity in some isolates of this clone (Table S6).  The ARGs bla NDM-1 (carbapenem resistance), floR (florfenicol/chloramphenicol resistance), sul2 (sulfonamide resistance), ble MBL (bleomycin resistance), msrE (macrolide and phenicol resistance), mphE (macrolide resistance), cmlB1 (phenicol resistance), and aphA6 gene (aminoglycoside resistance) were found only in AMA_NO (Figure 2). The aphA6 gene and the ISAba125 are located downstream of bla NDM-1 in the TnAphA6 transposon. The bla NDM-1 -TnAphA6 backbone was not found in any ST2 A. baumannii genomes, but it was present in 11 out of 3759 non-ST2 A. baumannii genomes. We analyzed the contig containing the bla NDM-1 -TnAphA6 structure (contig 55) and compared it with genomic sequences in the GenBank database. A total of 101 sequences were identified with 98-100% nucleotide identity and 98-100% coverage. Most sequences belonged to plasmids (88 sequences), mainly identified in the genus Acinetobacter (49 sequences). However, other plasmid sequences were also identified in other genera, such as Escherichia, Klebsiella, Citrobacter, and Providencia spp. (Table S5). Consequently, strain AMA_NO, as well as AMA166 are the first strains belonging to the ST2 clone to include the bla NDM-1 -TnAphA6 array. The co-existence of bla OXA-23 and bla NDM-1 A. baumannii ST2 genomes occurs in a small percentage of these strains (63 out of 2938 ST2 genomes). Furthermore, only 109 A. baumannii ST2 genomes carry bla NDM-1 . The bla OXA-23 gene has a greater representation among this group (521/2938 ST2 genomes). Both strains are resistant to ampicillin/sulbactam, cefepime, ceftazidime, ciprofloxacin, gentamicin, imipenem, and meropenem and are susceptible to minocycline, tigecycline, colistin, and amikacin ( Table 2).
A recent study reported the co-existence of blaOXA-23 and blaNDM-1 in six isolates with ceftazidime and imipenem MIC values greater than 256 µg/mL and 32 µg/mL [86]. Quantitative RT-PCR (qRT-PCR) assays using total RNA extracted from A. baumannii AMA_NO cells that were cultured in LB showed two-fold higher blaOXA-23 mRNA levels compared to blaNDM-1 (Figure 4). Although other factors may impact the contribution of each gene to the resistance phenotype, the higher expression of blaOXA-23 may reflect a higher contribution to carbapenem resistance. It is also of interest that the blaOXA-23 gene was found in numerous A. baumannii clinical isolates but rarely in other Gram-negative species. Conversely, blaNDM-1 is a promiscuous gene, which could indicate a recent adaptive process. The genetic location of this gene can explain NDM global dissemination. Plasmids carrying blaNDM have been described globally, with Klebsiella and Escherichia as the prevalent genus harboring them [87]. In Klebsiella pneumoniae, NDM has been reported in a wide variety of different STs worldwide, supporting its broad ability for dissemination [88].    Virulence factors refer to traits (i.e., gene products) that allow a microorganism to establish itself on or within a host and enhance its potential to cause disease. Several literature reports of genome comparative analyses confirm the multifactorial and combinatorial nature of A. baumannii virulence [65][66][67][68][69][70][71][72]. A total of 46 genes associated with virulence factors were found in AMA_NO and AMA166 (Table S6). A. baumannii produces a capsular polysaccharide (CPS) encoded by a gene cluster that is referred to as the K locus (KL). In contrast, the variable outer oligosaccharide of the lipopolysaccharides is encoded by the OC locus (OCL). CPS is an outer layer that is involved in protection against C3 deposition that occurs mainly in inhibiting macrophage phagocytosis. O antigen is responsible for the resistance of bacteria to complement mediated killing. Both components are essential to the blood passage of bacteria and the development of sepsis, but only CPS is involved in developing pulmonary infections [37,[73][74][75]. The OCL5 and KL2 were found in AMA_NO and AMA166, respectively.
The human body has specific and non-specific defenses against infection. Early studies on the latter led to the discovery of the "hypoferremic response," which consists of a reduction of free iron levels in blood and fluids in response to a bacterial invasion [76]. The consequence of the hypoferremic response is the iron starvation of invading bacteria. Later, it was found that iron is not the only essential element whose availability is reduced to interfere with bacterial growth. This understanding originated the concept of "nutritional immunity," the group of non-specific defense strategies based on deprivation of the metal ions [77,78]. Bacteria must circumvent these nutritional limitations to establish in the body and cause disease. A common mechanism to overcome the lack of available iron is the biosynthesis of a siderophore that is secreted to the environment and competes with iron with high-affinity iron-binding proteins. Then, the iron-siderophore complexes are recognized by specific bacterial receptors to uptake iron [79,80].
A. baumannii iron uptake systems compete with high-affinity iron-binding host proteins to capture essential iron for the survival and progress of the infection. Antunes et al. identified six iron-uptake systems in A. baumannii. They are coded for by the acinetobactin locus, baumannoferrin locus, fimsbactin locus, Heme uptake cluster 2, Heme uptake cluster 3, and feoABC genes [81]. The acinetobactin locus, baumannoferrin locus, Heme cluster 2, and feoABC genes were found in AMA_NO and AMA166 (Table S6). However, the acinetobactin and baumannoferrin loci were incomplete in both strains and are most probably nonfunctional. In the acinetobactin locus, the genes coding for basE (ACICU_02578) and a hypothetical protein (ACICU_02575) were missing. In the case of the baumannoferrin locus, there was a missing gene that encodes a siderophore synthetase (ACICU_01632). The four iron-uptake systems that were identified in A. baumannii AMA_NO and AMA166 were also found in 545 out of 2938 ST2 genomes (Table S7). The complete acinetobactin and baumannoferrin loci were present in 2273 out of 2938 (77.36%) and 2839 out of 2938 ST2 genomes (96.63%), respectively. The complete sets of other systems were found in different percentages of ST2 genomes (Table S7). While every ST2 genome contains at least one iron uptake system (Table S6), there is heterogeneity in the combinations of systems that are present in each strain. A previous study that included 111 A. baumannii ST2 genomes found that they all harbor incomplete baumannoferrin and acinetobactin loci [82]. This homogeneity can be explained by the small ST2 genomes sample, which might not have been sufficient to represent the iron-uptake cluster variability of the ST2 clone accurately.
It must be noted that only the acinetobactin iron-uptake system has been linked to a high virulence phenotype [83]. The absence of acinetobactin in 665 ST2 strains could be the reason behind the low pathogenicity in some isolates of this clone (Table S6).
A recent study reported the co-existence of bla OXA-23 and bla NDM-1 in six isolates with ceftazidime and imipenem MIC values greater than 256 µg/mL and 32 µg/mL [86]. Quantitative RT-PCR (qRT-PCR) assays using total RNA extracted from A. baumannii AMA_NO cells that were cultured in LB showed two-fold higher bla OXA-23 mRNA levels compared to bla NDM-1 (Figure 4). Although other factors may impact the contribution of each gene to the resistance phenotype, the higher expression of bla OXA-23 may reflect a higher contribution to carbapenem resistance. It is also of interest that the bla OXA-23 gene was found in numerous A. baumannii clinical isolates but rarely in other Gram-negative species. Conversely, bla NDM-1 is a promiscuous gene, which could indicate a recent adaptive process. The genetic location of this gene can explain NDM global dissemination. Plasmids carrying bla NDM have been described globally, with Klebsiella and Escherichia as the prevalent genus harboring them [87]. In Klebsiella pneumoniae, NDM has been reported in a wide variety of different STs worldwide, supporting its broad ability for dissemination [88]. compared to blaNDM-1 (Figure 4). Although other factors may impact the contribution of each gene to the resistance phenotype, the higher expression of blaOXA-23 may reflect a higher contribution to carbapenem resistance. It is also of interest that the blaOXA-23 gene was found in numerous A. baumannii clinical isolates but rarely in other Gram-negative species. Conversely, blaNDM-1 is a promiscuous gene, which could indicate a recent adaptive process. The genetic location of this gene can explain NDM global dissemination. Plasmids carrying blaNDM have been described globally, with Klebsiella and Escherichia as the prevalent genus harboring them [87]. In Klebsiella pneumoniae, NDM has been reported in a wide variety of different STs worldwide, supporting its broad ability for dissemination [88]. in the AMA_NO strain. Fold changes were calculated using ΔCt analysis. At least three independent samples were used, and four technical replicates were performed from each sample. Student's t-test analysis was performed using GraphPad Prism (GraphPad software, San Diego, CA, USA). **: A pvalue < 0.01 was considered statistically significant. Data are presented as the mean ± SD.
The AMA_NO was isolated from a patient that was also infected with SARS-CoV-2. Bacterial/viral coinfections are not rare and are often synergistic, i.e., they produce enhanced symptomatic manifestations. Numerous pairs of bacteria/viruses that act synergically to cause or enhance infection have been described in the literature. A Figure 4. Expression analysis by qRT-PCR of genes coding for β-lactamases (bla OXA-23 and bla NDM-1 ) in the AMA_NO strain. Fold changes were calculated using ∆Ct analysis. At least three independent samples were used, and four technical replicates were performed from each sample. Student's t-test analysis was performed using GraphPad Prism (GraphPad software, San Diego, CA, USA). **: A p-value < 0.01 was considered statistically significant. Data are presented as the mean ± SD.
The AMA_NO was isolated from a patient that was also infected with SARS-CoV-2. Bacterial/viral coinfections are not rare and are often synergistic, i.e., they produce enhanced symptomatic manifestations. Numerous pairs of bacteria/viruses that act synergically to cause or enhance infection have been described in the literature. A representative example is the Streptococcus pneumoniae and H1N1 influenza virus coinfection, which produce a lethal synergy [89]. The authors of this study observed that the bacterial infection induced a loss of lung repair responses. The fatal outcome was correlated with a loss of airway basal epithelial cells. Interestingly, the S. pneumoniae SirRH two-component system plays a role in the enhancement of bacterial survival inside lung cells that were previously infected with H1N1 [90]. Another mechanism that results in synergism is through the virus neuraminidase, which mediates the removal of sialic acid from the cell surface and facilitates the adhesion of invading bacteria [91]. Another significant observation is that some bacterial strains, when coinfecting with viruses, stimulated genetic recombination between two or more different viruses, acting as a driver of viral evolution. We hypothesize that the interactions between bacterial and viral elements could also result in bacterial acquisition of genes that increase virulence [92]. A future systematic study of bacterial isolates from SARS-CoV-2-infected patients may prove this hypothesis or not.

Conclusions
A. baumannii AMA_NO and AMA166 appear to belong to one clone. However, differences were identified in the pre-and post-pandemic strains, which could indicate adaptation to the environment found within the COVID patient. The SARS-CoV-2 viral infection could be a driver for evolutionary modifications. Expansion of genomic comparisons is needed to validate or disprove this hypothesis.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/biology12030358/s1. Figure S1: Co-linearity analysis between AMA_NO and AMA166 genomes. Pairwise alignment was performed using the progressive Mauve algorithm with the Mauve software. The LCBs are represented by different rectangles with different colors and numbers. The high-quality figure is deposited in the GitHub Repository. (https://github. com/germant13/Traglia2023_AbST2_SARS.git, accessed on 9 February 2023). Table S1: Sheet 1: List of A. baumannii ST2 clones' genomes from NCBI genomes downloaded to the study and the ST profile using Oxford scheme. Sheet 2: Clustering of ST2 genomes from NCBI genomes database. Yellow cells indicate the ST2 reference genome that was selected to perform the core-genome phylogenetic analysis. Sheet 3: Core-genes predicted by the Roary software.  Table S3: Sheet 1: ncRNA elements predicted in A. baumannii AMA166 using Infernal software. Sheet 2: ncRNA elements predicted in A. baumannii AMA_NO using Infernal software. Table S4: Sheet 1: List of shared and unique genes of A. baumannii AMA_NO and AMA166. Sheet 2: Unique genes with functional annotation in A. baumannii AMA166 using EggNOG 2.0v Software. Sheet 3: Unique genes with functional annotation in A. baumannii AMA_NO using EggNOG 2.0v Software. Table S5: Analysis of nucleotide identity and coverage of contig 55 containing the blaNDM-1-TnAphA6 genetic structure in the GenBank database. Table S6: Sheet 1: Virulence genes of A. baumannii AMA166 predicted using VFDB database. Sheet 2: Virulence genes of A. baumannii AMA_NO predicted using VFDB database. Sheet 3: Iron-uptake genes of A. baumannii AMA_NO and AMA166 Table S7: Matrix of absence/presence iron uptake locus in A. baumannii ST2 genomes recovered from NCBI GenBank database (n = 2938). Data Availability Statement: The Whole Genome Shotgun project has been deposited at GenBank with accession numbers JANKJZ000000000 and JANKKA000000000 for AMA166 and AMA_NO, respectively.

Conflicts of Interest:
The authors declare no conflict of interest.