1 Introduction
Select unicellular microalgae can utilize different trophic modes for metabolism and can switch between these modes in response to changing environmental conditions. Growth characteristics of microalgae have been described under photoautotrophic, chemoheterotrophic, and mixotrophic modes, particularly in the context of trying to improve growth rates in industrial applications. Mixotrophy in microalgae has often been described as a synergistic combination of autotrophy and heterotrophy and has been reported to enhance growth rates and biomass accumulation when compared to either trophic mode (Smith et al., 2015; Zhang et al., 2017; Abiusi et al., 2020). The hypothesis for this is that common hinderances experienced in strictly heterotrophic or photoautotrophic cultivations can be overcome by leveraging two energy sources within one cell (Wang et al., 2014; Abiusi et al., 2020). For example, in photoautotrophic cultivation, light availability can be the main growth limiting factor, but with supplementation of an organic carbon source this can be alleviated (Wang et al., 2014; Abiusi et al., 2020).
We have previously investigated mixotrophy of microalgal strains by adding raw, non-food, plant material to microalgal cultures as carbon source (Vogler et al., 2018; Schambach et al., 2020). Not only is this lignocellulosic biomass the most abundant biomaterial in the biosphere (Malhi, 2002), it is inexpensive compared to substrates like glucose, and recalcitrant to bacterial contamination in microalgal cultures; important aspects when considering industrial applicability. Plant substrate utilization has been observed in the freshwater green alga Auxenochlorella protothecoides and in two species of Nannochloropsis (Vogler et al., 2018; Schambach et al., 2020). These strains exhibited increased specific growth rates and biomass accumulation in the presence of switchgrass or corn stover. Glycome profiling of switchgrass after addition to A. protothecoides cultures compared to switchgrass in media alone demonstrated the utilization of the hemicellulose xyloglucan by the alga (Vogler et al., 2018). Moreover, differential expression analysis revealed enzymes likely involved in degradation of lignocellulose, including family 5 and 9 glycosyl hydrolases which catalyze hydrolysis of glycosidic linkages in cellulose (Vogler et al., 2018). SEM analysis and sugar analysis of corn stover revealed signs of phloem degradation and glucan utilization, respectively, by Nannochloropsis gaditana (Schambach et al., 2020).
Here, we examined potential mixotrophic growth of S. obliquus in the presence of plant substrate. Scenedesmus obliquus, a freshwater chlorophyte known for its robust overall growth, has been reported to grow mixotrophically using commonly investigated carbon sources like glucose or acetate (Mandal and Mallick, 2009; Shen et al., 2018; Song et al., 2021). This strain’s high tolerance to a wide range of environmental conditions suggests it is metabolically flexible (Msanne et al., 2020), making it an ideal strain to investigate utilization of lignocellulosic biomass. Through transcriptomic analysis we aimed to describe the molecular mechanisms, particularly regarding glycosyl hydrolase expression, employed by this strain to potentially degrade and utilize switchgrass. We also examine the overarching energy metabolism of S. obliquus based on differential expression of enzymes involved in pathways like oxidative phosphorylation, carbon fixation, and photosynthesis. In addition, we investigated the effects of increased CO2 concentration and scaling up to 50 L volume in mini-raceway ponds on plant substrate utilization.
2 Materials and methods
2.1 Cultivation conditions and initial plant substrate screen
Scenedesmus obliquus DOE0152z (S. obliquus) was obtained from Heliae Development, LLC. Stock cultures were grown in BG-11 media and maintained in exponential phase growth prior to the start of each experiment. The following BG-11 media composition was used: 17.6 mM NaNO3; 0.23 mM K2HPO4; 0.3 mM MgSO4∙7H2O; 0.24 mM CaCl2∙2H2O; 0.031 mM Citric Acid∙H2O; 0.021 mM Ferric Ammonium Citrate; 0.0027 mM Na2EDTA2∙H2O; 0.19 mM Na2CO3; 46 mM H3BO3; 9 mM MnCl2∙4H2O; 0.77 mM ZnSO4; 1.6 mM Na2MoO4; 0.3 mM CuSO4∙5H2O; 0.17 mM Co(NO3)2∙6H2O. Media was filter sterilized using 0.2 µm filters.
For all flask experiments the following conditions were used unless otherwise stated. S. obliquus cultures were grown in a Percival growth chamber (Percival Scientific, Perry, IA) in 100 mL volumes in 250 mL Erlenmeyer flasks and mixed via an orbital shaker (Benchmark Scientific, Sayreville, NJ) at 120 rpm. The light intensity was 100 µmol m-2 s-1 with a 12:12 h light/dark cycle. The temperature was maintained at 25°C. For each experiment three biological replicates were used for each treatment. The cell density of each inoculum was approximately 1-2 x 106 cells mL-1. Growth was monitored by manual cell counts of free-floating cells using a hemocytometer. Note this allowed for differentiation between microalgal cells and microscopic pieces of plant substrate. Cultures were grown until they reached early stationary phase (i.e., when cell counts began to plateau). Plant substrates were separated from microalgal cells using a 40µm Corning nylon cell strainer. Specific growth rates were calculated for log phase of growth using Equation 1:
μ
=
In
(
N
2
/
N
1
)
/
(
t
2
/
t
1
)
where N1 and N2 are the cell counts (in cells mL-1) at time 1 (t1) and time 2 (t2), respectively. To determine significant differences between growth rates we employed analysis of variance and least-squares means for multiple comparisons tests with a Tukey p-value adjustment. Data was plotted using R: A Language and Environment for Statistical Computing (https://www.R-project.org).
S. obliquus was inoculated into flasks with corn stover (CS; Zea mays), sugarcane bagasse (SB; Sacharum officinarum), switchgrass (SG; Panicum virgatum), and yard waste (YW) at a concentration of 0.1% weight by volume (w/v) as an initial screen for plant substrate utilization by S. obliquus compared to a control of the microalga alone. Specific growth rates were calculated from day 3 to 12. CS, SB, and SG were provided courtesy of Idaho National Laboratory (INL) Bioenergy Feedstock Library. CS comprises the stalks, leaves, leaf sheaths, tassels, cobs, and cob stems of the corn plant that are left on the field after the corn is harvested. CS was collected by multi-pass harvesting and ground to pass through a 2-in. sieve using a Vermeer BG480 grinder followed by a 1-in. sieve using a Bliss Hammermill (INL Bioenergy Feedstock Library). SB was processed to pass through a 6-in. screen (INL Bioenergy Feedstock Library). SG was ground to pass through a 1-in. sieve using a Vemeer BG480 grinder (INL Bioenergy Feedstock Library). YW was composed mostly of fresh monoculture grass clippings from an untreated lawn in Los Alamos, NM. These substrates were used “raw”, i.e., added to the flask in the format provided by INL (e.g., dried and milled) with no pretreatment other than autoclaving at 15 psi and 121°C for 20 minutes. Based on the results from this experiment reported below, SG was chosen as the primary substrate for further experimentation. We investigated the effect of plant substrate concentration on S. obliquus growth by comparing three concentrations of SG: 0.2%, 0.3%, and 0.4% w/v. Specific growth rates were calculated from day 1 to 12.
To investigate the molecular mechanisms as well as potential effects elevated CO2 may have on substrate utilization, cultures were first grown with and without 0.2% SG in ambient air (0.04% CO2) in the growth chamber. Then, the same stock culture was acclimated to 1% CO2 and the experiment was repeated. The cell density for each inoculum was approximately 5 x 106 cells mL-1. All other cultivation conditions were the same as described above. Specific growth rates were calculated from day 0 to 31 for ambient CO2, and day 0 to 19 for 1% CO2. To determine significant differences between cell density with switchgrass versus without at each timepoint we used the Welch two sample t-test. Samples of microalgal biomass were taken for transcriptomic analysis on day 14 and 35 in air, and day 7 and 21 in 1% CO2 representing mid- and late-log phase growth, respectively. These samples contained 4-9 x 108 cells mL-1, and were immediately pelleted, flash frozen in liquid nitrogen, and stored at -80°C until RNA could be extracted.
2.2 Pond cultivation
To determine how results at laboratory scale translate to a large volume cultivation, S. obliquus was grown with and without 0.2% w/v SG at 50 L volume in 100 L mini-raceway ponds (MicroBio Engineering, San Luis Obispo, CA) in a greenhouse. We had a total of six ponds for three biological replicates each. Each pond had a cylindrical bubble diffuser and was constantly aerated with ambient air (0.04% CO2) at 400 L min-1. CO2 was supplied on demand when pH exceeded 8.1 and turned off when pH fell below 8.0. The paddlewheel speed was 40 rpm. SG and media components for pond cultivation were not autoclaved. The greenhouse had automated light and temperature control. Greenhouse lights would turn on when outdoor light intensity fell below 300 W m-2 between the hours of 6 am and 9 pm. Samples were taken for cell counts, dry weight, and 16S amplicon sequencing. Specific growth rates were calculated from day 1 to 6.
2.3 Light and scanning electron microscopy
To observe physical degradation of the plant substrate when the alga was present, stem sections of SG and corn, at a total concentration of 0.2% w/v, were added to culture with S. obliquus or BG-11 media alone for subsequent SEM and light-microscopy imaging. Treated and untreated stems of a similar developmental stage were compared by microscopy. Because individual SG plants have stems at different degrees of maturation, corn was included due to more easily identifiable morphology compared to switchgrass. The observed surfaces of stem sections were prepared prior to the inoculation experiment according to the following procedure. Fresh stems of similar ontogenic stages, approximately 10 mm (corn) and 5 mm (SG) in diameter were selected. These were cut with a razor blade to 1-µm-thick cross sections and autoclaved before inoculation with algae as described earlier. After harvesting algal cells, plant cross sections were preserved in 1:1:1 glycerol:ethanol:water (Kitin et al., 2020) until they were processed for imaging.
For light microscopy imaging, the sections were stained with acridine orange according to Houtman et al., 2016 and observed by epifluorescence microscopy using a UV (365 nm) excitation and 435 nm long-pass fluorescence, which yielded orange-red fluorescence from unlignified cellulosic walls and green fluorescence from lignified portions of walls (Houtman et al., 2016; Kitin et al., 2020). Some sections were observed by confocal microscopy (LSM, Carl Zeiss 710) using objective lens C-Apochromat 40x/1.20 W Korr, and three excitation/emission channels: 405/460-480; 488/502-562; and 488/580-700.
For SEM imaging, sections were repeatedly rinsed in ethanol series (50%, 25%, 0%), frozen in liquid N2, freeze-dried, sputter-coated with gold, and observed at 10 kV. We observed more than 100 vascular bundles with microalgae and more than 100 vascular bundles without microalgae by both SEM and fluorescence microscopy.
2.4 DNA extraction and 16S amplicon sequencing
DNA was extracted from pond samples during early-log (1 day after inoculation) and late-log (7 days after inoculation) phase growth for 16S amplicon sequencing. DNA was extracted using the Zymo Research Quick-DNA Fungal/Bacterial Miniprep Kit (Zymo Research, Irvine, CA). Samples were homogenized in the ZR BashingBead Lysis Tubes (0.1- and 0.5-mm beads) using the Biospec Mini-Beadbeater-1 set to 4800 oscillations per minute for 30 seconds. Each sample was homogenized three times and rested on ice between runs. The remaining purification procedure followed the manufacturer’s instructions.
Purified DNA was amplified using the Illumina 16S protocol that includes primer pair sequences for the variable V3 and V4 regions with Nextera XT v2 adapters (Klindworth et al., 2013). Sequencing libraries were constructed and sequenced using an Illumina MiSeq (Illumina Inc., San Diego CA) resulting in DNA fragments of 600 pb. Pair-end sequencing was performed using Illumina MiSeq protocol based on instructions specified by the manufacturer resulting in 301 paired end reads. Read analysis was conducted in DADA2 software version 3.9 (Callahan et al., 2016). Pair-end sequencing data corresponding to ribosomal RNA amplicons were merged into contigs and resulting fragments were filtered by quality in non-overlapping regions as having no more than five sites with a Phred-value <= 30. Filtered paired-end sequences were clustered, merged and identified taxonomically using the SILVA 138.1 prokaryotic small subunit taxonomic database formatted specifically for DADA2 (http://arb-silva.de). The amplicon sequence variant (ASV) abundance and number of reads per ASV were integrated into a single table. The V3-V4 sequences with an average length of 200 base pairs were clustered into 569 species-level ASVs from V3-V4 sequences. Species represented by singleton ASVs were removed from the final analysis. A database containing the taxonomic assignment, the ASV table, and sample information was generated using the R package phyloseq (McMurdie and Holmes, 2013). Species represented by singleton ASVs were removed from the final analysis. Normalization and rarefying (less than 1,000 ASV sequences) of the dataset was performed before calculating the proportion of specific species present within each sample.
2.5 RNA extraction and RNA-seq
RNA was extracted using a hybrid method that combined TRIzol and bead beating for homogenization, followed by purification using Omega Bio-tek E.Z.N.A. Plant RNA Kit (Omega Bio-tek Inc., Norcross, GA). Frozen cell pellets were resuspended in 750 mL of TRIzol. The samples were transferred to 2.0 mL Precellys screw cap tubes containing 200 µL of a 1:1 mixture of 0.1 mm silicon carbide and 0.1 mm zirconium beads. The tubes were vortexed to ensure the beads didn’t stick to the bottom of the tube. The Precellys 24 bead beater (Bertin Instruments, Montigny-le-Bretonneux, France) was placed in a cold room and allowed to equilibrate. We used the programmed setting for 2 rounds of bead beating for 30 s at 6000 rpm, followed by a 2 min cooling time. This step was repeated 3-6 times. Homogenization progress was checked by loading a 2 µL aliquot onto a slide and observing lysis under a light microscope. Once it was determined that samples were sufficiently lysed the samples rested for 3 minutes at room temperature and were then transferred to a new 2.0 mL tube carefully avoiding transfer of the beads. 200 µL of chloroform was added and briefly vortexed to mix. The samples were then transferred to 5PRIME Phase Lock Gel Heavy tubes (Quantbio, Beverly, MA). They rested on ice for 15 minutes, inverted by hand every 5 minutes. Phases were separated by centrifugation at 17,000 x g for 20 minutes at 0-4°C. Samples rested at room temperature for 2 minutes after centrifugation. 80% of the top aqueous phase was transferred to a new tube. Equal volumes of 70% ethanol were added and vortexed for 20 seconds to mix. Purification of the RNA continued with step 7 of the Omega Bio-tek E.Z.N.A. Plant RNA kit. Note that during the wash steps we waited 1-2 minutes after adding the wash to the column to improve nucleic acid purity and waited 2 minutes after adding water to the column to improve yield during elution.
RNA was polyA selected using the Illumina Truseq RNA Library Prep kit (V2) according to the manufacturer’s instructions and sequenced on an Illumina NextSeq 500 (Illumina Inc., San Diego, CA). Raw read data (fastq) were aligned to the Scenedesmus obliquus DOE0152Z genome (Joint Genome Institute (JGI), retrieved January 2021). Alignments were performed using STAR v2.7.1a according to encode standard parameters with one exception (Dobin et al., 2013). Specifically, STAR was implemented using the basic two-pass mode to identify novel splice junctions/events given the genomic resources are draft-level and quant mode was used to directly produce gene-level counts data. All gene counts were imported into the R Bioconductor package, EdgeR (Robinson et al., 2010) and trimmed mean of M-values (TMM) normalization size factors were calculated to adjust for sample-specific differences. The TMM size factors and the matrix of counts were then analyzed using the R package LIMMA (Ritchie et al., 2015). Weighted likelihoods based on the observed mean-variance relationships were then calculated for all samples and genes using voomWithQualityWeights (Law et al., 2014) prior to the calculation of differential expression profiles between sample groups using LIMMA. Gene counts and differentially expressed genes were annotated using gene models for Scenedesmus obliquus DOE0152Z from JGI (Starkenburg et al., 2017) and heatmaps were constructed in R.
4 Discussion
We grew Scenedesmus obliquus in the presence of raw plant substrates: corn stover, switchgrass, sugarcane bagasse, and yard waste. Compared to photoautotrophic growth of the alga, we found that with switchgrass or yard waste present, the average specific growth rate increased 13% and 33%, respectively, although not statistically significant (p ≤ 0.05). We did observe a statistically significant increase in specific growth rate in the presence of switchgrass under ambient CO2 (p ≤ 0.05), with a 73% increase in dry weight. This difference in results is likely due to the greater concentration of plant substrate and greater inoculation cell density used in the latter experiment. As expected, under an increased CO2 concentration the dry weight of S. obliquus exceeded that of both conditions tested in air, regardless of switchgrass addition. However, with switchgrass present in the culture we observed an additional 7% increase in dry weight, although this was not statistically significant. We have previously reported increased growth rates and biomass accumulation of Auxenochlorella protothecoides and Nannochloropsis sp. when cultured with raw switchgrass or corn stover, respectively, with increases ranging from 13-40% in flask cultures. As with these other species, this overall increase in growth in the presence of plant substrate indicated the potential of S. obliquus to use the raw plant as a carbon source for mixotrophic growth.
When switchgrass utilization was tested at 50 L in mini raceway ponds, we did not see the expected increase in growth of S. obliquus with switchgrass present with no difference observed between treatments. In contrast, when Nannochloropsis gaditana was grown at 50 L in mini raceway ponds, specific growth rate was increased 30% with corn stover present when compared to cultures of the alga alone (Schambach et al., 2020). The discrepancy between laboratory scale results and pond scale results is an ongoing hurdle within applied microalgal research (da Silva and Reis, 2015). Many factors, such as culture depth and mixing, light intensity, temperature range, aeration, and on-demand CO2 for pH balance, can be markedly different from the laboratory experiments and likely play a major role in the outcome. Gaining a better understanding of how these factors play a role in plant substrate utilization and optimizing conditions based on that understanding is necessary to improve scalability of this strategy in the raceway pond setting.
Since we did not see the expected growth increase at 50 L, we hypothesized that we would not see a large difference in microbiome composition, and this was generally the case. We did see more representation from members with low abundance during late log phase. However, Paracoccus sp., Allorhizobium-Neorhizodium-Pararhizobium-Rhizobium sp., and Rudanella sp. were present in all cultures regardless of plant substrate or timepoint. This could indicate that bacteria from these genera could be key members of the microbiome that support growth of S. obliquus. Paracoccus is a genus whose members are ubiquitous to both natural and anthropogenic environments and play important roles in many types of microbiomes (Lasek et al., 2018). This genus was abundant in Nannochloropsis gaditana cultures both with and without plant substrate (Schambach et al., 2020). Allorhizobium-Neorhizodium-Pararhizobium-Rhizobium include diazotrophic bacteria commonly found in association with plant roots as well as some green algae including Scenedesmus (Kim et al., 2014). Rudanella is a genus in the family Spirosomaceae, however there is very little literature about the habitats and ecological functions of members of this genus. Characterizing the microbiome of S. obliquus under favorable conditions for plant substrate utilization is necessary for understanding how the microbiome may be playing a role in utilization and how to leverage it for increased microalgal productivity and culture resilience.
Increases in biomass production of S. obliquus under mixotrophic conditions have been reported with addition of low molecular weight, soluble organic substrates like glucose or acetate to the culture (Mandal and Mallick, 2009; Yang et al., 2014; Shen et al., 2018; Song et al., 2021). In addition, Yang et al., 2014 reported S. obliquus could utilize xylose, and when added at a 4 g L-1 concentration (0.4% w/v), cell density was increased 2.9-fold (Yang et al., 2014). S. obliquus could also successfully grow on filtrate containing soluble product of wheat bran fermented by two fungal species (EL-Sheekh et al., 2012). These studies support the ability of S. obliquus to uptake and assimilate several types of exogenous organic carbon. However, when considering the complexity and heterogeneity of lignocellulosic biomass, it may be necessary for S. obliquus to carry a suite of enzymes (e.g., glycosyl hydrolases and others) that, when working synergistically, hydrolyze the structural carbohydrates into free sugars for uptake and metabolism. Qualitative assessment of corn stem morphology via microscopy after incubation with S. obliquus indicated potential cellulolytic enzymatic production as there was noticeable degradation of the plant when compared to the plant incubated in just algal media. The pattern of degradation in these images was similar to observations in Nannochloropsis gaditana and corn stover, particularly regarding the missing and/or collapsed phloem (Schambach et al., 2020).
Multi-omic analysis of A. protothecoides revealed this alga has a suite of potential glycosyl hydrolases, including those from families 5 and 9, some of which were upregulated when grown in the presence of switchgrass or carboxymethylcellulose (Vogler et al., 2018). Proteins from glycosyl hydrolase families 5 and 9 contain members with known cellulolytic activity, suggesting A. protothecoides could be directly degrading cellulose. In the current study, the consistent downregulation of potential endo-1,4-beta-D-glucanases in the presence of switchgrass, regardless of CO2 concentration or time point, suggests S. obliquus may not be directly acting on the cellulose fraction of the plant. This finding could also indicate a lack of recycling or reorganizing of the cell wall of S. obliquus as it contains cellulose (Voigt et al., 2014). This would allow the alga to maintain cell wall rigidity when in the presence of the switchgrass. Anecdotally, lysis of S. obliquus cells that were in culture with switchgrass required 2 to 3 additional cycles of mechanical homogenization before efficient lysis was seen under the microscope.
Hemicellulose is another major component of plant cell walls, and can consists of xylan, xyloglucan, mannan, glucomannan, and beta-glucan (Scheller and Ulvskov, 2010).Considering that xylans, polysaccharides of beta-1,4 linked xylopyranosyl residues, are the major hemicellulose structure in switchgrass, and that S. obliquus was reported to utilize xylose, we expected to see significant upregulation of xylanases and/or xylosidases. These enzymes are endo- and exoglycosidases that release xylooligosaccharides and xylose, respectively, from the xylan backbone (Terrasan et al., 2016). We only observed one transcript encoding an endo-1,4-beta-xylanase significantly upregulated during mid-log phase in the presence of switchgrass in ambient CO2, whereas one xylan 1,4-beta xylosidase transcript was significantly downregulated in all other sample groups. However, significant upregulation of extracellular mannosidases, and beta-glucosidases under ambient CO2 could suggest S. obliquus is acting upon the mannan fraction of switchgrass hemicellulose. Beta-1,4-mannosidases and beta-1,4-glucosidases are the major enzymes involved in mannan hydrolysis with complete hydrolysis requiring a synergistic effect of these enzymes (Rodríguez-Gacio et al., 2012). While mannose content in most grasses is low, glucomannans are an important seed storage polysaccharide in grass species like switchgrass (Scheller and Ulvskov, 2010; Rodríguez-Gacio et al., 2012). Note that beta-1,4-glucosidases also act on the exposed terminal glycosyl residues of cellulose.
We hypothesized that when S. obliquus was grown in the presence of 1% CO2, there would not be a notable increase in growth rate or dry weight with switchgrass supplementation. The increased CO2 concentration allows for the carboxylation reaction catalyzed by RuBisCO to be more energetically favorable, leading to increased carbon fixation and ultimately contributing to faster overall growth. This is exemplified by the 2-fold increase in dry weight when S. obliquus alone was grown in 1% CO2 versus the ambient condition. Sforza et al., 2012 reported inhibition of mixotrophy in A. protothecoides and Nannochloropsis salina when grown under 5% CO2 likely because CO2 fixation was more energetically favorable under the high CO2 concentration. While not statistically significant, we did see slight increases in specific growth rate and dry weight with switchgrass present under 1% CO2 This could indicate the switchgrass derived carbon could be playing a role.
Despite the lack of a statistically significant growth response to switchgrass under 1% CO2, we saw several transcriptomic responses with switchgrass present, that were also contrary to what was found when compared to ambient CO2, particularly regarding differential expression of potential glycosyl hydrolases. Several mannosidases, and a xylanase were significantly downregulated with switchgrass present under 1% CO2, but several alpha- and beta- galactosidases were significantly upregulated. Furthermore, many alpha- and beta-amylases, starch synthase and a branching enzyme, all of which are involved in starch metabolism, were significantly upregulated in the presence of switchgrass and 1% CO2. Stimulation of starch metabolism in microalgae grown under high CO2 concentrations has been reported (Tanadul et al., 2014), however the presence of switchgrass also appears to influence this response. These transcripts were downregulated in the presence of switchgrass in ambient CO2. These results could indicate that under higher CO2 concentration, different carbon storage metabolic mechanisms are at play in the presence of switchgrass, compared to ambient air.
We explored differential expression within mitochondrial oxidative phosphorylation, photosynthesis, and carbon fixation pathways, particularly during late-log phase growth in ambient CO2 and mid-log in 1% CO2, to determine if there were any discernable patterns in the overarching energy metabolism of S. obliquus when switchgrass was present. It has been reported under mixotrophic conditions, the metabolism of some algae, like N. gaditana and C. reinhardtii, is dominated by respiration with an associated decrease in photosynthetic activity (Chapman et al., 2015; Bo et al., 2021). Differently, photosynthesis in C. sorokiniana is largely unaffected by the addition of acetate (Cecchin et al., 2018). Upregulation of transcripts encoding inorganic pyrophosphatases and proton exporting ATP synthases in S. obliquus in the presence of switchgrass under both CO2 concentrations, could be an indication of increased ATP production via respiration. In addition, significant upregulation of transcripts encoding key enzymes in the carbon fixation pathway, including RuBisCO, under both CO2 concentrations could suggest reutilization of the CO2 produced from respiration (Smith et al., 2015).
Interestingly, upregulation of PSI and PSII subunits in the presence of switchgrass was observed under ambient CO2, whereas, under 1% CO2 these or related subunits were slightly downregulated. It could be that increases in expression of these subunits is a response to increased light attenuation because of the large insoluble switchgrass particles. However, when inorganic carbon is less limited under 1% CO2, this appears to be alleviated and could indicate that under ambient CO2, cells are perhaps more reliant on NADPH and ATP produced from the light reactions of photosynthesis for substrate utilization. Conversely, upregulation of these subunits could be an artifact of increased energy derived from the plant carbon being shuttled into chloroplast production. Cecchin et al., 2018 reported upregulation of Ferredoxin-NADP+ reductase (FNR) when C. sorokiniana was grown with acetate, and we also observed this with S. obliquus in the presence of switchgrass under ambient CO2 (Cecchin et al., 2018). In linear electron flow, electrons are transferred from ferredoxin by FNR to produce NADPH in photosynthesis (Chitnis, 2001). For the case of C. sorokiniana, increased reducing power (i.e., NADH) derived from acetate consumption could be transferred from mitochondria to the chloroplast causing an overreduction of plastoquinones and increasing demand of FNR and ultimately contributing to increased overall growth (Cecchin et al., 2018). Similarly, upregulation of FNR could suggest increased electron transport in photosynthesis in S. obliquus in the presence of switchgrass but only under ambient CO2 and more work would be needed to determine if it is in fact due to increased reducing power derived from switchgrass consumption.