Introduction

In recent years, interest in microalgal fatty acids has increased as a potential new source for bioactive and high value lipid-based chemical precursors (Barbosa and Wijffels 2013; Coates et al. 2013), including polyunsaturated fatty acids (PUFAs) which have applications in the pharmaceutical, nutraceutical, and food industries (Tseng 2001; Tan 2007; Hu et al. 2008). Microalgae exhibit competitive advantages to terrestrial crops and fish, as sources of PUFAs and in general display a higher concentration of 16:0 over 16:l fatty acids, and abundant C18 fatty acids (Mendoza Guzmán et al. 2011; Seyfabadi et al. 2011; Li et al. 2011; Guedes et al. 2011; Ngangkham et al. 2012), as well as of n-3 (ALA) and n-6 isomers. Two of the most abundant fatty acids in Chlorophyta strains are linoleic acid (18:2n6) and octadecatrienoic acid (18:3n3) (Mendoza Guzmán et al. 2011; Li et al. 2011), which are considered as essential for human development, since humans are able to synthesize long chain PUFAs from them (Hu et al. 2008; Rubio-Rodríguez et al. 2010). However, one of the major commercial drawbacks in cultivation of microalgae is their poor biomass growth rate, which consequently leads to low PUFA productivities, even though PUFA may account for as much as 10–20 % of their cell weight (Guedes et al. 2011).

To overcome these disadvantages, culture condition variation to increase lipid productivity (Rodolfi et al. 2009; Gouveia et al. 2009; Li et al. 2011; Cheirsilp and Torpee 2012; Li et al. 2013) and/or to modify fatty acid composition (Renaud et al. 1991; Dunstan et al. 1993; Otero et al. 1997; Guihéneuf et al. 2010; Li et al. 2013; Huang et al. 2014) have been under scrutiny.

The mode of cultivation could be decisive for the quantity as well the quality of fatty acids, since it significantly influences the growth and metabolic pathways in microalgae (Wang et al. 2014). Autotrophic cultivation is the most common procedure for microalgae culture. However, in autotrophic culture it is difficult to reach a high density of microalgae biomass since light penetration is inversely correlated to the cell concentration (Chen and Johns 1995; Liang et al. 2009), which leads to a very low microalgal lipid productivity resulting from low biomass productivity (Martinez and Orus 1991; Liang et al. 2009).

An alternative to photoautotrophy is the use of heterotrophic culture in which organic carbon is used as carbon source in the absence of light (Borowitzka 1999; Perez-Garcia et al. 2011; Cheirsilp and Torpee 2012). Heterotrophic cultivation could avoid the defects associated with photolimitation in photoautotrophy, thus, providing high biomass productivities (Miao and Wu 2006; Liu et al. 2011). Lipid content under heterotrophic cultivation is generally at par with or higher than that under photoautotrophic mode (Miao and Wu 2006; Xu et al. 2006), resulting in even higher lipid productivity.

Recently, mixotrophic culture regime has been adopted as an effective mode for producing microalgal lipids. In mixotophic cultivation, CO2 and organic carbon are simultaneously assimilated and both respiratory and photosynthetic metabolism operate concurrently (Lee 2004).

Complementing photoautotrophy with organic substrates, mixotrophic cultivation of microalgae can reduce biomass loss in dark hours due to respiration (Rym et al. 2010; Park et al. 2012). Moreover, the CO2 released by microalgae via aerobic respiration can be trapped and reused for photosynthesis under mixotrophic cultivation, which further enhances biomass, and then lipid, productivities (Mohan et al. 2014). Furthermore, mixotrophic cultivation might switch to photoheterotrophic growth, where organic carbon is the main source of carbon and where the energy is provided mainly by the photosynthetic apparatus (Wilken et al. 2014). This combination of photoautotrophic and organoheterotrophic growth should result in a highly efficient carbon conversion and leads to an even higher biomass and lipid productivity.

The production of microalgal lipids under mixotrophic cultivation has not been fully investigated, especially for thermophilic species. To our knowledge, no one has examined the effect of trophic mode and growth phases on the lipid productivity and fatty acid composition of species related to the genus Graesiella. According to Richmond (1986), organisms with high optimal growth temperature possess wider possibilities for adaptation, as they have a rapid growth rate and a flexible metabolism to deal with various environmental conditions.

This study was aimed at evaluating the effect of trophic modes (autotrophic, mixotrophic, and heterotrophic) on growth, lipid production, and fatty acid variations in a strain of Graesiella sp., isolated from Tunisian hot spring water. In this study, we evaluated at first the kinetics of lipid production, cell growth, photosynthesis activity, and glucose consumption in autotrophic, mixotrophic, and heterotrophic cultures. Secondly, we examined the effects of growth phase and trophic modes on fatty acid variation.

Materials and methods

Algal strain

Samples were taken from “Aïn Echfa”, a hot spring (60 °C) located in the northern part of Tunisia (36° 49′ N, 10° 34′ E). Sampling materials were composed of microbial mats anchored to submerged stones. Collected mats were treated by filtration, centrifugation, and dilution techniques according to standard microbiological protocols (Stanier et al. 1971; Rippka et al. 1979). The purified strain was identified by phylogenetic analysis as Graesiella sp. (Mezhoud et al. 2014).

The strain was initially grown under different temperature and light intensity conditions to define the optimal autotrophic growth conditions. These preliminary lab experiments indicated that the species had the highest growth rate at temperature of 30 °C and light intensity of 75 μmol photons m−2 s−1 (Mezhoud et al. 2014).

Culture conditions

All cultures were incubated, in triplicate, at 30 °C in a temperature-programmable chamber. Autotrophic and mixotrophic experimental cultures were exposed to halogen lamps providing 75 μmol photons m−2 s−1 and with light/dark period of 16:8 h. Phyto-Claude halogen lamps (400 W) were used to illuminate chambers. The intensity of incident light was measured using the silicon sensor HD 8366.

Bold’s Basal medium (BBM) (Bischoff and Bold 1963) was used in all experiments. It was modified according to the Elser concept for freshwater microalgae with C/N/P ratio equal to 166:20:1 (Elser et al. 2000). The components of modified basal culture medium are as follows: NaNO3 2.5 g L−1, CaCl2. 2H2O 0.025 g L−1, MgSO4. 7H2O 0.075 g L−1, KH2PO4 0.175 g L−1, K2HPO4 0.075 g L−1, NaCl 0.025 g L−1, EDTA 0.05 g L−1, KOH 0.031 g L−1, H3BO3 0.011 g L−1, FeSO4·7H2O 0.048 g L−1, H2SO4 0.001 mL, NaHCO3 16 g L−1, and trace mineral solution 1 mL L−1. The initial pH was adjusted to pH 6.8. The pH of each culture was measured every day and after 7 days of cultivation; the pH increased to 10 and 10.2.

For mixotrophic and heterotrophic cultures, the modified BBM medium was supplemented with 40 g L−1 glucose. Heterotrophic cultures were cultivated in flasks wrapped with aluminum foil.

Cultivation was conducted for 7 days, in sterilized 2-L Erlenmeyer flasks containing 1.8 L of culture and equipped with a device that enabled the aseptic removal of samples. Each batch culture was inoculated with an initial biomass concentration of 0.07 g L−1 from pre-cultures in autotrophic mode. Daily microscopic observations were undertaken and testing for bacterial contamination was also performed by inoculating the cultures into glucose peptone agar plates.

Analytical procedures

Determination of biomass concentration

Graesiella sp. biomass concentration was monitored daily by gravimetric determination of algal biomass dry weight (dw). Daily aliquots of 80 mL algal suspension were centrifuged at 4000 rpm for 10 min at 4 °C and washed with deionized water to remove the salt. The pellets were freeze-dried and subsequently weighed and stored at −20 °C until analysis.

Determination of glucose concentration

Glucose concentration in the supernatant was determined with dinitrosalicylic acid (DNS) method according to Adney and Baker (1996). Reaction mixture contained 1.5 mL of the relevant reducing sugar solution (culture medium) and 3 mL of DNS reagent. The mixture was boiled 5 min in a water bath; subsequently, 0.2-mL sample was withdrawn and diluted with 2.5 mL of distilled water. The absorbance was measured at 540 nm.

Lipid extraction

For all samples, lipid analysis was conducted in triplicate. Pellets of each sample (20–80 mg) were added to 5 mL of a methanol/water/HCl (30:3:1, by vol) mixture into clean Teflon-lined screw-capped glass test tubes and held at 55 °C for 6 h (Rezanka et al. 2003). A total of 15 mL of cold water/hexane (2:1, by vol) mixture was added to the sample and vortex mixed for 20 s. The hexane layer was filtered and concentrated to dryness under a stream of nitrogen at 10 °C. The residue was extracted three times with 5 mL chloroform, then filtered and concentrated to dryness under a stream of nitrogen at 10 °C. The dried hexane and chloroform extracts were combined. The solvent was removed by being flushed with nitrogen, and the total lipid content was estimated gravimetrically.

Fatty acid analysis

To convert fatty acids to methyl esters, 0.5 mL of 14 % boron trifluoride in methanol was added to the dried whole lipids, and the mixture was refluxed for 1 h at 100 °C (Morrison and Smith 1964; Medina et al. 1992). Distilled water (0.5 mL) was added and fatty acid methyl esters (FAME) were extracted three times with hexane. The hexane extracts were rinsed twice with distilled water and evaporated under a stream of nitrogen at 10 °C. FAMEs were detected using a Hewlett-Packard HP 5890, series II gas chromatograph fitted with a flame ionization detector (FID) and equipped with a 60 × 0.32-mm DB-23 polar column. High-purity nitrogen was used as the carrier gas. The initial column temperature was set at 150 °C and was subsequently raised to 250 °C at 2°C min−1. The FID temperature was kept at 280 °C. The fatty acid content was determined by comparison of their integrated peak areas with that of the nonadecanoic acid methyl ester (19:0) as internal standard. The structures of FAMEs were confirmed by comparison of retention times with those of standard FAMEs (PUFA-1 Marine Source, Supelco). We used a fatty acid shorthand notation of the form A:BnX, where A is the number of carbon atoms, B is the number of double bonds, and X is the position of the first double bond relative to the methyl end of the molecule.

Measurement of the polyphasic chlorophyll a fluorescence transient

Variable fluorescence of chlorophyll was measured using the plant efficiency analyzer (PEA-Hansatech) with fixed excitation wavelength at 650 nm. Emission was measured at wavelengths beyond 720 nm and initial time resolution was 50 μs. Before measurement, three replicates of each sample were maintained in the dark for 20 min and then illuminated with an intensity of 3000 μmol photons m−2 s−1. All fluorescence transients were recorded within a time scan of 10 μs to 1 s with a data acquisition rate of 103 readings per second.

Apparent PSII activity parameter were determined from the F0 (initial fluorescence when QA is fully oxidized) and Fm (maximum fluorescence when QA is transiently fully reduced) values of the variable chlorophyll fluorescence. The photosynthesis activity was derived by the maximum quantum yield (Fv/Fm) according to the formula Fv/Fm = (Fm−F0)/Fm, where F0 is the initial fluorescence and Fm is the maximum fluorescence (Krause and Weis 1991).

Statistical analysis

Biomass concentration, lipid concentration, and fatty acid percentage were compared using one-way ANOVA, followed by the Tukey’s multiple comparison test that estimated the significance of the differences (P < 0.05).

Results

Biomass concentration, lipid content, and total glucose uptake

Graesiella sp. biomass and lipid production were evaluated by following the increase in biomass, lipid content, and glucose uptake kinetics of the batch cultures grown over 7 days under different trophic conditions: autotrophy (Fig. 1a), mixotrophy (Fig. 1b), and heterotrophy (Fig. 1c). Routine observation showed bacterial contamination at day 5 in heterotrophic cultures. Subsequently, these cultures were stopped.

Fig. 1
figure 1

Biomass concentration (black circle), lipid content (black up-pointing triangle), and glucose uptake (black square) under autotrophic (a), mixotrophic (b), and heterotrophic (c). Error bars represent means ± SD (n = 3)

Full size image

Batch culture data were used for assessing two parameters describing maximum lipid productivity (gL L−1d−1) and maximum biomass productivity (gdw L−1d−1) (Table 1).

Table 1 Biomass productivity and lipid productivity of Graesiella sp. culture under autotrophic, mixotrophic, and heterotrophic conditions. Values represent means ± SD (n = 3)
Full size table

The highest biomass concentration (0.40 ± 0.01; 0.47 ± 0.04 gdw L−1) was observed under both mixotrophic and autotrophic cultures, and it was three times higher than the maximum biomass obtained in heterotrophic conditions. Nevertheless, mixotrophic conditions showed (Table 1) the highest biomass productivity (0.17 ± 0.03 gdw L−1 d−1) which was 4.47 and 1.42 higher than heterotrophic and autotrophic conditions, respectively.

The maximum lipid content (45.8 ± 1.9 % dw) was produced mixotrophically, and it was respectively 2.32 and 1.28 times higher than those achieved autotrophically and heterotrophically (Fig. 1). Furthermore, the highest lipid productivity (0.07 gL L−1d−1) was observed for mixotrophic cultures, representing respectively 10 and 5.83 times the lipid productivity obtained in autotrophic and heterotrophic ones (Table 1). For both mixotrophic and heterotrophic cultures, the lipid content increased substantially at the early stationary growth phase. Conversely, lipid content in autotrophic growth conditions decreased during the whole time course of culture.

Kinetics of glucose uptake (Fig. 1b, c) showed that for the mixotrophic cultures, glucose uptake was maintained at 9 to 11.6 g L−1 during the exponentially growth phase. Then, a steep rise to 32.2 g L−1 took place at the stationary phase. A similar trend was observed in the heterotrophic cultures with lower glucose uptake, not exceeding 11.6 g L−1 at day 4. During the whole time course of mixotrophic growth, 81 % of the initial glucose concentration was consumed while only 30 % was used during the dark heterotrophic growth conditions.

Photosynthesis activity

Chl-a fluorescence induction curves (OJIP) enable calculation of several phenomenological and biophysical expressions for quantifying the energy fluxes and energy ratios through photosystem II (Mehta et al. 2010; Strasser et al. 2010; Mathur et al. 2011). Ting and Owens (1992) and Schreiber et al. (1995) suggested that chlorophyll fluorescence estimation is not an accurate method for the determination of absolute PSII activity; we have used in our study the maximum efficiency of PSII photochemistry assessed by (Fv/Fm) ratios as a tool only to confirm active photon capture in the light-harvesting antenna complexes of PSII.

At the beginning of the experiments (Fig. 2), mixotrophic and heterotrophic cultures acted similarly as autotrophic cultures and have Fv/Fm above 0.70 ± 0.08 at days 1 and 2. But, at the stationary stage, mixotrophic and heterotrophic maximal quantum yields of PSII dropped significantly to 0.09 ± 0.04, respectively, at day 5. The autotrophic cultures were not significantly affected by growth phase and the Fv/Fm remained steady at 0.77 ± 0.08.

Fig. 2
figure 2

Maximum quantum efficiency of PSII (Fv/Fm) of Graesiella sp. under autotrophic, mixotrophic, and heterotrophic conditions. Error bars represent means ± SD (n = 3)

Full size image

Fatty acid profile

Independent of the nutrition mode and growth phase, the fatty acid profile (Table 2) of this species was characterized by the predominance of unsaturated fatty acids (UFA): palmitoleic acid (16:1), linolenic acid (18:3n3), oleic acid (18:1), and linoleic acid (18:2n6) which represented more than 54 to 77 % of total fatty acids. Saturated carbon chain fatty acids (SFA) (23 to 46 % FA) were also identified with palmitic acid (16:0) and myristic acid (14:0) being the main components. Results also showed that Graesiella sp. synthesize 13:0, 15:1, 17:0, 18:0, 20:1n9, and 22:0 FAs whose content was smaller and did not exceed 10 %.

Table 2 Qualitative analysis of fatty acid profile (as percentages of dry weight) of Graesiella sp. under autotrophic, mixotrophic, and heterotrophic conditions. Values represent means ± SD (n = 3)
Full size table

For all nutritional modes of cultures, a substantial variation in the fatty acid proportions (Table 2) was observed with respect to the growth phases. The exponentially growth phase was mainly characterized by a considerable increase in the proportion of palmitoleic acid (16:1), (from 1.84 at the initial to 2.38–8.27 % dw) while palmitic acid (from 2.29 % dw at the initial to 0.89–1.82 % dw) and the subtotal of unsaturated C18 fatty acids (e.g., linolenic, oleic and linoleic acids) (from 2.85 to 0.84–2.24 % dw) decreased substantially.

This trend was inversed in the stationary growth phase compared to the exponential growth stage. Indeed, palmitoleic acid fraction decreased to 0.20–3.42 % dw, while unsaturated 18 carbon fatty acids became predominant and reached, particularly in heterotrophic and mixotrophic cultures, 6.84 and 16.98 % dw, respectively. The palmitic acid also showed a significant increase from 0.89 to 1.82 % dw at exponential stage to 0.99–7.9 % dw in stationary growth phase.

Comparing the different nutrition modes of the experimental cultures, the unsaturation degree of 18 carbon fatty acids in the stationary growth phase was more pronounced in mixotrophic and heterotrophic modes than it was in the autotrophic cultures. Indeed, the autotrophic mode was characterized by the predominance of 18:1n9 (0.75 % dw) and 18:2n6 (1.27 % dw) fatty acids, while in the mixotrophic and the heterotrophic cultures, the 18:2n6 (14.10 and 2.21 % dw, respectively) and the 18:3 n3 (2.56 and 3.62%dw, respectively) fatty acids became predominant.

Discussion

The current study revealed the extent to which Graesiella sp. had the capacity to grow and to produce lipids phototrophically and by utilizing glucose. The results clearly showed that supplementation of glucose led to a significant improvement of both growth and lipid production when the microalga was grown mixotrophically. The major feature of mixotrophic culture is the presence of two energy sources, organic carbon sources and light, so microalgae may benefit from both photoautotrophic and heterotrophic growth. In this study, in mixotrophic cultures, photoautotrophic and heterotrophic growth proceeds independently and in an additive manner, such that the maximum biomass productivity was about equal to the sum of the photoautotrophic and heterotrophic modes. These results are consistent with those obtained for several microalgae, including Chlorella vulgaris and Chlorella sp. (Ogawa and Aiba 1981; Orús et al. 1991, Ratha et al. 2012) and for several cyanobacteria (Marquez et al. 1993).

Conversely, previous reports showed that high glucose concentrations can reduce growth rates (Wan et al. 2011) and/or the photosynthetic efficiency (Liu et al. 2009) in several microalgae strains which was attributed to the presence of substrate inhibition (Liang et al. 2009; Pagnanelli et al. 2014). The inhibitory glucose concentration varies between 5 and 166 g L−1 depending on microalgal species and environmental conditions (Perez-Garcia et al. 2011). Interestingly, although large amounts of glucose were used in this work (40 g L−1), the photosynthetic efficiency was not affected during the exponential growth phase of mixotrophic cultures. Indeed, the rate of the photochemical efficiency of photosystem II was at the same level to autotrophic mode during the correspondent growth phase.

Negative effect of the glucose concentration on photosynthetic activity was observed during the mixotrophic stationary phase of culture. Indeed, the photosynthetic activity declined strongly after the first 2 days of culture, as shown by the rapid decrease of the maximal quantum yields of PSII in both mixotrophic and heterotrophic modes compared to autotrophic mode. This trend confirmed that Graesiella sp. was then undergoing completely heterotrophic metabolism. The substantial increase of the glucose uptake during the stationary growth phase under mixotrophic cultures supported this hypothesis. Thus, as shown in many eukaryotic microalgae (Wilken et al. 2014), besides dark respiration organic carbon can also be metabolized using energy derived from light which is converted to ATP via the cyclic electron transport around PSI (photoheterotrophy).

Compared to autotrophic cultures, the highest lipid productivity was reported for mixotrophic conditions. Under mixotrophic condition, the lipid productivity reached its maximum level at the stationary growth phase. Since glucose uptake in these conditions was also highly enhanced and since photosynthesis was suppressed, this leads us to suppose that the carbon excess under mixotrophic condition was then mainly used into photoheterotrophic lipid biosynthesis. Hence, in heterotrophic conditions, the consumption of sugars at a rate higher than the rate of cell generation could promote conversion of excess sugar into lipids (Chen and Johns 1991; Ratledge and Wynn 2002; De Swaaf et al. 2003).

This trend was not confirmed under dark heterotrophic cultures which showed very low growth capacity and where the lipid biosynthesis declined even when the growth declined and the glucose uptake was enhanced. This may suggests that the operation of enzymes related to growth and lipid synthesis by Graesiella sp. is light dependent. The effect of light on synthesis and accumulation of enzymes catalyzing lipid synthesis was not well investigated, especially in microalgae species. Nevertheless, light has been shown to induce the increase of MGD transcripts followed by the activation of the monogalactosyldiacylglycerol (MGDG) synthase in terrestrial plants (Yamaryo et al. 2000).

Under different trophic modes and different growth phases, the predominant Graesiella sp. fatty acids were 14:0, 16:0, 16:1, 18:0, 18:1n9, 18:2n6, and 18:3n3. In relative terms of the fatty acid composition of Graesiella sp., our results are in agreement with the earlier studies of Chlorella species, mainly concerning the predominance of UFA (Mendoza Guzmán et al. 2011; Seyfabadi et al. 2011; Li et al. 2011; Guedes et al. 2011; Ngangkham et al. 2012). Little variations could be recorded due to different growing conditions and the methods of extraction of lipids and fatty acids.

The ratio of UFA to SFA has been used as indirect indicator of membrane fluidity (Casadei et al. 2002). In this work, this ratio was unaffected under different growth phases for autotrophic mode. It reaches its highest level in exponential growth phase for both heterotrophic and mixotrophic modes. Thus, we can conclude that Graesiella sp. cells increase their membrane fluidity when the culture medium was supplemented with glucose which could be related to the glucose uptake efficiency. Moreover, at high growth rates, the demand for structural components increases, consequently increasing the proportion of UFAs (Kates and Volcani 1966). At the stationary stage for both mixotrophic and heterotrophic modes, the UFA/SFA ratio decreased mainly due to the high increase of 16:0. This was also observed by Fisher and Schwarzenbach (1978) for other microalgae, supporting the idea that 16:0 could represent an energy storage linked to extra production when the cell division is altered (Siron et al. 1989). Moreover, for all tested trophic modes in this work, the PUFA/MUFA ratio increased substantially in the stationary phase. This increase was essentially due to the reduction of 16:1 and the increase of 18:2 and 18:3 fatty acids, although biosynthesis of PUFA was observed mainly during intense cellular activity in a few microalgae species (Brown et al. 1996; Liang et al. 2006). Bigogno et al. (2002) and Mansour et al. (2005) reported that PUFA could be accumulated as reserve building blocks under high culture density.

In conclusion it was found that the addition of glucose in mixotrophic culture was able, compared to autotrophic mode, to enhance more than 1.4-fold the biomass productivity and more than 9-fold the lipid productivity for this thermophilic Graesiella strain. Moreover, mixotrophic mode showed also high content of polyunsaturated 18:2n6 and 18:3n3 fatty acids, exceeding 16.6 % dw with 14.1 % dw of 18:2n6. In dark heterotrophic mode, high level of 18:3n3 (3.2 % dw) was observed. Nevertheless, high lipid contents were associated with relatively low biomass productivity and, consequently, low overall lipid productivity. Hence, a two-phase cultivation strategy in mixotrophic mode can be suggested to promote the lipid productivity of this strain of Graesiella. A first phase could be conducted in reactor to optimize growth and biomass production followed by a second lipid induction phase, under a high biomass concentration, to promote linoleic acid (18:2n6) and linolenic acid (18:3n3).