Biofilm and Gene Expression Characteristics of the Carbapenem-Resistant Enterobacterales, Escherichia coli IMP, and Klebsiella pneumoniae NDM-1 Associated with Common Bacterial Infections

In light of the limited therapeutic options with Carbapenem-Resistant Enterobacterales (CRE) infections, understanding the bacterial risk factors, such as biofilm formation and related gene expression of CRE, is vital. This study investigates the biofilm formation and biofilm-related gene expression of two enteric Enterobacterales with major CR determinants Escherichia coli IMP and Klebsiella pneumoniae NDM-1, which were seen in high prevalence in most common bacterial infections over the past few years. To our knowledge, this is the first study that demonstrated the relationship between biofilm formation and the related gene expression, to understand the potential molecular mechanisms during the biofilm formation in CRE. Biofilms were quantified by tissue culture plate assay at the stages of the biofilm development: initial attachment (6 h), microcolony formation (12 h), maturation (24 h), and dispersion (48 h). In a dispersion, event bacteria detach without any mechanical means and colonise another area. To investigate the influence of different growth conditions on biofilm formation, biofilms were quantified under different growth conditions. In parallel, quantitative real-time PCR (qPCR) assessed the biofilm-related gene expression of a cluster of genes, including biofilm maturation, quorum sensing, stress survival, and antibiotic resistance. Structural changes during biofilm development were assessed via confocal laser scanning microscopy (CLSM). We observed that the biofilm formation of CRE is correlated with the biofilm development stages, with maximum biofilm observed at 24 h at the maturation stage. Our data also showed that biofilm growth, under the condition tested, is the major factor influencing the variability of biofilm gene expression quantification assays. qPCR analyses have demonstrated that the expression of biofilm-related genes is highly correlated with phenotypic biofilm development, and these findings can be further expanded to understand the variation in regulation of such genes in these significant CRE pathogens. Our study demonstrated that both CRE strains, E. coli IMP and K. pneumoniae NDM-1, are high biofilm formers, and genes involved in biofilm development are upregulated during biofilm growth. The characteristic of the increased biofilm formation with the upregulation of antibiotic-resistant and biofilm-related genes indicates the successful pathogenic role of biofilms of these selected CRE and is attributed to their multi-drug resistance ability and successful dissemination of CRE in common bacterial infections.


Introduction
The Enterobacterales, such as Escherichia coli and Klebsiella pneumoniae, are the most common causes of community and healthcare-associated urinary tract infections and are also associated with the most lethal lower respiratory and bloodstream infections [1][2][3]. There has been a vicious cycle of increasing resistance in Enterobacterales, specifically with in this study. Previous studies on biofilm analysis showed that there is a heterogeneity of biofilm formation in different media in a given strain [23,24].
To assess the effect of different growth media on biofilm development, we considered two types of growth media. Tested strains were cultured in two types of media, a nutritionally poor medium, Minimal Salts Medium (AB), and nutrient-rich Luria-Bertani Broth (LB). AB medium was used to mimic the low nutrient condition. For biofilm analysis collectively used three passages for each tested strain. Biofilms were quantified at different growth stages of biofilm development such as initial attachment (6 h), microcolony formation (12 h), maturation (24 h), and dispersion (48 h), at 37 • C under static and shaker conditions. The effect of aeration and distribution of nutrients on the biofilm formation process were also explored in this study by incubating test strains under shaking and static conditions.

Quantification of Biofilms
Using the 96-Well Tissue Culture Plate Assay E. coli IMP and K. pneumoniae NMD-1 biofilms were analysed using the 96-well tissue culture plate assay, according to the previously described methods with some modifications [18,25]. In summary, the strains were grown overnight in AB and LB media (5 mL) at 37 • C, where the optical density of each inoculum was adjusted to OD 595 , 0.08 nm, where OD 595 is the optical density measured at 595 nm. A 200 µL of inoculated cell suspension was then added to four 96-well plates and incubated at 37 • C in static and shaking conditions at 70 rpm, where the incubation was conducted for various time intervals of 6, 12, 24, and 48 h. After the intended period of incubation, the broth was eliminated from the plate wells using Pasteur pipettes, and the wells were rinsed twice with sterile phosphate-buffered saline (PBS) to remove any non-adherent planktonic cells. The biofilms in the wells were then stained by the addition of 125 µL of 1% crystal violet and incubated for 30 min at room temperature. The stain was rinsed with sterile water until clean and the dye was re-solubilised with 33% v/v acetic acid [26], which was then left at room temperature for 30 min. Finally, the optical density of each well was measured at OD 595 (measured at 595 nm) to evaluate the amount of biofilm formed. A medium-only negative control was included in each assay. The biofilms were classified in categories as strongly adherent (OD 595 > 0.5), moderately adherent (OD 595 0.3-0.5), weakly adherent (OD 595 < 0.3), lack of biofilm OD 595 < 0.15 and non-adherent (OD 595 = 0). All the tests for each microorganism were performed in triplicate at each growth stage, under shaking or static conditions, and for each growth media [27].

Biofilm Analysis Using Confocal Laser Scanning Microscopy (CLSM)
CSLM analysis was performed based on previously established protocols [25][26][27][28][29] E. coli IMP and K. pneumoniae NDM-1 were grown overnight in LB media (5 mL) at 37 • C, where the optical density of each inoculum was adjusted to OD 595 , nm 0.08, and 700 µL aliquots were placed into on µ-slide two wells with a chambered coverslip, (Ibidi, Gräfelfing, Germany). Each well with a lid was incubated statically at 37 • C for different periods (6, 12, 24, and 48 h). After incubation, the µ-slide was carefully washed with PBS buffer to remove all planktonic cells. An amount of 6 µL of Live/Dead staining (LIVE/DEAD BacLight TM Bacterial Viability Kits, Molecular Probes, Bleiswijk, The Netherlands) was added according to the manufacturer's instructions at room temperature and set for 15 min in the dark. The biofilms were fixed using 500 µL of Image-It fixative solution of 4% Formaldehyde (Thermo Scientific, Horsham, UK) and incubated at room temperature in the dark. The biofilms on the wells were washed with PBS three times and ProLong TM Gold Antifade reagents (Molecular Probes, Bleiswijk, The Netherlands). The test samples were then imaged via confocal laser scanning microscopy using an Olympus FV 1000 laser scanning microscope (Olympus, Tokyo, Japan). All the tests for each microorganism were performed in duplicate and representative images were selected for each growth condition.

RNA Isolation from Bacterial Biofilms and cDNA Synthesis
Bacterial strains were inoculated overnight in AB and LB media (5 mL) at 37 • C, where the optical density of each inoculum was adjusted to OD 595 , 0.08 nm, and 6 mL aliquots were transferred into the six-wells of a polystyrene tissue culture plate. The plate was covered with a lid and then incubated statically for 6, 12, 24, and 48 h at 37 • C. After incubation, the wells were carefully washed with sterile PBS buffer. The adhering bacterial cells in each well were disrupted and transferred into a cold sterile PBS buffer. The suspension was then treated with a double volume of RNA protection reagent (QIAGEN, Hilden, Germany) and immediately incubated for 5 min at room temperature and vortexed for 5 s. Cell pellets were obtained by centrifuging at 10,000× g for 15 min, then resuspended in 100 µL of RNase-free water (QIAGEN, Germany), after which the cell pellets were stored at −80 • C for subsequent RNA isolation [30].
Bacterial pellets were suspended in 100 µL of RNase-free water (QIAGEN, Hilden, Germany). The RNA isolation was performed using a RNeasy ® Power Biofilm Kit (QIAGEN, Hilden, Germany), according to the manufacturer's guidance [31], whilst DNA decontamination treatment was performed using Turbo DNase (QIAGEN, Hilden, Germany). The RNA was quantified by measuring the absorbance at 260-280 nm using a Nanodrop spectrophotometer ND (Thermo Fisher Scientific, Waltham, MA, USA), and the RNA quality was determined via agarose gel electrophoresis. The RNA was reverse transcribed to cDNA using RevertAid H Minue First Stand cDNA Synthesis (Thermo Fisher Scientific, Waltham, MA, USA) according to the manufacturer's protocol.

Primers and Their Specificities for qPCR
The NCBI database www.ncbi.nim.nih.gov (accessed on 1 December 2019) was used to select gene-specific targets, and primers were designed using the Integrated DNA Technologies, Coralville, IA, USA (IDT) website tools www.idtdna.com (accessed on 1 December 2019). Primers were purchased from IDT. The primers used for E. coli IMP and K. pneumoniae NDM-1 are listed in (Figure 1). Annealing temperatures were optimised for each primer pair using the melting curve analysis and post-PCR agarose gel electrophoresis for the PCR products obtained. Primer efficiencies were calculated using serial dilutions of cDNA (diluted for 100 ng/µL, 50 ng/µL, 10 ng/µL, 1 ng/µL, and 0.1 ng/µL).

Quantitative Real-Time PCR (qPCR)
The transcript levels of the eight genes in E. coli IMP and eight genes in K. pneumoniae NDM-1 which were responsible for antibiotic resistance and biofilm-related activity measured at four different growth stages (i.e., after 6, 12, 24, and 48 h of incubation) and in AB and LB media via qPCR. The strains were grown overnight in AB and LB media (5 mL) at 37 • C, where the optical density of each inoculum was adjusted to OD 595 , 0.08 nm. For each strain, selected genes were categorised into biofilm formation, quorum sensing, stress survival, and antibiotic resistance. The qPCR reaction was carried out using a SensiFAST SYBR NO-ROX Kit (BioLine, Memphis, TN, USA) and according to the previously published study [30]. The total reaction volume was set at 20 µL per sample and contained 200 ng/mL cDNA and 400 nM of each primer. All reactions were performed as technical triplicates using 96-well plates and on two biological replicates.

qPCR Data Analysis Using Modified ∆∆Ct Method
The expression of the eight selected genes in E. coli IMP and eight selected genes in K. pneumoniae NDM-1 at 12, 24, and 48 h was calculated relative to the sample at 6 h (calibrator sample) and normalised to the expression of the reference gene (16S rRNA for each organism). A calibrator sample was selected at 6 h, as at this time a minimal amount of biofilm formation was observed compared to 12, 24, and 48 h, where relatively high biofilm formation was observed [30]. The X-fold change of the transcription level was calculated using the modified ∆∆Ct method [32], with the modification of primer efficiencies that were calculated utilising the serial dilutions of cDNA (100 ng/µL, 50 ng/µL, 10 ng/µL,

Statistical Analysis
Biofilm analysis: The one-way ANOVA statistical analysis method was used to determine the statistically significant differences between the means of biofilm-forming capacity at 6 h, 12 h, 24 h, and 48 h group under the given media (AB media or LB media) and in a given condition (static or shaking) in a given strain (E. coli IMP or K. pneumoniae NDM-1). Prism 6 version 6.01 and Microsoft Excel 365 were used to perform statistical calculations. A p-value < 0.05 is considered a statistically significant maximum biofilm formation. The analysis was performed on three biological replicates and the representative values are the means of triplicates. Collectively, three passages were used for each tested strain.
Gene expression analysis: The one-way ANOVA statistical analysis method was also used to determine the statistically significant differences in relative gene expression of a target gene between 12 h, 24 h, and 48 h groups under the given media (AB media or the LB media) and in a static condition in a test strain (E. coli IMP or K. pneumoniae NDM-1). A p-value < 0.05 is considered a statistically significant increased fold expression, and when p-value > 0.05 is considered a statistically significant decreased fold expression value. The analysis was performed on two biological replicates and the representative values are the means of triplicates.

Biofilm Quantitative Assay
The varying amounts of biofilm formed by the non-resistant control, E. coli 12241 and the CRE strain E. coli IMP at different growth conditions is represented in Figure 2A,B respectively. Similarly, Figure 3A,B represent the biofilm quantities formed by the nonresistant control K. pneumoniae 9633 and CRE stain K. pneumonia NDM-1. Figures 4 and 5 show the mean values of the quantified biofilms including the statistical significance. Figure 6 compares the various adherence levels of non-resistant and resistant strains under the growth conditions tested. As shown in Figure 2, a high amount of biofilm formation was observed for antibioticresistant E. coli IMP compared to non-resistant E. coli 12241 control. As indicated in Figure 6 E. coli IMP was strongly adherent with values of OD 595 > 0.81 compared to the non-resistant control E. coli strain, with its weak adherence of OD 595 < 0.16. For both non-resistant and resistant E. coli strains, varying amounts of biofilm formation were observed corresponding to the biofilm development stages, where statistically significant (p < 0.05), maximum biofilm formation was observed at the 24-h incubation period under all the conditions tested ( Figure 4). E. coli IMP showed strong adherence at the 24 h, and under static conditions in AB with an OD 595 of 0.5, and LB OD 595 of 0.81, respectively ( Figure 4).     Figure 3, a similar pattern of biofilm development was observed for K. pneumoniae strains, where a greater amount of biofilm formation was observed for the CRE strain K. pneumoniae NDM-1 compared to the non-resistant K. pneumoniae 9633 (control) under all conditions tested. As indicated in Figure 6, K. pneumoniae NDM-1 was strongly adherent with an OD 595 > 0.73 compared to the non-resistant control K. pneumoniae strain with a weak adherence of OD 595 < 0.12. For both K. pneumoniae strains, varying amounts of biofilm formation were observed corresponding to the biofilm development stages, where statistically significant (p < 0.05), maximum biofilm formation was observed at the 24-h incubation period under all the conditions tested ( Figure 5). K. pneumoniae NDM-1 also showed strong adherence at the 24 h, and under a static condition with an OD 595 of 0.46 in AB media, and an OD 595 of 0.73 in LB media, respectively ( Figure 5).  Similarly, K. pneumoniae NDM-1 showed that cell surface-associated meshwork like structures is weak at 6 h, with the moderate formation of structures at 12 h. Around 24 h the meshwork structure becomes denser, with a dispersion of these structures observed at 48 h (Figure 8). Additionally, for K. pneumoniae NDM-1, the average optical thicknesses of 0.2 µm, 0.5 µm, 1.5 µm, and 1.1 µm, were observed corresponding to each time point at 6 h, 12 h, 24 h, and 48 h, respectively (Figure 8 lower panels). Interestingly, under all the growth conditions tested, the majority of morphological structures appeared live (green colour) in both E. coli IMP and K. pneumoniae NDM-1, even though they underwent the chemical fixation process during biofilm analysis. The quality and quantity of RNA extracted at different time points and growth conditions for both strains were found to be satisfactory. The most stable reference (housekeeping) genes from a set of selected candidates were determined by using the geNorm algorithm (https://genorm.cmgg.be/ (accessed on 12 December 2019)). From this, a gene expression normalisation factor was calculated for each target gene based on the geometric mean of a selected number of genes, according to the previously described method [33,34]. For E. coli IMP, 16SRNA, rssA, and galE were selected as reference genes and, according to geNorm, 16sRNA and rssA produced the lowest M-values (0.082) in combination, indicating they were the most stably expressed genes across all samples. Similarly, for K. pneumoniae NDM-1 16SRNA, recA and tonB were selected as reference candidates, and according to geNorm, 16S RNA and recA produced the lowest M-values (0.21) in combination, indicating they were the most stably expressed genes across all samples.

Confocal Laser Scanning Microscopy (CLSM)
The primer efficiencies for genes tested for both E. coli IMP and K. pneumoniae NDM-1 were evaluated using calculated serial dilutions of cDNA (100 ng/µL). The primer efficiency of 90-110% was accepted in the qPCR analysis, and the required efficiency was observed for all the primers to the target genes ( Figure 1).
Melting curve analysis was used at the end of qPCR products amplification to determine the specificity of each primer set for target genes in E. coli IMP and K. pneumoniae NDM-1. For both strains, it produced one peak and one product for each tested primer pair. The endpoint qPCR products visualised by agarose gel electrophoresis showed that all primer pairs resulted in amplification of a single product of the selected target gene (see Figures 9 and 10 for E. coli IMP and K. pneumoniae NDM-1, respectively).   Figure 11A,B and Figure 12 represents the relative changes in the expression levels of the selected gens of E. coli IMP, at different time points (12 h, 24 h and 48 h) of biofilm formation, in the two different growth media and indicates as fold expression values. A fluctuation of gene expression levels was observed over time, where initial upregulation was at the early adherence phase (12 h), and maximum upregulation at the mid-adherence phase (24 h), followed by a slight downregulation at the biofilm dispersion phase (48 h). This pattern of gene regulation was observed in both growth media, and all the gene categories (biofilms, quorum sensing, stress survival, and antibiotic-resistant genes), in E. coli IMP (Figure 1). In all genes tested, in both media, quorum sensing genes, such as luxS, were highly upregulated at 24 h; similarly, the highest upregulation of the antibioticresistant genes (marA and bla IMP ) were also observed at 24 h. All the categories of genes were highly upregulated in the nutritionally nutrient-rich LB media compared to the nutrient-poor AB media.

Expression Levels of Selected Antibiotic-Resistant and Biofilm-Associated Genes in K. pneumoniae NDM-1
The relative changes in the expression levels of the selected genes at each different time point (12 h, 24 h, and 48 h) of biofilm formation, and in the two different growth media are represented as fold expression values for K. pneumoniae NDM-1 and are represented in Figures 13 and 14. Similar to E. coli IMP, a fluctuation of gene expression levels was observed over time in K. pneumoniae NDM-1. An initial upregulation was observed in the early adherence phase (12 h), and maximum upregulation at the mid-adherence phase (24 h), followed by a slight downregulation at the biofilm dispersion phase (48 h). This pattern of gene regulation was observed in both growth media, and for all the gene categories (biofilms, quorum sensing, stress survival, and antibiotic resistance genes; Figure 1). In all genes tested, in both media, quorum sensing genes such as intl and luxS were highly upregulated at 24 h. Similarly, the highest upregulation of the antibiotic resistance gene, bla NDM , was also observed at 24 h. All the categories of genes were highly upregulated in K. pneumoniae NDM-1 in the nutritionally rich LB media compared to the nutrient-poor AB media, in a similar manner to E. coli IMP (Figure 12).

Discussion
Biofilm formation is a significant characteristic of Enterobacterales including Escherichia coli and Klebsiella pneumoniae [35]. Growth in biofilms promotes the survival of bacterial inhabitants in hospital settings and internal patients, rising the probability of occasional nosocomial infections [36][37][38]. Multidrug-resistant CRE, such as E. coli IMP and K. pneumoniae NDM-1, have been reported as a significant health challenge as they are associated with common community and hospital-acquired infections, and indeed CRE outbreaks have already been reported worldwide [21][22][23][24][25][26][27][28][29][30][31][32][33][34][35][36][37][38][39]. The clinical risk factors for bacterial infections due to CRE were reported as being well captured; however, bacterial risk factors such as antibiotic susceptibility, resistant mechanisms, and biofilm formation, for those infections due to CRE are limited. Hence, understanding the bacterial characteristics, such as biofilms is of utmost importance to the treatment of the biofilm-associated infections caused by CRE pathogens due to the associated poor clinical outcomes and limited therapeutic options associated with these pathogens.
Here, we observed the potential relationship between biofilm formation and the existence of antimicrobial resistance genes. Strong biofilm formation was observed for both E. coli IMP and K. pneumoniae NDM-1, which carry the bla IMP and bla NDM-1 resistance genes, compared to the non-resistant control strains. This rate of stronger biofilm formation was similarly observed in other studies of multidrug-resistant Enterobacteriaceae [40]. In addition to strong biofilm formation in E. coli IMP and K. pneumoniae NDM-1, the greatest upregulation of the resistance genes, bla IMP and bla NDM-1 at 24 h, where optimal biofilm formation was observed. This could be due to the fact the ability of these resistance genes to influence the process of biofilm formation, and that has been demonstrated that resistance genes in special plasmids can regulate biofilm formation in Klebsiella pneumoniae [28].
The effect of aeration and distribution of nutrients on the biofilm formation process were also explored in this study by incubating test strains under shaking and static conditions. Our data revealed that both CRE strains were high biofilm formers when grown under static incubation as indicated in Figure 5, and that characteristic could be attributed to their preferred physiological growth condition of facultative anaerobes [41].
Both CRE strains' biofilm analysis and CLSM revealed a varying amounts of biofilms correlating with the biofilm development stages, where the 6 h stage corresponded to attachment, the 12 h stage to exopolysaccharides secretion and microcolony formation, and the 24 h stage to maturation and stabilisation of a three-dimensional biofilm structure, with detachment and dispersion on the biofilm observed at the 48 h time point. This biofilm development was in accordance with the previously demonstrated studies on Gram-negative bacteria [42,43]. Moreover, both CRE strains demonstrated the maximum amount of biofilm at the biofilm maturation stage (24 h of incubation), as similarly observed in multidrug-resistant P. aeruginosa isolates [40]) and drug-resistant E. coli isolates [24].
In E. coli IMP, biofilm-related genes csgA and mqsR showed the highest upregulation at the biofilm maturation stage (24 h) to their role in the mechanism of biofilm formation [44,45]. csgA is part of the curli-specific csgBAC operon, which plays a role in curli-dependent biofilm development via encoding the protein Curli [44]. Their role of Curli in surface attachment and cell-to-cell contact makes them the main part of the biofilm matrices and a contributing factor to the biofilm development mechanism, for which it has been demonstrated that infection-associated bacteria are capable of developing curli-dependent biofilm structures [46][47][48]. Upregulation of the mqsR gene at the biofilm maturation stage could be attributed to its function in motility and biofilm formation, as has previously been described for E. coli [45].
For K. pneumoniae NDM-1, biofilm-related genes bssS, and also fimH, showed the highest upregulation at the biofilm maturation stage (24 h) to their role in the mechanism of biofilm formation [49,50]. bssS encodes a small cytoplasmic protein that is a significant regulator in biofilm formation via signal secretion [50]. fimH encodes for one of the most important virulence factors, fimbriae, which is important for attachment and adhesion and is also required for colonisation and biofilm formation [51].
The upregulation of the stress-responsive gene soxS in E. coli IMP is attributed to its response to the growth condition-related role in protecting bacteria from oxidative stress [23]. The presence of nitric oxide or superoxide-generating factors in growth conditions, such as controlled by locus soxRS. soxR senses the intracellular redox signals of these factors and acts as a potent activator of soxS transcription when being activated [52]. The evgA, another stress-responsive gene, encodes for proteins involved in bacterial signalling pathways that allow for a diversity of adaptive responses to their environment. These adaptive responses are often interceded by two-component regulatory systems which generally depend on a prospector kinase and a response regulator [53]. Collectively, both soxS and evgA play integral roles in biofilm development stages by adapting to different environmental stress responses and contributing to the virulence fitness of E. coli [54,55].
In K. pneumoniae NDM-1 the upregulating stress-responsive entB and ycfM could be implicated in the role of iron uptake in different growth/iron starvation conditions. This behaviour of upregulation of entB and ycfM genes in response to iron starvation was also observed in other parallel studies in poor nutrient and nutrient-rich environments [56,57].
For E. coli IMP, we observed maximum upregulation of the QS genes, tnaA and luxS, at the maturation stage of the biofilm development (24 h), which were highly upregulated irrespective of the growth medium. This finding is coincident with the observation that tnaA and luxS expression influences the microorganisms' adaption to diverse environmental conditions, and also influences the virulence of biofilm formation [17]. Interestingly we have observed the most extensive upregulation of tnaA is in nutrient-rich media (LB) which contradicts the previous findings on the induction of quorum-sensing genes in nutrient-poor conditions [58].
In K. pneumoniae NDM-1, we observed high upregulation of the QS genes, intl and luxS, at the biofilm maturation stage (24 h in all growth conditions tested), which is following a recent study in an extensive drug-resistant K. pneumoniae clinical isolate where luxS gene expression played a significant role in biofilm formation, polysaccharide production, and metabolic pathways, and was influenced by nutrient availability and other environmental conditions [59].
The high-level upregulation of bla IMP , encoding for carbapenem resistance, could be attributed to the hypervirulence/resistance capability that this strain process at the biofilm maturation stage, which could correlate with its successful dissemination in many common infections [60]. The upregulation of antibiotic-resistant gene marA can also be attributed to its role in transcription activation as a part of the marA/soxS/rob regulon, where it regulates more than 40 genes of the E. coli chromosome, resulting in different levels of resistance to a wide array of antibiotics and superoxide's [61]. Moreover, marA promotes the outflux and inhibition of the influx of antibiotics, giving rise to multiple antibioticresistant phenotypes [62].
Similarly, for K. pneumonia NDM-1, there is the highest upregulation of the carbapenemresistant bla NDM gene and transcription activator ramA at the mature biofilm development stage (24 h). The high-level upregulation of bla NDM-1 can be attributed to the hypervirulence/resistance that this strain process shows at the biofilm maturation stage. Collectively, the variation in the antibiotic-resistant genes tested in these strains could be attributed to the fact of their multi-drug resistant ability, as these strains harbour varieties of resistant determinants in their resistant plasmids that can easily conjugate with other bacteria, resulting in broad resistance to antibiotics [63]. The bla NDM gene is generally accepted by conjugative plasmids, resulting in multiple additional determinants and leading to multidrug resistance [64][65][66]. ramA code for ramA is an actual transcription manager in K. pneumoniae, which has been established to be significant to the bacterial response to antimicrobial challenges and wider the regulatory role in K. pneumonia during infection microbe-host or microbe-drug interactions [67][68][69][70].
In this study, we considered only a cluster of genes encoding for antibiotic resistance (bla IMP bla NDM-1 , marA, and ramA), genes encoding for biofilm maturation (csgA, mqsR, bssS, and fimH), genes encoding stress response (soxS, evgA, entB, and ycfM), and QS gene luxS, and their expression during the biofilm development stages. The formation of the biofilm confined into a selected cluster of genes on two selected resistant bacteria is a limitation of our study. Although present findings from our study indicated the potential role of the subset of genes selected on the biofilm formation, but performing additional studies, such as correlating the biofilm magnitude with the selected resistant determinants in CRE to understand the physiological impact of the biofilms is caused by differentiating gene expression, would allow us to understand the role of resistant genes as indicators for biofilm formation. This may help to better understand the molecular mechanism of biofilm formation over time. Our results showed that in both CRE strains, the gene expression is correlated with the biofilm development stages. The highest upregulation of all the gene clusters selected was observed at the biofilm maturation stage (24 h time point) and down-regulation was observed at the detachment stage (48 h time point) under all growth conditions tested. Therefore, the most interesting point in this study is the morphological imaging by CLSM coupled to qPCR which allowed correlating morphological modifications to gene expression profile.

Conclusions
This study evaluated the detailed gene expression analysis considering a cluster of genes that linked with the biofilm development in carbapenem-resistant enteric Gramnegative bacilli, E coli IMP and K pneumoniae NDM-1 strains. We observed that genes encoding for biofilm development are upregulated during biofilm growth in E coli IMP and K pneumoniae NDM-1 strains. Our data showed that biofilm growth, among the studied variables, is the major factor influencing the variability of biofilm gene expression quantification assays. CLSM combined with the qPCR has demonstrated that phenotypic biofilm development is correlated with the expression profile of the selected cluster of genes of biofilm, quorum sensing, stress survival, and antibiotic resistance. Therefore, this morphological analysis combined with the gene expression during the biofilm growth can be used to study variation in the regulation of such genes in these significant CRE pathogens. Considering our data and the published literature, it is thought that these CRE strains, E. coli IMP and K. pneumoniae NDM-1, with high biofilm-forming capacity in many growing conditions, indicate their ability to develop resistance to antibiotics and which is considered to be the major risk in terms of successful dissemination and risk in terms of the emergence of multidrug resistance among these CRE pathogens.