In summary, our results suggest that, upon onset of chloroplast proteotoxic stress, a signal transduction pathway, originating in the chloroplast, leads to activation of the cytosolic kinase Mars1, which in turn orchestrates the cpUPR transcriptional program. Activation of cpUPR through Mars1 mitigates photooxidative stress and delays photobleaching. However, loss of Mars1 does not impair expression of genes involved in non-photochemical quenching. Thus, the exact mechanism by which the cpUPR pathway confers photoprotection remains to be deciphered. Input conditions that activate the cpUPR, as well as cpUPR target genes identified to date, are phylogenetically conserved from C. reinhardtii to A. thaliana (Llamas et al., 2017; Zybailov et al., 2009; D’Andrea et al., 2018). Thus, despite not yet having identified a functional ortholog of MARS1 in higher plants, it is reasonable to assume that the cpUPR’s previously unknown role in protecting cells against photooxidative stress would likewise be conserved. This notion is particularly appealing in light of the observation that basal induction of the cpUPR conferred a protective effect in response to stress and that, conversely, loss of Mars1 activity profoundly sensitized cells towards HL and other chloroplast stressors. Hence, engineering plants with constitutive cpUPR activation may be a promising strategy to enhance their tolerance to environmental stresses.

All C. reinhardtii strains were maintained on Tris-Acetate-Phosphate (TAP) solid media (1.6% agar, USP grade, Thermo Fisher Scientific) with revised Hutner’s trace elements (Kropat et al., 2011) at 22°C in low light (~10–20 µmol photons m⁻² s⁻¹). Lines harboring the ClpP1 repressible gene were maintained in the media supplemented with 100 µg ml⁻¹ spectinomycin (Sigma). Lines harboring a mutagenic cassette disrupting the MARS1 gene were maintained in the media supplemented with 20 µg ml⁻¹ paromomycin (Sigma). Lines harboring a MARS1 transgene construct were maintained in the same conditions with solid media supplemented with 20 µg ml⁻¹ hygromycin (Thermo Fisher Scientific). Sulfur-depleted TAP liquid and agar pates were prepared as previously described (Davies et al., 1994). Generally, during liquid growth, no antibiotic was supplemented. For the experiments shown in Figure 4—figure supplement 2A–C, TAP liquid cultures and agar plates were supplemented with 5 µg/ml or 0.2 µg/ml tunicamycin (EMD Millipore), respectively. All strains used in this study are listed in Table 1.

The cpUPR reporter strain (CrPW1) was generated by nuclear transformation of the C. reinhardtii ClpP1 repressible strain (DCH16; Ramundo et al., 2014), using 300 ng of Nde1-linearized pPW3217 plasmid Table 2 for plasmids used in this study). Nde1 and all the other restriction enzymes described in this publication were purchased from NEB. The nuclear transformation was carried out via electroporation as described below (section on ‘Insertional mutagenesis’). Transformants isolated on TAP agar plates containing 20 µg ml⁻¹ hygromycin and tested by Flag immunoblot analysis upon ClpP1 repression and exposure to HL. As previously observed (Ramundo et al., 2014; Ramundo et al., 2013), during random insertion of a construct with regulatory regions in its promoter, less than 10% of the hygromycin resistant transformants preserved the correct gene expression pattern of the downstream coding sequence. Among them, we selected CrPW1 for further studies. The pPW3217 plasmid, containing a minimum region of the VIPP2 gene promoter, its 5’ untranslated region and its first exonexon intrintrone yellow fluorescent protein coding sequence (YFP CDS), C-terminally appended to a triple Flag epitepitope the 3’ untranslated region of the RBCS2 gene, was generated by In-Fusion cloning (Clontech). The VIPP2 genomic fragment was amplified from genogenomic with primers SR510 and SR502 (see Table 3 for primers used in this study). The YFP CDS fused to a 3x-Flag epitope was amplified from pLM005 (Chlamydomonas Resource Center) with primers SR503 and SR504. The 3’ untranslated region of the RBCS2 gene was amplified from pHyg3 (Berthold et al., 2002) with primers SR505 and SR506. All PCRsPCRse performed using Phusion Hotstart II polymerase (Thermo Fisher Scientific). PCR products were gel-extracted using NucleoSpin Gel and PCR Clean-Up Kit (Takara) according to manufacturer’s instructions, and they were further purified by ultrapurephenol:chloroform:isoamyl alcohol (25:24:1, v/v/v) (Life Technologies) extraction and ice-cold ethanol (Sigma) precipitation. These three purified DNA fragments were then mixed with a purified and linearized pHyg3 vector, previously digested by PciI and EcoRV, and incubated with the In-Fusion reagents (Takara) as per manufacturer’s instructions. The In-Fusion product was transformed in Stellar competent cells (Takara) according to manufacturer’s instructions and putative positive clones were selected in LB solid media (1.7% agar) supplemented with ampicillin after overnight incubation. The resulting plasmid, pPW3217, was purified using the Kit PureYield Plasmid Miniprep System (Promega) and verified by analytical digestion and sequencing. All constructs made by In-Fusion cloning for this publication follow this same protocol from transformation through plasmid isolation.

The mutagenic DNA cassette was isolated by restriction enzyme digestion of pMJ016c (provided by the Jonikas laboratory), which contains the HSP70-RBCS2 chimeric promoter, the paromomycin resistance gene AphVIII, and the PSAD and RPL12 chimeric terminator (Mackinder et al., 2016). Using the Mly1 enzyme, a blunt fragment of 2204 bp (containing the mutagenic DNA cassette) was isolated and extracted from a 1% agarose gel through the NucleoSpin Gel and PCR Clean-Up Kit (Takara) per the manufacturer’s instructions. To further remove possibly contaminating DNA, the mutagenic DNA cassette was electrophoresed on a 1% agarose gel, extracted and repurified as explained above.

A 1-liter liquid culture of cpUPR reporter strain (CrPW1) was grown in TAP medium in low light (~30 µmol photons m⁻² s⁻¹) to a density of about 2–4 × 10⁶ cells ml⁻¹. Cells were collected at room temperature (RT) by centrifugation at 3000 x g for 5 min and gently resuspended in TAP supplemented with 40 mM sucrose at 2 × 10⁸ cells ml⁻¹. Multiple aliquots of 250 µL of cell suspension were then transferred into Gene Pulser electroporation cuvettes (0.4 cm gap, Bio-Rad) and incubated at 16°C for 5–30 min. In each cuvette, about 20 ng of mutagenic DNA cassette was added to the cell suspension and quickly mixed by pipetting. Electroporation was performed immediately using a Gene Pulser II electroporation system (Bio-Rad) with the following parameters: capacitance = 25 μF and voltage = 800 V. Electroporated cells from each cuvette were then diluted into 8 ml TAP supplemented with 40 mM sucrose and allowed to recover overnight by gentle agitation in very dim light (5–10 µmol photons m⁻² s⁻¹). The next day, cells were collected by centrifugation at 1000 x g for 5 min, resuspended in 1 ml of TAP medium, plated on TAP agar plates containing 20 µg ml⁻¹ paromomycin and incubated in darkness for about three weeks before picking colonies.

Approximately 55,000 total mutants were picked and re-arrayed in a 384-colony format on rectangular agar plates (Singer Instruments) using a Norgren CP7200 colony-picking robot. In each 384-mutant array plate, the last two rows were kept empty to include internal positive and negative controls for the next stage of the screen (for details, read section ‘Execution of YFP mutant screen on agar plates’). This library of mutants (of approximately 150 agar plates) was grown in complete darkness at 22°C and propagated every 3–4 weeks by robotically passaging the mutant arrays to fresh 1.5% agar solidified TAP medium containing 100 µg ml⁻¹ spectinomycin using a Singer RoToR robot (Singer Instruments). Unfortunately, numerous mutants were lost during propagation due to a wide-spread contamination event.

To screen for YFP silencing or activating mutants, rectangular agar plates, each containing 97 ml of TAP medium -/+ Vit (400 μM thiamine-HCl (Sigma) and 80 ng ml⁻¹ of vitamin B₁₂ (Sigma)), were prepared. In each agar plate (-/+ Vit), 12 colonies of the cpUPR reporter strain (CrPW1), 12 colonies of the parental ClpP1 repressible strain (DCH16) and 12 colonies of 2 different positive YFP expressor strains (CrPW2 and CrPW3) were robotically spotted in the last two rows of the 384-colony array to be used as internal positive and negative controls during the YFP screen (see scheme in section ‘Semi-automated identification of YFP mutants through Image-J macroscripts’). Next, insertional mutants (freshly propagated) were spotted onto these same plates. Plates were incubated in dim light (20–30 µmol photons m⁻² s⁻¹) at 25°C and were imaged after 2 and 6 days using a Typhoon TRIO fluorescence scanner (GE Healthcare). For each round, 12 plates of insertional mutants (6 without and six with Vit) were simultaneously scanned with the settings described below. Chlorophyll Autofluorescence: Excitation 633 nm, Emission filter 670/30 nm, PMT 300, Sensitivity Normal, Pixel size 500 μm. YFP Fluorescence: Excitation 532 nm, Emission filter 555/20 nm, PMT 800, Sensitivity Normal, Pixel size 500 μm. The focal plane parameter was set at ‘plus 3 mm’ to focus the optics 3 mm higher than the glass plate. This last detail and the thickness of the agar plate (97 ml of liquid agar) were critical parameters to successfully detect the YFP signal.

We used macroscripts in ImageJ64 software (Schneider et al., 2012) to quantify the intensity values of colonies in the mutant library (for details, please refer to Source code 1). Each plate was imaged in the YFP and in the chlorophyll channel. To orient each image, the bottom two rows of each plate were spotted with characterized positive and negative YFP strains (CrPW1, DCH16, CrPW2, and CrPW3) in the specific order outlined in the scheme below, where (-) denotes lack of YFP signal and (+) denotes presence of YFP signal. The ordering of these colonies conferred a reproducible fluorescent pattern in the YFP channel, which was used to identify the bottom two rows of each image.

To quantify the YFP intensity of each mutant, a 16 × 24 array containing 384 Regions-Of -Interest (ROIs) was constructed for each image in the chlorophyll channel, where all living colonies exhibited signal. The same grid was applied to the corresponding image in the YFP channel. From each ROI on autothresholded images, the maximum intensity in the YFP channel was measured.

To account for variability in the magnitude of the YFP response on different plates (due to slight variations in agar thickness), it was necessary to normalize the YFP intensity of each mutant colony to the YFP intensity of its parental strain (CrPW1) from the same plate (average of n = 12). Colonies exhibiting YFP fluorescence higher than three standard deviations from the average of all colonies were labeled as potential activators, while colonies with YFP intensities below three standard deviations from the average were labeled as potential silencers.

Of the potential cpUPR silencers, we observed that many of these mutants grew to a larger colony size than the parental CrPW1 strain after 6 days. Their robust growth suggested suppression of vitamin-induced ClpP1 inactivation. To exclude these suppressor mutants, we analyzed the area of all the colonies at 2 and 6 days by measuring the particle area of autothresholded images in the chlorophyll channel. The average colony size increase was 1.38-fold for the 10000-plus colonies analyzed. Candidate colonies that increased 2-fold in colony area (more than one standard deviation away from the average) were regarded as suppressors. Of the remaining silencing candidates, mars1-1 exhibited the most attenuated YFP response in the presence of Vit.

We indicate mars1-2 with an asterisk (*) in Figure 2A because this mutant was identified in a secondary screen but its position in the original mutant library was lost due to contamination of the plate. To evaluate the YFP response and colony size of mars1-2 in the context of the entire mutant library, we re-spotted mars1-2 on a fresh agar plate containing the characterized cpUPR reporter strain and the other positive and control strains in the bottom two rows as described above. The normalized YFP intensity of mars1-2 in the absence and presence of Vit was then mapped onto the quantification of the original mutant colonies.

With the single exception of the DNA samples submitted for whole genome sequencing, all the other genomic DNA (gDNA) extractions were performed as described below. A 6 ml aliquot of a liquid TAP culture in mid-log phase were spun down, and the media was decanted. The pellet was resuspended in 400 µl of water and then 1 vol of 2x DNA lysis buffer was added (200 mM Tris HCl pH 8.0, 6% Sodium Dodecyl Sulfate (SDS), 2 mM EthyleneDiamineTetraAcetic acid (EDTA). To digest proteins, 5 µl of 20 mg/ml proteinase K (Life Technologies) was added and allowed to incubate at Room Temperature (RT) for 15 min. 200 µl of 5M NaCl was then added and mixed gently. Next, to selectively precipitate nucleic acids, 160 µl of 10% cetyltrimethyl ammonium bromide (CTAB) (EMD Millipore) in 0.7 M NaCl was added and allowed to sit for 10 min at 65°C with gentle agitation. Two or more consecutive rounds of DNA extraction using ultrapure phenol:chloroform:isoamyl alcohol (25:24:1, v/v/v) were performed to achieve a clean interphase. Then, the upper aqueous phase was retained and mixed with 1 vol of 2-propanol (Sigma). This was mixed gently for 15 min at RT. Then it was spun down for half an hour at 21,000 x g at 4°C. The supernatant was removed and 1 vol of ice-cold 70% ethanol was added and mixed with the pellet. This mixture was spun down for 15 min at 21,000 x g. The supernatant was removed and the DNA precipitate was dried in a speed-vac for about 10–25 min and resuspended in 40 µl of nuclease-free water (Ambion). To ensure complete removal of any potential RNA contamination, in most cases, the gDNA prep was then subjected to in-solution ribonuclease treatment using Rnase A/Rnase T1 mix (Thermo Fisher Scientific) according to manufacturer’s instructions. Finally, the gDNA was quickly repurified through an additional round of DNA extraction using ultrapure phenol:chloroform:isoamyl alcohol (25:24:1, v/v/v) and 2-propanol precipitation as described above.

The purity of the gDNA prep was assessed by Nanodrop (Thermo Fisher Scientific), ensuring absorbance ratios at 260/280 nm and 260/230 nm to be ~1.8 and ~2.0, respectively, prior to using the gDNA preparation for most of the follow-up applications. For the pooled (whole genome sequencing) DNA samples, the genomic DNA extraction was performed with the following protocol adapted from the Qiagen, DNeasy Plant Mini Kit using its proprietary buffers (Buffer P3, AW1, AW2, AE). A 25 ml culture of each progeny was grown for ~2 days to about 3 × 10⁶ cells ml⁻¹. Cells were then pelleted and resuspended in 0.5 ml of SDS-EB lysis buffer buffer (50 mM Tris-HCl, 200 mM NaCl, 20 mM EDTA, nuclease-free H₂O, 2% SDS, 1% Polyvinylpyrrolidone -average molecular weight = 40,000-, 1 mg/ml of proteinase K) and allowed to incubate for ~10 min. One volume of phenol:chloroform:isoamyl alcohol (25:24:1, v/v) was added and mixed vigorously. The mixture was spun at 13,500 x g for 5 min at RT. The upper phase was transferred into new Eppendorf tubes and 5 µl of 100 mg/ml of RNase A, was added and incubated at RT for 30 min. The lysate was mixed with 130 µl aliquot of Buffer P3 and incubated on ice for 5 min. This mixture was spun at 18,400 x g for 5 min at RT. The lysate was transferred to the QIAshredder Mini column and spun for 18,500 x g for 2 min. The flow-through fraction was transferred into a new tube and 1.5 volumes of Buffer AW1 was added to the cleared lysate and mixed well. This mixture was then transferred to a DNeasy Mini column and spun at 7600 x g for 1 min. The flow-through fraction was discarded and this step was repeated for any remaining mixture. The DNeasy Mini column was transferred to a new collection tube and 500 µl of Buffer AW2 was added, centrifuged at 7600 x g for 1 min and the flow-through discarded. Another 500 µl of Buffer AW2 was added and centrifuged at 18,400 x g to dry the membrane. The column was then transferred to a new Eppendorf tube. 90 µl of Buffer AE was added onto the DNeasy membrane and incubated for 5 min at RT. The DNA was eluted by centrifugation at 7600 x g for 1 min. This step was repeated again, using the 90 µl of Buffer AE collected after the first centrifugation. The quality of the DNA samples was assessed by Nanodrop as described above. The DNA samples were then stored in −20°C until use.

The protocol was optimized for single-colony DNA sequencing from the original protocol (Li et al., 2016). A pure genomic DNA preparation was assured by running the DNA on a 1.5% agar gel prior to starting. A single-stranded DNA fragment was generated by extending a biotinylated primer from the cassette to the flanking DNA using either primer oMJ598 or primer oMJ1234, which anneals to the 3’ or 5’ end of the mutagenic cassette, respectively. The linear extension mix was set in the following way: 500 ng of gDNA, 2 µl of 0.25 µM of biotinylated primer, 0.5 µl of Phusion Hot Start Polymerase (Thermo Fisher Scientific), 10 µl of Phusion GC buffer, 3 µl of DMSO, 1 µl of 50 mM MgCl₂, 1 µl of 10 mM dNTPs. Prior to starting the mix, the GC buffer was thawed, heated to 95°C for five minutes, vortexed and then put back on ice until the solution became completely clear. The linear extension reaction was carried out in a thermocycler with the following program: Stage 1) 98°C for 3 min; Stage 2) 98°C for 10 s, 65°C for 30 s, 72°C for 18 s (40 cycles). This program was run twice and, in between the first run and the second run, 0.5 µl of Phusion Hot Start Polymerase was added. The Dynabeads kilobase Binder Kit (Thermo Fisher Scientific) was used to purify the linear extension product. For each reaction, 8 µl of streptavidin-coupled magnetic beads) were transferred into an Eppendorf tube and washed in 100 µl of phosphate-buffered saline (PBS) up to four times using a DynaMag magnet (Thermo Fisher Scientific). The Dynabeads were then washed once more in 20 µl of binding solution and gently resuspended in 100 µl of binding solution, pipetting up and down only a few times. Next, the beads were transferred in the PCR tube from the linear extension reaction described above. To allow efficient binding of the linear extension product to the streptavidin-couple magnetic beads, the samples were incubated overnight at RT on an overhead-rotating platform. The following day, the linear extension product was isolated and ligated to a single-strand DNA adaptor, per the following procedure. The beads were washed three times with 100 µl of PBS allowing 8 min incubation in between washes. At the end of the final wash, we ensured that all PBS was carefully removed to avoid interference with the ssDNA ligation reaction. A 20 µl ssDNA ligation reaction was added and gently mixed with the magnetic beads. The ligation mix contained 11.25 µl of H₂O, 1 µl of 25 µM ssDNA adapter primer (oMJ619), 1 µl of 50 mM MnCl₂, 4 µl of 5 M betaine, 2 µl of CircLigase II reaction buffer, 0.75 µl of CircLigase II (Epicentre). The beads were transferred to the thermocycler, which was pre-heated to 60°C for 10 min. This mixture was incubated for 1 hr at 60°C. The beads were then washed three times with 100 µl of PBS as described above. Next, the ssDNA was converted to a dsDNA using primers annealing to the ligated adaptors at the ends of the ssDNA sequence. 1 µl of 25 µM of Primer 1 (see below), 1 µl of 25 µM of Primer 2 (oMJ621), 0.5 µl of Phusion HotStart, 10 µl of Phusion GC buffer, 32.5 µl of H₂O, 3 µl of DMSO, 1 µl of 50 mM MgCl₂, 1 µl of 10 mM dNTPs and the template DNA (beads) were mixed together. Primer one depended on whether the original extension from the cassette was in the 5’ or 3’ orientation. If the 3’ cassette flanking primer (oMJ598) was used during the linear extension, primer T3_3’oMJ016c 11/23 was chosen. Instead, when the 5’ cassette flanking primer (oMJ1234) was used during linear extension, oMJ1239 was chosen. Primer 2 (oMJ621) annealed to the ligated adaptor. Both primers were designed to contain a binding site for the mutagenic cassette and a binding site for a T3 sequencing primer. The following amplification program was used: Stage 1) 98°C for 3 min, Stage 2) 98°C for 10 s, 63°C for 25 s, 72°C for 20 s (10 cycles), Stage 3) 98°C for 10 s, 72°C for 45 s (13 cycles). The dsDNA products were then run on a 1% gel. The DNA smears were cut out of the agarose gel, purified using NucleoSpin Gel and PCR Clean-Up Kit (Takara) according to manufacturer’s instructions and subjected to Sanger Sequencing using a standard T3 sequencing primer. Finally, to identify the insertion site of the mutagenic cassette, the sequencing results were blasted in Phytozome, v5.5. The sequence of the aforementioned primers can be found in Table 3.

Note: mars1-1 and mars1-2 proved difficult to mate due to their genetic background. Therefore, extra measures were taken to increase the efficiency of mating. mars1-3 and mars1-4, obtained from the Jonikas library, did not have this problem, therefore, these extra measures were not taken. The following protocol will indicate the differences.

Cells were re-streaked onto fresh TAP agar and incubated in low light (<15 µmol photons m⁻² s⁻¹) for five days. They were then transferred onto TAP agar containing 1/10 of the usual NH₄Cl concentration and kept in this medium for four-five days under moderate light (~40 µmol photons m⁻² s⁻¹) to induce starvation. The gametes from each strain were then resuspended in a 24-well sterile transparent plate (Costar) using 150–200 µl of water or M-N/5 solution till a dark green resuspension is obtained. M-N/5 solution was used for the mars1-1 and mars1-2 backcrosses, and water was used for mars1-3 and mars1-4 backcrosses. M-N5 solution contained 1 ml of 10% sodium citrate, 0.2 ml of 1% FeCl₃, 0. 2 ml of 4% CaCl₂, 0.34 ml of 10% K₂HP0₄, 0.2 ml of 10% KH₂PO₄, 0.2 ml of Hutner’s Trace elements (Chlamydomonas Resource Center), H₂O to 1.25 liter. The solution was autoclaved in 100 ml aliquots per bottle. The plate was then transferred to a shaker under moderate light (~40 µmol photons m⁻² s⁻¹) and allowed to mix for ~1 hr. Gametes of the opposite mating type were mixed (~100 µl per gamete) in a separate well and the plate was placed under light with no shaking. For the mars1-1 and mars1-2 strains, after one hour of mixing, dibutyryl cyclic AMP (Sigma) was added to each mating mix to a final concentration of 30 mM. Mating efficiency was checked periodically (every 15–30 min) for fusion events and quadriflagellate formation. The gametes were mated for ~3 hr. Aliquots (100 µl) of the mating mixture were plated into TAP 4% agar. Plates were exposed to light (~50 µmol photons m⁻² s⁻¹) overnight and the next day wrapped in aluminum foil. After ~1–2 weeks, the vegetative cells were scraped off using a small rectangular soft razor blade (Personna, .009’', two-facet aluminum blade) with gentle pressure on the agar. Zygotes adhere to the agar surface and can be recognized under a light microscope due to their darker and larger appearance. A 100 µl aliquot of liquid TAP medium was then added on top of the zygotes and a more rigid scalpel (Feather, N.2) was used to scrape the zygotes off the agar. A line was drawn onto the center of a fresh TAP 1.5% agar plate, and the zygotes were spotted along this line. The cells were then allowed to dry. For mars1-1 and mars1-2, but not for mars1-3 and mars1-4, vegetative cells were killed by treating the plate with chloroform vapor for ~15–30 s. The plate was incubated under light (~80 µmol photons m⁻² s⁻¹) overnight to 1.5 days. At 24°C, germination typically occurred after ~20 hr. Under the dissection scope, tetrads and octads were found and dissected. Incomplete tetrads, full tetrads and octads were then re-arrayed onto TAP agar in a 96-array format and then replicated onto the appropriate drug resistances. When necessary, mating type-specific PCRs (Werner and Mergenhagen, 1998) were carried out to ensure that the progeny were in fact due to a sporulation event and were not mistakenly parental strains.

Genomic DNA from progeny derived upon crossing mars1-1 to CC-124 was obtained as outlined in ‘Genomic DNA extraction’ section. The insertion of the mutagenic cassette in the MARS1 locus was verified by PCR by using primers SR773 and T3_5'_oMJ016c 11/24, which anneal to exon 17 of MARS1, and to the 5’ side of the mutagenic cassette, respectively. The PCR reaction was run on 1–1.5% agarose, cut out of the agarose gel, purified using NucleoSpin Gel and PCR Clean-Up Kit (Takara) according to manufacturer’s instructions and subjected to Sanger sequencing to verify the expected sequence identity. The sequence of the aforementioned primers can be found in Table 3.

Progeny derived upon crossing mars1-2 x CC-124 were re-arrayed onto 96-well plates and replicated onto TAP supplemented with hygromycin, paromomycin or spectinomycin to determine the segregation patterns of the progeny. Mating type-specific PCRs were performed on almost all progeny (Werner and Mergenhagen, 1998). The progeny was then tested in high light and by Vipp2 immunoblot analysis two or three times to determine which one had silencing vs. wild-type phenotypes. Genomic DNA was extracted as described above in the section ‘Genomic DNA extraction’ and was subsequently pooled per the progeny’s phenotype, i.e. mars1-like or WT-like. Additional pools containing the parental strains were also analyzed likewise. The size of the pools of gDNA were of different proportions depending on the amount of progeny in that group (WT vs. mars1-like). The pooled gDNA was then fragmented using Covaris and Bioruptor Pico. The sequencing libraries were prepared with the aid of the PrepX DNA library kit (Takara). One cycle of PCR was used to linearize the library molecules. Fragment analyzer traces and Qubit values were assessed for each sequencing library as quality control checks. Sequencing was performed on the HiSeq2500 Rapid sequencer. The C. reinhardtii reference genome was downloaded from Phytozome, v5.5 onto the Geneious software (Kearse et al., 2012). The reads from each library were then aligned to the reference genome.

Genomic DNA from progeny of meiotic tetrads derived upon crossing mars1-2 to CC-124 was obtained as outlined in ‘Genomic DNA extraction’ section. The MARS1 locus was amplified by using primers SR789 and KP235, which anneal to exon 15 and the intron 19 - exon 20 junction of MARS1, respectively. The MARS1 deletion locus was amplified by using primers KP346, which anneals to intron 1 of the MARS1 gene, and KP347, which was derived from the a WGS read found only in the ‘mars1-like’ progeny pool. The KP347 primer sequence is a hybrid of telomeric sequence and MARS1 gene sequence - this sequence seems to have arisen after a genomic deletion at the end of chromosome 16 in mars1-2. The PCR reactions were then purified and sequenced as described above. The sequence of the aforementioned primers can be found in Table 3.

A MARS1 ‘midigene’ was generated by amplifying four different portions of this gene either from gDNA or cDNA using KOD Hot Start DNA Polymerase (Thermo Fisher Scientific) or Phusion Hotstart II polymerase (Thermo Fisher Scientific). In particular, the region spanning the promoter, the 5’UTR and the first 5 exons of this gene was amplified from gDNA using Phusion polymerase and the following primers: SR828 and SR818; the region spanning exon 5 to exon 15 was amplified from gDNA using KOD polymerase and the following primers: SR819 and HT7; the region spanning exon 15 to exon 28 was amplified from cDNA using KOD polymerase and the following primers: SR789 and SR797 and the 3’UTR was amplified from gDNA using KOD polymerase and the following primers: SR793 and SR829.

All PCR products were gel extracted and purified as described above in the section regarding the pPW3217 cloning. Next, these 4 PCR fragments were mixed with a purified and linearized and pRAM118/pPW3216 vector, previously digested by EcoRV and Not1, and incubated in presence of the In-Fusion reagents (Takara) as per manufacturer’s instructions. The resulting plasmid is notated as pPW3218. The sequence of the aforementioned primers can be found in Table 3.

The Phytozome v5.5 MARS1 transcript annotation is Cre16.g692228.t1.1.

Given the large size of the MARS1 transgene (>10 kbp), a different electroporation protocol was used to successfully integrate each aforementioned MARS1 transgene into the nuclear genome. The electroporation was performed using a NEPA21 electroporator (Nepagene) (Yamano et al., 2013). A 4–8 µl aliquot of purified, non-linearized, plasmid DNA at a concentration of 0.5–1 mg ml⁻¹ was used per transformation. In each case, the plasmid DNA was mixed together with 5 µl of a ready-to-use, sheared solution of salmon sperm DNA at a concentration of 10 mg ml⁻¹ (Thermo Fisher Scientific) prior electroporation. The electroporation parameters were set as follows: Poring Pulse (300 V; length = 6 ms; Interval = 50 ms; No = 1; D.Rate = 40%; + Polarity), Transfer Pulse (20 V; length = 50 ms; Interval = 50 ms; No = 5; D.Rate = 40%; +/- Polarity). Usually, during the electroporation, the impedance was measured to be around 400–700 ohms. Transformants were isolated on TAP agar containing 20 µg ml⁻¹ hygromycin and screened by Flag immunoblot analysis to identify MARS1 transgene expressors.

Due to the low abundance of the Mars1 protein, the Alexa Fluor 488 Tyramide SuperBoost Kit (Thermo Fisher Scientific) was utilized to amplify the signal. WT (mock control) and Flag-tagged strains were grown in TAP medium to logarithmic phase and a 5 ml aliquot of each cell cultures was harvested at 500 x g for 5 min at RT. The supernatant medium was decanted and the cell pellet was resuspended in 1.5 ml of PBS. A 35 µl cell suspension was added to wells on a slide pre-treated with 0.1% poly-L-lysine and allowed to adhere for 7 min. To solubilize the chlorophyll and maintain cell structure, slides were incubated in 100% methanol 2 times for 4 min at −20°C. Excess methanol was removed and the slides were dried for 2 min at RT. Slides were incubated in PBS supplemented with 0.1% Tween 20 (PBS-T) for 10 min to permeabilize the cell before adding the manufacturer-provided 3% H₂O₂ solution to quench endogenous peroxidase activity for 1 hr. Following three washes with PBS, non-specific signal was blocked with 10% goat serum in PBS (Jackson ImmunoResearch Laboratories) for 1 hr. Samples were incubated with commercial monoclonal mouse anti-Flag antibody (M2 Sigma, F1804, diluted 1:500) in combination with one of three subcellular markers: anti-Histone H3 (Agrisera, AS10 710, diluted 1:500), anti-AtpD (Agrisera, AS10 1590, diluted 1:500), or anti-Nab1 (Agrisera, AS08 333, diluted 1:500), and left overnight at 4°C in a humid chamber. The remaining steps were performed as per manufacturer’s instructions. In brief, samples were washed for 10 min in PBS for a total of three washes. Alexa Fluor 488 poly- horseradish peroxidase (HRP)-conjugated goat anti-mouse and Alexa Fluor 546 goat anti-rabbit secondary antibodies were utilized for detection of the Flag and subcellular marker signals respectively. Alexa Fluor 546 was diluted 1:500 in the manufacturer-prepared Alexa Fluor 488 solution and added to each well for 1 hr at RT protected from light. Slides were washed 3 times in PBS before the amplification step. To amplify the Flag signal, 35 μl of the Tyramide Working Solution was added to the cells and the reaction proceeded for 4 min before an equal volume of Stop Solution was added to end the reaction. Slides were rinsed in 1x PBS and covered with a thin layer of 0.5% low-melting point agarose dissolved in TAP medium before observation by 3D-Structured Illumination Microscopy (3D-SIM).

The microscopy samples were observed using an Elyra PS.1 SIM microscope (Zeiss) with objective lens alpha Plan-Apochromat 100x/1.46 oil (Immersol 518F/30°C, Zeiss), as described previously (Iwai et al., 2018). The fluorophores Alexa Fluor 488 and Alexa Fluor 546 were excited with a 488 nm laser and 561 nm laser, and the fluorescence was acquired through a 495–550 nm and 570–620 nm bandpass filters, respectively. Image acquisition was performed with ZEN software (Zeiss). Each focal plane for 3D-SIM image was captured sequentially by the excitation with the patterned light of 3 rotated angled, each of which contains five shifted phases. The optimal z-interval distance was set to 101 nm. Raw SIM images were processed to reconstruct 3D-SIM images using ZEN software. Extraction of the intensity data was done using the SIMcheck plugin for ImageJ software (Ball et al., 2015).

For localization studies, cytosol-enriched and cytosol-depleted fractions were isolated from mars1-3 cells (a cell-wall deficient strain) complemented with the MARS1-A transgene, according to the protocol described below, incorporating guidelines previously described in Klein et al., 1983 and Zerges and Rochaix (1998). A one liter liquid culture, synchronized by growth in dark-light cycles in minimum media and in early exponential phase (1–2 × 10⁶ ml⁻¹), was harvested at 3000 x g for 5 min at RT. Upon media removal, the cell pellet was resuspended by gentle hand-shaking of the centrifugation tube without using a pipette in 15 ml of autolysin freshly supplied with 1 mM potassium phosphate buffer (pH 6.0) and 0.5 mg/ml BSA (Solution A). Solution A was pre-warmed at 30°C for 30 min prior to use and, after resuspending the pellet in this solution, cells were transferred to a 200 ml 30°C-prewarmed beaker immersed in a water bath and incubated at 30°C for 50 min. Next, cells were transferred to a 50 ml Falcon tube and collected by centrifugation at 2000 x g for 5 min at RT. The autolysin was quickly removed using a 25 ml plastic pipette and the cell pellet was very gently resuspended in 20 ml of ice-cold Solution B consisting of 5 mM K-phosphate buffer (pH 6.5), 6% PEG (w/w) and 4 mg/ml BSA.

Cells were then transferred to a 32°C-prewarmed beaker immersed in a water bath at 32°C and 80 µl of freshly-prepared 1% digitonin (Calbiochem) (0.004% final concentration) was quickly added and well-mixed with the cell suspension.

Cells were then subjected to two rapid warming-cooling cycles using 32°C pre-warmed or ice pre-chilled beakers to induce plasma membrane rapture and cytosolic protein release without intracellular organelles breakage. Cycling was performed for 2 min at 32°C, 5 min on ice, 1 min at 32°C, 5 min on ice. Next, this suspension of permeabilized cells and released cytosolic proteins was transferred to a 50 ml pre-chilled Falcon tube and centrifuged at 800 x g for 3 min at 4°C. After centrifugation, the cytosol-enriched fraction was further purified from the supernatant fraction, while the cytosol-depleted fraction containing chloroplasts and mitochondria was further purified from the cell pellet.

To remove potential cell debris and obtain a clean cytosolic fraction, the supernatant was subjected to two further consecutive rounds of centrifugation: first, at 5000 x g for 15 min at 4°C and then at 23,000 x g for 1 hr at 4°C. Finally, the cytosol-enriched fraction was precipitated in ice cold acetone containing 10% of trichloroacetic acid (TCA).

To enrich the cytosol-depleted fraction with organelles, the pellet of permeabilized cells was kept in ice and resuspended in 2 ml of ice-cold 2x isotonic solution consisting of 0.6 M sorbitol, 10 mM MgCl₂ and 20 mM Tricine pH 7.8. At this point, a rather dark-green aggregate formed in the falcon tube. To resuspend this aggregate, a cut plastic pipette tip was used to gently pipette the pellet up/down for 20 times. Then, 2 ml of ice-cold milliQ water were added to bring the isotonic solution to 1x and the aggregates were further dissolved as described above.

Next, this suspension was loaded on a Percoll step gradient (10 ml 75% Percoll in isotonic solution/10 ml 45% Percoll in isotonic solution/4 ml cell lysate) in a Corex glass tube. The gradient was subjected to centrifugation using the HB4 swinging-rotor at 7000 x g for 15 min at 4°C. Chloroplasts and contaminating mitochondria were recovered from the interface between 45% and 75% Percoll and diluted in 20 ml of 1x ice-cold isotonic solution. The organelles were collected by 5 min centrifugation at 4000 x g at 4°C and, after removing the supernatant, the pellet was resuspended in 2 ml ice-cold isotonic solution and run through a second Percoll gradient as described above. Finally, the cytosolic-depleted fraction was precipitated in ice-cold acetone containing 10% TCA. Denatured proteins from each fraction were extracted after TCA precipitation, as described below. Prior to gel electrophoresis, the protein content of each fraction was normalized using a bicinchoninic acid (BCA) assay as described in the section below.

Proteins were extracted from whole cell lysate using a denaturing SDS extraction protocol for all experiments except for immunoblots that include the Mars1 protein, in which case TCA precipitation was used. For the SDS protein extraction, cells from a 5 ml culture in exponential growth phase were harvested at 3000 x g for 5 min and resuspended in 150 µl of SDS-lysis buffer (100 mM Tris-HCl pH 8.0, 600 mM NaCl, 4% SDS, 20 mM EDTA, freshly supplied with Roche Protease Inhibitors). Samples were vortexed for 10 min at RT and centrifuged at maximum speed for 15 min at 4°C to remove cell debris. The supernatant, containing a total extract of denatured proteins was transferred to a new Eppendorf tube, a 5 µl aliquot was saved for BCA quantification and 1/4 vol of 5X SDS-loading buffer (250 mM Tris-HCl pH 6.8, 5% SDS, 0.025% bromophenol blue, 25% glycerol), freshly supplied with 500 mM DTT or 5% of 2-mercaptoethanol prior use, was added to the extract and denatured at 37°C for >30 min. For the TCA precipitations, cell pellets were resuspended in 1 ml of 10% TCA in acetone, freshly supplemented with 0.5% β-mercaptoethanol. Samples were vortexed for 10 min at 4°C then left at −20°C for 1–2 hr for efficient protein precipitation. Samples were centrifuged at maximum speed for 10 min at 4°C and the TCA solution was carefully aspirated. Three washes with 1 ml of cold 100% acetone were performed (5 min of vortexing followed by 5 mins of maximum speed centrifugation) and the remaining pellet was dried for 5–10 min before resuspension in Lysis Buffer (same as above), achieved through vigorously shaking of the Eppendorf tube with the aid of a vortex at RT or with the aid of a thermomixer at 50°C for 10–15 min. The resuspended protein extract was isolated by a quick centrifugation and was transferred to a new Eppendorf tube. A 5 µl aliquot was saved for BCA quantification and 1/4 vol of 5x SDS loading buffer was added to the rest and denatured at 37°C for at least 30 min. Immunoblot analysis was performed on 20–60 µg of denatured protein extract. Proteins were separated by SDS-PAGE using Mini-PROTEAN or Criterion Precast Gels) (Bio-Rad) and transferred onto Protran nitrocellulose membrane, 0.2 µm pore (Perkin Elmer). Non-specific signal was blocked with PBS-T supplemented with 5% instant nonfat dry milk (Carnation, Nestlè) for 1 hr at RT or overnight at 4°C. All primary and secondary antibodies were diluted in this blocking buffer. The following antibodies (at the indicated dilution) were used for this publication: monoclonal mouse anti-Flag (1:3,000) (M2, Sigma F1804), monoclonal mouse anti-α-tubulin (1:5,000) (Sigma #T6074), polyclonal rabbit anti-DnaK (provided by Jean David Rochaix) (1:10,000) (H.Naver, K.Wilson and J.D.Rochaix, unpublished results) (Dauvillee, 2003), polyclonal rabbit anti-RpoA (1:10,000) (Ramundo et al., 2013), polyclonal rabbit anti-ClpP1 (provided by Francis-André Wollman and Olivier Vallon) (1:5,000) (Majeran et al., 2000), polyclonal rabbit anti-Hsp22E/F (provided by Michael Schroda) (1:5,000) (Rütgers et al., 2017), polyclonal rabbit anti-Vipp2 (1:3,000) (raised against a –CDPLERELEELRRRARE- peptide, developed during this study by Yenzym, South San Francisco), polyclonal rabbit anti-Aox1 (1:2,000) (Agrisera AS06 152), polyclonal rabbit anti-Sultr2 (provided by Arthur Grossman) (1:3,000) (Pootakham et al., 2010), polyclonal rabbit holo-Rubisco (provided by Jean David Rochaix) (1:10,000) (Borkhsenious et al., 1998), polyclonal rabbit anti-Hsp90C (1:10,000) (Agrisera AS06 174) and anti-Histone H3 (1:10,000) (Agrisera AS10 710). To detect the primary antibodies, HRP-conjugated anti-rabbit and anti-mouse secondary antibodies (Promega) were used at dilution 1:10.000 in PBS-T supplemented with 5% instant nonfat dry milk for 1 hr at RT. In between the incubation with primary and secondary antibody and after the incubation with the secondary antibody, three washes of about 10 min each time, at RT, were performed using PBS-T supplemented with 5% instant nonfat dry milk. The membranes were quickly rinsed three times with milliQ-water and a luminol-based enhanced chemiluminescence (ECL) method was applied to develop the signal. For most immunoblot analysis, the SuperSignal West Dura Extended Duration Substrate kit (Thermo Fisher Scientific) was used according to manufacturer’s directions. By contrast, for Flag immunoblot analysis to detect Mars1 protein, the SuperSignal West Femto Maximum Sensitivity Substrate kit (Thermo Fisher Scientific) was chosen, given the low expression level of this protein. The ECL signal was detected with the LI-COR Odyssey imaging system or using clear-blue X-ray films (CL-Xposure, Thermo Fisher Scientific).

To carry out the BCA assay, 5 µl of protein extract was added to 200 µl of BCA/copper sulfate solution (1:50 dilution of 4% CuSO₄ into BCA solution, Sigma) and incubated at 50°C for 5 min. Protein concentration was estimated by measuring the absorbance at 562 nm and comparing it to a BSA standard.

Note: for Figure 6—figure supplement 1B, the denaturing protein extraction was carried out as follows: cell cultures started in 7 ml of TAP were grown to mid-log phase and subsequently spun down at 3000 x g for 8 min. The pellets were resuspended in 150 µl of TAP. Then, an equal volume of 0.2M NaOH was added to the pellets, vortexed at RT for 5 min and pelleted at 15,000 x g for 5 min. The supernatant was removed, the pellet was resuspended in 280 µl of SDS samples buffer (0.06 M Tris-HCl pH 6.8, 5% glycerol, 2% SDS, 4% 2-Mercaptoethanol, 0.0025% bromophenol blue), boiled for 5 min and then pelleted again. A 28 µl aliquot was loaded onto a Criterion gel.

The following protocol was used for the HL assay described in Figure 4A–B. Slight modifications were applied during the HL assays described in Figure 6E and will be underscored below. Liquid cultures were started from TAP plates that had all been grown in dim light and in the same conditions. A small slab of cells was taken from the agar plate and resuspended in 28 ml of TAP media in 50 ml falcon tubes (Sarstedt). The Olympus Plastics brand product line from Genesee Scientific was avoided because there was a higher propensity for the cells to adhere to this plastic material. An equal slab of cells was used for each culture to approximate the same level of starting cells. Typically, the strains had been growing in fresh TAP agar for ~3–5 days. The cells were pre-conditioned in low light (~20–50 µmol photons m⁻² s⁻¹) for about 38–44 hr. For the HL assays shown in Figure 6E, the cells were preconditioned for slightly longer period,~3 days. The chlorophyll concentration of the cell cultures was measured using the methanol extraction method as described in Porra et al. (1989). At the above described time points, it was found to be ~13–18 µg/ml (HL assays described in Figure 4A–B), or ~25 µg/ml (in the HL assays described in Figure 6E). Cell cultures were then equally diluted to ~10 µg chlorophyll ml⁻¹ (in the case of the HL assays described in Figure 4A–B), or to 7 µg chlorophyll ml⁻¹ (HL assays described in Figure 6E). Chlorophyll concentrations were confirmed and, if needed, re-adjusted after dilution before HL was started. The final volume of cell culture used for high light treatment was ~26 ml in 50 ml Falcon tubes. During the high light treatment, the distance between the light source (Phlizon 2017, 2000W Plant LED Growth light) and the shaker was set to 25 cm. The HL intensity was measured at 1100 µmol photons m⁻² s⁻¹ at the beginning of the experiment but was reduced to ~900 µmol photons m⁻² s⁻¹ by the end of the experiment. On the right and left side of the shaker, two fans were turned on to keep the samples at RT and a Smart Sensor (SensorPush) was used to monitor temperature in real-time. Typically a 4°C increase in temperature (from 24.5°C to about 28.5°C) was measured after the cultures were shifted from the dim light growth setup to this HL setup. The light intensity at each position of the culture on the shaker was measured to ensure cells were getting the same number of photons (~50,000 LUX). The cultures in the Falcon tubes were taped (clear tape) onto the shaker and the shaker was set at 150 rpm. Chlorophyll measurements were taken at multiple time points of HL treatment (during HL assays described in Figure 4A and Figure 4—figure supplement 1A) and after 27 hr or 50 hr of HL treatment (during HL assays described in Figure 4B and Figure 6E, respectively). Serial dilutions performed on cultures before and after the HL treatment were spotted onto TAP plates. Photographs of these plates were taken over time to track cell recovery.

For immunoblot analyses, as shown in Figure 4D, cell cultures were grown in liquid TAP medium in a 50 ml Falcon tube for about two days to a chlorophyll concentration of 11 µg/ml in a volume of 30 ml. Cells were spun down and the pellet was resuspended in 1.5 ml of TAP. 0.3 ml of the resuspended pellet was then added to 10 ml of TAP with or without 1.1 mM metronidazole (Sigma). These cultures were then placed under white light (20–50 µmol photons m⁻² s⁻¹) for 12–15 hr and then spun down and saved at −20°C or directly used for denaturing protein extraction as described above.

For growth tests on TAP agar supplemented with metronidazole, as shown in Figure 4C and Figure 4—figure supplement 1B–E, strains were either manually re-streaked or robotically replicated from TAP agar plate to +/- 1.5 mM metronidazole TAP agar plates.

For dilution spot tests, shown in Figure 6D, cells were freshly inoculed from a 3-4 days old agar plate and grown in in falcon tubes for 2 days in a starting volume of 30 ml of TAP. After these 2 days of preconditioning, chlorophyll concentrations were normalized to be at ~18 µg/ml and serial dilutions of 1.5-fold were done in liquid TAP using a 96-well plate. Finally, 6 µl cells from each dilution series were spotted onto TAP agar plates with or without 2.2 mM metronidazole. Photographs were taken periodically to track growth over time.

Metronidazole is rather insoluble in aqueous solutions; therefore, it was always added (as powder) directly to the autoclaved liquid TAP medium at final concentrations of 1.1 mM, 1.5 mM or 2.2 mM.

Mars1-D and Mars1-D KD strains were subjected to metronidazole treatment for 15 hr. Culture were then harvested and subjected to Flag affinity purification followed by MS analysis according to the protocol described by {Mackinder, 2017 #30} and publically available through this link: https://docs.google.com/viewer?a=v&pid=sites&srcid=ZGVmYXVsdGRvbWFpbnxjaGxhbXlzcGF0aWFsaW50ZXJhY3RvbWV8Z3g6NzlkNjUzMTM0ZWEyYmI5. Preparation of samples for MS analysis and processing of MS raw data was performed by the Stanford Mass Spectrometry Facility in Palo Alto.

A 10–ml aliquot of a cell culture in early-mid exponential phase (1–5 × 10⁶ ml⁻¹), was harvested at 3000x g for 5 min at RT. After decanting the media, 1 ml of Trizol Reagent (ThermoFisher Scientific) was added to the cells. The cells were lysed in sample by vigorous shaking with the aid of a vortex for 5–8 min. Chloroform (1/5 vol,~200 µl) of was then added to the lysate and the tube was vigorously shaken up and down by hand for 60 s. The sample was then centrifuged at 11,000 x g for 7 min at RT. The upper aqueous phase (~350 µl) was removed with care to not draw any of the organic layer, transferred in a nuclease-free 1.5 ml tube (Ambion) and mixed well with 1 vol of 100% ethanol (Fisher Scientific) at RT. From this point onwards, the RNA purification was carried out with the aid of the Direct-zol RNA MiniPrep Plus kit (ZymoResearch) following the manufacturer’s protocol, including in-column DNase I treatment prior RNA washing and elution steps.

A semi-quantitative RT-PCR was carried out to qualitatively determine the presence or absence of MARS1 gene transcripts. Total RNAs were extracted as described in the previous section. Complementary DNA was synthesized from 1 µg of total RNA using PrimeScript 1 st strand cDNA Synthesis Kit (Takara) as per manufacturer’s instructions. Subsequently, RNA/DNA hybrids were removed by ribonuclease H treatment (NEB) as per manufacturer’s instructions. cDNAs were diluted two-fold and 1–2 µl were used as template for a 20 µl PCR reaction by Phusion High-Fidelity DNA Polymerase (Thermo Fisher Scientific) with the following parameters: initial melting 98°C for 30 s, amplification cycles 98°C for 10 s, 68°C for 30 s, 72°C for 1 min 15 s (35 times), final extension 72°C for 5 min. Primers SR834 and SR835 were used to amplify a fragment of the MARS1 coding sequence spanning from exon 16 to exon 28. Primers SR836 and SR837 were used to amplify the GBLP coding sequence, used as loading and positive control during the RT-PCR analysis. The sequence of the primers can be found in Table 3.

For qPCR analyses, the cDNA was prepared as described above but the cDNA was diluted six to eight-fold in nuclease-free water prior to use. Primers to amplify the target transcripts are indicated in the Table 3. The qPCR reactions were carried out using the iQ Sybr green supermix as per manufacturer’s instructions (Bio-Rad). To determine whether there was DNA contamination in the mix or whether there was primer dimer formation or misannealing, the same volume of the master mix, without cDNA, was added to a well in the 96-well plate. The raw Ct values were analyzed per the ‘eleven golden rules’, as previously described (Udvardi et al., 2008). GBLP was chosen as reference housekeeping transcript during normalization. Standard deviation was obtained for the 2–5 technical replicates and a minimum of 2 biological replicates were done per experiment, except for the experiment in Figure 3—figure supplement 1B, where only one biological replicate was analyzed.

The Phytozome v5.5 gene annotation for the target transcripts is as follows: VIPP2 (Cre11.g468050.t1.2), SULTR2 (Cre17.g723350.t1.2), LHCBM9 (Cre06.g284200.t1.2), HSP22F (Cre14.g617400.t1.1), SNOAL (Cre11.g478100.t1.2), LHCSR3.1 (Cre08.g367500.t1.1), PSBS1 (Cre01.g016600.t1.2), CPLD29 (Cre02.g088500.t1.1), GBLP (Cre06.g278222.t1.1).

Prior to RNA-seq analyses, the mars1-1 strain was backcrossed to the wild-type CC-124 (Chlamydomonas Resource Center). A full tetrad was selected and analyzed as shown in Figure 2E. A wild-type and MARS1 mutant progeny (CrPW8, indicated as E12, and CrPW9, indicated as F2 respectively, in Figure 2E and Figure 2—figure supplement 2B) were chosen for follow-up studies, based on the retention of the ClpP1 repressible system in their genetic background.

For each strain, two cultures (biological replicates) were inoculated in 30 ml liquid TAP medium using 50 ml Falcon tubes, starting from a fresh re-streak of cells propagated on TAP agar. These cultures were grown with mild agitation (150 rpm) on a shaker at around 22°C, under an illumination of 30–40 µmol photons m⁻² s⁻¹ for about three-four days till they reach mid-late exponential phase (4–7 × 10⁶ cells ml⁻¹). Next, cells were diluted to about 1 × 10⁶ cells ml⁻¹ in 30 ml of liquid TAP medium using 50 ml Falcon tubes and they were subjected to three different treatments: a) low light, i.e. they were kept at the same light intensity used during conditioning; b) HL, i.e. they were shifted to very high light (1200 µmol photons m⁻² s⁻¹) for 40 min or 70 min; and c) ClpP1 repression, i.e. they were diluted in liquid TAP medium containing 100 µM thiamine and 40 ng ml⁻¹ vitamin B₁₂ and incubated for 68 hr at the same light intensity used during conditioning. Cells were then shifted to ice and quickly harvested by centrifugation at 3000 x g for 5 min at 4°C. Cell pellets were snap-frozen in liquid nitrogen and saved at −80°C till use. RNA extraction was carried as described in the previous section. The following extra measures were taken to ensure a complete removal of DNA contaminants: An additional round of in-solution Dnase I treatment was performed using 1 unit of Rnase-free Dnase I (Roche)/1 µg of total RNA at RT for 20 min in presence of 1 unit of recombinant ribonuclease inhibitors (RNaseOUT, Thermo Fisher Scientific). Next, the RNA was re-purified using the same extraction protocol described above.

Each total RNA preparation was ran on an Agilent Bioanalyzer RNA 6000 Nano chip for quantification and quality control. PolyA mRNAs were purified and RNAseq libraries were prepared using the Kapa mRNA HyperPrep kit (Roche, KK8540) following manufacturer’s protocols. Libraries were pooled based on fragment analyzer concentrations. Sequencing was performed on Nextseq high-output flowcell, 1 × 75 bp run (Illumina).

We used a combination of publicly available tools and custom scripts for the processing of the raw demultiplexed Illumina sequencing data. Illumina adapter sequences were first trimmed off with TrimGalore! (version_0.4.1) (www.bioinformatics.babraham.ac.uk/projects/trim_galore) and contaminating ribosomal reads were removed by mapping against the Silva rRNA database using bbduk v37.32 (part of the BBTools suite, https://jgi.doe.gov/data-and-tools/bbtools/bb-tools-user-guide/). Quality control of raw and processed fastq files was performed using FastQC version 0.11.3 (https://www.bioinformatics.babraham.ac.uk/projects/fastqc/). The remaining reads were aligned to the unmasked C. reinhardtii genome (Phytozome, v5.5) using STAR v.2.5.3a (Dobin et al., 2013) and bam files were sorted with samtools 1.1 (Li et al., 2009). Count generation and downstream analysis were done in R (R project v3.4.0, www.R-project.org) using a combination of the packages Rsubreads (Liao et al., 2013) EdgeR (Robinson et al., 2010) plyr, ggplot2, gplots, and heatmap2. For differential expression analysis, genes with less than 0.5 counts per million reads in less than two samples were discarded, the data were fit to a negative binomial generalized linear model, and differential expression was determined using the quasi likelihood F-test with Benjamini-Hochberg correction of multiple testing in EdgeR. To subdivide genes into groups of MARS1-dependent and -independent genes, weakly-expressed genes (average RPKM <2.5 in at least two conditions) were discarded. Stress-responsive genes (defined as log₂-fold change >= |2| at FDR 0.001 upon treatment - high light, ClpP repression, or both as noted – in WT background) were considered MARS1-dependent when treatment in the MARS1 mutant did not lead to a greater than 2-fold change in expression (log₂-fold change < |1|). MARS1-independent genes were defined as genes with an at least 4-fold (log₂-fold change >= |2|) change upon treatment in the mars1 background, in the same direction (up or down) as the response in WT cells. For analysis of functional categories in MapMan 10.0 (Thimm et al., 2004, Usadel et al., 2005) the C. reinhardtii v5.5 proteome (downloaded from Phytozome, v5.5, protein sequences from primary transcript only) was binned using Mercator4 (www.plabipd.de/portal/web/guest/mercator4) and MARS1-dependent and MARS1-independent gene subsets were mapped onto plant pathways in MapMan.

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

We thank Jean David Rochaix for his generous donation of key reagents and critical feedback on this work, Olivier Vallon, Michael Schroda and Arthur Grossman for providing the ClpP1, Hsp22E/F and Sultr2 antibodies, respectively. We thank Fran Sanchez for preparing growth media for maintenance of our strains, Luke Mackinder, Nina Ivanova, Julia Wei for help with the genetic screen, and Lorenzo Costantino, Susan Dutcher, Stephane Gabilly, Ursula Goodenough, Robert Jinkerson, Yuval Kaye, Elif Karagoz, Xiaobo Li, Daniel Neusius, Jörg Nickelsen, Masayuki Onishi, Leif Palessen, Weronika Patena, Jirka Peschek, Shai Saroussi, Setsuko Wakao, Lan Wang and Tyler Wittkopp for invaluable advice in various aspects of this project. We thank Ryan Leib and Kratika Singhal who curated Mass-Spec analysis for this work and all Walter and Jonikas lab members for insightful discussions on this project.

© 2019, Perlaza et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.