Applications of Molecular Biology to Speed up The Processes of Crop Improvement

Plant breeding is based on the principles of Mendelian genetics. In the past, plant breeding was something of an art and selection of superior genotypes of a particular crop depended to a great extent on subjective decisions made by the breeder. With increasing knowledge of the genes underlying useful traits, plant breeding has become a more directed and scientific activity. This is in part a result of the generation of molecular maps of crop genomes, extensive sequencing of expressed sequences [expressed sequenced tags (ESTs)] and of genomic sequences and of study of genome organisation, repetitive and non-coding sequences and the ability to identify polymorphisms at particular loci which can be exploited as molecular markers if they are closely linked to a useful trait.
1. Molecular Maps of Crop Plants
Arabidopsis thaliana had been used as a model plant for mutagenic and genetic studies in the 1970s. The advantages of this species as a model for molecular studies became apparent in the 1980s because of its small nuclear genome size, low repetitive DNA content, short life-cycle, large seed production and later its amenability to transformation. As a result, through international collaborations, a genomic sequencing programme was established and the complete DNA sequence of the Arabidopsis genome was completed in 2000. The rice genome is only about four times the size of the Arabidopsis genome and, as a model cereal and important food crop, sequencing of this genome is also well advanced. Massive sequencing efforts of ESTs of wheat, barley, soybean, rice, Medicago truncatula and other crops are also in progress, mainly driven by major life sciences companies.
Much interesting information on genome organisation has resulted from this work, including knowledge of location of genes, gene clustering and repetitive and non-coding sequences. For cereals, gene arrangements show ‘synteny’ in that major blocks of genes are arranged in similar sequences in rice, maize, barley, wheat, etc. The major differences in genome size is the result of different amounts of repetitive/non-coding sequences and, for wheat, the fact that it is hexaploid and contains three sets of progenitor genomes.
For Arabidopsis, it emerges that there are about 22 000 genes required to contain all the information for this organism. For other plants, we might expect the number of genes present to be between this figure and about 50 000 genes.
2. Molecular Markers
A genetic marker is any character that can be measured in an organism which provides information on the genotype (i.e. genetic make-up) of that organism. A genetic marker may be a recognizable phenotypic trait (e.g. height, colour, response to pathogens), a biochemical trait (e.g. an isozyme) or a molecular trait (i.e. DNA based).Whereas phenotypic markers depend on expression of genes and are limited to those genes expressed at a particular time or under particular developmental or environmental conditions, DNA-based markers provide an almost unlimited supply of markers that identify specific sequences across the genome. Their advantages are:
(i) Single base changes in DNA can be identified, providing many potential marker sites across a genome.
(ii) They are independent of developmental stage, environment or expression.
(iii) Markers can be found in non-coding or repetitive sequences.
(iv) Most DNA marker sequences are selectively neutral.
Thus, for example, because about 80% of the wheat genome is noncoding DNA, only molecular markers can be used to identify polymorphisms and to map ‘loci’ in these regions of the genome.
3. Types of Molecular Markers
There are many potential approaches to identify molecular markers. Most are based on using the polymerase chain reaction (PCR) to amplify specific DNA sequences.3–5 They include:
(i) RFLPs (restriction fragment length polymorphisms);
(ii) RAPD-PCR (random amplified polymorphic DNA);
(iii) microsatellites or simple sequence repeats (SSRs);
(iv) AFLP (amplified fragment length polymorphisms).
RFLPs rely on the combination of a probe and restriction enzymes to identify polymorphic DNA sequences using Southern blotting. This approach requires either radioactive or non-radioactive detection methods to identify polymorphic DNA bands and is therefore more time
consuming than PCR-based methods.

RAPD-PCR does not require sequence information and involves amplifying random pieces of DNA in which PCR is primed by a single 10 base primer at low stringency, such that random sequences of DNA are amplified based on homologous sequences to the primer being present in the target DNA. It is a useful initial approach to identify polymorphisms, but is not regarded as reproducible enough between laboratories.
Microsatellites or SSRs are groups of repetitive DNA sequences that are present in a significant proportion of plant genomes. They consist of tandemly repeated mono-, di-, tri-, tetra- or pentanucleotide units. The number of repeats varies in different individuals and so the different
repeats can be regarded as ‘polymorphic’ alleles at that ‘locus’. To reveal polymorphic microsatellite sequences, it is necessary to sequence the conserved flanking DNA and to design PCR primers that will amplify the repeat sequences. (Because of the repetitive nature of the amplified sequences, typically the main amplified PCR band and additional ‘stutter’ bands are generated.) For example, at microsatellite locus Hspl76 of soybean, there is an AT repeat with 13 different numbers of bases in the repeated units in different soybean accessions.Microsatellites provide reliable, reproducible molecular markers.
AFLP is also a PCR-based technique, in which selective pre-amplification and amplification steps are carried out to amplify a subset of fragments of the genome, depending on the linkers added and primers used. Many potentially polymorphic fragments are generated by this approach. Polymorphic bands between parents can be identified and linked to useful traits.
Both microsatellite and AFLP markers can be analysed using autoradiography or a DNA sequencer, using fluorescent tags. The latter allows multiplexing such that three different coloured tags plus a size marker can be run in each lane. A single multiplexed AFLP gel can generate 100 polymorphic bands.
4. Marker-assisted Selection
Using one of the above approaches to identify molecular markers, in combination with an appropriate mapping population of plants plus or minus the trait of interest, many markers have been identified which are closely linked to genes for agronomic traits of interest. These include markers for genes for:
(i) pest and disease resistance (against viruses, fungi, bacteria, nematodes, insects);

(ii) quality traits (e.g. malting quality barley, bread or noodle wheats, alkaloid levels, etc.);
(iii) abiotic stresses (e.g. tolerance to salinity or toxic elements such as boron or aluminium);
(iv) developmental traits (e.g. flowering time, vegetative period).
If the molecular marker is in the target gene itself, it has been called a ‘perfect’ marker. Clearly, the closer the molecular marker is linked to the target gene, the better. The overall process of developing a marker thus involves setting up appropriate mapping populations, looking for polymorphic DNA sequences closely linked to the trait of interest, conversion of the polymorphism to a routine marker (usually PCR based), validation and implementation.
Quantitative trait loci (QTLs) are the genes which control quantitative traits such as yield for which the final character is controlled by several genes. To identify and map QTLs, a defined mapping population is required which is screened for polymorphisms by RFLPs, AFLPs and SSRs which can be mapped. Statistical approaches are then used to identify associations between the traits of interest and specific markers.
Although the location of QTLs is usually not known exactly, the association of a genotype at a marker/locus and a contribution to the trait indicates that there is a QTL near that marker. The promise of molecular marker-assisted selection for crop improvement is in the following: increased speed and accuracy of selection; stacking genes, including minor genes; following genes in backcross populations; and reduced costs of field-based selection. Thus, rather than growing breeding lines in the field and challenging or testing for important traits over the growing season, it is possible to extract DNA from 50mg of a seedling leaflet and test for the presence or absence of a range of traits in that DNA sample in one day. Plants lacking the required traits can then be removed early in the breeding programme. With the availability of more validated molecular markers, marker-assisted selection therefore becomes a highly cost-effective and efficient process.
5. Examples of Marker-assisted Selection
There are now many examples of the use of molecular markers for selection in plant breeding.3 Examples include (i) the microsatellite locus HSP176 of soybean , which is closely linked to a
gene (Rsv) conferring resistance to soybean mosaic virus, (ii) a perfect marker for noodle quality starch in wheat and (iii) a marker for early flowering in lupins.

Western Australia exports specialty wheat to the Asian market to make white alkaline salted noodles. This segment of the export trade is worth $250 million per annum. White noodles require specific swelling properties of starch. Noodle quality wheats all have two rather than three copies of granule bound starch synthase (GBSS), the enzyme which synthesizes amylose, the linear polymer of starch. This reduces the ratio of amylose to amylopectin by 1.5–2%, increasing the flour swelling volume. A ‘perfect’ PCR molecular marker was developed which identified presence or absence of the GBSS gene on chromosome A, i.e. ‘bad’ or ‘good’ noodle starch. This molecular marker test is now used as a primary screen for all noodle wheat breeding lines in Western Australia and has resulted in the accelerated production of a series of new noodle quality wheat varieties.
Narrow-leafed lupin is the major grain legume grown in Australia. Early flowering is required for the crop to complete its life-cycle before the rain limits growth in areas of Mediterranean climate where it is grown. Using fluorescent AFLPs, a marker linked to early flowering of lupins was identified (using a DNA sequencer). The AFLP was then run as a radioactive version, the polymorphic band isolated, cloned and sequenced and a co-dominant PCR-based marker developed for routine implementation.
6. Molecular Diagnostics
The same principles as used in developing molecular markers can be applied for a range of molecular diagnostic purposes in plants, including:
(i) identification of plant pathogens (viruses, fungi, nematodes,bacteria, insects);
(ii) studying population structure/variations in pathogens;
(iii) identifying the presence and quantifying the presence of transgenes in transgenic foods;
(iv) following possible pollen transfer of transgenes.
All that is needed is to identify specific nucleic acid (RNA or DNA) sequences unique to the target organism and then to develop a reliable extraction/PCR analysis system such that a DNA fragment is only amplified if the target organism or target sequence is present in a sample. The methods and scale by which such analyses (and also marker-assisted selection) can be carried out are advancing rapidly. Analysis can be by:
(i) PCR and gel electrophoresis;
(ii) real-time fluorescent PCR (e.g. using an ABI TaqMan 7700) to quantify the original amount of target sequence without gel electrophoresis;
(iii) matrix-assisted laser desorption/ionisation time-of-flight mass spectrometry (MALDI-TOF MS);
(iv) use of DNA chips and hybridisation of labelled samples to bound DNA sequences.
The last two methods (MALDI-TOF MS and chip technology), including microarrays ), promise to speed up all DNA fragment analysis applications by an order of magnitude over current gel-based DNA separation technologies.
Typical examples of applications of molecular diagnostics are the following. (1) Routine analysis of farmers’ seed samples (such as lupin) for the presence of seed-borne cucumber mosaic virus (CMV).8 This can done by RT-PCR (to detect viral RNA, sensitivity o1 infected seed per 1000 seeds) or by real-time fluorescent PCR. Farmers must buy clean seed if infection levels of CMV are above 0.5%. (2) Routine analysis of farmers’ lupin seed for the fungal disease anthracnose, caused by Colletotrichum acutatum. A PCR test, based on repeated ribosomal DNA sequences specific to the fungal pathogen, allows infection levels of one seed in 10 000 to be detected. In this case, only clean anthracnose-free seed can be sown.
7. DNA Fingerprinting, Variety Identification
The same processes of DNA fragment production and analysis can be applied to DNA fingerprint plants. The main use is in variety identification and quality control, but the techniques can equally be applied to study plant populations, taxonomy, conservation biology and rehabilitation of mine sites or cleared forests. For rehabilitation studies,DNA fingerprinting data can be used to ensure that an appropriate range of genotypes of species removed is used to rehabilitate cleared land.With the advent of end-point levies on delivered bulk grains rather than on royalties from seed sales (in some countries), there is a need for rapid and accurate identification of crop varieties at the receiving depots. This can be achieved by rapid analysis of DNA fingerprints of specific crop varieties.

8 .DNA Microarrays
DNA microarrays can be set up robotically by depositing specific fragments of DNA at indexed locations on microscope slides. With  current technology, cDNAs, EST clones or open reading frames (ORFs) from sequenced genomes can be set up in microarrays of 10 000 spots
per 3.24 cm2, thus the whole genome of Arabidopsis could be displayed on one microscope slide. Fluorescently labelled mRNA probes are hybridised on to the array and specific hybridising sequences are identified by their fluorescent signals. Microarray technology can be used to study gene expression patterns, expression fingerprints, DNA polymorphisms and, in theory, as a breeding tool to evaluate new genetic materials, for a specific trait (e.g. drought tolerance), together with phenotypic tests.
9.Bioinformatics
The generation of massive amounts of molecular information on plant genomes and their products from large-scale sequencing programmes, DNA fingerprinting and fragment analyses, mapping, molecular diagnostics, marker-assisted selection and DNA microarrays requires a concomitant increase in the ability to handle and analyse such data.
The handling and interpretation of molecular data are generally referred to as bioinformatics and marries requirements of computing power, data handling and appropriate software. Depending on the level of analysis, the area can be divided into ‘genomics’ (DNA level), ‘proteomics’ (protein expression level) and ‘metabolomics’ (metabolic level). In many cases, it is useful to ‘mine’ DNA or other databases to look for new or useful genes or sequences, using specific software, programs and algorithms.