The product of the psbA locus has been under intense investigation in several laboratories. This protein, usually referred to as the "32 kd" poly-peptide, is associated with photosystem Il. It turns over rapidly in the light (i.e., in cells active in photosynthesis) but not in the dark [Reisfeld et al., 1982]. Interestingly, this polypeptide appears to be the target site for several herbicides that block electron transport at the reducing side of photosystem II, presumably by interfering with quinone binding [Arntzen et al., 1982].
Labeled azido atrazine has been shown to bind preferentially to the 32 kd
polypeptide [Pfister et al., 1981]. Its amino acid sequence has been highly conserved in higher plants and algae. In contrast to higher plants, in which the psbA gene has been mapped in the single copy region, it is located within the inverted repeat and therefore present in two copies per genome in C.
reinhardii. As is shown in Figure 4, the gene contains four introns of 1.35, 1.4, 1.1, and 1.8 kb, and it spans a region of7 kb [Erickson et al., 1984a].
Uniparental mutants resistant to the herbicides diuron, atrazine, and bro-macil have been isolated in C. reinhardii [Galloway and Mets, 1982, 1984;
Tellenbach et al., 1983; Erickson et al., l 984b]. Atrazine resistance has also appeared in several weed species [ Arntzen et al., 1983]. The first mutant of C. reinhardii examined, DCMU4, was isolated by P. Bennoun as a diuron-resistant mutant. Further studies revealed that this mutant is also highly resistant to atrazine. A comparative sequence analysis of wild-type and mutant psbA revealed a single base pair change in the fifth exon at the ser 264 residue, which is replaced by ala. Both copies of psbA were found to be mutated [Erickson et al., 1984b]. Hirschberg and Mclntosh [1983] have previously reported that a similar change occurs in an atrazine-resistant biotype of Amaranthus hybridus, in which the same ser is replaced by gly.
0 10 11 kb
HHH X X fi H H
w t ( 2137) _:~L.._1.~_'.Ki..._, ...:.711er~vz0z::?ZZ2?Z:l+~/,22</22//22?/22V222:JIW1113 :z?z:z22:za:z:2:.~.::z!~z2?222?222??ZZ?VZ?0:Zv2:1•1115 L.K:...:, KÙ.1 ,:..:_K _ ___I~ _ _ __:,7:._
Fig. 4. Analysis of deletions in the psbA region. The psbA gene is indicated in the upper part of the figure with its five exons (•)and four introns (r::2). Transcription procecds from the left to the right. Restriction endonuclease sites are marked by R, EcoRI; K, Kpnl; H, HindlII; X, XbaI; B, BamHI. Mutants 11-lA, 11-40, and 8-36C were isolated by Sprcitzer and Mel~ (1981]; mutants FuDll-2, FuD7 and FuDl3 were obtained from P. Bennoun.
Deletions are indicatcd by open bars. The deletion end points were mapped within the regions marked by double arrows.
36 Rochaix et al.
Sequence analysis of psbA from two other uniparental herbicide-resistant mutants of C. reinhardii isolated by Galloway and Mets [1984) has revealed that other parts of the 32 kd polypeptide play a role in herbicide binding.
Mutant Dr2, which is 17-fold more resistant to diuron and twofold more resistant to atrazine, has the wild-type val 219 changed to ile; mutant Ar7, which is 15-fold more resistant to atrazine and more sensitive to diuron than wild type, has phe 255 changed to tyr (Erickson, Rahire, and Rochaix, unpublished results). These studies indicate that there is a considerable flexibility in the binding sites for these herbicides. It is noteworthy that electron flow ·is not affected by the mutations in Dr2 and Ar7, in contrast to DCMU4 and to the atrazine-resistant Amaranthus. Mutations of the former type might therefore be of considerable agronomie value.
The tightly packed structural organization of the polypeptides in photosys-tem II might introduce some complications in the interpretation of these results. The fact that amino acid changes occur in these herbicide-resistant mutants does not necessarily imply that the binding site is located entirely on the psbA product. Allosteric effects on neighboring polypeptides in the photosystem II unit involved in true herbicide binding cannot be ruled out [Gresse!, 1984]. lt is interesting to note that another PSII-associated polypep-tide, 02, displays partial sequence homology with the 32 kd polypeppolypep-tide, genome, one might expect that this gene is particularly sensitive to mutage-nesis. Among six mutants examined that are deficient in photosystem II [Spreitzer and Mets, 1981; P. Bennoun, unpublished results] all have deleted the entire psbA region (Fig. 4; Herz, Erickson, and Rochaix, unpublished results). Recause the psbA gene is within the inverted repeat, ail of these mutants contain double deletions. lt appears likely that one deletion was created first and then transmitted to the other copy of the inverted repeat, as seems to be the case for the single-site base substitutions in the herbicide-resistant mutants. The preliminary mapping data are consistent with the deletions being entirely within the inverted repeat. Double deletions at the other end of the inverted repeat have been described previously by Myers et al. [1982]. The mutations affecting psbA are valuable as genetic chloroplast markers. Linkage between the 8-36C mutation (Fig. 4) and other makers of the uniparental linkage group has been demonstrated [Mets and Geist, 1983), and several PSII mutations have been placed in a single chloroplast genetic locus [Spreitzer and Ogren, 1983].
IV. TRANSFORMATION IN CHLAMYDOMONAS REINHARDII The preceding sections have shown how useful defined photosynthetic mutations are in understanding the structure-function relationship of poly-peptides and how some of these mutants can be used to examine chloroplast-nucleocytoplasmic cooperation. lt is obvious that an efficient transformation system in C. reinhardii would be very helpful for developing this analysis further. In that numerous nuclear mutants with deficiencies in their photosyn-thetic apparatus are available [Harris, 1982), the genes affected could be isolated by complementation through transformation. The possible extension of the transformation to the chloroplast compartment would provide a very powerful tool for understanding the function and regulation of chloroplast genes.
Our first attempts with transformation used the ARG7 locus as selective marker, because this is one of the few nuclear loci of C. reinhardii that has been studied both at the genetic and biochemical levels. The ARG7 locus codes for arginino succinate lyase (ASL), the last enzyme in the arginine biosynthetic pathway. Several mutants have been isolated at this locus [Gill-ham, 1965; Loppes et al., 1972), and the enzyme has been partially purified [Matagne and Schlosser, 1977). A double mutant cw15, arg7B was used for most of the experiments. Because the cw15 mutant is cell-wall deficient [Davies and Plaskitt, 1971) and behaves like a natural protoplast, it is not necessary to use cell wall-degrading enzymes. The yeast ARG4 locus (which corresponds to the ARG7 locus of C. reinhardii), cloned in plasmic pYe arg4 [Clarke and Carbon, 1978), was used as transforming DNA. After incubation of cells with this DNA in the presence of poly-L-ornithine, colonies able to grow in the absence of arginine were recovered and some of these colonies were shown by Southern hybridization to have the foreign DNA integrated into their nuclear genomes [Rochaix and van Dillewijn, 1982). A serious limitation was the low transformation frequency, in the range of 10-6-10-7 transformants per treated cell, a value only slightly higher than the natural reversion rate. Attempts to clone the C. reinhardii ARG7 locus in yeast by using a cosmid bank of C. reinhardii DNA for transformation of yeast ARG4 strain were not successful. In that this locus appears to be very large based on genetic studies of interallelic complementation [Loppes et al., 1972], it is possible that the ASL gene contains introns that are not processed correctly in yeast.
A limiting step for transformation in this system might occur at the level of stabilization of the foreign DNA once it has entered the cells. One possibility is to use autonomously replicating plasmids as transformation vectors. Because no free plasmids exist in this alga except for chloroplast
38 Rochaix et al.
and mitochondrial DNA, they were constructed in vitro according to the strategy outlined in Figure 5. The 2.7 kb yeast HindIII fragment containing the ARG4 locus [Clarke and Carbon, 1978] was inserted into the EcoRI site ofpBR322 by blunt-end ligation, thereby producing the plasmid pJD2. Mbol and HindIII fragments from total DNA and purified chloroplast DNA were inserted into the BamHI and HindIII sites of pJD2. Pools of these recombi-nant plasmids were prepared and used to transform yeast or C. reinhardii by selecting for arginine prototrophy.
H1ndill fragment B of pYeargl
+
B pBR322 1 Eco RI)2100
ONA polymerase • dXTPs Blunt end ligation
pJD2 (ampR. TcR, argl)
PLASMID POOLS
Chlomydomonas remhordii arg selection
{ pJ02 • nuONA (HindII!,Mbol) pJD2 +et DNA ( HindIIT. Mbol I
E. co/1 amp" select1on
Yeast arg sel ectian
Fig. 5. Strategy for constructing plasmids that replicate autonomously in C. reinhardii and yeast. nu DNA, nuclear DNA; et, chloroplast DNA.
A. Relationship Between Chloroplast ARS Sequences and Authentic