Synthetic extreme environments: overlooked sources of potential biotechnologically relevant microorganisms

Summary Synthetic extreme environments like carwash effluent tanks and drains are potential sources of biotechnologically important microorganisms and molecules which have, however, remained unexplored. Using culture‐ and molecular‐based methods, a total of 17 bacterial isolates belonging to the genera Shewanella, Proteus, Paenibacillus, Enterobacter and Citrobacter, Aeromonas, Pseudomonas and Pantoea were identified. Hydrocarbon utilization and enzyme production screening assays showed that Aeromonas sp. CAC11, Paenibacillus sp. CAC12 and Paenibacillus sp. CAC13 and Citrobacter sp. PCW7 were able to degrade benzanthracene, naphthalene and diesel oil, Paenibacillus sp. CAC12 and Paenibacillus sp. CAC13 could produce cellulase enzyme, while Proteus sp. BPS2, Pseudomonas sp. SAS8 and Proteus sp. CAL3 could produce lipase. GC‐MS analysis of bacterial secondary metabolites resulted in identification of 107 different compounds produced by Proteus sp. BPS2, Paenibacillus sp. CAC12, Pseudomonas sp. SAS8, Proteus sp. CAL3 and Paenibacillus sp. CAC13. Most of the compounds identified by both GC‐MS and LC‐MS have previously been determined to have antibacterial, antifungal and/or anticancer properties. Further, microbial metabolites which have previously been known to be produced only by plants or microorganisms found in natural extreme environments were also identified in this study. This research has revealed the immense bioresource potential of microorganisms inhabiting synthetic extreme environments.


Introduction
Microbial life has been found to adapt to and thrive in extreme conditions ranging from extremely low water availability, intense solar radiation, high salinity and extreme temperatures, pH and pressure (Rampelotto, 2013;Azua-Bustos and Gonz alez-Silva, 2014). Natural extreme environments can be defined as habitats that experience steady or fluctuating exposure to one or more environmental factors such as salinity, osmolarity, desiccation, UV radiation, barometric pressure, pH and temperature (Seufferheld et al., 2008). Some extreme environments are, however, synthetic (Maes et al., 2016) and may be created by a sustained and consistent discharge of pollutants (like acid mine drainage, untreated municipal sewage effluents and industrial effluent discharges among others) into the environment (Selbmann et al., 2013), making it very selective and often hostile to major life forms.
Microorganisms that inhabit extreme environments are referred to as extremophiles and are broadly classified into extremophilic organisms (which require one or more extreme conditions to grow), and extremotolerant organisms (which can tolerate extreme values of one or more physicochemical parameters though growing optimally at normal conditions) (Rampelotto, 2013). Extremophiles from a wide range of natural extreme environments have been characterized, and indeed, a lot of biotechnological applications have been made stemming from their secondary metabolites (Seufferheld et al., 2008;Azua-Bustos and Gonz alez-Silva, 2014). Relatively few researchers (Chronakova et al., 2010;Perks, 2011) have characterized microorganisms from synthetic extreme environments with the aim to biotechnologically exploit them. Therefore, the varieties of microorganisms in synthetic extreme environments, together with the molecular mechanisms they have evolved to cope with abiotic stresses in their environments, still need to be elucidated.
The current challenge besetting the study of biological molecules produced by extremophiles is that their potential applications may not be well known, or even unknown (Azua-Bustos and Gonz alez-Silva, 2014). This, however, offers an immense potential for future development. To this end, evidence from past research suggests biotechnologically significant roles ranging from antimicrobials to anticancer molecules (Perks, 2011). Among the most explored applications of extremophiles are their production of enzymes (also referred to as extremozymes) (Demirjian et al., 2001;Enache and Kamekura, 2010), fatty acids and proteins (Cavicchioli et al., 2002;Reed et al., 2013), antibiotics or biomedicines (Azua-Bustos and Gonz alez-Silva, 2014), and the direct application of cultures in chemical processes like desulfurication of flue gases and in biohydrometallurgical processes (Huber and Stetter, 1998). Biohydrometallurgical processes include, but are not limited to, microbial mining of precious (and rare) metals, oil recovery, bioleaching and water treatment.
Among microbial groups most commonly found in extreme environments, actinobacteria are arguably the richest source of small molecule diversity, with widespread global and environmental dispersal (Jami et al., 2015;Kuang et al., 2015;Ettoumi et al., 2016). Synthetic extreme environments like carwash effluent tanks and drains remain unexplored, and their potential as a source of biotechnologically important molecules remains unknown. While synthetic extreme environments are (almost all) seriously polluted sites which are in need of remediation, they may conversely be harbouring life-saving agents and other biotechnologically important molecules. As the survival strategies of extremophiles are often novel and unique (Seufferheld et al., 2008), the necessity for microbial cell components to adapt to extreme environments (natural or synthetic) implies that a broad range of cellular products (genes and metabolites) are available for biotechnological applications.
While there exists no previous literature records elucidating the microbial composition of carwash effluents, Tekere et al. (2016) characterized the physicochemical properties of carwash effluents which they have shown to have high levels of, among other pollutants, petroleum hydrocarbons, polycyclic aromatic hydrocarbons (PAHs), phenols and heavy metals. Further, these effluents were shown to be toxic not only to animals and plants but were also found to inhibit light production by the bacterium Vibrio fischeri. PAHs are the most common environmental pollutants which enter both the terrestrial and aquatic environments in large volumes, posing a great hazard to ecosystems (Bayat et al., 2015). Research has shown that in the short to medium term, contamination of environmental media with heavy metals results in reduced microbial biomass, coupled with reduced enzyme activity (Wang et al., 2007;Akmal, 2009). In the long term, environmental media contaminated with PAHs showed reduced microbial biodiversity and species richness index (Markowicz et al., 2016). However, the same study by Markowicz et al. (2016) showed that while soils contaminated with heavy metals had reduced microbial activity and changes in microbial community structure, a high microbial evenness index was observed. This suggests that some populations of the microbial community are 'enriched' by the presence of the very same pollutants which diminish the population sizes of the less tolerant microorganisms. Assessment of the impact of long-term diesel contamination on soil microbial community structure, for example, showed both a decrease in the relative abundance of certain bacterial phyla and an increase in the relative abundance of other phyla (Sutton et al., 2013). The microbes that undergo natural attenuation to adapt to extreme environments could be sources of biotechnologically indispensable molecules with a wide range of applications ranging from bioremediation of polluted sites to producing industrially relevant enzymes and other products. The aim of this study therefore was to characterize bacterial isolates from carwash effluent samples and to determine their ability to produce some target enzymes as well as other secondary metabolites, previously known or unknown, and create a metabolic fingerprint.

Description of study area
Carwash effluent samples were obtained from six carwash stations in the Johannesburg Metropolitan Municipality located in Gauteng Province in South Africa. Three carwash stations were located in Johannesburg City, while the other three were in the affluent business area, Sandton City (Fig. 1).
For confidentiality purposes, the exact locations and/or real names of the carwashes are not mentioned. However, the sampling stations were code-named as BPS, CAL and SAS (Johannesburg), and CAC, PCW and TCWPM (Sandton).

Sample collection, bacterial isolation and characterization
One-litre (1 l) grab samples were obtained from the onsite treatment facility of each carwash using sterile plastic sampling bottles. The samples were immediately chilled by placing them in cooler boxes containing ice and transported to the laboratory at UNISA Florida Campus for analysis within 6 h of collection. Once in the laboratory, the samples were shaken to ensure homogeneity after which 100 ll aliquots of each sample were spread plated in triplicate onto freshly prepared nutrient agar (Sigma Aldrich, Pretoria, South Africa). The plates were incubated at 30°C for 48 h. The resultant (mixed) cultures were separated and purified by subculturing onto nutrient agar plates until axenic cultures were  for sequence analysis. The resultant sequence chromatograms were manually edited with Chromas software v2.6.1 (Technelysium Pty Ltd., South Brisbane, Qld, Australia) and subjected to BLAST analysis to compare the identity of the isolates. The sequences were then used for phylogenetic analysis using the MOLECULAR EVOLUTIONARY GENETIC ANALYSIS V6.0 (MEGA6) software (Tamura et al., 2013) using an alignment created with SINA Aligner. A phylogenetic tree was constructed using maximum-likelihood analysis with kimura 2-parameter model and 500 times of bootstrap replications. Finally, the sequences were submitted to Gene bank to obtain the accession numbers.

Screening for hydrocarbon utilization
The isolates were tested for the capacity to degrade hydrocarbons following a modified protocol previously described by Um et al. (2010). Minimal salt medium (MSM) containing in g/l: NaCl 5.0, K 2 HPO 4 1.0, NH 4 H 2 PO 4 1.0, (NH4) 2 SO 4 1.0, MgSO 4 . 7H 2 O 0.2 and KNO 3 3.0 supplemented with 1.25% (w/v) of agar, was autoclaved for 15 min and poured out onto Petri plates as 'bottom agar'. This bottom agar was then overlaid with 100 ll of a polycyclic aromatic hydrocarbon (PAH) solution prepared by dissolving 10 mg/50 ml naphthalene and 10 mg/10 ml benzanthracene (Sigma, Pretoria, South Africa) in methanol. The PAH solution was evenly spread over the agar surface using sterile disposable spreaders and the plates were left in the lamina flow for the solvent to evaporate, leaving behind a visible thin white layer of PAHs on the surface of the bottom agar. To inoculate the bottom agar, 100 ll of each bacterial isolate was mixed with 900 ll of 35-40°C molten 'top agar' medium containing 0.5% agar and same mineral composition as the bottom agar and immediately poured on Petri plates containing bottom agar. The plates were then swirled gently to spread molten top agar medium over bottom agar, which resulted in the white layer of PAHs to move to the surface of the top agar medium upon solidifying. The inoculated plates were incubated at 25°C and examined daily for the presence of growth and clear zones.
The isolates which were able to degrade PAHs (seen by formation of clear halos colonies) were selected for hydrocarbon-degrading assays following the method of Oliveira et al. (2012). The isolates were inoculated into 20-ml tubes containing 10 ml of fermentation broth made up of minimal salt medium supplemented with 10 g/l of glucose and 1.0 g/L of yeast extract and incubated at 25°C for 72 h with shaking at 120 r.p.m. After incubation, the inoculum was centrifuged at 10 000 r.p.m. for 5 min at 10°C. The cell pellet was resuspended in phosphate buffer (pH 7) and again centrifuged at 10 000 r.p.m. for 5 min at 10°C, discarding the supernatant. To remove all culture medium residues, this step was performed twice. After the final centrifugation, the isolates were resuspended in phosphate buffer and the OD adjusted to McFarland 0.5. The assay was carried out in a sterile 96-well microtitre plate by adding 20 ll of the isolate suspension in triplicate, to a row of wells containing 168 ll of minimal salt medium, 12 ll of 2,6-dichlrophenolindophenol (DCPIP) and 2 ll of diesel oil. Each isolate was inoculated into a row of wells containing three different hydrocarbons, the other two being naphthalene and benzanthracene. In addition to the test wells containing isolates and different hydrocarbons, the experiment had a positive control produced by mixing 148 ll of minimal medium with 12 ll of DCPIP solution, 20 ll of 10% glucose solution and 20 ll of the Pseudomonas cell suspension; and a negative control consisting of 168 ll of minimal medium, 20 ll of the isolate suspension and 12 ll of DCPIP solution. The plates were incubated at 30°C, and the readings were taken after 24, 48 and 72 h of incubation using a spectrophotometer at 600 nm. The percentage reduction in DCPIP was obtained using the following equation: where the value of Y was, sequentially, 24, 48 and 72 h.

Screening for enzyme production (lipase and cellulase)
Production of lipase was determined by growing the bacterial isolates in rhodamine-olive oil-agar medium following the method of Kumar et al. (2012) with little modifications. The agar medium contained, in g/l, agaragar 20, MgSO 4 0.2, CaCl 2 0.02, KPO 4 1.0, K 2 PO 4 1.0, NH 4 NO 3 1.0, FeCl 3 1.0 and yeast extract 5.0. The medium was adjusted to pH 7.0, autoclaved and cooled to about 50°C after which 31.25 ml of olive oil and 10 ml of rhodamine B solution (1.0 mg/ml distilled water and sterilized by filtration) were added with vigorous stirring. It was then poured into Petri plates under aseptic conditions and allowed to solidify. The plates were inoculated by smearing buttons of bacterial cultures on the agar surface. After incubation of the plates for 48 h at 30°C, they were viewed under UV irradiation and lipase-producing strains were identified by formation of orange fluorescent halos around bacterial colonies due to the hydrolysis of substrate. For cellulase screening, agar medium containing 0.2% (w/v) carboxymethylcellulose sodium salt (CMC), 1% agar and minimal salt medium was prepared and poured into Petri plates. Spot inoculation was then performed using axenic cultures of the test isolates followed by incubation at 30°C for 48 h. Hydrolysis zones were visualized by flooding the plates with 0.1% Congo red stain (Glass World, Johannesburg, South Africa) and allowing to stand for 15 min followed by destaining with 1 M NaCl.

Analysis of bacterial secondary metabolites
Bacterial isolates which were positive for enzyme production screens were grown in fermentation media consisting of minimal salt medium supplemented with olive oil and CMC-Na salt for 7 days in a shaking incubator at 30°C. After incubation, cell debris was separated from the supernatant under high-speed centrifugation at 4°C. The supernatant was subjected to solvent extraction of secondary metabolites. Briefly, the supernatant was a mixture of chloroform and methanol (1:1, v/v) and shaken for 12 h at 120 r.p.m. and 25°C. The ethyl acetate fraction was then separated from the aqueous fraction using a separating funnel. The solvent fraction was evaporated to dryness in vacuo at 80°C, and the residue was reconstituted in a mixture of 1:1 (v:v) acetonitrile and hexane. The acetonitrile fraction was used to analyse for the polar secondary metabolites using LC-MS, while the hexane fraction was used to analyse for nonpolar, volatile compounds using GC-MS. For GC analysis, ionization energy was set at 70 eV using He as the carrier gas. The flow rate was set at 1 ml/min, injection volume at 2 ll, split ratio at 10:1, injection temperature at 250°C, ion source temperature at 200°C and oven temperature at 110°C (isothermal at 2 min) with increase of 10°C/min to 200°C then 5°C/min to 280°C (with 9 min isothermal at 280°C). An HP-5MS fused silica capillary column (30 m, 0.25 mm i.d., 0.25lm film, cross-linked to 5% phenyl methyl siloxane stationary phase) was used. For MS, ionization energy was set at 70 eV, scan interval at 0.5 s, fragments at 45-450 kD and solvent delay at 0-2 min. The identification of compounds was based on comparison of their mass spectra with the National Institute of Standards and Technology (NIST 2005) library.
LC-MS was performed using an ultra high-performance liquid chromatography-mass spectrophotometer (Compass otofSeries 1.9, Bruker Instrument: ImpactII) system. For LC analysis, the column was Acquity UPLC BEH C18 1.7 lm, diameter 2.1 9 100 mm (Miscrosep Waters, Johannesburg, South Africa), and the solvents were 0.1% formic acid (FA) in water and 0.1% FA in acetonitrile. The column flow was set at 0.3 ml/min, column oven temp at 35°C and draw speed at 2 ll/s with a total injection volume of 2 ll. Mass spectrometer (MS) conditions were set at a mass range of 50-1600 m/z, capillary 4500 v, dry gas 8 l/m, gas temperature 220°C, ion energy 4.0 eV, collision energy 7.0 eV, cycle time 0.5 s. Data analysis was performed using the BRUKER Software (Bruker Compass DataAnalysis 4.3; Bruker Daltonik GmbH 2014, Bremen, Germany).

Isolation and characterization of bacterial isolates in carwash samples
A total of 17 different isolates were obtained from all the six carwash sampling sites. Phylogenetic comparison of PCR-amplified 16S rDNA sequence data of each isolate with the database of the databases of known species using the NCBI server revealed that isolated bacteria belonged to the following genera: Shewanella (35.32%), Proteus (11.76%), Paenibacillus (11.76%), Enterobacter (11.76%), Citrobacter (11.76%), Aeromonas (5.88%), Pseudomonas (5.88%) and Pantoea (5.88%) ( Table 1 and Fig. 2). The 1500-bp-long 16S rRNA gene containing nine hypervariable regions (V1-V9) (Chakravorty et al., 2007) was sequenced in single chain and, in all instances during phylogenetic analysis, the percentage similarity was ≥ 99%. Isolation of bacterial strains in the present work was performed under the conditions and culture medium explored, and the percentages observed only correspond to data observed in this work although other bacteria which could not grow under the explored conditions could exist in the same environment. There were no available data in literature to compare with the data obtained in this study, making this study potentially the first study to isolate and characterize bacterial isolates from carwash effluents.
Screening for enzyme production (lipase and cellulase) and hydrocarbon utilization Four isolates, three from carwash CAC namely Aeromonas sp. CAC11, Paenibacillus sp. CAC12 and Paenibacillus sp. CAC13 and one from carwash PCW namely Citrobacter sp. PCW7, were able to utilize hydrocarbon as a carbon source during preliminary screening (Table 2). Further, Paenibacillus sp. CAC12 and Paenibacillus sp. CAC13 were the only isolates to utilize the CMC sodium salt as a carbon source, indicating their ability to produce cellulase. Screening for lipase production showed that only isolate Proteus sp. BPS2, Pseudomonas sp. SAS8 and Proteus sp. CAL3 could hydrolyse olive oil.
There are environments which, consequent to their substrate profiles, are better suited for the growth and proliferation of microorganisms producing certain kinds of enzymes. For example, a municipal solid waste site is rich in lignocellulosic materials, thus providing a suitable environment for cellulase-producing organisms (Ali et al., 2013), while oil mill soils are likely to be endowed with lipase-producing organisms (Ramesh et al., 2014). However, extreme environments are known to support microbial life forms with highly flexible metabolic pathways (Stathopoulou et al., 2013;Dalmaso et al., 2015). It is this flexibility, which causes extremophilic microorganisms to be the best known producers of biotechnologically important molecules/products (Bisht and Panda, 2011;Bull, 2011;Azua-Bustos and Gonz alez-Silva, 2014). As in this study, production of cellulase by Paenibacillus sp. has previously been confirmed (Pandey et al., 2013). Similarly, production of lipase by Pseudomonas and Proteus has been confirmed in previous studies (Golani et al., 2016;Kumar et al., 2016). What has not been reported, however, is the occurrence of enzyme-producing microorganisms from carwash effluent samples. In the wake of growing demand for enzymes with improved catalytic performance and tolerance to process-specific parameters (Guazzaroni et al., 2015), it is imperative that every kind of environment be surveyed for enzyme-producing microbes.

Hydrocarbon degradation assays
Isolates Aeromonas sp. CAC11, Paenibacillus sp. CAC12, Paenibacillus sp. CAC13 and Citrobacter sp. PCW7 were further applied to hydrocarbon degradation assays using three classes of hydrocarbons namely benzanthracene, naphthalene and diesel fuel using 2,6-dichlorophenolindolphenol (DCPIP) as the indicator. DCPIP is an electron acceptor that becomes reduced (decolourized) when redox reactions occur, in this case when NADH is converted to NAD + during microbial degradation of hydrocarbons (Kubota et al., 2008;Bidoia et al., 2012). During the assay, the OD 600 at 24 and 48 h was still almost similar to the OD 600 at 0 h with, however, a sudden significant reduction in OD 600 at 72 h. The results are presented in Fig. 3. Ionescu et al. (2015) also observed that bacterial cells needed about 4 days for the activation of the transcriptional regulation of metabolic operons involved in producing the necessary enzymes to metabolize an unusual substrate.
Isolate Citrobacter sp. PCW7 showed the highest potential of the four test isolates to metabolize all three hydrocarbon classes, with the highest activity against benzanthracene (45% DCPIP reduction). Next was isolate Aeromonas sp. CAC11, which showed 25% reduction in DCPIP when metabolizing naphthalene and 15% when metabolizing benzanthracene. Isolates Paenibacillus sp. CAC12 and Paenibacillus sp. CAC13 did not result in any significant DCPIP reduction. High levels of polycyclic aromatic hydrocarbons in environmental media have generated considerable interest from a toxicological standpoint (Al-Daghri et al., 2014). Benzanthracene, for instance, is an environmentally recalcitrant pollutant, which is classified as a group 2A carcinogen by the International Agency for Research on cancer (Kunihiro et al., 2013). Besides naphthalene and benzanthracene, diesel fuel contains alkylbenzenes, toluene, naphthalenes and polycyclic aromatic hydrocarbons (Irwin et al., 1997), constituents which have already been implicated in asthma and cancer cases (Al-Daghri et al., 2013). As such, it is of public and environmental health interest to find microorganisms which can biodegrade such  hydrocarbons to rid or lessen the danger that they pose to public health, more so if such organisms are indigenous to the polluted environments. Isolates Citrobacter sp. PCW7 and Aeromonas sp. CAC11 showed much promise as potential microorganisms that can be used for the bioremediation of environments polluted by high molecular weight hydrocarbons. Other studies (Mandri and Lin, 2007;Affandi et al., 2014) have also proved the high diesel oil and grease degradation potentials of Citrobacter (C. freundii) and Aeromonas (A. hydrophila), which were, however, isolated from hydrocarbon-contaminated environments, as in the present study. Most other studies, which have assessed the hydrocarbon degradation potentials of Aeromonas and Shewanella, used test strains that were isolated from hydrocarboncontaminated environmental media (Mart ın-Gil et al., 2004;Ben Said et al., 2008;Hamzah et al., 2010). However, the study of Deppe et al. (2005) showed that Shewanella sp. isolated from a hydrocarbon-free environment were unable to degrade crude oil except when it was applied in consortia with other bacteria. This shows that these bacterial strains have evolved in the explored environments to degrade hydrocarbons. Bacteria of the genera Alcaligenes, Stenotrophomonas, Sphingomonas and Pseudomonas have also been documented to biodegrade benzanthracene (Kunihiro et al., 2013), although the metabolites were not documented, as was the case in this study. Elsewhere, Bonf a et al. (2011) described the isolation of aromatic hydrocarbon-degrading haloarchaea in hypersaline environments. As in this particular study, the hypersaline environment from which they isolated the organisms was synthetic, created as a result of the discharge of 'produced' hypersaline water from gas and oil extraction activities. One point comes to the fore; that environmental pollution of whatever nature may indeed eliminate most forms of life (including some microorganisms), but it will always result in selective proliferation of other groups of microorganisms that will utilize the pollutant as a food source or a means to it. It is that particular ability to not only survive but thrive in polluted/adverse environments that makes such microorganisms potential bioresources (or sources thereof) for the ecological recovery of polluted sites. Further to, the existence of polyaromatic hydrocarbon-degrading bacteria like Citrobacter sp. PCW7 and Aeromonas sp. CAC11 in the natural environment could be exerting a protective effect on higher organisms by continuously removing potentially harmful hydrocarbons from the environment.  Secondary metabolite mapping using GC-MS and UHPLC-MS GC-MS secondary metabolite elucidation was able to identify 107 different compounds produced by five bacterial isolates, which had tested positive when screened for production of either lipase or cellulase. These were Proteus sp. BPS2, Paenibacillus sp. CAC12, Pseudomonas sp. SAS8, Proteus sp. CAL3 and Paenibacillus sp. CAC13. The compounds were identified by comparison of their mass spectra with the NIST library based on their molecular weight and retention time. To ensure a higher degree of accuracy, the minimum similarity match cut-off for compound identification was pegged at 700. The compounds were configured into a map (Fig. 4) through multivariate analysis (seriation) using the PALEONTOLOGICAL STATISTICAL software (PAST3.13, Hammer 1999. Looking at the number of compounds that were produced, it would have been virtually impossible to discuss each of them in detail; hence, the data were also subjected to principal component analysis (PCA) using SIMCA (SIMCA 14.0 Ink, UMETRICS Company 1998 and the compounds which fell off the nucleus were chosen for further analysis and discussion. These are presented in Table 3. Isolate Paenibacillus sp. CAC12 produced the most number of biotechnologically important molecules. These included cyclopentane, silane (cyclohexyldimethoxymethyl-), spiro[2,4]hepta-4,6-diene and 3-(4-methylbenzoyl)-2-thioxo-4-thiazolyl 4-methylbenzoate. With the exception of silane, which is used as an important component of polymeric substances for filling dental cavities (Antonucci et al., 2005), all the other metabolites are known bioactive compounds exhibiting either antifungal or antibacterial activities. While no studies have previously reported production of any of these compounds by any member of the genus Paenibacillus, Pohl et al. (2011) reported microbial production of cyclopentane in the form of a C17 cyclopentane fatty acid from Pseudomonas. Zhang et al. (2015) have, however, previously extracted cyclopentane from Cunninghamia lanceolate wood extracts known for their rich drug compositions, using GC-MS. Peng et al. (2013) also extracted it from Illicium verum fruit, a medicinal plant that is known for having many active ingredients. As according to Zhang et al. (2015), the output of bioactive ingredients from herbaceous plants is too little to meet market demand, biotechnological approaches targeting enzyme-producing microorganisms or their genes can be employed to amplify production of such biotechnologically important molecules.
Production of the compound 3-(4-methylbenzoyl)-2thioxo-4-thiazolyl 4-methylbenzoate was of particular interest in this study as no research records could be found of its production by any microorganism. The only information that could be accessed on this compound was about its multiple uses which ranged from inhibition of tau fibril formation to inhibition of Bacillus subtilis virulence mechanisms to inhibition of parasitic mechanisms of the malaria parasite Plasmodium falciparum (NCBI, 2016). Formation of tau fibrils has been linked to Alzheimer's disease, which is one of the most common causes of dementia in adults (Morozova et al., 2013;Dinkel et al., 2015). Clearly, this is a very important compound with potential for many other applications, and further research on its production and applications is highly recommended.
A glance at Table 3 also shows that most of the compounds produced by the isolates are bioactive ingredients of commercial antimicrobials or plant extracts. The compound pyrrolo[1,2a]pyrazine-1, 4-dione, hexahydro-3-(2-methylpropyl), for instance, has antifungal, antibacterial, nematicidal and anticancer activities. Its antimicrobial profile suggests a need by the producing isolates (Pseudomonas sp. SAS8 and Proteus sp. BPS2) to competitively suppress the growth of other competing microorganisms, which could be likely due to restricted resources in their environment. Hibbing et al. (2010) allude to the fact that in cases of stiff competition for resources, microbial communities could resort to production of colicins (bacteriocins targeting enteric bacteria) and/or antibiotics among other strategies of limiting the growth of competing species. Microorganisms, which produce anticancer drugs, deserve a much bigger attention given the growing global incidence of this disease, and the discomfort of the current chemotherapeutic methods of treating it.
Although a few selected compounds were given extensive analysis and discussion, a potentially high number of biotechnologically important molecules from the metabolite map (Fig. 4) remain unexplored. This opens up new avenues for further research to fully tap into these microbial bioresources. Also, while compound determination by both GC-MS and LC-MS was only qualitative, the presence of any of the identified compounds points to the presence of the encoding gene(s) within the bacterial isolates. Application of biotechnological techniques like gene cloning can therefore be used to amplify production of the compounds to commercial levels.
The compound 5,7-dihydroxyflavone (chrysin), which was produced by all isolates, is a known natural flavonoid found in many plant extracts that are commonly used as wound-healing, skin-protective and anticancer medicines (Sathishkumar et al., 2015). Chrysin is also known for its anti-inflammatory and antioxidation properties (Engelmann et al., 2005). The research of Kim et al. (2011) also found chrysin to be an effective inhibitor against hyperpigmentation by directly inhibiting the activity of adenylyl cyclase, a key enzyme involved in cAMPinduced melanogenesis. Other researchers  have proven that chrysin is an antihypertensive drug, which acts by inhibiting the proliferation and migration of pulmonary artery smooth muscle cells, together with the associated extracellular matrix components, principally, collagens. Given its important range of biological activities, microbial production of chrysin could be manipulated to significantly increase its production both as a therapeutic agent and also as a dietary supplement.
The compound 2-(4-hydroxyphenyl)ethyl tricosanoate, which was also identified in the crude extracts of all isolates, was previously isolated by Acevedo et al. (2000)  as one of the components of the crude extracts of Buddleja cordata subsp. cordata known for antimycobacterial activity. In addition, the same compound is traditionally used in various other applications including treatment of diarrhoea, headaches and kidney ailments (Acevedo et al., 2000). Literature records of microbial production of 2-(4-hydroxyphenyl)ethyl tricosanoate are scarce, suggesting there could be many metabolites which are currently known to be produced by plants only when, in fact, they are produced by microorganisms also. One of the identified compounds, cyclo(phenylalanylprolyl), is a known cyclic dipeptide, which is used as a scaffold for drugs besides its use as an antiviral, antibiotic and antitumour drug (Wickrama Arachchilage et al., 2012). Microbial production of this dipeptide has previously been reported in a Streptomyces sp. isolated from marine sediments (Macherla et al., 2005). Vidarabine is yet another compound with antiviral properties that was identified in the crude extracts of all isolates. Vidarabine is a purine nucleotide analog that inhibits viral DNA synthesis and has previously been identified in the extracts of a Caribbean sponge Cryptotethia crypta (Suzuki et al., 2006). Literature is awash with evidence of production of vidarabine by marine bacteria and sponges (Sagar et al., 2010;Wang et al., 2013;Nadeem et al., 2016). To our knowledge, this is the first report of the production of vidarabine by bacteria isolated from a non-marine environment. Of late, however, use of vidarabine as an antiviral agent has been declining owing to its neurotoxicity, which has prompted research for more effective and less toxic alternatives (Vajpayee and Malhotra, 2000). Other identified compounds including 1-(2-hydroxyethyl)-2,5,5,8atetramethyldecahydro-2-naphthalenol, 4-hydroxy-1-phenylbenzimidazole hydrochloride, 2,2,3-trihydroxybutanoic acid and cyclohexyl(2,4-dimethylcyclohexyl)hydroxyacetic acid have no clear known uses, as does a host of other compounds which neither of the LC-MS linked online libraries could detect. This shows that exploration of novel microbial metabolites still has a long way to go. Dalmaso et al. (2015) estimated that between 1 and 10% of prokaryotes have so far been described, and this equates more or less to microbial secondary metabolites.
Most of the metabolites identified in this study have previously been documented to have antibacterial, antifungal and/or anticancer properties, although this was not experimentally proven in the present study. This finding, however, was to be expected since currently, the most marketed antimicrobials and anticancer drugs are natural products of microbial origin (Shaaban et al., 2013). It follows that microbial populations in different ecological settings including natural and synthetic extreme environments still have a lot to offer in terms of bioresources. In a bid to increase the success rate in the identification of novel microbial secondary metabolites, the use of GC-MS and UHPLC-MS in chemical profiling of crude fermentation extracts can be a very useful tool for assessing the chemical novelty of the crude extracts by comparing the mass spectra to in-built and online compound libraries such as Chemspider, Pubchem and KEGG.

Conclusion
Synthetic extreme environments have been overlooked as potential sources of biotechnologically relevant microbiota. The outcome of this research indicates that some microbial metabolites which have previously been known to be produced only by microorganisms in natural extreme environments like the marine environment can also be produced by microorganisms in synthetic extreme environments like carwashes. Further, this research has revealed the immense bioresource potential of microorganisms inhabiting synthetic extreme environments, which harbour potential as agents of bioremediation and/or producers of biotechnologically relevant molecules including enzymes.