Proteomic analysis reveals molecular changes following genetic engineering in Chlamydomonas reinhardtii

Abstract

Background

Chlamydomonas reinhardtii is gaining recognition as a promising expression system for the production of recombinant proteins. However, its performance as a cellular biofactory remains suboptimal, especially with respect to consistent expression of heterologous genes. Gene silencing mechanisms, position effect, and low nuclear transgene expression are major drawbacks for recombinant protein production in this model system. To unveil the molecular changes following transgene insertion, retention, and expression in this species, we genetically engineered C. reinhardtii wild type strain 137c (strain cc-125 mt+) to express the fluorescent protein mVenus and subsequently analysed its intracellular proteome.

Results

The obtained transgenic cell lines showed differences in abundance in more than 400 proteins, with multiple pathways altered post-transformation. Proteins involved in chromatin remodelling, translation initiation and elongation, and protein quality control and transport were found in lower abundance. On the other hand, ribosomal proteins showed higher abundance, a signal of ribosomal stress response.

Conclusions

These results provide new insights into the modifications of C. reinhardtii proteome after transformation, highlighting possible pathways involved in gene silencing. Moreover, this study identifies multiple protein targets for future genetic engineering approaches to improve the prospective use of C. reinhardtii as cell biofactory for industrial applications.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12866-024-03554-4.

Background

Chlamydomonas reinhardtii (Chlorophyta) is a unicellular green microalga that has historically been used as a model system to investigate fundamental aspects of algal biology, including photosynthesis, cell wall biogenesis, and nuclear/chloroplast interactions [1]. C. reinhardtii biomass is also currently produced for use in various industries, such as aquaculture, cosmetics, and nutraceuticals [2]. In addition, recent studies demonstrated the ability of C. reinhardtii to produce valuable secondary metabolites [3, 4] and recombinant proteins (RPs) [5–7].

C. reinhardtii presents many unique features suitable for industrial applications: (i) it can be grown in large volumes and requires a lower initial investment compared to bacterial or mammalian cell systems [8]; (ii) it does not produce endotoxins or harbour human pathogens [9]; and (iii) it can achieve high growth rates in inexpensive media, resulting in potentially lower production costs when compared to other organisms [9, 10]. In addition, the chloroplast, nuclear, and mitochondrial genomes have been fully sequenced [11], and transformation protocols for chloroplast and nuclear DNA are well established [1, 12, 13]. Chloroplast transgene expression can reach up to 20% of total soluble protein in C. reinhardtii [14]. However, the plastid lacks the machinery to perform post-translational modifications (PTMs) [14, 15]. Given the importance of PTMs for recombinant biopharmaceutical production, stability, and efficacy [16], chloroplast expression is not suitable for valuable complex glycoproteins (such as antibodies). In contrast, RPs produced in the nucleus can be directed to the secretion route and present PTMs [6].

Given that targeted nuclear transformation is still imprecise [17], nuclear transgene expression in C. reinhardtii is currently largely accomplished via random transgene insertion. Unfortunately, this approach is hindered by RP yields, and extensive variability in transgene expression among transgenic cell lines [18–21]. In fact, nuclear transgene expression does not go past 0.2% of TSP for certain strains [20]. Several issues contribute to low RP yields in C. reinhardtii, including those related to the chromosomal integration site in which the transgene is randomly inserted [21–23]. These compounding factors include the position effect, which is thought to be related to the position of the nucleosome on the DNA strand and its interaction with transcription repressors and activators [24–28]. Strong position effects result in high variation of transgene expression among transgenic cell lines generated with the same recombinant DNA. Unfortunately, these mechanisms are still not fully described nor understood [27]. Another reason behind low transgene expression is the activation of gene silencing mechanisms by the microalga [27, 28]. Transgene silencing is mainly regulated by histone modification [20, 23, 27, 29, 30]. At the transgene insertion locus, enzymes will add PTMs to the histones to prompt a rearrangement of chromatin, condensing its structure [20, 23, 30, 31]. Condensation of the chromatin structure at the site of insertion results in reduced transcription of the transgene [32–34]. Another mechanism negatively affecting transgene expression is methylation of recombinant DNA. In C. reinhardtii, it has been reported that methylation of transgene cytosine residues actively impedes recombinant DNA transcription [29, 34]. Transgene silencing is responsible for a reducing proportion of cells expressing the transgene in the transgenic population over time.

Multiple strategies have arisen over time to obtain efficient nuclear expression in C. reinhardtii through optimisation of the genetic design, the host strain, or the growth conditions. Transgene optimisation focused on a new generation of promoters [28, 35–37], on the insertion of native introns in the exogenous DNA sequence [38], and on the addition of functional peptides to enhance expression [39–41]. Optimisation of growth conditions [7], and the generation of modified and/or mutated C. reinhardtii strains showing enhanced transgene expression [20, 42] have led to improved final yields of secreted recombinant protein, up to 12–15 mg/L [43, 44]. However, these yields are still too low to be considered for industrial production [45]. Therefore, it is necessary to examine the cellular response to transgene insertion and expression, to better understand the mechanisms activated after genetic engineering, and ultimately achieve competitive RP yields.

In this study, we report the first proteomic analysis of five transgenic cell lines of C. reinhardtii strain 137c (also named strain cc-125 mt+) to investigate the cellular alterations correlated to the transformation process, transgene insertion, and heterologous protein expression. This study provides fundamental insights into the processes involved in RP expression, with the final aim of developing potential strategies to improve RP yields in C. reinhardtii for industrial applications.

Methods

C. reinhardtii cultivation, harvesting, and transformation

C. reinhardtii strain 137c (also named strain cc-125 mt+) was obtained from Invitrogen (GeneArt^® Chlamydomonas Protein Expression Kit). This strain was chosen because it is an unmodified C. reinhardtii strain that has shown very low yields of recombinant protein production. Therefore, a proteomic analysis of this strain might help to better elucidate the protein reprogramming that leads to low transgene expression and protein yield. The plasmid pOptimized (pOpt) pOpt_mVenus_Paro (NCBI: KM061060, Figure S1) containing the mVenus and the paromomycin selection marker expression cassettes [46] was purchased from the Chlamydomonas Resource Center (https://www.chlamycollection.org/).C. reinhardtii strains 137c was grown mixotrophically in Tris Acetate Phosphate (TAP) medium [47] under 50 µmol photons m^− 2 s^− 1 of continuous light in a 25 °C shaking incubator (100 rpm). Cells were harvested at early exponential phase (2–4 × 10⁶ cells mL^− 1) by centrifugation at 1,500 g for 5 min, washed twice with 10 mL of MAX Efficiency™ Transformation Reagent for Algae (Invitrogen™), and resuspended in MAX Efficiency™ buffer to a final concentration of 2–3 × 10⁸ cells mL^− 1. To prepare for electroporation, the concentrated cells (250 µL) and the linearized plasmid (2–4 µg, linearized with PsiI restriction enzyme) were transferred and mixed into an ice-cold electroporation cuvette (0.4 cm). The following settings on the Gene Pulser Xcell™ Electroporation System (Bio-Rad) were used: 500 V, 50 µF, and 800 Ω. Electroporated cells were left to rest for 10 min at room temperature and then transferred into 10 mL of TAP-sucrose (40 mM) at 25 °C under 50 µmol photons m^− 2 s^− 1 light and 100 rpm agitation for overnight recovery. After recovery, the cells were centrifuged for 5 min at 1,000 g. The supernatant was removed, and the cells were resuspended in 200 µL of TAP medium. The resuspended cells were transferred on solid TAP-paromomycin (10 µg mL^− 1) 1.5% agar plates. After one week, all the colonies growing on plates were all transferred to 5 mL of liquid TAP-paromomycin (10 µg mL^− 1) prior to selection by cell sorting.

Selection and screening of positive C. reinhardtii transformants

After one week of growth in selective liquid medium, the transformed cells were subjected to fluorescence-activated cell sorting (FACS) on a BD Influx flow cytometer (BD Biosciences). C. reinhardtii cells were first gated by chlorophyll fluorescence (488 nm excitation, 692 ± 40 nm detection channel), followed by single cell gating (area versus pulse width signals on trigger channel) to gate out cell aggregates. C. reinhardtii single cells were then visualised in FACS plot as chlorophyll fluorescence versus mVenus fluorescence (488 nm laser, 530 ± 40 nm detection channel). Untransformed wild type C. reinhardtii cells served as negative control to set a mVenus positive gate at ~ 0%. C. reinhardtii cells in the transformed samples that fell within the mVenus positive gate were selected and sorted using single cell sort mode (one cell per well) into a 96 well plate containing 200 µL of TAP-paromomycin (10 µg mL^− 1) medium per well. The selected cells were grown mixotrophically in TAP medium under 50 µmol photons m^− 2 s^− 1 of continuous light in a 25 °C shaking incubator (100 rpm) for one week. The surviving cultures were scaled-up to 4 mL TAP-paromomycin (10 µg mL^− 1) medium and grown for three days in triplicates. On the third day, C. reinhardtii transgenic cell lines and wild types were analysed on Beckman Coulter CytoFLEX S flow cytometer to measure their chlorophyll (488 nm excitation, 690 ± 50 nm detection channel) and mVenus (488 nm excitation, 525 ± 40 nm detection) fluorescence to quantify the positive cells in each transgenic cell line, to identify the best candidates for the proteomic analysis.

Cultivation and harvesting of C. Reinhardtii transformants for proteomic analysis

Fig. 1.

Schematic depicting the transformation, growth, and selection strategies for clonal lines of transgenic C. reinhardtii expressing YFP

Based on net fluorescence, the best five transgenic lines for strain 137c were scaled-up to 500 mL of TAP medium (1:250 inoculum ratio) and cultivated under mixotrophic conditions at 25 °C and 100 rpm with ~ 50 µmol photons m^− 2 s^− 1 of light. Three replicates of each transgenic cell line were used to analyse growth and mVenus fluorescence over time. Fluorescence of the transgenic cell lines was analysed by flow cytometry (on Beckman Coulter CytoFLEX S) each day for one week using the same method as described in Sect. 2.2. A fourth replicate was harvested for proteomic analysis when cell density reached ~ 1.5 OD_{750 nm} (mid-exponential phase, 72 h of growth). Fluorescence of the samples was measured before harvesting, to confirm the presence of mVenus-positive cells in the culture. Cells were removed from medium using tangential flow microfiltration with a 0.2 μm polyethersulfone (PES) membrane (Vivaflow 200, Sartorius). We previously demonstrated that the tangential flow microfiltration does not disrupt living cells [48]. Samples were stored at -80 °C prior to proteomics analysis.

Protein extraction and sample preparation

The harvested cells were resuspended (1:5 cell pellet to buffer ratio) in cell lysing buffer (Tris-HCl 50 mM, NaCl 0.4 M, Tween20 0.5%, pH 8.0) and sonicated on ice for 5 min (30 s on / 30 s off) at 30% amplitude (Sonicator Q500, QSonica). Cell debris was removed from the lysate by centrifugation at 10,000 g for 10 min at 4 °C. To eliminate non-protein contaminants, the intracellular proteins were precipitated using chloroform/methanol precipitation [20] and then resuspended in 100 mM of triethylammonium bicarbonate buffer (TEAB) containing 8 M urea. Protein concentration in each sample was measured using the bicinchoninic acid (BCA) protein assay (Pierce™ BCA Protein Assay Kit, ThermoFisher Scientific). The samples were further processed for mass spectrometry analysis. The proteins were reduced and alkylated using 5 mM of tris(2-carboxyethyl)phosphine (TCEP) and 20 mM of acrylamide monomers (AM). To halt the alkylation reaction, 20 mM of dithiothreitol (DTT) were added. To remove contaminants, 40 µg of total protein were processes using the single-pot, solid-phase-enhanced sample preparation (SP3) [49]. Following SP3 purification, proteins were resuspended in 100 µL of 200 mM ammonium bicarbonate (AMBIC), resulting in a final protein concentration of 0.4 µg/µL. The final step involved protein digestion, carried out by adding proteomic grade trypsin (Trypsin Gold, Promega) at a 1:50 w/w ratio, incubating overnight at 37 °C. The resulting peptides were quantified using the Pierce™ Quantitative Colorimetric Peptide Assay (ThermoFisher Scientific). A total of 2 µg of peptides were concentrated and resuspended in 5 µL of mass spectrometry loading buffer (2% acetonitrile (ACN), 0.2% trifluoroacetic acid (TFA)), and analysed by mass spectrometry.

Mass spectrometry proteome analysis

The samples were examined with an Acquity M-class nanoLC system (Waters, USA), using the same settings as described in Barolo et al. [48]. The obtained MS/MS data files were analysed using Peaks Studio X against the UniProt Chlamydomonas reinhardtii proteome (UP000006906, protein count: 18,829) and a contaminants database. The analysis parameters were: Fixed modifications: none; Variable modifications: oxidised methionine, deamidated asparagine; Enzyme: semi-trypsin; Number of allowed missed cleavages: 3; Peptide mass tolerance: 10 ppm; MS/MS mass tolerance: 0.05 Da. Results were filtered to include only peptides with a –log₁₀(P_value) score corresponding to a False Discovery Rate (FDR) of less than 1%. The score utilized was the one in which matches from the decoy database search were less than 1% of the total matches.

Protein annotation

Intracellular proteins were analysed using PEAKS label-free quantification, as already described in Barolo et al. [48]. Shortly, label-free quantification relies on peptide characteristics (such as mass-to-charge ratio and LC retention time) to match peptides across different samples, and calculates fold change between matches using peak areas (signal intensities). This quantification was obtained measuring 5 biological replicates of transgenic lines compared against the wild type. The statistical analysis was performed using PEAKS Q [50], an algorithm in part based on the Significance B method used by Cox and Mann [51]. The outcome of this analysis is a comparison between the two proteomes based on two fundamental values: the significance value (-10logPvalue), validating the quality of the data, and the fold change value, to quantify the difference between wild type and transgenic cell lines. Negative values refer to proteins that were less abundant in the transgenic lines, while positive values refer to proteins that were more abundant in the transgenic lines. The selection of differentially abundant proteins was based on three criteria: number of unique peptides detected (≥ 3), fold change value (-1 ≥ log₂ ≥ 1), and significance value (≥ 20, p-value ≤ 0.01), as recommended in PEAKS documentation (https://www.bioinfor.com/protein-quantification/). The protein annotation was performed as described in Barolo et al. [48]. In summary, the PEAKS software identification and UniProt database annotation were manually integrated to classify the significant proteins according to Gene Ontology (GO): biological process, molecular function, and cellular component [52].

Statistical analysis

Statistical analysis of the cell density data was performed using GraphPad Prism 5.0 (GraphPad Software, La Jolla, California USA), using Kolmogorov-Smirnov and Levene’s test to assess normality and homoscedasticity of the data. A parametric test (two-way ANOVA) was applied accordingly. Significant effects were then examined using post-hoc Fisher’s LSD test. These analyses assessed the null hypothesis of no significant difference in growth between the wild type and the transgenic lines. The results were considered significant at p-value < 0.05. All values for cell growth and fluorescence analysis are mean ± SEM (n = 3 biological replicates).

Results

High heterogeneity of mVenus expression in clonal lines suggest active impairment of transgene expression in C. reinhardtii

Following transformation of C. reinhardtii, the initial culture was heterogenous with a small proportion (20%) of cells expressing the transgene (Fig. 2A and B). Single paromomycin-resistant mVenus-positive cells were sorted by flow cytometry into a 96 well plate, to obtain clonal transgenic cell lines. After FACS sorting, only 11 cell lines survived (11%). The resulting 11 viable transgenic cell lines were scaled up to 4 mL of TAP-paromomycin medium and subjected to additional screening (Fig. 2C).

Fig. 2.

(A and B) Dot plots of fluorescence-activated cell sorting (FACS) for wild type (A) and transgenic cells (B) of C. reinhardtii. The positive cells showing expression of the transgene were gated in yellow based on wild type cells fluorescence. Cells included in the “mVenus” yellow gate were selected for single cell sorting; (C) Net mVenus fluorescence in 11 transgenic lines of strain 137c. The data was obtained by flow cytometry analysis after 72 h of growth under antibiotic selective pressure (n = 3)

There was a large distribution among the clonal lines with respect to transgene expression (Fig. 2C). Transgenic cell lines mV1, mV5, mV6, mV9, and mV10 were selected for proteomic analysis based on highest fluorescence values (Fig. 2C), to utilise cell lines expressing the transgene and obtain a proteomic profile that would be representative of RP expression.

Recombinant protein expression is highly dependent on growth phase

To elucidate the effects of transformation, transgene insertion, and RP expression on C. reinhardtii strain 137c fitness, the growth of the transgenic cell lines was analysed and compared to the wild type (Fig. 3A). The growth curves of strain 137c wild type and transformants showed comparable trends and similar growth rates and all cultures reached stationary phase after 72 h (Fig. 3A).

Fig. 3.

(A) Comparison of growth profiles for wild type and transgenic lines expressing mVenus. Cells were grown in 500 mL of antibiotic-free media (n = 3); (B) mVenus net fluorescence in transgenic cell lines of C. reinhardtii strain 137c over time. Cells were grown in 500 mL of antibiotic-free media. Net fluorescence of mVenus was detected based on wild type background fluorescence

Expression of mVenus varied among cell lines and over time (Fig. 3B). Differences in the extent and timing of mVenus expression may be attributed to the random nature of transgene integration into the endogenous DNA. However, expression of mVenus was maintained in all C. reinhardtii transformants throughout the 7-day experiment, despite being grown in media without selective pressure (Fig. 3B). The fact that all transgenic lines showed fluorescence higher than the wild type in antibiotic-free media suggests that the expression of mVenus did not have an immediate detrimental impact on host fitness.

All transgenic cell lines behaved similarly, reaching peak of fluorescence in mid-late exponential phase and with reduced fluorescence in stationary phase (Fig. 3B). The highest values of net fluorescence occurred between 48 and 72 h, and dropped after 168 h (Fig. 3B). Peak RP expression in mid-exponential phase was previously reported for 137c transgenic lines expressing the therapeutic protein interferon alpha 2a, suggesting that this pattern may be independent of the RP being produced [7].

C. reinhardtii shows proteomic differences after transformation

The analysis of proteins with significantly higher or lower abundance in the transgenic cell lines compared to the wild type revealed several pathways altered after transformation. 425 proteins showed different abundance after genetic engineering (all included in a table as supplementary material). These data are summarized in Fig. 4 and reflect the proteomic modification after the transformation process, the integration of DNA into the genome, and the expression of heterologous proteins.

Fig. 4.

Schematic of altered pathways in C. reinhardtii strains after transgene insertion and expression

Based on biological process, molecular function, and cellular component, the proteins were divided into groups. Among these groups, there were some possibly related to the cell machinery that activates during and after transgene insertion and expression. The ones involved in processes related to cell growth, gene expression, and protein assembly and transport were investigated further.

Cytoskeleton and morphogenesis-related proteins show lower abundance after genetic engineering

All the cytoskeleton proteins and the proteins involved in morphogenesis detected showed lower abundance in the transgenic lines compared to the wild type (Table 1). V-type proton ATPase (UniProt accession number: A8HYU2), flagellar outer dynein arm heavy chain beta (UniProt accession number: A8J1M5), actin (UniProt accession numbers: A8JAV1), and the tubulin beta-1/beta-2 chain (UniProt accession number: P04690), all reported a lower abundance in 137c after genetic engineering. All these proteins are engaged in cell shape and motility. Their lower abundance suggests that the transformation process and the transgenic insertion and expression may be correlated to morphogenesis and cell shape alterations.

Table 1.

Relative abundance of cytoskeleton proteins and proteins involved in morphogenesis after comparison of the transgenic lines with the wild type (n = 5)

Transgenic cell lines report lower abundance of proteins involved in DNA processing

Insertion of recombinant DNA in the endogenous genome is an invasive procedure, therefore after genetic engineering, multiple differences in proteins involved in DNA processing were expected.

All detected proteins in this group showed lower abundance in C. reinhardtii after genetic engineering (Table 2): transcriptional coactivator-like protein (UniProt accession number: A8J724), nucleolar protein, component of C/D snoRNPs (UniProt accession number: A8IA86), high mobility group protein (UniProt accession number: A8J3F0), and protein arginine N-methyltransferase (UniProt accession number: A8JGX5). In eukaryotes, these proteins are all deeply linked to gene expression, histone modification, and chromatin structure alteration [53–57]. The lower abundance of proteins engaged in DNA remodelling and processing after the insertion of exogenous transgenes is an indication that the organism DNA machinery is subjected to a stress after genetic engineering.

Table 2.

Relative abundance of proteins involved in DNA binding and chromatin remodelling after comparison of the transgenic lines with the wild type (n = 5)

Ribosomal proteins exhibit higher abundance after genetic engineering whereas translation initiation and elongation factors report lower abundance

All the ribosomal proteins detected were more abundant in transgenic lines than wild type, while proteins involved in initiation or elongation of translation were less abundant. Ribosomal proteins bind to rRNA, maintain the structural integrity of ribosomes, and are fundamental for translation processes [58]. Table 3 also contains multiple proteins involved in initiation or elongation of translation. These ubiquitous proteins can be found in the nucleus, the cytosol, or the chloroplast, and their role is to initiate the translation of mRNA and to elongate the polypeptide chain during its synthesis in the ribosome. Interestingly, these proteins all showed lower abundance in the transgenic lines.

Table 3.

Relative abundance of ribosomal proteins and proteins involved in translation initiation and elongation. Protein relative abundance was determined by comparing the transgenic lines with the wild type (n = 5)

Quality control proteins and transport processes display lower abundance in C. reinhardtii transgenic cell lines

Transgenic lines exhibited a lower abundance of proteins involved in protein quality control and transport (Table 4). Heat shock proteins (Accession numbers: A0A220IUF1 and A8IRV0), chaperonins (Accession numbers: A8JE91, I2FKQ9, and O48949), and calreticulin (Accession number: Q9STD3) are proteins that bind unfolded proteins (de novo or misfolded) and promote their correct folding. Another interesting result is the presence of multiple proteins involved in vesicle-mediated intracellular protein transport (Table 4). Small ARF-related GTPase (UniProt accession number: A8IL29), coatomer subunit gamma (UniProt accession number: A8IM71), and coatomer subunit beta (UniProt accession number: A8JGS8), and GTP-binding nuclear protein (UniProt accession number: A8IRX5), are all proteins involved in vesicular transport of proteins between organelles, and all show a lower abundance after transformation, transgene insertion, and RP expression.

Table 4.

Relative abundance of proteins involved in protein folding and transport. Protein relative abundance was determined by comparing the transgenic lines with the wild type (n = 5)

Discussion

This study presents the first in-depth proteome analysis of C. reinhardtii after genetic engineering, focusing on proteins associated with pathways potentially linked to the stress induced by the transformation process, the insertion of exogenous recombinant DNA, and RP expression.

To obtain a valuable amount of transgene cell cultures, and a valuable amount of recombinant protein production, it was necessary to screen the positive transformants using FACS. In fact, after transformation, only 20% of the cells showed fluorescence, i.e. successful insertion and expression of recombinant DNA. After cell sorting, only 11% survived the antibiotic pressure, and were available for scale-up to 4 mL in TAP-paromomycin medium and growth. The 11 transgenic cell lines showed a wide range of recombinant protein expression and mVenus fluorescence, reporting net fluorescence values 6- to 8-fold higher among cell lines. This variability could be caused by intrasample differences in cell age, unequal partitioning of RP during cell division, or asynchronous gene silencing mechanisms affecting individual cells in the population. The extensive variability of fluorescence intensity suggests heterogeneity of transgene insertion and expression within genetically identical clonal populations. Since the cells were selected and sorted based on fluorescence and kept under selective antibiotic pressure during the whole growth process, this result suggests gene silencing mechanisms [20, 23, 29] are extensively active and heavily oppose the expression of transgenes in C. reinhardtii. The transgenic cell lines showing higher fluorescence intensity were selected for proteomic analysis.

The 5 different transgenic cell lines selected were grown for one week. The transgenic cultures did not show differences in cell biomass and growth compared to the wild type. During this period, expression of mVenus was maintained in all C. reinhardtii transformants, despite the absence of selective pressure in medium. It is important to consider that all transgenic lines showed fluorescence higher than the wild type in antibiotic-free media and grew with OD comparable to the wild type. This indicates that the expression of mVenus did not have an immediate negative effect on host fitness. However, the expression of mVenus varied vastly among the transgenic cell lines. These differences in recombinant protein production may be attributed to the random nature of transgene integration into the endogenous DNA.

The whole proteome of the 5 transgenic cell lines was extracted and analysed, to investigate the protein alterations following the transformation process and the insertion and expression of recombinant DNA. These proteomic modifications are not directly caused by the transgene expression, they are correlated to the whole genetic engineering process. To be more specific, the transgenic cell lines exhibit protein alterations that allow the expression of RPs, but that are not directly caused by the transgene itself. It remains possible that these changes are caused by multiple genomic rearrangements happening during the whole transformation procedure. This approach was successful for other eukaryotic organisms, such as plants [59]. For all the cell lines, the single fold change values were reported. However, it did not seem to be present a correlation between the net fluorescence of the cell line, and the single fold change of the proteomic data.

Regarding cell growth, the transgenic cell lines showed lower abundance for all cytoskeleton and morphogenesis-related proteins. In more details, V-type proton ATPase, flagellar outer dynein arm heavy chain beta, actin, and the protein tubulin beta-1/beta-2 chain, all reported a lower abundance in 137c after genetic engineering. The V-type proton ATPase has the role of pumping protons through membranes using ATP [60]. This protein is involved in various biological processes, including flagellar beating and sensory signalling, by regulating the pH of the flagellar compartment. This may play a role in motility and various sensory functions. The transgenic cell lines showed a lower abundance of V-type proton ATPase, which confirmed the trend of proteins involved in motility and growth having lower abundance in these engineered algae cultures. The other 3 proteins are all essential components of the cytoskeleton involved in cell shape, growth, and motility [61, 62]. Given that we did not observe notable differences in 137c cell growth, it is possible that the lower abundance of these proteins is due to the use of electroporation, which has been reported to severely impair the tubulin cytoskeleton and motility of cells [63]. Electroporation is an aggressive transformation method that may cause a proteomic response from the cell host to recover after the violent procedure. However, strain 137c possesses a very strong cell wall that makes it impossible to use gentler methods such as glass beads. Regarding cell shape and motility, further experiments are needed to analyse the effect of lower abundance in cytoskeleton-related proteins on the organism.

The analysis was then focused on proteins involved in DNA processing. Multiple proteins reported a lower abundance after genetic engineering: transcriptional coactivator-like protein, nucleolar protein, component of C/D snoRNPs, high mobility group protein, and protein arginine N-methyltransferase. In eukaryotes, transcriptional coactivators are involved in gene expression [57]. These proteins regulate gene expression by interacting with transcription factors and modulating their activity. Depending on the regulatory elements present in their promoters or enhancers, these proteins can enhance or suppress the activity of specific genes. These transcriptional coactivator-like proteins should be investigated more, to fully understand their role after genetic engineering of eukaryotes. Regarding the other 3 proteins, in other eukaryotic organisms, they are all involved in histone modification and chromatin structure regulation [53–56]. Histones and their interaction with DNA play a major role in transcription regulation by condensing or extending the chromatin structure to affect DNA accessibility [32, 33]. Distension or condensation of the chromatin structure results in enhancement or diminishment of transcription and is regulated by histone PTMs [31]. The high mobility group protein, which was less abundant in transgenic lines, does not directly modify histones; however, it is involved in bending DNA to ease the insertion of histones during formation of the nucleosome (the octamer constituted of histones H2A, H2B, H3, and H4). Furthermore, the nucleolar protein component of C/D snoRNPs (gene NOP1) and the protein arginine N-methyltransferase (gene PRMT2), both less abundant in transgenic lines, are also two proteins involved in histone modification. More specifically, in yeasts, the activity of the NOP1 protein is directly related to enhanced transcription [53, 54]. The gene PRMT2 in Homo sapiens encodes for a protein related to both positive and negative regulation of transcription [56]. Histone accessibility and modification has been analysed in C. reinhardtii as well [64]. It has been observed that UV-mutated strains producing high yields of RPs (UVM4 and UVM11) [20] lack specific histone-modifying proteins, Sir2-type histone deacetylase, correlating increased RP production to lower histone post-translational modifications [64]. Depending on the type of modification of the histone, transgene expression can be activated or deactivated. It would be interesting to deeply understand the role of the nucleolar protein (gene NOP1) and the arginine N-methyltransferase (gene PRMT2) in C. reinhardtii, to evaluate their role in histone modification and RP production. Taken together, the differential production of all these proteins in the transgenic lines implies potential DNA restructuring after the whole genetic engineering process.

Subsequently, proteins involved in DNA translation were investigated. Several ribosomal proteins showed higher abundance after transgene insertion and expression. Ribosomal proteins bind to rRNA, maintain the structural integrity of ribosomes, and are fundamental for translation processes [58]. Interestingly, high abundance of these protein can be related to ribosomal stress. It has been reported that after genetic engineering of organisms, the extensive continuous production of secreted and intracellular RP can generate a ribosomal stress, leading to ribosomal proteins not being assembled into ribosomes and accumulating in the cytoplasm [58]. On the other hand, proteins involved in initiation or elongation of translation were less abundant. Interestingly, these proteins all showed lower abundance in the transgenic lines. It has been reported in numerous eukaryotes that endoplasmic reticulum (ER) stress can be caused by misfolded proteins accumulating in the ER during protein assembly [65, 66], including C. reinhardtii [67]. The organism response to ER stress is called unfolded protein response (UPR), a signalling cascade capable of restoring ER homeostasis. This stress response mechanism has multiple pathways, including phosphorylation of translation initiation factor 2 (eIF2α) [65–67]. This specific protein was not detected in this study; however, the lower abundance of other multiple initiation and elongation translation factors might be related to transgene expression stress and needs to be investigated further.

Lastly, proteins involved in protein quality control and transport were analysed. Transgenic lines exhibited lower abundance of specific heat shock proteins, chaperonins, and calreticulin, which are all involved in quality control. Overproduction of a RP could overwhelm the protein quality control and transport machinery, which could be shut down to divert energy into proteolysis pathways such as the ubiquitin–proteasome pathway (ERAD) and the lysosomal pathway (autophagy) [65]. Another interesting result is the presence of multiple proteins involved in vesicle-mediated intracellular protein transport. Small ARF-related GTPase, coatomer subunit gamma, and coatomer subunit beta, are all proteins involved in vesicular transport of proteins between the ER and Golgi [68, 69]. Moreover, the proteomic profiles of the transgenic cell lines show another class of transport protein, specifically proteins involved in nucleocytoplasmic transport. The GTP-binding nuclear protein is involved in transport of proteins into the nucleus and of RNA outside the nucleus into the cytosol. This protein is also responsible for the export of ribosomal subunits outside the nucleus. The lower abundance of more proteins involved in ER operations, especially in the ribosomal activity, highlights an interesting proteomic alteration after genetic engineering that requires additional investigation.

Conclusions

This study utilises an innovative screening method of Chlamydomonas reinhardtii transformants using fluorescence-activated cell sorting (FACS) to obtain positive transformant lines originating from single cells. Separation of single positive transgenic cells by FACS, and subsequent growth of the cultures as single transgenic cell lines, eliminated the need for time-consuming colony screening. However, even after FACS, not all transgenic cell lines survived the antibiotic pressure, and there was extensive variability in net fluorescence of mVenus in theoretically identical clones, suggesting that transgene silencing mechanisms are widely active in C. reinhardtii.

This study also delivers the first proteomic analysis and comparison of Chlamydomonas reinhardtii strains 137c following genetic engineering, to investigate the effect of recombinant DNA insertion and heterologous protein expression on the organisms, and possibly reveal pathways that are responsive to the whole transformation process (Fig. 4).

Transformation process, recombinant DNA insertion, and transgene expression had a strong impact on total protein production and altered multiple pathways. Surprisingly, C. reinhardtii transgenic cell lines reported lower abundance of proteins directly related to protein production, such as proteins involved in nucleic acid processing, translation initiation and elongation, protein quality control, and protein transport, while ribosomal proteins were detected in higher abundance (Fig. 4). Many of the proteins identified to be significantly differentially abundant in our dataset remain to be characterized in C. reinhardtii. However, these data offer first insights into the cellular responses to the stress of genetic transformation and heterologous protein expression. Future studies should include additional strains, diverse transformation methods, empty vector controls, and different plasmids for alternative protein expression to determine which proteomic changes are shared among these conditions. Additionally, genomic sequencing and analysis of insertion sites would shed further light onto the effects of random gene integration on the proteomic response of individual cell lines.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

LB, RMA, ASC, and MP designed the experiment; LB carried the experiment; LB and MPP performed the mass spectrometry analysis; LB, ASC, RMA, and MP performed the data analysis; LB prepared all the figures and wrote the manuscript with contributions from all authors.

Funding

This work was supported by funding from the Climate Change Cluster (C3) of the University of Technology Sydney (UTS). Lorenzo Barolo was supported by a C3 PhD scholarship.

Data availability

All proteomic data generated and analysed during this study are included in this published article as supporting material.

Declarations

Ethics approval and consent to participate

This article does not contain any studies with human participants or animals performed by any of the authors.

Consent for publication

This article does not contain any identifying images or other personal or clinical details of participants that compromise anonymity.

Competing interests

The authors declare no competing interests.