Overproduction of glycolipopeptide bio-amphiphile by Pseudomonas aeruginosa mutant generated through three-tier cocktail random mutagenesis

Abstract

A robust three-tier cocktail of random mutagenesis involving atmospheric and room temperature plasma, gamma-rays, and N-methyl-N’-nitro-N-nitrosoguanidine exposures was applied to induce glycolipopeptide overproduction in Pseudomonas aeruginosa strain IKW1 (GenBank Accession No.: PV664482.1). Auxanographic investigations suggested unrepaired metabolic blocking of the threonine, and hence, isoleucine branches of the aspartate pathway in the overproducing mutant, Thr⁻Ile⁻PGN4. Glycolipopeptide overproduction significantly correlated with increases in urease activity, Ni²⁺ uptake, cytosolic Ni²⁺ localization and upregulation of histidine biosynthesis. Scalability by batch fermentation in modified optimized medium in 5-L bioreactor led to 1.24-fold higher volumetric oxygen transfer coefficient, kLa, with an oxygen uptake rate (OUR) of 18.6 mmol O₂ gDCW^− 1 h^− 1, which resulted in 3.86-fold higher glycolipopeptide productivity. Periodic biochemical and thin layer chromatographic analysis revealed trophophasic synthesis of a functional surface-active rhamnolipid base with surface (SFT) and interfacial (IFT) tension reductions to 29.74 dynes cm^− 1 and 2.22 dynes cm^− 1, respectively. A late idiophasic peptidation of the rhamnolipid base further reduced SFT and IFT to 24.41 and 0.87 dynes cm^− 1, respectively. Superior catalysis by catechol 1,2-dioxygenase favored glycolipopeptide-mediated crude oil degradation by Pseudomonas putida strain ATCC 49,182 compared to commercial biosurfactants giving rate constants of 0.115 d^− 1 (glycolipopeptide), 0.087 d^− 1 (di-rhamnolipid), 0.096 d^− 1 (surfactin) and 0.065 d^− 1 (without surfactant). The technology is recommended for sustainable glycolipopeptide production toward efficient crude oil-impacted ecosystems’ bioremediation.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12866-026-04721-5.

Introduction

Glycolipopeptides are a new class of extracellular polymeric macromolecular bio-amphiphiles synthesized by specialty strains of a few genera of bacteria like Pseudomonas, Variovorax, Corynebacterium, Lactobacillus, Mycobacterium, Klebsiella and Myxobacteria [1, 2]. They have been described, by reason of their ability to reduce surface and interfacial tensions, as biosurfactants [3], and by reason of their high molecular weights, as bio-emulsifiers [4]. However, the glycolipopeptide is best described as a bio-amphiphile whose hydrophobic oligomeric lipid moiety is covalently linked to a short peptide sequence at its carboxylic end and a carbohydrate moiety linked at its hydroxyl end by a glycosidic bond [5].

The macromolecule was first reported by Koronelli et al. [6] in a strain of Pseudomonas aeruginosa, which was, until a few years back, thought to exclusively synthesize rhamnolipid biosurfactants. Ekpenyong et al. [7] reported the extraction of an unusual surface-active compound which they described as a glycolipopeptide from the culture filtrate of Pseudomonas aeruginosa strain IKW1 (GenBank Accession no. PV664482.1). Kerr et al. [5] gave a full description of the glycolipopeptide which they isolated from the culture broth of Variovorax paradoxus strain RKNM-096 in 2015, and for which they hold the patent. Recently, Pandey et al. [8], Gharaie et al. [9] and Kumar et al. [2] reported glycolipopeptide biosurfactant production in Aeromonas hydrophila RP1, Bacillus pumilus SG and Klebsiella pneumoniae subsp. ozaenae BK34, respectively.

Owing to their tremendous emulsifying capability, glycolipopeptides mostly find applications in the bioremediation of hydrocarbon-polluted sites and as antimicrobial, anti-adhesive, anti-cancer and anti-biofilm agents in cosmetic and pharmaceutical industries [10–12]. One of the draw-backs in biosurfactant application is poor availability and accessibility arising from low yield and high production cost compared to their chemical counterparts [13]. Several reports exist regarding successful media and fermentation conditions development for their enhanced production and recovery, most of which have revolved around response surface methodology and machine learning/artificial intelligence approaches [10, 14].

Strain engineering is potentially the most critical component of the yield improvement steps in bioprocess engineering owing to the central role played by the microbial entity in the process [15]. Different methods have been deployed to alter the genetic make-up of potential industrial microorganisms among which are protoplast fusion, cloning, recombinant DNA technology, and the trendy CRISPR/Cas9 gene editing technology [16, 17]. However, classical mutagenesis remains a reliable approach owing to its ease of use, cost-effectiveness, and environmental safety [13]. The major limitations of the approach are the tedium in screening a large number of mutants to obtain stable overproducing strains and the decreased vitality after several rounds of mutagenesis [18, 19].

Pseudomonas aeruginosa has emerged as a difficult-to-manipulate microorganism by random mutagenesis; the difficulty arising from its inherent efficient DNA repair mechanisms which renders replication errors unstable [20]. These repair mechanisms are coupled to the more than 5000 genes that are dedicated to regulation and metabolite synthesis, majority of which are under the efficient control of quorum sensing [21]. In recent times, plasma mutagenesis has evolved as an additional level of random mutagenesis [22]. Plasma is the 4th state of mater and refers to ionized gas which characteristically has the ability to generate a mixture of atoms, excited molecules, charged particles and UV photons, which interact with biological molecules to alter their sequences, thus providing this technique the unique ability to spontaneously and randomly introduce alterations in DNA sequences of microorganisms [23, 24]. The damage consequently induces the activation of an avalanche of repair mechanisms in the survived cells leading to development of mutants [25]. Still, only very few reports exist for its successful mutagenesis of Pseudomonas species.

In the present study, the glycolipopeptide-producing Pseudomonas aeruginosa strain IKW1 (GenBank Accession No.: PV664482.1) was exposed to alternating combinations of plasma, gamma and chemical mutagens to subvert the efficient genetic repair mechanisms of the bacterium, and enhance glycolipopeptide yield and productivity. The effectiveness of the biomolecule from the selected mutant was evaluated through comparative biosurfactant-mediated crude oil degradation by Pseudomonas putida strain ATCC 49,182. To the best of our knowledge, this is the first report on the successful combinatorial mutagenic engineering of a Pseudomonas aeruginosa strain for enhanced production of a biosurfactant.

Materials and methods

The producing organism and strain improvement

The lyophilized stock culture of Pseudomonas aeruginosa strain IKW1 (GenBank Accession No.: PV664482.1), was retrieved from University of Calabar Collection of Microorganisms (www.ccinfo.wdcm.com/details?regnum=652) and reactivated according to collection’s guidelines [26].

Combinatorial mutagenesis of strain UCCM 00004

An overnight culture of the bacterium was prepared in Luria-Bertani (LB) broth containing (g L^− 1) tryptone 10, yeast extract 5, and NaCl 10, and a cell population of 2.17 × 10⁸ cfu mL^− 1 was subjected to 27 combinatorial random mutagenesis of plasma (P), gamma-ray (G) and N-methyl-N’-nitro-N-nitrosoguanidine (N) (Table S1; Supplementary material). Survived cells from first exposure were re-exposed for the second and third time after a recovery period of 24 h.

Atmospheric and room temperature plasma (ARTP) mutagenesis

ARTP mutagenesis was conducted using the ARTP machine (ARTP-II Si Tsingyuan Biotechnology Co., Ltd, Beijing, China). A bacterial density of 2.17 × 10⁸ cfu mL^{− 1} was exposed to the plasma under slightly modified conditions of Xu et al. [13] as 2 mm from the irradiating plasma source, flow rate of 10 L min^{− 1}, mutagenic power of 120 W and variable exposure time of 0–240 s, at 30 s time difference to obtain a 90% kill.

Chemical mutagenesis with N-methyl-N’-nitro-N-nitrosoguanidine

This was conducted according to the protocol by Ohman et al. [27]. Briefly, 2.17 × 10⁸ cfu mL^{− 1} of exponentially growing cells of strain IKW1 were exposed to 0.1 mg mL^{− 1} of N-methyl-N’-nitro-N-nitrosoguanidine (Fluka-Chemika-Biochemika, Buchs, Switzerland; 99.7% purity) for 10–60 min to obtain a 90% lethality rate. The treatment was terminated by adding chilled modified minimal (minus D-glucose) Davis broth medium containing (g L^{− 1}): K₂HPO₄ 7, KH₂PO₄ 2, (NH₄)₂SO₄ 1, C₆H₅O₇Na₃ 0.5, MgSO₄.7H₂O 0.1, and pH adjusted to 7.0 to achieve 20X dilution. Mutagenized cultures were centrifuged and washed twice with double-distilled water and re-suspended in minimal medium. Flasks were incubated for 3 h before transferring onto complete Davis (plus D-glucose) agar (2% w v^{− 1}) medium for isolation of survivors.

Gamma-ray mutagenesis

Gamma mutagenesis was performed based on the protocol of Iqbal et al. [28]. A cell suspension of 2.18 × 10⁸ cfu mL^{− 1} was exposed to gamma rays at different doses to obtain the dose with 3 log kill. The selected dose was 300 Gy and the survived cells were diluted and appropriate dilutions plated onto LB agar (1.5% agar-agar; Oxoid). The plates were incubated at 37 °C for 36 h to isolate single colonies.

All survived cells were screened for biosurfactant overproduction in optimized glycolipopeptide production medium containing 5% (v v^{− 1}) waste frying sunflower oil as carbon source composed (% w w^{− 1}) of stearic acid 2.26, palmitic acid 6.18, volatile fractions 16.17, oleic acid 21.94 and linoleic acid 51.16. The minimal medium contained filter-sterilized urea (1.19 g L^{− 1}), Na₂HPO₄/KH₂PO₄ (4.5 g L^{− 1}; 2:1), MgSO₄.7H₂O (0.2 g L^{− 1}), NaCl (0.5 g L^{− 1}), CaCl₂ (0.5 g L^{− 1}), KCl (0.5 g L^{− 1}). It also contained trace element solution (1 mL) composed (mg L^{− 1}) of NiCl₂ 1.296, ZnCl₂ 0.125, FeCl₂ 0.125, K₃BO₃ 0.0104, CuSO₄.5H₂O 0.025. Medium pH was adjusted to 7.6 using 1 M HCl/1 M NaOH and sterilized by autoclaving (121 °C, 15 psi). Upon cooling, flasks were inoculated with 5% (v v^{− 1}) 18-h old LB broth culture of both mutant and wild type strains and incubated at 32 °C for 66 h with agitation at 130 rpm [10].

Crude biosurfactant was recovered by centrifugation at 7,155 x g for 10 min and filter-sterilized (0.22 μm Millipore) and was used for determination of biosurfactant overproduction by the oil displacement assay of Morikawa et al. [29].

Confirmation of chemical composition of the biomolecule

To confirm the glycolipopeptide nature of mutant biosurfactant, carbohydrate content was quantified by the phenol-sulfuric acid method of Dubois et al. [30] using glucose (Merck, USA) as standard. Rhamnose content was estimated by the orcinol method [31]. Protein was detected by the Bradford method [32] using bovine serum albumin (Sigma-Aldrich, USA) as standard and lipids by the method of Folch et al. [33]. The colorimetric results were confirmed by thin layer chromatography as outlined in Ekpenyong et al. [7].

Mutant stability evaluation

Overproducing strains were defined as survived cells with areas of displaced oil (ODA) significantly (p < 0.05) greater than 143.56 cm² of the wild strain. Each was tested for stability for 660 h corresponding to 10 generations at 66 h gen^− 1. Determinations were performed in triplicates and mean ODAs were compared at 5% significance level in GraphPad Prism 8 (GraphPad Software, Boston, Massachusetts, USA).

Mutant characterization

Auxanographic characterization of selected mutant

The selected stable mutant was characterized by auxanography using Holliday’s parallel plating method using 36 nutritional requirements for single growth factors [34].

Molecular characterization of selected stable mutant

The selected stable mutant was characterized by 16S rRNA partial-gene sequencing through DNA extraction, quantification, amplification, and sequencing, and the obtained sequence was evaluated by genetic alignment and phylogeny determination using the molecular evolutionary and genetic analysis (MEGA) software, version 12 [35].

Differential physiologics between mutant and wild type strains during fermentation

Glycolipopeptide fermentation was set up as described for stability testing, and urease activity, lipase activity and nickel uptake potentials of the wild and mutant strains were compared. Urease assay was performed by nesslerization method [36] using NH₄Cl (Merck, USA) as standard. Lipase assay was performed by spectrophotometric method at 405 nm measuring the amount of p-nitrophenol released from p-nitrophenol-palmitate [26]. Amount of Ni²⁺ in the fermentation broth before and after fermentation was quantified by atomic absorption spectrophotometer (AAS; Perkin Elmer, Model 2380, USA) at 232 nm wavelength and the difference was reported as amount taken up in intact bacterial cells. The intact bacterial nickel concentration was evaluated for distribution into periplasm, membrane and cytosol [37]. For the periplasmic protein, bacterial pellet was re-suspended in 30 mM Tris-HCl at pH 7.1 (30 mL) containing 20% sucrose and 1 mM phenyl-methyl-sulfonyl fluoride. The suspension was incubated for 30 min at room temperature (28 ± 2 °C), centrifuged at 11,000 x g for 10 min at 4 °C, and subsequently suspended in 30 mL of 0.5 mM ice-cold MgCl₂. The preparation was gently agitated for 5 min in an ice bath and finally centrifuged at 30,000 x g for 15 min at 4 °C. Amounts of protein and nickel were quantified by the Bradford assay and AAS, respectively, from the supernatant containing periplasmic proteins. Amount of nickel was expressed as nmol Ni²⁺ mg^{− 1} protein.

Cytosolic concentration of nickel was determined after disruption of osmotically-shocked cells that had lost periplasm by sonication at 30 W for 1 min. The procedure was repeated three times and the resultant suspension spun for 5 min at 27,190 x g. Amount of nickel in the cytosol was also expressed as nmol Ni²⁺ mg^− 1 protein. In summary, Ni²⁺ partitioning was determined using the following expressions;

1

2

3

where C_i is Ni²⁺ concentration in intact bacteria; C₀ is initial Ni²⁺ concentration in fermentation medium; C_f is Ni²⁺ concentration in sterile broth after fermentation; C_c is Ni²⁺ concentration in the cytosol; C_p is Ni²⁺ concentration in the periplasm and C_m is Ni²⁺ concentration in the cytoplasmic membrane. All data were obtained by triplicate determinations and comparison between mutant and wild type strain was made using analysis of variance (ANOVA) at 5% significance level in GraphPad Prism 8.

Comparative physiologics between mutant and wild type strains in minimal medium

Nickel uptake

The huge difference in nickel uptake and partitioning in mutant and wild type strains during fermentation necessitated evaluation of nickel uptake potential in minimal medium. HEPES-buffered minimal medium [38] was modified to rid the medium of inorganic phosphates and other metal chelators. Phosphate was replaced with 2 mM β-glycerophosphate and 10 mM HEPES (N-2-hydroxyethylpiperazine-N-2-ethanesulphonic acid) replaced Tris-HCl. Ni²⁺ was tested at varying concentrations ranging from 10 µM to 20 mM by exposing a cell density of 1.5 × 10⁸ cfu mL^{− 1} (0.5 McFarland) corresponding to 50 µg protein mL^{− 1} over a 1 h period. Determinations of nickel uptake were made every 10 min using the method described in Imperi et al. [37].

Effect of carbon source on nickel uptake

Carbon sources including glucose, sucrose, sodium succinate, sodium fumarate, gallic acid and phenol (Sigma Aldrich, USA) were tested for their effect on nickel uptake by the mutant and wild type strains in the modified HEPES-minimal medium. The medium was supplemented with each carbon source at 1.0% (w v^− 1) and incubated for 1 h. Determinations of nickel uptake were made every 10 min using the method described in Sect. 2.6.1 and data analyzed by one-way analysis of variance (ANOVA) at 5% significance level.

Effect of phosphate concentration on nickel uptake

The β-glycerophosphate concentration was varied by 2-fold dilution to obtain 4.0, 2.0, 1.0, 0.5, 0.25, 0.125 mM and nickel uptake assay repeated for 1 h. Assay for histidinol phosphate phosphatase (HPPA) and histidinol dehydrogenase (HDA) specific activities in wild and mutant strains were conducted according to Yadav et al. [39] using inorganic phosphate and nicotinamide adenine dinucleotide (NAD) as respective standards. One unit of HPPA, measured spectrophotometrically at 820 nm, was defined as the amount of enzyme required to release one nanomole of inorganic phosphate from L-histidinol phosphate at 37 °C per milligram protein per hour. On the other hand, one unit of HDA measured spectrophotometrically at 340 nm by NAD⁺ reduction, was defined as the amount of enzyme required to convert one micromole of L-histidinol to histidine and two micromoles of NAD reduced per minute at 25 °C.

Medium optimization for mutant glycolipopeptide production

Response surface optimization of levels of x₁ (glucose = 4.0)/0.5; x₂ (urea – 2.39)/0.6; x₃ (phosphates – 0.45)/0.15; and x₄ (nickel – 800)/100; toward glycolipopeptide production was carried out using central composite rotatable design (CCRD) with five levels of four factors and 6 center points in Design-Expert software version 13 (Stat Ease InC., Minnesota, USA). The developed quadratic model that established the relationship between the explanatory variables and glycolipopeptide concentration was optimized with the in-built response optimizer in Design-Expert using the desirability function following the quadratic expression below:

4

where β₀ denotes a constant coefficient, k = the k^th factor, β_ii = quadratic effect of the i^th factor and β_ij = effect of the interaction between the i^th and j^th factors, x₁, x₂ ………, x_k are the independent variables and ε is the error arising from the computation of the response variable Y.

The predicted optimized levels were validated in pentuplet shake flask experiments in real time and the validated optimized levels of the explanatory variables were used to modify the glycolipopeptide production medium previously developed for the wild strain [26].

Biosurfactant production by mutant Thr-Ile-PGN4 in modified optimized medium

Glycolipopeptide biosurfactant production by the mutant in the modified optimized medium was set up in a 5-L bioreactor with a 3 L working volume. Bacterial growth, glycolipopeptide production and substrate (waste frying oil) consumption were monitored periodically for 72 h. Briefly, the bioreactor was first sterilized at 121 °C for 20 min, allowed to cool to room temperature, then aerated to 100% oxygen saturation monitored by measuring dissolved oxygen concentration (DOC) using a DO probe (Mettler-Toledo; Thermo-Fischer Scientific, Porto Salvo). The oxygen transfer rate (OTR) was measured as an increase in DOC by sparging with pure nitrogen (99.98%; Sigma-Aldrich, USA) to bring DOC of the sterile system to zero. The system was gradually re-aerated and OTR monitored at different aeration and agitation rates. The volumetric oxygen transfer rate, kLa was determined as described in Ekpenyong et al. [40] while the oxygen uptake rate, OUR, was determined using the expression below:

5

where k_La is the volumetric oxygen transfer coefficient at the respective operating conditions, and [O₂] * is the oxygen saturation concentration, d [O₂]/dt is the oxygen transfer rate, which is used to calculate the current oxygen consumption (OUR) over a certain time period.

Bacterial growth (both wild type and mutant strains) was measured by the dry weight technique [41]. Residual waste frying oil was quantified by solvent extraction with petroleum ether and evaporation in vacuuo [42]. Biosurfactant was quantified by precipitation with 6 M HCl at 4 °C, solvent extraction with 2:1 mixture of chloroform: methanol, drying with anhydrous Na₂SO₄ and finally evaporating to constant weight in a rotary evaporator at 35 °C. The crude surface-active compound was further purified in silica gel (60–120 mesh) column, and fractions were eluted with a chloroform: methanol gradient mixture (18:1 to 2:1, v v^{− 1}). The obtained surface-active fractions were dried in a rotary evaporator to remove residual solvent and quantified by gravimetry [43]. All data were obtained in triplicates and used to determine glycolipopeptide yield and productivity.

Evaluation of mutant glycolipopeptide emulsifying ability, effectiveness and efficiency

The emulsification potential of the glycolipopeptide, expressed as emulsification index (E24) was evaluated as described by Kumar et al. (2024) using palm oil and kerosene. Surface and interfacial tensions which are measures of biosurfactant effectiveness were determined by the Du Nouy ring method at room temperature (28 ± 2 °C) using CSC Du Nouy tensiometer (CSC Scientific Company Inc, USA). The critical micelle concentration (CMC) which is a measure of biosurfactant efficiency was determined by plotting the logarithms of increasing concentrations of biosurfactant against surface tension.

Glycolipopeptide-mediated crude oil degradation

The produced glycolipopeptide was employed for enhanced degradation of Nigerian (Bonny) light crude oil (Exxon Mobil Nigeria Unlimited, Ibeno-Nigeria) using Pseudomonas putida strain ATCC 49,182. The bacterium was allowed to grow in Bushnell-Haas medium to an OD₆₀₀ of 1.0 in 250 mL Erlenmeyer flasks at 30 °C with agitation at 150 rpm for 18 h. Actively dividing cells were harvested by centrifugation at 3,075 x g for 10 min and resuspended in sterile Bushnell-Haas minimal medium to an OD₆₀₀ of 1.0.

Biosurfactant-enhanced crude oil degradation studies were monitored in two sets of triplicate 1000 mL Erlenmeyer flasks containing 100 mL sterile Bushnell-Haas minimal medium supplemented with 10 g L^{− 1} of Bonny light crude oil. The light crude oil had a density of 0.85 g/L at 15 °C and an American Petroleum Index (API) of 32.9. The biosurfactants were added at their critical micelle concentration: strain Thr⁻Ile⁻PGN4 glycolipopeptide at 20.8 mg L^{− 1}; di-rhamnolipid at 50 mg L^{− 1} and surfactin at 10 mg L^{− 1}. The preparations were inoculated with 2% (v v^{− 1}, OD₆₀₀ = 1.0) of active bacterial inoculum. One set was designated for determination of residual total petroleum hydrocarbons (TPH) while enzymes were determined from the second set to guard against crude oil loss during harvesting of samples. Triplicate uninoculated and inoculated-but-without biosurfactant controls were also set up. Flasks were incubated for 30 d on an orbital shaker at 200 rpm at 30 °C and sample withdrawals were made at 3 d interval for determination of residual TPH. Mean data was fitted to first-order kinetic model given below, using Excel Solver, to determine degradation kinetics and half-life of the degradation process [44]:

6

7

where C is the residual TPH concentration (mg L^− 1) at time t, C₀ the initial TPH concentration (mg L^− 1), k is the rate constant (d^− 1), t the time (d) duration of the study and Ln is the natural logarithm.

Determination of total petroleum hydrocarbon

Residual total petroleum hydrocarbon (TPH) was estimated by the gravimetric method after solvent extraction using n-hexane, dichloromethane and chloroform. The fractions were pooled and quantified gravimetrically. Triplicate data was subjected to one way ANOVA and significant means separated by Tukey HSD multiple comparison’s test at 5% significance level in GraphPad Prism 8.

Determination of total protein

Total protein was estimated by the Bradford method of protein determination using bovine serum albumin as standard [32]. Briefly, bacterial pellets were removed from 10 mL of degradation broth by centrifugation at 3,075 x g for 10 min at room temperature. The pellet was washed twice in 0.01 M phosphate buffer (pH 7.0), resuspended in same buffer, sonicated at 0 °C for 5 min and centrifuged at 20,376 x g for 25 min at 4 °C. The supernatants obtained from all sets were used for total protein and specific enzyme determinations.

Determination of catechol 1,2-dioxygenase

To detect cis, cis-muconate intermediate, the colorimetric assay method of Hegeman [45] was relied upon to detect the presence and concentration of catechol 1,2-dioxygenase (EC 1.13.11.1; catechol-oxygen 1,2-oxidoreductase; C12O) [46] at a pH of 7.0, specific wavelength, λmax of 260 nm; and a molar extinction coefficient, ε of 25,600 mol^{− 1} cm^{− 1}.

Determination of catechol 2,3-dioxygenase

To detect the presence of 2-hydroxymuconic semialdehyde, catechol 2,3-dioxygenase (EC 1.13.11.2; catechol-oxygen 2,3-oxidoreductase; C230) was assayed by the colorimetric method described in Adewale et al. [47] at a pH of 7.5, ε of 33,400 mol^{− 1} cm^{− 1} and λmax of 375 nm to detect the hydrolytic product.

Results and discussion

Glycolipopeptide overproducing mutants of Pseudomonas aeruginosa strain IKW1

The total number of survived strains obtained from the three-tier cocktail random mutagenesis was 1242 with 105 overproducing mutants. However, only 13 met the cut-off point of ≥ 200 cm² against the wild strain’s mean oil displaced area (ODA) of 143.56 ± 11.07 cm². The mutagen combinations that generated them and their associated oil displaced areas (cm²) are presented in Table S2 (Supplementary material). The most overproducing mutant, strain PGN4, gave a mean ODA of 268.64 ± 12.74 cm², indicating a 1.87-fold improvement in glycolipopeptide concentration. The overproduction screening relied on the oil displacement assay; a simple but sensitive technique that establishes a linear relationship between biosurfactant concentration and the ODA, and which can be used for preliminary quantitation of biosurfactants, even at microquantity levels [48]. This biosurfactant property is anchored on the ability of the compound to alter the contact angle encountered at the oil-water interface by reducing it and changing the wettability property of the surface from an oil-wet into a water-wet surface [49].

Results of stability studies revealed that only 2 of 13 glycolipopeptide overproducing mutants were stable over 660 h, namely strains PGN1 and PGN4 generated by first exposing the wild strain to plasma mutagenesis, followed by exposure to gamma irradiation before closing up with the chemical mutagen. From mean ODA results which presented a significant linear correlation (r = 0.9912, p < 0.0001) with glycolipopeptide concentration, mutant PGN4 was selected as most overproducing. It was observed that first exposure to plasma mutagenesis was the most important as it generated 9 of 13 overproducing mutants that gave ODA sizes ≥ 200 cm². Exposing the bacterium first to gamma irradiation generated zero mutant that met the cut-off. When strain IKW1 (wild strain) was first subjected to chemical mutagenesis, 4 mutants had ODA ≥ 200 cm², with the best (strain NPG1) displaying a mean ODA of 229.73 ± 21.16 cm². The success rate, therefore, of generating biosurfactant-overproducing strains of Pseudomonas aeruginosa strain IKW1 by alternate combinatorial random mutagenesis was highest by prior exposure to plasma mutagenesis compared to other tested mutagens. If plasma mutagenesis was followed with gamma mutagenesis, results were consistently better than when followed by the chemical mutagen. This therefore forced the conclusion that successful generation of overproducing mutants of strain IKW1 during combinatorial random mutagenesis is a function of the order of mutagenic exposures [50]. This could have arisen as a result of interactions between DNA lesions created by different mutagens, as well as the differences in the repair mechanisms evoked for specific mechanisms of mutagenesis. The type of mutation induced and timing of the cell cycle or DNA replication are other significant effects of mutagens mediated by the order of exposure. For instance, if the first mutant mediated cell cycle arrest or alteration in the timing of DNA replication, this may have significant, antagonistic, synergistic or additive influence on the second mutagenic effect [51].

The 16S rRNA partial gene sequences of both mutant and wild strains were aligned with those of other biosurfactant-producing strains of Pseudomonas aeruginosa retrieved from the GenBank (https://www.ncbi.nlm.nih.gov.). Results showed that the mutant strain shared an evolutionary clade with the wild strain with 100% sequence homology as Pseudomonas aeruginosa (Fig. 1a) indicating that they were still one and the same organism, and that the mutation was not of the 16S rRNA gene but on metabolic genes of the bacterium.

Fig. 1.

Phylogeny of mutant Thr-Ile-PGN4 mutant in comparison with its parent (a), comparative evaluation of hydrolytic enzyme activities (b), nickel uptake (c) and nickel sensitivities (d) of wild type and mutant strains of Pseudomonas aeruginosa strain IKW1

Auxanographic analysis of glycolipopeptide overproduction in mutant strain

Results of auxanography revealed that the mutant could grow only on one plate indicating auxotrophy for more than one growth factor, which were identified as threonine and isoleucine. The mutant was therefore fully named Pseudomonas aeruginosa strain Thr⁻Ile⁻PGN4. In Pseudomonas aeruginosa, three genes namely hom, thrB and thrC encoding homoserine dehydrogenase, homoserine kinase and threonine synthase respectively, are involved in threonine biosynthesis and any one of them could have been mutated [52]. Since protein synthesis was still going on, we assumed that methionine was still synthesized because methionine-specific aspartokinase-homoserine dehydrogenase, metL was still intact to catalyze the synthesis of methionine from homoserine through cystathionine. A defect in the gene, thrC encoding threonine synthase, ThrC which catalyzes the final step in threonine biosynthesis from the homoserine branch of the aspartate pathway led to threonine deficiency. Since isoleucine is derived directly from threonine in the aspartate pathway, it follows naturally that without threonine as substrate, there can be no isoleucine from that pathway. The successful use of auxanography to determine what growth factors were affected in a genetic alteration responsible for mutation was earlier reported by Wang et al. [53].

Involvement of hydrolytic enzymes in glycolipopeptide overproduction

Two major hydrolytic enzymes were found associated with glycolipopeptide production in the parent strain, namely, urease and lipase [26]. Co-production of biosurfactants with hydrolytic enzymes has been reported and is heralded as a viable and sustainable strategy to reduce overall cost of biosurfactant production [54]. In this study, urease activity of the two strains were compared and results showed a 1.82-fold increase in activity of strain PGN4 compared to the parent strain (Fig. 1b). However, there was no significant difference (p < 0.05) between the lipase activities of wild and mutant strains. We speculated that among other possible mechanisms, there was an upregulation of urease gene expression in the mutant leading to increased urea consumption, higher ammonium nitrogen availability, higher biomass concentration, hence, increase in glycolipopeptide titer. A more detailed genetic analysis would consolidate these hypotheses.

Involvement of nickel uptake and partitioning on glycolipopeptide overproduction

As nickel is required for urease maturation and activity, we evaluated the difference in amount of nickel taken up by the mutant in comparison with the wild strain during glycolipopeptide fermentation. Results showed a nickel uptake increase from 1.37 nmol Ni²⁺ mg^− 1 protein in the wild strain to 5.4 nmol Ni²⁺ mg^− 1 protein in the mutant corresponding to 3.94-fold increase (Fig. 1c). Interestingly, while the amount of nickel uptake became stationary between 66 and 72 h of fermentation in the parent, that in the mutant kept increasing beyond 72 h suggesting higher nickel tolerance hence, higher nutritional requirement which necessitated a nickel sensitivity evaluation. The results, as presented in Fig. 1d, showed that strain Thr⁻Ile⁻PGN4 could take up 272.64 nmol Ni²⁺ mg^− 1 protein in one hour compared to 148.34 nmol Ni²⁺ mg^− 1 protein; some 1.84-fold increase in nickel uptake by the mutant. This increase in nickel uptake significantly correlated (R = 0.9752; p < 0.00001) with increased urease activity. Approximately, 60% of total nickel was found in intact wild strain within 72 h while 99.6% was found in the mutant. This indicated superior nickel uptake ability of the mutant compared to the wild strain. While 80% of the nickel taken up by wild strain was found in the membrane (68.5%) and periplasm (11.5%), 70% of the nickel was partitioned to the membrane (42.3%) and periplasm (27.7%) in the mutant. The 30% of nickel found in the cytosol of the mutant bacterium inferred 2.41-fold higher nickel levels required for synthesis of molecules of nickel-requiring enzymes like urease and hydrogenase in the mutant.

Effect of carbon sources on nickel uptake toward glycolipopeptide overproduction

Figure 2a depicts a requirement for suitable carbon sources for increased nickel uptake which increased cytosolic nickel to 38.5% in the presence of glucose suggesting strong preferential exterior-interior partitioning of nickel in cells. The result also correlates differential nickel partitioning with its properties, thus giving clues to the processes that nickel may be involved in. A total nickel uptake concentration of 662.74 nmol Ni²⁺ mg^{− 1} protein was mediated by glucose in the mutant which was significantly higher than the 237.05 nmol Ni²⁺ mg^{− 1} protein in the wild type. Gallic acid and phenol also increased nickel uptake while sucrose and sodium succinate, acting as metal chelators, inhibited nickel uptake in both mutant and wild type strains, which corroborates earlier observations by Sar et al. [55].

Fig. 2.

Comparative effects of carbon sources (a) and phosphates (b) on nickel uptake, and the activities of histidinol dehydrogenase (c) and histidinol phosphate phosphatase (d) of wild and mutant strains of Pseudomonas aeruginosa strain IKW1

Effect of phosphates on nickel uptake toward glycolipopeptide overproduction

In our previous studies on glycolipopeptide production, we used inorganic phosphates as sources of phosphorus and buffering system [10, 26]. As inorganic phosphates are significant chelators of metals, the present study substituted the inorganic phosphate with an organic source (0.250 mM β-glycerophosphate) and observed significant (p < 0.05) increase in nickel uptake by the mutant from 662.74 to 855.73 nmol Ni²⁺ mg^{− 1} protein suggesting propensity for improved glycolipopeptide production via enhanced urease activity (Fig. 2b). By extension, we also compared the activities of phosphate-involving enzymes of the strains, particularly histidinol dehydrogenase and histidinol phosphate phosphatase. While the former catalyzes the last step in histidine biosynthesis, oxidizing L-histidinol to histidine, the later catalyzes the dephosphorylation of L-histidinol phosphate into L-histidinol, therefore serving to supply the substrate for the former. Our results showed significant upregulation of both enzymes in the mutant with histidinol dehydrogenase activity increasing from 223.78 to 298.64 NAD⁺ reduced min^{− 1} mg^{− 1} protein in the presence of 0.25 mM of β-glycerophosphate. Substrate concentration above 0.25 mM led to significant (p < 0.05) decrease in enzyme activity in both mutant and wild stains (Fig. 2c).

A similar situation was observed for histodinol phosphate phosphatase which significantly reduced with increase in phosphate concentration in both mutant and wild type strains of Pseudomonas aeruginosa strain IKW1. Peak enzyme activity occurred at 0.25 mM of β-glycerophosphate with 1,799 nmol of inorganic phosphate h^{− 1} mg^{− 1} protein (Fig. 2d). Two-way analysis of variance of data revealed no significant interaction (p = 0.5735 > 0.05) between phosphate concentration and bacterial strain used. However, the phosphate concentration and bacterial strain significantly (p < 0.0001) influenced histidinol phosphate phosphatase activity individually. The increased activities of these enzymes lend credence to the influence of the hydrolytic enzymes in glycolipopeptide production by the mutant. These enzymes together, make histidine available for urease maturation; a process which involves post-translational insertion of nickel into urease at the catalytic site using carbamylated lysine [56]. This therefore substantiates why glycolipopeptide overproduction would metabolically flux materials away from threonine and isoleucine, in favour of the lysine route of the aspartate pathway. The enhanced uptake of nickel by the mutant could be exploited in bioremediation of nickel-impacted agricultural soils toward improved soil fertility for sustainable agricultural practice as summarized in the review by Datta et al. [57].

Response surface optimization of medium variables for glycolipopeptide overproduction

By reason of the involvement of glucose, nickel and phosphate in urea hydrolysis, or these factors were optimized by RSM for inclusion in a previously optimized medium for the wild strain to cater to the nutritional requirements of the improved strain. The developed quadratic model, given as Eq. 8, was significant (adjusted R² = 0.9861, p < 0.0001) with a non-significant lack-of-fit of F = 1.13, p = 0.4753 > 0.05 suggesting model adequacy and accuracy (Table 1).

8

Table 1.

ANOVA for glycolipopeptide production RSM modeling

Source	Sum of Squares	df	Mean Square	F-value	p-value
Model	1251.96	14	89.43	147.88	< 0.0001	significant
A-Glucose	68.95	1	68.95	114.02	< 0.0001
B-Urea	874.11	1	874.11	1445.48	< 0.0001
C-Nickel	3.81	1	3.81	6.30	0.0241
D-Phosphate	252.98	1	252.98	418.34	< 0.0001
AB	5.38	1	5.38	8.90	0.0093
AC	3.63	1	3.63	6.00	0.0271
AD	12.53	1	12.53	20.72	0.0004
BC	1.68	1	1.68	2.77	0.1166
BD	1.10	1	1.10	1.82	0.1970
CD	0.4830	1	0.4830	0.7988	0.3856
A²	3.72	1	3.72	6.15	0.0255
B²	2.52	1	2.52	4.17	0.0592
C²	17.04	1	17.04	28.17	< 0.0001
D²	0.2574	1	0.2574	0.4257	0.5240
Residual	9.07	15	0.6047
Lack of Fit	6.29	10	0.6285	1.13	0.4753	not significant
Pure Error	2.79	5	0.5571
Cor Total	1261.03	29

where Y is the glycolipopeptide concentration; x₁ to x₄ indicate design codes for the explanatory variables corresponding to glucose, urea, nickel and phosphate concentrations respectively.

Figure S1 (supplementary material) summarizes the optimized perturbations in the medium during glycolipopeptide fermentation, the model accuracy plot and the optimized 3-dimensional relationship (surface plot) between explanatory variables and the response. The coefficient estimates, which were used to develop the model, clearly put urea (6.035) as the most significant variable to which the bioprocess was most sensitive, followed by phosphate (3.247). The urea coefficient, for instance, estimates the expected change in glycolipopeptide concentration when there is a unit change in urea concentration, while holding all other variables constant. The developed model was optimized to give a glycolipopeptide concentration of 92.75 g L^− 1 at explanatory variable levels of 4.22 g L^− 1 glucose, 2.97 g L^− 1 urea, 900 µM nickel and 0.55

g L^− 1 phosphate (β-glycerophosphate). The optimized glycolipopeptide production medium was validated in real time to give a glycolipopeptide biosurfactant concentration of 91.84 g L^− 1 which was not significantly different from the predicted optimized solution. The final modified optimized medium was therefore composed (g L^− 1) of filter-sterilized urea 2.97, β-glycerophosphate 0.55, MgSO₄.7H₂O 0.2, NaCl 0.5, CaCl₂ 0.5, KCl 0.5 and NiCl₂ 0.053. It also contained trace element solution (1 mL) composed (mg L^− 1) of ZnCl₂ 0.125, FeCl₂ 0.125, K₃BO₃ 0.0104, CuSO₄.5H₂O 0.025 and was supplemented with 5% (v v^− 1) waste frying oil and 4.22 g L^− 1 glucose. The medium was buffered by HEPES (N-2-hydroxyethylpiperazine-N′-ethane-sulphonic acid) to maintain pH at 7.6.

Glycolipopeptide overproduction in bench-scale bioreactor

Mutant glycolipopeptide biosurfactant production in 5-L bioreactor revealed a higher oxygen demand as evidenced in the significantly higher kLa value of 104.54 h^{− 1} compared to 84.51 h^{− 1} of the wild strain at an aeration and agitation rates of 1.5 vvm and 350 rpm, respectively (Tables 2 and 3). This led to superior biomass accumulation by the mutant which translated to 3.87-fold higher glycolipopeptide productivity corresponding to 3.83 ± 0.21 g L^{− 1} h^{− 1} compared to 0.99 ± 0.01 g L^{− 1} h^{− 1} in the wild type. This indicates successful engineering of the strain by the technique of three-tier cocktail mutagenesis combined with bioprocess optimization in a bioreactor. Bazsefidpar et al. [58] reported a rhamnolipid productivity of 0.9 g L^{− 1} h^{− 1} using fed-batch mode of fermentation. The higher productivity of the biomolecule in this study may be attributed to the huge success recorded by strain engineering.

Table 2.

Effect of volumetric oxygen transfer coefficient, k_La, on glycolipopeptide biosurfactant production by wild strain

Agitation (rpm)	Aeration (vvm)	kLa (h− 1)	OSR (mmoles L− 1)	Duration (h)	WFO utilization (g L− 1)	Bacterial growth ([g L− 1] @ OD600)	Biosurfactant concentration (g L− 1)	Biosurfactant productivity (g L− 1 h− 1)	Biosurfactant yield on biomass (g g− 1)
200	0.5	24.67 ± 1.53	5.92 ± 0.39	72	37.00 ± 0.83	20.88 ± 1.23	15.44 ± 0.53	0.21 ± 0.01	0.74 ± 0.02
200	0.75	37.34 ± 1.11	8.96 ± 0.47	72	33.40 ± 1.07	20.29 ± 0.85	16.65 ± 0.36	0.23 ± 0.02	0.82 ± 0.01
200	1.0	44.16 ± 1.25	10.60 ± 0.55	60	35.00 ± 1.12	20.73 ± 1.27	16.76 ± 0.62	0.28 ± 0.01	0.81 ± 0.02
200	1.25	51.55 ± 1.46	12.37 ± 0.68	60	30.20 ± 0.71	20.90 ± 1.42	17.44 ± 0.77	0.29 ± 0.02	0.83 ± 0.01
200	1.5	58.53 ± 1.79	14.05 ± 0.49	48	30.27 ± 0.53	19.61 ± 1.22	18.37 ± 0.28	0.38 ± 0.01	0.83 ± 0.02
250	1.5	68.45 ± 1.55	16.43 ± 0.94	36	29.18 ± 0.45	22.94 ± 1.34	18.97 ± 0.43	0.53 ± 0.02	0.83 ± 0.02
300	1.5	76.46 ± 1.94	18.35 ± 0.62	36	28.37 ± 1.26	27.36 ± 0.78	22.24 ± 0.39	0.62 ± 0.02	0.81 ± 0.01
350	1.5	84.51 ± 2.16	20.29 ± 0.88	24	30.46 ± 0.63	28.05 ± 0.67	23.86 ± 0.64	0.99 ± 0.01	0.85 ± 0.02
400	1.5	85.75 ± 1.72	20.58 ± 1.15	24	23.74 ± 1.13	27.20 ± 0.74	18.96 ± 1.21	0.79 ± 0.02	0.68 ± 0.01
450	1.0	87.13 ± 1.71	20.91 ± 1.36	24	18.62 ± 1.12	25.52 ± 0.51	15.92 ± 1.45	0.66 ± 0.01	0.62 ± 0.03

Table 3.

Effect of volumetric oxygen transfer coefficient, k_La, on glycolipopeptide biosurfactant production by mutant strain

Agitation (rpm)	Aeration (vvm)	kLa (h− 1)	OSR (mmoles L− 1)	Duration (h)	WFO utilization (g L− 1)	Bacterial growth ([g L− 1] @ OD600)	Biosurfactant concentration (g L− 1)	Biosurfactant productivity (g L− 1 h− 1)	Biosurfactant yield on biomass (g g− 1)
200	0.5	34.61 ± 1.03	5.92 ± 0.39	72	65.00 ± 0.63	50.88 ± 1.03	42.29 ± 0.23	0.59 ± 0.11	0.83 ± 0.11
200	0.75	41.24 ± 0.87	8.96 ± 0.47	72	63.40 ± 1.59	55.29 ± 0.47	45.66 ± 0.42	0.63 ± 0.32	0.83 ± 0.01
200	1.0	49.77 ± 1.52	10.60 ± 0.55	60	64.00 ± 1.16	60.73 ± 1.32	48.54 ± 0.32	0.81 ± 0.14	0.80 ± 0.12
200	1.25	60.43 ± 1.22	12.37 ± 0.68	60	62.20 ± 0.44	70.90 ± 0.52	57.44 ± 0.33	0.96 ± 0.02	0.81 ± 0.05
200	1.5	64.42 ± 1.56	14.05 ± 0.49	48	61.27 ± 0.52	82.61 ± 0.41	68.33 ± 0.32	1.42 ± 0.13	0.86 ± 0.07
250	1.5	75.39 ± 1.82	16.43 ± 0.94	36	59.18 ± 0.55	94.14 ± 1.11	78.11 ± 0.13	2.17 ± 0.02	0.83 ± 0.12
300	1.5	89.86 ± 1.66	18.35 ± 0.62	36	57.37 ± 1.22	97.36 ± 0.55	82.64 ± 0.26	2.30 ± 0.34	0.85 ± 0.14
350	1.5	104.54 ± 3.1	20.29 ± 0.88	24	60.46 ± 0.86	102.05 ± 0.87	91.82 ± 0.23	3.83 ± 0.21	0.90 ± 0.01
400	1.5	92.66 ± 1.57	20.58 ± 1.15	24	53.74 ± 1.04	96.20 ± 0.66	83.44 ± 0.15	3.48 ± 0.18	0.87 ± 0.02
450	1.0	95.06 ± 1.25	20.91 ± 1.36	24	48.62 ± 1.15	90.52 ± 0.72	75.94 ± 0.27	3.16 ± 0.19	0.84 ± 0.04

k_La Volumetric oxygen transfer coefficient, OSR Oxygen supply rate, OD Optical density, C* 0.24 mmoles L

Biosurfactant yield is often reported to improve significantly with strain improvement but this study did not find that perhaps owing to some terminology mix-up in what actually constitutes a yield. Only very few studies clarify what yield is presented. In our study, biosurfactant yield on biomass (Yp/x) was calculated as 0.90 g g^{− 1} which did not differ significantly from that by the wild type bacterium under similar conditions. However, glycolipopeptide yield on substrate (Yp/s) was 1.52 g g^{− 1} which was significantly different (by 1.95-fold) from 0.78 g g^{− 1} by wild strain. Xu et al. [13], while using a somewhat similar technique, improved a strain of Pseudomonas with 2.7-fold higher unspecified yield than the parent strain. Recently, Liu et al. [59], using a similar technique of combinatorial random mutagenesis involving atmospheric and room temperature plasma on a strain of Pseudomonas aeruginosa obtained a high-yielding strain with 4.67-fold unspecified yield increase compared to the parent suggesting great promise for the alternate combinatorial random mutagenesis technique in this study.

Confirmation of surface activities and chemical composition of glycolipopeptide

Confirmatory surface-active properties of the glycolipopeptide revealed that surface tension decreased significantly from 72.63 ± 0.01 dynes cm^{− 1} to 29.72 ± 0.01 dynes cm^{− 1} within 30 h while the interfacial tension reduced from 21.88 ± 0.02 dynes cm^{− 1} to 3.02 ± 0.02 dynes cm^{− 1} within the same period (Table 4). This period also shows increased biomass from 0.342 at time 0 h to 4.825 gDCW L^{− 1} at 30 h. Interestingly, the period showed presence of rhamnose sugar from 6 h where surface and interfacial tensions started to drop. However, there was a very clear absence of peptides (or proteins) in the surface-active compound, indicating that the surface-active compound contained rhamnose as carbohydrate linked to the lipid moiety. At 36 h, presence of protein was detected and was confirmed by TLC. As the rhamnose test continued to be positive in the presence of peptide moieties, it suggested that the rhamnolipid was simply peptidated into a glycolipopeptide compound [5]. This brought about further decrease in SFT from 29.72 dynes cm^{− 1} at 30 to 24.41 dynes cm^{− 1} at 48 h, above which no further decrease in SFT was found. The table also indicates a similar scenario for interfacial tension (IFT) where the late peptidation of the rhamnolipid molecule decreased IFT from 3.02 dynes cm^{− 1} at 30 h to 0.87 dynes cm^{− 1} by 48 h. The non-significant (p > 0.05) difference among biomass concentration between 30 and 72 h indicates late exponential-to-stationary phase of growth of the mutant. This implies that the rhamnolipid base was synthesized during trophophase while its peptidation into glycolipopeptide occurred during the idiophase of bacterial growth. Our glycolipopeptide is very well corroborated by the work of Koronelli et al. [6] on the pioneer peptidoglycolipid. However, a detailed chemical and structural characterization of the biomolecule using state-of-the-art instrumentation such as nuclear magnetic resonance and liquid chromatography-mass spectrometry is required for complete elucidation of structure-function of the bio-amphiphile.

Table 4.

Periodic confirmation of chemical composition of mutant biosurfactant and its surface properties during fermentation

Time (h)	BMC (gDCW L-1)	SFT (dynes cm-1)	IFT (dynes cm-1)	kE24 (%)	pE24 (%)	tCHO	Rhamnose	Proteins	Lipids
0	0.342 ± 0.01	72.63 ± 0.01	21.88 ± 0.02	-	-	+	-	-	+
6	0.553 ± 0.01	68.34 ± 0.02	18.53 ± 0.03	3.27	9.73	+	+	-	+
12	0.992 ± 0.01	62.62 ± 0.01	13.23 ± 0.01	17.46	17.36	+	+	-	+
18	1.594 ± 0.01	54.76 ± 0.01	8.84 ± 0.02	28.54	28.33	+	+	-	+
24	2.731 ± 0.01	38.51 ± 0.01	4.11 ± 0.01	42.48	41.63	+	+	-	+
30	4.825 ± 0.01	29.74 ± 0.02	3.02 ± 0.02	68.56	77.38	+	+	-	+
36	4.921 ± 0.01	26.46 ± 0.02	1.10 ± 0.02	76.58	79.36	+	+	+	+
42	5.001 ± 0.02	24.82 ± 0.01	0.98 ± 0.02	82.41	82.36	+	+	+	+
48	5.022 ± 0.01	24.41 ± 0.01	0.87 ± 0.02	82.36	88.64	+	+	+	+
54	5.034 ± 0.01	24.41 ± 0.02	0.87 ± 0.01	82.37	91.38	+	+	+	+
60	5.033 ± 0.01	24.41 ± 0.01	0.87 ± 0.02	82.38	91.41	+	+	+	+
66	5.001 ± 0.01	24.41 ± 0.02	0.87 ± 0.01	82.38	91.42	+	+	+	+
72	4.981 ± 0.01	24.41 ± 0.01	0.87 ± 0.01	82.38	91.39	+	+	+	+

The emulsification activities of the biomolecule also followed the same pattern and were not significantly different from those earlier reported for the wild type [10]. Future work on the chemical and structural elucidation of the biomolecule is required to unravel further inherent potencies of the macromolecule and mechanisms of action to direct suitable applications.

Comparative biosurfactant-mediated degradation of Nigerian light crude oil revealed significant differences among the three tested biosurfactant types namely glycolipid (di-rhamnolipid), lipopeptide (surfactin) and the new class of glycolipopeptide biosurfactant. The well documented crude oil degrading bacterium, Pseudomonas putida was able to degrade 5.424 g L^{− 1} of 10 g L^{− 1} light crude at a rate constant, k, of 0.065 d^{− 1} in the absence of an extraneous biosurfactant, giving a process half-life of 10.66 d. Degradation of the oil followed a similar pattern in the presence of di-rhamnolipid and surfactin added at their critical micelle concentrations (cmc) over the 30-d degradation study. Biodegradation rates followed first-order kinetic model [60], and revealed a k of 0.096 d^{− 1} for surfactin and 0.087 d^{− 1} for di-rhamnolipid (Fig. 3a), giving degradation half-lives of 7.22 and 7.97 d, respectively. However, glycolipopeptide-mediated degradation was superior in terms of amount of oil removed and the rate at which degradation occurred; by some 2.67-fold higher degradation rate using first-order kinetics. The glycolipopeptide facilitated the degradation of 9.994 g L^{− 1} of crude oil at a rate of 0.33 g L^{− 1}d^{− 1}. At a rate constant, k, of 0.115 d^{− 1}, the half-life of oil degradation by the organism in the presence of glycolipopeptide biosurfactant, added at its cmc of 20.86 mg L^{− 1}, was 6.03 d; some 1.77-fold fewer days than in its absence. This superior mediation of crude oil degradation may be attributed to its superior emulsification property compared to other biosurfactant classes, where the biosurfactant demonstrated 82.4% emulsifying activity in kerosene and 91.4% in palm oil, underscoring its potential for application in detergent formulations.

Fig. 3.

Kinetics of biosurfactant-mediated degradation of Nigerian light crude oil by Pseudomonas putida strain ATCC 49,182 (a), total protein kinetics (b), Catechol 1,2-dioxygenase kinetics (c) and catechol 2,3-dioxygenase kinetics (d) over a 30-day degradation period

Glycolipopeptides are more commonly classified as polymeric (high molecular weight) biosurfactants, making them better emulsifiers and therefore better solubilizers of crude oil than other classes. Accordingly, in the presence of a bio-emulsifier, a significantly more hydrophobic carbon substrate is made available at the boundary for uptake and metabolism. Additionally, the availability of more substrates triggered early synthesis and activity of hydrocarbon utilizing enzymes and induction of the appropriate catabolic pathway. Figure 3b reveals that total protein increased periodically until it reached its peak of 2.08 mg in systems containing glycolipopeptide, which was significantly higher than the total protein accumulated in the absence of other extraneous bio-emulsifier (0.81 mg).

In this study, catechol-1,2-dioxygenase activity was induced in the bacterium to its peak at 6 d with a specific activity of 42.81 U mg^− 1 in the presence of glycolipopeptide. Other biosurfactants also mediated peak activity of the enzyme on the same day (Fig. 3c). However, activities mediated by surfactin and di-rhamnolipid were not statistically different from each other as surfactin gave 35.69 U mg^− 1 while di-rhamnolipid gave 34.20 U mg ^− 1. The activity of catechol 2,3-dioxygenase reached its peak on the 9th day and its peak was highest in the glycolipopeptide-mediated system with a specific activity of 7.76 U mg^− 1 (Fig. 3d). In the absence of extraneous biosurfactant, catechol 1,2-dioxygenase was induced on the 9th day while catechol 2,3-dioxygenase was had its peak at 18 d. From the foregoing results, it implies that Pseudomonas putida strain ATCC 49182 predominantly degrades aromatic hydrocarbon using the ortho-pathway, although the meta-pathway is also operable.

Conclusion

Our three-tier cocktail alternate combinatorial random mutagenesis approach successfully generated a glycolipopeptide hyperproducing mutant, Pseudomonas aeruginosa strain Thr⁻Ile⁻PGN4, when the wild strain was first exposed to ARTP, then gamma rays and finally to the chemical mutagen. Physiological and biochemical characterizations revealed that the mutant was auxotrophic for threonine and isoleucine, with enhanced urease activity, nickel uptake, histidinol dehydrogenase and histidinol phosphate phosphatase activities. Response surface optimization revealed urea as the explanatory factor to which glycolipopeptide production process was most sensitive. Bioreactor production of the biomolecule revealed higher oxygen uptake rate in the mutant leading to significant increase in glycolipopeptide productivity. Glycolipopeptide-mediated degradation of Nigerian light crude oil by Pseudomonas putida was superior to other biosurfactants at a rate of 0.33 g L^− 1 d^− 1 and a rate constant of 0.115 d^− 1. The mutant is recommended for bioremediation of nickel-contaminated systems and the biosurfactant for sustainable bioremediation applications of hydrophobic compound-impacted systems.

Authors’ contributions

M.E. and S.A. conceived the idea; M.E., P.E. & L.A. handled materials acquisition; M.E., P.E., L.A. & I.O. handled the methodology; M.E. & B.E. handled data analysis; M.E., P.E. wrote the first draft; B.E. & S.A reviewed and edited the work; All authors reviewed the final version of the manuscript.

Funding

None was received.

Data availability

Sequence data that support the findings of this study have been deposited at the National Center for Biotechnology Information (NCBI) with the GenBank Accession No.: PV664482.1. The parent and mutant strains have been deposited in the University of Calabar Collection of Microorganisms (UCCM). All other data are available in the main text and in the supplementary material.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.