Introduction

The microalgae, a diverse group of eukaryotic photosynthetic organisms, occupy diverse habitats including freshwater, marine and more extreme environments such as hot springs and frozen regions (Li et al. 2012a, b). This ability to adapt to different environmental conditions is due to the plasticity of algal physiology. Microalgae have been used as a source of biodiesel (Chisti et al. 2007) and they are considered promising candidates for large-scale biofuel production due to their high photosynthetic efficiency and oil accumulation ability (Borowitzka and Moheimani 2013; Moody et al. 2014).

Abiotic stress leads to a variety of cellular responses in eukaryotic photosynthetic algae, including changes to the photosynthetic apparatus (Glaesener et al. 2013). In addition, stress leads to increased accumulation of lipids, particularly triacylglycerides (TAG) (Juergens et al. 2015). Lipid accumulation is thought to help microalgae withstand adverse environmental conditions (BenMoussa-Dahmen et al. 2016). Nitrogen starvation, which is normally detrimental to algal growth, was found to cause increased TAG accumulation and lipid droplet formation (James Gabriel et al. 2011). Another abiotic stress, salinity, was also shown to significantly enhance the synthesis of neutral lipids, specifically TAG. The limitation of micronutrients (e.g. zinc and iron) can result in the alteration of membrane lipids and the formation of lipid droplets (Chen et al. 2011). In conditions of iron and zinc deprivation, the growth of D. tertiolecta is severely inhibited and the cells accumulate neutral lipids (Chen et al. 2011; Urzica et al. 2013). Microalgae have two iron-rich organelles: the chloroplast and mitochondrion. Both organelles contain numerous iron-dependent proteins whose functions are essential in electron transfer pathways. An inadequate supply of iron limits the photosynthetic process leading to lipid accumulation (Glaesener et al. 2013).

The unicellular green microalgae, Chlamydomonas (C.) reinhardtii, is a model organism that has been used in the study of basic physiology and for investigating the regulation of primary metabolism (Merchant et al. 2007). It is genetically well characterized and displays a significant response to adverse conditions, so it was chosen as a suitable organism to investigate changes in lipid metabolism caused by iron deficiency. According to the current model, TAG accumulation in C. reinhardtii and related species occurs in the endoplasmic reticulum (ER), plastids, and at contact site between plastids and the ER (Fan et al. 2011). The final step of TAG biosynthesis is catalysed by diacylglycerol acyltransferase (DGAT), which transfers an acyl group to sn-1,2-diacylglycerol (DAG). Chlamydomonas has at least two types of DGAT that use acyl-CoA as the acyl donor: type I represented by DGAT1 and type II represented by DGTT1–DGTT5 (Boyle et al. 2012). Another type of acyltransferase, phospholipid:diacylglycerol acyltransferase 1 (PDAT1), transfers acyl groups from a wide range of glycerolipids to TAG (Yoon et al. 2012). Recently, monogalactosyldiacylglycerol (MGDG), a signature lipid class of photosynthetic membranes, was identified as a plastidic source of certain acyl groups in TAG biosynthesis (Li et al. 2012a, b). The mechanism of lipid biosynthesis in C. reinhardtii and the genes/proteins involved have not been extensively studied under Fe deficiency. Moreover, little is known about their expression under stress conditions such as micro-nutrient deficiency. Due to its potential influence on TAG accumulation, iron deprivation stress is likely to have wide-ranging effects not only on iron-rich organelles like chloroplasts and mitochondria but also on ER and membrane structures.

Under iron-deficient conditions, photosynthetic eukaryotes become chlorotic and their plastids are often smaller and less well developed (Moseley et al. 2002). The abundance of photosynthetic machinery is decreased due to chlorosis, which is preceded by structural remodelling of photosystems (PS) I and PSII (Moseley et al. 2002; Busch et al. 2008; Yadavalli et al. 2012a; Devadasu et al. 2016). In higher plants, the greatly decreased level of photosynthetic proteins also alters electron transport in PSII and PSI (Andaluz et al. 2006; Timperio et al. 2007; Msilini et al. 2011; Chen et al. 2015). In cyanobacteria deprived of iron, the light-harvesting phycobilisomes were reduced and an iron-stress-induced operon isiAB is expressed (Bibby et al. 2001; Boekema et al. 2001). Iron stress-induced protein A (IsiA) displays homology to CP43, a chlorophyll a-binding protein of PSII. A ring of 18 molecules of IsiA forms around the PSI trimeric reaction centers (Bibby et al. 2001; Boekema et al. 2001). Another report states that two centric layers of IsiA form around trimeric PSI following prolonged growth in nanomolar Fe levels (Chauhan et al. 2011). In addition, the major effect of iron deprivation has been reported to be monomerization of the PSI trimers of whole cells in vivo, but not the formation of IsiA ring around the PSI trimers. This is quite spectacular as observation, but represent only a small fraction of the PSI reaction centers under iron-limited conditions (Ivanov et al. 2006).

In conditions of Fe deficiency, the PSI-associated light-harvesting complex I (PSI-LHCI) of C. reinhardtii undergoes remodelling which results in a decreased PSI antenna size (Naumann et al. 2005, 2007). These structural changes decrease the efficiency of excitation energy transfer between LHCI and PSI. PSII is also damaged by iron deprivation but to a lesser extent than PSI (Yadavalli et al. 2012a; Devadasu et al. 2016). Iron stress also induces TAG accumulation and lipid droplet formation in C. reinhardtii. In this study, C. reinhardtii was subjected to progressive iron deficiency conditions in order to understand the changes induced in photosynthetic efficiency and lipid metabolism. Furthermore, TAG accumulation and lipid droplet formation under iron stress were explored and the expression of key genes/enzymes connected with photosynthesis efficiency was studied.

Materials and methods

Culture and growth conditions

Chlamydomonas reinhardtii CC125 wild-type (wt) and the mutant strain NPQ4 (npq4) were obtained from the Chlamydomonas resource centre (http://www.chlamy.org/). Cells were grown photoheterotrophically in Tris–acetate phosphate medium at 25 °C with a photon flux density of 50 µmol photons m−2 s−1 in 250-mL conical flask shaken at 120 rpm in an orbital shaker (Terauchiet al. 2010; Yadavalliet al. 2012a).

Iron starvation time course

To induce iron starvation, cells (wt and npq4) were grown to mid-logarithmic phase (3 × 106 cells mL−1), then collected by centrifugation (3000 rpm for 5 min at RT) and washed three times in Fe-less TAP medium. The cell suspensions were then used to inoculate 500 mL lots of Fe-less TAP medium in 1-L conical flasks. Cultures were incubated as described above and samples were collected at 0, 12, 24, 48, and 72 h. Parallel cultures were grown in standard TAP medium and samples were taken according to the same schedule as controls for subsequent analysis. Further, 50 mL of Fe deficiency cells [first generation of Fe deficiency (1st batch)] was centrifuged and these cells were again inoculated to the same amount of Fe deficiency [2nd generation of Fe deficiency (2nd batch)] medium (50 mL) and allowed them to grow up to 5–6 days. Similarly, the first and second generation of Fe deficiency cells were centrifuged and inoculated to the same amount of Fe containing medium and allowed them to recover till 5–6 days.

Biomass determination

For total biomass determination, 5 mL of the + Fe and − Fe cultures were collected after 72 h incubation and centrifuged in weighed tubes at 3000×g for 5 min. The pelleted cells were washed once with distilled water and then lyophilized at 4 °C for 8 h. The tubes were then weighed again and the total dry masses calculated.

Nile Red staining and microscopy

Samples collected from the cultures at 0, 12, 24, 48 and 72 h of incubation were immediately mixed with iodine solution (5 µL; 0.25 g Iodine in 100 mL of 95% ethanol) to immobilize the cells. The cell suspensions were then stained with Nile Red (NR) (1 µg mL−1 final concentration, Sigma). Staining was performed in darkness for 20 min at room temperature with constant gentle agitation. The cells were then concentrated by centrifugation at 500×g for 3 min. Fluorescence and transmission images of the stained cells were acquired using a Zeiss LSM 8 PASCAL confocal microscope system with PASCAL version 4.0 software. The microscope was fitted with LASOS HeNe 543 nm and Argon 488 nm scanning lasers. Scans of the algal cells were taken with a × 60 objective at a pixel resolution of 1024 × 1024 in 8-bit format (pixel intensity range 0–255). Laser transmission and scan settings remained constant in all scans. Leica confocal software was used for image analysis.

Chlorophyll a OJIP fluorescence measurements

Chlorophyll a fluorescence fast induction curves (O–J–I–P) were measured using a plant efficiency analyzer (PEA, Hansatech, Kings Lynn, Norfolk, UK). The cells were dark adapted for 20 min and then exposed to a photon flux density of 3000 µmol photons m−2 s−1 produced by an array of three light-emitting diodes peaking at 650 nm.

Pulse amplitude modulated (PAM) fluorescence measurements

Culture samples were adjusted to a density of 3 × 106 cells mL−1 and dark adapted for 20 min before fluorescence measurement performed using a pulse amplitude modulated 101 fluorometer (Walz GmbH, Effeltrich, Germany) connected to a PAM Data Acquisition System 100 controlled by PAM WINCONTROL v2.08 software (Heinz Walz GmbH, Effeltrich, Germany). The maximum potential quantum efficiency of PSII [Fv/Fm = (Fm − Fo)/Fm] was assessed after 20 min of dark incubation. Subsequently fluorescence rapid light curves (RLCs) were obtained using eight different actinic light intensities (Kosourov et al. 2007) (145, 221, 344, 536 and 830 µmol photons m−2 s−1 for 30 s each). Subsequently, the actinic light (250 µmol photons m−2 s−1) was maintained for 5 min, before a final saturating pulse was applied. The fluorescence rapid light curve (RLC) parameters were calculated according to Kosourov et al. (2007). Similarly, the Chlorophyll a fluorescence measurements were carried out from Handy PEA and PAM fluorescence for two generations of Fe deficiency cells and also from recovery for both generations of Fe sufficient cells.

Quantification of photosynthetic pigments and carotenoids

After 72-h growth, cultures of C. reinhardtii at a cell density of approx. 3 × 106 cells mL−1 were harvested by centrifugation. The cell pellets were resuspended in 80% acetone and held for 20 min in the dark at − 20 °C to extract the pigments. Total chlorophyll was estimated according to the method of Porra et al. (1989). For the estimation of total carotenoids, an equivalent number of cells were resuspended in 95% ethanol and held for 4 h in the dark at 4 °C. The extracts were then centrifuged and total carotenoids in the clarified supernatants were determined by measuring the optical density at 470 nm as previously described (Wellburn 1994).

Nile Red fluorescence assay for triacylglycerides (TAG)

The C. reinhardtii cultures were diluted to a density of 3 × 106 cells mL−1 and placed in the wells of a 96-well microplate. To each well, a 5 µL aliquot of 50 µg mL−1 Nile Red (NR; Sigma; prepared in acetone) was added, mixed in by agitation and the plate incubated in the dark for 20 min. The neutral lipids were then measured by fluorescence (excit. 485 nm/emiss. 595 nm) using a plate reader (Tecan infinite 200, Magellan). Quantification was achieved using a standard curve prepared with the lipid Trioleine (Sigma T4792).

FACS analysis of neutral lipids

To further quantify neutral lipids stained with Nile Red, the fluorescence intensity was measured by flow cytometry. A 10 µL aliquot of 100 µg mL−1 NR in acetone was added to 3.0 × 106 cells in 1 mL of TAP medium, gently mixed in and the tube held in the dark for 30 min. The cells were then harvested by centrifugation and resuspended in a small volume. Flow cytometry analysis was performed using a FACS Calibur Flow Cytometer (BD Falcon, USA) fitted with a 488-nm excitation filter and a 545-nm emission filter. Ten thousand cells were analysed without gating and auto fluorescence (from chlorophyll) was nullified before measuring the NR fluorescence. The fluorescence of NR was plotted on both dot plot and histogram.

Neutral lipid extraction and TLC analysis

The identification of TAG was performed by thin layer chromatography. To extract lipids, a 5-mg sample of lyophilized biomass was resuspended in a mixture of methanol, chloroform and 88% formic acid (2:1:1) plus 0.9% KCl (Andaluz et al. 2006). After vortexing to mix, the homogenate was centrifuged at 3000×g for 5 min to cause a phase separation. The organic layer was transferred to a clean glass tube and evaporated by nitrogen gas purging. The dried extract was then dissolved in 50 µL chloroform. As a standard for TAG identification, Triolein (Sigma-Aldrich) was dissolved in CHCl3 for mg mL−1 concentration. The samples or Triolein standard (10 µL) were applied to silica gel plates by spraying using a Linomate 5 (Camag). The TLC plates were then run in a chamber filled with a mixture of petroleum ether–diethyl ether–acetic acid (80:20:1) for 20 min. Bands were visualized by brief staining with iodine.

Gas chromatography–mass spectrometry analysis

For extraction of lipids, the procedure is same as that of extraction procedure for TLC analysis. Extracted fatty acids were converted to their corresponding methyl esters by treatment with diazomethane (Cakmak et al. 2014). Heptadecanoic acid (C17:0) was added to samples as an internal standard during derivatization for both full scan and selected ion monitoring (SIM) mode analysis. Fatty acid methyl esters (FAMEs) were analysed by Gas Chromatography with an Electron Impact Mass Spectrometry (GC-EIMS) Agilent 7890B gas chromatograph (Agilent Technologies, Palo Alto, CA) equipped with a 5977A mass selective detector. An AB-5MS capillary column (length 30 m, i.d. 0.25 mm, film thickness 0.25 µm) was used for chromatographic separation. Samples of 1 µL were injected into the GC–MS instrument in split less mode. The GC oven was programmed to hold the temperature at 50 °C for 2 min followed by a 10 °C min−1 temperature ramp to 210 °C. After a 1 min hold at 210 °C, there followed a 2 °C min−1 ramp to 220 °C with a 2 min hold, and then a 10 °C min−1 ramp to 280 °C with a 2-min hold. Helium was used as the carrier gas in constant flow mode at a flow rate of 1 mL min−1. The inlet and GC–MS interface temperatures were maintained at 250 °C and 280 °C, respectively. The solvent delay time was set to 4 min. The EI source and quadrupole temperatures were maintained at 230 °C and 150 °C, respectively. The mass spectrometer was scanned from m/z 29 to 600 for a full scan and SIM analysis. All fatty acid methyl esters (FAME) were identified and confirmed by NIST-Wiley library. FAME were further quantified by SIM mode analysis of specific ions m/z 74, m/z 87, for saturated and monoenoic acids, as well as m/z 79 and m/z 81 for FAMES having more than one double bond.

Protein identification by immunoblotting

Chlamydomonas reinhardtii cells were resuspended in SDT lysis buffer (4% SDS and 0.1 M DTT in 0.1 M Tris–HCl, pH 7.6), vortexed for 2 min and incubated at 95 °C for 5 min. The lysates were further centrifuged at 14,000×g for 10 min at room temperature and the supernatant transferred to a fresh tube. The total protein in each supernatant was quantified using the Bradford method (Bio-Rad) and samples containing 20 µg of protein were analysed by polyacrylamide gel electrophoresis (PAGE) using 10% or 15% Bis-Tris gels for DGAT2A and PDAT1, or 12% SDS gels for other proteins. Using a standard methodology (Towbin et al. 1979), the separated proteins were then transferred to nitrocellulose membrane, incubated with primary antibodies, and specific binding was disclosed. Specific primary antibodies (at different dilutions) were used to detect photosynthetic proteins: PsaA (1:2000), PsbA (1:5000), Rubisco small subunit, SSU (1:1000), and LHCSR3 (1:1000). Other antibodies were used to evaluate the expression patterns of acyl-CoA:diacylglycerol; acyltransferase (DGAT2A, 1:1000) and phospholipid:DAG acyltransferase (PDAT1, 1:250) (Boyle et al. 2012). An antibody against Histone H3 (1:5000) was used to confirm equivalent loading of the gels and efficient protein transfer to the membrane. The identification of major lipid-droplet protein (MLDP) was performed using an antibody (1:250) provided by Anthony H.C. Huang, USA. All other primary antibodies were obtained from Agrisera (Sweden). The secondary antibody anti-chicken (1:10000) was used for identification of PsbA (Agrisera) and anti-rabbit (1:10000) was used for identification of other proteins (Cell Signalling Technology). Immunoblot images were obtained using a Chemidoc-touch imaging system (Bio-Rad).

Results

Growth characteristics and pigment analysis

To study the impact of iron deficiency on cell growth and structure, as well as on the photosynthetic apparatus, C. reinhardtii cells were propagated in iron-sufficient (control) and iron-deficient TAP medium. Growth curves were plotted by monitoring the change in A750 of the cultures (Supplementary Fig. 1). In the iron-sufficient conditions, the absorbance value increased twofold after 2 days and then reached a plateau at approximately 72 h. Similarly, the absorbance of the iron-deprived culture doubled over the first 2 days in culture but then it decreased to approximately half of the maximum value over the following 4 days. Notably, the cells grown in iron-deficient (− Fe) condition grew considerably slower in comparison to the control (+ Fe) cells. Further the second generation of Fe deficiency cells were grown very slowly (more than 6 days) than the first generation. These cells were centrifuged and re-inoculated to Fe sufficient medium to observe the recovery. In second generation cells, the recovery took place longer (more than 6 days) (data not shown). This difference in growth rate highlighted the importance of iron nutrition to Chlamydomonas cell growth.

To determine the effect of iron deficiency on the level of pigments associated with photosynthetic complexes of C. reinhardtii, total chlorophyll and pigment levels were quantified. Total chlorophyll levels were reduced by around 50% in iron-deprived cells compared with iron-sufficient cells (Supplementary Fig. 2a). The recovery also followed with growth pattern that the second generation cells accumulated the Chl to the control level after 6–7 days, whereas the first generation recovery was faster (Supplementary Fig. 2b). A change in the level of total carotenoids was also observed (data not shown). It has previously been shown that during iron starvation the membrane lipids of chloroplasts are remodelled and the levels of photosynthetic pigments and proteins are decreased in C. reinhardtii (Yadavalli et al. 2012a; Urzica et al. 2013). The degradation of the photosynthetic apparatus and decreased protein turnover are common metabolic changes that occur during iron deprivation. Our results confirm the importance of iron as an essential micro-nutrient for growth and chlorophyll biosynthesis in C. reinhardtii.

Chlorophyll a fluorescence analysis

Chl a fluorescence induction curves were used to monitor several photosynthetic events in C. reinhardtii strains (wt and npq4 mutant) grown under Fe deficiency. When measured under saturating light Chl a fluorescence induction curves show a fast multiphasic rise occurring within the µs to second(s) range known as the O–J–I–P transients (Kodru et al. 2015). In dark-adapted C. reinhardtii cells, it was shown that QA, an electron acceptor of PS II, is in the oxidized state at the O transient (the initial, minimum fluorescence). The O–J fluorescence rise (J is an inflection between O and the peak P) is a photochemical phase which is relevant to the reduction of QA to QA−, whereas J–I (I being a second inflection between O and P) and I–P are thermal phases involving the reduction of the PQ pool (Stirbet and Govindjee 2011, 2012) as well as that of the electron acceptor side of PS I (Schansker and Strasser 2005; Kodru et al. 2015). As in our previous report (Kodru et al. 2015), typical OJIP curves were obtained from wt and the npq4 mutant C. reinhardtii strains. In the cells grown under Fe deficiency (4 days), the O phase (initial fluorescence, Fo) was increased in the wt, whereas in the npq4 mutant it was reduced. Furthermore, the OJ, JI and P phases were shorter in both strains grown in iron-deficient conditions, although this truncation was more drastic in the npq4 mutant (Fig. 1a, c). Normalization of the fluorescence curves at O and P indicated that the changes occurred equally in all phases (Fig. 1b). In addition the components of fluorescence, Fm and Fv/Fm were drastically changed in both wt and npq4 mutant under iron deficiency. The substantial reduction in the Fm value in the mutant indicated that non-photochemical quenching is very important in protection against Fe deficiency. From Fe deficiency, OJIP curves show that there was dramatic change in photosynthetic efficiency but still the cells were showing significant photosynthetic activity. In order to understand the severity of the Fe deficiency on cells, the cells grown under Fe deficiency were re-inoculated to Fe deficiency medium. After 6 days of growth, the photosynthesis efficiency was drastically reduced when compared to the first generation of Fe deficiency. We know that the first generation of Fe deficiency cells could recover the photosynthetic activity to the control cells after 4 days (Devadasu et al. 2016), whereas the second generation cells restored the photosynthetic activity after 6–7 days (Fig. 1f, g). It is interesting to note that both generations of Fe deficiency cells were in shock but they were alive.

Fig. 1
figure 1

a Chl a fluorescence transient (OJIP) curves from cultures of wt and npq4 mutant C. reinhardtii grown in iron-sufficient control (+ Fe) medium or iron-deficient (− Fe) medium; b double normalized values of OJIP curves; c Fo values; d Fm values; e Fv/Fm; f shows Chl a fluorescence transients (OJIP) curves from control, recovery first (R1) cells, recovery second (R2) cells; g shows the Fv/fm value from control (c), recovery first (R1), recovery second (R2). All samples were excited at 650 nm with 3000 µmol photons m−2 s−1 light. The measurements were repeated three times using different individual cultures

Full size image

Photosynthetic efficiency

Photosynthetic parameters were recorded for the cells grown from iron-sufficient and iron-deficient conditions (Fig. 2). Notably, the electron transport rate (ETR) decreased faster with increasing light intensity (250 µmol photons m−2 s−1) in the iron-deprived cells. This indicates that the electron flow is altered in cells grown under iron-deficient conditions. In addition, PSII quantum efficiency [Y(II)] values in increasing light intensities were decreased in both the wt and the npq4 mutant under iron-deficiency compared with the iron-sufficient control (Fig. 2b). The first generation recovery of ETR and Y(II) values was reached to the control level after 4 days, whereas in second generation the recovery was after 6–7 days (Supplementary Fig. 3). This is due to changes in the redox state of the primary electron-accepting plastoquinone of the PSII centre in Fe-deficient conditions. Furthermore, compared with the control cultures, NPQ increased with rising light intensities in wt cells, and to a lesser extent npq4 cells grown under iron deficiency (Fig. 2c). Alterations in the PQ pool in iron-deficient cells could be attributed to the failure of PSI to draw electrons out of the pool or the activation of a mechanism to increase electron flow into the pool. Taken together, these results suggest that NPQ4/LHCSR3 modulates photosynthetic acclimation to iron-deficient conditions in C. reinhardtii.

Fig. 2
figure 2

Cells grown in iron-deficient (− Fe) conditions shows decreased photosynthetic electron flow compared to those grown with sufficient iron (+ Fe). a Irradiance dependence of relative electron transport through PSII (ETR); b quantum yield of PSII Y(II); c non-photochemical quenching (NPQ) values for wt and npq4 mutant cells grown in control (+ Fe) and iron-deficient (− Fe) medium. The values are the means ± SD from three independent experiments

Full size image

Lipid accumulation observed by confocal microscopy

The accumulation of lipid bodies under iron-deprived conditions was examined by confocal microscopy (Fig. 3a). Iron-deficient cells were half the normal size, displayed an abnormal morphology and a reduced chlorophyll content (data not shown). Lipid body production was clearly induced within 48 h of transfer to TAP − Fe medium and the cells accumulated a considerable number of lipid bodies. In contrast, cells grown in iron replete medium did not show any visible accumulation of lipid bodies. Thus, Fe deficiency caused C. reinhardtii cells to synthesize additional lipids. The lipid accumulation in recovery cells was the same like in control (data not shown).

Fig. 3
figure 3

Lipid accumulation in C. reinhardtii cells examined by Nile Red staining and confocal microscopy. a Lipid droplets detected by fluorescence microscopy. Scale bar, 20 µM. DIC differential interference contrast, scale bar 10 µM; the images are representative of three independent replicates. b Formation of lipid droplets in cells exposed to 4 days of iron starvation

Full size image

Identification of neutral lipids by Nile Red fluorescence

The neutral lipid content of cells during the growth of cultures for 4 days in control or iron-deficient conditions was determined by staining with Nile Red and measuring fluorescence (Fig. 4). Nile Red fluorescence has previously been shown to be proportional to the level of neutral lipid and cell density (Chen et al. 2009, 2011). After staining, numerous fluorescent lipid droplets (LDs) were observed in iron-deprived cells, but none or only a few were present in control cells (Fig. 3b). The LDs were highly fluorescent when viewed using the FITC filter set and were clearly distinguishable from the large chloroplast. A time course experiment using a TECAN fluorescence spectrophotometer revealed that the formation of LDs occurred within 48 h of transferring cells to iron-deficient medium, as gauged by the increase in Nile Red fluorescence (Fig. 4a). However, this was accompanied by a reduction in the optical density at 750 nm, a parameter indicative of cell growth. This OD750 value (0.395 ± 0.05) was around half that of the culture grown in control iron-sufficient conditions (0.822 ± 0.10), suggesting growth inhibition and a consequent decline in biomass. Overall, iron-deficient cells exhibited a fourfold increase in their neutral lipid content over the course of 4 days. For analysis of lipids, we restrict to the wt since the effect of Fe deficiency was more in npq4 mutant.

Fig. 4
figure 4

Neutral lipid content of C. reinhardtii evaluated by Nile Red fluorescence. a Cultures grown in control (+ Fe) and iron-deficient (− Fe) medium stained with Nile Red on days 1–4, then examined using a plate reader fluorometer. b FACS-two-dimensional dot plots of iron-deficient culture: unstained (red), Nile Red-stained (blue). c Histogram plot for iron-deficient unstained cells (background, black line), and iron-sufficient (red line) and iron-deficient cells stained with Nile Red (blue line). Statistical analysis of the median FITC-A values is presented

Full size image

Analysis of neutral lipids by FACS

Neutral lipid accumulation in cells grown for 4 days in iron-sufficient (+ Fe) and -deficient (− Fe) conditions was monitored by Nile Red staining and flow cytometry. Significant amounts of neutral lipids were accumulated in iron-deprived cells (mean fluorescence value, 29.27 ± 0.93) compared to control cells (7.293 ± 1.09) (Fig. 4b, c). These results demonstrate that the metabolism of iron and carbon in C. reinhardtii is interlinked, so that under conditions of iron deficiency a greater proportion of carbon fixed by photosynthesis (i.e. the output of the Calvin–Benson cycle) is channelled to produce neutral lipids (Velmurugan et al. 2014).

Impact of iron deprivation on photosynthetic proteins

The C. reinhardtii stress response protein LHCSR3 is important for energy-dependent quenching (qE), and its expression is induced under high light conditions that up-regulate qE (Peers et al. 2009; Bonente et al. 2011). C. reinhardtii has three genes for LHCSR3 isoforms (lhcsr1, lhcsr 3.1, and lhcsr 3.2), but lhcsr 3.1 and lhcsr 3.2 encode the same 259-amino acid polypeptide (Merchant et al. 2007). The LHCSR1 isoform has 253 residues with 87% sequence similarity to LHCSR3.1/LHCSR3.2. LHCSR3 expression is induced by low pH, the trigger of non-photochemical quenching. Surprisingly, we found that levels of this protein were increased in Fe-deficient cells. These results are consistent with the increase of NPQ fluorescence in wt and decrease in the npq4 mutant (Fig. 3). Furthermore, we observed marginal changes in the levels of both RUBISCO subunits (SSU). As expected, and in agreement with the Chl a fluorescence data (Figs. 2, 3), the levels of the photosynthetic reaction centre proteins psaA (PSI) and D1 (PSII) were significantly reduced under Fe-deficient conditions, which indicates decreased photosynthesis (Fig. 5a).

Fig. 5
figure 5

Effect of iron deficiency on C. reinhardtii protein abundance monitored by western blotting. a For photosynthetic polypeptides, soluble or membrane protein extracts (20 µg protein/lane) were separated by denaturing gel electrophoresis, transferred to PVDF or nitrocellulose membranes, and probed with specific antibodies for the indicated proteins. Histone H3 was used as a loading control. b Western blot analysis of DGAT2A (Acyl-CoA:diacylglycerol acyltransferase) and phospholipid:diacylglycerol acyltransferase 1 (PDAT1) from cells grown for 0–72 h in iron-deficient conditions. Protein extracts (10 µg protein/lane) were resolved on denaturing 10% Bis-Tris gels and equal loading was confirmed by CBB staining. c MLDP protein level in cells grown under iron-deficient conditions

Full size image

Impact of iron deprivation on fatty acid metabolism

Following the observation that iron starvation resulted in TAG accumulation and changes in FA levels, we examined levels of two key enzymes involved in lipid synthesis in C. reinhardtii. Phospholipid:diacylglycerol acyltransferase (PDAT1) (Boyle et al. 2012) is an enzyme that catalyses TAG synthesis via two pathways: transacylation of diacylglycerol with acyl groups from phospholipids and galactolipids, and DAG:DAG transacylation (Yoon et al. 2012; Kobayashi et al. 2013). Another enzyme, diacylglycerol acyl-transferase (DGAT2A) (Boyle et al. 2012), contributes to the incorporation of hydroxylated fatty acids into TAG (Zhang et al. 2009). There are six genes encoding DGAT in Chlamydomonas and previous studies have shown that the expression of DGAT2A is increased in nitrogen-starved cells (Boyle et al. 2012). We analysed DGAT2A and PDAT1 protein levels by western blotting using histone H3 as a control. After 72 h of growth in iron-deprived conditions the level of both enzymes was increased, although the expression of DGAT was lower than that of PDAT1.

Chlamydomonas reinhardtii under nitrogen-deficient conditions show an increase in the number of lipid droplets in the cell (Kropat et al. 2011), which indicates that TAG are accumulated as an energy source in times of nutrient stress. Above we demonstrate that this also occurs in conditions of iron deficiency. In primitive and higher plants, intracellular neutral (TAG) storage lipid droplets (LD) are stabilized by a surface layer of phospholipids and a structural protein oleosin. Oleosins have proven difficult to detect in chlorophytes (green algae) (Huang et al. 2013) and the predominant protein associated with LD is termed the major lipid-droplet protein (MLDP). Immunoblot analysis using a MLDP-specific antibody showed increased levels of this protein in cells grown under iron-deprived conditions (Fig. 5c). Taken together, these results confirm that C. reinhardtii produces increased levels of neutral lipids that are stored in LD under iron-deficient conditions.

Quantification of total lipids by TLC

Biomass production occurs via a series of highly coordinated cellular processes, starting with the capture of light by photosystems, followed by the fixation of CO2 through the Calvin–Benson cycle, and resulting in cell growth and division. In comparison with growth in iron-sufficient conditions, we observed a twofold decrease in total biomass in cells subjected to iron starvation (Fig. 6a). To characterize the overall change in lipid content caused by iron deficiency, we prepared a neutral lipid extract from equal dry weights of cells cultured on both control and Fe-depleted medium, and these were analysed by TLC. The TAG concentration was quantified based on a standard applied to the same TLC plates. The analysis of lipid samples from cells collected on each day of culture showed that the TAG content was increased significantly by Fe deficiency (Fig. 6b), although there was no change in polar lipid content.

Fig. 6
figure 6

Lipid production by C. reinhardtii cells grown in control (+ Fe) and iron-deficient (− Fe) conditions. a Cell biomass in the analysed cultures. b TLC analysis showing TAG accumulation caused by iron deficiency. Lipids extracted from cells were resolved on TLC plates and TAG were visualized by brief iodine staining. Triolein (5 µg) was applied as a standard

Full size image

Identification and quantification of fatty acids

The fatty acids of C. reinhardtii grown under control or iron-deficient conditions were extracted, converted to fatty acid methyl esters (FAMEs) by methylation with diazomethane and then identified and quantified by GC/MS (Lewis et al. 2000) (Table 1). The FAME mass spectra (GC-EIMS) were compared with those from the lipid library (NIST-Wiley library) to identify individual fatty acids, and heptadecanoic acid (17:0) was used as an internal standard for quantification. Under conditions of iron deficiency, the cells were found to contain up to 50% fatty acids by total dry weight. The major fatty acids in these iron-depleted cells were 14:0, 16:0, 18:0, 18:2 and 18:3Δ 9, 12, 15. In addition, some essential fatty acids were identified including omega-3 and α linolenic fatty acids. Although TLC analysis of lipids showed an increased TAG content in iron-deficient cells (48–72 h) (Fig. 6b), there was no change in the total FA content. This may be due to recycling of membrane fatty acids into neutral lipid stores. FAME analysis of total fatty acids showed a marginal change in saturated FAs (e.g. 16:0 and 18:2) and interestingly levels of 18:0 were altered slightly, suggesting a distinct effect on membrane lipids. These results indicate that iron deficiency in C. reinhardtii leads to an altered FA profile and the accumulation of TAG. Essential fatty acids like omega-3, α-linolenic acid (18:3Δ9, 12, 15), linoleic acid (18:2Δ9, 12) and the omega-9 fatty acid oleic acid (18:1Δ9) were identified but their level was not altered significantly in comparison to iron-sufficient condition.

Table 1 Free fatty acid composition of C. reinhardtii cells grown under TAP + Fe for 72 h and TAP − Fe medium for 1–4 days
Full size table

The overall proportion of unsaturated fatty acids remained the same in both control (iron sufficient) and iron-depleted cells (ƩUNSAT/ƩFAME = 0.14–0.12) (Table 1). Notably, the following fatty acids were not identified in cells cultured in either condition: 16:1Δ3, 16:3Δ4,7,10, 18:3Δ6,9,12 or 18:3Δ5,12,15.

Discussion

In this study, we investigated the impact of iron deficiency on photosynthesis and lipid droplet formation in C. reinhardtii grown under photoheterotrophic conditions. Previously, we reported that the photosynthetic yield of PSI and PSII was reduced equally in C. reinhardtii subjected to iron deficiency (Yadavalli et al. 2012a; Devadasu et al. 2016). However, the damage to PSI function was more severe since PSI reaction centre assembly and electron transport require more Fe. The abundance of several proteins is altered in iron-depleted cells, especially the PsaC, D and E subunits of PSI and also the LHCI subunits. These alterations are likely to disrupt the assembly of the PSI–LHCI supercomplexes and similar changes were observed in iron-deficient rice (Yadavalli et al. 2012a, b). Moreover, electron transport and PSII supercomplex organization were also disturbed under Fe-deprived conditions (Devadasu et al. 2016). Since iron is crucial for both PSI and PSII functions, Fe deficiency was accompanied by many changes in cell structure.

In the present study, a drastic reduction in the photosynthetic yield of PSII was detected in iron-depleted C. reinhardtii cells. The yield was further decreased in a npq4 mutant lacking the activity of a light-harvesting stress-related gene (Fig. 3). The pronounced increase in Fo and decline in the I–P phase of fluorescence induction that reduced Fv/Fm and Fm indicate an impact on the electron transport chain (Fig. 2). An increase in Fo is a recognized effect of abiotic stress related to formation of inactive PSII centers or the dissociation of LHCII from the PSII core (Aro et al. 1993). Increased Fo in the dark Chl fluorescence could also result from a reduction in the PQ pool (Cournac et al. 2002). As observed in our previous study (Devadasu et al. 2016), the change in electron transport and quantum yield could indicate some alteration in reaction center proteins and PSII antenna proteins such as CP43 and 47. As expected, the second generation of Fe deficiency cells diminished the photosynthetic activity. However, the recovery of these cells has taken longer duration (6–7 days) indicating that these cells were in severe nutrient shock. Though the cells were alive, the cell division and restoration of photosynthetic activity took longer time as Fe deficiency results in multiple changes in the cell. In-depth studies are required to elucidate the mechanism that trigger the recovery process after iron starvation. As anticipated, the non-photochemical change was increased in wild-type cells grown in Fe-deficient medium and unsurprisingly there was no alteration in NPQ in the npq4 mutant (Figs. 2, 3). Notably, the expression of the LHCSR3 protein was significantly induced under Fe stress conditions (Fig. 5a). This change is in line with the finding that plants experiencing stress adapt or are protected by changes that increase heat dissipation from their photosynthetic apparatus. These changes include the increased expression of specific proteins (i.e. LHCSR3 in C. reinhardtii and PsbS in higher plants) that bind to Chl and carotenoids. Although cells subjected to iron-deficiency can generate a thylakoid lumen pH that is sufficiently acidic to induce the xanthophyll cycle, this stress decreases the capacity of photosynthetic electron transport so that it is unable to maintain a lumen pH acidic enough to induce NPQ and increase heat dissipation. It was reported earlier that energy-dependent quenching (qE) mechanism in algae is significantly different from that of higher plants due to the absence of PsbS (Bonente et al. 2011), whereas it has been recently reported that PsbS protein is transiently expressed in C. reinhardtii and is essential for the activation of non-photochemical quenching. It indicated that PsbS protein could be promoting conformational changes required for the activation of LHCSR3-dependent quenching in the antenna of PS II (Correa-Galvis et al. 2016). Hence, this protein LHCSR3 or Lhc gene compound called LHCSR3 (formerly known as LI818) is essential for qE. Of the three LhcsR isoforms present in C. reinhardtii, LHCSR3 is the only one responsive at low pH, the trigger of non-photochemical quenching. Upon acidification under high light conditions, the energy dissipation properties of this protein are increased (Bonente et al. 2011). LHCSR3 expression is characteristic of high light stress conditions, so it was surprising to find that iron deficiency also induces the production of this protein. Thus, it is apparent that the increase in non-photochemical quenching occurring in Fe stress conditions causes an increase in LHCSR3 expression that is similar to that under high light stress conditions.

Taken together, the results of this study show that carbon fixation is the preferred bioenergetic pathway in Chlamydomonas. These photoheterotrophic cells, given the option to grow heterotrophically, suppressed the expression of iron-containing photosynthetic proteins (PsaA and PsbA) in response to iron-deficient conditions, but no significant changes in the level of Rubisco large and small subunit proteins were observed.

When grown under iron-starvation conditions, C. reinhardtii cells displayed an abnormal morphology and tended to form clumps (Fig. 1). These multicellular aggregates resemble palmelloids that are observed in salt stress condition (Neelam and Subramanyam 2013). In addition, the growth rate of the iron-limited photoheterotrophic cells was severely decreased. It is interesting that although the photosynthetic efficiency decreased, lipid droplet accumulation was increased by Fe stress. Earlier reports noted increased lipid droplet formation in C. reinhardtii grown under conditions of nitrogen deficiency (Nguyen et al. 2011). During photoheterotrophic growth, the carbon is fixed into cellular biomass, which would explain why more lipid droplets are produced by these cells than by those employing photosynthesis (Ravenel 1988). The increase of lipid accumulation can be a result of imbalance between photosynthetic efficiency and utilization of assimilates under Fe deficiency and this could be one of the reasons to slow-down the biomass production. Thus, we assume that the photosynthesis seems to be not the primary reason of the lower biomass production in conditions of iron deficiency.

MLDP is a protein specific to algae that was identified in a previous study as being induced by nitrogen deficiency and associated with lipid droplets in C. reinhardtii (Huang et al. 2013). MLDP appears to have a structural role, but its specific molecular and biochemical functions remain uncharacterized. We found that MLDP expression is highly induced in response to iron deprivation, and its abundance correlates with the number and size of lipid droplets produced, supporting a role for this protein in their formation.

Chlamydomonas reinhardtii cells completely deprived of iron also contained more saturated FA. Since TAG is naturally more abundant in saturated FA, the accumulation of TAG under iron starvation conditions might account for this increase in saturated FA (Urzica et al. 2013). Iron-depleted cells showed a slight increase in the most abundant FA, C16:0 (Table 1). However, the total FA content in C. reinhardtii remained relatively stable in iron-deficient conditions, indicating that another mechanism is responsible for the elevated content of saturated FA. A shift from highly polyunsaturated species to less unsaturated FAs has been observed under several stress conditions as a protective mechanism. Polyunsaturated FAs are major components of photosynthetic membranes that are prone to oxidation under Fe stress conditions, and this could cause a loss of photosynthetic activity. A similar phenomenon was observed in C. reinhardtii subjected to different abiotic stresses (Nguyen et al. 2013). The rate of lipid oxidation might be reduced in cells containing low levels of polyunsaturated FAs, making them more tolerant of stress conditions. Acyltransferases responsible for de novo TAG biosynthesis (type 1 DGAT1 or type 2 DGTTs) in C. reinhardtii were previously identified based on homology to A. thaliana (DGAT1 and DGAT2) and yeast (ARE1, ARE2, and DGA1, respectively) enzymes (Yoon et al. 2012). PDAT1 and DGAT1 of A. thaliana are both involved in seed oil accumulation (Zhang et al. 2009). In the present study, we found that C. reinhardtii PDAT1 and DGAT2A show greatly increased expression upon iron starvation that is associated with lipid droplet formation, which strongly supports a role for these enzymes in TAG accumulation.

Iron starvation affects the expression of PDAT1 and DGAT2A in a similar manner to nitrogen starvation with the abundance of both enzymes increasing considerably after 48–72 h of iron deficiency condition (Fig. 5b). Iron deficiency-induced TAG synthesis pathway is dependent on both acyl-CoA-dependent and -independent acyltransferases (Yoon et al. 2012). Our data corroborate a previous report that PDAT1 from C. reinhardtii mediates lipid turnover by using its hydrolase activity to catabolize galactolipids and phospholipids for TAG synthesis. In cyanobacteria, Synechococcus sp. PCC7942, the thylakoids lipids which are PG, MGDG and DGDG were reduced in order to maintain the dynamics of membrane (Ivanov et al. 2007). However, little is known about the TAG accumulation during iron starvation.

Fully iron-deprived C. reinhardtii cells contain raised levels of saturated FA. Because TAG is naturally rich in 16:0 and 18:0 FA, TAG accumulation might account for the increase in saturated FA. However, our data demonstrate that the total FA content in C. reinhardtii cells remains relatively stable under iron-deficient conditions. This indicates that another mechanism is responsible for the elevated content of saturated FA, which leads to TGA accumulation (Table 1). Furthermore, the expression of both Rubisco subunits was unchanged in iron-deprived cells (Fig. 5a), which indicates that a coordinated process, dependent on the capture of light by the photosystems and fixation of CO2 via the Calvin–Benson cycle, is responsible for the increased TAG biosynthesis.

In conclusion, the results of this study confirm that iron deficiency induces the formation of lipid droplets, since the photosynthetic parameters and lipid profiles studied are similar to the control (Petroutsos et al. 2009). However, the iron-depleted cells show different photosynthetic phenotypes, as revealed by the notable reduction in Fv/Fm during the experiment indicative of a lowering of both the effective quantum yield of photosystem II (ΦPSII) and the photosynthetic capacity (ETR) when compared with the control. Taken together, the findings of this study suggest the existence of complex feedback interactions between the Rubisco enzyme and carbon fixation in C. reinhardtii and should help inform strategies for tuning carbon metabolism toward lipid biosynthesis and altered lipid composition. Importantly, the expression of light-harvesting stress protein LHCSR3 under Fe deficiency conditions indicate that non-photochemical quenching prevails in order to photo-protect. Furthermore, our results indicate that cells grown under iron-deficient conditions could be useful for biodiesel production because they show an increase in lipid content, and especially neutral lipids over polar lipids. Furthermore, the lipid composition of iron-deprived cell indicates a reduced content of unsaturated FAs, which is another desirable characteristic for biodiesel production.