Identification of Corn Chaff as An Optimal Substrate for the Production of Rhamnolipids in Pseudomonas aeruginosa Fermentations

1. Introduction

The production chains of the agricultural sector generate abundant by-products and organic waste (herein also indicated as “agro-waste” or “waste agricultural biomass”). The quantities of the above are often evaluated only marginally by the official statistics of the sector, causing underestimates and inconsistent data, which do not provide a clear and exhaustive cognitive framework. The latest available data on the production of vegetable waste of agricultural origin in Italy refers to 1997, the year in which the Italian production exceeded 20 million tons per year of dry matter, of which the majority (approximately 13 million tons per year) came from cereal production. This corresponds to 2–4 % of the product obtained (a percentage which, in the case of the presence of mycotoxins, can grow up to 30%), which are mainly sent to anaerobic digestion plants for the production of biogas [1,2]. Due to the composition, mainly consisting of cellulose, hemicellulose and lignin [3] agro-waste has the characteristics that allow the growth of suitable microorganisms and the potential production of their metabolites [4,5,6,7]. To convert agro-waste into substrates for microbial fermentation, different processes were developed including grinding and milling, acid hydrolysis, alkali pretreatment, hydrothermal treatment, and/or enzyme hydrolysis. The aim of the pre-treatment is to break down the lignocellulosic complex and to increase the bioavailability of sugars (and eventually of other relevant nutrients) present in cellulose and hemicellulose [8]. After the pre-treatment, the agro-wastes (in the form of liquid solutions of nutrients and/or solid residues) can be eventually combined with other nutrients to be used as substrates for microbial growth and production of microbial products[9,10,11,12,13].

Among the molecules which can be conveniently produced by microbial fermentation of agro-waste, worth of note are biosurfactants (BS) and bioemulsifiers (BE) [14,15,16]. Microbial-derived BS and BE can act as surface-active compounds (SAC). These compounds possess amphiphilic properties, characterized by their distinct hydrophobic and hydrophilic regions, which facilitate the emulsification and dispersion of hydrophobic substances in a hydrophilic environment. BS are known for their excellent surface activity which involves lowering the surface and interfacial tension (ST) between different phases (liquid-air, liquid-liquid, and liquid-solid) and possess a low critical micelle concentration (CMC) which allows the formation of stable emulsions.

Structurally, the biosurfactants are highly diverse and can be classified as glycolipids, lipopeptides, phospholipids, lipopolysaccharides, fatty acids, and polymers. They find applications in cosmetics, personal care products, and household cleanings, while their potential in pharmaceuticals, environmental clean-up, agriculture, and food industries are also being explored. Glycopeptides and lipopeptides are much represented in commercial products and are the most studied [17].

BE are higher than BS in molecular weight as they are complex mixtures of heteropolysaccharides, lipopolysaccharides, lipoproteins and proteins. BE are also known as high molecular weight biopolymers or exopolysaccharides. Similar to BS, BE can efficiently emulsify two immiscible liquids such as hydrocarbons (or other hydrophobic substrates) and water but are less effective than BS at ST reduction. Therefore, BE are generally described as possessing emulsifying activity without surface activity [18].

BS and BE have several advantages over the chemical surfactants such as lower toxicity, higher biodegradability, better environmental compatibility and higher selectivity and specificity at extreme temperature, pH and salinity. [19,20]. BS and BE of microbial origin are produced by genera such as Bacillus, Pseudomonas, and Candida, that produce low molecular weight SAC, and Cromobacter, Mucor and Acinetobacter, that produce high molecular weight SAC [11,21,22]. At present, members of the genera Pseudomonas, Bacillus, Rhodococcus and Candida are the most widely used in the industrial production of these biomolecules [17,23].

Most BS are low molecular weight microbial amphiphilic molecules that typically consist of a hydrophilic head group and a hydrophobic tail and can include a diversity of subunits such a sugars, fatty acids, amino acids, and carboxylic acid groups [24]. Pseudomonas and Bacillus are great surfactants producer[25]. In particular, Pseudomonas strains have been reported as efficient producers of the glycolipids rhamnolipids, the most intensively studied biosurfactants. [26,27] while bacteria of the Bacillus genus mainly produce surfactin, a lipopeptide biosurfactant.

Rhamnolipids are composed of one or two rhamnose moieties (mono-rhamnolipids or di-rhamnolipids) linked to β-hydroxy fatty acid chains that vary in number, length and degree of unsaturation [28]. Approximately sixty rhamnolipids congeners and homologues have been described. The predominant rhamnolipid species and the relative concentrations of the congeners are dependent on the rhamnolipids producing strains [29,30]. At present, Rhamnolipids are used mainly in the petrochemical industry, in the bioremediation of different pollutants, in agricultural chemicals, and in personal care products. In addition, they present antimicrobial activity, and they show low human and environmental toxicity[29,31]. They are currently also considered for application in the pharmaceutical industry [32].

Surfactin is the best-known lipopeptide biosurfactant produced by Bacillus. Surfactin is a cyclic lipopeptide composed of a heptapeptide with the following sequence: L-Glu1-L-Leu2-D-Leu3-L-Val4-L-Asp5-D-Leu6-L-Leu7, forming a lactone ring structure with a b-hydroxy fatty acid chain. Bearing both, a hydrophilic peptide portion and a lipophilic fatty acid chain. Bacillus subtilis, Bacillus amyloliquefaciens and Bacillus licheniformis are the main producers of this compound. In particular, B. subitlis, can produce natural surfactins as a mixture of isoforms A, B, C and D with various physiological properties. They contain at least eight depsipeptides with the number of carbon atoms between 13 and 16 as part of the ring system. Surfactin (the most powerful biosurfactant discovered so far) lowers the surface tension of water from 72 to 27 mN m^-1, at a critical micelle concentration (CMC) of 20 mg L^-1. It thereby displays strong emulsifying and foaming activities [33].

Among BE, the most relevant is Emulsan, a lipopolysaccaharide bioemulsifier with a molecular weight of 1000 kDa produced by Acinetobacter calcoaceticus RAG-1. It is one of the most widely studied emulsifiers from bacteria [34]. In pure form, emulsan shows emulsifying activity at low concentrations (0.01–0.001%). It increases the bioavailability of poorly soluble substrates in aqueous environments for microbial access and degradation by coating the hydrophobic substrate to form minicapsules. The producing bacterium can also have direct access to hydrophobic substrates, but the emulsifying activity is exhibited by secreted emulsan.

Draw backs in the large-scale production of microbial SAC, can be the low production yield, expensive recovery and purification, and the high costs of the substrates required for fermentation. On the other side, SAC of microbial origin have a range of applications (from medical, to cosmetics to crude oil recovery, to bioremediation) [17,35], which enhances the possibility to use different kinds of raw materials for fermentation depending on the value of the application of the final product. Bioremediation applications, in example, limit the requirements of standardized raw materials in the production of BS and consequently reduce the costs. Furthermore, as vegetable waste is commonly used as amending agent in soil bioremediation, the fermentation of vegetable waste to produce biosurfactants and the subsequent application of the whole fermentation broth to the contaminated soils, could shorten the chain of the process and consistently reduce the costs. Besides, the microorganisms used to produce BS could also be selected among those able to degrade important environmental pollutants such as Poly Aromatic Hydrocarbons (PAH) [36,37,38].

In conclusion, even if the BS market remains small compared to that of synthetic surfactants, mainly due to higher production costs, the use of agricultural wastes such as the abundant cereal waste is considered a promising strategy to make these products economically more competitive [39]. In this study we report the cross-identification of microorganisms able to produce biosurfactants and suitable agro-waste for the fermentation media formulation. Surprisingly, high productivities of Rhamnolipids and Surfactin were obtained by use of corn chaff and emmer and oat spelts respectively. Most relevant in our study was the fact that no pre-treatment of the agro-waste was required to reach high productivity of BS.

This study was realized within the frame of the collaborative project “Rifiuti cerealicoli per il biorisanamento”, acronym “RICREA” (https://www.progetto-ricrea.org/). The RICREA project was funded by the Italian Ministry for the Environment. The goal of the RICREA project was to evaluate the possibility of recovering and valorising wastes and scraps from the production and processing of cereals and legumes. Specifically, these wastes were ideally to be used as substrates to produce biosurfactants which, in turn, had to be used for the bioremediation of hydrocarbons contaminated soils.

2. Materials and Methods

2.1. Microbial Strains, Culture Media and Culture Conditions

The microorganisms used in this study were supplied by the company Madep SA and have been isolated from soils contaminated by hydrocarbons (Table 1). The isolated strains were able to produce biosurfactants at industrially interesting levels in standard media for fermentation.

The media used for the growth of the microbial strains, and for biosurfactant production were obtained from Bioc-CheM Solutions proprietary media database (BCSMedDat). All the media were prepared in 500 mL baffled flasks and for each flask a volume of 100 mL was dispensed. The media were sterilized at 121–123 °C for 20–25 minutes. The pH of the media was measured before and after sterilization by use of a pHmeter (Mettler-Toledo, MP 120).

To perform viable cell count and revitalize the microbial strains, Luria-Bertani (LB) agar medium, Malt Extract (ME) agar medium and Nutrient-broth (NB) agar medium were used. For LB agar medium 25 g of LB powder (10 g yeast extract; 10 g sodium chloride; 5 g tryptone) (Difco) and 18 g of agar (Hi-Media) were dissolved in 1 L of ultrapure water and sterilized at 121 °C for 15 minutes. The pH post sterilization was 7.00 ± 0.1. For ME agar medium, 20 g of Malt Extract powder (17 g malt extract; 3 g peptone) (Sigma)-Aldrich, and 18 g of agar (Hi-Media) were dissolved in 1 L of ultrapure water and sterilized at 121 °C for 15 minutes. The pH post sterilization was 5.6 ± 0.2. For NB agar medium, 8 g of NB powder (3 g beef extract; 5 g peptone) (Difco) and 18 g of agar (Hi-Media) were dissolved in 1 L of ultrapure water and sterilized at 121 °C for 15 minutes. The pH post sterilization was 6.8 ± 0.2. All agar media prepared were poured in Petri dishes before use. The seed (vegetative) phase of growth for the bacterial strains was carried out in LB or NB broth. 25 g of LB powder (Difco) (10 g yeast extract; 10 g sodium chloride; 5 g tryptone) were dissolved in 1 L of ultrapure water and sterilized at 121 °C for 15 minutes. The pH post sterilization was 7.00 ± 0.1. After sterilization, 100 mL of LB broth were dispensed into sterile 500 mL baffled flasks. To control foam formation during the fermentation, 50 µL of sterile antifoam (O-10, Sigma-Aldrich) were added to each flask. The seed (vegetative) phase of growth for yeast C. bombicola was carried out in BMGY medium (10 g L^-1 yeast extract, 20 g L^-1 peptone, 10 g L^-1 glycerol, 400 μg L^-1 biotin, and 0.1 M potassium phosphate buffer at pH 6.0) [40].

The fermentation (production) media used for production of biosurfacants are detailed in the experimental section. Media are mainly composed of glycerol, glucose and soybean oil as carbon source, and yeast extract or sodium nitrate (NaNO₃) as the nitrogen source. Potassium dihydrogen phosphate (KH₂PO₄) was added as P source or to control pH whenever required. Each production medium (except the controls) was also supplemented with the different agro wastes (Figure S1) which were finely grinded before use (through the use of a cereal mill equipped with a 40 - mesh sieve (Figure S2).

Saline solution (NaCl 0.9%) was prepared dissolving 9 g of sodium chloride (NaCl) in 1 L of ultrapure water and sterilizing at 121 °C for 20 minutes. The saline solution was used to prepare serial dilutions for viable cell count and to suspend the microorganisms for the inoculum.

The nutrient glycerol solution was prepared dissolving 20 g L^-1 of nutrient broth and 200 g L^-1 of glycerol in ultrapure water. The solution was sterilized at 121 °C for 15 minutes.

For the preparation of the Master Cell Banks (MCBs) and Working Cell Banks (WCBs) strains were grown for 48–72 hours on LB or NB agar solid media (for bacteria) or ME agar (for Candida) at 28–30 °C. After the growth, the colonies were suspended in Nutrient Glycerol. The suspension was homogenized until the bacterial or yeast pellet was completely suspended. 1 mL of the solution was dispensed in cryovials and stored at -80 °C.

The fermentation was performed according to the steps reported in Figure S3. In brief, one vial of the WCB was used to inoculate one agar plate which was incubated at 28 °C for 18 hours. A loop of cells from the agar plate was used to inoculate 100 mL of liquid medium dispensed into a 500 mL baffled flask. The flask was then incubated at 28 °C and 200 rpm till the OD600 (measured with a spectrophotometer Shimadzu UV-160) reached a value of 2.0–2.5 (bacteria) and 10–15 (Candida). 1 to 5% of the grown culture was used to inoculate 100 mL of the production medium dispensed in 500 mL baffled flasks. Incubation of the production flasks was then at 28 °C and 200 rpm for up to 200 hours.

To estimate growth on the production media, the viable cell count (as CFU m L^-1) was performed on serially diluted cultures which were plated on agar medium.

2.2. Analytical Methods

2.2.1. Acid hydrolysis and HPLC Quantification of Rhamnolipids and Sophorolipids (Glycolipids)

Glycolipids concentration was routinely estimated based on sugar equivalents (rhamnose for rhamnolipids or glucose for Sophorolipids) obtained upon acid hydrolysis of the culture sample. Sugars quantification was performed on an Agilent technology 1260 infinity HPLC. An isocratic HPLC method was used for rhamnose analysis. The samples for analysis were prepared as described below.

1 mL of sample from the fermentation broth was dispensed in a 2 mL Eppendorf tube. 160 µL of 37% HCl were added and the tube was shaken (HCl final concentration 5% v v^-1). The sample was incubated on Thermomix for 4 hours at 95 °C (hydrolysis). After hydrolysis, the sample was centrifuged for 5 minutes at 16000 rcf to remove the precipitate. 1000 µL of chloroform (CH₃Cl) was eventually added before centrifugation to remove the oily phase (for media containing oil). 900 µL of the supernatant were transferred in a clean 2 mL Eppendorf tube and 100 µl of 35% v v^-1 perchloric acid (HClO₄) were added. The sample was vortexed and was placed at -20 °C for 10 minutes. 55 µL of 7 M potassium hydroxide (KOH) w v^-1 were added and the tube was shaken. The sample was centrifugated at 16000 rcf for 2 minutes and was filtered on 0.22 µm PES pore membrane. The sample was analysed with the HPLC method reported in Table 2.

The glycolipids concentration was defined starting from the L-rhamnose or D-Glucose value obtained after acid hydrolysis of the fermentation broth. The value obtained was multiplied for the correction factor as reported by Kobayashi and co-workers [41].

2.2.2. LC-MS Analysis of Rhamnolipids

500 mg of crude fermentation broth were extracted with 500 mL of water/methanol (H₂O/MeOH) (50:50). After centrifugation at 16000 rcf, the supernatant was diluted with 10 volumes of (H₂O/MeOH) (50:50). LC-MS analysis was performed using a 1290 Infinity Agilent Instrument according to the analytical method reported in Table 3.

2.2.3. HPLC Analysis of Surfactin

The samples used for the analyses were prepared from the culture broth according to the following steps. The broth was corrected at pH to 2 with 6 N HCl, 1 volume of methanol (MeOH) was added and the sample was stirred for 10 minutes. After stirring, the sample was centrifuged at 16000 rcf for 5 minutes and the supernatant was transferred to HPLC vials. The samples were analyzed by use of a 1260 Agilent HPLC according to the method described in Table 4 [42].

The surfactin concentration was calculated based on a calibration curve obtained from the use of a surfactin standard (Sigma-Aldrich).

2.3. Oil Displacement Test (ODA)

The oil displacement assay is a rapid and effective qualitative method to evaluate the surfactant activity of the molecule under investigation. This assay, developed by Morikawa in 2000 [43], exploits the ability of biosurfactants to create circular zones (halos) in which the oil is displaced once added on top of oil deposited on top of water. The size of the displacement halo is roughly proportional to the activity and to the concentration of the biosurfactant. The assay is performed in Petri dishes with a diameter of 5 cm, in which 30 μL of light crude oil are layered upon 3 mL of ultrapure water. For the test, 3 μL of the biosurfactants containing sample are dropped on top of the crude oil layer. The diameter of the halo obtained after dropping the sample gives a qualitative indication of the surfactant efficacy. The results of this assay are expressed as diameter size of the concentric halo or simply using the (+) or (-) sign to indicate the development or not of halos even if not concentric with respect to the Petri dish.

2.4. Emulsification Index (EI24(%))

The emulsification index (EI₂₄(%)) [44] is a parameter used for determining the emulsifying power of a surfactant molecule. The measurement was performed by mixing equivalent volumes (usually 2 mL), of the solution containing the biosurfactant and of n-hexadecane. The mixture was then stirred with the help of a vortex for 2 minutes and then placed at 25 °C for 24 hours. The emulsification index is calculated as the ratio (expressed as percentage) between the height of the emulsified phase and the total height of the liquid column. Crude oil can be used as an alternative to n-hexadecane.

2.5. Qualitative Analysis of the Biosurfactants by Thin Layer Chromatography (TLC)

Crude extracts of biosurfactants were qualitatively analysed by Thin Layer Chromatography (TLC) on silica gel. TLC for Rhamnolipids was developed with a mobile phase composed of chloroform (CHCl₃), methanol (MeOH) and ultrapure water (H₂O) in a ratio of 65:15:1. The rhamnolipids were then visualized with the orcinol reagent (suitable for detecting the presence of sugars, glycolipids and glycosides). TLC for surfactin was developed with a mobile phase composed of chloroform (CHCl₃), methanol (MeOH) and ammonium hydroxide (NH₄OH 30%) in a ratio of 65:25 + 4%. Surfactin spots were visualized with ultrapure water (suitable for hydrophobic molecules).

2.6. pH Analysis

pH is an essential parameter analysed during fermentation. pH was determined on a sample of culture broth with a Mettler-Toledo pHmeter. The pH value can give hints on the metabolism of the microorganism during fermentation.

2.7. Microscopic and Macroscopic Monitoring

During fermentation, the bacterial culture was monitored both macroscopically and microscopically. Macroscopical analysis was performed to evaluate the growth of the microorganism by turbidimetry and viscosity monitoring, and to evaluate the biosurfactant production by foam observation. Microscopically, the oil emulsification and uptake, possible contamination, biosurfactant production and microorganism growth was monitored.

2.8. Extraction of Rhamnolipids from Fermentation Broths

To extract rhamnolipids from fermentation broth, the protocol from Zhang et al. [45] was used. The protocol uses (NH₄)₂SO₄ to make the water (fermentation broth) and a water miscible solvent (2–propanol) immiscible. More details are reported in the experimental section.

3. Results

3.1. Chemical Composition of the Agro-Wastes

The agro-wastes used in this study were oat and emmer hull, corn chaff, and pea pod hull. The details of the origin of the agro-wastes are reported in Figure S2. These agro-wastes were chosen based on wide availability in Italy, low cost and relative ease of use in a fermentation process (after milling and sewing or without treatment). The composition of the agro-wastes was determined as described in the section materials and methods and is reported in Table S1. The analysis evidenced that the agro-wastes used in this study were rich in nutrients potentially useful for bacterial and fungal growth. The analysis also evidenced a similar composition of the used agro-wastes with analysis reported in the literature on the same material [46,47,48,49].

3.2. Identification of Microorganisms Suitable for Growing on Agricultural Waste

The preliminary tests for a culture medium that allowed the growth of microorganisms, were performed by suspending 100 g of each of the milled agro-waste in 1 L of ultrapure water followed by heat sterilization (123 °C for 20 minutes). Based on the chemical analyses (Table S1), all the agro-wastes considered were rich in nitrogen and carbon sources and also phosphates were present. Therefore, it was considered that a cultivation medium formulated with this type of material alone, could be suitable for microbial growth. The model microorganisms (Table 5), selected among those identified as producers of biosurfactants and available from the company MADEP SA, were then grown on these media. Among the model microorganisms used, all demonstrated the ability to grow on the media formulated as above (colony forming units > 10⁹) (Table 5). Based on pH trend and on the consumption of detectable carbon sources (starch and reducing sugars), we argued that different types of metabolism characterized growth on the different substrates (Figure S4).

In Acinetobacter sp. a growth comparable to that observed in control media can be observed on emmer and oat hull and on corn chaff. Good growth was also observed on pea pod hull (Table 5). As regards the evolution of pH during growth, a similar profile can be observed in the analysis of the pH trend for all the agro-wastes with a slight basification on emmer and oat chaff and on pea pod hull. A remarkable acidification was instead observed on corn chaff (Figure S4(A)). Acidification in the presence of corn chaff could be due to the presence of reducing sugars (3.6 g L^-1 for corn chaff vs. 1.3 g L^-1 for emmer and oat hull,) and starch. Notably, most Acinetobacter sp. are not capable of utilizing glucose as a carbon source [50]. For those strains able to degrade glucose, to gluconolactone/gluconate pathway is used and the reaction can be readily detected by the acidification of medium in the presence of D-glucose[51].

Bacillus subtilis showed the best growth on pea pod hull and emmer and oat hull, while on corn chaff CFU mL^-1 were lower and required longer incubation times (up to 160 hours vs. 50 hours for the other agro-wastes) to reach the stationary phase. As the growth is generally affected by the surrounding environment as the medium composition, the balance of the nutrients and other parameters as pH, it was possible that this slow growth was due to suboptimal pH conditions and to sub-optimal nutrient availability. As regards pH, a progressive increase was observed for the tests carried out on emmer and oat hull and on pea pod hull; on the opposite, a constant pH value around 6 was observed on corn chaff (Figure S4(B)). Poor growth on corn chaff could also be due to the low level of free nitrogen and proteins compared to the other agro-wastes (see Tables S1 to S3) [52].

Candida bombicola was selected for its ability to produce Sophorolipids [53]. Candida bombicola showed good growth on corn chaff as already reported [54]. As regards to the pH, a trend compatible with the physiology of the strain (acidification) can be observed on corn chaff, while on emmer and oat hull and on pea pod hull the culture pH was stable around neutrality (Figure S4(C)).

Pseudomonas aeruginosa was chosen for its ability to produce rhamnolipids as well as for its nutritional versatility. The growth tests carried out on Pseudomonas aeruginosa showed values above control media on corn chaff and pea pod hull, while on emmer and oat hull a lower growth was achieved (Table 5). The determination of the pH showed a similar trend for all three agro-wastes under analysis, reaching basification, up to pH 9, at the end of the exponential growth phase (Figure S4(D)). Worth of note is also the fact that the maximum growth was reached in around 100 hours for P. aeruginosa while for the other strains the lag-phase of growth was usually reached within 50 hours.

Rhodococcus sp. was chosen for its versatile metabolism and its use in numerous soil bioremediation processes [55,56]. Rhodococcus sp. it is also capable of producing biosurfactants. For this strain, it was possible to appreciate good growth on pea pod hull and corn chaff while less abundant growth occurred on emmer and oat hull (Table 5). The analysis of the pH showed a trend towards basification on pea pod hull and on emmer and oat hull, while on corn chaff a constant pH value of 6 was observed (Figure 4(E)).

In conclusion, the agro-wastes used in this study were a suitable growth medium for microorganisms able to produce biosurfactants. The possibility of using agro-waste biomass as a substrate for microbial growth was reported by other authors and relies on the ability to use hemicelluloses and cellulose as carbon and energy source, thanks to the presence of a suitable enzyme array [57].

3.3. Production of Biosurfactants from Microbial Strains Grown on Agro-Wastes

In some of the experimental growth tests described above, the media formulated with agro-waste were also suitable for the production of biosurfactants (Table 5). The oil displacement activity test (ODA) was qualitatively used to ascertain the presence of biosurfactants in microbial cultures. With the ODA test it was possible to identify the production of surfactants in the culture broths of Bacillus subtilis MAD3 (Figure 1) and Pseudomonas aeruginosa MAD10 (Figure 2), while the other strains did not give any ODA activity (both whole fermentation broth and supernatant were tested). Specifically, Bacillus subtilis showed positive ODA results upon growth on all the agro-wastes tested, while Pseudomonas aeruginosa showed positive results only on corn chaff. Negative controls were performed by dropping the supernatant from the abiotic media on plates prepared with water and crude oil. The EI₂₄ (%) test was consistent with the above results (Figure 1 and Figure 2).

In a preliminary characterization, the biosurfactants were solvent extracted and run on thin layer chromatography (TLC) as described in the section Materials and Methods (Figure 3). Figure 3A shows the results obtained on Bacillus subtilis samples, compared with a surfactin standard (Sigma-Aldrich). Figure 3B shows the results obtained on corn chaff from Pseudomonas aeruginosa cultures compared with a rhamnolipid standard (Sigma-Aldrich).

The quantitative determination of surfactin by B. subtilis MAD3 and of rhamnolipids by P. aeruginosa MAD10 was performed by HPLC as described in the section materials and methods and is reported in Table 6.

3.4. Formulation of a Suitable Fermentation Medium Based on Corn Chaff for the Production of Rhamnolipids

As above reported, P. aeruginosa MAD10 was able to produce rhamnolipids when grown on media formulated with corn chaff and water and displayed a fast and abundant growth. In cultures of P. aeruginosa MAD10, the oil displacement activity (ODA) was also tested with samples of broth culture from mixtures of the different agro-wastes. The only mixture that gave a positive ODA result was that with oat and emmer hull / corn chaff, while a negative result was obtained with oat and emmer hull / pea peel, effectively confirming the data obtained with individual agricultural waste and the peculiarity of corn chaff in stimulating the biosurfactant production (Table 6). As corn (Zea mays L.) is one of the most produced cereals in the world and the by-products of corn cultivation are estimated to account globally for 1.64 × 10⁸ tons, we considered this result of particular interest in view of the development of a large scale production process based on this agro-waste. For this reason and in consideration of the high productivity in rhamnolipids obtained (see below), we concentrated our efforts on corn chaff as a fermentation medium nutrient.

The corn chaff is mainly composed of cellulose and hemicellulose (accounting for ca. 75%), lignin (accounting for ca. 20%), starch (0.3%), and has a good content of proteins (2.3%) [58,59]. Hemicellulose sugar composition is mainly represented by arabinose (16.4%), galactose (5.3%), xylose (75.7%) and glucuronic acid (1.9%). From the practical point of view, corn chaff is abundant (21% of the corn waste) [58] and easy to degrade due to its delicate structure (described as bee-wings) and can be used “as is” in the formulation of the fermentation media. In contrast, the other agro-wastes of this study had to be grinded and sewed before entering the fermentation medium. Due to the interesting industrial characteristics of corn chaff and to the ability to sustain rhamnolipids production, we investigated it further by designing an improved fermentation medium. Mixtures composed of corn chaff and standard nutrients used in fermentation were tested. The combinations of nutrients and the production of rhamnolipids are reported in Table 6. Quantification of rhamnolipids was performed as described in the section Materials and Methods.

A titration of the rhamnolipids present in the culture broth at the time of harvest (ca. 144 hours for the industrial medium) was performed by HPLC. Under the conditions tested, the best results were obtained with trial C (below also identified as medium BCS388), and trial D which gave 18 and 16 g L^-1 of rhamnolipids respectively (vs. a maximum of 2 g L^-1 with media formulated with corn chaff and inorganic salts only and a maximum of 2 g L^-1 with the corn chaff in water). The media of those tests indicated that both vegetable oil (both soybean oil and exhausted sunflower oil, a waste cooking oil of considerable interest from a circular economy perspective) and glycerol, when in combination with inorganic salts and corn chaff, are important for the maximization of rhamnolipids production. On the other side, glycerol or soybean oil in combination with inorganic nitrogen and phosphorous are not enough to warrant production of rhamnolipids (trials G and I). The amount of rhamnolipids produced is in the range of the best productivities reported in the literature for batch fermentations and for wild type strains [26], indicating that corn chaff is an excellent substrate for production.

From the qualitative point of view, the mono-rhamnolipids and di-rhamnolipids percentages were determined by LC-MS (Table 4) with mono- and di- rhamnolipids about 5 : 95 in all the tested conditions. This result suggested that the ratio between mono-/di-rhamnolipids is strain specific rather than fermentation medium specific as already reported in the literature [30].

3.5. Evaluation of the Optimal Corn Chaff Concentration and of the Effect of a-Amylase on Rhamnolipids Production

A drawback of the fermentation media formulated with corn chaff was the high viscosity post sterilization which limited the corn chaff concentration in the fermentation medium to 100 g L^-1. Above 100 g L^-1 the medium reached a jelly consistence which was not suitable for liquid phase fermentations. To reduce the viscosity of the broth, and to verify the optimal corn chaff concentration for rhamnolipids production, trials were initially performed with decreasing concentrations of corn chaff. BCS388 medium with 25 g L^-1, 50 g L^-1 and 75 g L^-1 of corn chaff was prepared, and the resulting fermentations were compared with the best medium (BCS388, which was formulated with 100 g L^-1 of corn chaff). After strain revitalization and initial growth in vegetative media, 1% of the grown vegetative culture was inoculated in 500 mL baffled flasks containing 100 mL of medium, as described in the section Materials and Methods. Figure 4 shows that the optimal condition for rhamnolipids production was 100 g L^-1 (higher corn chaff concentrations were not tested due to excessive viscosity). Media formulated with reduced corn chaff displayed a consistent decrease in viscosity (170, 37 and 2.5 cP respectively for 75, 50 and 25 g L^-1 corn chaff). However, a decrease in rhamnolipids production of ca. 23%, 13% and 41% from the maximum productivity achieved in the BCS388 medium, was observed respectively for the 75 g L^-1, 50 g L^-1 and 25 g L^-1 corn chaff concentrations. The viscosity issue could explain the productivity trend with an increase with 50 g L^-1 of corn chaff. The 50 g L^-1 concentration could indeed be a compromise between nutrient availability (including oxygen availability) and viscosity.

Figure 4. Effect of corn chaff concentration on RLs production by P. aeruginosa MAD10. The productivity scale indicates the % of the value obtained with 100 g L^-1 of corn chaff (ca. 18 g L^-1).

Figure 4. Effect of corn chaff concentration on RLs production by P. aeruginosa MAD10. The productivity scale indicates the % of the value obtained with 100 g L^-1 of corn chaff (ca. 18 g L^-1).

As the high viscosity of the broth was considered a challenging issue in view of large-scale fermentations, it was investigated further by treating medium BCS388 with a - amylase. The hypothesis was that starch and/or other polysaccharides with α-linked D-glucose units and released during the sterilization of the medium could contribute to the viscosity. Therefore, to reduce the viscosity of the broth, α-amylase was used under controlled conditions to hydrolyse the polysaccharides. The α-amylase was added to the fermentation broth and, after hydrolysis, was heat inactivated by the sterilization of the medium. Concentrations of α-amylase equivalent to 1.25 U mL^-1, 2.5 U mL^-1 and 5 U mL^-1 were added to Medium BSC388 as described in the section Materials and Methods. After strain revitalization and initial growth in vegetative media, 1% of the vegetative culture was then inoculated in 500 mL flasks containing 100 mL of media treated with different α - amylase concentrations, as described in the section Materials and Methods.

The treatment with α-amylase at all tested concentrations compromised rhamnolipids production. The addition of 1.25 U mL^-1 of α-amylase reduced the viscosity from ca. 2850 cP to ca. 20 cP with a parallel rhamnolipids production decrease by 20%. Increased addition of α-amylase proportionally decreased rhamnolipids production (Figure 5). The inhibition of rhamnolipids production could be due to high glucose concentration originating from the starch and/or other polysaccharides composed of α-linked D-glucose units hydrolysis. Indeed, previous studies performed in our experimental conditions demonstrated (in contrast with literature data where glucose is routinely used for the production of rhamnolipids) that glucose has an inhibitory effect on rhamnolipids production by Pseudomonas aeruginosa MAD10.

3.6. Study of the Fermentation of P. aeruginosa in Medium BCS388

Glycerol is a fundamental nutrient for the production of biosurfactants and specifically of rhamnolipids [60]. Its main function lies in osmoprotection: cell cultures grown with the addition of glycerol show a cellular physiology less subjected to stress when compared to cultures grown in the presence of only sugars as a carbon source [61]. To better evaluate the physiology of production in the newly formulated medium, the consumption of glycerol was monitored through HPLC as described in the section materials and methods. As shown in Figure 6, a correlation can be observed between the onset of rhamnolipid production and the decrease in the glycerol present in the culture medium. The increase in production of the rhamnolipids ends with the exhaustion of glycerol.

3.7. Purification of Rhamnolipids from Medium BCS388 and Identification of the Different Congeners

After the optimization of the fermentation conditions for rhamnolipids production, trials in rhamnolipids extraction and purification were carried out to verify how much the presence of corn chaff could affect the purification of rhamnolipids and if an eco-friendly method of extraction could be applied. At present, rhamnolipids are extracted from the fermentation broth with expensive methods using organic solvents [26,62]. The goal of this trial was to identify an environmental friendly and efficient method to extract rhamnolipids. The extraction/purification system used, was an alcohol-inorganic salt system. 2 - propanol was selected as alcoholic phase and (NH₄)₂SO₄ was selected as salt. The addition of salt facilitates the localization of rhamnolipids in the organic phase. The mechanism involved is related to the salting out effect already described in the literature[45] and the purification scheme is reported in Table 7. Along the different steps of purification, the rhamnolipids concentration was monitored by HPLC as described in the section materials and methods. Table 7 shows the recovery of rhamnolipids in the different steps of the purification.

After the first separation through centrifugation and filter paper filtration of culture broth, 92% of rhamnolipids are separated in the supernatant. The crude rhamnolipids organic extract appeared as a viscous oil with a content in pure rhamnolipids of 28%. Overall, the yield of the process was 63.4% of the initial amount of rhamnolipids in fermentation broth. Although encouraging, more studies will have to be dedicated to this DSP method to improve the yields and the purity which are compromised by the complexity of the corn chaff matrix.

The semi purified sample described above was used for the identification of the different congeners of rhamnolipids. Indeed, rhamnolipids are produced by P. aeruginosa as a group of related molecules (congeners or complex). The analysis was performed by LC-MS (Figure 8) according to the method described in the section Materials and Methods and the different congeners are reported in Table 8. The results showed that the complex composition was like the one which we obtained in different media during our studies with the same strain (data not shown) thus suggesting the neutrality of the developed medium in the synthesis of the different congeners and that the synthesis of the congeners is linked to the strain rather than to the fermentation conditions [30].

4. Discussion

Agricultural waste has long been known as a substrate for the growth of microorganisms and to produce microbial biomass, metabolites, and enzymes [63,64]. While the advantage of using agricultural waste impacts consistently on the costs and is a driving factor in a zero-waste economy, the drawbacks can seriously hamper the success of the approach. In particular, the use of agro-waste in fermentation requires costly pre-treatments, gives low yields in production, and the variations in the seasonal and non-seasonal factors affecting the quality of feedstocks impacts on quality, yield and costs of the product [11].

In this study, we have considered the use of agro-waste, with reference to waste from cereals, for the production of biosurfactants. As similar studies have already been reported and all were hampered with low yields of production, we hypothesized that it was due to the non-correct selection of the producer microorganisms and/or the correct fine tuning of the additional micro and macro-nutrients required for optimal production media [15]. Therefore, we have approached the challenge by initially coupling the agro-wastes with the correct microorganisms. For this purpose, we screened a few microorganisms, known for their ability to produce biosurfactants, for their growth on the different agro-wastes. Surprisingly, the agro-wastes supported microbial growth in all cases and were also able to support the production of biosurfactants. We have indeed identified corn chaff as a suitable substrate for the growth of P. aeruginosa MAD10 strain and for the production of rhamnolipids, and oat and emmer chaff as supporting growth and production of surfactin by Bacillus subtilis MAD3. The fact that P. aeruginosa MAD10 was able to grow on the different substrates but produced rhamnolipids only when the substrate was corn chaff confirmed our initial idea of the necessity to couple the correct microorganism with the agro-waste. On the other side B. subtilis MAD3 was less selective in this sense as it was able to produce surfactin upon growth on the three different substrates with highest yield on oat and emmer chaff.

Besides the yields in biomass and biosurfactants (particularly relevant in P. aeruginosa MAD10), the advantage of our approach was: i) the easy pre-treatment of the agro-waste (pre-treatment is indeed based on milling, which could eventually be performed directly at the source), ii) the possibility of relying on a fermentation medium exclusively based on agro-waste and water, and iii) production of surfactants in a range of interest for the industry (up to 17 g L^-1 in batch fermentations for rhamnolipids production) using fermentation media which had costs far below those of standard industrial media (corn chaff has a cost of 1 € per ton). As a meter of comparison, the production of rhamnolipids from agro-waste (including from corn chaff) was already reported in the literature but occurred with low yield (51.6 mg L^-1) and only when the agro-waste was treated prior to fermentation[65].

As concerning the reason why corn chaff was selectively able to stimulate rhamnolipids production, this remains largely unknown. The treatment of corn chaff with a - amylase compromised rhamnolipids production at all tested concentrations, indicating that free sugars are detrimental for production. On the other side, we were able to increase production in rhamnolipids by only adding to the medium additional carbon sources (soybean oil in example increase production to more than 6 g L^-1 in medium M and N) and this indicated that corn chaff was already an excellent supply for nitrogen, phosphate and other micronutients. Small amounts of nitrates and phosphates together with additional carbon sources pushed productivity towards the maximum levels (medium C and D). Finally, media devoid of corn chaff did not show any production (medium I and G). This suggested that corn chaff had a balanced composition for the stimulation of production of rhamnolipids or that it contained specific inducers which are at present unknown.

5. Conclusions

In addition to adding value to organic leftovers, the production of biosurfactants from agricultural waste encourages sustainable practices and reduces the environmental impact of waste disposal. To successfully produce biosurfactants from these renewable resources, process optimization and sorting of agricultural waste and microorganisms will be the key. The synthesis of biosurfactants could see a considerable drop in price if agro-industrial waste were used efficiently as raw materials for fermentation. In conclusion, agro-waste could be an interesting substrate for the production of biosurfactants and we are investigating it further in P. aeruginosa (by use of different strains which are known to give different rhamnolipids complex compositions) and by applying DOE to the medium improvement. Additionally, i) as the production of biosurfactants is of interest for bioremediation of hydrocarbons, ii) considering that the strains used in our work were also selected for their ability to degrade PAH, and iii) that cereal waste is commonly used as amendant in the bioremediation of soils, our fermentation products are being directly applied to contaminated soils in bioremediation treatments within the frame of the RICREA project. This approach consistently shortens the chain between waste and reuse or recycle.