|Home | About | Journals | Submit | Contact Us | Français|
The barley chloroplast mutator (cpm) is an allele of a nuclear gene that when homozygous induces several types of cytoplasmically inherited chlorophyll deficiencies. In this work, a plastome Targeting Induced Local Lesions in Genomes (TILLING) strategy based on mismatch digestion was used on families that carried the cpm genotype through many generations. Extensive scanning of 33 plastome genes and a few intergenic regions was conducted. Numerous polymorphisms were detected on both genic and intergenic regions. The detected polymorphisms can be accounted for by at least 61 independent mutational events. The vast majority of the polymorphisms originated in substitutions and small indels (insertions/deletions) in microsatellites. The rpl23 and the rps16 genes were the most polymorphic. Interestingly, the variation observed in the rpl23 gene consisted of several combinations of 5 different one nucleotide polymorphisms. Besides, 4 large indels that have direct repeats at both ends were also observed, which appear to be originated from recombinational events. The cpm mutation spectrum suggests that the CPM gene product is probably involved in plastome mismatch repair. The numerous subtle molecular changes that were localized in a wide range of plastome sites show the cpm as a valuable source of plastome variability for plant research and/or plant breeding. Moreover, the cpm mutant appears to be an interesting experimental material for investigating the mechanisms responsible for maintaining the stability of plant organelle DNA.
Spontaneous genetic variability in the plastid genome or plastome is scarce and the plastome is usually considered as a highly conserved genome (Palmer 1985, 1990; Wolfe et al. 1987; Clegg 1993; Wicke et al. 2011). Results from the artificial induction of plastome mutations are so far very limited, contrasting with the amount of information coming from experiments on nuclear gene-induced mutations (van Harten 1998; Prina et al. 2012a). The use of genetically unstable genotypes as a tool for expanding plastome variability has been pointed out as a promising alternative to artificial mutagenesis, however it has received little attention (Kirk and Tilney-Bassett 1978; Börner and Sears 1986; Gressel and Levy 2010; Prina et al. 2012a). Furthermore, in comparison with our knowledge about DNA repair mechanisms in bacteria and eukaryotic nuclei, little is known about these mechanisms in plant organelles, especially chloroplasts (Rowan et al. 2010; Ruhlman and Jansen 2014). In those regards, the barley chloroplast mutator (cpm), previously described as inducing several types of maternally inherited chlorophyll deficiencies (Prina 1992; Prina et al. 2009), appears to be an interesting experimental material that deserves further investigations. Only a few of the mutants that originated in cpm plants, which were previously isolated by their phenotypes, have been experimentally identified as plastome mutants (Rios et al. 2003; Landau et al. 2007, 2009, 2011). To better understand the mechanisms of the cpm and to evaluate it as a source of plastome variability, in this study we aimed to obtain more molecular information on the types of induced mutations and their distribution throughout the plastome. Nowadays, the successful use of reverse genetic strategies based on mismatch digestion called Targeting Induced Local Lesions in Genomes (TILLING), (McCallum et al. 2000) for high-throughput searching for induced nuclear DNA polymorphisms, led us to pursue a similar strategy for the screening of polymorphisms in the chloroplast genome of cpm plants (cpTILLING).
To fulfill the objectives mentioned above, we took advantage of experimental material previously available that was maintained by natural self-pollination of plants carrying the cpm genotype through numerous generations. One group of plants consisted of 2 families that have kept the mutator genotype from 12 to 17 generations, while the other one consisted of 4 families that carried the mutator genotype for 5 generations. We conducted an extensive sequence scan of 33 chloroplast genes and a few intergenic regions. In this way, we detected a substantial number of polymorphisms that originated from at least 61 independent mutational events and were widely distributed in different segments of the plastome and affecting sequences involved in a variety of functions. All results show the cpm mutant as a valuable source of plastome variability for plant research and/or plant breeding and suggest that the nuclear gene responsible for the cpm syndrome encodes a protein involved in the mismatch repair system of plastid DNA.
The plant material consisted of chloroplast mutator (cpm) (Prina 1992, 1996) and wild-type (WT) barley seedlings. The latter is the parental genotype, which was mutagenized and later used to isolate the cpm mutant (Prina 1992). The cpm seedlings analyzed in this work belong to 2 groups of progenies. Group A consisted of 2 families that carried the homozygous mutator genotype (cpm/cpm) for many generations (from 12 to 17). One of these families (A-M2) originated from the very same cpm/cpm M2 plant, which through self-pollination, produced the progenies used for isolating the cpm mutant at the M4–M5 generations (Prina 1992). The other one (A-F2) was genotyped as a cpm/cpm F2 plant, which came from self-pollination of an F1 derived from a cross of a WT plant as female with a cpm/cpm plant as the male parent (see Prina 1992, 1996; Landau et al. 2011). Group B consisted of 4 families that carried the mutator genotype for 5 generations, each one from a different F2 plant (B-F2-1, B-F2-2, B-F2-3, and B-F2-4), obtained as described above for A-F2 progenies (see Supplementary Material 1).
Both groups of families were already available before the start of this work, and we chose this experimental material for the present investigation with the idea that in these families there would be a greater chance to find cpm-induced plastome mutants than in families that carried the cpm activity during fewer generations.
Regarding the seedlings from which the control DNA was isolated, they belong to families that have gone through many more generations of self-pollination than those that produced the cpm seedlings. Indeed, they come from a genotype that had already undergone many generations of natural self-pollination when it was used to mutagenize and to isolate the cpm mutant (Prina 1992). Thereafter, this mother genotype was multiplied by natural self-pollination in parallel with the cpm analyzed families. It is worth noting that in this material there was no phenotypic evidence (presence of chlorophyll deficiencies and/or conspicuous morphological changes) suggestive of mutational changes. Those analyses were carried out in the glasshouse at second leaf seedling stage and in plants grown at the field nursery until maturity.
Genomic DNA was isolated from 1 or 2 leaves of individual seedlings using the micromethod described in Dellaporta (1994) with modifications. The tissue was ground with Dellaporta isolation buffer in the The Fast Prep®-24 Instrument (MP Biomedicals, USA) and extracted with chloroform before DNA precipitation.
DNA concentrations were measured using a spectrophotometer (Nanodrop, Thermo Scientific, Wilmington, DE, USA) and standardized to a concentration of 80ng/µL.
Primers were designed based on the chloroplast genome sequence of barley (Hordeum vulgare) [GenBank: NC_008590.1] using Primer3 software (v. 0.4.0, http://frodo.wi.mit.edu/). The primers were designed to have an amplicon size of 1–1.5kb (see Supplementary Material 2) and checked for single-band amplification on agarose gels before the TILLING screen. We took into account that most of the chloroplast-encoded genes are smaller than the required size for an adequate amplicon for TILLING (1–1.5kb). For this reason, many amplicons contained 2 or more genes and intergenic regions. On the other hand, due to their larger size, some genes as 16S rRNA, 23S rRNA, ycf3, psaA, and psaB, required the design of more than one pair of primers.
The 31 amplicons were distributed throughout different regions of the plastome, i.e. large single copy (LSC), small single copy (SSC), and inverted repeat (IR) regions. They covered 33 genes that according to their function can be grouped as follows: group 1: genes of the photosynthesis apparatus: psbA, psbC, psbD (PSII); psaA, psaB (PSI); rbcL (Rubisco large subunit), and ycf3 (chaperone involved in PSI assembly); group 2: genes related to the chloroplast translation machinery: rps2, rps3, rps4, rps7, 3’ rps12, rps14, rps16, rps18, rps19, rpl2, rpl16, rpl20, rpl22, rpl23, rpl33 (ribosomal proteins), infA (translation initiation factor 1), 16S rRNA, 23S rRNA (ribosomal RNAs) and tRNA His, tRNA Arg, tRNA Met, tRNA Leu (transfer RNAs), and group 3: genes whose functions are not very well understood and that do not belong clearly to any of these groups: matK (maturase K), ccsA (cytochrome c biogenesis protein), clpP (ATP-dependent Clp protease proteolytic subunit), and cemA (envelope membrane protein) (for details see Supplementary Material 2).
The DNA samples from individual cpm seedlings were mixed with DNA from WT control seedlings in a 1:1 ratio. A PCR reaction was performed in a final volume of 12 µL using 40ng of genomic DNA, 1× Taq buffer (50mM KCl, 10mM Tris–HCl, pH 9.0, 0.1% Triton X-100), 1.25mM MgCl2, 0.6 µM of each primer, 0.6mM dNTP mix, and 1.25 units kit T-plus 5U/µl Taq DNA polymerase (Inbio Highway, Tandil, Argentina). After denaturation at 94 °C for 3min, the reaction mixtures were heated to 94 °C for 30s, 60 °C for 1min, and 72 °C for 2min, in 35 cycles and 10min at 72 °C. Denaturation consisted of heating the samples at 99 °C for 10min; this was followed by a 70-step re-annealing series starting at 70 °C for 20s, decreasing 0.3 °C in each step, to favor the formation of a heteroduplex at the end of the PCR reaction. After re-annealing, the whole PCR reaction was digested with celery juice extract (CJE) obtained with the protocol described in Till et al. (2006). Digestions and electrophoresis were performed according to Uauy et al. (2009). Samples were run on 3% polyacrylamide gels (19:1 Acrylamide:bis ratio) in 0.5× TBE running buffer and stained with ethidium bromide after electrophoresis. We used a Protean II xi Cell vertical electrophoresis system (gel size: 20cm tall by 1.5mm thick; Bio-Rad Laboratories, CA, USA) and samples were run at 350V for 90min. Gel images were acquired with a G:BOX Chemi gel documentation system (Syngene, Cambridge, UK). The putative mutants were identified by the presence of cleaved products whose combined size was similar to that of the original PCR product. These samples were then sequenced together with the same fragment of the WT barley as a PCR product with their specific primers. The sequence alignments were done for the putative mutants, our WT barley and the sequence of reference [GenBank: NC_008590.1]. Vector NTI Software (Life Technologies, USA) was used for contig assembling, alignments, and subsequent sequence analyses. In order to discard errors due to the Taq polymerase and/or sequencing, once a pattern of digestion compatible with a putative polymorphism was found, the PCR and digestion were repeated to confirm the pattern of digestion before resequencing was conducted. Furthermore, the results of sequencing the putative mutants were always compared with both the corresponding sequence found in the WT control and the reference barley sequence.
The location of the analyzed amplicons within the barley plastome is displayed in Figure 1. Polymorphic amplicons found in the families belonging to group A or B are shown in different colors in Figure 1. The detailed molecular changes composing the polymorphisms and the number of times each polymorphism was observed in independently analyzed seedlings are listed in Tables 1 and and2,2, for groups A and B, respectively. Due to the peculiar pattern of variation observed in the rpl23 gene, these data are shown separately in Table 3.
In order to estimate the number of mutational events that would cause the observed polymorphisms, it should be noted that each of the 6 families came from a different M2 or F2 plant (see Material and Methods) ensuring that the polymorphisms observed in each family would come from mutations that occurred independently from those observed in other families. Thus, identical polymorphisms observed in different families were considered to have originated from independent mutational events, as the case of the A387G transition in the rps18 gene that was observed in both families of group A, while all identical polymorphisms found within a family were considered as having originated from a single mutational event. According to this reasoning, all the polymorphisms detected after the screening of 182 seedlings belonging to group A were interpreted as coming from at least 35 independent mutational events, while the polymorphisms detected through the analysis of 122 seedlings of group B would correspond to at least 26 mutations. Altogether, 61 independently induced mutations were detected after analyzing 304 seedlings. Fourteen different polymorphisms were found in intergenic regions and, without considering the rpl23 gene, 47 polymorphisms were located in 19 of the 33 analyzed genes (summarized in Table 4). The latter affected protein coding, intron, tRNA, and rRNA sequences (see Tables 1 and and22).
The largest number of different polymorphisms was found in the rps16 gene, which was polymorphic in 5 of the 6 analyzed families, with 9 different polymorphisms that were all located in the intron. Identical polymorphisms were repeatedly found in different seedlings within families, a fact that was especially observed in families of group A, which kept the mutator genotype (cpm/cpm) over many generations (see Table 1). Those observed in a larger number of seedlings were found in family A-F2: the insertion in the rps16 gene of 1 G at position 805bp, which was independently observed in 5 seedlings; the rbcL gene polymorphisms T 456 C; and the polymorphism in an intergenic region near the rps14 gene T 795 C observed in 4 seedlings each. In group A families, there were also 10 cases of polymorphisms that were independently detected 2 or 3 times in different seedlings of the same family, but only 4 cases of identical polymorphisms were found twice in group B families (see Table 1 and and22).
As mentioned above, the rpl23 gene showed a peculiar pattern of variation (Table 3). It consisted of different combinations of the very same 5 molecular changes: 2 transitions (G118A: missense; A203G: missense), 2 transversions (G115T: missense; T132A: silent), and 1 nucleotide deletion (G at position 133: frameshift). In some seedlings the 5 polymorphisms were observed together, whereas other different combinations of 4, 3, 2 or only 1 polymorphism were also observed.
In Table 5, the mutations detailed in Tables 1 and and22 are grouped according to the type of molecular changes. Fifty-seven of the 61 polymorphisms detected correspond to point mutations. They consisted of 43 transitions (31 were A/T to G/C vs. 12 that were G/C to A/T), 3 transversions and 11 small insertions/deletions (indels). The small indels consisted of 1 or 2 nucleotides localized in microsatellites (mononucleotide repeats of 9–10 bases).
Besides these, 4 large indels (1 insertion of 15bp and 3 deletions of 45, 79 and 620bp) were detected, all of them having direct repeat sequences flanking the inserted/deleted fragments ranging from 7 to 25bp. The largest deletion of 620 nucleotides was found in the coding sequence of the psbA gene. This deletion was detected in heteroplastidic condition and was observed in the gel as 2 bands of PCR products (Figure 2). The higher molecular weight band (1390bp) corresponded to the entire amplicon and the lower one (770bp) to the fragment without the deleted 620bp. The absence of this big portion of the sequence was confirmed by isolating and sequencing the 770bp band from the gel. Two additional smaller bands of approximately 430bp and 340bp could be observed in the gel corresponding to the digestion of the 770-bp fragment (Figure 2). These fragments probably originated from the digestion of a loop generated during the heteroduplex (770 bp: 1390bp) formation at the step of re-annealing, which was recognized and cleaved by the CJE.
The data presented in Tables 1 and and22 result from the analysis of 42466bp per seedling. As mentioned above, by analyzing 182 seedlings corresponding to group A, 35 mutations were detected that accumulated after an average of roughly 14.5 generations carrying the cpm syndrome, while in group B, after 5 generations, 26 mutations were detected by analyzing 122 seedlings. Considering only the substitutions, which were 24 for group A or 22 for group B, the substitution rate per bp and generation were 2.1×10–7 and 8.5×10–7, respectively.
The barley chloroplast mutator (cpm) is to our knowledge the only example in monocots in which a wide spectrum of cytoplasmically inherited mutants is induced (Prina 1992, 1996; Greiner 2012; Prina et al. 2012a). Previous investigations carried out on mutants that were isolated by direct genetics, allowed us to discover only a few plastome mutations caused by the cpm (Rios et al. 2003; Landau et al. 2007, 2009, 2011). The main goal of the present work was to detect a more significant number of polymorphisms in cpm plants in order to gain knowledge about the molecular types of mutations induced by the cpm and its target sites. We applied a TILLING (McCallum et al. 2000) approach based on enzymatic mismatch cleavage for the detection of plastome polymorphisms. This approach, which can be carried out on small samples of tissue, is appropriate for analyzing the experimental material used, that consist of seedlings supposed to be in a wide variety of heteroplastidic conditions. A TILLING strategy was previously used to detect plastome polymorphisms in natural populations (ORG-EcoTILLING, Zeng et al. 2012), but we found no reports of its use for searching “newly induced” mutants in the plastome. The success of the chosen strategy was a priori not obvious, because after a plastome mutation occurs, the growth of the mutant sector to an appropriate size to be detected is slower and less predictable than that of a nuclear one (Kirk and Tilney-Bassett 1978; Birky 2001; Prina et al. 2012a, 2012b).
As was pointed out in Materials and Methods, plastome polymorphisms were detected by mixing the DNA sample isolated from each cpm seedling with control DNA. In order to discard errors due to the Taq polymerase and/or to sequencing, once a pattern of digestion compatible with a putative polymorphism was detected, the PCR and digestion were repeated to confirm the pattern once again before sequencing. Furthermore, the results of sequencing the putative mutants were always compared with both the corresponding sequence found in the WT control and the reference barley sequence. Interestingly, all the polymorphisms were found in cpm DNA samples, but not in those of the WT control. Moreover, the fact that the designed amplicons in many cases comprised more than one gene and also some intergenic regions suggests that it is highly unlikely that we are amplifying nuclear or mitochondrial inserts of plastid DNA (nupts, mipts) with exactly the same size of our amplicons. Besides, we aligned all the amplicons against the Hordeum vulgare sub sp. vulgare Ensembl Database and although, we identified part of the amplicons sequences in the nuclear genome, the sequences of both primers of each amplicon never matched together in that small sequences (not more than 400 pb in the majority of the cases).
From the results in Figure 1 and Tables 1, ,2,2, and and4,4, it can be concluded that the polymorphisms were widely distributed along the plastomes of cpm plants and were located in both genic and intergenic regions. Without considering the peculiar situation observed for the rpl23 gene, 19 genes were found to be polymorphic, comprising sequences involved in a wide variety of functions, i.e. protein coding sequences, introns, tRNAs and rRNAs (see Tables 1 and and2).2). Table 4 summarizes the number of mutations detected in each gene. The rps16 gene was the most polymorphic, showing 9 different polymorphisms all of them localized in its intron, a result that is consistent with the use of this intron in phylogenetic studies given its potential for variation (Downie and Katz-Downie 1999; Popp and Oxelman 2004).
The situation observed in the rpl23 gene was very peculiar and therefore these results were not included in Tables 1 and and2.2. In the rpl23 gene, 5 different polymorphisms were observed in a variety of combinations (Table 3) that can be hardly explained by independently induced mutations, but rather is more likely they originated in recombination events between the rpl23 gene and the rpl23 pseudogene (NCBI Reference Sequence: NC_008590.1), which already in nature contains those 5 polymorphisms. Interestingly, studies of the grass family have shown that this pseudogene is being maintained by gene conversion with the functional gene (Morton and Clegg 1993).
Regarding the types of DNA molecular changes observed in cpm seedlings, the vast majority were substitutions, mostly transitions and small indels (Table 5), suggesting that the CPM gene is highly specialized in correcting these kind of mutations.
All the 11 small indels detected consisted of 1 or 2 nucleotides located in single sequence repeats (mononucleotide repeats of 9–10 bases), which are the predominant type of microsatellites in chloroplast genomes (Rajendrakumar et al. 2002). Besides the above-mentioned point mutations 4 big indels were detected, 3 of them exceeding the size of the largest indel previously detected by nuclear TILLING (Comai et al. 2004). Interestingly, direct repeats of variable length were detected at both ends of the 4 big indels, suggesting that recombination events were the cause of these large indels. In this regard, intramolecular recombination mediated by short direct repeats has been reported as a mechanism for producing plastome genetic diversity in wheat species (Ogihara et al. 1988, 1991).
Regarding the mechanism responsible for the cpm syndrome, the present results support the early idea about the malfunction of a nuclear encoded enzyme responsible for plastome DNA integrity (Prina 1992), which in accordance with Hasting et al. (1976), when recessive indicates that it corresponds to a gene affecting the DNA repair capabilities. Besides, the mutational spectrum of molecular changes described above strongly suggests that the CPM gene is involved in the DNA mismatch repair (MMR) system, which plays a fundamental role in mismatch repair during replication and has also been reported as having anti-recombination activity (Harfe and Jinks-Robertson 2000; Jun et al. 2006; Lin et al. 2007). In this sense, it is widely accepted that an increase in tandem repeat indels is a typical indicator of a failure in the DNA-MMR system (Thibodeau et al. 1993; Vaish and Mittal 2002). It is interesting to note that an increase in short indels associated with microsatellite DNA was observed by Stoike and Sears (1998) in Oenothera plants carrying a plastome mutator, which is one of the few mutator genotypes previously reported as inducing a wide spectrum of cytoplasmically inherited mutants in dicots (Epp 1973).
Another mutator genotype inducing a wide spectrum of mutants was originally reported in Arabidopsis as a chloroplast mutator (chm) mutant (Redei 1973) and later noted to induce rearrangements in the mitochondrial genome (Martínez-Zapater et al. 1992). This mutant was proven to be homolog of the MutS gene, which in E.coli is involved in MMR and recombination, and it was renamed as AtMSH1 (Abdelnoor et al. 2003). Its failure causes mitochondrial genomic shifting involving rapid and dramatic changes in the relative copy number of portions of the mitochondrial genome instead of fixing mismatches (Abdelnoor et al. 2003; Maréchal and Brisson 2010). Despite extensive research, no modifications were detected within the plastome of chm plants in the early investigations (Martínez-Zapater et al. 1992; Mourad and White 1992), while more recently a low frequency of DNA rearrangements mediated by recombination were found in the plastome of MSH1 disrupted plants (Xu et al. 2011). Other previous observations taken on MSH1 Arabidopsis mutants that differ from our observations in barley cpm plants are the faster sorting out of the mutant clones that was inferred based on leaf variegation and the usual presence of plants carrying distorted leaves and high levels of sterility (see Redei 1973; Sakamoto et al. 1996).
Considering the results summarized above, the MSH1 defective plants seem to differ from our observations of cpm plants, suggesting different roles of AtMSH1 and CPM. Our present results strongly suggest that the defective protein encoded by the CPM nuclear gene plays a role in the chloroplast MMR system. However, an extensive study of the influence of the CPM on the mitochondrion is still lacking in order to determine if the CPM protein has dual targeting to plastids and mitochondria.
In conclusion, through a TILLING strategy we demonstrated that cpm seedlings carried numerous plastome polymorphisms, mostly point mutations that were widely distributed over the plastome. The slight molecular impact of the majority of the cpm induced mutations could be helpful if we aim to obtain allelic series that could include vital alleles that retain functionality and not only gene knock-outs. In this regard, advances in plastome gene functionality based on plastome mutations are so far scarce (Greiner 2012), whereas most of the current knowledge about plastome gene functions came from reverse genetics studies based on gene knock-outs (reviewed by Rochaix 1997; Rochaix 2003; Leister 2003; Day 2012). In this sense, it is remarkable that from cpm seedlings it was possible to isolate mutants in essential genes like in matK or in ribosomal subunits genes, for which no gene knock-outs so far exist (Scharff and Bock 2014). Genetically unstable genotypes like the cpm can be a useful alternative to artificial mutagenesis for expanding plastome variability, as it has been previously pointed out (Kirk and Tilney-Basset 1978; Börner and Sears 1986; Gressel and Levy 2010; Greiner 2012; Prina et al. 2012a). It appears as a very valuable source of plastome variability for research, classical plant breeding and/or biotechnology, either for use in direct selection experiments (see Rios et al. 2003) or by applying an adequate reverse genetic strategy by TILLING or by massive sequencing. Further research on cpm can greatly contribute to improve the limited knowledge about DNA repair mechanisms that maintain the integrity of plant organellar DNA (Rowan et al. 2010; Ruhlman and Jansen 2014). Finally, it is worth mentioning that organelle mutators like cpm could play an important role in plant evolution, as it was pointed out by Gressel and Levy (2010) especially with regard to the strategies of plant adaptation to environmental stresses.
International Atomic Energy Agency Research Contract No 15671: Isolation and Characterization of Genes Involved in Chloroplast Genes Mutagenesis. Agencia Nacional de Promoción Científica y Tecnológica, Fondo para la Investigación Científica y Tecnológica 2007 Nº 620: The barley chloroplast mutator as a tool to originate plastome genetic variability. Instituto Nacional de Tecnología Agropecuaria, Proyecto Específico Area Estratégica Biotecnología 244631: Mutagenesis techniques for diversity generation on characters of agricultural and/or agro-industrial interest.
Data deposited at Dryad: http://dx.doi.org/doi:10.5061/dryad.j0bd4
We are very grateful to Profs Barbara Sears, Barbara Hohn, Anne Britt, and Thomas Börner for the invaluable suggestions on the original manuscript. We also want to thank Mr. Abel Mario Moglie for skillful handling of the plant material for such a long time since the cpm mutant was evident for the first time in 1984.