Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2018 Jun;30(6):1178-1198.
doi: 10.1105/tpc.18.00071. Epub 2018 May 9.

Comprehensive Discovery of Cell-Cycle-Essential Pathways in Chlamydomonas reinhardtii

Affiliations
Free PMC article

Comprehensive Discovery of Cell-Cycle-Essential Pathways in Chlamydomonas reinhardtii

Michal Breker et al. Plant Cell. .
Free PMC article

Abstract

We generated a large collection of temperature-sensitive lethal mutants in the unicellular green alga Chlamydomonas reinhardtii, focusing on mutations specifically affecting cell cycle regulation. We used UV mutagenesis and robotically assisted phenotypic screening to isolate candidates. To overcome the bottleneck at the critical step of molecular identification of the causative mutation ("driver"), we developed MAPS-SEQ (meiosis-assisted purifying selection sequencing), a multiplexed genetic/bioinformatics strategy. MAPS-SEQ allowed us to perform multiplexed simultaneous determination of the driver mutations from hundreds of neutral "passenger" mutations in each member of a large pool of mutants. This method should work broadly, including in multicellular diploid genetic systems, for any scorable trait. Using MAPS-SEQ, we identified essential genes spanning a wide range of molecular functions. Phenotypic clustering based on DNA content analysis and cell morphology indicated that the mutated genes function in the cell cycle at multiple points and by diverse mechanisms. The collection is sufficiently complete to allow specific conditional inactivation of almost all cell-cycle-regulatory pathways. Approximately seventy-five percent of the essential genes identified in this project had clear orthologs in land plant genomes, a huge enrichment compared with the value of ∼20% for the Chlamydomonas genome overall. Findings about these mutants will likely have direct relevance to essential cell biology in land plants.

Figures

None
Figure 1.
Figure 1.
Distinct Arrest Morphologies Differentiate Mutants. Nitrogen-depleted synchronized samples were plated on TAP agar plates and were transferred to restrictive temperature (33°). Images were taken every hour by light microscopy. (A) Wild-type sample. Cells grow over the course of the first 10 h and then enter rapid cycles of division and reach clusters of cells, as demonstrated at the 12-h time point, followed by hatching of new born cells. (B) Some mutants fail in cell growth (Non-growers). Other mutants (Large round) grow but fail to make cleavage planes (CP). Mutants making cleavage planes (arrows) frequently lose cell integrity a few hours later (Tulin and Cross, 2014); some mutants (Arrested clusters) stay morphologically intact. White arrows mark the initiation of CPs. *Due to differences of focal planes of the images of ess123-1, cell size may seem erroneously shrinking.
Figure 2.
Figure 2.
Combinatorial Framework to Efficiently and Accurately Sequence Pools of Mutants. (A) A set of mutants is distributed in overlapping fashion across many pools (in the illustration, 28 mutants in eight pools, A–H). (B) Each mutant has a unique occupancy pattern. Mutant numbers occupying these pools (C and D) are represented in black and gray. Minority reads shared exclusively between two pools allow assignment of a mutation uniquely to the relevant mutant (marked in black). (C) Representative numbers of mutations identified in a set of eight pools. “All assigned” (single-nucleotide polymorphisms) are the number of total genetic lesions that were identified in the set. “CDS-involved” is the number of genetic lesions that change the annotated coding sequence. The complementary number is the “Non-CDS.” The number in parentheses is the average per mutant. (D) Example for a detected single-nucleotide polymorphism. Minority reads “C” were identified uniquely in C and D pools (compared with the majority reads “T”) and were assigned uniquely to mutant 14. Together with the adjacent mutation, the lesion (C>T) results in a conversion from leucine to serine in amino acid 151 of CDC20.
Figure 3.
Figure 3.
Mass Mating of Ts− Mutants Followed by Selection of Ts+ Meiotic Segregants: Behavior of Causative Mutations and of Linked and Unlinked Passengers. (A) Sets of 42 Mat+ and 42 Mat– Ts− mutants were crossed in a mass mating (see Methods) and allowed to form zygospores (premeiotic diploids; ∼500 zygospores per mutant). Transfer to light triggers meiosis and the formation of four haploid progeny per zygospore. According to Mendelian inheritance for unlinked lesions, the expected result is that 50% of these haploids will inherit a single Ts-lethal mutation from one parent (mutant A or mutant B), 25% will inherit Ts-lethal mutations from both parents, and 25% will inherit no Ts-lethal mutations and will therefore be Ts+. (B) Passenger (neutral) mutations should appear in the Ts+ pool at a frequency of the reciprocal of the total number of mutants in the pool (since any neutral chromosome in the pool is equally likely). However, the rare causative lesions should be entirely depleted, and passenger mutations linked on the same chromosome with the causative lesion will be detected at progressively decreasing frequency as the mutation gets closer to the causative mutation. Rarely, very tightly linked passengers exhibit complete depletion and are not distinguishable from linked causative mutations. In the diagram, the red star represents a causative mutation from some mutant; green or orange indicates a chromosome lacking or containing a causative mutation. Mutations from the causative chromosome that are distant from the causative mutation can readily recombine onto a neutral chromosome. The closer the neutral mutation is to the causative mutation, the lower its frequency in the Ts+ pool (schematics at right). (C) Tetrad analysis of random zygospores of the pools confirms that most parental diploids contain two unlinked Ts− mutations, as expected. PD, parental ditype: 0 Ts+:4 Ts−; TT, tetratype: 1 Ts+:3 Ts−; NPD, nonparental ditype: 2 Ts+:2 Ts−. The net yield of Ts+ is 23% (expected 25%). (D) Plotting frequencies of all predefined lesions on two chromosomes (12 and 16) within the Ts+ pool (far left) and after computational extraction of the mutations on these chromosomes specific to mutants A and B. For Chr. 12-Mutant B or Chr. 16-Mutant A, all mutations are passengers; Chr. 12-Mutant A and Chr. 16-Mutant B exhibit the pattern indicative of a causative genetic lesion: decreasing frequency along the chromosome, up to complete depletion (a V shape) (compared with expectation sketched in Figure 2B). The expected slope of this V shape is dependent on the mutant proportion within the pool and the conversion between physical and genetic distance. In Chlamydomonas, the conversion is ∼10 cM per 1 Mb (Merchant et al., 2007; Tulin and Cross, 2014), resulting in an expected slope of ∼0.6 for an average coverage of 3%, approximately as observed.
Figure 4.
Figure 4.
Sequencing of Ts+ Meiotic Progenies Pool Allows Multiplexed Identification of Causative Genetic Lesions. (A) In most (77%) mutants, such a V shape is observed in exactly one of the 17 chromosomes. Mutants were classified as “single-depleted region” mutants as shown. Once the depleted region is identified, close examination of the mutations within this region gave two possible results. In most cases, one or more CDS-changing mutation (marked in blue) was identified and scored as candidate causative mutations. (B) Mutant Seq4-AH chromosome 12. Arrow indicates candidate causative CDS-changing mutation. This is the only CDS-changing mutation that is completely depleted. (C) Rarely, no CDS-changing (marked in red circles) mutations were identified in the depleted region (Seq1-CE chromosome 7). The latter constitutes the “unsequenceable” class.
Figure 5.
Figure 5.
Representative Yield from Applying MAPS-SEQ to a Large Collection of Ts-Lethal Mutants. (A) Seventy-seven percent of 340 mutants tested with this procedure were identified with a single depleted (Dep.) region, and for 82% of them, MAPS-SEQ enabled the identification of the exact genetic lesion. Thirteen percent of these were “known genes” (KG), essential genes identified in our previous work (Tulin and Cross, 2014); 50% are “new genes” (NG), candidate essential genes identified in this study; 14% of mutants are “unsequenceable” (US), with a single depleted region, but with no identified CDS-changing mutation. The rest of the mutants (23%) represent poor maters, mutants without depletion, or mutants with multiple depleted regions (PM, ND, and MD, respectively). These are discussed in the text. (B) Within the new genes group, 90% have a single candidate causative lesion, and for 10%, we have detected more than one candidate lesion (usually two), likely representing a tightly linked passenger(s) (see Figure 3).
Figure 6.
Figure 6.
Conservation and Evolutionary Constraints Confirm the Likelihood of Candidate Causative Lesions. As described (Tulin and Cross, 2014), we classify mutations according to sequence conservation and severity. Category A: mutation falls within a segment of BLAST alignment to Arabidopsis (HSP) and alters a conserved residue (Blosum62 > 0 for Chlamydomonas versus Arabidopsis) within this segment. Category B: mutation falls within an overall conserved region but alters an unconserved residue (Blosum62 ≤ 0). Category C: mutation lies N-terminal or C-terminal to all detected HSPs. Category D: mutation in gene with no Arabidopsis BLAST hit. Bar graphs demonstrate the distribution of the identified mutations in this work according to the classified groups. Within each category, mutations are classified according to severity: severe mutation (Blosum62 <−1); less severe mutation (Blosum62 ≥−1). (A) Distribution of all mutations (including both the depleted ones from the Ts+ pool and presumably passenger mutations that appear in the pool). (B) Distribution of mutations depleted in the Ts+ pool (potential drivers) within genes that were previously verified to be essential (Tulin and Cross, 2014) (known genes). (C) Distribution of mutations depleted in the Ts+ pool that are in genes newly identified in this work (new genes). (D) Distribution of depleted mutations for mutants that were assigned with more than one candidate (in nearly every case these mutations are closely linked). The distribution in (D) suggests bimodality (a combination of the patterns in [A] and [B]), as would be expected for a mixture of driver and linked passenger mutations in this class (see text). (E) The essential genes set (middle) is strongly enriched (78%) for orthologous genes in Arabidopsis compared with the total Chlamydomonas proteome (left) (P < 0.00001). In contrast, the nonessential gene set (right; defined as genes with passenger mutations that are candidate null alleles; Supplemental Data Set 1) is depleted for orthologous gene content compared with the total proteome (left) (P < 0.00001).
Figure 7.
Figure 7.
Functional Grouping of Mutants Based on Hierarchical Clustering of DNA Content Profiles. Dot plots of DNA content (intensity of SYTOX green staining) and cell size measured by forward scatter (FSC). Gates are applied to extract the number of cells in each DNA peak. Gate colors: gray, newborns; cyan, 1C large (pre-S phase); red, 2C; orange, 4C; purple, 8C; brown, 16C; black (dotted circle), cell aggregates. (A) In a wild-type sample there are multiple peaks of DNA content, reflecting sequential rounds of replication. (B) The cdc20 mutant arrests with mostly 2C. (C) Numbers of cells in each gate are extracted; accordingly, each sample has a characteristic vector of numbers. (D) Hierarchical clustering of all samples based on the DNA content profiles. Letters on the right indicate the definitive groups that emerge from the clustering. Letters below the clustergram indicate the discreet peaks of DNA content. NB, newborn; Agg, aggregates; Db, debris; Int, intermediates. The color key on the left provides the standard deviations from the mean (=0) for the normalized values in the clustergram.
Figure 8.
Figure 8.
Mutants in APC Subunits and Some DNA Replication Factors Arrest after the First Round of Replication. DNA-stained samples were analyzed by flow cytometry, and gates were applied as described in Figure 7. Hierarchical clustering of all samples tested reveals distinct functional groups, among them arrested samples at 2C that are mostly composed of APC subunits (A) and DNA replication factors (B). NB, newborn; Agg, aggregates; Db, debris; Int, intermediates. The color key on the left provides the standard deviations from the mean (=0) for the normalized values in the clustergram.
Figure 9.
Figure 9.
Imaging Flow Cytometry Used to Determine a Ploidy Ratio. (A) Diagram of the Chlamydomonas cell cycle. (B) and (C) Wild-type (B) and spc25 (C) samples analyzed by image stream provide a similar pattern of discrete peaks, suggesting a complete rounds of replication, as analyzed by conventional flow cytometry. The table provides the values for the average number of nuclei (AveNuc.) under each peak and the calculated DNA content per nucleus in each peak accordingly. Representative images for the wild type demonstrate the nuclei distribution and size within one cell before and after segregation (segregation was essentially unobserved in the spc25 mutant at any nuclear ploidy level). (D) Distribution of DNA content per nucleus calculated for all mutants. Black arrow marks wild-type value. (E) Mutants whose nuclei have average DNA content >2C were enriched for microtubule synthesis and spindle formation components (Supplemental Data Set 2; orange diamonds in [F]). Their DNA content per nucleus in each peak increases with each DNA replication round (colored lines), unlike the wild type, which maintains normal nuclear ploidy throughout the replication cycles (black lines). (F) For sensitive detection of defective ploidy control, we examined 71 mutant samples for which at least ∼10% of cells were in the 4C peak. We examined images captured by Amnis for these 4C cells specifically and determined the proportion of mononucleates. These images provide one orientation and one plane of focus from a cell; a binucleate cell will necessarily get called as mononucleate at some frequency. We attribute the large signal at ∼0.2 to this background; therefore, a signal above 0.25 is likely specific to a defect in ploidy control. Some of the discussed genes are marked. Orange diamonds correspond to mutants in (E) with average DNA content >2C per nucleus.

Comment in

Similar articles

See all similar articles

Cited by 4 articles

Publication types

MeSH terms

LinkOut - more resources

Feedback