G-quadruplex stabilization in the ions and maltose transporters gene inhibit Salmonella enterica growth and virulence.

The G-quadruplex structure is a highly conserved drug target for preventing infection of several human pathogens. We tried to explore G-quadruplex forming motifs as promising drug targets in the genome of Salmonella enterica that causes enteric fever in humans. Herein, we report three highly conserved G-quadruplex motifs (SE-PGQ-1, 2, and 3) in the genome of Salmonella enterica. Bioinformatics analysis inferred the presence of SE-PGQ-1 in the regulatory region of mgtA, SE-PGQ-2 in ORF of entA, and SE-PGQ-3 in the promoter region of malE and malK genes. The G-quadruplex forming sequences were confirmed by biophysical and biomolecular techniques. Cellular studies affirm the inhibitory effect of G-quadruplex specific ligands on Salmonella enterica growth. Further, PCR inhibition, reporter based assay, and RT-qPCR assays emphasize the biological relevance of G-quadruplexes in these genes. Thus, this study confirmed the presence of G-quadruplex motifs in Salmonella enterica and characterized them as a promising drug target.

Introduction

Salmonella enterica belongs to the Enterobacteriaceae family and is known to cause typhoid fever and food poisoning. Salmonella enterica contains six subspecies among those, S. enterica is known to infect humans and further divided into two subclasses (Fig. 1a) [1,2]. Typhoidal class of S. enterica included S. enterica subsp. enterica ser. Typhi (S. ser. Typhi) and S. ser. Paratyphi and are known to causes typhoid fever, whereas non-typhoidal class includes S. ser. Enteritidis, and S. ser. Typhimurium causes food poisoning in humans [3]. As per the Centers for Disease Control and Prevention (CDC), typhoid fever causes ~22 million new cases and ~ 200,000 deaths every year across the world [4]. The emergence of anti-microbial resistance for chloramphenicol, co-trimoxazole, ampicillin, ciprofloxacin, ofloxacin, azithromycin, and cephalosporin leads to increased death rate due to clinical treatment failure and make the situation more dangerous [[5], [6], [7]]. Due to its high prevalence and rapid emergence of drug resistance, Salmonella rings a global alarm for the development of novel and promising therapeutic approaches.

Fig. 1

Salmonella genus classification and G-quadruplex topologies. a) Flowchart depicting the classification of the Salmonella genus. b) G-quartet structure showing hoogsteen hydrogen bond formation between Guanines and cation binding. c) Different types of topologies formed by intermolecular and intramolecular G-quadruplex structures.

S. enterica is an intracellular pathogen that grows in phagocytes and macrophages. During the growth phase, the host innate immune system generates various nitro-oxidative and oxidative stresses to eradicate this pathogen. However, S. enterica possesses a magnesium homeostasis mechanism that is mainly controlled by mgtA gene and helps bacteria to survive in nitro-oxidative stressed conditions [8]. The mgtA encodes for an Mg2+ transport ATPase that neutralizes the reactive nitrogen stress (RNS, nitro-oxidative) and plays a vital role in the bacterial survival inside macrophages (Supplementary data 1: Fig. S1a) [9]. Hence, targeting mgtA gene expression may serve as a promising therapeutic approach against S. enterica pathogenesis.

Similar to Mg2+, Iron(Fe) is also an essential nutrient element for the S. enterica. The S. enterica contains an entABCDEF operon that encodes for two low molecular weight siderophores (enterobactin and salmochelin) that have a high affinity for iron than the host iron-binding proteins and import iron inside the bacteria cell (Supplementary data 1: Fig. S1b) [10]. The formation of these siderophores involves the conversion of 2,3-dihydro-2,3-dihydroxybenzoate (2,3 DHB) to 2,3-dihydroxybenzoate (DHB) in the presence of 2,3-dihydro-2,3-dihydroxybenzoate dehydrogenase enzyme (EntA), that is encoded by entA of entABCDEF operon. The inhibition of entA gene expression has been observed to abolish the DHB formation leading to the reduced production of enterobactin and salmochelin [11]. Targeting the production of siderophore has previously shown to have anti-microbial activity against Mycobacterium tuberculosis [12], Aspergillus fumigatus [13], Yersinia pestis [14], Pseudomonas aeruginosa [15], and Vibrio chloreae [16]. Therefore, targeting entA gene may prove as another therapeutic approach to combat the infection and virulence of S. enterica.

Similar to magnesium and iron, maltose is another essential nutrient for the S. enterica that serves as a significant source of carbon for this bacterium. The uptake of maltose from the host environment is tightly regulated by two genes, malK and malE (Supplementary data 1: Fig. S1c) [17]. Inhibiting the malK and malE synthesis has shown to decline the growth of S. enterica [18]. Henceforth, targeting these three genes (mgtA, entA, malK, and malE) will make S. enterica unable to grow inside the gastrointestinal tract.

G-rich regions having a motif G>2NG≥2NG≥2NG≥2 present in the genome tends to form specific non-canonical secondary structures known as G-quadruplex (G4). These G4s are stabilized by the presence of monovalent and some divalent cations and can adopt various topologies (Fig. 1b & 1c) [19,20]. This structural diversity has been exploited for the diagnosis and therapeutic targeting [21,22]. The G-quadruplex structure forming nucleic acid motifs have been shown to be evolutionarily conserved in eukaryotes [23], prokaryotes [24], protista [25], plants [26] and viruses [27] and work as promising drug targets in various diseases such as cancer, neurological disease, and virus infection. The G4s are reported to play a regulatory role in various biological processes such as regulating DNA replication by the specification of origin of replication (ORI) sites, telomere maintenance in human cells, antigenic variations by regulating recombination, transcription, and translation.

Recently, G-quadruplexes are being investigated for their involvement in virulence and survival mechanisms of various human pathogens [28]. Stabilization of G4s in protozoans: Plasmodium falciparum, Trypanosoma brucei and Leishmania donovani have shown anti-protozoal activities [29]. Further, G4s have emerged as a promising drug target in viruses like SARS coronavirus, Human Papilloma virus (HPV), Zika, Ebola, Herpes simplex virus (HSV), Epstein-Barr virus (EBV), Hepatitis B virus (HBV), Hepatitis C virus (HCV), Human immunodeficiency virus 1 (HIV-1), Nipah virus, Adenoviruses, etc. where they play an essential role in proliferation and pathogenicity [27,30,31]. In bacteria, G4 present at the upstream of pilE locus, B31 vlsE locus and tprK antigen protein in Neisseria gonorrhoea, Borrelia burgdorferi, and Treponema pallidum, respectively acts as an activator for the initiation of antigenic variation and helps the pathogens in bypassing the immune system of the host [28]. In Deinococcus radiodurans, G4 sequences present in the regulatory regions of various genes contribute to radioresistance [32]. Similarly, G4 present at the upstream of nasT in Paracoccus denitrificans is involved in nitrite assimilation [33].

All these reports demonstrate the pivotal role of G-quadruplex in human pathogens, and their conserved-ness suggests them as a promising drug target for both drug-susceptible and drug-resistant strains of pathogens. Therefore, a comprehensive study that discovers highly conserved G-quadruplex in the essential genes of S. enterica may provide a more suitable therapeutic approach for fighting against infection and overcome the emergence of the drug-resistant problem in this deadly pathogen.

In the present study, we sought to explore the highly conserved potential G-quadruplex forming sequences (SE-PGQs) in completely sequenced strains of S. enterica. Bioinformatics analysis revealed the presence of three SE-PGQs in three different gene locations of S. enterica genome. SE-PGQ-1 was found to be present in the regulatory region of mgtA, SE-PGQ-2 in the open reading frame of entA, whereas SE-PGQ-3 in the regulatory regions of malK and malE genes (Fig. 2a). In order to confirm the formation of G-quadruplex structure by SE-PGQs, CD, NMR, and EMSA were employed and to validate these SE-PGQs as a potential drug target, CD melting, and polymerase stop assays were performed. Disc diffusion assay and MTT assay confirmed the growth inhibition of S. enterica cells by G4 binders BRACO-19 (9-[4-(N,N-dimethylamino)phenylamino]-3,6-bis(3-pyrrolodinopropionamido) acridine) and 9-Aminoacridine. Further, Real-time PCR (RT-PCR) revealed the reduced expression of genes that harbor these SE-PGQs in either their coding region or regulatory region upon the treatment with both BRACO-19 and 9-Aminoacridine. mTFP reporter based assay further strengthens the role of SE-PGQs in the gene expression regulation. This change in expression of PGQ harboring genes in the presence of G4 binding ligand suggested a G4 mediated regulatory mechanism in the expression of these genes. Therefore, G-quadruplex motifs found in these genes can be utilized as a potential drug target to develop a promising anti-microbial therapeutics. Moreover, the high conserved-ness of these SE-PGQs, even in the drug-resistant strain would be able to overcome the problem of the emergence of drug-resistance in S. enterica.

Fig. 2

Representation and localization of three highly conserved PGQs. a) Diagrammatic representation of SE-PGQ-1 harbored at upstream of mgtA, SE-PGQ-2 in the open reading frame of entA gene, and SE-PGQ-3 in the intergenic regulatory region of malK and malE in the Salmonella enterica genome. b) Details of three most conserved PGQs, including location, Gene id, Gene locus, G-quadruplex strand, and gene direction in the reference genome of Salmonella enterica (CT18 strain). c) Consensus Sequence of the most conserved PGQs predicted by Glam2 tool of MEME Suite.

Result and discussion

Salmonella enterica genome harbors three most conserved G-quadruplexes

Since the last several decades, the scientific community has been witnessed a rapid increase in the number of such human pathogenic bacterial species that acquired resistance to multiple anti-bacterial agents. Currently, the emergence of multidrug-resistant strains remains a significant public health concern for clinical investigators that rings a global alarm to search for novel and highly conserved drug targets. Recently, conserved G-quadruplexes and their binding with small molecules are being extensively investigated as a promising therapeutic approach for combating the various type of human pathogenic infection [27,28]. Considering the suitability of G-quadruplex structure as a promising drug target in both drug-susceptible and drug-resistant strains of pathogens, here we sought to search for G-quadruplex motifs in S. enterica strains.

Comprehensive mining of potential G-quadruplex forming motifs (SE-PGQs) was performed on 412 completely sequenced strains of S. enterica (Supplementary data 1: Table S1). The bioinformatics analysis observed a total of 109,400 PGQs in 412 strains of S. enterica (Supplementary Dataset 2). Given that, a similar sequence may correspond to the same structure and evolutionarily conserved function, all the predicted PGQs were further clustered by Unweighted Pair Group Method with Arithmetic Mean clustering method using Clustal Omega tool. The conserved-ness is an essential parameter that makes these PGQ motifs suitable to work as promising drug targets. Therefore, next, we examined the conservation of each PGQ clustered using the following equation:where p is the frequency of occurrence, n = number of strains with specific G4 sequence, and N represents the total number of strains of S. enterica. These conservation analysis revealed 187 PGQ clusters that were observed to possess conversed-ness in more than 90% strains of S. enterica (Supplementary data 1: Table S2). G-quadruplex with loop length 1–7 and G tract of ≥3 forms a more stable G-quadruplex [34]. Therefore, for further study, we selected only those PGQs that satisfied the aforesaid criteria of G-quadruplex formation and were listed in Supplementary data 1: Table S3.

These predictions were further crosschecked with QGRS Mapper and PQSFinder (Supplementary data 1: Table S4 and S5). Interestingly, out of 18 PGQ clusters, three PGQs (SE-PGQ-1, SE-PGQ-2, and SE-PGQ-3) were found to be conserved in more than 98% strains of S. enterica (Supplementary data 1: Table S3) and present in the four essential genes namely mgtA, entA, malK and malE (Fig. 2a-c). The consensus sequence depicted the conserved Guanine residues of SE-PGQ motifs during the evolutionary process (Fig. 2b).

In vitro 1H NMR analysis affirms the formation of G-quadruplex in SE-PGQs

NMR spectroscopy is considered as the most reliable technique for confirming the formation of G-quadruplex structure formation by the nucleic acid sequences. Therefore 1D 1H NMR spectroscopy was performed to confirm the formation of G-quadruplex conformation by SE-PGQs. The presence of a chemical shift in the range of 10–12 ppm in 1D 1H NMR spectra depicts the presence of hoogsteen base pairing in characteristic G-tetrads of G-quadruplex structure whereas, canonical G-C Watson Crick base pairing can be characterized by a chemical shift in the range of 12–14 ppm. All the three SE-PGQs showed an imino proton resonance between 10 and 12 ppm and clearly affirmed the formation of G-quadruplex structure (Fig. 3 and Supplementary data 1: Fig. S2). Whereas, mutant sequence NMR spectra analysis depicted an absence of imino proton peaks between 10 and 12 ppm. (Supplementary data 1: Fig. S3).

Fig. 3

NMR Spectra. 1D 1H NMR spectra of SE-PGQ-1(mgtA), SE-PGQ-2(entA), SE-PGQ-3(malE and malK) in the presence of K+ Buffer.

Evaluating the topology and stability of the PGQs using circular dichroism

Circular dichroism is one of the widely used techniques for analyzing the topology of the G-quadruplex structure. G-quadruplex, depending upon its sequence, loop length, and bound cation, can form either a parallel, anti-parallel or hybrid conformation. A positive peak at ~260 nm and a negative peak at ~240 nm signifies for parallel G-quadruplex topology. However, a positive peak at ~290 nm and a negative peak at ~260 nm signifies for anti-parallel G-quadruplex topology. Two positive peaks at 260 nm and 290 nm with a negative peak at 240 nm depicts the mix or hybrid topology. Additionally, all the G-quadruplex conformers forms a positive peak at 210 nm [35]. Different cation affects the stability of the G-quadruplex structure to a different extent. The stabilizing ability of some well studied cations is ranked as follows: K+ > Na+ > Mg2+ > Li+ [20]. Therefore, we performed the CD spectroscopy of SE-PGQs in these four different cations (K+, Na+, Li+, and Mg2+) containing buffers (Fig. 4 and Supplementary data 1: Fig. S4).

Fig. 4

Circular Dichroism Analysis. a) Circular Dichroism spectra of the three most conserved PGQs in the presence of Tris-Cl buffer (10 mM) containing either of 50 mM K+ (red), 50 mM Na+ (green), 50 mM Li+(yellow), 50 mM Mg2+ (blue) or mutant of the same length(Pink). b) Melting spectra obtained by Circular Dichroism in different buffers (K+, Na+, Li+, and Mg2+) for conserved PGQs predicted in Salmonella enterica. In the absence of any buffer(Black), K+ 50 mM(Red), K+ 200 mM (Green), Na+ 50 mM (Yellow), Li+ 50 mM (Blue)and Mg2+ − 50 mM (Pink). c) Bar-graph is depicting Tm of SE-PGQ-1(mgtA), SE-PGQ-2(entA), SE-PGQ-3(malE, and malK) in the absence and presence of various cations. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

CD spectra analysis revealed the predominant parallel G-quadruplex topology exhibited by SE-PGQ-1 and SE-PGQ-3 in the presence of the K+ ion, whereas SE-PGQ-2 showed hybrid G-quadruplex topology (Fig. 4a). As expected, CD spectral scanning performed in the increasing concentration of K+ ion showed the maximum molar ellipticity in the highest K+ ion concentration (Supplementary data 1: Fig. S5). Fig. 4b shows the CD melting curve of SE-PGQs in various cation conditions. CD melting analysis revealed the higher stability of the SE-PGQs in the K+ ion concentration (Fig. 4b & c). To evaluate the significance of G-tracts for their G-quadruplex forming ability, the central Guanine was mutated to Adenine, and CD spectra analysis was performed in 50 mM K+ ion (Supplementary data 1: Table S6). Mutants (mut-PGQ-1, mut-PGQ-2, and mut-PGQ-3) failed to show the characteristic CD signal of G-quadruplex i.e., a positive band at 210/260/290 nm and a negative band at 240 nm suggesting the mutation in G tract disrupted the G-quadruplex formation (Fig. 4a).

Electrophoretic mobility shift assay (EMSA) supports intramolecular conformations of SE-PGQs

Next, Electrophoretic Mobility Shift assay (EMSA) was performed to check the molecularity (inter or intra molecular G-quadruplex) of SE-PGQs in the solution. An intramolecular G-quadruplexes possess a compact topology and migrate faster than their linear counterpart, whereas intermolecular G-quadruplex contains a comparatively wider topology and exhibited slow migration than their linear counterpart [36]. All three SE-PGQs and positive control (Tel22 DNA G-quadruplex) showed faster mobility than their respective linear counterpart. They, therefore, suggested the formation of intramolecular G-quadruplex by SE-PGQs (Supplementary data 1: Fig. S6).

G-quadruplex specific ligands inhibit Salmonella enterica growth

Similar to the G-quadruplex motif present in the genome of another human pathogen, a G-quadruplex motif present in mgtA, malK, malE, and entA genes of S. enterica strains may also serve as a potential target for developing anti-bacterial therapy. These G-rich targets can also overcome the problem of the drug resistance due to their high conserved-ness in both drug-susceptible and drug-resistant bacterial strains. Recently, conserved G-quadruplexes and their binding with small molecules are being extensively investigated as a promising therapeutic approach for combating the various type of human pathogen infection [27,28]. For example, HIV-1 promoter region possessed a G-quadruplex motif in the long terminal repeat (LTR) region of their genome and observed to be critical for its proliferation. BRACO-19, a tri-substituted Acridine derivative, has shown anti-HIV-1 activity by stabilizing the G-quadruplex motif present in the LTR region [27]. BRACO-19 is also reported to inhibit the viral multiplication in Adenoviruses and Herpes Simplex virus by stabilizing G-quadruplex motifs in their genome [30,37].

Similarly, stabilization of G-quadruplex structure present in the core gene of HCV genome by PDP, halts its replication, translation and, therefore, can be used as potential anti-hepatitis therapeutics [38]. Pyridostatin is also observed to stabilize the G-quadruplex structure formed in the mRNA of the nuclear antigen 1 protein of EBV, leading to its translation suppression [39]. Recently, BRACO-19 has shown an inhibitory effect on Mycobacterium tuberculosis, and Klebsiella pneumoniae, while Quarfloxin inhibited Plasmodium falciparum, by stabilizing G-quadruplexes present in various regions of their genome [25,40,41]. Previously, 9-Aminoacridine and its derivatives have been observed for their anti-proliferative properties in cancer cells [42] by binding to the telomeric region [43], the c-Myc gene [44] and c-Kit promoter [45]. Therefore, here we analyzed the effect of G-quadruplex specific ligands, BRACO-19, and 9-Aminoacridine on the S. enterica growth and performed MTT assay. An increase in color intensity with the decrease in the concentration of G4 specific ligands was observed giving maximum in control well (untreated) that suggested the inhibitory effect of both G4 ligands, BRACO-19 and 9-Aminoacridine on the S. enterica growth with an IC50 value of 15.877 μM and 10.5 μM respectively (Supplementary data 1: Fig. S7 and S8).

Acridine derivatives have a higher affinity towards Salmonella enterica PGQs

Isothermal titration calorimetry (ITC) is one of the most reliable techniques for molecular interaction analysis as it relies on the change in enthalpy (ΔH) and entropy (ΔS) upon the interaction between two molecules. Herein, to check the specificity and affinity of BRACO-19 and 9-Aminoacridine with the SE-PGQs, we performed the ITC experiment in 50 mM potassium phosphate buffer and fitted the obtained thermogram in the two-site binding model (Fig. 5 and Supplementary data 1: Fig. S9). A duplex DNA was taken as a control. Negative change in free Gibbs energy (ΔG) depicted the thermodynamically favorable interaction between BRACO-19/9-Aminoacrdine and SE-PGQs and thus the formation of a stable complex. Briefly, in the interaction analysis of SE-PGQs with BRACO-19, ΔH was observed to be −6.88 × 103 cal/mol, 4.49 × 106 cal/mol, and 3.14 × 102 cal/mol for SE-PGQ-1, SE-PGQ-2, and SE-PGQ-3 respectively. Negative ΔH for SE-PGQ-1 depicted that the reactions were enthalpy driven. Though SE-PGQ-2 and SE-PGQ-3 on interaction with BRACO-19 showed positive ΔH, a negative change in ΔG for both the reactions strengthens there biologically feasibility and were entropically driven. The first association constant (Ka1) for BRACO-19 was found to be 7.78 × 105 M−1, 2.73 × 106 M−1, and 9.37 × 105 M−1 for SE-PGQ-1, SE-PGQ-2, and SE-PGQ-3 respectively, which was 27.20, 95.5 and 32.76 folds higher than the duplex DNA (2.86 × 104 M−1) [Fig. 5 and Supplementary data: Table S7] Similarly, with 9-Aminoacridine, the association constant for SE-PGQ-1, SE-PGQ-2 and SE-PGQ-3 was −2.55 × 105 M−1, −2.11 × 104 M−1, and − 2.48 × 105 M−1respectively and were enthalpy driven (Supplementary data 1: Table S8). For 9-Aminoacridine, the association constant (Ka1) was found to be 61.99, 52.02, and 62.98 folds higher for SE-PGQ-1, SE-PGQ-2, and SE-PGQ-3 respectively than the association constant for the negative control duplex DNA (Supplementary data 1: Fig. S10 and Table S8). On comparative analysis, it was observed that the BRACO-19 showed the highest affinity for SE-PGQ-2 (2.73 × 106 M−1) while 9-Aminoacridine showed the highest binding against SE-PGQ-2 (2.81 × 106 M−1). In summary, ITC analysis depicted the higher affinity and specificity of BRACO-19 and 9-Aminoacridine with SE-PGQs as compared to that of the duplex DNA [46,47].

Fig. 5

ITC Analysis of SE-PGQs with BRACO-19. Isothermal titration calorimetry (ITC) binding isotherms for the interaction of BRACO-19 with SE-PGQ-1(mgtA), SE-PGQ-2(entA), SE-PGQ-3 (malE and malK) and Duplex DNA (Control). Raw data is shown in the upper panel and curve fit using a two-site binding model in the bottom panel. The graph represents one of the fitted curves out of the three repeats.

Acridine derivative stabilizes the SE-PGQs, and thereby stalled the movement of polymerase

To understand the role of SE-PGQs in the cytotoxic effect of acridine derivatives, binding affinity of BRACO-19 and 9-amino acridine with these SE-PGQs were analyzed by performing CD Melting studies. An increase in the melting temperature (ΔTm) was observed upon the addition of BRACO-19 compared with alone SE-PGQs (Fig. 6a). This indicated that BRACO-19 increased the thermodynamic stability of SE-PGQs (Fig. 6a). To check the stabilizing effect of BRACO-19 on the non G-quadruplex motifs, CD melting analysis was also performed on SE-PGQ-mutant sequences. The addition of BRACO-19 leads to a slight change in the Tm of the mutant sequences, depicting the higher affinity and stabilizing effect of BRACO-19 with SE-PGQ sequences as compared to that of their mutants. (Supplementary data 1: Fig. S10), CD melting analysis with 9-Aminoacridine was in line with that of BRACO-19 (Supplementary data 1: Fig. S11a). Further, we employed a PCR stop assay to investigate whether acridine derivatives complex formation with SE-PGQs, makes it possible to stop the movement of polymerase replication machinery or not. In order to investigate this hypothesis, we incubated PCR reaction mixture with BRACO-19 / 9-Aminoacridine in a concentration-dependent manner and then performed PCR amplification. We observed diminished intensity of bands with increase in concentration of acridine derivatives due to the formation of G-quadruplex motif leading to the unavailability of a template. However, in the absence of the BRACO-19 / 9-Aminoacridine, the band intensity was maximum indicated that the availability of template and Taq polymerase was able to extend the SE-PGQs motifs (Fig. 6b and Supplementary data 1: Fig. S11b). It shows that binding of the BRACO-19 and 9-Aminoacridine to the SE-PGQs motif stabilized the G-quadruplex structure and inhibited the movement of replication machinery over the untreated SE-PGQs. On the contrary, when mutant PGQs lacking G-tract were used as a DNA template, BRACO-19 and 9-Aminoacridine were not able to bind resulting in template availability and thus, could not inhibit the PCR amplification and produced a PCR product in the reaction(Fig. 6b and Supplementary data 1: Fig. S11b).

Fig. 6

Interaction of BRACO-19 with SE-PGQs. a) Circular Dichroism melting curve depicting the change in Tm of SE-PGQ-1 (mgtA), SE-PGQ-2 (entA), and SE-PGQ-3 (malE and malK) in the presence and absence of BRACO-19. b) Gel image of PCR stop assay for SE-PGQ-1(mgtA), SE-PGQ-2(entA), and SE-PGQ-3(malE and malK) and linear DNA with the increasing concentrations of BRACO-19.

BRACO-19 inhibits the reporter gene expression by stabilizing the PGQ motifs

Highly stable G-quadruplexes in the regulatory regions and open reading frames of genes were previously shown to down-regulate the expression of genes [[48], [49], [50]]. Therefore, to analyze the effect of BRACO-19 on the translation of the harboring genes, mTFP based reporter assay was performed. The three SE-PGQ motifs were inserted in the open reading frame of teal fluorescence protein (mTFP) [Supplementary data 1: Fig. S12]. The plasmid was constructed with SE-PGQ motifs at the upstream of mTFP gene and transfected in HEK293 genes. On treating the transfected HEK293 with BRACO-19, a decrease in the fluorescence intensity was observed in comparison to the untreated cells depicting the stabilization of G4 motifs. An increase in the BRACO-19 concentration resulted in a further decrease of mTFP expression. This decrease in the mTFP expression shows the stabilization of G-quadruplex motifs by BRACO-19. The G4_mutant plasmid transfected HEK293 cells showed an almost similar level of mTFP expression in both treated and non-treated conditions depicting the G-quadruplex binding specificity of BRACO-19 (Supplementary data 1: Fig. S12).

9-Aminoacridine and its derivative decrease the transcription rate of the genes harboring PGQs

Furthermore, we performed qRT-PCR assay to check the effect of BRACO-19 and 9-Aminoacridine on the expression of the PGQs harboring genes. The transcripts of the genes were quantified with respect to the 16 s rRNA gene, and the fold change in the expression of the transcripts was analyzed for the treated cells in comparison to non-treated cells (control culture). The samples treated with 15.877 μM BRACO-19 upon analysis showed 6.28, 3.42, 6.03, and 6.22 fold change in expression for mgtA, entA, malE and malK, respectively (Fig. 7 ). Similarly, the change in the gene expressions was also analyzed in the presence of another acridine derivative, 9-Aminoacridine. Inline with BRACO-19, 9-Aminoacridine reduced the rate of transcription of mgtA, entA and malK by 1.86, 3.03, and 2.94 fold, respectively (Supplementary data 1: Fig. S13). malE gene showed the highest suppression by 7.16 fold decrease in its expression in the presence of 9 amino acridine. In conclusion, all the four genes showed a decrease in the expression level, thus strengthening the G4 mediated inhibition mechanism in their promoter/regulator or open reading frame region (Fig. 7 and Supplementary data 1: Fig. S12).

Fig. 7

RT-qPCR. The normalized fold change of mgtA, entA, malE and malK transcripts in Salmonella enterica determined by quantitative PCR in the presence of BRACO-19.

Highly stable G-quadruplexes in the regulatory regions and open reading frames of genes were previously shown to down-regulate the expression of genes. The treatment of S. enterica cultures with Acridine derivatives, BRACO-19, and 9-Aminoacridine led to a decrease in the expression of entA, mgtA, malK and malE genes, suggesting a G-quadruplex mediated inhibition mechanism is involved in this process. As a schematic model elaborated in Fig. 8 , G-quadruplex mediated inhibition of entA, mgtA, malK, and malE genes is expected to increase the innate immune response of host cells and reduced survival of bacterium inside the host cell. The Inhibited expression of the entA and mgtA proteins would make the bacteria unable to respond against Reactive oxygen/nitrogen (ROS/RNS)species and reduced their survival inside the host macrophages.

Fig. 8

Effect of Acridine derivatives on mgtA, entA and maltose operon mediated mechanisms. Schematic representation of G-quadruplex loci and the effect of their stabilization with BRACO-19 or 9-Aminoacridine on the survival and virulence of Salmonella enterica.

As a concluding remark, the current study shows the presence of stable and highly conserved G-quadruplex structures in essential genes of Salmonella enterica. The present study is a proof-of-concept analysis for the identification and characterization of G-quadruplex motifs in the essential genes of Salmonella enterica. These SE-PGQs can provide a novel platform for therapeutic development against the infection of Salmonella enterica. They can be targeted in both antibiotic susceptible and antibiotic-resistant strains due to their conserved nature throughout the Salmonella enterica genus. Most of the commonly used G-quadruplex binders share a common structural feature of a planer hetero-aromatic chromophore that helps in π-π interaction with the G4 motif. A short alkyl chain substituents in this heterocyclic ring, usually terminated by an amino group, enhances the G4 specificity to a large extend [51]. 9-Aminoacridine satisfies this condition of the hetero-aromatic chromophore and can be substituted with the desired functional groups. This has been widely exploited for the synthesis of numerous di-substituted and tri-substituted Acridine Derivatives that have shown very high specificity towards G-quadruplexes as compared to the duplex DNA [42,43,45,[52], [53], [54], [55], [56], [57], [58]]. Thus, this Acridine scaffold can be used as a starting point for the synthesis of novel G4 specific ligands with minimum cytotoxicity to the host cells. Herein, two Acridine derivatives, BRACO-19 and 9-Aminoacridine were observed to bind and reduced the expression of these G-quadruplex structures possessing genes and thereby proposed as novel G4 mediated therapeutic approach for combating the infection of Salmonella enterica in humans.

Materials and methods

Prediction, conserved motif enrichment and functional analysis of G-quadruplex motifs in Salmonella enterica strains

Completely sequenced strains of S. enterica (Supplementary data 1: Table S1) were downloaded from National Center for Biotechnology Information (NCBI). These strains were then extensively mined for the potential G-quadruplex motifs in both sense and antisense strand using our previously developed G-quadruplex predictor tool [59]. This prediction tool used the following regular expression.where T1 represents consecutive tracts of Guanine that can be any number from 2 to 7, X is any nucleotide (A, T, G, C), L1, L2, L3 represents the variable loop region and can be any number from 1 to 20. For our prediction, we used G-tracts – 3 or 4 and loop length 1–20 nucleotide [59]. The results were further cross-verified by using QGRS Mapper [60] and PGSFinder [61] tools.

To find the conserved PGQ's that are available in all the strains, multiple sequence alignment (MSA) was performed by using Clustal Omega tool, and clustering was done using UPGMA method. Consensus sequences representing the whole G4 sequence with −5 and + 5 flanking regions were constructed using the Glam2 tool of MEME Suite [62].

The resultant PGQ clusters were then mapped for their gene location in the genome of the individual S. enterica strains using the coordinates extracted from our G4 prediction tool by using Graphics mode of GenBank Database (https://www.ncbi.nlm.nih.gov/nuccore/).

Salmonella genus G4 homolog prediction

In order to check the conservation of predicted PGQs at the Salmonella genus level, NCBI nucleotide BLAST was performed by taking each consensus PGQ as a query sequence and Salmonella bongori genome sequences as a target (NCBI taxid: 590). The threshold e-value was set as 1e-3 to remove any results by chance.

Oligonucleotide preparation for CD and ITC analysis

Predicted G4 oligonucleotides sequences were procured from Sigma Aldrich Chemicals Ltd. (St. Louis, MO, USA). 100 μM stock solutions were prepared as per the manufacturer's instructions. Before each set of experiments, oligonucleotides were subjected to re-anneal by heating at 95 °C for 10 min and slow cooling at room temperature for 2 h. All these oligonucleotides were dissolved in four different Tris-buffer (pH = 7.0, 10 mM) containing 50 mM of K+, Na+, Li+ and Mg2+ separately.

CD spectroscopy and melting analysis

CD experiments were performed for each oligonucleotide in different buffer conditions at 25 °C with a scanning rate of 20 nm/min from 220 to 320 nm using Jasco J-815 Spectropolarimeter (Jasco Hachioji, Tokyo, Japan). The instrument was equipped with a Peltier Junction temperature controller. Spectra were recorded in a cuvette of 1 mm path length for a final concentration of 20 μM of all oligonucleotides in Tris buffer (10 mM, pH = 7.4) containing four different cations viz. K+, Na+, Mg2+ and Li+ (50 mM each) in separate experiments. To avoid signal contribution from the buffer, a blank spectrum containing only buffer was recorded before each measurement and subtracted from CD spectrum of the sample prepared in the respective buffer.

CD Melting analysis was performed for every PGQs (final concentration of 20 μM) with a temperature range of 25 °C to 98 °C in four different cations (K+, Na+, Mg2+ and Li+, 50 mM each) containing buffers. The heating rate was set as 1 °C/min for each melting experiments

Electrophoretic mobility shift assay

Native PAGE was run using 20% polyacrylamide gel in 1× TBE buffer. Each sample was dissolved in Tris buffer(pH = 7, 10 mM) containing four different cations K+, Na+, Mg2+, Li+ (50 mM each) separately. For each PGQ, an oligonucleotide of similar length (G mutated with T nucleotide) was taken as a negative control, and standard G-quadruplex (Tel22 DNA) was taken as a positive control. 20 μL of each oligonucleotide sample were loaded, and electrophoresis was performed at 40C, 90 Voltage in a vertical gel unit system. The gels were visualized by staining with ethidium bromide and analyzed on ImageQuant LAS 4000 gel doc (GE Healthcare Biosciences Ltd., Sweden).).

One dimensional 1H NMR spectroscopy

AVANCE 500 MHz BioSpin International AG, Switzerland equipped with a 5 mm broadband inverse probe was used to perform NMR spectroscopic analysis. The SE-PGQs were dissolved in 50 mM potassium phosphate buffer. All the NMR experiments were performed using H2O/D2O solvent at 9:1 ratio. Temperature of 298 K with 20 ppm spectral width and 3 - (Trimethylsilyl) propionic-2, 2, 3, 3-D4 acid sodium salt (TSP) as an internal reference were used. NMR data processing, integration, and analysis were done by using Topspin 3.1 software.

Isothermal titration calorimetric (ITC) analysis

ITC analysis was performed for all three PGQs and a linear DNA (negative control) using MicroCal iTC200 calorimeter (GE Healthcare, Biosciences Ltd., Sweden. All the oligonucleotides were dissolved in 50 mM potassium phosphate buffer at the final concentration of 20 μM. The oligonucleotide was used in the cell of ITC and 200 μM of 9-Aminoacridine and BRACO-19 dissolved in the same buffer were titrated 21 times by using syringe with the initial injection of 0.4 μL followed by 1.8 μL of the ligand at each step for the duration of 3.6 s with the decay of 90 s between each step. The heat of dilution of the oligonucleotides was determined by adding the same amount of 9-Aminoacridine/BRACO-19 into the 50 mM potassium buffer and was subsequently subtracted from the binding isotherms of the oligonucleotides before the curve fitting analysis. Data were acquired in triplicates and was analyzed or the determination of association constant by its two-site binding mode analysis using origin scientific software version 7 (Microcal Software Inc. Northampton, MA, USA).

Bacterial strain culture and growth conditions

The S. ser. Typhimurium strain ATCC 14028 was procured from HiMedia and streaked on Nutrient Agar (HiMedia). A single colony was inoculated in the in Nutrient Broth (HiMedia) and kept overnight at 37 °C and 220 rpm in the incubator shaker [63]..

MTT growth inhibition assay

MTT assay was performed for cytotoxic analysis of 9-Aminoacridine and BRACO-19 on S. enterica. 50 μL of the overnight grown culture of S. enterica, was inoculated in 5 mL Nutrient Broth (NB) at 220 rpm, 37 °C and allowed to grow till the O.D600 = 0.5. After that, 50 μL was transferred in a fresh 5 ml NB tube, and 100 μL was transferred in each well of 96 well plates. Dilution was prepared from the stock solution (200 μM) of 9-Aminoacridine/BRACO-19 of the following concentrations 100 μM - 0.09 μM and added to the respective wells, last well served as blank (without 9-Aminoacridine). The plates were kept at 37 °C, 220 rpm for 3 h. Afterward, 10 μL of MTT (5 mg/mL) was added to each well and incubated for 3 h. Finally, 20 μL of DMSO was added in each well to dissolve formazan crystal, and the plate was examined using a microplate reader (BioTek) at 590 nm [64].

PCR stop assay

Templates and Primers were procured from Sigma-Aldrich Chemicals Ltd. (St. Louis, MO, USA) (Supplementary data 1: Table S9). The reaction was performed in the master mix consisting of 1× PCR buffer, 0.33 mM dNTPs, 2 μM templated, 2 μM reverse primer, 2.5 units Taq DNA polymerase (Sigma-Aldrich Chemicals Ltd. St. Louis, MO, USA) and dose titration of BRACO-19 and 9-Aminoacridine. The following thermal cycling conditions were used: initial denaturation at 95 °C for 5 min, followed by 25 cycles of 95 °C for 30 s, 50 °C for 30 s, 72 °C for 0.5 min and finally held at 4 °C. The 6× DNA loading dye was added to the amplified products and resolved in 3% agarose gel. Ethidium bromide was used for staining, and gel images were analyzed using ImageQuant LAS 4000 (GE Healthcare, Biosciences Ltd., Sweden).

mTFP reporter based assay

Reporter based assay was performed by utilizing a revised protocol described elsewhere [65]. Briefly, G-quadruplex harboring mTFP plasmid was constructed by inserting the G4 sequences and a mutant sequence just before the start codon of mTFP coding region in pCAG-mTFP plasmid [a generous gift from Dr. Debasis Nayak, IIT Indore]. Insertion was performed using an Overlap extension PCR based cloning strategy with overlapping forward and reverse primers. The constructed plasmids were transfected in HEK293 cell lines using Lipofectamine2000 as per the manufacturer's protocol. The transfected cells were incubated for 4 h, followed by the treatment of 10 μM and 20 μM BRACO-19. The treated cells were further incubated for 24 h and then visualized under Fluorescence microscope for mTFP expression analysis. ImageJ software was used for comparative intensity analysis.

Total RNA isolation and cDNA synthesis

S. enterica was grown overnight at 37 °C, 220 rpm in 5 mL of a test tube containing NB. Two flasks of 500 mL containing 50 mL of NB each were inoculated with 1% inoculum from overnight culture. Both the flasks of untreated and treated cells were allowed to grow according to previously mentioned conditions till the O.D600 nm reached 0.5. At O.D600 = 0.5, 20 μM of BRACO-19 (dissolved in water) and 10.35 μM of 9-Aminoacridine (dissolved in DMSO) were added to MIC flask. For 9-Aminoacridine treatment assay, control culture DMSO was added as 9-Aminoacridine was dissolved in it. All the flasks were kept at 37 °C, 220 rpm for 45 min. Subsequently, samples were centrifuged at 12000 rpm and immediately preceded for RNA isolation. S. enterica culture prior to RNA isolation was treated with RNA protect reagent (Qiagen, USA) to stabilize RNA and prevent it from degradation. Treated and untreated Salmonella cultures were used for total RNA isolation. RNA isolation for all samples was carried out using TRIZOL reagent (Invitrogen) according to manufacturer's instructions. After RNA isolation, its concentration and purity were measured using NanoDrop (Thermoscientific) as ng/μL and A260/A280, respectively. Subsequently, all the RNA samples were treated with RNase-free DNaseI (Invitrogen) as mentioned by the manufacturer. Finally, 5 μg of total RNA from each sample was used for cDNA synthesis. cDNA synthesis was performed in a 20 μL reaction by using Invitrogen Superscript IV kit (Cat # 18091050) according to the manufacturer's instruction.

Gene expression profiling using real-time quantitative PCR

Quantitative Real-Time PCR was used to elucidate the variation in gene expression profiling of Salmonella treated with 9-Aminoacridine and BRACO-19. All the aliquots of cDNA from untreated (control) and treated (9-Aminoacridine/BRACO-19) were used in real-time PCR. qPCR was carried out in PCR master reaction mix containing 1× PowerUp™ SYBR® Green Master Mix (Applied Biosystem, USA), 0.5 μM of each primer and 11 μL of cDNA sample in a final reaction volume of 25 μL in a 96 well PCR plate in Step One Plus (Applied Biosystem, USA) Thermal Cycler. All the samples for real-time were analyzed in three dilutions and duplicates; normalization was done with respect to 16 s rRNA (housekeeping gene). The proportionate change in gene expression was assessed by change in expression of the target gene in treated as compared to control. List of primers used in real-time PCR and thermo-cycling conditions are mentioned in Supplementary data 1: Table S10 and S11, respectively. Briefly, the thermo-cycle used in qRT-PCR was 94 °C for 2 min, subsequently 40 cycles of 94 °C for 15 s and 57 °C for 1 min.

Authors contribution

Data conceptualization and methodology were performed by A.K. Bioinformatics analysis was done by N.J., and U.S. Biophysical analysis was performed by N.J., U.S.. N.J. and A.J. performed MTT assay. A.J., S.K.M. and N.J. performed RT-PCR under the supervision of P.K., and A.K.. N.J., and U.S. wrote the manuscript. A.K. T.K.S. and P.K. did the review and editing.

Funding

The work was supported by the , Govt. of India [EMR/2016/3897 to AK], [EMR/2016/003547 to P.K.], and Department of Biotechnology, Govt. of India [BT/PR20319/BBE/117/189/2016 to P.K.].

Declaration of Competing Interest

The authors declare no conflict of interest.