Establishment of Optogenetic Modulation of cAMP for Analyzing Growth, Biofilm Formation, and Virulence Pathways of Bacteria Using a Light-Gated Cyclase

In bacteria, cyclic adenosine monophosphate (cAMP) signaling plays an essential regulatory role whose modulation via optogenetic tools would provide researchers an immense opportunity to control biological processes simply by illumination. The cAMP signaling in bacteria is a complex network of regulatory pathways, which utilizes distinct proteomic resources under different nutrient environments. We established an optogenetic modulation of cAMP and studied important cellular process of growth, biofilm formation, and virulence in the model bacterium E. coli using a light-gated adenylate cyclase (LgAC) from Beggiatoa sp. Blue light-induced activation of LgAC elevated the cAMP level in a blue light-dependent manner in E. coli. Quantitative proteomics revealed a decrease in the level of certain proteins governing growth (PTS, Adk, AckA, GlnA, and EFP), biofilm formation (IhfA, flagellin, YajQ, YeaG, and HlfC), and virulence (ClpP, YebC, KatE, BtuE, and Zur) in E. coli cells expressing LgAC upon blue light illumination. This optogenetic modulation of cAMP would be useful for deciphering cAMP-associated host–pathogen signaling of bacterial systems. Proteome knowledge established by this research work would also be useful for the scientific community while adapting LgAC-based optogenetic modulation for studying other relevant cAMP-driven bacterial physiology (e.g., energy metabolism). The systematic utilization of the established method and more extensively designed experiments regarding bacterial growth, biofilm, survival, and virulence might provide a road map for the identification of new targets for developing novel antibacterial drugs.


Introduction
Cyclic adenosine monophosphate (cAMP) is a secondary messenger, which plays a central role in a myriad of signaling processes and is synthesized by a specific membrane or soluble protein, adenylate cyclase. This secondary messenger coordinates complex signaling by controlling intracellular metabolites like fructose 1,6-bisphosphate, phosphoenolpyruvate, acetyl-Coenzyme A, α-ketoglutaric acid, pyruvate, and oxaloacetate, known to be essential for enzyme-level regulation, transcriptional control, and nutrient homeostasis [1,2]. In bacteria, cAMP directly binds and activates the transcription factor of the cAMP receptor protein (CRP) family [3]. In Escherichia coli, cAMP is primarily known to be involved in mediating the glucose response or catabolite repression [4], but its role has now been expanding into the domains of biofilm formation, microbial virulence, and pathogenesis [3].
The cAMP-CRP complex controls extracellular matrix production and biofilm formation through transcriptional regulation of the csgD gene [5]. Hufnagel and coworker showed that cAMP may act as a signaling molecule, which drives the uropathogenic E. coli invasion from the bladder lumen to the interior of epithelial cells [5]. Muller et al. [6] suggested that the metabolic sensor cAMP-CRP regulates type 1 fimbriation, which is required for attachment and colonization of the uroepithelium. The surface attachment and biofilm formation are key factors for the onset of microbial pathogenesis and are also the major contributors to human diseases. The biofilm growth of E. coli might be a cAMP-dependent (mediated by phosphoenolpyruvate-dependent phosphotransferase, PTS) or -independent (pH-dependent) process [7]. Recently, Ritzert et al. [8] also demonstrated the association of cAMP-CRP complex with virulence, metal acquisition, and quorum sensing in addition to carbon metabolism. Several crucial bacterial processes may be managed by the modulation of the intracellular cAMP concentration, which might play an important role in the development of novel antibacterial drugs.
Researchers have tried to modulate the cAMP level in cellular systems to regulate the cyclic nucleotide(s)-mediated signaling in a controlled manner. Agonists activating adenylate cyclases [9], and phosphodiesterase (PDE) inhibitors [10], have been employed to modulate cAMP-dependent signaling; however, these methods lack spatial and temporal resolution. It was suggested that genetically encoded photo-switchable cyclases would provide scientists an opportunity to achieve cell-targeted control of cAMP levels by illumination [11]. In early 2005, channelrhodopsin (ChR) and other rhodopsins were used as genetically encoded photo-switches, for controlling the neural activity simply by illumination and such applications have paved the way for the development of the field of optogenetics [12].The successful application of these light-sensitive proteins as an optogenetic tool has encouraged scientists to identify and apply photoreceptor proteins for light-dependent regulation of cyclic nucleotide(s), protein-protein interaction, and protein translocation events in a cell.
The photoactivated adenylate cyclases (PACs) were genetically engineered into the heterologous systems and blue light illumination was used to alter the cAMP-mediated cellular and behavioral responses of the systems [13][14][15]. PACs possess a BLUF (blue light using flavin) domain-coupled adenylate cyclase and is utilized for blue light-dependent modulation of cAMP [16,17]. In 2007, Euglena PAC (ePAC) was expressed in a cell model system as well as in transgenic animals, and the cAMP level was manipulated simply by illumination with blue light [15,18]. However, these large-sized (110 kDa) PACs showed significant cyclase activity in the dark. Tanwar and coworkers characterized smaller-sized (38 kDa) PAC, which showed fast kinetics for light-regulated cyclase activity [13]. PACs have been used for altering cAMP levels in Drosophila [15,17] and for finding the mechanism underlying cAMP-dependent axonal morphogenesis [14]. In the current study, a light-gated adenylate cyclase (LgAC) from Beggiatoa sp. (bPAC) was utilized as an optogenetic tool to decipher the cAMP-modulated bacterial processes like growth, biofilm formation, and virulence in E. coli. A label-free quantitative proteomic approach was adapted to analyze the light-induced cAMP-dependent variations in the proteome regulating growth, biofilm formation, and virulence in E. coli.

Bioinformatics Analysis of LgACs
Protein sequences having a BLUF-cyclase domain were retrieved from the NCBI database (https://www.ncbi.nlm.nih.gov/), and were subjected to a conserved domain search using the Conserved Domain Architecture Retrieval Tool (CDART; https://www.ncbi.nlm.nih.gov/Structure/lexington/ lexington.cgi) [19]. Subsequently, 22 BLUF-cyclase domain-containing proteins were selected and aligned with a well-characterized PAC protein for the structure-function analysis using the BioEdit 7.2 [20]. The three-dimensional structure of LgAC from Beggiatoa sp. was predicted by I-Tasser (http://zhang.bioinformatics.ku.edu/I-TASSER) by employing an integrated approach comprising of threading, structural assembly, model selection, refinement of predicted structure, and structure-based functional annotation [21]. The accuracy of the predicted structure was estimated in terms of a confidence score (C-score) of the modeling. Docking analysis of ATP with the predicted 3D structure was performed using COACH I-TASSER to find the amino acid residues important for the interaction of ATP with LgAC. The docking output was visualized using Discovery studio visualizer version 16 (Dassault systems Biovia Corp., San Diego, CA, USA) [22].

Biochemical and Functional Characterization of the Recombinant LgAC
The commercially synthesized LgAC gene was cloned in the pASK vector (gifted by Prof. Peter Hegemann, Humboldt University, Germany) and was further confirmed by automated sequencing. The LgAC was expressed in Escherichia coli BL21 (λDE3). The recombinant protein was purified and the photophysical and biochemical characterizations of the purified protein were performed as per the methods employed previously [18]. The recombinant protein of interest was purified from the soluble fraction obtained after centrifugation of the cell lysate in the dark by immobilized metal affinity chromatography (IMAC) using Co 2+ -IMAC resins (Clontech Laboratories Inc., Mountain View, CA, USA). Subsequently, the purified protein was further separated using SDS-PAGE and transferred to a nitrocellulose membrane (Mdi Membrane Technologies, Harrisburg, PA, USA). Blotted membranes were blocked using 5% fat-free milk (Titan Biotech Ltd., Delhi, India) and incubated with a penta-His antibody. The membrane was washed with 1× PBS containing 0.1% Tween-20 (Sigma, St. Louis, MO, USA), followed by immunolabelling detection using an HRP-conjugated anti-mouse secondary antibody (Sigma, St. Louis, MO, USA). Signals were visualized by the chemiluminescence method and captured on X-ray films.

Measurement of Cyclase Activity of Recombinant LgAC Expressed in E. coli BL21 (λDE3) Cells
The cAMP levels were quantified immunologically to determine the light-regulated adenylate cyclase activity in E. coli cells expressing recombinant LgAC incubated in the dark and illuminated with blue light (BL). The bacterial cells were harvested by centrifugation at 6000 rpm for 10 min and resuspended in 1× phosphate buffer saline (PBS) containing lysozyme (50 µgmL −1 ; Bio Basic Inc., Markham, ON, Canada) and phenylmethylsulfonyl fluoride (PMSF) (200 µM; Sigma, St. Louis, MO, USA). The cells were then sonicated and clarified by centrifugation at 13,000 rpm for 1 h. The soluble fraction containing 60 µg of total cell protein was used for measuring the cyclase activity. All samples were acetylated to increase the sensitivity of the assay. The light-dependent cyclase activity was then quantified in terms of the levels of cAMP produced in the cell by a competitive ELISA-based kit (Cayman, Ann Arbor, MI, USA) using the manufacturer's instructions. The concentration of cAMP was quantified using a standard curve and the data are presented as mean ± SD. The data were statistically tested employing student's t test using SPSS version 16.0.

Confocal Microscopy of E. coli Cells Expressing LgAC
Harvested cells of E. coli grown until the late exponential phase were incubated in the dark and BL and were pelleted and washed 5 times with 1× PBS. The cells were fixed with 0.4% paraformaldehyde (Sigma, St. Louis, MO, USA; prepared in 1× PBS) for 20 min at room temperature after which they were pelleted down by centrifugation and re-dissolved in 1× PBS. Then, 200 µL of cells were seeded on poly-L-lysine (Sigma, St. Louis, MO, USA)-coated coverslips. After 1 h, the coverslips were washed 3 times with 1× PBS and mounted onto clean slides with glycerol and visualized by exploiting the intrinsic fluorescence property of the LgAC. For the detection of biofilm formation, E. coli cells expressing LgAC were grown on fresh cover slips with the addition of glucose (Hi Media, West Chester, PA, USA) in media and incubated at 37 • C for 24 h in the dark and blue light. After incubation, cover slips were washed mildly with 1× PBS followed by fixation with 0.4% paraformaldehyde (Sigma, St. Louis, MO, USA; in 1× PBS) for 20 min. Mounting and visualization was done as mentioned previously [10].

Label-Free Quantitative Proteomics
The comparative label-free quantitative proteomics of E. coli cells expressing LgAC incubated in the dark and illuminated with BL was performed. The cells were lysed using lysis buffer (8.0 M urea in 25 mM ammonium bicarbonate) supplemented with protease inhibitors (Roche Applied Science, Penzberg, BY, Germany). The total protein in the lysate was quantitated using Brad-ford assay (Bio-Rad, Hercules, CA, USA). The sample was first reduced using 5 mM tris (2-carboxyethyl) phosphine and later alkylated using 50mM iodoacetamide. The sample was then digested using trypsin (1:50 trypsin: lysate ratio) for 16 h at 37 • C. The digested sample was cleaned using a C18 silica cartridge (The Nest Group, Southborough, MA, USA) according to the manufacturer's protocol and dried using a speed vac. The dried pellet was resuspended in buffer A (5% acetonitrile, 0.1% formic acid) and analyzed on nano-LC/MSMS. Label-free sample was analyzed using an EASY-nLC system (Thermo Fisher Scientific, Waltham, MA, USA) coupled to an LTQ Orbitrap-Velos mass spectrometer (Thermo Fisher Scientific, Waltham, MA, USA) equipped with a nano-electrospray ion source as per the method described in Jhingan et al. [23]. Four biological replicates were processed for each sample, and the RAW files were analyzed with MaxQuant (version. 1.4.1.2) against the E. coli UniProt reference proteome database. Functional annotation and pathway validation of differentially expressed proteins (DEPs) were conducted using the KEGG pathway analysis (http://kobas.cbi.pku.edu.cn/anno_iden.php). Protein networking of selected DEPs were performed using String version 11 (Academic Consortium 2020; [24]) and the output was analyzed with Cytoscape 3.7.2 (San Diego, CA, USA) using the betweenness centrality algorithm [25]. A Venn diagram representing the distribution of proteins identified by label-free comparative quantitative proteomics was created using Venny 2.1 (Calle Darwin, MD, Spain) [26]. A heat map was constructed with Multi Experiment Viewer (MeV 4.9.0, Rockville, MD, USA) to show the comparative expression profile of differentially expressed proteins [27]. A pie diagram was created utilizing Past 3.11 (Natural History Museum, University of Oslo, Sars gt., Oslo) [28].

Biofilm of E. coli Expressing LgAC
The biofilm formation by E. coli cells in the dark and BL was estimated as described by O'Toole and Kolter [29]. An overnight grown culture (100 µL) was inoculated in 900 µL of LB medium containing ampicillin (10 µgmL −1 ; Hi Media, West Chester, PA, USA) and anhydrous tetracycline (0.3 mM; IBA Lifesciences, Göttingen, LS, Germany) in a polystyrene plate, and incubated at 37 • C for 24 h in the dark and BL. After 24 h of incubation, the cell suspension was removed, and the walls of the plate were rinsed with autoclaved distilled water and dried at room temperature. Further, 1 mL of crystal violet (0.1% v/v) was added to the plate for 20 min. The stained wells were later washed several times with autoclaved distilled water and allowed to dry at room temperature. The stained cells were extracted with 1 mL of 20% (v/v) acetone in ethanol and the optical density at 550 nm (OD 550 ) was estimated using the Perkin Elmer Lambda 465 Spectrophotometer (Waltham, MA, USA). Experiments were performed with three biological triplicates and two technical replicates for each biological replicate (n = 6) and data was statistically verified by ANOVA using SPSS (ver. 16.0). The qualitative assay for biofilm detection was also performed under dark and BL following the protocol of Christensen et al. [30] with slight modifications. An overnight grown culture [100 µL culture + 900 µL of LB medium supplemented with 100 µg mL −1 ampicillin (Hi Media, West Chester, PA, USA) and 3 mM anhydrous tetracycline (IBA Lifesciences, Göttingen, LS, Germany)] was incubated at 37 • C for 24 h. After 24 h of incubation, the cells were removed, and the plate was washed several times with phosphate saline buffer (1× PBS). The biofilm formed on the walls of the polystyrene plate was stained with safranin for 1 h. The stained plates were later rinsed twice with PBS to remove the excess stain and further air dried. Finally, the stained biofilm formed was visualized on the wall and the bottom of the plates.

LgAC is an Optogenetic Modulator of cAMP in E. coli
The homology analysis of the BLUF domain ( Figure S1A), linker region ( Figure S1B), and cyclase domain ( Figure S1C) from 22 BLUF cyclases was performed against the well-characterized PAC from Oscillatoria acuminata. The residues important for BLUF photodynamics and ATP binding pocket of the cyclase domain were mostly found to be conserved. LgAC from Beggiatoa sp. showed the maximum conservation for the ATP binding pocket residues (ASP6, LEU8, ALA9, PHE10, SER11, THR12, ILE48, GLY49, ASP50, CYS51, ARG133) (Figure S1C and S2; [31]). Therefore, the LgAC from Beggiatoa sp. was selected as the PAC to be used as an optogenetic tool for the regulation of cAMP-dependent bacterial physiological processes. Further, to study the impact of light-gated adenylate cyclase on cAMP-dependent processes in bacteria, the heterologous expression of recombinant LgAC was induced in E. coli. The LgAC was expressed in E. coli cells as an N-terminal Histidine (His) tag fusion protein (5X His:LgAC). The spectroscopic analysis of the purified recombinant LgAC suggested the functional expression of LgAC in E. coli ( Figure 1A). In addition, Western blotting with an anti-His antibody ( Figure 1B) and confocal analysis ( Figure 1C-E) displayed a robust expression of recombinant LgAC in E. coli cells. An ELISA-based immunoassay was performed to estimate the cAMP level in the total lysates prepared from the LgAC-expressing cells incubated under dark and BL to confirm the photoactivation of LgAC. A significant increase (~10 fold; p < 0.001) in the cAMP level was detected in E. coli after BL illumination when compared to dark-adapted cells ( Figure 1F); however, insignificant cyclase activity of LgAC under dark corresponding to the basal cyclase activity in E. coli cells was also observed.

Label-Free Quantitative Proteomics Revealed cAMP-Dependent Modulation of Growth, Biofilm Formation, and Virulence in E. coli
The comparative proteomics data of LgAC-carrying E. coli cells incubated in the dark and illuminated with BL revealed quantitative changes in the protein profile. A total of 732 proteins were identified (Figure 2A), out of which 57 were differentially expressed, and are represented by a heat map ( Figure 2B; Table 1). The KEGG pathway analysis categorically linked differentially expressed proteins (DEPs) with 29 different pathways ( Figure 2C; Table 2). The DEPs involved in growth, biofilm formation, and virulence were studied in detail.

Label-Free Quantitative Proteomics Revealed cAMP-Dependent Modulation of Growth, Biofilm Formation, and Virulence in E. coli
The comparative proteomics data of LgAC-carrying E. coli cells incubated in the dark and illuminated with BL revealed quantitative changes in the protein profile. A total of 732 proteins were identified (Figure 2A), out of which 57 were differentially expressed, and are represented by a heat map ( Figure 2B; Table 1). The KEGG pathway analysis categorically linked differentially expressed proteins (DEPs) with 29 different pathways ( Figure 2C; Table 2). The DEPs involved in growth, biofilm formation, and virulence were studied in detail.  Table 1. (C) A pie diagram representing the KEGG (Kyoto Encyclopedia of Gene and Genome) pathway analysis output data; tick labels represent the KEGG identifiers (IDs) of the corresponding pathways. Details of the pathways are given in Table 2.  Table 1. (C) A pie diagram representing the KEGG (Kyoto Encyclopedia of Gene and Genome) pathway analysis output data; tick labels represent the KEGG identifiers (IDs) of the corresponding pathways. Details of the pathways are given in Table 2. Abv.-Abbreviations, nk-not known, *-growth associated proteins, #-biofilm associated proteins, $-virulence associated proteins. The KEGG pathway analysis suggested that most DEPs are involved in growth and energy metabolism ( Figure 2C; Table 2). Proteomic analysis showed a~2-fold reduction in the expression of growth-related proteins, i.e., phosphoenolpyruvate-dependent phosphotransferase (PTS), adenylate kinase (Adk), and acetate kinase (AckA) in recombinant E. coli illuminated with BL ( Figure 2B; Table 1). In E. coli, the cyclic AMP (cAMP)-catabolite receptor protein (CRP) complex regulates (positively or negatively) the catabolic repression and transcription of several genes involved in carbohydrate catabolism [2,32]. The inactivation of either adenylate cyclase or CRP would lead to highly pleiotropic phenotypes [33]. Bettenbrock and coworkers showed the correlation between the intracellular cAMP concentration and growth in E. coli [33]. Phosphoenolpyruvate-dependent phosphotransferase (PTS) regulates carbon catabolism via phosphoenolpyruvate-dependent protein kinase enzyme IIA Crr (EIIA Crr )-mediated inducer exclusion and cAMP-CRP-based catabolic repression in E. coli. A reduction in PTS expression is explained by the fact that phosphorylated EIIA Crr activates bacterial adenylate cyclase, which further synthesizes cAMP [33]. However, in the present experiment, adenylate cyclase activation was light induced, which bypasses the requirement of EIIA Crr . Adenylate kinase is another crucial enzyme involved in regulating the growth of E. coli by controlling the cellular ATP levels [34]. In Streptococcus pneumoniae, the adenylate kinase (SpAdk) level and growth rate were correlated, where mutation in SpAdk caused growth defects in vivo [35]. The ratio of phosphorylated and non-phosphorylated AckA on the other hand regulates the phosphoenzyme I/enzyme I ratio of PTS as well as the ATP/ADP ratio, respectively. This in turn controls the rate of sugar transport and cellular growth [36]. Apart from regulating bacterial growth, AckA is also considered abroad-spectrum novel drug target in several pathogenic bacteria [37].
The cAMP modulation also controls the molecular complexes that regulate coordination between carbon metabolism and other nutrient homeostasis that are essential for the growth and survival of E. coli [1]. The downregulation of glutamine synthetase (GlnA;~2 fold) and a specific potassium ion (K + )-binding protein (Kbp;~2.5 fold) was observed after BL illumination ( Figure 2B; Table 1). In many bacteria, GlnA is an important enzyme of the central nitrogen metabolic circuit other than glutamate synthase (GOGAT), and glutamate dehydrogenase (GDH), which are responsible for the assimilation of ammonia/ammonium (NH4 + ) into glutamine [38]. In bacteria, the glutamate/glutamine ratio serves as a sensor of external N availability and is essential for their proper growth and sustenance [39]. Yan et al. [38] reported a glnA mutant SK3112 that showed slow growth on NH4 + as an N-source due to a limiting internal glutamine pool. In adverse conditions, GlnA in coordination with Adk and Ack also provide a phosphorelay system, which behaves as a potent ATP generator [40]. This phospho-network helps in acquiring ATP by oxidative phosphorylation for bacterial growth in adverse conditions [41]. Potassium ion is involved in several key processes like the maintenance of intracellular pH and osmolarity [42]. The cytoplasmic protein Kbp, a member of RpoS regulon, is essential for normal growth in the presence of high levels of external K + . Ashraf et al. [42] compared the growth of wild-type and ∆kbp strains at different K + concentrations and found that at higher K + concentrations, a mutation in kbp resulted in a delayed exponential phase. Non-canonical elongation factor (EF-P) was downregulated (~2.5 fold) in E. coli illuminated with BL ( Figure 2B; Table 1). In many organisms, short sequences ending with a Pro residue (PPG, PPP, or longer Pro clusters) in nascent peptides may lead to ribosome pausing (stall the translation process), which can be alleviated by EF-P [43,44]. The regulated synthesis of proteins having terminal Pro residues in the nascent peptide by EF-P is essential for different processes. Inactivation of EF-P or its modifying enzymes YjeA or YjeK causes pleiotropic phenotypes, including a reduced growth rate and motility [45].
Growth curve analysis of LgAC-expressing E. coli incubated in the dark and illuminated with BL was performed, and the results obtained showed a reduction in the growth of BL-illuminated E. coli cells ( Figure 3A). The untransformed E. coli cells (without LgAC expression; negative control) did not show any pronounced changes in growth between the cells grown in the dark and under BL illumination ( Figure S3). For E. coli, it is already established that glucose modulates the intracellular concentration of cAMP [46]; therefore, we considered the glucose-supplemented condition as a positive control. When the growth of E. coli cells supplemented with glucose (0.1%) was compared with dark-adapted and BL-illuminated E. coli cells expressing LgAC, the growth of E. coli cells in the glucose-supplemented condition is comparable to dark-adapted cells ( Figure S3). In E. coli, adenylate cyclase catalyzes the cyclization of adenosine triphosphates (ATP) into cyclic adenosine monophosphates (cAMP) by releasing a pyrophosphate (PPi) molecule. Therefore, activation of adenylate cyclase and cAMP generation may indirectly inhibit bacterial growth by reducing cellular ATP levels.
A comparative quantitative Western blot analysis using protein-specific antibodies of the differentially expressed protein(s) (DEP) and quantitative proteomics as well as transcriptomics (RNA-seq) would be required for targeted validation of the quantity of these protein(s). These studies would establish a direct molecular correlation (transcriptional and/or translational) between the optogenetic modulation of cAMP levels and its role in alteration of the quantity of the protein(s) associated with biofilm formation, growth, and virulence of E. coli.

Functional Expression of LgAC Modulates Biofilm Formation and Associated Processes in E. coli
The cAMP-based regulation of the expression of genes controlling biofilm formation in many bacteria is well established [47]. In Vibrio cholerae thriving under a low glucose concentration, the cAMP-CRP complex regulates biofilm formation by modulating CdgA and HapR (quorum sensing regulator) [48,49]. Biofilm formation is an important transitional event occurring in the bacterial lifestyle, wherein the inhibition of motility promotes biofilm formation. Motility-to-biofilm transition is regulated by controlling either flagellar function or genes of the flagella-biofilm regulatory network [50]. Proteomic analysis revealed a ~2-fold downregulation of two proteins associated with the flagella-biofilm network (i.e., IhfA [51] and flagellin (FliC) [52]) in BL-illuminated E. coli ( Figure 2B; Table 1). IhfA is one of the global regulators (GRs) of biofilm formation in E. coli and modulates the expression of genes (rpoS, matA, and csgD) involved in the flagella-biofilm network [51]. In E. coli, flagellin is a structural protein encoded by the fliC gene, which is involved in the polymerization of the long helical filament of bacterial flagella [52]. Mutation in the fliC gene causes a severe defect in biofilm formation [53]. In Pseudomonas aeruginosa

Functional Expression of LgAC Modulates Biofilm Formation and Associated Processes in E. coli
The cAMP-based regulation of the expression of genes controlling biofilm formation in many bacteria is well established [47]. In Vibrio cholerae thriving under a low glucose concentration, the cAMP-CRP complex regulates biofilm formation by modulating CdgA and HapR (quorum sensing regulator) [48,49]. Biofilm formation is an important transitional event occurring in the bacterial lifestyle, wherein the inhibition of motility promotes biofilm formation. Motility-to-biofilm transition is regulated by controlling either flagellar function or genes of the flagella-biofilm regulatory network [50]. Proteomic analysis revealed a~2-fold downregulation of two proteins associated with the flagella-biofilm network (i.e., IhfA [51] and flagellin (FliC) [52]) in BL-illuminated E. coli ( Figure 2B; Table 1). IhfA is one of the global regulators (GRs) of biofilm formation in E. coli and modulates the expression of genes (rpoS, matA, and csgD) involved in the flagella-biofilm network [51]. In E. coli, flagellin is a structural protein encoded by the fliC gene, which is involved in the polymerization of the long helical filament of bacterial flagella [52]. Mutation in the fliC gene causes a severe defect in biofilm formation [53]. In Pseudomonas aeruginosa PAO1, the FliC protein is associated with the dispersal of the biofilm [54], where post-translational modification (phosphorylation of T27 and S28 residues) of FliC alters biofilm formation and dispersal. Proteomic analysis revealed a significant reduction (~2 fold) in the expression of the YeaG protein in BL-illuminated E. coli ( Figure 2B; Table 1). Ibanez-Ruiz et al. [55] reported the yeaG gene as a member of the rpoS regulon that plays a crucial role in biofilm formation and its regulation in Salmonella entrica Serovar Typhimurium. Further, yeaG was identified as a putative stress response gene that was associated with the stress response occurring during biofilm growth [56].
A c-di-GMP receptor protein, YajQ, showed a significant decrease (~2 fold) in expression in BL-illuminated E. coli ( Figure 2B; Table 1). An and coworkers performed an affinity pull-down assay using cyclic di-GMP-coupled magnetic beads and identified the YajQ family protein as a potential cyclic di-GMP receptor, and mutation of the gene manifested a reduction in biofilm formation [57]. A reduction (~2 fold) in expression of the HflC protein (FtsH modulator) was observed in E. coli after BL illumination ( Figure 2B; Table 1). In bacteria, the matrix-producing cells are essential for effective biofilm formation [58]. In E. coli cells, HflC and HflK (chaperons) oligomerizes FtsH protein and make it a functional unit [59]. Yepes and coworkers showed a direct interaction between FloT/YqfA (HflC/HflK homologues) and FtsH. Inhibition of FloT/YqfA negatively affected the FtsH concentration and further reduced the differentiation of matrix producer cells and biofilm formation [59].
In this study, the cAMP-dependent biofilm formation ability of E. coli cells expressing LgAC was estimated, and the result showed a significant reduction (p < 0.05) in the biofilm formation ability in BL-illuminated cells ( Figure 3B). However, the alteration in the cAMP-dependent biofilm formation ability of untransformed E. coli culture (negative control) was not significant (p > 0.069) when compared between dark-adapted and BL-illuminated conditions ( Figure S3), suggesting that the reduction in the biofilm formation ability of E. coli was partially associated with the light-dependent elevation of cAMP. Safranin-based qualitative detection and confocal microscopy revealed a reduction in the biofilm formation ability of E. coli expressing LgAC upon BL illumination ( Figure 3C,D). In the confocal microscopy, the fluorescent property of the protein was exploited for the visualization of biofilm formation. Cell aggregation, which is characteristic of biofilm-forming microorganisms, was reduced or absent in the BL-illuminated LgAC-expressing E. coli culture ( Figure 3D).

Optogenetic Modulation of cAMP Regulates Virulence Signaling in E. coli
The involvement of cAMP in the regulation of bacterial virulence is a well-known phenomenon [60]. In Mycobacterium tuberculosis, 10 functional adenylate cyclases (ACs) and CRP were reported to be involved in the complex regulation of virulence [61]. Quantitative proteomics showed a reduced expression of proteins (ClpP, YebC, KatG, BtuE, and Zur) associated with virulence in E. coli after BL illumination. The caseinolytic protease P (ClpP), a serine protease, was significantly reduced (~2 fold) in the LgAC-expressing BL-illuminated E. coli ( Figure 2B; Table 1). In bacterial pathogens, ClpP plays an important role in the onset of pathogenesis and virulence [62]. The Clp mutant of Salmonella typhimurium was unable to grow in peritoneal macrophages, which is a crucial step for the establishment of S. typhimurium virulence [63]. Knudsen and coworkers demonstrated that S. typhimurium virulence was attenuated in the Clp mutant due to the mis-regulation of RpoS as well as the indirect regulation of CsrA and the genes of the Salmonella pathogenicity island 1 [64]. YebC, a PmpR (pqsR-mediated PQS regulator) family protein, which regulates quorum sensing in the bacterial system [65], was significantly downregulated (~1.5 fold) in E. coli after BL illumination ( Figure 2B; Table 1). Bacterial virulence depends on the type III secretion system (T3SS), type VI secretion system (T6SS), and quorum sensing system (QS), which controls cell-to-cell communication and helps the bacterial population to invade and interact with hosts [66][67][68]. Liang et al. [49,69] also reported the role of the cAMP-CRP complex in the regulation of HapR expression (a major regulator of quorum sensing) in a cell density-dependent manner in V. cholerae via control of the Fis and RpoS pathways. Therefore, modulating cAMP-dependent quorum sensing may control the bacterial pathogenesis and compromise host-pathogen interactions.
Bacterial pathogenesis elicits a host immune response to generate and release toxic reactive oxygen species (ROS) by phagocyte cells. However, many pathogens have evolved an arsenal of effective oxidative stress defense systems to cope with such conditions, such as molecular scavengers, proteins, and DNA repair enzymes [70]. Quantitative proteomics in the current study revealed the downregulation of proteins associated with oxidative stress in the LgAC-expressing E. coli after BL illumination. An oxidative stress defense protein was significantly downregulated (~1.5 fold) in E. coli after BL illumination. Furthermore, a~2-fold downregulation of BtuE protein (glutathione peroxidase) and specific catalase (KatE) was also observed ( Figure 2B; Table 1). In E. coli, BtuE has non-specific peroxide activity, and it utilizes thioredoxin A for the decomposition of different hydroperoxides [71]. In Xanthomonas axonopodis pv. citris, KatE is required for full virulence in Citris plant, where KatE is involved in the colonization and survival of X. axonopodis during pathogenesis [72]. Significant downregulation (~4 fold) of the Zur protein (zinc uptake regulator) was observed in E. coli after BL illumination ( Figure 2B; Table 1). Nutritional immunity is one of the most common responses posed by the host upon pathogen invasion, wherein the host intoxicates the invading microbes with essential metals, of which zinc is an important component. Several microbes have evolved ways to adapt hypozincemic or hyperzincemic conditions [73]. In bacteria, Zur regulated the expression of genes involved in the control of zinc homeostasis and the onset of oxidative stress [74].

Protein Networking Deciphers Functional Links among the DEPs
We constructed a protein network of DEPs associated with growth, biofilm, and pathogenesis, and the resulting network revealed different nodes showing connections between proteins within as well as between the relevant pathways ( Figure 4). The betweenness centrality analysis depicted dominating nodes and paths in the constructed protein network. Interestingly, most of the DEPs (HflC, IhfA, ClpP, BtuE, GlnA, Adk, and AckA) identified through quantitative proteomics were principal effectors that modulate the functioning of respective bacterial signaling pathways. Among the growth regulating proteins, Adk, AckA, and GlnA were the main node, which exert a determining impact on the protein network ( Figure 4). In bacteria, cellular stoichiometry between Adk, AckA, and GlnA is crucial for the maintenance of the energy budget by invoking alternative ATP-generating pathways, which is an important factor regulating bacterial growth [40]. Similarly, IhfA, HflC, and MotA (flagellar motor) were the main nodes influencing the biofilm network ( Figure 4). IhfA and MotA are important regulators of the flagella-biofilm network, whereas HflC regulates the differentiation of matrix-producing cells, two of the important steps of bacterial biofilm formation. ClpP and BtuE are the main nodes influencing the bacterial pathogenicity network (Figure 4). Michel et al. [75] demonstrated the regulatory role of ClpP protease in the oxidative stress response by controlling the antioxidative enzymes like catalase and peroxidases, respectively. In bacteria, as in other organisms, physiological processes do not occur in an isolated and exclusive manner. They are part of an interlinked and complex network driven by different metabolic pathways in parallel. Likewise, growth, biofilm formation, and pathogenesis are interrelated phenomena in bacteria; therefore, the machinery regulating these processes might be linked via some regulatory proteins.

Figure 4.
Growth-biofilm-pathogenicity protein network. Protein-protein interaction analysis was performed by using String version 11 (https://string-db.org/;) and further analyzed by Cytoscape 3.4.1 applying the betweenness centrality algorithm. Color range represents betweenness values in the network. Protein network framed under blue, pink, and green lines belongs to biofilm formation, virulence, and growth, respectively. Blue, pink, and green solid arrows represent proteins that showed differential expression in quantitative proteomic analysis of LgAC-expressing E. coli incubated in the dark and under blue light illumination, respectively.

Conclusions
The establishment of a new optogenetic modulation is an innovative approach to reprogram gene expression and cell signaling with spatiotemporal precision. A variety of optogenetic tools have been developed for controlling neural activity, behavioral response, and cellular events in cells and the whole organism. In bacteria, cAMP signaling regulates a variety of crucial processes essential for its growth, survival, and pathogenesis. We established optogenetic modulation of cAMP in a model bacterium (E. coli) and studied the optogenetically modulated cAMP-dependent bacterial processes like growth, biofilm formation, and virulence using a light-gated adenylate cyclase (LgAC). The blue light-dependent activation of LgAC and the corresponding increase in cellular cAMP level demonstrated a significant decrease in the growth and levels of relevant proteins (PTS, Adk, AckA, GlnA, Kbp, and EF-P). Proteins regulating the motility to biofilm transition and flagella-biofilm network were modulated upon optogenetic control of cAMP. Photoactivation of LgAC modulated the amount of IhfA, YajQ, and HflC, which regulates crucial steps in biofilm formation. Photoinduced elevation in the cAMP level influenced the levels of proteins (ClpP, YebC, BtuE, KatE, and Zur), which are responsible for conferring virulence. ClpP and YebC proteins are known to regulate different aspects of bacterial virulence and pathogenesis. BtuE and Zur represent the group of proteins that helps pathogens counter oxidative stress imposed by the host immune system. It is evident that LgAC is an optogenetic photoswitch, which can modulate cAMP-dependent bacterial growth, biofilm formation, and pathogenesis simply by illumination with blue light. The established optogenetic modulation of bacterial cAMP opens avenues for studying cAMP-driven physiological aspects, including the host-pathogen signaling of bacterial systems. The systematic utilization of the established method and more extensively designed experiments regarding the bacterial growth, biofilm, survival, and virulence might help in providing a road map for the identification of new targets for developing antibacterial drugs. . Growth-biofilm-pathogenicity protein network. Protein-protein interaction analysis was performed by using String version 11 (https://string-db.org/;) and further analyzed by Cytoscape 3.4.1 applying the betweenness centrality algorithm. Color range represents betweenness values in the network. Protein network framed under blue, pink, and green lines belongs to biofilm formation, virulence, and growth, respectively. Blue, pink, and green solid arrows represent proteins that showed differential expression in quantitative proteomic analysis of LgAC-expressing E. coli incubated in the dark and under blue light illumination, respectively.

Conclusions
The establishment of a new optogenetic modulation is an innovative approach to reprogram gene expression and cell signaling with spatiotemporal precision. A variety of optogenetic tools have been developed for controlling neural activity, behavioral response, and cellular events in cells and the whole organism. In bacteria, cAMP signaling regulates a variety of crucial processes essential for its growth, survival, and pathogenesis. We established optogenetic modulation of cAMP in a model bacterium (E. coli) and studied the optogenetically modulated cAMP-dependent bacterial processes like growth, biofilm formation, and virulence using a light-gated adenylate cyclase (LgAC). The blue light-dependent activation of LgAC and the corresponding increase in cellular cAMP level demonstrated a significant decrease in the growth and levels of relevant proteins (PTS, Adk, AckA, GlnA, Kbp, and EF-P). Proteins regulating the motility to biofilm transition and flagella-biofilm network were modulated upon optogenetic control of cAMP. Photoactivation of LgAC modulated the amount of IhfA, YajQ, and HflC, which regulates crucial steps in biofilm formation. Photoinduced elevation in the cAMP level influenced the levels of proteins (ClpP, YebC, BtuE, KatE, and Zur), which are responsible for conferring virulence. ClpP and YebC proteins are known to regulate different aspects of bacterial virulence and pathogenesis. BtuE and Zur represent the group of proteins that helps pathogens counter oxidative stress imposed by the host immune system. It is evident that LgAC is an optogenetic photoswitch, which can modulate cAMP-dependent bacterial growth, biofilm formation, and pathogenesis simply by illumination with blue light. The established optogenetic modulation of bacterial cAMP opens avenues for studying cAMP-driven physiological aspects, including the host-pathogen signaling of bacterial systems. The systematic utilization of the established method and more extensively designed experiments regarding the bacterial growth, biofilm, survival, and virulence might help in providing a road map for the identification of new targets for developing antibacterial drugs.
Supplementary Materials: The following are available online at http://www.mdpi.com/2076-3417/10/16/5535/s1, Figure S1: Multiple sequence alignment of the BLUF (blue light using flavin) domain (A), linker region (B), and cyclase domain (C) from different organisms with that of the well-characterized BLUF-Cyclase proteins from Oscilatoria accuminata depicting the functionally important amino acids in the BLUF domain (red star represent the amino acids important for photocycle and photodynamics of the BLUF while the orange square denotes the residues essential for the flavin binding pocket), in the cyclase domain [green stars represent important amino acids required for ATP binding, red arrows depict metal-binding residues (MBR), blue arrows correspond to the essential adenine-binding residues (ABR), and orange arrows denote transition state-stabilizing residues (TSSR)], respectively. The length of the linker region is crucial for light-gated cyclase activity of all the BLUF; cyclases were shown to be similar except in Leptonema illini, where the length of the linker region was surprisingly very short, Figure S2