Characterization of the EYE2 Gene Required for Eyespot Assembly in Chlamydomonas reinhardtii

Corresponding author: Carol L. Dieckmann, Department of Biochemistry, University of Arizona, P.O. Box 210106, Tucson, AZ 85721-0106., dieckman{at}email.arizona.edu (E-mail)

Communicating editor: M. D. ROSE

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The unicellular biflagellate green alga Chlamydomonas reinhardtii can perceive light and respond by altering its swimming behavior. The eyespot is a specialized structure for sensing light, which is assembled de novo at every cell division from components located in two different cellular compartments. Photoreceptors and associated signal transduction components are localized in a discrete patch of the plasma membrane. This patch is tightly packed against an underlying sandwich of chloroplast membranes and carotenoid-filled lipid granules, which aids the cell in distinguishing light direction. In a prior screen for mutant strains with eyespot defects, the EYE2 locus was defined by the single eye2-1 allele. The mutant strain has no eyespot by light microscopy and has no organized carotenoid granule layers as judged by electron microscopy. Here we demonstrate that the eye2-1 mutant is capable of responding to light, although the strain is far less sensitive than wild type to low light intensities and orients imprecisely. Therefore, pigment granule layer assembly in the chloroplast is not required for photoreceptor localization in the plasma membrane. A plasmid-insertion mutagenesis screen yielded the eye2-2 allele, which allowed the isolation and characterization of the EYE2 gene. The EYE2 protein is a member of the thioredoxin superfamily. Site-directed mutagenesis of the active site cysteines demonstrated that EYE2 function in eyespot assembly is redox independent, similar to the auxiliary functions of other thioredoxin family members in protein folding and complex assembly.

CHLAMYDOMONAS reinhardtii is a unicellular green alga that responds to light. It swims toward low intensity light, which is termed positive phototaxis, and away from high intensity light, which is termed negative phototaxis (for reviews see WITMAN 1993 ; KREIMER 1994 ; HEGEMANN 1997 ; SINESHCHEKOV and GOVORUNOVA 1999 ; KREIMER 2001 ). Phototaxis requires two components: the eyespot, which is the light-sensing structure, and the paired flagella, which change their breast stroke beat pattern to alter swimming direction in response to light-induced signals originating in the eyespot region. Our interest focuses on the assembly and placement of the eyespot in the cell.

The eyespot is observed by light microscopy as a bright reddish-orange spot located approximately equatorially in the cell relative to the flagellar pole and offset from the plane of the flagella by 45° (HOLMES and DUTCHER 1989 ; see Fig 1). The orange color of the eyespot is due to an array of carotenoid-filled lipid granules tightly packed between the inner envelope and thylakoid membranes of the chloroplast (MELKONIAN and ROBENEK 1980 ). Rhodopsin-like photoreceptors and associated signal transduction components (calcium channels, G-proteins, and kinases) are located in the plasma membrane overlying the carotenoid granule patch area of the chloroplast (FOSTER et al. 1984 ; BECKMANN and HEGEMANN 1991 ; HARZ and HEGEMANN 1991 ; KROGER and HEGEMANN 1994 ; DEININGER et al. 1995 ; LINDEN and KREIMER 1995 ; CALENBERG et al. 1998 ). The carotenoid layers confer directionality by reflecting light back onto the photoreceptors in the plasma membrane (KREIMER and MELKONIAN 1990 ; KREIMER 1994 ; SCHALLER and UHL 1997 ) and by blocking light passing through the cell (CRESCITELLI et al. 1992 ; KREIMER et al. 1992 ). Both functions of the carotenoid layers are dependent upon the asymmetric quarter wave-plate sandwich arrangement of alternating thylakoid membranes and granules (see Fig 1; FOSTER and SMYTH 1980 ).

View larger version (20K): In this window In a new window Download PPT slide   Figure 1. Eyespot organization in Chlamydomonas. The eyespot is located equatorially between the flagellar pole and the opposite pole of the cell, on a meridian that is 45° from the flagellar plane as shown on the left (HOLMES and DUTCHER 1989 ). The diagram shown on the right is a representation of an electron micrograph of a cross section through the eyespot (LAMB et al. 1999 ). The plasma membrane and outer and inner chloroplast membranes are tightly apposed to a layered sandwich of carotenoid granules and thylakoid membranes in the stroma of the chloroplast. The photoreceptors and signal transduction components are in the plasma membrane. The layered granule assembly reflects light that strikes it orthogonally back onto the photoreceptors in the plasma membrane, whereas it absorbs light that strikes it from the backside, after traveling through the cell (as reviewed in KREIMER 2001 ). eye2 mutant cells have no eyespot visible by light microscopy and do not have layered eyespot structures in the chloroplast.

When light is absorbed by the retinal-based photoreceptors, calcium channels open in the plasma membrane overlying the pigment granule layers (HARZ et al. 1992 ). The ion flux at the eyespot is sensed by the flagella, which change their beat pattern. At lower light intensities, the cis-flagellum (eyespot side) beats more weakly while the trans-flagellum beats more strongly to turn the cell toward the light; at higher intensities, the flagellar responses are reversed and the cell turns away from the light (RUFFER and NULTSCH 1991 , RUFFER and NULTSCH 1997 ). The asymmetric location of the eyespot relative to the flagellar plane is a requirement for phototaxis (FOSTER and SMYTH 1980 ). During cell division, the eyespot disappears from the cleavage plane and two newly formed eyespots appear 180° from each other and 90° from the cleavage plane (HOLMES and DUTCHER 1989 ). How is the eyespot assembled de novo at every cell division from signal transduction components in the plasma membrane of the cell and the underlying sandwich of chloroplast envelope, thylakoid, and carotenoid granule layers? Does photoreceptor localization in the plasma membrane depend on pigment granule layer assembly in the chloroplast?

In an ongoing effort to identify genes that affect eyespot placement and assembly, four loci, MIN1 (mini-eyed), MLT1/PTX4 (multi-eyed), EYE2, and EYE3 (eyeless) were identified by mutations that render Chlamydomonas unable to swim toward light in simple assays (PAZOUR et al. 1995 ; LAMB et al. 1999 ). A fifth locus, EYE1, was identified almost 50 years ago (HARTSHORNE 1953 ). eye2 and eye3 mutants appear eyeless by light microscopy, and electron microscopy revealed that the carotenoid granule layers are not assembled in these strains, whereas eye1 mutants can assemble small, ordered arrays of granules (LAMB et al. 1999 ).

Isolation of the eye2 and eye3 mutants has allowed us to ask interesting questions about the assembly of the eyespot from components in two cellular compartments. Can photoreceptors assemble properly in the plasma membrane when the underlying chloroplast layers are gone? In the absence of the carotenoid granule layers the assembly of the plasma membrane components may be disrupted completely, rendering the eyeless mutants unable to respond to light with a productive phototactic response. Alternatively, the photoreceptors may be intact and properly localized, but loss of the underlying carotenoid layers with their reflective/absorptive properties diminishes the phototactic responsiveness of the cells. To distinguish between these two possibilities, we have used sophisticated cell-tracking equipment and a directional laser light source to measure phototaxis by the eyeless eye2-1 mutant strain. We found that this mutant can phototax, but has a 100-fold reduced sensitivity to light and cannot orient as precisely as wild type to the direction of light. These data are suggestive that the EYE2 protein is required for the assembly of the pigment granule layers in the chloroplast, but is not required for the localization and function of the plasma membrane components of the eyespot.

Isolation of the genes mutated in eyeless strains will facilitate the elucidation of the function of proteins that govern carotenoid granule layer organization. To that end, we recovered an insertion mutation in EYE2, which allowed the isolation of the gene. Sequence analysis indicates that the EYE2 protein is a member of the thioredoxin superfamily. Members of this family are involved in a variety of cellular redox reactions via the two active site cysteines. In addition, several family members facilitate the assembly or function of protein complexes; these modulatory activities do not require redox activity. We demonstrate here that a redox-independent activity of EYE2 is required for eyespot assembly.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Chlamydomonas strains and media:
Chlamydomonas strains (Table 1) were grown in modified Sager and Granick medium I with Hutner's trace elements (HARRIS 1989 ), either with 0.1% (w/v) sodium acetate (R) or without acetate (M). Arginine-requiring strains were grown with supplemental arginine at a final concentration of 0.2 mg/ml (HARRIS 1989 ) in either acetate-containing medium (RNA) or acetate-free medium (MNA). Cultures were grown at 21°, under continuous light and agitation. Strains grown for the motion analysis phototaxis assay were grown in liquid M medium on a 14-hr light:10-hr dark photoperiod and supplemented with the bubbling of a mixture of CO₂ and air at a final concentration of 5% CO₂. Solid media contained 1.5% washed agar.

Chlamydomonas genomic DNA isolation:
Chlamydomonas genomic DNA was isolated following the protocol of ROCHAIX 1980 , with the omission of the second CsCl gradient required for the purification of chloroplast DNA.

Chlamydomonas transformation:
Chlamydomonas strains were transformed according to the silicon carbide whisker method of DUNAHAY 1993 , with the following modifications. A loopful of freshly grown cells from a plate was used to inoculate 5-ml RNA precultures. Overnight 5-ml precultures were used to inoculate 250-ml cultures at 1-day intervals. After the third culture was started, all cultures were grown 3 days at 21° in continuous light, with constant agitation. When the density of the middle culture attained between 4 x 10⁶ and 7 x 10⁶ cells/ml, all cultures were concentrated and washed with arginine-free medium and 1 x 10⁸ cells from each culture were transformed with 10 µg of linearized DNA and plated on acetate medium lacking arginine to select for transformants.

Light microscopy:
Cells were scraped from fresh plates into liquid medium and, when necessary, fixed with one-twentieth volume of tincture of iodine (HARRIS 1989 ). The presence of eyespots was confirmed by observation at x1500 magnification under oil immersion.

Cosmid library screen:
The pARG7.8cos cosmid library (PURTON and ROCHAIX 1995 ) was screened by making pools of between 500 and 1000 cosmid clones per pool. Twenty pools, representing a total of three genome equivalents, were screened by Southern analysis using a restriction fragment probe from the sequence flanking the insertion of ARG7 in the eye2-2 strain. Pools that contained the same 2.5-kb HindIII fragment detected by the probe in wild-type genomic DNA were plated for single colonies. Approximately 2000 single colonies from each pool were screened by colony hybridization using oligonucleotide #63395, designed from a sequenced portion of the eye2-2 flanking sequence. Southern hybridization with the eye2-2 flanking sequence probe confirmed the presence of the 2.5-kb HindIII fragment in individual cosmids.

DNA sequencing:
Automated DNA sequencing was performed at the DNA sequencing facility, Laboratory of Molecular Systematics and Evolution, University of Arizona (Tucson, AZ).

Southern analysis:
Southern analyses were performed according to standard methods (AUSUBEL et al. 1994 ), using uniformly labeled DNA fragments as probes. For Southern analysis of Chlamydomonas DNA, approximately 10 µg of digested genomic DNA was loaded. For Southern analysis of PCR reactions, one-tenth of the volume of the PCR reaction was loaded.

PCR amplifications:
PCR amplifications were performed according to standard methods (AUSUBEL et al. 1994 ) in a total volume of 100 µl. Reactions contained 1 ng of template DNA for amplifications of plasmids or 100 ng of template DNA for amplifications of genomic DNA. For amplifications using cDNA libraries as templates, the reaction typically contained 100 ng of template and contained DNA from 10⁶ clones. The annealing temperature for amplification reactions was 2° lower than the calculated T_m of the primer with the lower T_m.

Oligonucleotide synthesis:
The oligonucleotides used in this study are shown in Table 2. Oligonucleotides were synthesized by National Biosciences Inc. (Plymouth, MN) or Genosys Biotechnologies (The Woodlands, TX).

Plasmid construction:
The plasmids used in this study are shown in Table 3. Plasmids were constructed by standard methods (AUSUBEL et al. 1994 ). All plasmids were verified by restriction analysis, by sequencing across cloning junctions, or by sequencing entire inserts. The EYE2 flanking sequence containing plasmid pBR329-10C#1 was obtained by the plasmid rescue method of TAM and LEFEBVRE 1993 . Plasmid pKS-10C S/P was constructed by ligating a 1.2-kb SalI/PstI fragment of the EYE2 flanking sequence in pBR329-10C#1 into SalI/PstI digested pBluescriptII-KS(+). EYE2 containing cosmids A3-6-1, 3-10, 7-8-1, and 8-3 were isolated from a Chlamydomonas cosmid library (PURTON and ROCHAIX 1995 ).

Subclones of cosmid 3-10 containing the EYE2 gene were generated by partial digestion of cosmid 3-10 with either Sau3AI or a combination of Sau3AI and TaqI. Cosmid fragments were size selected and ligated to either BamHI-digested (B-15 and B-42) or BamHI/ClaI-digested pARG7.8 (C/B-10). Plasmid pKS+C/B-10R5/H was prepared by isolating the 2.2-kb EcoRV/HindIII fragment from the EYE2 gene in C/B-10 and ligating that fragment to pBluescriptII-KS(+) digested with HindIII and HincII.

Plasmids containing amplified EYE2 cDNA products, pGEM+3400 and pGEM+670, were constructed by ligating PCR products from a synchronized cDNA library using primers A3 and CB3400S (pGEM+3400) or S6 and CB670A (pGEM+670) to the pGEM-Teasy vector.

EYE2 site-directed mutant plasmids C193S, C190A, CYS12, and P191G were generated by the method of KUNKEL 1985 using single-stranded DNA from the plasmid pKS+C/B-10R5/H as template, with oligonucleotides Newdcys2, C190AA, Delcys12, or PDIA as the mutagenic primers, respectively. Mutated plasmids were verified by restriction analysis and by sequencing the mutated region.

Plasmids pEY2C193S, pEY2P191G, pEY2C190A, and pEY2CYS12 were generated by replacing the wild-type SgfI/BglII fragment of the EYE2 gene in C/B-10 with the mutant SgfI/BglII fragment from the site-directed mutant plasmids mentioned above. These plasmids introduce the thioredoxin motif mutations into the context of the genomic EYE2 clone.

Phototaxis assays:
For rapid screening of the phototactic ability of strains, the assay described in LAMB et al. 1999 was performed. For a more detailed examination of phototactic ability, the phototaxis assays described in MOSS et al. 1995 and PAZOUR et al. 1995 were performed. Aliquots of early log phase cells were darkadapted for 3 hr and placed between 18- x 18-mm coverslips mounted 2 mm apart on a microscope stage in a dark room for analysis (MOSS et al. 1995 ). Cells were subjected to directional light stimuli from a laser (500 ± 10 nm) in which the laser beam was passed through neutral density filters to alter the intensity of the stimulus beam. The individual swimming paths of cells were recorded by the ExpertVision motion analysis system (Motion Analysis Corp., Santa Rosa, CA). The duration of the stimulus beam was 20 sec, and the 10-sec recording period began 10 sec after sample illumination. ExpertVision software was used to calculate the mean swimming direction for each cell path for the duration of the recording. Data were collected at different light intensities, with a minimum of 100 cell paths collected for a given light intensity. Cells were allowed 2 min of dark adaptation between stimuli (ZACKS et al. 1993 ).

Prediction of exon structure in the EYE2 gene:
GeneMark prediction of coding potential (BORODOVSKY and MCINNICH 1993 ; LUKASHIN and BORODOVSKY 1998 ) was done using all default settings and the default Chlamydomonas coding sequence training sets.

Construction of figures:
Autoradiograms were scanned and cropped as necessary using Adobe Photoshop. Cropped images were imported into Microsoft Powerpoint, version 7.0, for final figure generation.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

eye2-1 mutant can phototax, but has 100-fold reduced sensitivity to light and orients imprecisely:
One question central to understanding eyespot assembly and localization is whether the plasma membrane photoreceptors and signal transduction components can be organized correctly in the absence of the underlying carotenoid pigment granule layers. Observing phototactic behavior by an eye2 strain would provide evidence that the assembly of functional photoreceptors is not strictly dependent upon the presence of the layered granules in the chloroplast. Previously, we assayed the eye2-1 mutant for the ability to swim to a slit in a masked test tube exposed to light at 3000 erg/cm²/sec. After 20 min, the mutant strain is evenly distributed throughout the culture, whereas wild-type cells form a tight band at the illuminated slit (LAMB et al. 1999 ). In the present study, we sought to discover whether the eye2-1 mutant had any capacity for phototaxis by sensitizing the cells to light by prior dark adaptation, varying light intensity over five orders of magnitude, and precisely recording the direction of movement of many cells relative to light direction with the quantitative motion analysis system. In the motion analysis phototaxis assay, the paths of at least 100 individual cells are recorded (MOSS et al. 1995 ). The chamber is illuminated from one direction, 0°, with an optic fiber attached to a laser. Populations of cells showing positive phototaxis have a distribution of path directions centered at 0° and populations of cells displaying negative phototaxis have a distribution of path directions centered at ±180°.

At low light intensity, the wild-type strain clearly shows positive phototaxis; 50% of the population is swimming on paths with angles near 0° (see Fig 2). The eye2-1 mutant population is randomly oriented at this intensity (see Fig 2A). At 100-fold higher light intensity, the wild-type population is even more strongly oriented toward the light source. The eye2-1 mutant population shows orientation toward the light, but the cells are evenly distributed at angles between -80° and +80° (see Fig 2B). At an additional 100-fold increase in light intensity, both wild-type and eye2-1 mutant populations display almost exclusively negatively phototactic orientations, although, as for positive phototaxis, the eye2-1 mutant is not as precise in its orientation (see Fig 2C). These data indicate that the eye2-1 mutant can phototax, but is 100-fold less sensitive to light than wild type and is imprecise in its orientation. Finding that the mutant strain is able to respond to light suggests that assembly of the carotenoid layers in the chloroplast is not required for proper localization of the photoreceptors in the plasma membrane of the cell.

View larger version (40K): In this window In a new window Download PPT slide   Figure 2. Phototactic orientation of eye2-1 and wild type at various light intensities. Represented are polar histograms of the mean swimming direction for each cell in a population subjected to a directional light stimulus of varied intensity; the direction of the laser beam is defined as 0° and is shown with an arrow. The bin size in each plot is 20°, with the middle angle of each bin indicated by the angle at the terminus of each radial line. Thus, the bin on the radial line denoted 170° represents the percentage of cells in the population traveling with paths of angles between 160° and 180°. Populations of cells traveling with a mean angle centered at 0° are positively phototactic, while populations of cells with paths of angles centered around 180° are negatively phototactic. The solid bars indicate the number of paths for a wild-type strain traveling at the angles defined by the 20° bin, while the shaded bars indicate the same for the eye2-1 mutant strain. Each annulus of the histogram represents 10% of the population. (Top) Represents the lowest light intensity used in the experiment, and the difference between each subsequent panel (top and middle) and (middle and bottom) shows the behavior of the population at a light intensity 100-fold higher than that of the previous panel.

Insertional mutagenesis yields the eye2-2 mutant strain:
The complete inability of the eye2-1 strain to assemble the carotenoid granule layers in the chloroplast portion of the eyespot suggests that the EYE2 protein is important in the assembly and/or integrity of this structure. To facilitate the analysis of the EYE2 protein, an insertional mutagenesis screen was used to "tag" and isolate the wild-type EYE2 gene. The auxotrophic arg7-8 strain was mutagenized by transformation with a plasmid containing the wild-type ARG7 gene (DEBUCHY et al. 1989 ). Arg⁺ transformants were enriched for phototaxis mutants as described in LAMB et al. 1999 . Only 1 of 26 Arg⁺ phototaxis-deficient (Ptx^-) strains obtained, H9-8, had the eyeless phenotype. Linkage analysis demonstrated that a mutation causing the eyeless phenotype was linked to the UV-induced mutation eye2-1; 31 progeny from 10 incomplete tetrads in a cross of H9-8 to 10-18 (eye2-1) were all unable to phototax. The new eyeless allele in H9-8 was named eye2-2.

Ten complete tetrads from an outcross of H9-8 to wild-type 137c scored 2 Ptx⁺:2 Ptx^-, suggestive that only one mutation affecting phototaxis segregated in the cross. Seventy-three progeny from 21 incomplete tetrads of a cross of H9-8 to an arg7-8 strain were scored; 31 were Arg^-Ptx⁺ and 41 were Arg⁺Ptx^-. That arginine prototrophy segregated with the phototaxis defect demonstrated that only one active ARG7 gene segregated with the eye2-2 mutation. To make sure that the only plasmid DNA in H9-8 is that inserted in the EYE2 gene, representative progeny from the crosses of H9-8 to wild type and arg7-8 were analyzed by Southern blot. The integrated plasmid DNA segregated with the Ptx^- phenotype, as shown by the hybridization pattern of progeny from both crosses (Fig 3). With the assurance that there was only one copy of the plasmid in the H9-8 strain, Chlamydomonas genomic sequence flanking the ARG7 plasmid insertion site was recovered by the method of TAM and LEFEBVRE 1993 .

View larger version (22K): In this window In a new window Download PPT slide   Figure 3. Insertion DNA segregates with the phototaxis defect and arginine prototrophy. Approximately 10 µg of PstI-digested genomic DNA from wild-type 137c (lanes 1 and 21), H9-8 (eye2-2: ARG7 arg7-8; lanes 2 and 20), and several progeny from a cross of H9-8 to an arg7-8 strain (lanes 3–14) and an outcross of H9-8 to wild-type 137c (lanes 15–19) were resolved on an agarose gel, blotted onto Nytran, and hybridized with a probe that detects the insertional plasmid backbone. The 500-bp PstI fragment detected by the probe contains both plasmid sequence and Chlamydomonas sequence flanking the plasmid insertion site. The phototaxis (Ptx) phenotype was scored by the lighted-slit method (LAMB et al. 1999 ). Arginine auxotrophy was scored on plates lacking arginine. The fainter bands in lanes 3 and 15 are due to a reduced amount of digested DNA in these lanes. The strain in lane 11 (m) was found to be a mixed population of Ptx+Arg- and Ptx-Arg+ cells.

EYE2 gene isolated from an ARG7 cosmid library:
To identify genomic sequence containing the wild-type EYE2 gene, a unique flanking sequence probe was used to screen a pARG7.8cos cosmid library (PURTON and ROCHAIX 1995 ; see MATERIALS AND METHODS). Four cosmids were verified by Southern blot to contain a wild-type 2.5-kb HindIII band, which is shifted to 11.5 kb by insertion of the ARG7 plasmid in H9-8 (data not shown). Two of the four cosmids rescued the eye2-1 mutant strain to wild-type phototaxis (see Table 4A). The shortest rescuing subclone was the 4.0-kb C/B-10 (Table 4B).

EYE2 genomic sequence analysis aided by GeneMark predictive algorithm:
The complete sequence of C/B-10 yielded no clues as to the organization of the EYE2 gene, and numerous attempts to identify the EYE2 transcript by Northern analysis and conventional cDNA library screening were unsuccessful. Therefore, the GeneMark program was used to predict regions of the C/B-10 insert that had a high probability of being exons (BORODOVSKY and MCINNICH 1993 ; LUKASHIN and BORODOVSKY 1998 ). The algorithm predicted five exons in three different reading frames on one strand of C/B-10 (see Fig 4A) and did not find any regions of high coding potential on the other strand (data not shown). Predicted exon sequences were used to design primers to amplify EYE2 cDNA. Because the transcripts for another gene involved in phototaxis (PTX2) were observed to accumulate just after cell division (G. PAZOUR and G. WITMAN, unpublished observations), EYE2 cDNA was amplified from a library made from mRNA at the same time in the cell cycle. Southern analysis verified that the PCR products contained EYE2 sequence (Fig 4B).

View larger version (60K): In this window In a new window Download PPT slide   Figure 4. GeneMark algorithm aids in isolation of EYE2 cDNA. (A) The 4-kb sequence of C/B-10 was analyzed using the GeneMark program (BORODOVSKY and MCINNICH 1993 ; LUKASHIN and BORODOVSKY 1998 ). Default settings for window size and step size were used. The output of the GeneMark prediction is shown for the three reading frames of one strand. The solid bars indicate a low stringency prediction of regions with high probability of containing exons, while the spikes indicate a more stringent window of high coding potential. PCR primers were designed based on predicted coding regions. One primer, A1, was chosen as an anchor, and antisense primers within that candidate exon (S2) and in the other candidate exons (S3, S4, S5, S6) were used for PCR. This process was reiterated using a different anchor primer, A3 or A4, with the appropriate downstream primers, to confirm the observations made from the first PCR series using the A1 primer as the anchor. (B) Southern hybridization of PCR products from primer pairs indicated by the schematic in A. DNA (100 ng) representing 5 x 106 cDNA clones from a synchronized population of wild-type Chlamydomonas at cell division served as templates for the amplification. PCR products were resolved on a 0.8% agarose gel, blotted onto Nytran, and probed with a random primed 2.3-kb HindIII/EcoRV fragment from C/B-10. Bands with an underlying bar were verified as genuine EYE2 cDNA products by sequencing. (C) A schematic summary of the organization of the EYE2 gene in the 4-kb insert of the C/B-10 plasmid is displayed. Solid bars represent coding sequence. Striped bars denote the 5' and 3' untranslated regions (UTRs). A short upstream ORF is shown as an open bar interrupting the 5'-UTR.

Amplifications using exonic anchor primers were used in combination with various other upstream and downstream primers to amplify as much of the 5' and 3' portions of the cDNA as possible. In combination with the exonic primer A3, primers with 5' ends at positions 786, 746, 646, and 596 successfully amplified EYE2 cDNA products, while all primers more upstream, 565, 556, 516, consistently failed. Likewise, on the 3' end of the cDNA, primers with 5' ends at positions 3093, 3155, and 3324 all successfully amplified EYE2 cDNA products in combination with exonic primer S6, while all primers more downstream, 3367, 3386, 3395, 3417, consistently failed to amplify EYE2 cDNA products. The longest EYE2 cDNA sequence was assembled and used to determine exon/intron boundaries.

EYE2 has five exons and an upstream open reading frame:
The EYE2 gene has five exons spanning over 2700 base pairs of genomic DNA (see Fig 4C). The 5' junctions for the four EYE2 introns are either G/GTGAG (introns 1 and 3) or G/GTGGG (introns 2 and 4), which closely match the proposed 5' splice site consensus G/GTGAG for Chlamydomonas (PURTON and ROCHAIX 1995 ). The four EYE2 3' splice sites are all different: GTGCAG/T, TCTCAG/T, GCCCAG/G, and GCGCAG/T, matching only the ultimate CAG intron nucleotides of the degenerate 3' Chlamydomonas consensus (G/A)CAG/G (PURTON and ROCHAIX 1995 ).

The spliced EYE2 sequence creates an open reading frame (ORF) of 503 codons. Interestingly, examination of the EYE2 cDNA sequence reveals the existence of a short ORF of 39 codons upstream of the EYE2 initiation codon (see Fig 5). Upstream ORFs (uORFs) have been observed in some mRNAs from other organisms (HINNEBUSCH 1984 ; KOZAK 1991 ; GEBALLE 1996 ; VILELA et al. 1998 ). In cases where the function of these uORFs has been elucidated, a role has been shown for the uORFs either in determining the efficiency of translation initiation at the coding ATG (KOZAK 1991 ; PINTO et al. 1992 ) or in the control of the stability of the mRNA (RUIZ-ECHEVARRIA et al. 1996 ; LINZ et al. 1997 ; VILELA et al. 1998 ).

View larger version (35K): In this window In a new window Download PPT slide   Figure 5. Organization of the 5' region of the EYE2 gene. The nucleotide sequence of the 5' region of the EYE2 gene including the first three codons of the EYE2 reading frame is shown. The exact 5' end of the EYE2 cDNA is unknown, but must be near the most upstream primer that successfully amplified the cDNA; the position of this primer is marked with asterisks above the sequence. Possible promoter elements are underlined. The position of the presumed initiation codon for the EYE2 protein (nucleotide 872) is boldface and underlined. The nucleotide sequence of the upstream open reading frame (uORF) has been italicized and the stop and start codons for the uORF are in boldface italics. The deduced amino acid sequence of the uORF is provided on the lines just below the nucleotide sequence in one-letter code. The complete sequence of the C/B-10 insert of Chlamydomonas genomic DNA containing the EYE2 gene is available from Genbank/EMBL/DDBJ under accession no. AF233430.

Three possible TATA promoter elements can be found upstream of the 5' end of the EYE2 cDNA, at positions 272–278, 329–336, and 582–588 (Fig 5). The element at 582–588 is likely to be the in vivo promoter for the EYE2 gene because of its proximity to the 5' end of the cDNA amplified from the synchronized library.

The signal for mRNA polyadenylation in Chlamydomonas is almost universally conserved; all but one known mRNA contains the sequence TGTAA 20 to 30 nucleotides (nt) upstream of the site of poly(A) addition (SILFLOW et al. 1985 ; MERCHANT and BOGORAD 1987 ). An exception to this rule is the actin mRNA, which contains the related sequence TGTAG 13 nt upstream of the poly(A) addition site (SUGASE et al. 1996 ). Two candidates for poly(A) signal can be found downstream of the 3' end of the amplified EYE2 cDNA. The most proximal signal is located at nt 3352–3361, and this signal is similar to the one observed for the Chlamydomonas actin mRNA. The other potential poly(A) signal sequence is located at nt 3741–3745 and is a perfect match to the canonical Chlamydomonas poly(A) signal sequence. On the basis of its proximity to the observed 3' end of the amplified EYE2 cDNA, the actin mRNA-like signal at 3352–3361 is likely to be the genuine signal for the EYE2 mRNA.

Mutations verify EYE2 ORF assignment and gene organization:
Two lines of evidence indicate that the major ORF in the EYE2 cDNA encodes the EYE2 protein. First, an ARG7 plasmid bearing the EYE2 gene with a nonsense mutation created at codon 186 was used to transform an arg7-8 eye2-1 strain. All 40 Arg⁺ transformants failed to produce eyespots. Second, PCR amplification and sequencing of genomic fragments from the eye2-1 mutant strain reproducibly revealed a G to A mutation at nucleotide 1326, the first nucleotide of the first intron. In other introns, mutations of this conserved nucleotide completely abolish splicing (BREATHNACH and CHAMBON 1981 ). Data from both types of mutant strains support the conclusion that the 503 amino acid (aa) ORF encodes the EYE2 protein.

EYE2 is a member of the thioredoxin superfamily:
Analysis of the EYE2 amino acid sequence, shown in Fig 6A, revealed that the protein is rich in arginine (13.3% relative to 5.8% for a set of 200 nuclearly encoded proteins) and proline (8.8% relative to 5.3% for the reference set). The theoretical pI of the protein is 11.3, and the predicted mass is 56.6 kD. EYE2 does not contain any regions that resemble transmembrane spanning helices. Simple ungapped or gapped BLAST analyses (ALTSCHUL et al. 1997 ) of the EYE2 protein sequence revealed a region with a highly significant match (expected by chance at a frequency of 3 x 10^-10) to sequences encompassing the active sites of proteins in the thioredoxin superfamily (EKLUND et al. 1991 ). The thioredoxin superfamily includes three major subclasses: 5'-adenylylsulfate reductases (ASRs) are required for sulfur metabolism (SETYA et al. 1996 ; BICK et al. 1998 ), protein disulfide isomerases (PDIs) chaperone protein folding (FREEDMAN et al. 1994 ), and thioredoxins (TRXs) are involved in numerous cellular redox reactions (THELANDER and REICHARD 1979 ; see Fig 6B). Most proteins in the thioredoxin superfamily contain two active site cysteines that catalyze oxidation/reduction reactions. The superfamily includes proteins with a variety of functions, some of which do not require the redox activity of the active site cysteines. We sought to discover whether the active site cysteines in the EYE2 protein are required for eyespot assembly or are dispensable as they are for some superfamily functions.

View larger version (28K): In this window In a new window Download PPT slide   Figure 6. Comparison of the thioredoxin active site motif of the EYE2 protein with thioredoxin superfamily members. (A) The complete sequence of the EYE2 protein is shown in one-letter code. The region of the protein similar to other thioredoxin superfamily members is shown in boldface. The putative active site cysteines are capitalized. (B) The Pileup program from GCG (Genetics Computer Group, University of Wisconsin) was used to align 26 residues of the EYE2 protein with regions of other thioredoxin superfamily proteins identified by BLAST searches. Ungapped BLAST was used for adenylyl (phospho)-sulfate reductases and protein disulfide isomerases. One iteration of PSI-BLAST was required to identify the thioredoxins with similarity to EYE2. A dendrogram generated by Pileup shows the relatedness of the family members to EYE2 and each other. The proteins displayed are the 5' adenylyl (phospho)-sulfate reductase-like proteins: apsr-athaliana (AF016282, aa 377–402, Arabidopsis thaliana), apsr-bjuncea (AJ001208, aa 376–401, Brassica juncea), apsr-acepa (AF212155, aa 354–379, Allium cepa), apsr-croseus (U63784, aa 375–400, Catharanthus roseus), apsr-lminor (AJ249831, aa 371–396, Lemna minor), apsr-eintestinalis (AF069951, aa 334–359, Enteromorpha intestinalis); thioredoxins: trx-soleracea (X51462, aa 22–47, Spinacia oleracea), trx-psativum (X76269, aa 89–114, Pisum sativum), trx-ccaldarium (D63676, aa 20–45, Cayanidium caldarium), txr-caurantiacus (A55124, aa 23–48, Chloroflexus aurantiacus), trx-afulgidus (AE001015, aa 51–76, Archaeoglobus fulgidus); protein disulfide isomerases: pdi-aniger (X98748, aa 161–186, Aspergillus niger), pdi-ncrassa (Y07562, aa 162–187, Neurospora crassa), pdi-athaliana (AC002535, aa 44–69, A. thaliana), pdi-msativa (M80235, aa 50–75, Medicago sativa), pdi-spombe (Z98597, aa 162–187, Schizosaccharomyces pombe), pdi-lmajor (AL160493, aa 13–38, Leishmania major), pdi-ecoli (M77746, aa 41–66, Escherichia coli dsbA); and EYE2: AF233430, aa 182–207, C. reinhardtii.

Mutagenesis of the thioredoxin active site motif in EYE2 does not affect eyespot assembly:
To test whether the conserved cysteines in the thioredoxin motif of EYE2 are required for eyespot assembly, mutant constructs were generated containing changes of the more N-terminal cysteine to alanine (C190A), the more C-terminal cysteine to serine (C193S), and of both cysteines to serines (C190S/C193S). Previous work indicated that the redox functions of thioredoxins require both cysteine residues (RUSSEL and MODEL 1986 ; HUBER et al. 1986 ), whereas some protein disulfide isomerases retain activity when the more C-terminal cysteine is replaced by serine or alanine (TACHIBANA and STEVENS 1992 ; ZAPUN et al. 1994 ; WUNDERLICH et al. 1995 ; WALKER et al. 1996 ; WALKER and GILBERT 1997 ). Additionally, because the identity of the two residues between the catalytic cysteines is known to influence the redox potential of the active site, a mutant was constructed with the proline at residue 191 (like Escherichia coli periplasmic PDI DsbA) changed to glycine (like most other PDIs).

The mutations were introduced into the EYE2 gene in plasmid C/B-10. These plasmids were linearized and transformed into strain eye2-1 arg7-8. Arg⁺ transformants were tested for their ability to undergo phototaxis and observed for the presence of eyespots. As shown in Table 5, all EYE2 thioredoxin active site mutants tested are capable of phototaxis and possess eyespots. The frequency of restoration of eyespots and phototaxis in the Arg⁺ transformants is similar to that of the wild-type EYE2 gene tranformed in the same manner (Table 4). Thus, catalytic activity of the thioredoxin motif is not required for the function of EYE2 in eyespot assembly.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Eyespot pigment granule layers are necessary for optimal sensing of light intensity and direction, but not for photoreceptor localization:
The pigment granule layers of C. reinhardtii eyespots have two functions: they enhance light detection through their reflective properties and they allow sensing of light direction through their absorptive properties. The granules are arranged in multiple layers of two, three, or four. The first layer of granules is tightly packed between the chloroplast envelope and a layer of thylakoid while the subsequent granule layers are subtended by single layers of thylakoid (MELKONIAN and ROBENEK 1980 ). The alternating arrangement of granules and thylakoid membranes is crucial for the function of the eyespot as a quarter-wave interference reflector (FOSTER and SMYTH 1980 ), as the regularity of the spacing between the layers contributes to the reflective properties of the eyespot.

If the light source is on the same side of the cell as the eyespot, light will strike the eyespot and be reflected back onto the area containing the photoreceptors, enhancing light detection. Reflection by the eyespot has been demonstrated using confocal microscopy and microspectrophotometry (KREIMER and MELKONIAN 1990 ; KREIMER 1994 ; SCHALLER and UHL 1997 ). If light enters the cell from the side opposite the eyespot, it must pass through the entire cell body before activating the photoreceptors. As it passes through the chloroplast, some light is absorbed by the photosynthetic machinery. CRESCITELLI et al. 1992 and KREIMER et al. 1992 have demonstrated that the carotenoid pigment granules absorb light that travels through the cell body and strikes the back of the eyespot. Together, the eyespot and the cell body provide the cell with an eightfold enhancement in directional sensitivity (HARZ et al. 1992 ).

Eyespots are formed de novo after cytokinesis in C. reinhardtii. A central question is how the cell can coordinate the positioning of components in the plasma membrane and the chloroplast to produce a fully functional visual apparatus. Previous work is suggestive that the presence of eyespot pigment granules is not required for the appropriate positioning of the photoreceptor molecules (MOREL-LAURENS and FEINLEIB 1983 ; MELKONIAN and ROBENEK 1984 ; KREIMER et al. 1992 ; DEININGER et al. 1995 ). MOREL-LAURENS and FEINLEIB 1983 used the microscopic motion analysis system (refined by MOSS et al. 1995 ) to analyze the sensitivity and precision of phototaxis by the eyeless eye1-3 strain. They inferred that the photoreceptors were properly positioned relative to the flagella, because the mutant could phototax to some degree. However, eye1 strains do assemble eyespots in stationary phase, and in logarithmic cultures, small, organized carotenoid granule layers can be observed by electron microscopy (LAMB et al. 1999 ). In contrast to eye1-3, the eye2-1 strain is completely devoid of ordered pigment granule arrays. Therefore, we used the motion analysis system to analyze the phototactic behavior of the eye2-1 mutant and determine whether the strain positions photoreceptors in the absence of pigment granule layers.

The eye2-1 strain displayed a positive phototactic response at 100-fold higher light intensity than that required to induce positive phototactic responses in the wild-type strain. Moreover, the orientation of the eye2-1 mutant was less precise than that of wild type at all light intensities tested. The imprecise orientation of the eye2-1 mutant is responsible most likely for the inability of this mutant to swim to the illuminated slit of a masked tube (LAMB et al. 1999 ) even though the light intensity used for that assay (3000 erg/cm²/sec) was 10-fold higher than that which induces a net bias toward the light in the motion analysis phototaxis assay. Interestingly, the eye2-1 strain displayed negative phototactic behavior at the same light intensity threshold as wild type.

Three conclusions can be drawn from the phototaxis assay data. First, the phototactic responses displayed by eye2-1 indicate the strain has some degree of organization of the photoreceptors even in the absence of the eyespot pigment granules, because an asymmetrically localized photoreceptor is required for phototactic behavior (FOSTER and SMYTH 1980 ). Second, the lack of eyespot pigment granules decreases the sensitivity of the visual apparatus; that more intense light is required to elicit positive phototaxis in the eye2-1 mutant is suggestive that the reflection provided by the eyespot is critical for signal enhancement at lower light intensities. In principle, this decrease in sensitivity for positive phototaxis could be due to a decrease in the number of photoreceptor molecules present in this strain. However, the observation of a wild-type threshold for negative phototaxis in the eye2-1 mutant indicates the eye2-1 strain is unlikely to have a decreased number of photoreceptor molecules. Third, the absence of eyespot pigment granules reduces the ability of Chlamydomonas to orient precisely with respect to the direction of the light source. Photoorienting eye2-1 populations clearly display a much broader range of swimming angles than do wild type. This is likely due to illumination of the photoreceptors by light passing through the cell body, which is absorbed by the pigment granule layers in wild-type cells.

EYE2 function in eyespot assembly is redox independent:
Similarity searches can often provide hints as to the function of a newly sequenced gene. The part of the EYE2 protein that shows significant similarity to other proteins comprises a putative thioredoxin-like active site. We found that several different site-directed mutations in the thioredoxin motif that should abolish redox function did not interfere with eyespot assembly. There are several possible explanations for this observation. First, retention of the thioredoxin motif in EYE2 over evolutionary time is due to chance, and it has no function in this protein. We do not favor this hypothesis, since the one segment of the protein that appears to have survived evolutionary change is that with the characteristic active site signature. A second hypothesis is that while the catalytic function of the thioredoxin active site is dispensable for EYE2 function, the motif has been conserved because it is important for the folding and stability of the EYE2 protein. It is reasonable to assume that the conservative site-directed mutations tested (C190A, C193S, P191G, and C190S/C193S) do not affect EYE2 structure. A third hypothesis is that the EYE2 protein has two functions, a redox role that is nonessential for growth or eyespot function and a separate function essential for assembly of the eyespot.

Several thioredoxin superfamily members have functions that are known to be independent of the disulfides in the active site. T7 phage DNA replication and f1 phage filament formation are dependent on E. coli thioredoxin. However, mutagenesis of either of the two active site cysteines fails to abolish the role of thioredoxin as a processivity factor for the T7 DNA polymerase (HUBER et al. 1986 ) or as an assembly factor for f1 filaments (RUSSEL and MODEL 1986 ). Dib1, a highly conserved subunit of the U5 snRNP, has a thioredoxin fold as determined by NMR and X-ray crystallography (REUTER et al. 1999 ; ZHANG et al. 1999 ). Its essential role in the large splicing complex is preserved despite a mutated active site. In addition to these known redox-independent functions of thioredoxin, protein disulfide isomerases can act as chaperones for protein folding (HAYANO et al. 1995 ; QUAN et al. 1995 ) and dimer assembly (VUORI et al. 1992 ; LAMBERG et al. 1996 ) in the absence of an active redox site. Prolyl 4-hydroxylase and microsomal triacylglycerol transfer protein (MTP) are both ß-heterodimers. The ß-subunit of each dimer is a protein disulfide isomerase. Mutagenesis of the active site cysteines of the PDI subunit has no effect on prolyl hydroxylase activity (VUORI et al. 1992 ) or the role of MTP in assembly of lipoproteins in the liver (LAMBERG et al. 1996 ). Expression of the -subunit in the absence of the PDI subunit results in aggregation and inactivity (LAMBERG et al. 1996 ), suggestive that the PDI subunit is required for complex assembly. We envision that EYE2 is involved in a similar redox-independent chaperone or protein-complex modulatory function that is required for the assembly of the eyespot.

	FOOTNOTES

¹ Present address: Department of Cell Biology, University of Massachusetts Medical Center, Worcester, MA 01655.

	ACKNOWLEDGMENTS

We thank Dr. Saul Purton for the pARG7.8cos library, Drs. Greg Pazour and George Witman for the synchronized cDNA library and use of motion analysis equipment, and Drs. Michel Goldschmidt-Clermont, Jeanette Quinn, Sabeeha Merchant, and Patrick Ferris for other useful cDNA libraries not published in this study. We thank Charles Quinton for the isolation of the eye2-2 mutant. We thank Drs. Georg Kreimer, John Little, Telsa Mittelmeier, and Mr. Michael Rice for reading the manuscript at various stages. We also thank Drs. Greg Pazour, George Witman, Georg Kreimer, Kim Sparks, and Tim Ellis for helpful discussions and Drs. Kirsten Krause and Lorraine Marnell for assistance with figures. This work was supported by National Institutes of Health grants GM34893 and GM60933 and National Science Foundation grant MCB-9806135 to C.L.D.

Manuscript received February 22, 2001; Accepted for publication April 12, 2001.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ALTSCHUL, S. F., T. L. MADDEN, A. A. SCHÄFFER, J. ZHANG, and Z. ZHANG et al., 1997  Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25:3389-3402[Abstract/Free Full Text].

AUSUBEL, F., R. BRENT, R. KINGSTON, D. MOORE, J. SEIDMAN et al., 1994 Current Protocols in Molecular Biology. John Wiley & Sons, New York.

BECKMANN, M. and P. HEGEMANN, 1991  In vitro identification of rhodopsin in the green alga Chlamydomonas. Biochemistry 30:3692-3697[Medline].

BICK, J. A., F. ASLUND, Y. CHEN, and T. LEUSTEK, 1998  Glutaredoxin function for the carboxy-terminal domain of the plant-type 5'-adenylylsulfate reductase. Proc. Natl. Acad. Sci. USA 95:8404-8409[Abstract/Free Full Text].

BORODOVSKY, M. and J. MCINNICH, 1993  GeneMark: parallel gene recognition for both DNA strands. Comput. Chem. 17:123-133.

BREATHNACH, R. and P. CHAMBON, 1981  Organization and expression of eucaryotic split genes coding for proteins. Annu. Rev. Biochem. 50:349-383[Medline].

CALENBERG, M., U. BROHSONN, M. ZELDACHER, and G. KREIMER, 1998  Light and Ca²⁺-modulated heterotrimeric GTPases in the eyespot apparatus of a flagellate green alga. Plant Cell 10:91-103[Abstract/Free Full Text].

CRESCITELLI, F., T. W. JAMES, J. M. ERICKSON, E. R. LOEW, and W. N. MCFARLAND, 1992  The eyespot of Chlamydomonas reinhardtii: a comparative microspectrophotometric study. Vision Res. 32:1593-1600[Medline].

DEBUCHY, R., S. PURTON, and J.-D. ROCHAIX, 1989  The arginosuccinate lyase gene of Chlamydomonas reinhardtii: an important tool for nuclear transformation and for correlating the genetic and molecular maps of the ARG7 locus. EMBO J. 8:2803-2809[Medline].

DEININGER, W., P. KRÖGER, U. HEGEMANN, F. LOTTSPEICH, and P. HEGEMANN, 1995  Chlamyrhodopsin represents a new type of sensory photoreceptor. EMBO J. 14:5849-5858[Medline].

DUNAHAY, T. G., 1993  Transformation of Chlamydomonas reinhardtii with silicon carbide whiskers. Biotechniques 15:452-460.

EKLUND, H., F. K. GLEASON, and A. HOLMGREN, 1991  Structural and functional relations among thioredoxins of different species. Proteins 11:13-28[Medline].

FOSTER, K. W. and R. D. SMYTH, 1980  Light antennas in phototactic algae. Microbiol. Rev. 44:572-630[Free Full Text].

FOSTER, K. W., J. SARANAK, N. PATEL, G. ZARILLI, and M. OKABE et al., 1984  A rhodopsin is the functional photoreceptor for phototaxis in the unicellular eukaryote Chlamydomonas. Nature 311:756-759[Medline].

FREEDMAN, R. B., T. R. HIRST, and M. F. TUITE, 1994  Protein disulphide isomerase: building bridges in protein folding. Trends Biochem. Sci. 19:331-336[Medline].

GEBALLE, A. P., 1996 Translational control mediated by upstream AUG codons, pp. 173–197 in Translational Control, edited by J. HERSHEY, M. MATHEWS and N. SONENBERG. Cold Spring Harbor Laboratory Press, Plainview, NY.

HARRIS, E. H., 1989 The Chlamydomonas Sourcebook: A Comprehensive Guide to Biology and Laboratory Use. Academic Press, San Diego.

HARTSHORNE, J. N., 1953  The function of the eyespot in Chlamydomonas. New Phytol. 52:292-297.

HARZ, H. and P. HEGEMANN, 1991  Rhodopsin-regulated calcium currents in Chlamydomonas. Nature 351:489-491.

HARZ, H., C. NONNENGASSER, and P. HEGEMANN, 1992  The photoreceptor current of the green alga Chlamydomonas. Philos. Trans. R. Soc. Lond. B Biol. Sci. 338:39-52.

HAYANO, T., M. HIROSE, and M. KIKUCHI, 1995  Protein disulfide isomerase mutant lacking its isomerase activity accelerates protein folding in the cell. FEBS Lett. 377:505-511[Medline].

HEGEMANN, P., 1997  Vision in microalgae. Planta 203:265-274[Medline].

HINNEBUSCH, A. G., 1984  Evidence for translational regulation of the activator of general amino acid control in yeast. Proc. Natl. Acad. Sci. USA 81:6442-6446[Abstract/Free Full Text].

HOLMES, J. A. and S. K. DUTCHER, 1989  Cellular asymmetry in Chlamydomonas reinhardtii.. J. Cell Sci. 94:273-285[Abstract/Free Full Text].

HUBER, H. E., M. RUSSEL, P. MODEL, and C. C. RICHARDSON, 1986  Interaction of mutant thioredoxins of Escherichia coli with gene 5 protein of phage T7. J. Biol. Chem. 261:15006-15012[Abstract/Free Full Text].

KOZAK, M., 1991  An analysis of vertebrate mRNA sequences: intimations of translational control. J. Cell Biol. 115:887-903[Abstract/Free Full Text].

KREIMER, G., 1994  Cell biology of phototaxis in flagellate green algae. Int. Rev. Cytol. 148:229-310.

KREIMER, G., 2001 Light reception and signal modulation during photoorientation of flagellate green algae, pp. 193–227 in Comprehensive Series in Photosciences, edited by M. LEBERT and D. HÄDER. Elsevier/North Holland, Amsterdam.

KREIMER, G. and M. MELKONIAN, 1990  Reflection confocal laser scanning microscopy of eyespots in flagellated green algae. Eur. J. Cell Biol. 53:101-111[Medline].

KREIMER, G., C. OVERLÄNDER, O. A. SINESHCHEKOV, H. STOLZIS, and W. NULTSCH et al., 1992  Functional analysis of the eyespot in Chlamydomonas reinhardtii mutant ey627 mt^-.. Planta 188:513-521.

KRÖGER, P. and P. HEGEMANN, 1994  Photophobic responses and phototaxis in Chlamydomonas are triggered by a single rhodopsin photoreceptor. FEBS Lett. 341:5-9[Medline].

KUNKEL, T. A., 1985  Rapid and efficient site-specific mutagenesis without phenotypic election. Proc. Natl. Acad. Sci. USA 82:488-492[Abstract/Free Full Text].

LAMB, M. R., S. K. DUTCHER, C. K. WORLEY, and C. L. DIECKMANN, 1999  Eyespot-assembly mutants in Chlamydomonas reinhardtii.. Genetics 153:721-729[Abstract/Free Full Text].

LAMBERG, A., M. JAUHIAINEN, J. METSO, C. EHNHOLM, and C. SHOULDERS et al., 1996  The role of protein disulphide isomerase in the microsomal triacylglycerol transfer protein does not reside in its isomerase activity. Biochem. J. 315:533-536.

LINDEN, L. and G. KREIMER, 1995  Calcium modulates rapid protein phosphorylation/dephosphorylation in isolated eyespot apparatuses of the green alga Spermatozopsis similis.. Planta 197:343-351.

LINZ, B., N. KOLOTEVA, S. VASILESCU, and J. E. MCCARTHY, 1997  Disruption of ribosomal scanning on the 5'-untranslated region, and not restriction of translational initiation per se, modulates the stability of nonaberrant mRNAs in the yeast Saccharomyces cerevisiae.. J. Biol. Chem. 272:9131-9140[Abstract/Free Full Text].

LUKASHIN, A. V. and M. BORODOVSKY, 1998  GeneMark.hmm: new solutions for gene finding. Nucleic Acids Res. 26:1107-1115[Abstract/Free Full Text].

MELKONIAN, M. and H. ROBENEK, 1980  Eyespot membranes of Chlamydomonas reinhardtii: a freeze fracture study. J. Ultrastruct. Res. 72:90-102[Medline].

MELKONIAN, M., and H. ROBENEK, 1984 The eyespot apparatus of green algae: structure, development and function, pp. 193–268 in Progress in Phycological Research, Vol. 3, edited by F. E. ROUND and D. J. CHAPMAN. Biopress, Bristol, UK.

MERCHANT, S. and L. BOGORAD, 1987  The Cu(II)-repressible plastidic enzyme cytochrome c: cloning and sequence of a complementary DNA for the pre-apoprotein. J. Biol. Chem. 262:9062-9067[Abstract/Free Full Text].

MOREL-LAURENS, N. and M. E. FEINLEIB, 1983  Photomovement in an "eyeless" mutant of Chlamydomonas. Photochem. Photobiol. 37:189-194.

MOSS, A. G., G. J. PAZOUR, and G. B. WITMAN, 1995  Assay of Chlamydomonas phototaxis. Methods Cell Biol. 47:281-287[Medline].

PAZOUR, G. J., O. A. SINESHCHEKOV, and G. B. WITMAN, 1995  Mutational analysis of the phototransduction pathway of Chlamydomonas reinhardtii.. J. Cell Biol. 131:427-440[Abstract/Free Full Text].

PINTO, I., J. G. NA, F. SHERMAN, and M. HAMPSEY, 1992  Cis- and trans-acting suppressors of a translation initiation defect at the cyc1 locus of Saccharomyces cerevisiae.. Genetics 132:97-112[Abstract].

PURTON, S. and J.-D. ROCHAIX, 1995  Characterization of the ARG7 gene of Chlamydomonas reinhardtii and its application to nuclear tranformation. Eur. J. Phycol. 30:141-148.

QUAN, H., G. FAN, and C. C. WANG, 1995  Independence of the chaperone activity of protein disulfide isomerase from its thioredoxin-like active site. J. Biol. Chem. 270:17078-17080[Abstract/Free Full Text].

REUTER, K., S. NOTTROTT, P. FABRIZIO, R. LÜHRMANN, and R. FICNER, 1999  Identification, characterization and crystal structure analysis of the human spliceosomal U5 snRNP-specific 15 kD protein. J. Mol. Biol. 294:515-525[Medline].

ROCHAIX, J.-D., 1980  Restriction fragments from Chlamydomonas chloroplast DNA. Methods Enzymol. 65:785-795[Medline].

RÜFFER, U. and W. NULTSCH, 1991  Flagellar photoresponses of Chlamydomonas cells held on micropipettes: II. Changes in flagellar beat pattern. Cell Motil. Cytoskeleton 18:269-278.