Screening of phospholipase A activity and its production by new actinomycete strains cultivated by solid-state fermentation

Reagents and media

Gum arabic, bacteriological agar, β-cyclodextrin (β-CD), casein, cresol red, L-α-phosphatidylcholine from egg yolk ≈60% TLC (PC), N-lauroylsarcosine (NLS), butylhydroxytoluene (BHT), oleic acid, rhodamine 6G, sodium taurodeoxycholate (NaTDC), 3-(N-morpholino)propanesulfonic acid (MOPS), ethylenediaminetetraacetic acid (EDTA), Tween 80, porcine pancreatic PLA₂ (ppPLA₂), PLA₁ from Thermomyces lanuginosus (tlPLA₁) and honey bee venom (Apis mellifera) PLA₂ (hbPLA₂) were purchased from Sigma-Aldrich-Fluka (St. Louis, MO, USA). Recombinant human pancreatic lipase (rHPL) was expressed in Pichia pastoris and purified from culture media as described by Belle et al. (2007). Recombinant guinea pig pancreatic lipase-related protein 2 (rGPLRP2) was expressed in Aspergillus oryzae and purified according to Hjorth et al. (1993). Yeast and malt extracts were purchased from BD (Denver, CO, USA). 1,2-α-eleostearoyl-sn-glycero-3-phosphocholine (EEPC) was synthesized as described by El Alaoui et al. (2014). Thin-layer chromatography (TLC) aluminum sheets, coated with 0.2 mm silica gel 60 F₂₅₄, were purchased from Merck. Sugar-cane bagasse support was kindly donated by the sugar refinery San Francisco Ameca S.A. de C.V. (Ameca, Jalisco, México). All other reagents were obtained from local suppliers.

Microbial collections

Microorganisms analyzed in this investigation were isolated from different Mexican environments. Collection 1 was obtained from lake sediments (1.5 m depth) of regions of Lake Chapala located in Michoacán (Bahía de Cojumatlán) and Jalisco (San Luis Soyatlán, Tuxcueca and San Juan Cosalá), states of Mexico where ambient temperature ranging from 10 to 35°C. Samples from Collection 2 were obtained from uncultivated soil of a wet region within the area called Las Tortugas located in Jalisco with temperatures from 16 to 24°C. Microorganisms from Collection 3 were kindly donated by Dr. Ali Asaff Torres (CIAD, Sonora, México) and obtained from soil and dry crusts of the Sonoran desert, a region in the north of Mexico with extreme temperatures ranging from 0 to 50°C. Strains were classified according to macroscopic and morphological characteristics into different actinomycete strains.

Strain conservation

Strains were cultured on yeast malt extract agar medium composed of yeast extract (4 g/L), malt extract (10 g/L), dextrose (4 g/L), bacteriological agar (20 g/L) and pH adjusted to 7.0. The cultures on Petri dishes were incubated for 10 days at 30°C (Collections 1 and 2) or 37°C (Collection 3). Biomass or spores were then gently scraped and placed into a cryovial containing 1 mL sterile glycerol solution 80% before being stored at −80°C.

Screening agar-plate method for PLA activity

Extracellular production of PLA activity was detected using an agar-plate method implemented in this work and based in that reported by Kouker & Jaeger (1987) with modifications. Minimal medium composed of soluble starch (1 g/L), casein (0.3 g/L), KNO₃(2 g/L), NaCl (2 g/L), K₂HPO₄ (2 g/L), MgSO₄⋅7H₂O (0.05 g/L), CaCO₃ (0.02 g/L), FeSO₄⋅7H₂O (0.01 g/L), CaCl₂ (0.5 g/L), bacteriological agar (18 g/L), pH 7.0 was supplemented with 1% (w/v) egg yolk PC for induction and 1 mg/mL rhodamine 6G for detection of PLA activity (Okada, 2000). Substrate was previously dispersed separately with 3% (w/v) gum arabic at 10% (w/v) and rhodamine 6G was previously dissolved in ethanol and sterilized by filtration. Cultures were incubated during 10 days at conditions mentioned above. PC hydrolysis by extracellular PLA activity produced by cultures was detected by the formation of yellow fluorescent halos around actinomycete colonies visible upon UV irradiation.

Molecular identification of microorganisms

Molecular identification was done by amplifying and sequencing a 1.3 kb fragment of the 16S rRNA gene. PCR reactions were carried out with primers 63F and 1387R (Marchesi et al., 1998) on a Veriti™ 96-well thermal cycler (Applied Biosystems, Foster City, USA) using OneTaq Hot Start DNA Polymerase (New England Biolabs). Sequencing of both strands was done at Macrogen USA Corp. Consensus sequences (Supplemental Information) were generated with the CLC Main Workbench 5.5 software package (CLCBio, Aarhus, Denmark) and compared to recorded sequences from GenBank database using the BLASTn algorithm.

PLA activity production by SSF

Sugar-cane bagasse, polyurethane foam (PUF)/perlite (1/3 mixture), and commercial textured soy were used as SSF supports. The SSF impregnation medium for sugar-cane bagasse and PUF/perlite was composed of dextrose (5 g/L), urea (4 g/L), K₂HPO₄ (5 g/L), MgSO₄ (1 g/L), CaCl₂ (0.5 g/L) and PC inductor (40 g/L). The initial pH was adjusted to 7.0 and the medium was then finally dispersed using a mechanical homogenizer. When textured soy was used, the SSF impregnation medium consisted of glycerol (40 g/L). The agricultural residue sugar-cane bagasse and perlite were sieved to provide particles ranging between 0.8 to 1.7 mm and around 0.42 mm in size, respectively. Low density PUF was cut into 0.5 cm³ cubes. The three materials were washed with distilled water and then dried for two days at room temperature. Commercial textured soy received no previous treatment. SSF was performed based on the method described by Raimbault & Alazard (1980). The support was impregnated with a 1.5 times concentrated impregnation medium in the ratio of 1 g per 2 mL, respectively, and sterilized at 121°C for 15 min. Inoculum was added to obtain a final concentration of 2 ×10⁷ spores/gS and the humidity was adjusted to 75%. The inoculum was prepared by growing each strain in 250 mL Erlenmeyer flasks containing 50 mL of yeast malt extract agar medium, at 30°C (Collections 1 and 2) or 37°C (Collection 3) for 10 days. Spores were harvested with 20 mL of a solution of Tween 80 (0.01%, w/v). 125-mL Erlenmeyer flasks were packed with 20% relative volume of support, i.e., 5 g textured soy, 2.5 g sugar-cane bagasse or 0.5 g PUF with 1.5 g perlite. Culture conditions were 30 or 37°C during 10 days. Non-inoculated solid medium was incubated under the same conditions as a control.

PLA activity extraction

Phospholipases produced in SSF cultures were extracted by adding 2.5 mM MOPS buffer (pH 7.2) containing 0.5% NLS, in a proportion of 2 mL per gram of support. The impregnated support was incubated on ice during 1 h and the liquid content was then removed gently, manually squeezed and collected. The liquid fraction was centrifuged at 13,000 rpm during 20 min to eliminate residual solids in order to obtain enzymatic extracts ready to be analyzed and were preserved on ice during treatment.

Protein determination

Protein concentration was determined using the Bradford (1976) microassay procedure with Bio-Rad Dye Reagent. BSA was used as the reference protein and the concentration was expressed in mg/mL or µg/µL. Enzyme amount in a microplate was expressed in µg/well.

Spectrophotometric methods for PLA activity measurements

For all enzyme activity methods applied here, one enzyme unit was defined as the amount of protein releasing 1 µmol of fatty acid per minute under assay conditions. Enzyme activity milliunits (mU) of extracts coming from SSF and determined using PC substrate, were expressed per gram of support (gS) in mU/gS. Activity measurements using EEPC substrate were also expressed in mU/gS or as specific activity as mU per mg of protein of extract in mU/mg.

Extraction of lipolysis products

Lipid extraction was performed immediately after the kinetic measurements from the microplate reaction medium using Folch’s method (Folch, Lees & Stanley, 1957). 20 µL of 0.1 N HCl solution were added to each microplate well to stop reaction and all the content of each microplate well was transferred into a 2 mL glass vial with a screw-cap and 500 µL of chloroform/methanol (2/1) (v/v) were added. After vigorous shaking and phase separation, the lower organic phase was collected using a Pasteur pipette and transferred to a 2 mL tube, where it was dried over anhydrous MgSO₄. MgSO₄ was removed by centrifugation for 1 min at 1,000 g and the clear organic phase was transferred to a 2 mL vial with a screw-cap and kept at −20°C until the analysis was performed.

Thin-layer chromatography of lipolysis products

Residual phospholipids and lipolysis products were analyzed by thin-layer chromatography (TLC) onto 0.2 mm silica gel 60 F₂₅₄plates using a single migration either for samples resulting from the cHTS-PLA assay or for samples resulting from the UV assay. Mobile phase used was composed by chloroform/methanol/acetic acid/NaCl 0.9% (50/25/8/4) (v/v/v/v) and 0.001% (w/v) BHT was added when EEPC was employed. Lipolysis products for samples resulting from the cHTS-PLA assay were revealed by charring the plate after spraying it with a mixture (50:50) (v/v) of saturated solution copper acetate in water and 85.5% phosphoric acid. EEPC lipolysis samples were revealed using UV light at 254 nm.

Screening of actinomycete strains on agar-plate for extracellular PLA activity

About 200 strains from different soil/sediment samples from humid (Collection 1), mild (Collection 2) and extreme (Collection 3) Mexican environments were isolated by using selective microbiological culture media for actinomycetes and their ability to produce extracellular phospholipases was analyzed. Primary screening of strains to detect PLA activity was done using a sensitive agar-plate method containing the organic fluorophore rhodamine 6G and egg yolk PC substrate, as described in Materials and methods. Hydrolysis of PC substrate in plates was detected by a yellow fluorescent halo around strain colonies, visible upon UV irradiation. More than 40% of strains from each collection produced PLA activity: n = 53 in Collection 1 (46% of strains), n = 30 in Collection 2 (70% of strains), and n = 37 in Collection 3 (75% of strains) (Fig. 1A). According to this primary screening, 120 strains were found as important phospholipases producers. To identify the strains with PLA activity belonging to Actinobacteria phylum, 16S rRNA analysis was done. About 42% (n = 22) of strains from Collection 1, 80% (n = 24) of strains from Collection 2 and 22% (n = 8) of strains from Collection 3, were identified as actinomycetes. Genus distribution of actinomycetes for each collection is shown in Fig. 1B. In Collection 1, Streptomyces and Micromonospora were the predominant genera with about 50 and 27% of occurrence, respectively, followed by Actinomadura and Mycobacterium with a prevalence of around 9% each, and Nonomuraea with a lower prevalence of 5%. Results for Collection 2 showed a predominant proportion of the genus Streptomyces with around 79% of occurrence, followed by Amycolaptosis (13%), Nocardia and Saccarothrix (4% each). In Collection 3, 100% of microorganisms belonged to the genus Streptomyces.

During primary screening, different fluorescence intensities were observed in positive strains (Fig. S1), thus allowing an estimation of the relative levels of PLA activity in each collection (Fig. 1C).

Level 1 (+) corresponding to a low PLA activity was principally observed in Collection 3 with about 49%, followed by Collection 1 with about 35%, and finally Collection 2 with about 30% of positive strains. Moderate PLA activity represented by level 2 (+ +) was found mostly in strains from Collection 1 reaching about 48% of total strains, followed by Collection 2 with about 37% and Collection 3 with about 30%. Level 3 (+ +  +) established for high PLA activity was predominant in Collection 2 with about 33% of total strains showing an intense yellow fluorescent halo, followed by Collection 3 with about 21% and to a lesser extent, Collection 1 with 17%. Actinomycete strains that presented qualitative high PLA activity level (+ +  +) were kept for further enzyme production studies.

Development of the colorimetric HTS assay for general PLA activity detection (cHTS-PLA)

With the aim to have a sensitive and robust method to easily quantify PLA activity and to perform the screening of a large number of samples from actinomycete strains, a continuous spectrophotometric assay referred as the cHTS-PLA assay, using microtiter plates, PC as substrate and cresol red as pH indicator was implemented based on described methods (Mateos-Diaz et al., 2012; Camacho-Ruiz et al., 2015) with some modifications. The solubility of PC (up to 100 mg/mL) was tested in several alcohols such as ethanol, 2-methyl-2-butanol, 2-methyl-2-propanol, and 2-propanol, this latter alcohol being the most suitable (Fig. S2).

To validate the assay specificity, various purified or commercial phospholipases with sn-1 (PLA₁) or sn-2 (PLA₂) stereoselectivity toward phospholipids were used. tlPLA₁ (Yang et al., 2006) and rGPLRP2 (Carriere et al., 1994; Hjorth et al., 1993) are known to display a stereopreference for the sn-1 position of phospholipids. hbPLA₂ and ppPLA₂ are known to be sn-2 stereoselective phospholipases (Dudler et al., 1992; Okimasu, Sasaki & Utsumi, 1984). In order to prove the specificity to detect phospholipase activity, a lipase such as the rHPL was chosen as a negative control to validate the assay. The rHPL has been reported to possess no lipolytic activity against phospholipids such as phosphatidylcholine, phosphatidylethanolamine and phosphatidylglycerol (Abousalham & Verger, 2000). Figure 2 shows the typical kinetics obtained when high amount of rHPL (21.5 µg) and low amount of rGPLRP2 (0.115 µg) were used. Significant changes in the OD decrease (ΔOD/min) were observed for rGPLRP2 in comparison with blank reaction, while with rHPL, changes in the OD with time were similar to those observed in the blank reaction and maybe attributed to a slow spontaneous hydrolysis of PC at the assay pH (8.2). The cHTS-PLA assay therefore appeared to be sensitive enough to detect PLA activity.

The effects of the enzyme amount on the steady-state reaction rate are shown in Fig. 3. A linear relationship was observed up to a maximum protein amount tested of 0.5 µg/well for tlPLA₁ (R² = 0.991, Fig. 3A), 23 µg/well for rGPLRP2 (R² = 0.992, Fig. 3B), 28.8 µg/well for ppPLA₂ (R² = 0.995, Fig. 3C) and 13.4 µg/well for hbPLA₂ (R² = 0.997, Fig. 3D).

Production of PLA activity by solid-state fermentation (SSF)

PLA production with actinomycete strains having high PLA activity (Level +++ in primary screening) was first attempted by submerged fermentation (SmF) using PC inductor. PLA activity from strains analyzed was however hardly detectable (Fig. S3). The actinomycete strain A119 (close to Streptomyces djakartensis) was then cultivated by SSF using different supports such as textured soy, sugar-cane bagasse and PUF/perlite impregnated with PC inductor. The strain A119 was able to grow in all tested supports by SSF and the PLA activity of the extracts obtained from fermented solids was measured using the developed cHTS-PLA assay with PC substrate and expressed as milliunits (mU) per gram of support (gS). The PLA activity of A119 was detected with all tested supports (Table 1); around 70 mU/gS were detected with textured soy, while with PUF/perlite, around 128 mU/gS were reached. The maximal activity value of around 240 mU/gS was obtained with sugar-cane bagasse, being about 1.8 and 3.4 times higher in comparison with PUF/perlite and textured soy, respectively. These results demonstrate for the first time, a higher production of extracellular PLA activity by actinomycete strains cultivated by SSF. Despite the fact of having done a liquid extraction from fermented solids, the PLA activity from crude enzymatic extracts obtained from these cultures was easily detectable using the implemented cHTS-PLA assay.

Subsequently, once having selected the more convenient support, PLA production by selected actinomycete strains from the primary screening (level +++) was performed by SSF. The PLA activity analysis of crude extracts from SSF was made after 10 days of growth for every isolate and results are shown in Fig. 4. PLA activity of strains from Collection 1 ranging from 207 ± 4 mU/gS for 5C-11 strain to 2,591 ± 49 mU/gS for 2F-1 strain. Besides 2F-1 strain, 6C-3, 1F-7 and 2F-2 strains were also high PLA producers reaching 1,929 ± 1, 1,533 ± 49 and 1,300 ± 250 mU/gS, respectively. In Collection 2, the highest values of PLA activity were obtained for AC104-18 strain with 2,217 ± 32 mU/gS and AC3-9 strain with 440 ± 7 mU/gS; conversely, ACS102-12 strain had the lowest PLA activity value with 259 ± 11 mU/gS. In Collection 3, strains with major PLA activity were A99 with 2,495 ± 234 mU/gS, followed by A3 strain with 497 ± 46 mU/gS, and A119 with the lower value (240 ± 4 mU/gS).

The phylogenetic tree of actinomycete strains producing PLA activity by SSF was obtained by neighbor joining analysis based on their 16S ribosomal RNA gene sequences (Fig. 5). There is no report of PLA activity for the actinomycete strains presented in the phylogenetic tree, which increases the chance to find novel interesting enzymes for potential applications.

Analysis of PLA activity of selected strains

A TLC analysis of PC hydrolysis from the cHTS-PLA assay by phospholipases of actinomycete strains produced by SSF with sugar-cane bagasse was performed. Results showed the production of lyso-PC by the action of phospholipases A from actinomycete strains of all collections, where non accumulation or different level of accumulation of lyso-PC was observed (Fig. S4). Figure 6 displays the results obtained with 3 selected strains from Collection 1 (2F-1, close to Streptomyces champavatii/sampsonii; 2F-2, close to Streptomyces caouvurensis; and 6C-3 close to Streptomyces olivaceus/pactum) that stood out for the high production of lyso-PC, and thus evidence for high PLA₁ or PLA₂ activities in the corresponding cultures. PC hydrolysis by ppPLA₂ was used as a positive control. PLA activity of crude extracts from 2F-1, 2F-2 and 6C-3 strains was also tested using the cHTS-PLA assay and various amounts of enzymatic extracts (Fig. 7). Typical kinetics of the decrease in the OD at 580 nm during the PC hydrolysis are shown in Figs. 7A, 7C and 7E. Variations in the steady-state reaction rate (ΔOD/min) with increasing amounts of extracts were found to be linear with time. This linear relationship was observed up to maximum protein amounts of 39 µg for 6C-3 strain (Fig. 7B, R² = 0.988), 56 µg for 2F-1 strain (Fig. 7D, R² = 0.985) and 42 µg for 2F-2 strain (Fig. 7F, R² = 0.985) which was the maximum injected volume (20 µL) for each sample.

To confirm PLA production of these three actinomycete strains, a sensitive and continuous UV assay for PLA activity based on the use of the specific substrate EEPC esterified at both sn-1 and sn-2 positions with α-eleostearic acid (9Z, 11E, 13E octadecatrienoic acid) (El Alaoui et al., 2014) was applied. Once the EEPC substrate is coated onto a microplate well, it is stable in the presence of reaction buffer without enzyme (no increase in OD at 272 nm). When a phospholipase is injected in a well, it binds to the surface coated with EEPC and can hydrolyze it. Thanks to the presence of β-cyclodextrin, the α-eleostearic acid released from EEPC is solubilized in the aqueous phase. This leads to an increase in the OD at 272 nm that is specific to α-eleostearic acid and allows estimating PLA activity (El Alaoui et al., 2014; El Alaoui et al., 2016; Mendoza et al., 2012; Serveau-Avesque et al., 2013).

The PLA activities on EEPC of extracts from 2F-1, 2F-2, and 6C-3 strains were measured using 20 µL of crude enzymatic extract (Fig. 8) and expressed as milliunits (mU) per gram of support (gS). Under these conditions, the activities of 6C-3, 2F-1 and 2F-2 were estimated to be 535, 52 and 193 mU/gS, respectively. An initial absorbance (Fig. 8B) was probably due to some detached substrate molecules that disturbed in some degree the UV measurement after sample injection into the microwell, until they finally were dispersed in the surface. This was more evident when the activity of sample was lower, and was also always observed at the beginning of blank measurements.

The reaction mixtures were recovered from the microtiter plates for lipid extraction and TLC analysis to confirm that the synthetic EEPC substrate coated in microplates was hydrolyzed by PLA present in crude enzymatic extracts from actinomycete strains (Fig. S5). The UV assay with EEPC substrate can therefore be used to confirm and measure PLA activity in crude microbial extracts.

Screening of actinomycete strains on agar-plate for extracellular PLA activity

Soil samples were collected from different Mexican environments to isolate actinomycetes with the final purpose to determine their ability to produce PLA activity. The detection of phospholipases in three microbial collections was successfully achieved using a sensitive method with agar-plates containing rhodamine 6G and PC as a substrate. This method implemented in this work was effective and reliable to easily detect PC hydrolysis by extracellular phospholipases. Interestingly, in this primary screening, around 50% of strains from each collection presented this activity. Although previous works have shown determination of extracellular phospholipases using agar-plates containing either soybean lecithin or PC from fresh frozen egg yolk (Kim & Rhee, 1994; Song, Kim & Rhee, 1999), none of them included the addition of PC hydrolysis indicator. The organic fluorophore rhodamine 6G (Okada, 2000) in the agar-plate assay was added and developed here, inspired from the use of rhodamine B, another dye of the same nature, for the detection of extracellular lipase activity on agar-plates (Kouker & Jaeger, 1987). By means of the agar-plate assay containing rhodamine 6G, it was demonstrate that extracellular phospholipase activity can be detected by the PC hydrolysis producing a fluorescent halo surrounding colonies (Fig. S1) allowing the classification of strains into the three categories with low, moderate and high PLA activity (Fig. 1C). The mechanism by which rhodamine acts as an indicator of hydrolysis upon complexation with free fatty acids is not fully understood. A conceivable mechanism may be the generation of excited dimers of rhodamine which fluoresce at longer wavelengths than the excited monomer (excimer fluorescence) (Kouker & Jaeger, 1987).

Development of the cHTS-PLA assay

With the aim to search PLA activity directly from microbial enzymatic extracts, a continuous assay was developed using microtiter plates, with commercially available enriched PC (≈60%, TLC) as substrate and cresol red as pH indicator. PLA activity measurement was based on the decrease of pH indicator optical density due to protonation caused by the release of free fatty acids during the hydrolysis of substrate and thus acidification of the reaction media (Camacho-Ruiz et al., 2015). A lipase (rHPL) and enzymes with PLA activity were used to validate the method. As expected, rHPL was not able to hydrolyze the PC (Fig. 2), while in the case of enzymes with PLA activity, the rate of lipolysis was found to be linear with time and proportional to the amount of enzyme unto the maximum amount added in each case (Fig. 3). Since each of these enzymes has a different specific activity toward the PC substrate, the range of enzyme amount to have a proportional velocity was different between them. A similar HTS assay has been reported by Mateos-Diaz et al. (2012) to determine esterase/lipase activity using tributyrin or trioctanoin substrate and p-nitrophenol as pH indicator at pH 7.2. They applied it to determine lipase activity in crude extracts from fungal strains Rhizopus sp.8a and Aspergillus ochraceus. Although previous PLA activity assays using phospholipid substrate combined or not with chromogenic molecules have been proposed (Lobo de Araujo & Radvanyi, 1987; Petrovic et al., 2001; Price, 2007), they were not designed to be used with crude microbial extracts and some of them were not compatible with HTS. Hence, it was showed a clear evidence that the continuous cHTS-PLA assay using PC substrate and cresol red at pH 8.2 developed in this work is a convenient tool for high-throughput screening.

PLA activity in actinomycetes

Analysis of 16S rRNA sequences provided enough information to identify actinomycete strains up to the genus level. Those from the genus Streptomyces were found as notable sources of phospholipase activity. Our findings (Figs. 4 and 5) are in agreement with some literature reports indicating the presence of bacterial phospholipase activity in several Streptomyces species based on multiple sequence alignment (Yang & Roberts, 2002; McMahon et al., 2012). In particular, a secreted PLA₂ produced by Streptomyces violaceoruber A-2688 isolated from a soil sample, was described as the first PLA₂ identified in prokaryote (Sugiyama et al., 2002; Takemori et al., 2012). Similarly, a PLA₁ and a phospholipase B have also been reported in the Streptomyces genus (Sugimori, Kano & Matsumoto, 2012; Matsumoto et al., 2013). Some authors have also reported the presence of lipases in actinomycete strains (Sztajer, Maliszewska & Wieczorek, 1988) while others did not observe the production of lipolytic enzymes (Basha et al., 2009).

Production of PLA by solid-state fermentation

PLA production was initially done using submerged fermentation cultures, however, low production yields were obtained. To deal with this problem, the SSF was proposed as an alternative since several studies have stressed the advantages of SSF over SmF, attributed to the higher biomass growth rates and the lower enzyme proteolysis showing that SSF gives higher enzyme yields (Aguilar et al., 2001; Diaz-Godinez et al., 2001; Romero-Gomez, Augur & Viniegra-Gonzalez, 2000). In order to achieve higher PLA production yields, SSF was performed using three different supports and PC inductor contained into the impregnation media. The A119 strain (close to Streptomyces djakartensis) cultured on three different supports produced PLA activity in all of them. SSF cultures on sugar-cane bagasse support showed the highest yield reaching 240 mU/gS compared with 128 mU/gS and 70 mU/gS produced on PUF/perlite and textured soy supports, respectively (Table 1). Based on these findings, sugar-cane bagasse was selected as a convenient support for actinomycete SSF cultures for PLA production. Our results are in line with publications proposing SSF as a potent tool for enzyme production and an alternative to SmF (Raimbault & Alazard, 1980; Aguilar et al., 2001; Hansen et al., 2015; Pandey et al., 1999; Thomas, Larroche & Pandey, 2013; Viniegra-Gonzalez et al., 2003). The direct comparison of SSF and SmF using various parameters such as growth rate, productivity or volume activity for the production of enzymes is in favor of SSF in most cases (Holker, Hofer & Lenz, 2004). It has been reported that SSF system achieved higher productivity for some enzymes mainly because of three factors: higher levels of biomass production, apparent resistance to catabolite repression and reduced breakdown of enzymes by proteases (Viniegra-Gonzalez et al., 2003). The higher production of enzymes in SSF may be due to the fact that it takes place in the absence or near absence of free water, thus being close to the natural environment to which microorganisms are adapted (Pandey, Soccol & Mitchell, 2000). Moreover, many reports of SSF have been recently published in which emphasis is given to the application of agricultural by-products for the production of fine chemicals and lipolytic enzymes (Treichel et al., 2010). Utilization of agro-industrial wastes provides alternative substrates and may help solving pollution problems, which otherwise might be caused by their disposal (Treichel et al., 2010). Particularly, the use of sugar-cane bagasse has been proposed to produce interesting molecules (Lakshminarayana et al., 1975) and was also applied for lipase production (Dutra et al., 2008; Mateos-Diaz et al., 2006; Rodriguez et al., 2006). Phospholipase production by actinomycetes using SSF using sugar-cane bagasse is however reported here for the first time (Fig. 4).

Production of PLA activity by bacteria have been only reported using SmF and different protocols for enzyme quantification have been used. In that regard, a native membrane-bound PLA₁ purified from Escherichia coli (Scandella & Kornberg, 1971) showed 0.0084 U/mL of phospholipase activity in the crude extract determined using the radioactive substrate [³²P]phosphatidylglycerol. A native PLA₁ from Serratia sp. (Kim & Rhee, 1996) produced around 1.2 U/mL of phospholipase activity determined using the pH titration method at 50 °C with egg yolk lecithin as a substrate. Similarly, a native PLA₁ from Streptomyces albidoflavus produced 1.5 U/mL of phospholipase activity in the culture supernatant (Sugimori, Kano & Matsumoto, 2012) determined using a fluorometric method with a BODIPY(R) fluorescent dye-labeled substrate. In other hand, a microbial PLA₁ was isolated from a mutant strain of the ciliated protozoan Tetrahymena thermophile and had 0.068 U/mL of phospholipase activity in the crude extract (Guberman et al., 1999) determined with the radioactive substrate 1-palmitoyl-2-[1-¹⁴C]linoleoylphosphatidylcholine. Despite the differences in activity quantification, comparison of these values with the PLA activities obtained here with the cHTS-PLA assay using PC as a substrate (2.59 U/gS (3.45 U/mL) for 2F-1, 1.93 U/gS (2.57 U/mL) for 6C-3 and 1.53 U/gS (2.04 U/mL) for 2F-2 strain), clearly demonstrates that their production by SSF with sugar-cane bagasse at the conditions employed here, appears to improve profiles for native microbial PLA production.

By means of the TLC analysis of PC hydrolysis from the cHTS-PLA assay, it was demonstrate the presence of PLAs in the crude extracts of actinomycete strains produced by SSF with sugar-cane bagasse as support. Moreover, it was shown that the synthetic and specific EEPC substrate coated into microtiter plates can be used to confirm the presence of PLA in crude microbial extracts (Fig. 8). As proposed previously (Mendoza et al., 2012; Pencreac’h et al., 2002; Serveau-Avesque et al., 2013), synthetic substrates containing α-eleostearic acid appears to be stable, sensitive and convenient tools for measuring the activity and specificity of lipolytic enzymes using UV detection of the released fatty acids, either on TLC plates or microtiter plates. Finally, the value of activity detected with EEPC assay was smaller and different to activity detected with the cHTS-PLA assay for the three selected microbial extracts. It was found that the PLA activity detected with cHTS-PLA assay on PC substrate was about 3.6, 49 and 8 times more compared with the PLA activity on EEPC for 6C-3, 2F-1 and 2F-2 strains extracts, respectively. These kinds of differences has been observed for example in the PLA activity determination of rGPLRP2 on C8-PC (1,2-dioctanoyl phosphatidylcholine) and egg PC with values of 551 ± 50 and 812 ± 100 U/mg, respectively, using the pH-stat technique in which the conditions assay consisted of a mechanically stirred emulsion of substrate previously emulsified by ultrasonic treatment (Dridi et al., 2013). In other hand, the PLA activity of rGPLRP2 on coated EOPC (1-α-eleostearoyl-2-octadecyl-rac-glycero-3-phosphocholine) was reported with a value of 0.3 ± 0.02 U/mg (El Alaoui et al., 2016), while the PLA activity of rGPLRP2 with the cHTS-PLA assay developed in this work was of 25.2 ± 0.9, about 84 times more compared with EOPC, and 32 times less compared to pH-stat technique using the same substrate.

These differences can be attributed to the specificity of each enzyme for each kind of substrate, the presentation of substrate (e.g., coated monocapes for EEPC and an emulsion for PC substrate) or the conditions of the assay (e.g., mechanical stirring, orbital shaking).