Managing gene expression in Pseudomonas simiae EGD-AQ6 for chloroaromatic compound degradation

Pseudomonas simiae EGD-AQ6 is capable of utilizing chloroaromatic compound i.e., 2–4-D efficiently in its biofilm phenotype. The differential accumulation of intermediate 4-chlorocatechol rates were significant in planktonic and biofilm phenotypes, as well as in the increased biofilm adapted cell numbers. Interestingly, response surface analysis demonstrated the combined positive effects of 2–4-D degradation and 4-CCA accumulation rates and the gene expression profiles, with significant up-regulation of degradative and biofilm genes, and greater participation of pellicle genes in the biofilm phenotypes than their planktonic counterparts, thereby revealing a phenotype variation. It positively validated the physiological data. Furthermore, the sequence similarity of the 2–4-D catabolic and biofilm-forming proteins (pel ABCDEFG and pga ABCD), which are responsible for building carbohydrate rich extracellular matrix, were significant with the respective organisms. This is the first study, which endorses this strain to be unique in efficient chloro-aromatic degradation through phenotype variation, thereby proving a potential candidate in the improvement of bioremediation technologies.


Introduction
Chloro-aromatic compounds are released to the environment through runoffs from solvents, dielectric hydraulic fluids, and in the form of pesticides as agrochemicals. It causes significant pollution of soil and groundwater throughout this time (Field et al., 1995). 2-4-dichlorophenoxyacetic acid (2-4-D) belongs to the group of chlorophenoxy herbicides widely used as plant growth regulators, most often for the treatment of tomatoes (Kumar et al. 2016). It is significantly toxic, mutagenic, and carcinogenic effects on animals and humans (Zharikova et al. 2018). Although many chlorophenoxy herbicides are recalcitrant, there are bacteria capable of aerobic degradation (Feld et al. 2016). Nevertheless, 2-4-D is used as a model chloro-aromatic compound to investigate the role, on account of its evolution, acute toxicity, and prolonged pollution history in the contaminated niches (Song 2014). But the degradative ability of the authochthonous or augmented planktonic bacteria often declines due to suboptimal conditions such as temperature, pH, partial pressure of oxygen, toxic substances, resulting in diminished bacterial activity and viability which impedes degradation (Ghosh et al. 2019b). However, biofilm-mediated remediation is inescapably efficacious and gaining interest worldwide. The spatial structure in a biofilm matrix grants heterogeneous microenvironments, sharing of intermediates, horizontal gene transfers, maintains optimal physical conditions due to mass transfer balance (Palmer et al. 2007;Flemming et al. 2016;Abe et al. 2020). This ensures enhanced bioavailability of substrates and deterrence against stress to the underlying members in the biofilm, leading to the evolution of altered physiology in a population of the same strain as well as in a community (Ghosh et al. 2017b;Tsang and Simpson 2020). The cells are generally characterized by one or more phenotype(s), each with distinct degradative functionality (Ghosh et al., 2017a). In addition, the variety of phenotypes with distinct characteristics can often be observed depending on the presence of multiple gradients i.e., nutrient and oxygen within the biofilm microenvironment (Caro-Astorga et al. 2020). This is implicated to variety of metabolic options via sharing of substrates and intermediates, leading to elevated utilization potential of the toxic aromatic substrates, without inhibiting the underlying cells (Flemming et al. 2016;Ghosh et al. 2017a). In view of above literature reports, this study is the first to report the influence of such phenotypic variation towards 2-4-D degradation and conditional accumulation of intermediate i.e., 4-chlorocatechol (4-CCA). Also, combined positive effects of the 2-4-D and 4-CCA utilization rates on differential expression of biofilm matrix and degradative genes were observed with escalated up-regulation in the biofilm adapted phenotypes comparative to its planktonic counterpart through response surface analysis. Furthermore, the sequence similarity of the biofilm and 2-4-D catabolic genes in P. simiae EGD-AQ6 were found significant with the respective bacteria, thereby advocating the strain to be an efficient candidate in the development of bio-augmentation technologies.

Biofilm formation analysis by Pseudomonas simiae EGD-AQ6 in media amended with 2-4-D
To analyze biofilm formation by Pseudomonas simiae EGD-AQ6, 0.01O.D 600 /ml cells were inoculated in NMM medium supplemented with 2-4-D (44.2 to 442 mg/L) for shaken and static flask assays (Verhagen et al. 2011;Ghosh et al. 2017a). Crystal violet binding assay in 6 ml microplates was performed as described in Ghosh et al. (2017a) after every 24 h of interval upto 288 h under static and shaking conditions. For the enumeration of attached cells, standard curve with a positive correlation between crystal violet retention O.D and CFU count was determined (r 2 = 0.965 at p < 0.05) as per the method described in Ghosh et al. (2019a). The rate of biofilm formation were also determined according to Ghosh et al. (2019a). The suspended cultures from the microplates i.e., unattached cells (planktonic) growth rate was determined as per Wood et al. (2019). µ = growth rate, O.D t and O.D t0 are the respective optical densities taken at times t and t 0.

Analytical methods
To analyze 2-4-D utilization and intermediate metabolite formation high-performance liquid chromatography (HPLC) was performed. The supernatants from static and shaking cultures were obtained after every 24 h interval, which were filtered through a 0.22 µm nylon filters (Randisc, Inc). Filtrates (20 µl) were injected through a Lichrosorb RP18, 5 mm (250 X 4.6 mm) column. Metabolites were detected in 284 nm with an isocratic flow rate of 0.6 ml/min in Waters PDA detector (Waters, PDA LC) with LC pump 5000 using mobile phase (80% Acetonitrile + 19% Acetic acid + 1% water). The 2-4-D degradation and 4-CCA accumulation rates were determined according to the Oberoi et al. (2015).

Gene expression analysis: qRT-PCR
The expression of respective genes i.e., biofilm (pelA, pgaA) and catabolic genes (arhd, tfdC) (Franklin et al. 2011;Camara et al. 2009) were obtained under static and shaking conditions. Real-time primers were designed (Table. S1) using DNA star (Lasergene, USA) and checked for their specificity by PCR amplification (data not shown). Total RNA extraction, qRT-PCR and relative quantification of EGD-AQ6 cells were performed as described by Ghosh et al. (2017a) using the 2 ΔΔCt method using rpoB as a housekeeping gene (Livak et al. 2013).

Design of experiments: Response surface methods (RSM)
RSM approach was used to test the effect of 2-4-D degradation and 4-CCA accumulation on the expression of biofilm and catabolic genes (pgaA and pelA; arhd and tfdC) under static and shaking conditions. Here, 2-4-D degradation rate (factor A), 4-CCA accumulation rate (factor B) were regarded as two input variables to study their effect on four response functions pgaA, pelA, arhd, tfdC expression rates separately under static and shaking phases. The optimal ranges of the two factors used were based on preliminary experiments ( Fig. 1) to determine the range of values of parameters for effective responses. RSM with central composite design (CCD) was used as a factorial experimental design in this study. Design of experiments, mathematical modeling, and optimization was performed by Minitab v19 (trial version) software. Here, the NMM media were amended with the amounts of 2-4-D and 4-CCA for the four responses, i.e., on the above-mentioned gene expression rates up to 144 h according to the combinations suggested by the software. Calculations and CCD analyses were then performed. For two variables (n = 2) and five levels [low (-1), high (1), 0, + α, -α] as given in Table 1, the total number of the experimental run was 13 as determined by the software with 4 factorial points, 4 axial points and only 5 center points (five replications) as given in Table S2. The fitted regression model was checked using regression analysis and ANOVA. The significance and robustness of the model were determined using (R 2 -Value), F-values, and p-values. The response variables were considered to be significant when p < 0.05. In general, the response for the quadratic polynomials is described below: where Y = response, X i and X j are the variables, β 0 is the constant coefficient, βii, βi, and βij are the coefficients of the quadratic, linear, and interaction effect, and k is the no of factor variables.

Pseudomonas simiae EGD-AQ6: Whole genome analysis
The whole-genome sequences of Pseudomonas simiae EGD-AQ6, was assembled as per Ghosh et al., 2017a, b and submitted to NCBI under accession no: AVQG00000000. NCBI has previously characterized the strain to be P. fluorescens EGD-AQ6. However, the taxonomic statistics of the strain were updated based on the ANI index which 99.4% identical to P. simiae than 88.4% with P.fluorescens. Hence, NCBI has renamed the strain as Pseudomonas simiae EGD-AQ6. Here in this report, the genome sequences were selectively analyzed for sequence similarity of biofilm and 2-4-D catabolic genes with the reported organisms. (2)

2-4-D degradation and biofilm formation analysis of P. simiae EGD-AQ6 under static and shaking phases.
The influence of specific growth conditions renders bacteria transfigure into variable phenotypes (Mangwani et al. 2015). These are adaptation strategies used to defend against the stress imposed (Jefferson et al. 2004). Reports point out that static and shaking conditions allow bacteria to evolve into different phenotypes such as biofilm (attached) and planktonic (unattached) cells, which influence aromatic degradation efficiency and other biotechnological reactions (Verhagen et al. 2011;Mangwani et al. 2015;Ghosh et al. 2019a). This was observed in our study; as shown in Fig. 1a, the rates of 2-4-D degradation were similar in both static and shaking conditions. On the contrary, 4-chlorocatechol (4-CCA) accumulation was significantly different under both the phases, where the culture supernatants of cells adapted under shaking conditions showed significant 4-CCA accumulation (0.014-0.096 mg −L /hr) than the cells evolved in static conditions (0.0011-0.0026 mg −L /hr) at all 2-4-D concentrations (Fig. 1b). The cell adherence rates were found to be elevated (Fig. 2a) Alternatively, the planktonic cell growth rates were found to be signifcantly lower, thereby suggesting a phenotype switch in the biofilm associated cells (Fig. 2b). A similar degradation trend was observed during the phenol and benzoate biodegradation by Pseudomonas sp. The intermediate catechol accumulated significantly higher in the planktonic cells than the biofilmadapted phenotype of the same strain (Yong and Zhong 2013). Benzoate biodegradation at elevated concentrations (40 mM), catechol accumulation was minimum in the culture supernatants of biofilm-associated cells, which comply with our results (Ghosh et al. 2017a).

Quantitative expression of biofilm and catabolic genes: Response surface analysis
The aromatic stress in the chemical surroundings influences phenotype transitions through differential gene expression (Jefferson et al. 2004;Flemming et al. 2016). To test this possibility in our strain, RSM was used to check the combined effects of factors, 2-4-D degradation (A), and 4-CCA accumulation (B) rates on expression rates of biofilm matrix (pelA, pgaA) and catabolic genes (arhd, tfdC) (responses) separately under static and shaking conditions. Real-time PCR analysis showed differential expression levels of the biofilm matrix and catabolic genes in biofilm and planktonic adapted phenotypes, as shown in the contour plots (Fig. 3, Fig. 4), which were similar to the predicted values calculated by the Minitab software (Table: S2). The calculated R 2 > 0.90 in all the genes indicated that the predictions of the response function were in accordance with the experimental ones at the confidence level of higher than 90% (See ANOVA Table.2). The absolute coefficients of A and B were significantly higher in pelA, pgaA, arhd, and tfdC expression in biofilm-adapted phenotype than its planktonic counterpart (Eg.3a-3 h). The p-value for interaction terms is < 0.05 stipulating their significant effect on the regression model. The contour plots illustrate the interaction effect between independent variables on the responses, where the favorable regions for catabolic and biofilm matrix genes were significantly more in the biofilm phenotype than the planktonic counterpart. It can be observed that expression rates of biofilm matrix genes (pelA and pgaA) had a maximum value of 0.09, 0.015, fold change/hr at the high levels (+ 1) of A 1 (1.46) and B 1 (0.00461), respectively (Table: S2). (Table 2). Likewise, the favorable regions for catabolic genes were comparatively different in both the phenotypes, with tfdC showing a noteworthy response in the biofilm phenotype with expression rate ~ 0.08 fold change/hr. It appears that, as the 4-chlorocatechol accumulation rate is lowered, tfdC expression rate increases indicating its utlization in the biofilm-adapted cells (Table S1; S2). For every 0.008 fold change/hr of arhd expression, tfdC expression rate was increased by ~ 2.8 times, as that of pelA and pgaA by ~ 9.3 and 6.1 times (Table: S2). On the contrary, this trend is reversed in planktonic associated cells (Fig. 4). At high levels (+ 1) of A 2 (1.36) and B 2 (0.132), the expression rates for arhd, tfdC, pelA and pgaA were 0.03, 0.06, 0.0007, 0.02. It indicates the decline in tfdC expression rate by 0.2fold change/hr. The declination of matrix genes expression rates was significant by 73 and 1.29 times comparative to arhd expression. The interaction between the factors A1 and B1 on the expression of pelA and pgaA were significant, which culminates in the involvement of 2-4-D and 4-CCA degradation in the building of biofilm matrix. Similarly, at low levels (-1) of A 2 and B 2 , the catabolic and biofilm matrix genes expression rates were significantly dissimilar in both the phenotypes of EGD-AQ6. It also corroborates to the greater participation and importance of the matrix protein pellicle than pgaA in building the biofilm at increasing chloroaromatic compound stress. It is similar to the results obtained by Meliani and Bensoltane (2014), indicates that pellicle synthesis was increased during xylene degradation and conferred an increased biofilm mass to P. fluorescens and P. aeruginosa. It indicates the combined impact of toxic metabolites, leading to phenotype switch of the strain EGD-AQ6 to combat the stress imposed, resulting in elevated biodegradation potential (Fig. 3). In this way, we have managed to unravel the gene expression patterns, regulating the 2-4-D biodegradation in both the phenotypes. Similarly, Yong and Zhong. (2013) reported that a different central cleavage pathway was initiated on account of a phenotype switch during phenol degradation. Low values in lack of fit indicate the involvement of other genes despite from degradation and biofilm genes in the biodegradation process. As reported, diverse catabolic, matrix synthesizing pathways and signal transduction mechanisms are involved in the phenotypic switch that allows a bacterium to attain physiological and metabolic heterogeneity for executing a particular metabolic process. Furthermore, these pathways also appear to be interconnected. Therefore, it is very likely that depending on the prevailing conditions, a number of variants may appear simultaneously in the same strain (Allison et al. 2011;Amato et al. 2013). In addition, this invulnerability of the underlying cells in the biofilm is due to diffusion limitations imparted by local variations in pH, nutrient and oxygen availability, and concentrations of bacterial metabolites leading to physiological heterogeneity, slow growth (Sauer et al. 2002: Jefferson et al. 2004Flemming et al. 2016). It further influences the activity of dioxygenases through regulating the partial pressure of oxygen in the fluid surroundings (Ding et al. 2008), resulting in differential gene regulation leading to phenotype transition (Paliwal et al. 2014). The interactive effect of the factors on the two responses (Table: S2; Table 2) was statistically quantified, and the approximate functions for attached cells (Eq. 3a, 3c) and planktonic cells (Eq. 3b, Eq. 3d) under static and shaking conditions was obtained as follows.  Adj R 2 adjusted R-squared; Taken together, the gradient-based physical and chemical microenvironments imparted by the matrix to the inhabitants (Palmer et al., 2007;Ghosh et al., 2017b) may have regulated the catabolic activity through oxygen fluctuations and additional alterations in gene expression in the biofilm. It could be a possible explanation for the results obtained in our study.

Conclusion
In conclusion, Pseudomonas simiae EGD-AQ6 showed phenotype variation towards degradation and utilization of toxic chloro-aromatic compound, 2-4-D. Biofilm phenotype showed comparatively less intermediate 4-CCA accumulation than its planktonic counterpart. Furthermore, RSM studies indicated the combined effect of 2-4-D and 4-CCA utilization in the increased expression of biofilm and degradative genes with a relatively different expression profile in the biofilm and planktonic phenotypes. Greater up regulation of biofilm matrix genes, particularly pelA was observed in biofilm-adapted cells. Sequence similarity analysis further validated the physiological and RSM data. This is the first report to exhibit efficient chloro-aromatic degradation through phenotype evolution, thereby envisaging Pseudomonas simiae EGD-AQ6 to be an effective candidate in bioremediation technology.