The medium components were obtained from HiMedia Laboratories (Mumbai, India) and solvents used in this study were all of reagent grade. Fatty acid standard and Sudan black B were obtained from Sigma-Aldrich, Bangalore, India.

Cunnighamella elegans CFR C07 isolated from the Western Ghats of Karnataka and Nilgiris region, India, is used in the study. The strain was stored on potato dextrose agar (PDA) plates (HiMedia Laboratories) at 4 °C. Culture was maintained by repeated subculturing.

The culture medium designed by Somashekar et al. (12) contained (in g/L): KNO₃ 1.00, KH₂PO₄ 2.50, ZnSO₄·7H₂O 0.01, CuSO₄·5H₂O 0.002, MnSO₄ 0.01, MgSO₄·7H₂O 0.50, FeSO₄·7H₂O 0.02, yeast extract 5.00, glucose 30.0 and CaCl₂ 0.10, pH=5.5. Seed culture was prepared in potato dextrose broth (PDB) using spore solution (10⁴ spores per mL). A volume of 20 mL of the spore solution was added to 80 mL of PDB and incubated for 2 days at 30 °C and 200 rpm. A volume fraction of 20% of the inoculum was added to the medium and the culture was grown in Erlenmeyer flasks (250 mL) incubated with shaking at 200 rpm for 5 days at (28±2) °C.

Four inorganic carbon sources (30 g/L each), namely maltose, galactose, sucrose and fructose were evaluated in the fermentation medium and the production of biomass from different carbon sources was determined on dry mass basis by separating the biomass by centrifugation at 6000×g (model C-24BL; REMI, Maharashtra, India) and drying at 60 °C in a hot air oven (model 101; RRT NC, Trivandrum, Kerala, India).

The fungal strains were stained with Sudan black B according to the method of Burdon (13) and du Preez et al. (14) in order to check the oil accumulation. The fungal biomass was separated after fermentation by centrifugation at 6000×g (model C-24BL; REMI), suspended in 1 mL of Sudan black B staining solution and washed with distilled water. A smear of culture solution was made on a clear glass slide. The cells were then washed with 70% alcohol 3-4 times to remove excess stain and observed under optical microscope with oil immersion objective (model DM 2000; Leica, Wetzlar, Germany).

In order to study the effect of incubation period on GLA production, fungal culture was fermented for varying incubation time (3 to 5 days). Biomass was collected, oil was extracted and GLA was estimated by gas chromatography (model GC1000; Chemito Instruments Pvt. Ltd., Nashik, India) after converting it into methyl esters.

The previously described medium (12) was used for studies in the fermentor. The medium was transferred into 3-litre submerged bioreactor (Infors HT, Bottmingen, Switzerland) and sterilized. The pH was maintained in the fermentor using 1 M HCl and 1 M NaOH and frothing was controlled using antifoam. Pre-inoculum was prepared in PDB. Fermentation was continued for 4 days and the fermentor conditions were: total volume 1.5 L, temperature 28 °C, pH=5.5, aeration 1 vvm, shaking speed of 200 rpm and inoculum size 20%. GLA production was evaluated by GC as described previously.

For dry cell mass determination, cell samples in the liquid medium were harvested by centrifugation at 6000×g (model C-24BL; REMI) for 15 min. The obtained cell pellet was then kept for freeze drying in a lyophilozer (ScanVac; Labogene, Allerød, Denmark).

Biomass obtained after 4 days of incubation was subjected to four different solvent extraction methods (11). In method 1, biomass and acid washed sand (1:2, by mass) were homogenized for 1 h. Lipids were extracted by chloroform/methanol (2:1, by volume), centrifuged at 6000×g (model C-24BL; REMI) to get a clear supernatant, and anhydrous sodium sulphate was added to remove any residual moisture. Solvent was removed by flushing with nitrogen. In method 2, biomass was lyophilised and extracted three times by chloroform/methanol (2:1, by volume) using automatic solvent extraction system B-811 (BÜCHI Labortechnik AG, Flawil, Switzerland).

Method 3 was Soxhlet extraction (11). Lipids were extracted with 250 mL of hexane for 8 h in automatic solvent extraction system B-811 (BÜCHI Labortechnik AG). In method 4, biomass and glass beads (1:2, by mass) were homogenised for 1 h and lipids were extracted by chlorofom/methanol (2:1, by volume). The extracted lipids were centrifuged at 6000×g (model C-24BL; REMI) to give a clear supernatant, and anhydrous sodium sulphate was added to remove the residual moisture. Solvent was removed by flushing with nitrogen.

Thin layer chromatography (TLC) of lipids was done after the extraction and the bands were evaluated. The samples were spotted on TLC plates coated with silica gel (HiMedia), which were run in a solvent system containing n-hexane and ethyl acetate in a volume ratio of 9:1. After air drying the plates, the fractions were observed in an iodine chamber containing silica gel. For identification of triacylglycerol (TG) and free fatty acid (FFA), corresponding standards (Sigma-Aldrich) were used.

Before the estimation of GLA by gas chromatography (GC), the extracted lipids were converted to methyl esters. The extracted lipids (500 µL) were mixed with 2% H₂SO₄ and 10 mL of methanol and refluxed for 5 h (at approx. 60 °C) in rotavapor (model R-210; BÜCHI Labortechnik AG). After 5 h, the methanol was evaporated and 10 mL of ethyl acetate were added and transferred into separating funnel. Distilled water (10 mL) was added to the separating funnel and it was shaken gently. The aqueous layer was separated from the organic layer and washing was repeated twice. The ethyl acetate layer was separated and dried over sodium sulphate. It was then filtered, evaporated and resuspended in ethyl acetate/hexane (1:1). Then these samples were analysed by TLC to confirm the presence of free fatty acid methyl esters and then used for GC analysis.

The samples were analyzed using gas chromatograph equipped with flame ionization detector (model GC1000; Chemito Instruments Pvt. Ltd) and Stabilwax capillary column (Restek Corporation, Bellefonte, PA, USA). The analysis started at 40 °C for 1 min and the temperature was increased to 240 °C at the rate of 40 °C/min. After reaching 240 °C, the temperature was held stable for 5 min before the analysis was terminated. The column temperature was 140 °C and the injection volume was 1 μL with split injection ratio of 1:100. The injector temperature was 240 °C and the carrier gas was nitrogen at rate of approx. 5 mL/min. The peak area percentages were recorded. Individual fatty acid methyl esters were identified by comparing their retention times with fatty acid standards.

Glucose was found to be a good carbon source for the production of biomass with a yield of (20.5±1.2) g/L, whereas galactose and maltose were found to be poor substrates with yields of 9.7 and 10 g/L, respectively. Fructose and sucrose gave almost similar yields (19.8 and 20 g/L, respectively) of biomass as that of glucose. Studies by Shrivastava et al. (7) and Ahmed et al. (9) reported glucose as the major carbon source for the production of GLA. Previous studies showed an increased yield of GLA after the addition of glucose (15). After staining with Sudan black B, the presence of blue or greyish oil globules was observed within the mycelium under microscope (Fig. 1). Medium with glucose gave oilier mycelia than other carbon sources. Glucose was selected as the carbon source in the production medium to increase oil accumulation and 
biomass yield.

Optimum incubation period for Cunninghamella elegans CFR C07 was found to be 4 days, with GLA production of (882±10.8) mg/L from (20.5±1.2) g/L of biomass. In incubation period longer than 4 days, GLA production seemed to decrease even though biomass increased (Fig. 2). The peak at retention time of 21.8 min indicates the GLA. The GLA production decreased to 866 mg/L on the 5th day of incubation. Earlier studies on Cunninghamella sp. also reported optimum incubation period of 4-5 days (4, 16) and other fungal strains like Mucor sp. take 7 days (9).

Biomass obtained after 4 days of incubation was subjected to different solvent extraction methods. Among them, hexane used as a solvent for 8 h in automatic extraction system gave the best oil yield, so this method was selected for further studies. Microbial oils are considered to be used for human consumption, so the solvent used in the extraction must be acceptable in terms of toxicity, handling, safety and cost (8). Therefore, hexane was selected, as being less toxic than chloroform/methanol mixture (11). The presence of GLA in the extracted oil was confirmed by TLC in which a clear band of GLA was observed against standard (Fig. 3).

In flask cultivation, 20.5 g/L of biomass were obtained after 4 days and glucose was used as the carbon source. The results obtained here were better than those obtained by Manoh and Seto (4) with C. elegans NRRL 1378 and the medium containing more glucose, malt extract, yeast extract and peptone. In the study done by Murad et al. (17), 12 g/L of biomass were obtained using C. baineri 2A1. In another study by Shrivastava et al. (7) on Cunninghamella echinulata var. elegans MTCC552, 10.90 g/L of biomass were produced. In a study by Gema et al. (18) on Cunninghamella echinulata, 8.9 g/L of biomass were obtained.

In the present study, 882 mg/L of GLA were obtained using the strain after 4 days of incubation, which was higher than the yields reported by Murad et al. (17), Shrivastava et al. (7) and Gema et al. (18), i.e. 650, 209.5 and 720 mg/L, respectively.

Production in a 3-litre fermentor was evaluated and maximum GLA production of 733 mg/L was observed. Fermentation studies by Saad et al. (16) resulted in 747.72 mg/L of GLA with the optimization of aeration and agitation. The optimization of physical parameters such as aeration and agitation in the fermentor used in this study may give even better results, but it can still be concluded that Cunninghamella elegans CFR C07 can be used for the production of GLA and that it could be a potent organism for further scale-up studies and commercialization.