Present address: Genome Center, UC Davis, 451 East Health Sciences Drive, Davis, CA 95616-8816, USA.
Molecular karyotyping and aneuploidy detection in Arabidopsis thaliana using quantitative fluorescent polymerase chain reaction
Article first published online: 22 SEP 2006
The Plant Journal
Volume 48, Issue 2, pages 307–319, October 2006
How to Cite
Henry, I. M., Dilkes, B. P. and Comai, L. (2006), Molecular karyotyping and aneuploidy detection in Arabidopsis thaliana using quantitative fluorescent polymerase chain reaction. The Plant Journal, 48: 307–319. doi: 10.1111/j.1365-313X.2006.02871.x
- Issue published online: 22 SEP 2006
- Article first published online: 22 SEP 2006
- Received 5 April 2006; revised 19 June 2006; accepted 5 July 2006.
- Arabidopsis thaliana;
- quantitative fluorescent-polymerase chain reaction;
- quantitative genotype;
Certain cellular processes are sensitive to changes in gene dosage. Aneuploidy is deleterious because of an imbalance of gene dosage on a chromosomal scale. Identification, classification and characterization of aneuploidy are therefore important for molecular, population and medical genetics and for a deeper understanding of the mechanisms underlying dosage sensitivity. Notwithstanding recent progress in genomic technologies, limited means are available for detecting and classifying changes in chromosome dose. The development of an inexpensive and scalable karyotyping method would allow rapid detection and characterization of both simple and complex aneuploid types. In addition to the problem of karyotyping, genomic and molecular genetic studies of aneuploids and polyploids are complicated by multiple heterozygous combinations possible at loci present in more than two copies. Quantitative scoring of allele genotypes would enable large-scale population genetic experiments in polyploids, and permit genetic analyses on bulked populations in diploid species. Here, we demonstrate that quantitative fluorescent-polymerase chain reaction (QF-PCR) can be used to simultaneously genotype and karyotype aneuploid and polyploid Arabidopsis thaliana. Comparison of QF-PCR with flow cytometric determination of nuclear DNA content indicated near perfect agreement between the methods, but complete karyotype resolution was only possible using QF-PCR. A complex karyotype, determined by QF-PCR, was validated by comparative genomic hybridization to microarrays. Finally, we screened the progeny of tetraploid individuals and found that more than 25% were aneuploid and that our artificially induced tetraploid strain produced fewer aneuploid individuals than a tetraploid strain isolated from nature.
Certain cellular processes are sensitive to changes in gene dosage. Often, the proper function of a protein complex requires strict stoichiometry between the different gene products involved and altering the dosage of any one of these gene products can disrupt this stoichiometry (Birchler et al., 2001, 2003, 2005; Papp et al., 2003; Veitia, 2003, 2004, 2005). Such alterations in gene dosage are not uncommon. Recent improvements in the resolution of genomic analyses have revealed extensive copy number variation (CNV) between humans (van Ommen, 2005; Sebat et al., 2004). The observation that CNV for DNA elements smaller than whole chromosomes may play a role in human variation has rekindled an interest in the molecular and phenotypic consequences of varying gene dosage and the mechanisms underlying dosage sensitivity (Aitman et al., 2006; Fortna et al., 2004; van Ommen, 2005; Sebat et al., 2004; Snijders et al., 2005).
Another common alteration of gene dosage involves whole chromosomes. Sutton's proposal that genes reside on chromosomes (Sutton, 1903) found critical confirmation in Bridges and Blakeslee's description of aneuploidy, the presence of unbalanced sets in which chromosome copies are in excess or deficiency (Blakeslee, 1922; Bridges, 1916). Aneuploidy introduces dosage imbalance on a chromosomal scale and is proposed to alter phenotypes by simultaneously disrupting the stoichiometry of all dosage-sensitive gene products encoded by a chromosome, or subset of chromosomes (Birchler et al., 2001). As a result, aneuploidy is often associated with distinct and usually deleterious phenotypes that are specific to the type of aneuploidy (Blakeslee, 1922; Khush, 1973; McClintock, 1929). For example, all 12 trisomics of Datura exhibit distinct phenotypes, depending on which type of chromosome is present in three copies (Blakeslee, 1922). Similarly, in humans, patients diagnosed with Down's syndrome (trisomy of chromosome 21) consistently exhibit specific developmental defects (Epstein, 1986). The deleterious nature of chromosomal imbalance is also highlighted by the fact that cancerous cells are aneuploid and that aneuploidy may be a cause of cancer (Matzke et al., 2003; Pihan and Doxsey, 2003). The study of the molecular mechanisms behind aneuploid syndromes is therefore critical for a deeper understanding of dosage-sensitive genetic processes.
Notwithstanding the progress in genomic technology, we still have limited means for the detection and description of variations in chromosome number. This is particularly important for the study of many plant species because of their intrinsic tolerance to altered chromosome number, which frequently results in viable individuals with chromosomal combinations other than diploid. Polyploidy, the presence of more than two full chromosome sets, is very common in plants and as many as 70% of plant species may have polyploid origins (Masterson, 1994; Otto and Whitton, 2000; Ramsey and Schemske, 2002; Soltis et al., 2003). Polyploidy does not alter the relative dosage of genes or chromosome types and often has mild phenotypic consequences (Singh, 2003) but meiotic fidelity is lower in polyploids, resulting in frequent aneuploid offspring in polyploid populations (Doyle, 1986; Randolph, 1935). In polyploid backgrounds, relative chromosome imbalance is reduced due to the overall higher chromosome copy number and aneuploidy can have minor phenotypic consequences (Ramsey and Schemske, 1998). The precision of any study of polyploidy, or of polyploid populations, would therefore be improved by the detection of aneuploids to control for phenotypic effects because of altered chromosomal dosage.
Several methods for detection of aneuploidy have been developed, primarily for pre-natal diagnostics of aneuploidy in humans (Dudarewicz et al., 2005). Some methods, such as chromosome banding and fluorescent in situ hybridization (FISH), result in complete karyotype information, the determination of both chromosome number and type. Other methods, such as quantitative fluorescent-polymerase chain reaction (QF-PCR) (Ogilvie et al., 2005), multiplex amplifiable probe hybridization, multiplex ligation-dependent probe hybridization (Schouten et al., 2002), melting curve analysis of single-nucleotide polymorphisms and real-time quantitative PCR (Dudarewicz et al., 2005), have been used to detect changes in gene dosage and to infer the dosage of specific chromosomes. Recently, comparative genomic DNA hybridization to microarrays has been used to determine gene copy number and whole chromosome aneuploidy in several organisms such as yeast (Bond et al., 2004; Hughes et al., 2000) and humans (Sebat et al., 2004). For example, this method was recently used to detect and characterize aneuploidy in cancerous cells (Lucito et al., 2003). Similarly, the human single nucleotide polymorphism (SNP) arrays allow the identification of changes in copy numbers over short sequences (Rauch et al., 2004) and could be used to detect variation in whole chromosome dosage.
Fluorescent in situ hybridization and flow cytometric analyses have been the methods of choice for detection and characterization of aneuploidy in plants (Burton and Husband, 2001; Galbraith, 2004; Henry et al., 2005; Roux et al., 2003). The use of flow cytometric measurement of nuclear DNA content cannot classify the type of aneuploidy and often cannot unequivocally distinguish individuals differing by a single chromosome because the difference in total nuclear DNA content is too small. As a result, the construction of a complete karyotype, i.e. the determination of the exact number of each chromosome type, is not possible on the basis of DNA content alone. Fluorescent in situ hybridization can determine complete karyotype (Wang et al., 2006), but is laborious and requires considerable expertise and sophisticated probes. In addition to improving the resolution of studies on polyploids, an inexpensive, scalable, high-throughput karyotyping method would allow larger-scale experimentation to elucidate the mechanism and consequences of dosage sensitivity in both population genetic and clinical settings.
In addition to the problem of karyotyping, genomic and molecular genetic studies of aneuploids and polyploids are complicated by increased gene and allele copy numbers. Specifically, genotyping is complicated by the possibility of more than two alleles at each locus and the existence of different heterozygous states. For example, even with only two alleles, a trisomic heterozygous for A and a can exhibit allelic ratios of 1:2 (Aaa) and 2:1 (AAa). Similarly, in autotetraploids, AAAa (3:1), AAaa (1:1), and Aaaa (1:3) are all possible heterozygous combinations. Aneuploid individuals within polyploid populations further increase the number of heterozygous classes. For example, loss of one chromosomal copy in autotetraploids results in trisomy, adding two heterozygous combinations (1:2 and 2:1). Similarly, gain of one chromosomal copy adds two more heterozygous combinations associated with pentasomy (2:3 and 3:2). Correct identification of allelic dosage is required when the two alleles, A and a, have different additive values because the Aaaa genotype in a tetraploid would be phenotypically different from the AAAa genotype.
Most genotyping technologies detect polymorphisms as dominant or co-dominant markers (D'Surney et al., 2001). Dominant markers only detect the presence or absence of one of the two alleles. Co-dominant markers detect both alleles, but most of the methods used to distinguish between the alleles do not provide a quantitative measure of relative allelic ratios. Yet the relative ratio of two or more alleles must be determined in order to infer the genotype of a polyploid. In addition to allowing the genotyping of polyploids, quantitative genotyping can be used to estimate allelic frequencies in populations or perform bulk segregant analyses in diploid species. In plants, only a few studies have focused on the quantitative aspect of genotyping polyploid or aneuploid genomes. Pyrosequencing was used to quantitatively genotype SNPs in tetraploid potato (Rickert et al., 2002) and to quantify allele-specific gene expression in strawberry (Schaart et al., 2005). Quantitative genotyping of SNPs using microarrays was also recently performed in tetraploid potato (Rickert et al., 2005) and has been used for bulked segregant analysis in Arabidopsis (Borevitz et al., 2003).
Methods that allow the easy measurement of allelic dosage in polyploids and the easy detection and classification of aneuploids would be advantageous in basic and applied studies, especially if they were to be amenable to high-throughput applications. For example, statistical methods for analysis of quantitative trait loci in polyploids will have greater power when complete marker genotype information is available (Cao et al., 2004; Luo et al., 2006). Development of a scalable and inexpensive approach to quantitatively genotype polyploid populations would allow large-scale population genetic experiments in polyploids as well as improve our ability to perform genetic analyses on bulked populations in diploid species. In addition, it would allow rapid detection and characterization of complex aneuploid types, a necessary step towards a deeper understanding of aneuploid syndromes and the mechanisms underlying dosage sensitivity. In the present report, we demonstrate that QF-PCR can be used to simultaneously genotype and karyotype both aneuploid and polyploid populations of A. thaliana. Comparison of our QF-PCR results with flow cytometric determination of nuclear DNA content indicated near-perfect agreement between the methods, but complete karyotype resolution was only possible using QF-PCR. A complex karyotype, determined by QF-PCR, was validated by comparative genomic hybridization to microarrays. Finally, we demonstrate the utility of this method by screening a tetraploid population of A. thaliana and finding that more than 25% of the offspring from a tetraploid cross were aneuploid.
Quantitative genotyping in polyploid plants using QF-PCR
We wanted to develop a reliable method for quantitative genotyping and apply it to the study of karyotypic variation. We reasoned that a quantitative method should be able to identify and describe different aneuploid karyotypes and that it should be consistent with the measurement of genomic content provided by flow cytometry. We tested QF-PCR as a tool for distinguishing between the different allelic ratios present in aneuploids and autotetraploids.
In order to produce a swarm of aneuploid plants, an A. thaliana triploid termed ‘CWW’ (the F1 or diploid Col-0 × tetraploid Wa-1; see Experimental procedures) was used as the starting material. Meiosis in a triploid produces gametes of varying chromosomal content, many of which are aneuploids (Henry et al., 2005; Johnsson, 1942, 1945; Levan, 1942; McClintock, 1929; Punyasingh, 1947; Satina and Blakeslee, 1938). The aneuploid progeny swarm from a selfed CWW triploid of A. thaliana has been described previously (Henry et al., 2005), but the progeny karyotypes remain elusive. In order to reduce the genotypic complexity of the aneuploid swarm, triploids were used in pseudo-backcrosses (pBC) to either a diploid Col-0 or a tetraploid 4×-Col. A limited number of allelic ratios are expected, corresponding to the following %Wa-1 values: 0 (CC, CCC genotypes), 0.25 (CCCW genotype), 0.33 (CCW genotype), 0.5 (CCWW or CW genotype) and 0.66 (CWW genotype). Fluorescent genotyping was performed on 245 pBC1 progeny. The fluorescence intensity of the peaks associated with each allele was recorded and analysed for relative intensities (see Figure 1a and Experimental procedures for details). For all markers, the relative intensities of the two allelic peaks were expressed as the percentage of the parental Wa-1 PCR product (%Wa-1; Figure 1b). For 14 of the 16 markers analysed, the %Wa-1 values were distributed into four distinct heterozygous categories, allowing us to infer the number of copies of each allele at these loci (Figure 1c).
From genotype to karyotype
The expected chromosome number was inferred from each genotype at each marker. In the case of heterozygous marker genotypes, the number of each type of chromosome could be easily derived from the genotype. For example, a Col-0 to Wa-1 ratio of 2 to 1 at a given marker indicates trisomy for the chromosome carrying that marker. Data from multiple markers on each chromosome were considered. The karyotype of an individual was constructed from the genotypes at all 12 markers, including two markers on each chromosome and two additional markers on chromosome 1. An example of the genotypes at all 12 loci and the inferred karyotype is shown for plant TC211 in Figure 2(b). When data from markers on the same chromosome were not consistent, the karyotype at that chromosome was regarded as unresolved, except for chromosome 1 which was considered resolved if data from three of the four markers were consistent. In the progeny of the pBC, there were 67 unresolved chromosomes out of 1590 comparisons (4.2%).
Homozygous genotypes in the progeny of the pBC were always of the C type and posed a special challenge. Although loci that appear homozygous for C in the gamete are expected to be encoded by monosomic chromosomes, a triploid CWW individual can produce a gamete with homozygous CC loci on disomic chromosomes through a process called double reduction. For each individual marker, this process occurs at low frequency through recombination between C and W chromatids and co-segregation of C chromatids to a meiosis II product. Therefore, double reduction in a triploid requires trivalent pairing of the three replicated chromosomes. In a CCC triploid this was observed in 70–90% of microspore mother cells, depending on the chromosome (Steinitz-Sears and Lee-Chen, 1970). Next, trivalents must separate and the two chromatids carrying the C alleles must co-segregate to the same pole. The frequency of co-segregation depends on the distance of the marker from the centromere such that it varies from 1/15 to 0/15 for a centromere-unlinked and centromere-linked marker, respectively. Despite the low frequency of double reduction, C genotypes that appear homozygous could therefore be consistent with one or two copies of the corresponding chromosome in the gamete produced by the CWW triploid.
In some cases, the uncertainty was resolved by querying markers located on the same chromosome. For example, markers scored as W (gamete genotype) located on the same chromosome as C genotypes confirmed the monosomic nature of the C allele and the absence of double reduction. We determined that chromosome 5 was the most likely to undergo double reduction at both markers. The expected frequency of double reduction at both markers MN5.1 and MSAT5.19 was approximately 1 out of 1100 meioses. The possibility of double reduction was therefore ignored in cases where chromosomes exhibited C genotypes at both (or all) markers. Finally, in instances in which a marker scored as WW or CW (disomics) was located on the same chromosome as a C genotype, it was impossible to distinguish between double reduction and a genotyping error. To maintain a conservative estimate of frequency of aneuploidy, data for these chromosomes were scored as unresolved. Of the 67 chromosomes scored as unresolved, 24 instances of possible double reduction at a single marker on a chromosome were observed.
Detection of aneuploidy using comparative genomic hybridization
Comparative genomic hybridization of nuclear DNA content was used to confirm the karyotype of one of the aneuploid plants. Plant TC211 was chosen because of its complex karyotype (two copies of chromosomes 2 and 5 and three copies of chromosomes 1, 3 and 4). Labelled DNAs from TC211 and from a diploid Col-0 control were co-hybridized to the microarray slides and the ratios of fluorescence intensity from the two labelled DNA samples were recorded. Averaging these ratios over a sliding window of 101 features along the chromosomes (Figure 2a) highlighted differences in the relative dosage of each chromosome type and confirmed the karyotype obtained through quantitative genotyping (Figure 2b).
Relationship between nuclear DNA content and karyotype
We analysed the plants originating from the pBC by flow cytometry to determine the total nuclear DNA content. We experimented with a simplified protocol allowing for up to 20 experimental samples and four control samples to be prepared in 1 h (see Experimental procedures). Instead of using a heterologous internal size standard, we employed control samples derived from A. thaliana individuals of known genome content that were analysed along with each set of experimental samples.
For each plant, two values for genome content were obtained: one from the flow cytometric analysis (fGC) and one calculated from the molecular karyotype (kGC) as described below. The relative size of each chromosome type was determined based on version 6.0 of the Col-0 genome (http://www.arabidopsis.org/) and genome content was calculated according to the following formula: kGC = (24.5 × no. of Chr. 1 + 15 × no. of Chr. 2 + 20 × no. of Chr. 3 + 18.5 × no. of Chr. 4 + 22 × no. of Chr. 5)/100. On this scale, the kGC of a diploid individual would be 2. Comparison of kGC and fGC demonstrated that the two sets of values were highly correlated (Figure 3). The regression was significant and the data fit the following regression model kGC = 0.038 + 0.970 fGC (P < 0.0001; R2 = 0.983). These results suggest that both molecular karyotyping and flow cytometric measurement of nuclear DNA content are robust methods for determining the relative genome content within populations. As discussed earlier, flow cytometric analysis provided estimates of genome content but was uninformative as to the relative dosage of each type of chromosome (karyotype).
Chromosome counts using FISH confirm the inferred karyotypes
From the plants that were karyotyped, anthers were harvested from eight aneuploid individuals and used to perform chromosome counts using FISH. Although the exact karyotype of each plant could not be obtained using this technique, the number of chromosomes could be determined by counting the number of brightly staining spots after hybridization with a FITC-labelled centromeric DNA probe. For each of these eight individuals, the number of chromosomes observed in somatic cells using FISH was in agreement with the karyotypes previously determined using our karyotyping method. Hybridization results for four of these individuals are shown in Figure 4.
Tetraploid A. thaliana produces more than 25% aneuploid progeny
Quantitative genotyping was used to estimate the level of aneuploidy in a tetraploid population of A. thaliana. Tetraploid 4×-Col was crossed to the natural tetraploid Wa-1 in both directions (three independent crosses in each direction). F1 seeds were planted and 49 WWCC and 47 CCWW plants were genotyped at the 12 loci listed in Table 4. No homozygous individuals were detected, indicating that there was no contamination from accidental selfing. For each plant, a complete karyotype was inferred from the genotyping data, as described earlier. Nineteen out of 475 chromosomes (4%) were initially unresolved. These PCR reactions were re-run, allowing the resolution of all of them. Aneuploidy was indicated by a deviation in the %Wa-1 from the 1:1 ratio expected for CCWW and corresponded either to a missing or an additional chromosome. The origin of the aneuploid gametes was inferred from the genotype of the aneuploid individual. For example, the CCW genotype in an individual derived from a 4×-Col × Wa-1 cross indicated lack of a Wa-1 chromosomal copy in the pollen. Similarly, a CCCWW genotype from the same cross indicated an additional copy of a Col-0 chromosome in the ovule. Pooling of the results obtained in the different crosses allowed us to compare the rate of production of aneuploid gametes by 4×-Col and Wa-1 as well as by the female and male gametophytes.
|Name||Chr.||Physical location (bp)*||Primer sequence (5′ to 3′)||Size of PCR fragment (bp)||Parameter used for quantitative analysis|
|MN1.5||1||898627||F: HEX-TTATTATCAAGATCAAAGATTGTATGGTTT R: CTTGTTTTTATATCTGTTTGGTTTAATTGT||308–320||Height of highest peak|
|MN1.6||1||11362256||F: ROX-TAGTGGAAAGCTGTGCGATG R: CTCACCATGTTGTCCGAATG||189–209||Height of highest peak|
|MN1.7||1||13634035||F: NED-GCAAATCGCCTGTTTTCTTG R: CATCTGCGACTGAGAGTTCAA||217–227||Height of highest peak|
|MN1.2||1||21949869||F: ROX-TCAATCTCAACATCGGATCAA R: TGCTTCCAACAAGTGACAATG||179–189||Height of highest peak|
|nga1145||2||683000||F: HEX-GCACATACCCACAACCAGAA R: CCTTCACATCCAAAACCCAC||202–213||Sum of the height of the two highest peaks|
|nga1126||2||Approximately 9000000||F: 6-FAM-GCACAGTCCAAGTCACAACC R: CGCTACGCTTTTCGGTAAAG||191–196||Height of highest peak|
|nga162||3||4608284||F: HEX-CTCTGTCACTCTTTTCCTCTGG R: CATGCAATTTGCATCTGAGG||107–122||Sum of the height of the two highest peaks|
|MN3.3||3||15770898||F: 6-FAM-CGCAAGATCGTCATAATACAAAAGT R: AAGAACTCAAAGCAGCTATTCACAC||338–358||Height of highest peak|
|MN4.2||4||4197691||F: 6-FAM-TAAGGTCAGACTATATGTTTACGTTTCATT R: GTCATCCTCGTTTAAGTTACGATTG||347–357||Height of highest peak|
|MST4.18||4||10931201||F: NED-TGTAAATATCGGCTTCTAAG R: CTGAAACAAATCGCATTA||159–165||Height of highest peak|
|MN5.1||5||7843029||F: NED-GAAAACCATGTCTATTAACAACAACAAC R: AGCTCTAACACGTTTCCCAAGTATAA||430–410||Height of highest peak|
|MSAT 5.19||5||25911521||F: HEX-AACGCATTTGCTGTTTCCCA R: ATGGTTATCTCATCTGGTCT||199–210||Height of highest peak|
A total of 36 instances of aneuploidy were detected (Table 1). Comparison of the rates of aneuploidy obtained in the progeny of replicate crosses (4×-Col × Wa-1 A1 to A3 as well as Wa-1 × 4×-Col B1 to B3) suggested that two of the six crosses probably involved an aneuploid parent. Specifically, eight of the 20 plants analysed from cross A1 contained an extra copy of chromosome 5 from Col-0, suggesting that the Col-0 seed parent used in that cross was pentasomic for chromosome 5. Similarly, five of the 16 plants analysed from cross A3 contained an extra copy of chromosome 1 from Wa-1, suggesting that the Wa-1 pollen parent used in that cross was pentasomic for chromosome 1. Therefore, data from crosses A1 and A3 were omitted from further analyses. With these aneuploids excluded, 16 of the 56 remaining F1 plants (27.6%) were determined to be aneuploid.
As described in Table 2, most instances of aneuploidy were associated with an additional copy of a chromosome type (hyperploidy) as opposed to the lack of a chromosome type (hypoploidy). Specifically, only four of the aneuploid genotypes corresponded to the lack of a chromosome while the remaining 14 aneuploid genotypes corresponded to one or two additional chromosomes (4/18 versus 14/18, χ2P-value < 0.0001). Our data suggest that the natural autotetraploid Wa-1 produced more aneuploid gametes than the synthetically induced tetraploid of Col-0 (12/18 versus 6/18, χ2P-value = 0.0027). This was true both for pollen grains and ovules (Table 3). The frequency of aneuploidy in the functional pollen grains was higher than in the functional ovules but the difference was not significant (7/18 versus 11/18, χ2P-value = 0.053). Finally, the frequency of altered number of chromosome copies depended on the chromosome type. Specifically, six of the aneuploid individuals were aneuploid for chromosome 1 while only one individual was aneuploid for chromosome 3.
|Chromosome no.||No. of individuals||% Individuals|
|Col-0 pollen||Col-0 ovule||Wa-1 pollen||Wa-1 ovule||Extra chr.||Missing chr.||All|
QF-PCR is a rapid and low-cost method for quantitative genotyping in polyploid species
Our results demonstrate that allelic dosage and complete karyotypes can be determined by QF-PCR in A. thaliana. We used alleles distinguishable by small size polymorphisms and analysed them by capillary electrophoresis of PCR products in which one of the primers was fluorescently labelled. Analysis of the relative intensity of the two allelic peaks (Figure 1a,b) allowed the distinction between the three expected heterozygous states, and corresponding genotypes, associated with tetraploidy as well as those associated with aneuploidy (Figure 1c).
Similar to conventional non-quantitative genotyping methods, QF-PCR has the advantage of being easy to perform, produces data that are easily interpreted and uses robust, widely available technology. The method can be applied to any species for which any fragment size polymorphism exists and is flexible, as more markers can be added to the analysis at any point. Additionally, the method is scalable to higher throughput as PCR reactions can be run in microtitre plates and samples can be robotically loaded. Also, the number of electrophoresis runs can be drastically reduced by choosing fragment sizes and fluorophores that do not overlap and running these markers in the same capillary. We typically run six markers together, but higher levels of multiplexing could easily be achieved.
Quantitative genotyping using QF-PCR can therefore be a very useful tool both for breeding and basic genetic studies in the many economically important autopolyploid species such as potato, coffee or alfalfa. Moreover, it can also be used in diploid populations, where it provides a tool for performing bulk segregant analysis and measuring allelic ratios at the population level.
Aneuploidy detection is required in polyploid populations
In addition to the problem of polysomic inheritance of alleles, mapping in polyploid populations can be complicated by the presence of aneuploid individuals. Although aneuploidy is rare post-birth in humans, it is common in other species. For example, systematic aneuploidy detection using comparative microarray hybridization demonstrated whole chromosome or segmental aneuploidy in 8% of 300 deletion mutant strains of Saccharomyces cerevisiae (Hughes et al., 2000). Using the same technique, several instances of aneuploidy were detected in the genome of two lager yeast strains (Bond et al., 2004).
In plants, aneuploid individuals appear spontaneously within diploid populations at low frequency (Khush, 1973). In most species, these are easily detectable because they are phenotypically distinct from the diploid individuals and sometimes exhibit reduced fertility (Khush, 1973). In polyploid populations, on the other hand, the frequency of aneuploid individuals can increase drastically. We found that more than 25% of the progeny of tetraploid A. thaliana were aneuploid (Table 1) and that most of the aneuploid plants observed in our study carried an extra copy of a chromosome while only a few were lacking a chromosome copy (Table 2). Depending on the species, the percentage of aneuploid individuals detected within tetraploid populations is variable: 50% in maize (Randolph, 1935); 44% in barley (Rommel, 1961); 25% in Datura (Belling and Blakeslee, 1924); 15–23% in rye (Hagberg and Ellerstrom, 1959; Morrison, 1956; Muntzig, 1951); 20% in lettuce (Einset, 1947) and 3–6% in alfalfa (Bingham, 1968). This suggests that different species might produce aneuploid gametes at different rates or exhibit varying levels of tolerance of aneuploidy. Consistent with our data, the majority of the aneuploids in these studies were also found to be hyperploid. These data are consistent with the idea that extra chromosomal copies are less deleterious than the lack of a chromosomal copy (Birchler et al., 2001). Contrary to trisomic individuals, which exhibit obvious phenotypes in Arabidopsis (Koornneef and Van der Veen, 1983; Lee-Chen and Steinitz-Sears, 1967; Steinitz-Sears, 1962; Steinitz-Sears and Lee-Chen, 1970), most of the aneuploid plants with near-tetraploid genome contents were not readily identifiable by casual phenotypic inspection, probably because of a reduced relative chromosome imbalance in polyploid backgrounds. If this fraction of the population is undetected, their phenotypes can distort the analysis of many traits. For example, the presence of aneuploids within tetraploid populations of many crops has been shown to results in reduced seed set (Einset, 1944, 1947; Hagberg and Ellerstrom, 1959; Morrison, 1956; Rommel, 1961; Shaver, 1962). Quantitative genotyping makes it possible to tackle this problem by allowing the rapid scoring of a previously unexplored quantitative trait: aneuploid production.
Our experimental design allows the aneuploidy to be traced to a specific parent and gametophyte (Table 3). Of the two tetraploids used in this study, Wa-1 is a natural autotetraploid and 4×-Col originates from a colchicine-induced line. Surprisingly, the rate of aneuploid gametes produced was higher in Wa-1 than in 4×-Col arguing against a stabilization of tetraploid meiosis in Wa-1. It is also possible that the higher incidence in aneuploidy originates from an increased tolerance to aneuploidy in Wa-1 compared to Col-0. Seed and pollen parents produced aneuploid gametes at similar rates but aneuploidy was more often attributed to certain chromosome types than others. Specifically, chromosome 1 was most often found to be unbalanced with respect to the other chromosome types.
QF-PCR allows complete karyotyping of aneuploid individuals
Quantitative genotyping using QF-PCR analysis also allows karyotyping. We were able to infer full karyotypes from genotypes at two markers per chromosome. As a tool for aneuploidy detection and characterization, QF-PCR is faster, easier to perform and much more amenable to large-scale experiments than FISH. Moreover, in A. thaliana, like in many other species, karyotypes cannot be obtained from conventional cytological analysis because of the small size of the chromosomes and the lack of chromosome-specific probes. Quantitative fluorescent-PCR, on the other hand, can be performed in any organism for which polymorphic loci are known or can be identified, enabling the study of the effect of karyotype on phenotype in organisms not amenable to classical cytological studies. Finally, QF-PCR is well suited for the detection of segmental aneuploidies, as additional markers can be selected from a genomic region of interest. It does not, however, allow the detection of other types of chromosomal abnormalities visible by FISH, such as chromosomal rearrangements or translocations. Additionally, it is possible that in aneuploid individuals, aberrant chromosome segregation during mitosis produces mosaics, in which a small percentage of cells contain a different number of chromosomes from the rest of the cells. While beyond the resolution of our technique, such a phenomenon could decrease the sensitivity of the quantitative genotyping. Finally, another limitation of QF-PCR is that it requires heterozygosity. It may be possible to develop a modification of this method in which the DNA of a queried inbred individual is mixed with a known amount of a related polymorphic DNA from a diploid so that allelic ratios can be detected for each chromosome and used as measures of the karyotype.
As presented in this work and in others earlier (Bond et al., 2004; Hughes et al., 2000; Sebat et al., 2004), comparative genomic hybridization can be used to measure chromosomal dosage relative to a control sample. In our case, the karyotype of one of the aneuploid plants was confirmed using comparative genome hybridization to microarray slides (Figure 2). The use of microarrays for comparative genomic hybridization has the advantage of simultaneously determining gene dosage over small genomic regions and spanning the whole genome. Quantitative fluorescent-PCR, on the other hand, allows the processing of a large number of individuals at once and is significantly less expensive than microarrays. In our study, we calculated that producing a single karyotype costs less than $10 using QF-PCR, compared to approximately $500 using a microarray analysis involving a dye swap.
Finally, flow cytometry has been used in many polyploid species to measure nuclear genome content in polyploid and aneuploid plants (Burton and Husband, 2001; Roux et al., 2003). Our data demonstrate that the flow cytometric measurement of nuclear genome content can provide data consistent with QF-PCR (Figure 3). Nevertheless, flow cytometry can only identify aneuploidy that results in differences that are larger than 5% of the euploid state and does not provide a karyotype.
Applications to the study of aneuploid syndromes
The mechanisms behind aneuploid syndromes are still generally poorly understood. In humans, only trisomies of chromosomes 13, 18 and 21 are described clinically (Epstein, 1986), suggesting that trisomies of other chromosomes or more complex aneuploidies are lethal. For reasons that remain unclear, plants are more tolerant of extra chromosomes than animals (Matzke et al., 2003). Individuals trisomic for one of each individual chromosome have been described in several species (Blakeslee, 1922; Goodspeed and Avery, 1939; Khush, 1973; McClintock, 1929; Steinitz-Sears, 1962; Tsuchiya, 1967) and individuals with multiple trisomics have been reported in some species as well (Henry et al., 2005; Johnsson, 1942; Levan, 1942; McClintock, 1929; Rick and Notani, 1961; Satina and Blakeslee, 1938). Because most of these studies were performed at times when molecular technologies were unavailable, chromosomes were counted for aneuploid individuals but full karyotypes were usually not determined. Moreover, the numbers of individuals analysed were limited by the labour-intensive microscopic observations of chromosome spreads. Using QF-PCR, we were able to successfully, rapidly and robustly karyotype over 250 aneuploid individuals of A. thaliana. This method opens a door for a more in-depth study of aneuploid syndromes, the phenotypes associated with specific aneuploidies and the molecular mechanisms underlying these phenotypes on a population level.
Plant materials: lines, growth condition and crosses
All plants were grown on soil (Sunshine Professional Peat-Lite mix 4; SunGro Horticulture, Vancouver, BC, Canada) in a growth room lit by fluorescent lamps (model TL80; Phillips, Sunnyvale, CA, USA) at 22 ± 3°C with a 16-h:8-h light:dark photoperiod or in a greenhouse at similar temperatures and light regimes, with supplementary light provided by sodium lamp illumination as required.
Tetraploid lines were generated as previously described (Henry et al., 2005). Col-0 represents the diploid ecotype Columbia; 4×-Col represents tetraploidized Col-0 and Wa-1 represents the naturally occurring tetraploid ecotype Warschau-1 (CS6885). C and W refer to basic genomes or alleles of Col-0 and Wa-1, respectively. The CWW triploid was generated by crossing Col-0 as the seed parent to Wa-1. The CCWW and WWCC tetraploid plants were generated by the Wa-1 and 4×-Col tetraploid plants, with the seed parent genomic symbol listed first. Plant TC211 (Figure 2) and TC496, TC493, TC468 and TC473 (Figure 4) is one of the aneuploid plants produced by a crossing of a CWW triploid to Col-0.
Flow cytometric determination of genome content
Unopened buds from the apices of three inflorescences were harvested from an individual plant and laid in a round Petri dish containing 1.5 ml of ice-cold chopping buffer (15 mm HEPES, 1 mm EDTA, 80 mm KCl, 20 mm NaCl, 300 mm sucrose, 0.20% Triton-X, 0.5 mm spermine, pH = 6.1; β-mercaptoethanol was added to 0.1% v.v immediately before use). The tissue was chopped by hand using a fresh carbon-steel razor blade (VWR catalogue no. 55411-050) for approximately 1 min. The tissue suspension was transferred to a 5 cm3 luer-loc syringe (Becton-Dickinson, Franklin Lake, NJ, USA) fitted with a 13 mm filter holder (Millipore, Billerica, MA, USA) containing a 30 μm nylon mesh filter (CMN-30 monofilament cloth, Small Parts Inc., Miami Lakes, FL, USA). The fluid containing the nuclei was gently passed from the syringe through the filter and into a microcentrifuge tube that was kept on ice. Samples were centrifuged at 500 g for 7 min at 4°C to pellet the nuclei and the supernatant was discarded. Nuclei were resuspended in 0.4 ml of ice-cold staining solution (chopping buffer supplemented with 50 mg l−1 RNAse A and 10 mg ml−1 propidium iodide) by vortexing the tubes briefly. The tubes were stored in the dark and on ice for at least 2 h prior to analysis to allow the binding of propidium iodide to DNA to approach equilibrium.
All samples were analysed with a Becton-Dickinson FACScan flow cytometer using the cellquest software. Data from a minimum of 10 000 events were collected from each sample. For each sample, the mean fluorescence intensity of the basal somatic DNA content peak (2C peak) produced by the cellquest software was saved for calculation of genome content. In a given day, up to 20 samples were analysed for genome content and at least two Col-0, one CWW triploid and one tetraploid plant (4×-Col, CCWW or Wa-1) were used as control samples. These control samples were all run three times on each day: at the beginning, in the middle and at the end of the data collection. Using data from these control samples, a standard curve describing the relationship between genome content and mean fluorescence intensity of the 2C peak was calculated and used to determine the genome content of the experimental samples (Henry et al., 2005). All genome content values were expressed as a multiple of the haploid genome content of Col-0. On this scale, a genome content of 2.0 corresponds to a diploid individual and a genome content of 4.0 corresponds to a tetraploid individual.
Quantitative genotyping and molecular karyotyping
Genomic DNA was extracted from leaf or bud tissue using the Q-Biogene Fast DNA kit (Q-Biogene, Carlsbad, CA, USA) according to the manufacturer's recommendations, and diluted 1:5 in water (v.v). Genetic marker loci were amplified by PCR. Reactions contained 2 μl of diluted DNA, 2 μl of 10× PC2 buffer (ABPeptides, St Louis, MO, USA), 0.8 μl of 2.5 mm deoxynucleotide triphosphates (dNTPs), 0.2 μl of fluorescently labelled forward primer (20 μm), 0.2 μl of unlabelled reverse primer (20 μm), 1 unit of Taq polymerase (Klentaq; ABPeptides) and H2O for a total volume of 20 μl per reaction. A touchdown PCR program was used with annealing temperature sequentially decreasing from 62 to 52°C (1°C/cycle), followed by 25 cycles at 48°C. Each cycle included 10 sec at 94°C, 30 sec of annealing and 40 sec at 72°C. Finally, the program included an initial 2 min at 94°C and a final 4 min at 72°C. Information concerning the sequence and modification of the primers is summarized in Table 4.
Polymerase chain reaction products were diluted in TE (10 mm Tris–HCl, 1 mm EDTA, pH = 8.0) before loading on a 3100 Genetic Analyzer (Applied Biosystems, Foster City, CA, USA). The efficiency of amplification and the fluorescence intensity of the labelled primers varied between markers. To help equalize the intensities of all markers, a dilution factor was determined for each marker. Polymerase chain reaction products from five independent reactions were diluted 1:10, 1:25 and 1:100 in TE and the 15 samples were run side by side. The average fluorescence intensity of the allelic peaks was recorded and used to infer the value of a dilution factor that result in measured peak fluorescence intensities between 1000 and 6000 units. Depending on the marker, dilution factors ranged from 1:2 to 1:50. Once a dilution factor was determined for each marker, it was used in all subsequent experiments using this marker.
Polymerase chain reaction products from reactions using different primer sets were mixed and diluted in TE using these dilution factors. One microlitre of the mix was transferred to a final 96-well plate and supplemented with 15 μl of Hi-DiTM Formamide (Applied Biosystems). The plate was incubated at 90°C for 1 min before being immediately chilled on ice. The PCR products were electrophoretically separated by size on a 3100 Genetic Analyzer using a 36-cm array and POP6 polymer (Applied Biosystems).
Data obtained were analysed using GeneMarker v1.3 (SoftGenetics LLC, State College, PA, USA) without using a size standard. After baseline subtraction, one or several peaks were identified for each allele (Col-0 or Wa-1). Homozygous individuals were scored directly (in our pseudo-backcross experiments, homozygous individuals were always homozygous for the Col-0 allele). Heterozygous individuals were scored by comparing the height of the Col-0 and Wa-1 allelic peaks. Depending on the marker, the height of the highest peak or the sum of the heights of the two highest peaks for each allele were recorded (Table 4). For each individual and each marker, the allelic ratio (R), expressed as the percentage of Wa-1 allele, was calculated using the following formula: R = HW/(HW + HC) where HW and HC refer to the Wa-1 and Col-0 peaks, respectively. The ratios of fluorescence intensity measured for the two alleles at each locus were converted into genotypes using a standard curve derived from control samples of known allelic ratios (tetraploid CCWW and triploid CWW plants). A different standard curve was built for each marker and at least four control samples were included in each 96-well plate run. In the tetraploid population, any deviation from the 1:1 ratio indicated aneuploidy. The four possible %Wa-1 values associated with aneuploidy were 33% (CCW), 40% (CCCWW), 60% (CCWWW) and 66% (CWW). In order to unambiguously distinguish between these aneuploid types, two additional steps were taken. For each marker, data obtained from the analysis of the pBC were used to calculate a standard curve describing the relationship between %Wa and genotype. The %Wa-1 value obtained for the aneuploid genotypes within the tetraploid population were than compared with those obtained in the pBC, allowing the identification of the trisomic individuals (pentasomic individuals were not detected in the pBC). Next, a new set of PCRs was run for each marker that included most of the aneuploid individuals, four control CWW individuals, three control CCW individuals and two control CCWW individuals. As expected, the trisomic individuals exhibited %Wa-1 values similar to the CCW and CWW controls while the assumed pentasomic individuals exhibited %Wa-1 values intermediate between trisomy and the CCWW controls.
We made several observations about the use of fluorescent PCR for this purpose. Firstly, primers surrounding microsatellite sequences (MSAT and nga primers in Table 4) produced signals that were more difficult to score quantitatively than primers amplifying a region containing a short deletion in one of the two genomes (MN primers in Table 4). This was due to the presence of multiple peaks (stutter) for each allele when using microsatellite markers compared to a single peak per allele for the other markers. Similarly, longer primers (30-mers instead of 20-mers) generated more reliable results. Secondly, a separate standard curve describing the relationship between allelic ratios in sample DNA and allelic ratios of the PCR product must be calculated for each primer because one of the two alleles often amplified more efficiently than the other. For example, a CCWW individual generated allelic peaks corresponding to varying %Wa-1 values, depending on the marker that was amplified (from 0.41 for MN5.1 to 0.84 for MSAT5.19). Finally, it is best to run as many samples as possible at once and to include standard samples of known allelic ratios within each PCR run because slightly different PCR conditions may result in peaks of slightly different relative intensities even when using identical templates.
Comparative genome hybridization
Genomic DNA was extracted from clusters of buds of a Col-0 plant and plant TC211 using the Q-Biogene Fast DNA kit (Q-Biogene, Morgan Irvine, CA, USA) following the manufacturer's recommendation. Two separate aliquots of 300 ng of genomic DNA from each plant were supplemented with water to a final volume of 10 μl. The DNA was labelled with either Cy3 or Cy5 dUTP through random priming. Firstly, 20 μl of 2.5× Random Primers Solution (Bioprime Array CGH Genomic Labelling System catalogue no. 18095-12; Invitrogen, Carlsbad, CA, USA) was added to each tube and the mix was denatured for 10 min in boiling water before being immediately chilled on ice. A total of 50 μl was obtained by adding 11 μl of water, 5 μl of 10× dUTP Nucleotide Mix (Bioprime kit; Invitrogen), 3 μl of Cy3-dUTP or Cy5-dUTP (catalogue no. PA53022 or PA55022; Amersham Biosciences, Piscataway, NJ, USA) and 1 μl of Exo− Klenow fragment (Bioprime kit; Invitrogen) to the 30 μl of the DNA–primer mix. The content of each tube was gently mixed and incubated at 37°C overnight. Probe sets for array hybridization were designed to carry out a dye swap using a total of two slides (Churchill, 2002; Quackenbush, 2002; Yang et al., 2002). Cy5-labelled Col-0 DNA and Cy3-labelled TC211 DNA were combined in one tube while Cy5-labelled TC211 DNA and Cy3-labelled Col-0 DNA were combined in a second tube. Unincorporated nucleotides were removed using the Qiaquick PCR purification kit (catalogue no. 28104; Qiagen, Valencia, CA, USA) with a final elution volume of 50 μl of EB buffer. Finally, 48 μl of water, 18 μl of 20× SSC and 6 μl of yeast tRNA (10 μg μl−1, catalogue no. 15401-029; Invitrogen) were added to reach a final volume of 120 μl. Hybridization, washes and scanning were performed at the Microarray Facility of the Fred Hutchinson Cancer Research Center (Seattle, WA, USA). Arrays contained 26 090 70-mer oligonucleotides (GEO accession no. GPL1911; Arabidopsis Genome Oligo Set Version 1.0; Operon Biotechnologies, Huntville, AL, USA) spotted on epoxy-coated glass slides, and corresponding to 26 029 A. thaliana genes. Image conversion was performed using GenePix Pro 6.0 (Axon Instruments/Molecular Devices Corp, Sunnyvale, CA, USA). The data were subsequently analysed using the open-source software packages express converter 1.7 and midas (Saeed et al., 2003). Features for which the signal to noise ratio was less than 2 were discarded. Red and green fluorescent intensities for each feature were normalized using a locally weighted linear regression (LOWESS). Dye-swap pairs were checked for consistency and features with low reproducibility across the dye swap (two standard deviations) were excluded. For the remaining features, merged intensities were calculated as the geometric mean of the flip-dye replicates. Finally, the ratio of intensities was calculated by dividing the merged intensities and these values were plotted against the genomic location of their corresponding feature (Figure 2) using CGH-Explorer (Lingjaerde et al., 2005).
Fluorescent in situ hybridization (FISH)
Developing flower buds were harvested and used for chromosome counts by FISH with centromeric probes as previously described (Comai et al., 2003) except that the following steps were performed prior to chromosome denaturation: the slides were first incubated in 100 μg ml−1 ribonuclease A (catalogue no. R-4642; Sigma-Aldrich, Wilwaukee, WI, USA) in 2× SSC for 30 min at 37°C. The slides were subsequently washed three times for 5 min each in 2× SSC and finally for 1 min in 10 mm HCl. The slides were incubated with 1 μg ml−1 pepsin (catalogue no. P6887; Sigma-Aldrich) in 10 mm HCl for 10 min at 37°C before being rinsed in water and washed three times for 5 min each in 2× SSC. The slides were then treated with 4% formaldehyde for 10 min at room temperature (instead of 65°C for 15 min in 70% formaldehyde) and washed three times for 5 min each in 2× SSC before being dehydrated in a series of ethanol dilutions (70%, 90% and 100%). After hybridization, the slides were stained with 0.25 μg ml−1 4′,6-diamidino-2-phenylindole (DAPI).
We thank the Comparative Genome Center (Department of Biology, University of Washington) and the Murdock Foundation for use of the ABI3100 Genetic Analyzer and of software for data analysis. We thank the Cell Analysis Facility (Department of Immunology, University of Washington) and Biology Greenhouse (Biology Department, University of Washington) for material support. We thank the Fred Hutchinson Cancer Research Center microarray facility for carrying out the hybridizations. We thank the Magnus Nordborg laboratory for the sequence database of A. thaliana ecotypes. We thank Hsin-Yi Lin for help with the flow cytometry analysis, Jerry Davison for help with the microarray analysis and Anand Tyagi for performing the FISH experiments. This work was supported by the National Science Foundation Polyploidy Project (NSF DBI 0077774 and DBI 0501712 to LC), the USDA National Research Initiative (2003-35300-13248 to BPD) and the Washington Research Foundation (to IMH).
- 2006) Copy number polymorphism in Fcgr3 predisposes to glomerulonephritis in rats and humans. Nature, 439(7078), 851–855. , , et al. (
- 1924) The distribution of chromosomes in tetraploid daturas. Am. Nature, 58(654), 60–70. and (
- 1968) Aneuploids in seedling populations of tetraploid alfalfa, Medicago sativa L. Crop Sci. 8, 571–574. (
- 2001) Dosage-dependent gene regulation in multicellular eukaryotes: implications for dosage compensation, aneuploid syndromes, and quantitative traits. Dev. Biol. 234(2), 275–288. , , and (
- 2003) Dosage dependent gene regulation and the compensation of the X chromosome in Drosophila males. Genetica, 117(2–3), 179–190. , and (
- 2005) Dosage balance in gene regulation: biological implications. Trends Genet. 21(4), 219–226. , , and (
- 1922) Variation in Datura due to changes in chromosome number. Am. Nature, 56, 16–31. (
- 2004) Aneuploidy and copy number breakpoints in the genome of lager yeasts mapped by microarray hybridization. Curr. Genet. 45(6), 360–370. , , and (
- 2003) Large-scale identification of single-feature polymorphisms in complex genomes. Genome Res. 13(3), 513–523. , , , , , , , and (
- 1916) Non-disjunction as a proof of the chromosomal theory of heredity. Genetics, 1, 1–52. (
- 2001) Fecundity and offspring ploidy in matings among diploid, triploid and tetraploid Chamerion angustifolium (Onagraceae): consequences for tetraploid establishment. Heredity, 87, 573–582. and (
- 2004) Correct estimation of preferential chromosome pairing in autotetraploids. Gene Res. 14, 459–462. , and (
- 2002) Fundamentals of experimental design for cDNA microarrays. Nature Genet. 32(Suppl.), 490–495. (
- 2003) FISH analysis of meiosis in Arabidopsis allopolyploids. Chromosome Res. 11(3), 217–226. , and (
- 2001) Genetic markers and genotyping methodologies: an overview. Ecotoxicology, 10(4), 201–204. , and (
- 1986) Aneuploidy and inbreeding depression in random mating and self-fertilization autotetraploid populations. Theor. Appl. Genet. 72, 799–806. (
- 2005) Molecular methods for rapid detection of aneuploidy. J. Appl. Genet. 46(2), 207–215. , , , and (
- 1944) Cytological basis for sterility in induced autotetraploid lettuce (Lactuca sativa L.). Am. J. Bot. 31(6), 336–342. (
- 1947) Aneuploidy in relation to partial sterility in autotetraploid lettuce. Am. J. Bot. 34(3), 99–105. (
- 1986) The Consequences of Chromosomal Imbalance. Cambridge: Cambridge University Press. (
- 2004) Lineage-specific gene duplication and loss in human and great ape evolution. PLoS Biol. 2(7), E207. , , et al. (
- 2004) Cytometry and plant sciences: a personal retrospective. Cytometry A, 58(1), 37–44. (
- 1939) Trisomics and other types in Nicotiana sylvestris. J. Genet. 38, 381–458. and (
- 1959) The competition between diploid, tetraploid and aneuploid rye. Theoretical and practical aspects. Hereditas, 45, 369–415. and (
- 2005) Aneuploidy and genetic variation in the Arabidopsis thaliana triploid response. Genetics, 170(4), 1979–1988. , , , , and (
- 2000) Widespread aneuploidy revealed by DNA microarray expression profiling. Nature Genet. 25(3), 333–337. , , et al. (
- 1942) Cytological studies of triploids progenies of Populus tremula. Hereditas, 28, 306–612. (
- 1945) The triploid progeny of the cross diploid × tetraploid Populus tremula. Hereditas, 31, 411. (
- 1973) Cytogenetics of Aneuploids. New YORK: Academic Press. (
- 1983) Trisomics in Arabidopsis thaliana and the location of linkage groups. Genetica, 61, 41–46. and (
- 1967) The location of linkage groups in Arabidopsis thaliana. Can. J. Genet. Cytol. 9, 381–384. and (
- 1942) The effect of chromosomal variation in sugar beets. Hereditas, 28, 345–399. (
- 2005) CGH-Explorer: a program for analysis of array-CGH data. Bioinformatics, 21(6), 821–822. , , , and (
- 2003) Representational oligonucleotide microarray analysis: a high-resolution method to detect genome copy number variation. Genome Res. 13(10), 2291–2305. , , et al. (
- 2006) Modeling population genetic data in autotetraploid species. Genetics, 172(1), 639–646. , , , , , and (
- 1994) Stomotal size in fossil plants: evidence for polyploidy in majority of angiosperms. Science, 264, 421–424. (
- 2003) Does the intrinsic instability of aneuploid genomes have a causal role in cancer? Trends Genet. 19(5), 253–256. , , and (
- 1929) A Cytological and genetical study of triploid maize. Genetics, 14, 180–222. (
- 1956) Chromosome behavior and fertility of Tetra Petkus rye. Can. J. Agr. Sci. 36(3), 157–165. (
- 1951) Cytogenetic properties and practical value of tetraploid rye. Hereditas, 37, 17–84. (
- 2005) Rapid prenatal diagnosis of aneuploidy using quantitative fluorescence-PCR (QF-PCR). J. Histochem. Cytochem. 53(3), 285–288. , , , and (
- 2005) Frequency of new copy number variation in humans. Nature Genet. 37(4), 333–334. (
- 2000) Polyploid incidence and evolution. Annu. Rev. Genet. 34, 401–437. and (
- 2003) Dosage sensitivity and the evolution of gene families in yeast. Nature, 424(6945), 194–197. , and (
- 2003) Mutations and aneuploidy: co-conspirators in cancer? Cancer Cell, 4(2), 89–94. and (
- 1947) Chromosome numbers in crosses of diploid, triploid and tetraploid maize. Genetics, 32, 541–554. (
- 2002) Microarray data normalization and transformation. Nature Genet. 32 (Suppl.), 496–501. (
- 1998) Pathways, mechanisms, and rates of polyploid formation in flowering plants. Annu. Rev. Ecol. Syst. 29, 467–501. and (
- 2002) Neopolyploidy in flowering plants. Annu. Rev. Ecol. Syst. 33, 589–639. and (
- 1935) Cytogenetics of tetraploid maize. J. Agric. Res. 50, 591–605. (
- 2004) Molecular karyotyping using a SNP array for genomewide genotyping. J. Med. Genet. 41, 916–922. , , , , , , , and (
- 1961) The tolerance of extra chromosomes by primitive tomatoes. Genetics, 46, 1231–1235. and (
- 2002) Genotyping of SNPs in a polyploid genome by pyrosequencing. Biotechniques, 32(3), 592–593, 596–598, 600 passim. , , and (
- 2005) Quantitative genotyping of single-nucleotide polymorphisms by allele-specific oligonucleotide hybridization on DNA microarrays. Biotechnol. Appl. Biochem. 42, 93–96. , , , and (
- 1961) Aneuploidy, seed set and sterility in artificially induced autotetraploid Hordeum vulgare L. Can. J. Genet. Cytol. 3, 272–282. (
- 2003) Rapid detection of aneuploidy in Musa using flow cytometry. Plant Cell Rep. 21(5), 483–490. , , , and (
- 2003) TM4: a free, open-source system for microarray data management and analysis. Biotechniques, 34(2), 374–378. , , et al. (
- 1938) Chromosome behavior in triploid Datura. III. The seed. Am. J. Bot. 25, 595–602. and (
- 2005) Quantification of allele-specific expression of a gene encoding strawberry polygalacturonase-inhibiting protein (PGIP) using Pyrosequencing. Plant J. 41(3), 493–500. , and (
- 2002) Relative quantification of 40 nucleic acid sequences by multiplex ligation-dependent probe amplification. Nucleic Acids Res. 30(12), e57. , , , , and (
- 2004) Large-scale copy number polymorphism in the human genome. Science, 305(5683), 525–528. , , et al. (
- 1962) A study of aneuploidy in autotetraploid maize. Can. J. Genet. Cytol. 4, 226–233. (
- 2003) Plant Cytogenetics. Boca Raton: CRC Press. (
- 2005) Mapping segmental and sequence variations among laboratory mice using BAC array CGH. Genome Res. 15(2), 302–311. , , et al. (
- 2003) Advances in the study of polyploidy since Plant Speciation. New Phytol. 161, 173–191. , and (
- 1962) Chromosome studies in Arabidopsis thaliana. Genetics, 48, 483–490. (
- 1970) Cytogenetic studies in Arabidopsis thaliana. Can. J. Genet. Cytol. 12, 217–223. and (
- 1903) The chromosomes in heredity. Biol. Bull. 4, 231–251. (
- 1967) The establishment of a trisomic series in a two-rowed cultivated variety of barley. Can. J. Genet. Cytol. 9(4), 667–682. (
- 2003) A sigmoidal transcriptional response: cooperativity, synergy and dosage effects. Biol. Rev. Camb. Philos. Soc. 78(1), 149–170. (
- 2004) Gene dosage balance in cellular pathways: implications for dominance and gene duplicability. Genetics, 168(1), 569–574. (
- 2005) Gene dosage balance: deletions, duplications and dominance. Trends Genet. 21(1), 33–35. (
- 2006) High-resolution single-copy gene fluorescence in situ hybridization and its use in the construction of a cytogenetic map of maize chromosome 9. Plant Cell, 18(3), 529–544. , and (
- 2002) Normalization for cDNA microarray data: a robust composite method addressing single and multiple slide systematic variation. Nucleic Acids Res. 30(4), e15. , , , , , and (