Characterization of a newly isolated freshwater Eustigmatophyte alga capable of utilizing far-red light as its sole light source
- 687 Downloads
Oxygenic phototrophs typically utilize visible light (400–700 nm) to drive photosynthesis. However, a large fraction of the energy in sunlight is contained in the far-red region, which encompasses light beyond 700 nm. In nature, certain niche environments contain high levels of this far-red light due to filtering by other phototrophs, and in these environments, organisms with photosynthetic antenna systems adapted to absorbing far-red light are able to thrive. We used selective far-red light conditions to isolate such organisms in environmental samples. One cultured organism, the Eustigmatophyte alga Forest Park Isolate 5 (FP5), is able to absorb far-red light using a chlorophyll (Chl) a-containing antenna complex, and is able to grow under solely far-red light. Here we characterize the antenna system from this organism, which is able to shift the absorption of Chl a to >705 nm.
KeywordsLight harvesting complex Stramenopila Eustigmatophyte Far-red light Antenna
The spectral range that can be utilized to drive oxygenic photosynthesis in most plants, algae, and cyanobacteria is usually limited to the visible range, 400–700 nm (Chen and Blankenship 2011). While light in the visible region is abundant in unfiltered sunlight and is thus an efficient way to drive photosynthesis for most species, those organisms growing under filtered light tend to experience illumination enriched in the transmitted far-red wavelengths (beyond 700 nm). A well-characterized example of this sort of light filtering is found in ascidians that inhabit coral reefs, where their symbionts such as Prochloron, containing chlorophylls (Chls) a and b, absorb most of the incoming visible light, transmitting far-red light (FRL) (Kühl et al. 2005; Behrendt et al. 2011). In addition to marine coral reefs, such environments also exist in terrestrial and freshwater systems. Some organisms adapted to growth under these conditions possess specialized antenna systems able to absorb light in the far-red range (Kühl et al. 2005; Behrendt et al. 2011; Chen and Blankenship 2011; Kotabová et al. 2014).
Multiple strategies have been discovered by which specialized organisms can perform oxygenic photosynthesis using FRL. These include expression of antennas containing the red-shifted Chls d and f (Miyashita et al. 1996, 1997; Kühl et al. 2005; Behrendt et al. 2011; Chen et al. 2012), red-shifted phycobilisomes (Gan et al. 2014; Li et al. 2016), or transmembrane light-harvesting complexes (LHCs) that utilize the protein environment to red-shift the absorption spectrum of Chl a (Bína et al. 2014; Kotabová et al. 2014; Pazdernik 2015). Such organisms tend to be found in habitats enriched in FRL, such as would be found located below a layer of other phototrophs that absorb visible light but transmit FRL. For example, the Chl-d-containing cyanobacterium Acaryochloris marina was originally isolated from extracts of colonial ascidians (Miyashita et al. 1996, 2003). The Alveolate Chromera velia, which contains a FRL-absorbing LHC-like protein (Bína et al. 2014; Kotabová et al. 2014), was isolated as a coral symbiont off the coast of Australia (Moore et al. 2008; Tichy et al. 2013). While many of the described FRL-utilizing species are marine (Miyashita et al. 2003; Chen et al. 2012; Tichy et al. 2013), there have been several descriptions of freshwater organisms containing FRL-absorbing antennas (Pazdernik 2015).
In some marine and freshwater algae, these far-red-adapted antennas do not contain specialized pigments, but rather use the protein environment to shift the absorption properties of Chl a. For example, in C. velia, multimerization of the antenna complex results in the observed red-shift in absorption and fluorescence (Bína et al. 2014; Kotabová et al. 2014). Several species of the Chlorophyte algal genus Ostreobium contain significantly red-shifted antennas, which have been shown to perform thermally activated uphill energy transfer to drive charge separation in the special pair of photosystem (PS) II (P680) (Wilhelm and Jakob 2006). Unlike higher plants and most algae that exhibit a so-called “red drop,” where oxygen evolution falls sharply beyond a certain wavelength, this is not observed in Ostreobium; it continues its photosynthetic productivity even in the far-red range (Wilhelm and Jakob 2006).
In both higher plants and algae, integral membrane protein-pigment complexes known as LHCs capture solar energy and transfer the resulting excitation to the reaction center, with which they form a supercomplex wherein photochemistry is performed (Kühlbrandt 1994; Wientjes and Croce 2011; Wei et al. 2016). While plants have not been shown to grow under far-red illumination, their PSI antennas have some red-shifted Chl a, which is accomplished by tuning the protein-pigment interaction (Morosinotto et al. 2003). The observed red-shift can be significant, with the fluorescence of these red-shifted chlorophylls at 77 K peaking at 730 nm, but the absorption, at 700 nm, is not significantly shifted to the far-red (Morosinotto et al. 2003). It has also been demonstrated that heterodimerization of several Lhca polypeptides is responsible for the red-shifted characteristic of Chl a in these antennas (Wientjes and Croce 2011). Many niche-adapted photosynthetic microbes have absorption bands shifted much farther to the red (Wilhelm and Jakob 2006; Mohr et al. 2010; Chen et al. 2012; Bína et al. 2014; Kotabová et al. 2014; Pazdernik 2015). While far-red absorbing chlorophylls like Chls d and f can extend in vivo absorption out to nearly 750 nm (Chen and Blankenship 2011), some LHC systems that contain only Chl a are able to absorb well beyond 700 nm (Wilhelm and Jakob 2006; Bína et al. 2014; Kotabová et al. 2014; Pazdernik 2015).
Besides Ostreobium and Chromera, FRL-absorbing antennas have been found in some Stramenopiles, including diatoms (Herbstová et al. 2015) and a Eustigmatophyte (Pazdernik 2015). Eustigmatophytes are Stramenopiles that lack Chl c, containing only Chl a as well as carotenoids including violaxanthin and vaucheriaxanthin (Litvín et al. 2016). Eustigmatophytes are of particular interest to the biofuel industry because of their ability to produce large amounts of triacylglycerol (Iliev et al. 2010; Li et al. 2014). The LHCs of Eustigmatophytes are known as Viola-/Vaucheriaxanthin Chlorophyll Proteins (VCPs), as contrasted with the typical Fucoxanthin Chlorophyll Protein (FCPs) found in most Stramenopiles (Sukenik et al. 2000; Litvín et al. 2016).
The substantial diversity of these long-wavelength adaptations suggests that there are likely additional adaptations of this sort yet to be discovered. Since they are found in so many radiations of both prokaryotes and eukaryotes, we posited that they may be found in any shaded location with filtered light. We therefore adopted a broad approach to obtain as many different species as possible. By collecting samples from varied environments and growing the samples using far-red light as the sole light source, these organisms were enriched. To accomplish such selection, we constructed FRL LED growth chambers and isolated several algae and cyanobacteria capable of utilizing FRL using these selective conditions.
One particularly well-adapted isolate has been identified as a previously uncharacterized Eustigmatophyte, which we have provisionally named Eustigmatophyceae sp. FP5. To better understand how this specialized alga captures far-red light and uses it to support its growth and survival, we have characterized its antenna system spectroscopically as well as through imaging and protein analysis.
Environmental sampling and culturing
Water samples were collected in spring in a circulating water system in Forest Park, St. Louis, MO. The sample from which Eustigmatophyceae sp. FP5 was cultured originated as a water sample from one of the system’s partially shaded, slow-moving sections. Glass tubes containing 10 mL of BG-11 growth medium were inoculated with a small amount of each sample. Inoculum size was determined based on the sample’s apparent turbidity. The tubes were placed into a growth chamber dimly illuminated using far-red (740 nm) LEDs (Fig. S1) and maintained at room temperature. When green growth appeared, the organisms within were analyzed by fluorescence and absorption spectroscopy, and shaking flasks were inoculated from the tubes and maintained under FRL.
To purify the culture and obtain a clonal culture, FP5 was streaked on 0.4% Agarose BG-11 plates and grown under FRL (Shirai et al. 1989; Ferris and Hirsch 1991), and streaked again onto a similar plate to obtain single colonies. Wells of a 12-well culture plate containing BG-11 were each inoculated with a single colony, and these were grown under FRL until a dense green growth appeared, at which time some wells were used to inoculate larger cultures. The resulting clonal culture was scaled up in large (1–2 L) flasks, and light was provided by 740 nm LEDs at a power density of 2.3 mW cm−2. Light intensity was measured using a Thorlabs PM200 power meter with a S120VC Si-UV Photodiode sensor. Flasks were either gently shaken on an orbital shaker or stirred using a stir bar and bubbled with air sterilized using a Whatman Hepa-Vent filter. Cultures were regularly monitored spectrally and by microscopy to ascertain their quality. While FP5 is capable of growth under white light, we have focused on the properties of the organism when grown under the far-red light condition described above to ascertain its far-red light absorbing characteristics.
Pigment analysis was performed using an Agilent 1100 HPLC with a C18 resin as the stationary phase and pure methanol as the mobile phase. Absorption was measured using an in-line diode array detector. The flow rate was 1.5 mL/min and 100 µL sample injections were used. The cell walls of FP5 are recalcitrant, presumably due to the presence of algaenan found in other Eustigmatophytes (Scholz et al. 2014), and resist the extraction of pigments using methanol treatment, so the cells were disrupted prior to extraction of pigment. 1 mL of a culture started from a single colony and grown under FRL was rinsed with and exchanged into Buffer A (50 mM MES, 2 mM KCl, pH 6.5), and 1.4 mm stainless steel beads were added to the culture in a screw-top vial to half the culture volume. The cells were then disrupted by five passes on a Biospec Mini Bead-Beater running at full speed for 20 s. Cells were cooled between runs using an ice bath.
The broken cells were then subjected to pigment extraction: 0.5 mL broken cell suspension, 2 mL 7:2 acetone/methanol, 200 µL 4 M NaCl, and 3 mL petroleum ether were mixed vigorously and allowed to separate. The ether layer was removed and dried using an Eppendorf Vacufuge vacuum centrifuge, then resuspended in HPLC grade methanol immediately prior to injection on the HPLC.
The 18S ribosomal DNA sequence of FP5 was amplified using polymerase chain reaction (PCR). Whole cells provided template for the reaction: 1 µL of a thick cell suspension was pipetted into the PCR reaction containing GoTaq Green master mix. A 5 min 94 °C denaturation at the beginning of the PCR reaction was used to disrupt cells. The forward primer was an algal-specific primer (5′-CGGTAATTCCAGCTCC-3′) and the reverse was a universal primer (5′-GGGCGGTGTGTACAARGRG-3′). For sequencing, forward and reverse sequences were generated and aligned to produce a longer sequence. The amplicon was sequenced through the commercial vendor GeneWiz, and compared with other algal 18S rDNA sequences using the NCBI Basic Local Alignment Search Tool (BLAST) (Altschul et al. 1990). Sequences were aligned using MUSCLE in MEGA6 (Tamura et al. 2013), and excess base pairs that were longer than the FP5 sequence were removed. Using these sequences, a neighbor-joining phylogenetic tree was created using MEGA6 with 500 bootstrap replicates (Tamura et al. 2013).
Phase contrast light microscopy images were taken on a Nikon Labophot 2 light microscope using an oil-immersion 100× objective or on a Wild Heerbrugg light microscope using a 40× objective with a 3× adjustable second lens. Scale bars were added using a stage micrometer and ImageJ software (Schindelin et al. 2015). Confocal images were taken using a PerkinElmer Ultraview Vox spinning disc confocal system on a Zeiss Observer Z1 microscope using a Zeiss Plan Apo 100 × 1.46 oil-immersion objective. 0.1 uM Z-slices were flattened and images processed using FIJI software (Schindelin et al. 2012). For deep-etch freeze-fracture cryogenic electron microscopy, samples were prepared and imaged as described previously (Weiss et al. 2012), with the following modification: FRL-grown cells were first pelleted and 5 µL of this material was applied to a cushioning material before flash freezing.
The clonal FP5 culture was frozen by free air convection at −80 °C in Buffer A, which is composed of 50 mM MES, 2 mM KCl, pH 6.5. The light-harvesting complex was purified as follows, modified from previously described methods (Tichy et al. 2013; Bína et al. 2014; Kotabová et al. 2014; Pazdernik 2015). FP5 was grown under FRL at room temperature and harvested by centrifugation. The cell pellet was washed once in Buffer A. Approximately 1–2 g of the wet cell pellet were resuspended in 30 mL of Buffer A and either frozen at −80 °C or used immediately. Following thawing of cells, if frozen, and addition of DNAse (Sigma-Aldrich), cells were disrupted via three passes through a French Press Pressure Cell at 20,000 PSI. Unbroken cells and cellular debris were separated by centrifugation at 1000×g for 5 min in a Thermo Scientific Legend RT+ centrifuge with a Thermo-Heraeus 75006445 swinging-bucket rotor. The supernatant was then subjected to centrifugation at 45,000×g for 2 h at 4 °C in a Beckman Coulter Optima L-80 XP Ultracentrifuge using a Type 45 Ti fixed-angle rotor.
The membrane pellet was gently resuspended in 1–2 mL of Buffer A using a paint brush and the Chl a concentration was measured. For this, 10 µL of the cell suspension was added to 990 µL pure methanol. These were mixed and the insoluble portions were separated using an Eppendorf 5415 D benchtop centrifuge at full speed for 1 min. The Chl a concentration, in mg/mL, was calculated from the Abs665 value. The suspension was diluted to a Chl a concentration of 0.5 mg/mL using Buffer A. At this point, membranes were either frozen at −80 °C or solubilized immediately. The material was solubilized by adding a stock of 10% (w/v) n-Dodecyl β-d-maltoside (DDM) to a final concentration of 1%. After 1 h on ice while shaking in the dark, insoluble material was immediately removed by centrifugation at 30,000×g for 20 min at 4 °C in a Sorvall Evolution or RC5 centrifuge using a SS34 rotor, and the supernatant stored until subjected to gel filtration. Sample storage was optimal at 4 °C in 25–50% glycerol (Fig. S2), but storage overnight at 4 °C without glycerol was also used with very little sample degradation.
Gel filtration chromatography
Gel filtration was performed at 4 °C under dim green light using Sephacryl S-300 resin (GE Healthcare Life Sciences) at a 1 mL/min flow rate using Buffer A with 0.02% DDM. The column was 30 cm long with a 1.5 cm internal diameter. One millilitre fractions were collected and fluorescence spectra were recorded using a Varian Cary Eclipse fluorimeter. For analysis through CN-PAGE, all fractions containing significant FR-LHC were pooled and concentrated using a 30 kDa MWCO Millipore centrifugal filter unit.
Clear native polyacrylamide gel electrophoresis (CN-PAGE)
Preparative CN-PAGE gels were cast using the Sturdier setup (Hoefer) with a 1.5 mm thickness, and analytical gels for 2D electrophoresis were cast using a Bio-Rad mini-gel system. The gel and buffer compositions were modified from Schägger and von Jagow (1991). Coomassie was omitted from the cathode buffer and instead 0.05% sodium deoxycholate was added (Litvín et al. 2016). To increase the separation of the protein complexes of interest, the density of the gel was 9–16% polyacrylamide. Gradients were made using an acrylic gradient maker and a Dynamax peristaltic pump running at 5 mL/min for the Sturdier setup or 1 mL/min for the Bio-Rad system. The preparative gel was run at 80 V overnight at 4 °C in the dark, and the analytical was run for 3 h under the same conditions. For imaging, the gel was placed between two transparent sheets of plastic and was imaged using a visible light Epson Perfection V30 scanner and a Syngene PXi fluorescence imager. A blue LED excitation and a 705 nm emission filter were used to observe far-red fluorescence of the bands. When required, a Novex Nativemark native protein standard was included in the gel and was stained using 0.2% Coomassie G-250 in 40% methanol and 7% acetic acid.
2D polyacrylamide gel electrophoresis (2D PAGE)
A 12% SDS-PAGE gel was poured in the Bio-Rad system using the pre-mixed Protogel (National Diagnostics) reagent set. A modified comb allowing one large well for a CN-PAGE gel slice and one small well for the ladder was used to form the stacking gel. The desired lane was carefully excised from the CN-PAGE gel and denatured in a solution of 1% β-mercaptoethanol (βME) and 1% Sodium dodecyl sulfate (SDS) for 1 h (Wittig et al. 2006). The gel slice was then inserted into the large well on the SDS-PAGE gel. The gel was silver stained using a Pierce Sliver Stain Kit (Thermo Scientific) and imaged as described above.
Elution of native proteins from native gel
Native proteins were extracted through simple diffusion using a modification of a previously described method (Wittig et al. 2006). Bands were cut from the gel and diced using a razor blade into fragments approximately one cubic mm in size. The fragments were placed into a 1 Dram glass vial and enough Buffer A with 25% glycerol (v/v) and 0.02% DDM was added to double the volume of the gel slices. The tubes were wrapped in foil and placed on a rotator at 4 °C for 2 days. The green eluent was pipetted off and centrifuged at 4 °C in an Eppendorf 5415 D benchtop centrifuge at full speed for 5 min to remove gel debris. The supernatant was used for subsequent spectroscopic analysis.
Fluorescence and absorption spectroscopy
Absorption measurements were taken using a Shimadzu UV1800 spectrophotometer, which was equipped with a Janis Research Company VNF-100 optical liquid nitrogen cryostat for 77 K measurements. Fluorescence emission spectra were taken using a Shimadzu RF-6000 or Varian Cary Eclipse fluorimeter. For 77 K fluorescence measurements, an optical dewar filled with liquid nitrogen was used. For all 77 K spectroscopic analysis, sample material was mixed with glycerol to a concentration of ~60% (v/v).
For spectral analysis of native gel bands, bands were cut from the gel using a razor blade. For absorption measurements, each slice was placed on the side of the cuvette such that it did not move before placing the cuvette in the instrument. An area of the gel within an unused lane was cut out and used as a blank. For fluorescence measurements, gel slices were placed in a 1 cm square plastic cuvette with Buffer A containing 0.02% DDM. The slices were placed diagonally in the fluorimeter cuvette such that the edge of the gel faced the excitation light and the wide edge faced the detector. Fluorescence spectra were taken using an excitation wavelength of 435 nm with excitation and emission slit widths of 5 nm.
Eluted pigment-protein complexes were subject to room temperature and 77 K measurements of absorption and fluorescence with excitation at 435 nm and excitation and emission bandwidths of 5.0 nm. Additionally, thermally activated energy transfer was analyzed using fluorescence excitation at 750 nm with an excitation bandwidth of 1.5 nm and an emission bandwidth of 5.0 nm. For fluorescence measurements of eluted proteins, the OD was maintained below 0.1. For 77 K and thermal activation fluorimetry, a blank sample was subtracted from the raw data to correct for dewar effects and excitation light scattering, respectively.
Spectra were analyzed using Origin 2016 software. Spectra were normalized to a range of 0–1 for uniformity. Prior to calculation of the second derivative, spectra were normalized to unity at the maximum within the region shown and a Savitzky–Golay smoothing algorithm was applied. The same normalization was applied to spectra before the difference was calculated. This algorithm was also applied to smooth 77 K absorption and thermally activated fluorescence spectra.
Whole cell spectroscopy was performed using a culture grown under 740 nm LED light under stirred, aerated conditions. To disrupt cell clumps, 1 mL of culture was resuspended by shaking and was then subjected to 10 s of sonication using a Fischer Sonic Dismembranator Model 300 fitted with a micro probe at 35% power. To reduce scattering of light, the cell suspension was diluted with a sucrose solution to a final concentration of approximately 60% sucrose for room temperature measurements of absorption and fluorescence. For fluorescence, an excitation wavelength of 435 nm was used with excitation and emission slit widths at 5.0 nm. For 77 K fluorescence measurements, the cell suspension was diluted with a glycerol solution to a final concentration of approximately 60% glycerol and frozen using an optical dewar containing liquid nitrogen. Absorption measurements were taken using a Shimadzu UV1800 and fluorescence measurements were taken using a Varian Cary Eclipse fluorimeter.
Isolation and identification of Eustigmatophyceae sp. FP5
Environmental isolates were left in the FRL chamber for several months and periodically checked. Culture Forest Park #5 (FP5) was found to contain an antenna system with substantial FRL absorption, peaking at 705 nm (Fig. 1) as measured by fluorescence and absorption spectroscopy. Since the only chlorophyll present is Chl a (see below), this additional peak cannot be explained by the presence of any far-red chlorophyll such as Chls d or f. Room temperature fluorescence spectroscopy of whole cells indicated that a single fluorescence maximum is present at 709 nm (Fig. 1b), which represents emission from the red-shifted antenna. At 77 K, this maximum shifts to 718 nm in whole cells (Fig. 1b).
Phase-contrast light microscopy revealed spherical non-motile cells with a morphology similar to the Eustigmatophyte Vischeria (Gartner et al. 2012) (Fig. 2a, b), ranging in size from 5 µm to (rarely) greater than 20 µm. A thick cell wall is visible. The cells appear to be primarily occupied by green photosynthetic material, but in many cells, a punctate red organelle is present (Fig. 2a, b arrows), also seen in other Eustigmatophytes and considered a hallmark trait of the class (Hibberd and Leedale 1972; Gartner et al. 2012).
Fluorescence confocal microscopy (Fig. 2c, d) revealed that most of the cell volume is taken up by a large, irregular chloroplast that fluoresces in the far-red region of the spectrum (shown red in these false-color images). A small point of red fluorescence was observed in the middle of each cell (shown green in these false-color images) when a green excitation light was used, presumably corresponding to the red punctae observed by phase microscopy.
Quick-freeze deep-etch electron microscopy (Fig. 2e–g) revealed a massive chloroplast that comprised a large portion of the cell’s volume. The tightly appressed thylakoids form laterally stacked groups, also observed in other Eustigmatophytes such as several Vischeria species (Santos 1996; Gartner et al. 2012). The cells also contain a central organelle containing a stack of plates (Fig. 2e, enlarged in Fig. S3), which we propose corresponds to the red/fluorescent body seen in the phase/confocal images.
18 S rDNA amplification produced a usable sequence of 1008 base pairs. When these same primers were compared in silico to the 18S ribosomal RNA gene of a model Eustigmatophyte, Nannochloropsis gaditana strain CCMP526, the sequence length was 1075 base pairs of the total 1796 base pairs in the gene for a coverage of 60% of the total sequence. Phylogenetic analysis revealed that Isolate FP5 is a Eustigmatophyte alga, in the Stramenopile radiation, most closely related to several other cultured species in the Eustigmatophyceae genus. Using this analysis, FP5 has been provisionally placed into this genus (Fig. 3).
While FP5 is capable of growing under FRL using wavelengths typically not absorbed by Chl a, reverse phase HPLC of extracted FP5 pigments revealed that the organism contains only Chl a and various unidentified carotenoids (Fig. 4a, b), implicating pigment-protein interactions as the primary cause of the red shift observed in the absorption of the light-harvesting complexes of FP5. Four carotenoid peaks were observed on the HPLC chromatogram and only a single chlorophyll peak with a spectrum corresponding to Chl a (Fig. 4a, c). The time range shown encompasses all peaks observed, and extension of the run to 1 h did not reveal any additional carotenoid peaks.
Gel filtration chromatography
Gel filtration chromatography of solubilized membrane material generated a profile of protein complexes, and these were monitored by fluorescence excitation. Highly separated bands do not form, as was also the case with sucrose gradient centrifugation (data not shown), but several distinct zones were detected. The longest wavelength portions elute from the column first and the shorter wavelength portions elute last (Fig. S4A-B). The FR-LHC purified by gel filtration fluoresces at approximately 705 nm at room temperature when excited at 435 nm (Fig. S4D). The pigment-protein complex is unstable and breaks down over time, and when breakdown occurred, the fluorescence emission was blue-shifted to 697 nm.
Clear native polyacrylamide gel electrophoresis
FR-LHC fractions from gel filtration were subjected to clear native gel electrophoresis (native PAGE). Five bands, B1-B5, were visible and well separated, with two, B2 and B3, representing the most far-red fluorescing complexes (Fig. 5a). When stained with Coomassie and compared with a native protein mass standard, approximate molecular weights were: B1: 300 kDa, B2: 200 kDa, B3: 120 kDa, B4: 110 kDa, and B5: 50 kDa (Fig. 5a). Two additional low molecular weight bands were also stained, but were not observed on the visible or fluorescence images. These bands, at approximately 20 kDa, are putatively considered to be antenna monomers. Fluorescence and absorption spectra of the native PAGE bands showed that the FR-LHC complex contains two forms: one with primarily red-shifted pigments and a second with less red-shifted forms. B2 and B3 exhibit more far-red absorption than B4 and B5, shown by the difference spectrum between B3 and B5 (Fig. 5c), and increases in their red-shifted fluorescence emission characteristics (Fig. 5e).
Comparing the second derivative of absorbance for B3 and B5, the positions of the negative peaks are significantly different (Fig. 5d). Negative second derivative peaks indicate positions of individual absorbing components within a complex spectrum. In B3, a significant negative peak is observed at 703 nm, compared to the 693 nm negative peak of B5. This indicates the presence of a species absorbing at 703 nm in B3. Three components were visible in the absorption spectrum at room temperature (Fig. 5b, c). The absorption difference spectrum showed that the difference between the larger far-red absorbing complex and the smaller red-absorbing bands is a sharp peak at 706 nm, representing the FR-LHC (Fig. 5c). This was seen as a far-red shoulder on the second peak of the Q y band of the absorption spectra (Fig. 5c).
A 683 nm component in the fluorescence emission spectra of B1 was observed in addition to a far-red peak, while B2 and B3 contain only a single emission peak at 709 nm, which is consistent with the fluorescence emission of whole cells. B4 and B5 had fluorescence maxima near 700 nm (Fig. 5e).
2D polyacrylamide gel electrophoresis
A second, denaturing dimension was run from a vertical slice of a native gel. The first dimension was the non-denaturing CN-PAGE, which maintained the pigment-protein complexes in their native state. A vertical slice of the gel was excised, chemically denatured with SDS and reduced using βME, and run in the second, denaturing dimension along with a molecular mass ladder.
Regardless of the spectroscopic properties of the four native gel bands, all appeared to have similar polypeptide compositions: two major bands migrating as 18 and 21 kDa species, with the addition of two weaker, lower molecular weight bands (Fig. 6). These same two major bands also appear to be the primary constituents of whole solubilized membrane material as well as the various gel filtration fractions (Fig. S5). The 18 kDa polypeptide appeared to be more abundant than the 21 kDa polypeptide in the native complexes. A faint, high molecular weight bands (B1 and B2) on the native gel additionally contained 30 and 40 kDa polypeptides, which likely correspond to components of photosystem II, which we would expect to see complexed in higher oligomers with our purification protocol.
Purification of protein complexes using preparative CN-PAGE
Protein-pigment complexes were eluted from each slice of the CN-PAGE gel, yielding several protein complexes, including forms of the FR-LHC (B2 and B3), as measured by absorption and fluorescence emission spectroscopy (Fig. 7). The 77 K absorption spectrum of the eluted proteins showed several features in the Q y region (Fig. 7a). Comparing B3 and B5 by their difference spectrum, the far-red and red antenna components are enriched in B3 (Fig. 7b). Comparing the second derivative spectra of these two bands, five negative peaks are present in B3 at 660, 668, 682, 693, and 709 nm, while negative peaks in B5 are observed at 671 and 697 nm (Fig. 7c). These peaks indicate that the FR-LHC complex in B3 includes pigments absorbing light at far-red wavelengths, while other, lower molecular weight antennas also present in the FP5 light-harvesting system notably lack pigments absorbing in the far-red. The farthest red-shifted negative peak in B3 is located at 709 nm, which likely corresponds to the most red-shifted form of Chl a in the FR-LHC.
The room temperature fluorescence maximum of B3 after elution was at 707 nm (Fig. S6), but when the sample was cooled to 77 K, this maximum shifted to 716 nm (Fig. 7d), consistent with measurements taken in vivo, which also show a 718 nm fluorescence peak at 77 K (Fig. 1b). However, other bands lacking far-red fluorescence at room temperature in the eluted material (Fig. S6) or in-gel form (Fig. 5e) still maintained this far-red fluorescence at 77 K. While B2 and B3 were made up of almost entirely a single component at 716 nm, B1, B4, and B5 contained notable features in the red region in addition to this far-red component (Fig. 7d).
Eustigmatophyceae sp. FP5 is able to utilize the normally unavailable wavelengths of light in the far-red region using only Chl a and integral membrane proteins. The red-shift of Chl a observed in FP5 appears to be due to the protein environment surrounding the pigments. Such a protein-induced red-shift of Chl a has also been seen in several different Stramenopiles and other algal clades (Wilhelm and Jakob 2006; Tichy et al. 2013; Bína et al. 2014; Kotabová et al. 2014), and protein-environment-induced red-shifts that lead to the existence of a small amount of far-red absorbing chlorophyll have been reported in the antennas of plant PSI (Wientjes and Croce 2011).
In plants, LHCII, associated with PSII, is found as a trimer in which each monomer contains eight Chl a, six Chl b, and various carotenoids (Liu et al. 2004). LHCI, associated with PSI, which also binds Chls a and b as well as carotenoids, forms dimers rather than trimers (Wientjes and Croce 2011). FP5 does not contain any Chl b, and it is likely that that the 18 and 21 kDa antenna polypeptides making up its FR-LHC are similar to the light harvesting proteins found in the vaucherioxanthin-chlorophyll proteins, (VCP) of the related genus Nannochloropsis, which are themselves similar to the FCP proteins in most diatoms (Litvín et al. 2016). While structural modeling, sequence alignment, and pigment analysis have been done on FCP, no crystal structure is available (Durnford et al. 1996; Gundermann and Büchel 2014; Llansola-Portoles et al. 2016). Phylogenetic analysis of the 18S rDNA sequence, pigment composition, and polypeptide sizes suggest that the FP5 antenna may be similar to the VCPs found in Nannochloropsis, which have not been reported to utilize FRL to support growth of the organism (Litvín et al. 2016).
The far-red absorption phenotype of another protein-pigment complex, the red form of the LHC in the Alveolate C. velia, has been shown to require multimerization to exhibit a red-shift (Bína et al. 2014). This was observed in FP5 as well (Fig. S2). FP5 contains several low molecular weight monomers, 18 and 21 kDa, which are similar to the FCP-like proteins found in C. velia (Tichy et al. 2013). Trimers of some forms of FCP have been shown to form higher order oligomers as well. In the diatom Cyclotella meneghiniana, FCPs have been shown to form both trimers made up of 18 and 19 kDa subunits and oligomers that are made up of trimers of 19 kDa subunit only (Büchel 2003). While C. meneghiniana FCPs were not shown to exhibit far-red absorption characteristics, multimerization has also been shown to be important to the red-shifting of Chl a absorption in other species (Bína et al. 2014; Kotabová et al. 2014), indicating that it may also be at play in the FP5 system.
FP5 is capable of growing under pure far-red light (Fig. S1B), indicating that it is able to utilize long wavelength light to drive charge separation at both PSI and PSII. Typically, LHCs absorb light at a shorter wavelength than the absorption of the special pair of chlorophylls that drives photochemistry, allowing downhill energy transfer from LHC pigments to the RC pigments. However, exceptions to this rule have been reported in some of the LHCs associated with PSI (Wientjes and Croce 2011): red-shifted chlorophylls in the LHCs associated with PSI, called Lhcas, were found to absorb wavelengths beyond 700 nm, which is the absorption of the special-pair chlorophylls of PSI (P700) (Wientjes and Croce 2011). Several functions were proposed, the most interesting being the possibility that these far-red forms allow leaves to absorb far-red light in shaded environments, where FRL is enriched, and transfer it to the RC using thermally activated energy transfer (Trissl 1993; Rivadossi et al. 1999; Wientjes and Croce 2011). The far-red forms have been shown to account for up to 40% of the total photons absorbed under shade conditions (Rivadossi et al. 1999). This process of increasing the absorption cross-section of a leaf allows for increased utilization of this low-energy light (Trissl 1993; Rivadossi et al. 1999). However, these chlorophylls only form a small portion of the overall antenna pigments (Rivadossi et al. 1999).
Thermally activated energy transfer in the isolated FP5 FR-LHC was demonstrated by measuring fluorescence emission upon long-wavelength excitation that was significantly red-shifted compared to the emission band (Fig. S7). Thermal activation has also been observed in Ostreobium and C. velia, although these measurements were taken using whole cells and not purified complexes (Wilhelm and Jakob 2006; Kotabová et al. 2014).
Because leaves also transmit FRL, it is likely that if such an adaptation were to be engineered into crop plants, this wasted energy could be utilized in leaves below the canopy layer (Ort et al. 2015). It is theoretically possible to absorb 19% more photons in this range that contain enough energy to drive oxygenic photosynthesis (Chen and Blankenship 2011; Blankenship and Chen 2013). It may be possible to modify the response of plants to far-red light hitting the lower leaves to optimize their efficiency (Ort et al. 2015). While plants already do this to some extent (Pettai et al. 2005a, b), the far-red absorption cross-section is severely limited compared to FP5. By understanding how these far-red antenna systems work, we hope to provide a basis for the future transformation of these systems into crop plants to improve the efficiency of photosynthesis.
The newly isolated Eustigmatophyte alga Eustigmatophyceae sp. FP5 contains a specialized photosynthetic antenna system that extends the range of light absorption into the far-red region. This adaptation allows FP5 to grow under a solely far-red light source. We have shown that this antenna system is made up of a complex of several antenna monomers, which must interact to facilitate this far-red light absorption phenotype. The antenna contains only Chl a and several carotenoids, and therefore we conclude that the FP5 antenna system uses the protein environment to shift the absorption of Chl a to the far-red.
The ability of some specialized organisms to absorb far-red light to drive oxygenic photosynthesis is challenging the notion that energies lower than that absorbed by plants are not capable of sustaining oxygenic photosynthesis. Indeed, these longer wavelengths, typically wasted in many ecosystems, are a potentially ignored source of photosynthetic energy. The antenna system of the newly isolated Eustigmatophyte Eustigmatophyceae sp. FP5 demonstrates the ability of Chl a antennas to absorb these useful wavelengths. Understanding the function of these antennas may pave the way for modifications in plants that allow these useful, but typically wasted, wavelengths to be used in driving photosynthesis.
Funding for this work was from the Photosynthetic Antenna Research Center (PARC). PARC is a Department of Energy (DOE) Energy Frontier Research Center (EFRC) funded by Grant #DE-SC 0001035. Benjamin Wolf was supported by the William H. Danforth Plant Science Fellowship. Confocal microscopy was performed by Zuzana Kocsisova (Division of Biology and Biomedical Sciences, Washington University in St. Louis). We also acknowledge Jeremy D. King (Department of Biology, Washington University in St. Louis) for his contributions to the original sampling protocols and helpful discussions and Gregory S. Orf (Department of Chemistry, Washington University in St. Louis) for instruction on fluorimetry.
- Durnford DG, Aebersold R, Green BR (1996) The fucoxanthin-chlorophyll proteins from a chromophyte alga are part of a large multigene family: structural and evolutionary relationships to other light harvesting antennae. Mol Gen Genet 253:377–386. doi: 10.1007/s004380050334 CrossRefPubMedGoogle Scholar
- Iliev I, Petkov G, Lukavsky J et al (2010) The alga trachydiscus minutus (Pseudostaurastrum Minutum): growth and composition. Gen Appl Plant Physiol 36:222–231Google Scholar
- Pazdernik M (2015) Light harvesting complexes and chromatic adaptation of Eustigmatophyte alga Trachydiscus minutus. Master’s Thesis, University of South BohemiaGoogle Scholar
- Santos LMA (1996) The Eustigmatophyceae: actual knowledge and research perspectives. Nova Hedwig Beih 112:391–406Google Scholar