Chlamydomonas reinhardtii as a plant model system to study mitochondrial complex I dysfunction

Abstract Mitochondrial complex I, a proton‐pumping NADH: ubiquinone oxidoreductase, is required for oxidative phosphorylation. However, the contribution of several human mutations to complex I deficiency is poorly understood. The unicellular alga Chlamydomonas reinhardtii was utilized to study complex I as, unlike in mammals, mutants with complete loss of the holoenzyme are viable. From a forward genetic screen for complex I‐deficient insertional mutants, six mutants exhibiting complex I deficiency with assembly defects were isolated. Chlamydomonas mutants isolated from our screens, lacking the subunits NDUFV2 and NDUFB10, were used to reconstruct and analyze the effect of two human mutations in these subunit‐encoding genes. The K209R substitution in NDUFV2, reported in Parkinson's disease patients, did not significantly affect the enzyme activity or assembly. The C107S substitution in the NDUFB10 subunit, reported in a case of fatal infantile cardiomyopathy, is part of a conserved C‐(X)11‐C motif. The cysteine substitutions, at either one or both positions, still allowed low levels of holoenzyme formation, indicating that this motif is crucial for complex I function but not strictly essential for assembly. We show that the algal mutants provide a simple and useful platform to delineate the consequences of patient mutations on complex I function.


List of Supplemental
Supplemental Figure S7. The wild-type NUOB10 gene restores heterotrophic growth to an amc5 mutant.
Supplemental Figure S8. Alignment of NUOB10 / NDUFB10 / PDSW subunits. for at least 12 -16 hr. The meiotic progeny was obtained through bulk germination or tetrad dissection of zygotes.
Individual diploids or meiotic progeny were subcloned to a single colony and their mating type was determined by diagnostic PCR (Werner & Mergenhagen, 1998).

Supplemental Method S3: Ten-fold dilution series and growth curve analysis
The SID phenotype was tested by ten-fold dilution series as follows. One loop of cells grown on solid TARG plates (for 3-5 days) was resuspended in 500 µl of liquid TARG medium. The cell density was measured spectrophotometrically at 750 nm and normalized to an OD750 = 2.0 by dilution. This normalized suspension was used as the starting material [1] for making five serial ten-fold dilutions [10 -1 , 10 -2 , 10 -3 , 10 -4 , and 10 -5 ]. A volume of 8 µl for each dilution was plated on solid TARG plates. For scoring the SID phenotype, two plates were prepared simultaneously and incubated at 25°C, one in continuous light and another in continuous dark, for at least 7 days. The light-incubated plate served as a control for confirming equal cell density amongst multiple strains.
To determine the generation time, liquid cultures were inoculated with a starting cell density of 10 5 cells. mL -1 in 50 mL TARG medium. Growth in liquid culture was observed by evaluating cell density at A750.
Measurements were taken every 8 hr over a period of 10 days. For each strain, three biological replicates were inoculated in continuous light at 25 µmol m -2 s -1 or in continuous darkness. Growth rate µ was calculated as 3.3*[log10N -log10N0] / (tN-t0), where N is the final cell density at time tN and N0 is the initial cell density at time t0. The generation time was calculated as 1/ µ (Harris, 1989, Kropat et al., 2011. The time periods used for calculating doubling times are as follows: For wild-type (4C -) and [amc9; NUO5] strains in the light tN = 86 hr and t0 = 26 hr, and in the dark tN = 110 hr and t0 = 38 hr; for the amc9 strain in the light tN = 110 hr and t0 = 38 hr, and in the dark tN = 170 hr and t0 = 86 hr; for wild-type (3A + ) in the light tN = 96 hr and t0 = 24 hr; for wild-type in the dark tN = 120 hr and t0 = 36 hr; for amc5 in the light tN = 85 hr and t0 = 37 hr; for amc5 in the dark tN = 123 hr and t0 = 49 hr; for [amc5; NUOB10] in the light tN = 85 hr and t0 = 37 hr and for [amc5; NUOB10] in the dark tN = 123 hr and t0 = 37 hr.

Supplemental Method S4: Nucleic acid extraction, diagnostic PCRs and real-time qPCRs
Genomic DNA was extracted from Chlamydomonas by the phenol-chloroform method with some modifications (Sambrook et al., 1989). One or two loops of cells, grown for two to three days in continuous light at 50 µmol m -2 s -1 on TARG solid medium, was harvested and resuspended in buffer (10 mM Tris-HCl, pH 8, 10 mM EDTA, 10 mM NaCl, 15% (w/v) glycerol). Cells were lysed by sonication for 5 s at 9 W output. Proteins were degraded by treatment with 100 µg proteinase K (Invitrogen, 25530049) and the extract was incubated at 55 ºC for one hour. RNA was degraded with 50 µg RNase A (Amresco, 0675-250mg). Nucleic acids were extracted twice by phenol-chloroform and DNA was precipitated by adding 2.2 volumes of ethanol and .1 volume of 3 M Na acetate pH 5.5.
The sequences of primers used for diagnostic PCR are shown in Table S1. For diagnostic PCR analysis, GoTaq Polymerase (Promega, M3008) was used as recommended by the manufacturer. To enhance the amplification of GC-rich regions, 2.5% (v/v) DMSO was used in each reaction in addition to a denaturation temperature of 98 ºC (instead of 95 ºC). For sequencing analyses, PCR products were gelpurified using the NucleoSpin Gel Extraction Kit (Machery Nagel, 740609.25) as per the manufacturer's instructions. The purified PCR product was then cloned into pGEM-T Easy Vector Systems (Promega, A1360) and then sequenced with T7 and SP6 primers flanking the cloning site.
For real-time quantitative PCR, RNA was prepared as in (Newman et al., 1990). RNA was extracted from 2 x 10 8 cells grown in liquid culture. Nucleic acids were extracted twice with equal volume of phenolchloroform (pH 5.0) and RNA was precipitated after overnight incubation at -20°C with 1/3 rd volume of 8 M LiCl. Eight micrograms of RNA were treated with RQ1 RNase-free DNase I (Promega, M6101).

Supplemental Method S5: Plasmid construction
Plasmids expressing the NUO5 and NUOB10 wild-type and mutant genes were generated by recombination-based cloning in yeast. The pRS426-ble plasmid, containing three markers allowing for selection in algae, yeast, and bacteria, was used as the vector (Noor-Mohammadi et al., 2014). It contains the ble gene conferring zeocin resistance for selection in Chlamydomonas, the URA3 gene for selection in S. cerevisiae, and the bla gene conferring ampicillin resistance in E. coli. The pRS426-ble plasmid was linearized with NotI and AleI (pB-NA). The NUO5 and the NUOB10 genes were cloned between the NotI and AleI sites in pRS426-ble. The NUO5 and NUOB10 genes were expressed under their putative native promoters by including the entire intergenic region upstream of each gene (1220 bp and 1075 bp for NUO5 and NUOB10, respectively). Each gene, including its promoter region, was amplified from purified genomic DNA using Velocity DNA Polymerase (Bioline, BIO21098) as consecutive overlapping fragments (Table S2). The sequence corresponding to the C-terminal FLAG-tag and site-directed mutations were introduced by PCR with appropriately designed primers (Table S2). The overlapping PCR fragments and linearized vector (pB-NA) were assembled via in vivo molecular recombination in yeast (Table S3). The S. cerevisiae strain CW04 (Banroques et al., 1986) was used for recombination. Two hundred nanograms of each linear fragment was introduced into the strain by the one-step transformation method (Chen et al., 1992, Saint-Georges et al., 2002 and cells containing the recombinant clones were selected based on uracil prototrophy. Successful recombinants were identified based on diagnostic PCR and restriction digestion, verified by sequencing, and introduced into the respective Chlamydomonas strains by biolistics. However, the ble selection in Chlamydomonas was not successful as it led to the emergence of spontaneous zeocin-resistant colonies in which the transforming DNA was absent. Hence, the ble selection marker was substituted with the APHVII selection marker (iHyg3) conferring HyB resistance (for the NUOB10 clones) or APHVIII selection marker (iPm) conferring Pm resistance (for the NUO5 clones). The ble marker (1.18 kb) was excised from each plasmid with EcoRI and NotI restriction enzymes. The iHyg3 and iPm markers were amplified from the plasmids pHyg3 and pSL18 (Berthold et al., 2002, Depege et al., 2003, respectively, using Velocity DNA Polymerase (Bioline, BIO21098) with primers including a 25 bp overlap with the digested vector (Table S2). Cloning of the new selection markers were achieved using the In-Fusion HD Cloning Kit (Clontech, 639648) according to the manufacturer's protocol (Table S3). The recombinant clones finally used for Chlamydomonas Tables S3 and S4).

Supplemental Method S6: Complex II+III and Complex IV enzymatic assays
Activity assays were conducted on crude membrane extracts that were prepared as described in Materials and Methods section. Complex II+III activity assay was conducted in the presence of 20.25 mM succinate (Acros organics, 158751000), 1 mM KCN (FisherScientific, P223I-100) and 56 µM equine heart cytochrome c (Sigma, 2506-500mg). The activity was determined as the rate of cytochrome c reduction and measured spectrophotometrically at A550. Complex II+III activity was calculated using molar extinction coefficient for cytochrome c at ∆ε550nm = 19.6 mM -1 cm -1 , in the absence and presence of complex III-specific inhibitor myxothiazol (3 µM) (Sigma, T5580). Complex IV activity assay was conducted in the presence of 1% Triton X-100 and 56 µM reduced cytochrome c. Cytochrome c was reduced with two times the amount of sodium dithionite and purified with a PD10-desalting column with Sephadex G-25 resin (GE Lifesciences, 17085101) according to the manufacturer's protocol. Complex IV activity was calculated using molar extinction coefficient for cytochrome c at ∆ε550nm = 19.6 mM -1 cm -1 , in the absence and presence of complex IV inhibitor KCN (1 mM).

Supplemental Method S7: Immunoblotting analysis
BN-PAGE was completed as described in Materials and Methods section. For SDS-PAGE, 10 µg of crude membrane proteins was separated by 12.5% acrylamide gel and immunoblotting was performed according to established protocols (Sambrook et al., 1989). The separated proteins were electro-blotted onto PVDF membranes and custom-made rabbit polyclonal antibodies specific for Chlamydomonas complex I subunits (from Genescript, as described in (Barbieri et al., 2011)), were used. Membranes with electroblotted proteins separated in BN-PAGE gels were probed overnight at 4°C with 1:3000 diluted α-51 kDa, a polyclonal antibody that detects the soluble arm 51 kDa subunit. Membranes containing proteins resolved by SDS-PAGE immunoblots were probed at room temperature for 3 hr with 1:3000 diluted α-51 kDa, 1: 3000 diluted α-49 kDa, 1:2000 diluted α-TYKY and for 1 hr with 1:12,000 diluted α-cytochrome f (Dreyfuss et al., 2003). For rabbit polyclonal antibodies, a HRP-conjugated anti-rabbit Goat IgG (Biorad, 170-6515) at 1:10,000 dilution was used as the secondary antibody. For detecting the FLAG-tagged proteins, membranes were incubated with monoclonal anti-FLAG antibody (Sigma, F1804) at 1:5000 dilution overnight at 4°C followed by a HRP-conjugated anti-mouse Goat IgG (Pierce 18584_3) at a dilution of 1:2,500 as the secondary antibody. Intensity of bands in immunoblots were quantified using ImageJ (Schneider et al., 2012). Figure S1. Additional Blue-Native PAGE immunoblotting analyses.

Supplemental
Blue-Native PAGE followed by immunoblotting was conducted on crude membrane extracts using a polyclonal antibody to detect the 51 kDa subunit of complex I. The white vertical line between the lanes of WT and amc9 denotes the assembly of lanes from the same blot.
Supplemental Figure S2. The amc8 to amc13 mutations are recessive.
To test if the amc mutations causing the complex I-deficient phenotype are recessive or dominant with respect to the wild-type allele, heterozygous diploids (amc/+) were constructed by crossing amc mutants with a wild-type strain as described in Method S2. Two independent diploids from each cross were tested for growth in the dark by ten-fold dilution series. The dilution series were plated on acetate-containing medium and incubated in the light or dark for 10 days. The heterozygous amc/+ diploids were restored for growth in the dark, thereby indicating that all the amc mutations are recessive. White vertical lines indicate pictures of strains grown from the same plate and assembled for the display in the figure. The growth of the following diploids is depicted in panels A-F wherein WT (4C -) and the respective amc mutants, used for constructing the diploids, were used as controls: A, amc8/+; B, amc9/+; C, amc10/+; Figure S3. The amc9 mutation is linked to the insertional cassette.
Meiotic zygotes were obtained by crossing the complex I proficient strain 141 (mt + ; arg9-2) with the complex I-deficient amc9 (3 -) strain (mt -; nuo5::APHVII). The meiotic zygotes, obtained from the same genetic cross, were used for both tetrad dissection (A) and bulk germination (B). (A) The growth phenotype of two out of seven tetrads dissected from the above-mentioned cross is shown here by tenfold dilution series. The dilutions were plated on acetate-containing medium and incubated in the light and in the dark for 15 days. The WT strain and the amc9 strain shown here are the original parental strains.
(B) Fifty recombinant hygromycin B resistant progeny, were analyzed from bulk germination of meiotic zygotes. The progeny were replica-plated on TARG solid medium in the light or dark and the growth phenotype was scored after 15 days of incubation. The WT and amc9 strains shown here are 4Cand amc9 (41D9), respectively. Out of 50 antibiotic-resistant meiotic progeny that were tested, all of them displayed a SID phenotype confirming that the amc9 mutation is tightly linked to the insertional cassette.
13 Supplemental Figure S4. The NUO5 gene encoding the 24 kDa complex I subunit is disrupted in the amc9 strain.
(A) A diagram of the NUO5 gene with the approximate position of the insertional cassette in the amc9 mutant is depicted here. The brown arrows indicate primers APH7R5 and APH7F8 that bind to the APHVII gene in the iHyg3 cassette. The black arrows represent NUO5-specific primers NUO5E1L, NUO5E2R, NUO5E2L, NUO5E2L2, NUO5E3R, NUO5E4L, and NUO5E5R (Table S1)    NUO5 / NDUFV2 / 24 kDa subunit proteins were aligned using Clustal Omega (BLOSUM 62 scoring matrix) and Bioedit (Sievers et al., 2011, Hall, 1999. The conserved cysteines that coordinate the 2Fe-2S

T.thermophilus 175 hvhevev-----------------E.coli 166 ------------------------
(N1a) cluster in NUO5 are highlighted in green. In E. coli, the N1a cluster has a high midpoint redox potential implying a possible role in electron transfer through a flavosemiquinone intermediate (Verkhovskaya et al., 2008). However, in mammalian complex I this cluster exhibits a low midpoint redox potential implying inability for electron relay and hence NDUFV2 is alternatively proposed to function in the structural stability of the enzyme (Birrell et al., 2013, Verkhovskaya et al., 2008. The two cysteines that provide increased stability by forming a disulfide in T. thermophilus ortholog are highlighted in magenta (Sazanov & Hinchliffe, 2006). The conserved lysine residue that is mutated to arginine in the Parkinson's disease patients (Nishioka et al., 2010)  NUOB10 / NDUFB10 / PDSW subunit proteins were aligned using Clustal Omega (BLOSUM 62 scoring matrix) and Bioedit (Sievers et al., 2011, Hall, 1999. The conserved PDSW sequence at the N-terminus is highlighted in green. The two cysteines of the C-(X)11-C motif are highlighted in yellow, the C-(X)6-C reveals the presence of the molecular lesion as previously described in (Barbieri et al., 2011). The [amc5; NUOB10] strain contains an intact copy of the NUOB10 gene as expected. (C) Real -time RT-qPCR was used to assess the quantity of NUOB10 mRNA relative to three reference genes TUA2, CBLP and EIFA.

20
The average was obtained from three biological replicates, each including two technical replicates. The error bars represent standard deviation of the mean. The results are represented as percentage of fold change relative to WT (WT set to 100). Statistical significance was determined with respect to wild-type by two-tailed unequal variance t-test. ** indicates p-value < .01. All primer sequences described in (A -C) are detailed in Supplementary Table S1 Figure S8A. The sequence of primers used to amplify the overlapping DNA fragments by PCR is provided here. The nucleotide sequence encoding the FLAG-tag is in red. The site-directed mutations introduced into the primers are indicated with an underline. Each amplicon is given an alphabetical label for easy reference and its size is also provided in the last column.
The amplicons A to N were generated using Chlamydomonas genomic DNA (WT, 4C -) as template. The iHyg3 and iPm amplicons were amplified using plasmids pHyg3 and pLS18 as templates, respectively (Berthold et al., 2002, Depege et al., 2003.
The PCR fragments described in Table S2 carry overlapping segments that will enable recombinant cloning. The fragments used for generating each clone are detailed here. pB-NA refers to the vector pRS426-ble vector linearized (6864 bp) with NotI and AleI restriction enzymes.
The NUO5 and NUOB10 genes were cloned into the pRS426-ble vector by in vivo recombination in yeast. Subsequently, the ble selection marker was substituted by in vitro recombination using the In-Fusion HD enzyme.
The amc9 and amc5 recipient strains were plated on the specified selective medium (column five) and bombarded with tungsten particles coated with the respective plasmids. Resulting transformants were screened by diagnostic PCR for the presence of the transgene and the site-directed mutations were confirmed by sequencing. The transformant chosen for further analysis is indicated in column one. The recipient strain used for generating each transformant is provided in column two. The plasmids used to generate the transformants and the respective selection markers are detailed in columns three and four, respectively. The Chlamydomonas Resource Center reference numbers for the strains are provided in square brackets.