wild-type cells (4A+) were grown on norflurazon to determine the expected phenotype of pds1
mutants. Norflurazon is a bleaching herbicide that specifically inhibits PDS activity and therefore carotenoid biosynthesis 
. When wild-type cells were grown on norflurazon, dark green cells became light green to almost white with increasing concentrations of norflurazon (). Cell growth was inhibited by norflurazon concentrations above 10 µM in the dark, whereas in low light cell growth was completely inhibited at 5 µM and higher. HPLC analysis of dark-grown cells showed that norflurazon-treated cells accumulated phytoene and had severe reductions in chlorophyll and carotenoids levels (). Phytoene was identified by its absorbance spectrum at 296 nm and its retention time (). Chlorophylls and other carotenoids were detected at 445 nm and also identified by their absorbance spectra and retention times ().
Phenotype of wild-type C. reinhardtii cells grown norflurazon.
A phytoene accumulating mutant: pds1-1
Based on the results of norflurazon inhibition, C. reinhardtii
mutants that are defective in PDS activity were predicted to have a light to very pale green color. From a UV mutagenesis screen, 135 light green, pale green, white, and green/brown color mutants were picked and analyzed by HPLC for pigment abnormalities. The pds1-1
mutant was identified from this screen—it was light green and accumulated phytoene ( and ). However, pds1-1
still produced carotenoids downstream of phytoene including ß-carotene, lutein, antheraxanthin, violaxanthin, and neoxanthin () at ~5% the levels found in wild-type cells (). The pds1-1
mutant accumulated 5-fold more chlorophyll than lts1
mutants but only ~12% the level detected in wild-type cells. The chlorophyll to colored carotenoid ratio for wild-type cells was 3.2
1, whereas in pds1-1
the ratio was 8.7
1. Wild-type and lts1
mutants did not accumulate any phytoene, whereas pds1-1
mutants accumulated significant levels of phytoene ().
Light sensitivity of wild-type, lts1 and pds1 mutants.
Chlorophyll and carotenoid profiles of PDS-activity deficient mutants.
Quantification of chlorophyll and carotenoid content of dark-grown lts1 and pds1 mutants.
Similar to lts1 mutants, the pds1-1 mutant was found to be very light sensitive. After growth in the dark for four days, pds1-1 died after being exposed to more than 24 hours of vLL (). In vLL cells died and turned brown, whereas at higher light intensities (LL and HL) cells bleached completely and turned white (). In contrast, wild-type cells grew well at all light intensities including HL.
Genetic analysis of pds1-1
revealed that the pds1
phenotype is caused by a single, recessive nuclear mutation. Crosses between pds1-1
and wild-type cells produced tetrads that segregated 2
2 for the pds1-1
mutant phenotype (light colored, phytoene accumulation, and reduced levels of colored carotenoids) and the wild-type phenotype (dark green, no phytoene, and normal levels of carotenoids) (). Dominance testing using heterozygous pds1-1/PDS1
vegetative diploids showed the pds1-1
mutation is recessive.
Tetrad analysis of pds1-1 and pds1-3 crossed to wild-type.
The pds1-1 mutant was crossed to the polymorphic wild-type strain S1D2 in order to map the mutation relative to the annotated PDS gene. A total of 21 progeny were isolated from this cross: 12 from complete tetrads and 9 from incomplete tetrads. A marker for the PDS locus on chromosome 12 amplified a 268 bp PCR product from both pds1-1 and S1D2. Digestion of the 268 bp PCR product with ScrFI yielded 215 and 52 bp fragments from pds1-1, whereas 111, 104, 26, and 25 bp fragments were produced from S1D2. DNA fragments smaller than 100 bp could not be visualized. When the PDS marker was tested on DNA isolated from the progeny, the light green phenotype cosegregated with the polymorphism found in pds1-1 (215 and 52 bp), while dark green progeny yielded fragments similar to S1D2 (111, 104, 26, and 25 bp) (). This result shows that the light green phenotype is linked to the PDS locus and that a mutation in PDS is likely to be responsible for the light green color and phytoene accumulation in pds1-1.
The PDS gene is genetically linked to the pds1-1 mutant phenotype.
The PDS locus was sequenced from pds1-1 to discover if a mutation in this locus was responsible for the phytoene-accumulating, light green phenotype. The Chlamydomonas nuclear genome sequence of PDS is 4030 bp and the predicted protein is 564 amino acids long. Amplification and sequencing of the PDS locus identified a single base pair change in exon two of pds1-1. The point mutation consisted of a G/C to A/T transition, resulting in an E143K missense change in deduced PDS protein sequence (). A multiple sequence alignment of predicted PDS protein sequences from wild-type C. reinhardtii, O. tauri, Synechocystis sp PCC6803, and A.thaliana and revealed that the amino acid change occurred in the conserved dinucleotide (FAD)-dependent oxidoreductase/amine oxidase domain of the PDS protein ().
Multiple sequence alignment of PDS protein sequences.
Isolation of pds1-2 as an intragenic enhancer of pds1-1
Because pds1-1 still synthesizes colored carotenoids, a second round of UV mutagenesis was conducted to find enhancer mutants that eliminated PDS activity. Light green pds1-1 cells were mutagenized, and a white mutant, P3-84, was isolated ( and ) that had a similar pigment profile as null psy (lts1-210) mutants, except that P3-84 accumulated a low level of phytoene (). When P3-84 was crossed to wild-type cells, the tetratype tetrad progeny gave unexpected pigment phenotypes: two white progeny with phytoene accumulation and no other carotenoids, one dark green mutant with wild-type carotenoid composition and levels, and one light green mutant with lower carotenoid levels (). The original light green, phytoene-accumulating pds1-1 pigment phenotype was not recovered. To determine if the P3-84 mutant phenotype was due to mutations in either the PSY or PDS gene, both genes were sequenced. Sequencing results revealed new mutations in both PSY and PDS genes in P3-84. An in-frame deletion of 24 bp removed eight amino acid residues from positions 7 to 14 (H7SAQTCPA14) in the putative chloroplast transit peptide of PSY (). This new allele of PSY was named lts1-301. The PDS locus was found to carry two point mutations: the original pds1-1 mutation (E143K) and an additional T to C transition resulting in the conversion of a leucine residue at position 64 to a proline residue (L64P) (, ). This double mutant allele of PDS was named pds1-2.
Analysis of enhancer strain P3-84 (lts1-301 pds1-2) and intragenic suppressors of pds1-2 mutants.
Summary of mutants described in this work.
The PDS and PSY sequencing results from P3-84 explained the unexpected tetratype phenotypes recovered in the cross between P3-84 and wild-type. The two parental phenotypes were represented: dark green wild-type and white P3-84 (). For the two unexpected phenotypes, the light-green, no phytoene accumulating phenotype belonged to progeny with reduced PSY activity (). Sequencing of PSY and PDS genes from this progeny (lts1-301) revealed that it has the eight amino acid chloroplast transit peptide deletion in PSY and no mutations in PDS (). The lts1-301 strain synthesizes wild-type carotenoids, but at reduced levels, indicating that PSY function is reduced or “leaky” () possibly because of inefficient transport of the PSY protein into the chloroplast. The second unexpected phenotype, white plus phytoene accumulation, belonged to progeny with wild-type PSY and the two mutations in PDS (L64P and E143K) (). This second white progeny, pds1-2, is an intragenic enhancer mutant for pds1-1 since the only carotenoid detected was phytoene (). Both P3-84 (lts1-301 pds1-2) and pds1-2 survive only in the dark. Similar to lts1-210 and pds1-1 mutants, they die when cultured under very low light. In contrast, lts1-301 is very light tolerant, growing almost as well as wild-type cells in HL ().
pds1-3 is a null allele derived from DNA insertional mutagenesis
An additional white, phytoene-accumulating mutant, pds1-3, was isolated from a DNA insertional mutagenesis screen based on its sensitivity to light and white color. Similar to pds1-1 mutants, pds1-3 bleached and died at vLL intensities (), and it also accumulated phytoene (). Unlike pds1-1, however, it does not synthesize any colored carotenoids ().
Tetrads from crosses with wild-type segregated 2
2 for the pds1-3
and wild-type phenotypes (), indicating that the pds1-3
mutant phenotype is controlled by a single gene. Co-segregation of the mutant phenotype with paromomycin resistance also indicated that the mutation is tagged by the transforming plasmid ().
To identify the mutation responsible for the white, phytoene-accumulating phenotype of the pds1-3
mutant, RESDA-PCR was used to recover a flanking sequence tag for one end of the vector insert in pds1-3
. The flanking sequence was used as a query in a BLAST search for homologous sequences 
against the Department of Energy (DOE) Joint Genome Institute (JGI) Chlamydomonas reinhardtii
v4 genome (www.jgi.doe.gov/chlamy
) and found to have significant identity to a 331 bp sequence on Chromosome 12. Flanking sequence analysis indicated that the insertion interrupts an intron in PDS
(). A primer was designed within the Chlamydomonas
genomic DNA flanking the putative insert location obtained from RESDA-PCR for pds1-3
, and PCR with this primer and three nested primers within the vector was performed (). Successful amplification confirmed the location of the plasmid vector in pds1-3
genomic DNA (). Recovery of flanking sequence at the other end of the insert was unsuccessful, however. One reason may be because the insertion of foreign DNA into Chlamydomonas
genomic DNA is often accompanied by a deletion 
. To determine whether a significant deletion was present in pds1-3
, PCR primers were designed within the PDS
genomic DNA on the side of the insertion for which no flanking sequence could be recovered. MS031A and MS031B primers amplified a 498 bp product from pds1-3
genomic DNA, 500 bp downstream from the insertion point, and primers MS041A and MS041B amplified a 200 bp fragment 2.5 kb distant from the site of insertion. Successful amplification and DNA sequencing of these PCR products indicated that a large deletion did not accompany the plasmid insertion ().
Analysis of pds1-3 DNA insertional mutant.
The PDS transcript is present in pds1-1 (), although at a reduced level: ~13% of the level found in wild-type (). In contrast, no PDS transcript was detectable by RT-PCR or qPCR in pds1-3 ().
Growth defects of pds1 mutants
Multiple independent alleles of lts1
but no pds1
mutants were isolated in a previous screen for white mutants of C. reinhardtii
. To understand why pds1
mutants were not found, the growth rates of pds1
mutants were compared to lts1-210
and wild-type cells. Comparison of growth rates in liquid TAP medium in the dark showed that pds1-1
mutants grew more slowly than either lts1-210
or wild-type cells ().
Results of plating assays of wild-type, lts1-210, pds1-3, and pds1-1.
Differences in growth between pds1 mutants and white lts1-210 or wild-type cells were more pronounced in plating assays. First, the plating efficiency of pds1, lts1, and wild-type strains was measured as colony-forming units (CFU). After 3 weeks growth in the dark, the plating efficiency of pds1-1 was calculated as 53.5%±2.8; pds1-3 was 48.4%±2.9; lts1-210 was 86.7%±10.2; and wild type was 83.3%±3.5 ().
A second plating experiment was performed to assess how visible pds1
mutants are in a background of wild-type cells when grown in a ~1
1 ratio (). After accounting for the ~50% and ~80% observed plating efficiencies for pds1
strains and wild-type/lts1-210
strains, respectively, the expected CFU was 1250 CFU/plate for pds1
mutants and 1320 CFU/plate for wild-type and lts1-210
strains. On TAP-agar plates with a 1
1 ratio of lts1-210
to wild-type cells, white lts1-210
colonies were easily identified; they were as densely populated and equal in diameter to wild-type colonies (). In contrast, it was difficult to identify light green pds1-1
and white pds1-3
mutants among wild-type colonies because their colonies were frequently half the diameter or smaller than wild-type colonies and fewer in number (). Of the carotenoid mutants tested, pds1-3
colonies were the smallest and the least dense. Because of their small size, it was difficult to determine the color of some pds1
colonies and as a result, they could have been mistaken for extremely small wild-type colonies or not been detected at all in a screen for white mutants 
Intragenic pds1-2 suppressor mutants
To gain further insight into amino acid residues important for PDS structure and function, mutations that suppressed the white phenotype of pds1-2 were isolated. Sixteen light green pds1-2 suppressor mutants falling into three allelic classes were isolated from UV mutagenesis of the white P3-84 strain (lts1-301 pds1-2).
PSY and PDS genes were both sequenced from the pds1-2 suppressor mutants to identify any revertants and/or additional mutations. All 16 suppressor mutants retained the chloroplast transit peptide mutation in the PSY gene (lts1-301) from strain P3-84 (). Four of these strains, csp6, csp10, csp14, and csp15, had a reversion of the pds1-1 mutation: a transition from “A/T” (Lys143) back to wild-type “G/C” (Glu143) ( and ; ). These suppressors retained the L64P mutation, which was named pds1-4. The second class of intragenic pds1-2 suppressor mutants, pds1-5, had the original pds1-1 mutation (E143K) plus a new pds1 mutation, which converted the pds1-2 mutation (L64P) in exon one to L64F (, ). Seven pds1-5 strains were isolated: csp3, csp4, csp5, csp9, csp11, csp16 and csp17 (, csp17 not shown; ). The third class of intragenic pds1-2 suppressor mutants had three mutations in PDS: E143K from pds1-1, L64P from pds1-2 and a new mutation, K90M. The methionine at position 90 resulted from a transversion mutation that changed the wild-type “A” to a “T ( and ). In this third allelic class, pds1-6, five strains were isolated: csp1, csp7, csp8, csp12, and csp13 (, csp8 not shown; ).
Analysis of intragenic suppressors of pds1-2 mutants.
Light green intragenic suppressor mutants of pds1-2
were more light tolerant than light green pds1-1
and white pds1
mutants but less light tolerant than medium green lts1-301
single mutants (). All suppressor mutants grew well under vLL, but only pds1-4
mutants could survive in LL. The pds1-5
mutants bleached and died in LL (). No suppressor mutant survived in HL. Although the suppressor mutants were less light tolerant than lts1-301
single mutants, pigment analysis of pds1-2
suppressor mutants showed that all three classes synthesize the full spectrum of wild-type colored carotenoids, but at ~20% of the levels found in wild-type cells, similar to lts1-301
(). Comparison of total xanthophylls, known for their photoprotective properties 
, did not reveal any significant differences between suppressor mutants and lts1-301
. Zeaxanthin, a xanthophyll particularly important for photoprotection 
, was elevated 1.6 fold in lts1-301
compared to suppressor mutants and 1.4 fold higher than in wild-type cells. The levels of lutein were significantly lower than in wild-type cells but not significantly different among suppressor mutants or lts1-301
mutants. Two of the three pds1-2
suppressor mutant classes still accumulated phytoene. The pds1-5
strains accumulated phytoene at ~17% and ~6%, respectively, of the levels present in the starting strain P3-84 (). Like wild-type cells, pds1-4
mutants did not accumulate any phytoene ().
PDS structural prediction
The 3DLigandSite web server predicted that the C. reinhardtii
PDS protein has a structure most similar to a human monoamine oxidase, C2c70B. Both C2c70B and the C. reinhardtii
PDS proteins were classified as oxidoreductases and had 12% identity to each other. 3DLigandSite predicted the presence of a dinucleotide-binding motif [NAD(P) or FAD] in the center of the PDS protein (cyan blue, ) and potential ligand binding sites (indigo, ) 
. The C-terminus of bacterial carotenoid dehydrogenases was proposed to contain a hydrophobic carotenoid-binding pocket 
, which was also found conserved among cyanobacteria, algae, and plants 
. In C. reinhardtii
this region spans amino acid residues 492–517 (, lavender).
In the predicted PDS protein structure, amino acid residue 143 (mutated in both pds1-1 and in pds1-2 suppressor mutants) and amino acid residue 90 (mutated in pds1-6), were adjacent to one another, and in spacefilling mode, in physical contact (, Glu143, red and Lys90, orange). In the wild-type PDS protein these amino acids are Glu143 and Lys90. Amino acid residue 64, which is affected in pds1-2 and the suppressor mutants could not be visualized, because no structure was predicted for the first 71 amino acid residues of the N-terminus (or the last 22 residues of the C-terminus) of C. reinhardtii PDS.