Abstract

   Despite the widespread use of antiseptics such as chlorhexidine
   digluconate (CHX) in dental practice and oral care, the risks of
   potential resistance toward these antimicrobial compounds in oral
   bacteria have only been highlighted very recently. Since the molecular
   mechanisms behind antiseptic resistance or adaptation are not entirely
   clear and the bacterial stress response has not been investigated
   systematically so far, the aim of the present study was to investigate
   the transcriptomic stress response in Streptococcus mutans after
   treatment with CHX using RNA sequencing (RNA-seq). Planktonic cultures
   of stationary-phase S. mutans were treated with a sublethal dose of CHX
   (125 µg/mL) for 5 min. After treatment, RNA was extracted, and RNA-seq
   was performed on an Illumina NextSeq 500. Differentially expressed
   genes were analyzed and validated by qRT-PCR. Analysis of differential
   gene expression following pathway analysis revealed a considerable
   number of genes and pathways significantly up- or downregulated in S.
   mutans after sublethal treatment with CHX. In summary, the expression
   of 404 genes was upregulated, and that of 271 genes was downregulated
   after sublethal CHX treatment. Analysis of differentially expressed
   genes and significantly regulated pathways showed regulation of genes
   involved in purine nucleotide synthesis, biofilm formation, transport
   systems and stress responses. In conclusion, the results show a
   transcriptomic stress response in S. mutans upon exposure to CHX and
   offer insight into potential mechanisms that may result in development
   of resistances.

   Keywords: antibacterial, CHX, RNA-seq, Streptococcus mutans, stress
   response, transcriptomics

1. Introduction

   Despite the current COVID-19 pandemic, antimicrobial resistance (AMR)
   remains one of the greatest challenges for public health in the 21st
   century [[44]1,[45]2]. For instance, a 2018 report of the Organization
   for Economic Co-operation and Development (OECD) predicted that 2.4
   million people in Europe, North America and Australia will die from
   infections associated with AMR within the next 30 years, resulting in
   costs of up to USD 3.5 billion per year for the healthcare services of
   the 33 countries included in the analysis [[46]3]. Despite many reports
   about antibiotic resistance, antiseptics or biocides have only recently
   come into spotlight in the context of AMR [[47]4,[48]5,[49]6,[50]7].

   Chlorhexidine is a twofold positively charged bis-biguanide and is
   mostly used as its digluconate salt (chlorhexidine digluconate, CHX) in
   clinical practice [[51]4,[52]8]. Since its introduction into dentistry
   in the late 1960s [[53]9], CHX has come to be considered the
   gold-standard antiseptic in oral care [[54]4,[55]10]. It is used for
   plaque control and management of gingivitis [[56]11]; in patients with
   high caries risk, such as those with fixed orthodontic appliances
   [[57]12]; or in patients after periodontal or implant surgery [[58]13].
   Furthermore, CHX is included in a wide range of over-the-counter oral
   care products, such as toothpastes or mouthwashes
   [[59]11,[60]14,[61]15].

   Although there is ample evidence that rinsing with CHX can decrease the
   salivary bacterial load [[62]16] and consequently reduce oral biofilm
   formation [[63]17,[64]18], the antibacterial efficacy of CHX toward
   mature oral biofilms may be limited [[65]4,[66]19,[67]20]. For
   instance, we recently showed that microcosm biofilms from human saliva
   cultured for 72 h and treated with 0.1% or 0.2% CHX for 1 min resulted
   in colony-forming-unit reduction rates of less than 1 log[10] step
   [[68]20]. Accordingly, treatment with 0.2% CHX for 1 min only affected
   the outer layers of biofilms formed in situ for 48 h [[69]19]. This
   limited efficacy of CHX in mature oral biofilms may be mainly due to
   the biofilm matrix, which acts as a diffusion barrier for positively
   charged molecules, such as CHX [[70]21]. Consequently, it is reasonable
   that bacteria in deeper layers will be exposed to sublethal
   concentrations of CHX upon application of a CHX-containing mouthwash
   [[71]4,[72]20]. Several studies from recent years showed that oral, as
   well as non-oral, bacteria can phenotypically adapt upon multiple
   exposure to sublethal concentrations in vitro
   [[73]20,[74]22,[75]23,[76]24,[77]25] and potentially also develop cross
   resistances toward antibiotics [[78]23,[79]25]. However, the actual
   molecular mechanisms causing these adaptations, as well as the stress
   response in bacteria upon exposure toward sublethal concentrations of
   CHX, are not well understood yet [[80]15].

   Therefore, the aim of the present study was to investigate the
   transcriptomic stress response after sublethal treatment with CHX using
   an RNA sequencing approach with Streptococcus mutans as model organism
   due to its key role in biofilm formation and pathogenesis of dental
   caries [[81]26,[82]27].

2. Materials and Methods

2.1. Chlorhexidine Digluconate

   Chlorhexidine digluconate (CHX) was obtained from Sigma-Aldrich (Merck,
   Darmstadt, Germany) in a concentration of 20% and was dissolved in
   distilled water to yield the final concentration of 125 µg/mL
   (0.0125%).

2.2. Bacterial Culture and Treatment

   A-type strain of S. mutans (DSM 20523; ATCC 25175) was obtained from
   the German Collection of Microorganisms and Cell Cultures GmbH (DSMZ;
   Braunschweig, Germany) to be used in this study. S. mutans was grown in
   brain heart infusion broth (BHI; Sigma-Aldrich, St. Louis, MO, USA) and
   on BHI agar plates. Planktonic cultures (5 mL) were grown aerobically
   overnight at 37 °C. Afterwards, suspensions were harvested by
   centrifugation (ROTINA 420 R, Hettich Lab Technology, Tuttlingen,
   Germany) and resuspended in BHI, yielding an optical density (OD) of
   2.0 per mL, as measured by means of a spectrophotometer at 600 nm
   (Ultrospec 3300 pro; Amersham Biosciences, Amersham, UK). Samples were
   centrifuged, supernatants were discarded and bacterial pellets were
   resuspended and incubated with 200 μL BHI broth or 200 µL CHX (final
   concentration: 125 µg/mL) for 5 min.

   To evaluate the bacterial ability to replicate, samples were
   centrifuged, BHI broth or CHX solution was carefully removed, and the
   bacterial pellet was brought to suspension with 200 μL
   phosphate-buffered saline (PBS; Biochrom, Berlin, Germany) by frequent
   pipetting. Tenfold serial dilutions (10^−2 to 10^−7) were prepared in
   PBS, and aliquots (20 μL) were plated on BHI agar (both provided by the
   Institute of Medical Microbiology and Hygiene, University Hospital
   Regensburg, Germany) according to the method described by Miles et al.
   [[83]28]. Samples were incubated aerobically for 72 h, and subsequently
   colony forming units (CFUs) were evaluated (n = 5).

   For RNA-Seq, samples were centrifuged after treatment, and distilled
   water or CHX solution was carefully removed. Pellets were resuspended
   in 500 μL prewarmed BHI broth and incubated at 37 °C for 30 min.
   Afterwards, 1 mL of RNAprotect bacteria reagent (Qiagen, Hilden,
   Germany) was added to the samples and incubated for 5 min at room
   temperature. Then, samples were centrifuged at 17,035× g for 5 min,
   supernatants were discarded and pellets were stored at −80 °C until
   further use (n = 5).

2.3. RNA Extraction, Library Preparation and RNA Sequencing

   RNA extraction and sequencing were performed by the NGS Competence
   Center Tübingen (NCCT), the sequencing partner of the Quantitative
   Biology Center (QBiC), Tübingen, Germany. RNA of all samples was
   isolated with the Quick-RNA fungal/bacterial miniprep kit (Zymo
   Research, Freiburg, Germany). RNA extraction was followed by DNase I
   treatment in solution using the DNase I recombinant kit (Roche, Basel,
   Switzerland), as well as purification and concentration of RNA using
   the RNA Clean & Concentrator-5 kit (Zymo Research, Freiburg, Germany).

   Library preparation was performed using an Illumina Stranded Total RNA
   Prep Ligation with the Ribo-Zero Plus kit (Illumina, San Diego, CA,
   USA) with 100 ng total RNA. Sequencing was performed on an Illumina
   NextSeq 500 using a high-output kit (version 2.5; 75 cycles) with 1%
   PhiX spike-in (Illumina, San Diego, CA, USA).

2.4. RNA Sequencing Data Analysis

   Data management and bioinformatic analysis were performed at the
   Quantitative Biology Center (QBiC), Tübingen, Germany. A Nextflow-based
   nf-core pipeline nf-core/rnaseq (version 1.4.2;
   [84]https://github.com/nf-core/rnaseq, accessed on 22 January 2022) was
   used for the RNA-seq bioinformatic analysis. As part of this workflow,
   FastQC (version v0.11.8) was used to determine the quality of the FASTQ
   files [[85]29]. Subsequently, adapter trimming was conducted with Trim
   Galore (version 0.6.4) [[86]30]. STAR aligner (version 2.6.1, [[87]31])
   was used to map the reads that passed quality control against the
   latest assembly (RefSeq assembly accession GCF_000347875.1) of strain
   ATCC 25175, downloaded from NCBI. Note that the gtf file for the S.
   mutans genome had to be modified to be used for the nf-core/rnaseq
   workflow by removing empty fields for the tag “transcript_id” in the
   ninth column of that gtf file. RNA-seq data quality control was
   performed with RSeQC (version 3.0.1) [[88]32] and read counting of the
   features (e.g., genes) with featureCounts (version 1.6.4) [[89]33]. An
   aggregation of the quality control for the RNA-seq analysis was
   performed with MultiQC (version 1.7; [90]http://multiqc.info/, accessed
   on 22 January 2022) [[91]34].

   The analysis of the differential gene expression was performed in R
   language (version 3.5.1) using DESeq2 (version 1.22) through the
   Nextflow-based workflow qbic-pipelines/rnadeseq
   ([92]https://github.com/qbic-pipelines/rnadeseq, accessed on 22 January
   2022, version 1.3.2). Genes were considered differentially expressed
   (DE) when the Benjamini–Hochberg multiple testing adjusted p-value
   [[93]35] was smaller than 0.05 (p[adj] ≤ 0.05). Multiple testing
   correction helps to reduce the number of false positives (not real DE
   genes). In the case of a threshold of 0.05, the proportion of false
   discoveries in the selected group of DE genes is controlled to be less
   than the threshold value—in this case, 5%. Genes were further filtered
   for biological relevance if the log[2] fold change (FC) in expression
   between the two considered groups was above the threshold of 1.0 and
   less than −1.0. Final reports were produced using the R package
   rmarkdown (version 2.1) with the knitr (version 1.28) and DT (version
   0.13) R packages. The sample similarity heatmaps were created using the
   edgeR (version 3.26.5) R package. BioCyc database was used for pathway
   analysis [[94]36]. All DE genes were included, and enrichment was
   calculated using Fisher’s exact test (p ≤ 0.05).

2.5. Primer Design and qRT-PCR

   To validate RNA-Seq data, randomly selected genes were measured by
   qRT-PCR. The nucleotide sequence of each gene ([95]Table 1) was
   downloaded from BioCyc [[96]36]. Primers were designed using the
   Primer3 tool (version 4.1.0, Whitehead Institute for Biomedical
   Research, Cambridge, MA, USA) and synthesized by Eurofins MWG Synthesis
   (Ebersberg, Germany). Primer efficiencies were checked by qRT-PCR using
   different cDNA dilutions. Primers gave a single PCR product, which was
   verified by gel electrophoresis and melt curves at the end of each run.
   Total RNA (200 ng) used for RNA-Seq was reverse-transcribed with the
   Quantitect reverse transcriptase kit (Qiagen, Hilden, Germany) in a
   reaction volume of 20 μL. Polymerase chain reaction was carried out
   with 20 pmol of each primer, 5 µL cDNA (0.2 ng) diluted and qPCR master
   mix with SYBR^®Green in a final volume of 20 µL (Applied Biosystems^TM
   SYBR^®Select Master Mix, Thermo Fisher Scientific, Waltham, MA, USA) on
   the QuantStudio 3 real-time PCR system (Thermo Fisher Scientific,
   Waltham, MA, USA) following an initial denaturation of the samples at
   95 °C for 10 min and 40 cycles of alternating denaturation (95 °C for 1
   s), annealing and elongation (60 °C for 20 s).

Table 1.

   Primer sequences for qRT-PCR.
   Target Gene Sequence 5′-3′
   (F = Forward; R = Reverse) Product Size
   gyrB F: GCACAAGAGTACGATGCCAGT 119 bp
   R: TCCCAAACAAGGTGATGCAGC
   D820_RS03005 F: CGTGGTTATCAAGTATCGTGTGA 148 bp
   R: AAAGAATTGGTCCTGCATCCA
   D820_RS09005 F: CAGTAGGTGCCGCTCAAACT 128 bp
   R: AAGTCCGCCGCCAAACATAT
   glgA F: GTGCCTTGCCCAAATCCCTT 145 bp
   R: ACATATTCACGACGCCAGCC
   ssrS F: CGGAAGCAACTAAAGTCAGAGCG 80 bp
   R: TGGCACCGATGATTCACGTT
   [97]Open in a new tab

   Gene expression was quantified by the comparative threshold cycle (Ct)
   method (2^−∆∆Ct) [[98]37]. First, Ct values of all expressed genes were
   normalized by the housekeeping gene gyrB. Furthermore, gene expression
   of treated cells was calculated relative to the untreated control cells
   (log[2] fold change). Median values and quantiles of all experiments
   were calculated and compared to log[2] fold changes obtained from
   RNA-seq (n = 5).

3. Results

3.1. Sublethal Concentration of CHX

   The bacterial ability to replicate was investigated for planktonic
   cultures of S. mutans after treatment with 125 µg/mL CHX by a CFU
   assay. The chosen concentration of 125 µg/mL was determined by
   experiments screening multiple concentrations of CHX (data not shown).
   For the purpose of this study, sublethal doses of CHX were defined as
   the treatment that resulted in a < 0.5 log[10] step reduction in CFU.
   Untreated controls exhibited 1.0 × 10^7 to 1.4 × 10^7 CFU. Samples
   treated with 125 µg/mL CHX for 5 min showed 4.5 × 10^6 to 8.5 × 10^6
   CFU, resulting in a reduction in CFU of 0.2 log[10] steps, as compared
   to the untreated control.

3.2. RNA-Seq and Analysis of Differentially Expressed Genes

   RNA-seq was performed to investigate the molecular stress response of
   S. mutans upon sublethal treatment with CHX. Five independent
   experiments were performed on CHX-treated and -untreated bacteria.
   Principal component analysis (PCA) indicated that all biological
   replicates grouped together, suggesting that gene expression was
   significantly influenced by the treatment with CHX compared to
   untreated samples, explaining 89% of the variance observed in the
   complete dataset ([99]Figure 1). The distribution of gene expression
   between the untreated and treated samples is represented by a volcano
   plot ([100]Figure 2). The genes located outside the borders of the line
   fold change ≥ 1.0 were considered DE genes. The number of DE genes with
   an adjusted p-value ≤ 0.05 was 718. Genes situated on the left boundary
   are downregulated, whereas those on the right are upregulated genes. As
   compared to the untreated control, 404 genes were upregulated and 271
   genes were downregulated in S. mutans after treatment with 125 µg/mL
   CHX for 5 min. A list of all genes detected by RNA-seq can be found in
   the [101]Supplementary Material (Table S1).

Figure 1.

   [102]Figure 1
   [103]Open in a new tab

   Principal component analysis (PCA) of gene expression in S. mutans
   untreated (UC, orange) or treated with CHX (turquoise). The plot shows
   the first two principal components (PC1 and PC2), which account for 79%
   and 10% of the total variation of the data, respectively. UC: untreated
   control.

Figure 2.

   [104]Figure 2
   [105]Open in a new tab

   Volcano plot of differential gene expression of S. mutans treated with
   CHX versus untreated bacteria. Each point represents the average value
   of one transcript in five replicate experiments. The expression
   difference is considered significant for a multiple (FDR-based)
   adjusted p-value of 0.05 (light grey broken horizontal line). The list
   of DE genes was then further filtered for biological relevance by
   filtering on a log[2] fold change of ≥1 and ≤−1 (red points, outer blue
   broken vertical lines). The top 20 DE genes based on log[2] fold change
   are labelled.

   [106]Table 2 and [107]Figure 2 show the 10 most up- and 10 most
   downregulated genes. DE genes were selected by their log[2] fold
   change. It was observed that CHX treatment significantly regulated
   expression of genes involved in transport systems (ABC transporter and
   PTS system). [108]Table 3 shows selected DE genes involved in oxidative
   stress (ahpC, ahpF, sod, trxP and tpx), general stress (groEL, groES
   and D820_RS07050) and acid stress response (aguA and mleP), which were
   significantly upregulated.

Table 2.

   The 10 most up- and downregulated genes in S. mutans upon sublethal CHX
   treatment. The full list of DE genes can be found in the
   [109]Supplemental Table S1.
   Gene ID Gene Name Product Log[2] FC p [adj]
   D820_RS03005 NA PTS fructose transporter
   subunit IIB −3.2 1.60 × 10^−6
   D820_RS01705 NA tRNA-Thr −3.1 2.46 × 10^−4
   D820_RS03010 NA PTS mannitol/fructose,
   IIC component −3.0 2.26 × 10^−5
   D820_RS03000 NA PTS mannitol transporter
   subunit IIB −3.0 6.57 × 10^−6
   D820_RS02990 lacC Tagatose-6-phosphate kinase −3.0 2.40 × 10^−5
   D820_RS02995 lacD Tagatose-bisphosphate aldolase −2.9 6.31 × 10^−6
   D820_RS08540 NA tRNA-Ser −2.8 6.90 × 10^−3
   D820_RS02980 lacA Galactose-6-phosphate isomerase −2.7 1.24 × 10^−4
   D820_RS02985 lacB Galactose-6-phosphate isomerase −2.7 1.43 × 10^−4
   D820_RS03015 NA Lactose-specific phosphotransferase enzyme
   IIA component −2.6 2.85 × 10^−8
   D820_RS02805 glgA Glycogen synthase 3.9 0.00
   D820_RS04685 NA ABC transporter
   (ATP-binding protein) 4.0 0.00
   D820_RS06900 NA Hypothetical protein 4.0 0.00
   D820_RS01630 NA 16S ribosomal RNA 4.1 2.44 × 10^−7
   D820_RS02745 NA ABC transporter
   (ATP-binding protein) 4.1 0.00
   D820_RS02750 NA ABC transporter permease 4.2 0.00
   D820_RS06905 lrgB Antiholin 4.4 0.00
   D820_RS07700 pflB Formate C-acetyltransferase 4.5 0.00
   D820_RS0109785 NA 23S ribosomal RNA 5.0 1.79 × 10^−6
   D820_RS0109610 NA 23S ribosomal RNA 6.5 5.73 × 10^−6
   [110]Open in a new tab

   FC: fold change; p[adj]: adjusted p-value; NA: not applicable.

Table 3.

   Differentially expressed genes in S. mutans upon CHX treatment related
   to stress response.
   Gene ID Gene Name Product Log[2] FC p [adj]
   D820_RS06120 ahpC Peroxiredoxin 1.1 1.50 × 10^−6
   D820_RS06115 ahpF Alkyl hydroperoxide reductase subunit F 1.8 6.00 ×
   10^−17
   D820_RS06685 sod Superoxide dismutase 1.6 1.71 × 10^−6
   D820_RS07435 trxB Thioredoxin disulfide reductase 1.1 5.15 × 10^−4
   D820_RS05415 tpx 2-Cys-peroxiredoxine 1.0 1.02 × 10^−2
   D820_RS07050 NA DNA protection protein 1.9 2.41 × 10^−12
   D820_RS00990 groES Co-Chaperone GroES 1.1 8.15 × 10^−6
   D820_RS00995 groEL Chaperone GroEL 1.5 0.00
   D820_RS05270 NA ATP-dependent Clp protease ATP-binding subunit 1.1 1.04
   × 10^−7
   D820_RS03310 clpB Clp proteinase ATP-binding subunit ClpB 2.4 4.29 ×
   10^−15
   D820_RS06955 NA ATP-dependent Clp protease ATP-binding subunit 1.4 3.43
   × 10^−10
   D820_RS08310 aguA Agmatine deiminase 1.3 1.27 × 10^−14
   D820_RS08960 mleP Malate permease 3.7 0.00
   [111]Open in a new tab

   FC: fold change; p[adj]: adjusted p-value; NA: not applicable.

3.3. Pathway Enrichment Analysis with Differentially Expressed Genes

   Pathway enrichment analysis was performed using the BioCyc database,
   including all upregulated DE genes or all downregulated DE genes after
   CHX treatment. Significantly enriched pathways that were upregulated
   were composed of genes involved in carboxylate degradation, glycan
   biosynthesis, purine nucleotide biosynthesis, L-histidine biosynthesis
   and L-ascorbate degradation ([112]Table 4). Lactose and galactose
   degradation, protein modification and L-phenylalanine biosynthesis were
   found to be downregulated after CHX treatment ([113]Table 4).

Table 4.

   Significantly up- and downregulated pathways in S. mutans upon CHX
   treatment.
   Pathway Genes Involved in Pathway
   (Ensemble IDs) p-Value
   Upregulated
   Carboxylate degradation D820_RS00885, D820_RS07700, D820_RS07305,
   D820_RS08215, D820_RS08265, D820_RS08275, D820_RS08270, D820_RS04975,
   D820_RS04980, D820_RS08965 5.00 × 10^−5
   5-Aminoimidazole ribonucleotide biosynthesis D820_RS09340,
   D820_RS09355, D820_RS09335, D820_RS09300, D820_RS09345 2.00 × 10^−3
   Glycan pathway D820_RS07040, D820_RS02810, D820_RS02685, D820_RS02800,
   D820_RS02795, D820_RS02790, D820_RS02805 3.00 × 10^−3
   L-ascorbate degradation D820_RS08215, D820_RS08265, D820_RS08275,
   D820_RS08270 6.00 × 10^−3
   L-histidine biosynthesis D820_RS03965, D820_RS03970, D820_RS03955,
   D820_RS03950, D820_RS03960, D820_RS03940, D820_RS03930 7.00 × 10^−3
   Purine nucleotide biosynthesis D820_RS04105, D820_RS04110,
   D820_RS04100, D820_RS09285, D820_RS09290, D820_RS09325, D820_RS09250,
   D820_RS08295, D820_RS09340, D820_RS09355, D820_RS09335, D820_RS09300,
   D820_RS09345 1.00 × 10^−2
   Inosine-5’-phosphate biosynthesis D820_RS09285, D820_RS09290,
   D820_RS09325, D820_RS09250 2.00 × 10^−2
   Downregulated
   Lactose degradation D820_RS02995, D820_RS02980, D820_RS02985,
   D820_RS03025, D820_RS02990 4.00 × 10^−6
   Galactose degradation D820_RS02995, D820_RS02980, D820_RS02985,
   D820_RS03025, D820_RS02990 4.00 × 10^−6
   Protein modification D820_RS05715, D820_RS02255, D820_RS08010 2.00 ×
   10^−2
   L-phenylalanine biosynthesis D820_RS03740, D820_RS07095 2.00 × 10^−2
   [114]Open in a new tab

3.4. Validation of RNA-Seq Data Using qRT-PCR

   To validate RNA-Seq data, randomly selected genes were measured by
   qRT-PCR to evaluate the transcription levels. Overall, the expression
   levels of all selected genes were similar to those of the RNA-seq, with
   a Spearman correlation coefficient of 0.8. The results showed that the
   D820_RS03003 expression levels were downregulated in CHX-treated
   bacteria compared to the untreated control ([115]Table 5). In
   comparison, expression levels were upregulated in D820_RS09005, ssrA
   and glgA in CHX-treated bacteria compared to the untreated control
   ([116]Table 5), which is in line with the DE results from the RNA-Seq
   experiment.

Table 5.

   Validation of differentially expressed genes using qRT-PCR. Transcript
   levels of selected genes ([117]Table 1) were corrected to gyrB. Each
   value (log[2] fold change) is the median of five replicate PCR
   reactions.
       Gene     RNA-Seq qRT-PCR
   D820_RS03005  −3.2    −2.4
   D820_RS09005   3.3     3.4
      ssrS        3.4     2.5
      glgA        3.9     4.0
   [118]Open in a new tab

4. Discussion

   Despite the widespread use of antiseptics such as CHX in dental
   practice and oral care, the risks of potential resistance toward these
   antimicrobial compounds in oral bacteria have only been highlighted
   very recently [[119]4,[120]5,[121]20,[122]22,[123]24]. Since the
   molecular mechanisms behind antiseptic resistance or adaptation are not
   entirely clear yet and the bacterial stress response upon sublethal
   exposure toward CHX has not been investigated systematically so far
   [[124]4], the aim of the present study was to investigate the
   transcriptomic stress response in S. mutans after sublethal treatment
   with CHX.

   S. mutans was chosen as a model organism for the present study due to
   its key role with regard to biofilm formation and pathogenesis of
   dental caries [[125]26,[126]27]. A type strain was used in order to
   ensure that the transcriptomic stress response of a sublethal treatment
   with CHX was investigated on a strain not pre-adapted to CHX. For this
   purpose, it was necessary to choose “sublethal” treatment stress
   conditions but not kill the bacteria in order to allow for
   transcriptomic adaptations. Thus, sublethal conditions were defined as
   conditions that result in a reduction in bacterial ability to replicate
   in <0.5 log[10] step as measured by CFU, which means that less than 50%
   of the bacteria are influenced in their ability to replicate. Bacteria
   were treated with CHX at a concentration of 125 µg/mL for 5 min, which
   resulted in a CFU reduction of 0.2 log[10] steps. After treatment,
   bacteria were cultured in fresh nutrient broth for 30 min to allow for
   regulation of transcriptomic stress response. Afterwards, RNA was
   extracted and used for RNA-seq. RNA-seq data were validated using
   qRT-PCR to evaluate the transcription levels of randomly selected
   genes.

   Sublethal treatment of S. mutans with CHX showed a strong treatment
   effect that led to a clear separation of samples into the two
   experimental groups, thereby explaining 89% of the variance in the
   dataset. The reason for this is most likely the differential regulation
   of gene expression, with 404 genes upregulated and 271 genes
   downregulated. Further analysis of DE genes and significantly regulated
   pathways showed regulation of genes involved in purine nucleotide
   synthesis, biofilm formation, transport systems and stress responses
   ([127]Figure 3).

Figure 3.

   [128]Figure 3
   [129]Open in a new tab

   Schematic representation of a possible stress response in S. mutans to
   CHX treatment found in this study. Shown are the mechanisms described
   in the discussion in response to CHX treatment. Green: upregulated;
   red: downregulated; purple: up- and downregulated. Black arrows
   represent the links between different mechanisms. ROS: reactive oxygen
   species; PTS: phosphotransferase system; ABC: ATP-binding cassette;
   eDNA: extracellular DNA.

   Amongst the DE genes and significantly regulated pathways, an
   upregulation in the expression of genes involved in purine nucleotide
   synthesis was detected. For the bacterial cell, purines primarily
   contribute to cell division and synthesis of necessary energy and
   cofactors [[130]38]. For biofilm formation, purines are required for
   the production of extracellular DNA [[131]39,[132]40], as well as
   synthesis of secondary messengers from guanosine monophosphate (GMP)
   and adenosine monophosphate (AMP), which also play a role in the
   overall stress response [[133]41]. Purine nucleotide biosynthesis has
   been shown to be the most regulated pathway in biofilm formation in
   Gram-positive bacteria [[134]42]. Further studies have found a link
   between increased purine nucleotide biosynthesis and the formation of
   “persister” bacteria and the emergence of antibiotic resistance
   [[135]43,[136]44]. This is explained by the fact that the increase in
   purines leads to an increase in ATP concentration, which is needed for
   a stronger formation of polymers, such as for the construction of the
   peptidoglycan layer [[137]43].

   After treatment with CHX, regulation of ATP-binding cassette (ABC)
   transporters occurred. These are transport systems responsible for the
   uptake of nutrients, amino acids, ions or peptides, as well as the
   release of hydrophobic substances or toxins [[138]45]. In addition, ABC
   transporters play a role in DNA repair and translation of mRNA
   [[139]46,[140]47]. Regulation of ABC transporters upon exposure to CHX
   has been discussed and linked to resistance formation in other studies
   [[141]48,[142]49,[143]50,[144]51]. Treatment of Acinetobacter baumannii
   with CHX resulted in the upregulation of the efflux system AdeAB and
   AceI, which were responsible for increased tolerances [[145]50]. A
   study by Liu et. al. also demonstrated that deletion of the ABC
   transporter LmrB resulted in enhanced biofilm formation and a higher
   resistance to acid, H[2]O[2], CHX, penicillin and erythromycin
   [[146]52]. Another group of transporters constitutes the
   phosphotransferase system (PTS), members of which were also
   significantly regulated in S. mutans after treatment with CHX in the
   present work. The role of these transporters is mainly the uptake and
   phosphorylation of sugar derivatives [[147]53,[148]54]. Accordingly,
   regulation of PTS may be responsible for reduced lactose and galactose
   degradation. In addition, it may be related to the upregulation of
   glycan synthesis [[149]55,[150]56]. The accumulation of intracellular
   polysaccharides (IPS) can be enhanced by nutrient deficiency, thus
   helping the cell to survive by promoting the emergence of “persister”
   bacteria [[151]57]. Glucan synthesis outside the cell indicates
   formation of extracellular polymeric substances (EPS), hinting to
   increased biofilm formation [[152]58,[153]59]. The increased biofilm
   formation after treatment of S. mutans with CHX was also demonstrated
   in a study by Dong et al. [[154]60].

   Sublethal CHX treatment also resulted in increased expression of
   specific genes encoding for the synthesis of enzymes, such as
   superoxide dismutase, thioredoxin, peroxiredoxin, and alkyl
   hydroperoxide reductases, indicating intracellular oxidative stress
   [[155]61]. Oxidative stress may lead to damage of proteins or DNA
   [[156]62,[157]63]. This is confirmed by the regulation of genes
   encoding for chaperones (groEL and groES) and a DNA protective protein
   after treatment with CHX. In addition, increased expression of Clp
   proteases was observed, which play a major role in cell survival after
   stressful events, such as heat shock, acid stress or oxidative stress
   [[158]64,[159]65]. In a study by Deng et al., it was shown that
   deletion of the protease ClpP in S. mutans leads to a higher
   sensitivity to CHX [[160]66]. In addition, measurement of metabolic
   activity showed increased tolerance to toxic H[2]O[2] and CHX after
   pre-incubation with sublethal levels of the corresponding compounds in
   the wild-type strain but not in the mutant [[161]66]. Accordingly, the
   increase in regulation of Clp proteases observed in the present work
   may be related to oxidative stress.

   In addition to the suggestion that the regulatory changes in gene
   expression occur due to oxidative stress, there is also evidence of
   intracellular acid stress in S. mutans after treatment with CHX, which
   may be explained by the acidic pH of CHX (pH 5.5). This assumption can
   be supported by a study of Svensäter et al. which showed that exposure
   of planktonic growing S. mutans to pH values between 6.0 and 3.5
   results in the induction of an acid tolerance response [[162]67]. To
   protect themselves from low pH, oral bacteria attempt to maintain
   intracellular pH by alkalinization [[163]68]. In the present work,
   malate permease (mleP) was found to be upregulated, which plays a role
   in malolactic fermentation. This involves the conversion of L-malate to
   L-lactate with the acquisition of ATP and CO[2], leading to
   alkalinization of the environment [[164]69,[165]70]. Furthermore, an
   increased expression of agmatine deiminase (aguA) was found in this
   study. Agmatine is a decarboxylated derivative of arginine that is
   taken up by the cell and degraded by agmatine deiminase. The
   degradation produces ATP, CO[2] and NH[3], which eventually leads to an
   increase in intracellular pH [[166]71,[167]72].

5. Conclusions

   In conclusion, we have identified and analyzed a considerable number of
   genes and pathways significantly regulated in S. mutans after sublethal
   treatment with CHX ([168]Figure 3). RNA-seq showed increased expression
   of genes related to oxidative stress. Furthermore, regulation of
   transporters and increased biofilm formation were found, which could
   support the risk of development of resistance. Due to the acidic pH of
   CHX, acid stress, in addition to oxidative stress, occurred. This is
   the first attempt to assess the transcriptomic stress response
   following sublethal treatment of S. mutans with CHX. Further
   investigations should focus on the transcriptomic stress response in
   clinical isolates or strains exhibiting phenotypic adaptation toward
   CHX in order to assess potential resistance mechanisms. Furthermore,
   whereas the present study investigated the transcriptomic stress
   response in planktonic bacteria, future studies should include bacteria
   grown in biofilms, which may exhibit different stress responses.

Acknowledgments