Appendix 4. Plant breeding techniques

Table of Contents

1 Plant breeding techniques at crop and plant level A 4-2
1.1 Selection A 4-2
1.1.1 Mass selection A 4-2
1.1.2 Pedigree selection A 4-3
1.1.3 Site-determined selection A 4-3
1.1.4 Change in surroundings A 4-4
1.1.5 Change in sowing time A 4-4
1.1.6 Ear-bed method A 4-4
1.1.7 Indirect selection A 4-5
1.1.8 Test crosses A 4-5
1.2 Inducing variation A 4-6
1.2.1 Combination breeding A 4-6
1.2.2 Species crosses A 4-7
1.2.3 Bridge crosses A 4-7
1.2.4 Temperature treatment of the style A 4-7
1.2.5 Cut style A 4-8
1.2.6 Grafting on the style A 4-8
1.2.7 Mentor pollen technique A 4-8
1.2.8 Repeated backcrossing A 4-9
1.2.9 Hybrid breeding A 4-9
1.2.10 Parthenocarpy and cucumber A 4-10
1.2.11 Synthetic varieties A 4-11
1.2.12 Mutation breeding A 4-12
1.3 Propagation and maintenance A 4-12

2 Tissue culture techniques A 4-14
2.1 Propagation A 4-14
2.1.1 In vitro propagation A 4-14
2.1.2 Meristem culture A 4-15
2.1.3 Somatic embryogenesis A 4-15
2.2 Variation and selection A 4-16
2.2.1 In vitro selection and variation A 4-16
2.2.2 Polyploidisation A 4-17
2.2.3 Anther culture A 4-18
2.2.4 Microspore culture A 4-18
2.3 Recombination A 4-19
2.3.1 In vitro pollination A 4-19
2.3.2 Ovary and embryo culture A 4-19
2.3.3 Protoplast fusion A 4-20

3 Molecular techniques A 4-21
3.1 'Natural' gene transfer A 4-22
3.1.1 Transformation using Agrobacterium tumefaciens A 4-22
3.1.2 Viral vectors A 4-23
3.2 Direct DNA transfer A 4-23
3.2.1 PEG-mediated transformation, electroporation and micro-injection A 4-23
3.2.2 Particle gun transformation A 4-24
3.3 Anti-sense technology A 4-24
3.4 DNA marker selection A 4-25

1 Plant breeding techniques at crop and plant level

Genetic variation is the source of all biodiversity in the world. New genes, gene combinations and gene frequencies result from mutations, from selective advantages of one genotype compared to other genotypes and from recombination by pollination and fertilisation. Plant breeders have homed in on these basic principles, creating (crossing) and finding genetic variation, followed by the selection of the most desirable genotypes. Characteristics for optimal processing and consumption are selected, but also traits which enable the genotype to be maintained (stability) and propagated.

The three main factors in plant breeding are:


creating new varieties and combinations;

maintenance and propagation.

1.1 Selection
The purpose of selection is to point out the most promising genotypes in a group of plants. However, variation in plants is determined to a large extent by environmental factors such as growing conditions, therefore statistics and indirect selection methods are often used to account for the influence of these factors.

1.1.1 Mass selection

How it works
Mass selection is based on the ability to recognise desirable or undesired characters in plants of a population. What appear to be the best plants are maintained in bulk (positive mass selection), while plants with too little of the desired characteristic are eliminated (negative mass selection). The selected plants flowersimultaneously and are harvested in bulk for the next generation, ie. next selection. In positive mass selection, only a small part of the initial population passes to the next round, while in negative mass selection the greatest part passes through.

Mass selection does not require special facilities or techniques. Since selection is based on phenotype alone, the technique is particularly effective for those characteristics which are barely influenced by environmental factors and which are not inherited as dominant or recessive traits but rather as complementary traits.

The technique is mostly applied during the early stages of a breeding programme, when there is not enough representative plant or seed stock for repeated testing, when seed stock needs to be improved at short notice, or when little capital is available for a breeding programme for a certain crop. However, breeding based only on mass selection is a lengthy process. This is the selection method that most resembles natural selection.

1.1.2 Pedigree selection

How it works
In pedigree selection, a plant is maintained or eliminated on the basis of the performance of its progeny. It often follows mass selection in a population. Each selected plant is then harvested separately. In the next generation, is the seeds from each are planted as distinct lines. Lines will only be maintained following a favourable assessment of performance throughout the line. The best plants of the line will be identified and their seed will again be harvested separately for the next round.

By splitting into lines a general view of the parent plant's genotype is gained. Negative recessive characteristics, for example, often crop up with this method. The selection is made on the basis of general impressions (phenotype) and heredity of the desired characteristics (genotype).

Pedigree selection requires a sound infrastructure and administration, since seed is not harvested and processed in bulk but in small quantities, per line and per plant.

Pedigree selection enables selection for phenotype and progeny and thus provides new cultivars sooner than mass selection. The method is particularly appropriate for self-pollinators. In cross-pollinators, some cross-pollination between selected plants andother plants in the line or between selected and non-selected plants is bound to occur in the trial plot. This can be a problem when the most important characteristics (for example, grain formation or crop quality) cannot be assessed until after flowering.

1.1.3 Site-determined selection

How it works
This method is applied especially in organic breeding programmes for cereals such as wheat, barley, rye and spelt. However, in theory it is also suitable for use with other crops. The goal of this selection method is to select varieties, usually population varieties, which are optimally adapted to specific regional conditions. The parent lines are also selected with the final growing site (region) of the crop in mind. Crosses are carried out at a central location, where the F1 is also sown. The F2 and F3 plants are assessed and selected at different locations by means of pedigree selection. This is a combination of natural selection and artificial selection. In other words, environmental factors determine which characteristics are expressed and which are not and this may influence the breeder's selection of promising phenotypes.

Suitable trial plots must be available in each region where varieties are to be developed.

This method ensures that natural and artificial (human) selection work towards the same goal, namely, a variety with optimal ecological stability (Kunz & Karutz, 1991). The method differs from conventional methods because the seed stock is assessed and selected at different locations at a relatively early stage in the breeding programme. The goal is to find optimal types for specific locations. Conventional breeders usually select plants that perform acceptably at all locations.

1.1.4 Change in surroundings

How it works
In this method, subsequent generations are grown in vastly different environments in alternate years; for example, on sandy soil one year and on clayey soil the next; or in different climates, dry-wet, mountain-valley, coastal-inland etc. Thus, types with adaptability to extreme conditions are selected within a variety or population. Subsequent selection and variety development of these types is performed at one location.

Breeders must have trial plots in vastly different environments. The seed stock/variety should still have considerable variation in characteristics which are expressed in the different surroundings. This method depends on a breeder's knowledge of the plant in all its facets, otherwise specific types will not be identified. As with any selection method, the breeding goal should be clearly defined.

This method appears to work best with cross-pollinators, since plants which are fertilised in the open maintain considerable genetic variation. This method makes the most of the potential qualities of a variety. The method was first developed for breeding biodynamic rye. However, if only existing varieties are used in the selection process, the different types may differ in appearance but will have a similar genetic make-up. In the long term, genetic erosion of the cultivar could creep in unnoticed. To prevent this, new crosses will occasionally need to be made.

1.1.5 Change in sowing time

How it works
A change in the sowing time is usually applied in organic breeding to improve quality characteristics in cross-pollinating winter wheat. Sowing in the autumn (September-October) appears to improve the nutritional quality (baking quality), while sowing towards the beginning of winter (late November-early December) improves seed quality. The morphology and quality of the plant are influenced by the sowing time in subsequent years.

No special preconditions apply.

Sowing time may be varied per crop and per goal. It is vital that a breeder knows his plant and his goal from years of experience. For example, in winter rye, three years of late-spring sowing results in a tall and limp plant with small ears containing many small grains. By contrast, late-autumn sowing results in a much stockier plant with few leaves (Muggli et al., 1990). Alternating late-spring sowing with late-autumn sowing yields a long-straw plant with plenty of vigour. According to Spiess (1996), winter wheat varieties are best maintained by always sowing in late-autumn.

1.1.6 Ear-bed method

How it works
This organic breeding method was developed especially for cereals, but it may also be used for other crops. An ear bed is a seedbed in which the grains from one ear are sown in the same sequence in which they were arranged in the ear. Thus the plants in the seedbed (ear-bed) reflect the quality of the original ear. The plants' vigour, nutritional quality, resistance, production etc. will differ depending on whether the plants are grown from grains from the middle, upper or lower part of the ear. To induce additional variation, the plants originating from the upper part of the ear are selected in one year and plants from the lower part in the next year. The ear-bed method may also be used in the early stages of variety development. In this case, plants are not selected from the different ends of the ear but from the middle (Schmidt, 1994).

For plants with a different form of inflorescence, such as podded crops, seeds from the pod are planted in order, also taking into account the order of the pods. Thus the bed starts with the seeds from the lowest pod and ends with the seeds from the highest pod.

This method demands an accurate trial plot design and administration.

This method utilises the variation between plants as well as intra-plant variation.

1.1.7 Indirect selection

How it works
Some characteristics are not easily identified. Sometimes this is because of great variability in the crop, requiring repeated testing with large numbers of plants, a procedure which is only feasible with a limited number of selections during the advanced stages of a breeding programme. In all other cases, testing is expensive, complex or bound to restrictions (such as testing for a notifiable disease). In these cases, the solution may be to use indirect selection markers. Markers are identified more easily and are strongly linked to the desired characteristic. Thus selecting for the marker characteristic will also improve the plant with respect to the desirable characteristic. In some cases, linkage is due simply to the proximity of genes on the chromosome; these genes have little else in common but almost always occur simultaneously, for example flower colour and resistance to disease. In other cases, it is not clear whether the linkage has agenetical or a physiological basis.

Biochemical and molecular techniques are often used in indirect selection. Isozymes have a slightly different composition from common enzymes and can be shown as a banded pattern on a gel. A certain band links up with a desirable characteristic. For example, the gliadin and glutenin bands in wheat are linked to baking quality. Isozyme analyses in seedlings can largely replace extensive testing for disease and baking trials.

Before indirect selection can be carried out, a linkage must be found between a marker and the desired characteristic through genetic research. Only then can it be incorporated in a breeding programme. When a marker is an easily identified qualitative characteristic of the plant, no special facilities are needed.

Indirect selection can contribute considerably to breeding programmes.

1.1.8 Test crosses

How it works
In breeding, it is not enough merely to select the best cultivars, but parents with good crossing ability must also be available, so that a number of important characteristics are combined and passed on to progeny. This is especially crucial for asexually propagated plants and hybrids.

Promising parent genotypes are crossed with a number of other known genotypes. Progeny is grown separately and assessed for the desired characteristics. The measurements are used to determine General Combining Ability (GCA) and Specific Combining Ability (SCA), and these are needed if effective crosses are to be made. The most extreme type of test cross is the diallel, in which a range of genotypes is intercrossed in all possible combinations, both as female and as male parent. Diallelcrossing schemes are not much used in the later stages of breeding practice, but are limited more to the research phase.

A lot of time goes into test crosses and extensive testing facilities are needed. New cultivars are rarely to be expected directly after a cross, but these crosses do provide valuable information about a genotype's potential in future crosses with other promising genotypes. More than anything, test crosses are an investment in the future.

Building up sound knowledge about the value of parent lines in a cross-breeding programme is typically a long-term investment, especially with self-pollinating and cross-pollinating plants. Often, only the larger breeding programmes will be able to afford such an investment in knowledge, on top of the required working capital. However, the investment pays back in the end, with a solid cross-breeding programme based on in-depth knowledge of the seed stock.

1.2 Inducing variation

One is only able to select if there is variation in a plant population. Variation can be stimulated by various means. The methods below induce variation at crop and plant level.

1.2.1 Combination breeding

How it works
Most plant breeders start their selection programme by crossing two genotypes of the same species, for example two accepted cultivars. Generally, these are easy crosses and since the two genotypes in the cross have already proven their worth as cultivars, interesting new genotypes are to be expected in the next generation. Depending on the plant, whether it is asexually propagating, self-pollinating or cross-pollinating, various generations are propagated and selected after the cross. Usually one of the selection methods described above is used.

No special facilities are required for combination breeding. It is crucial that pollination and fertilisation are effected only by the intended paternal plant. The cross may be carried out by hand, for example after emasculation, or by pollinator insects caged in with the plants.

The recombination of characteristics through cross-breeding is a generally accepted technique and requires no special skills.

1.2.2 Species crosses

How it works
Sometimes a desired characteristic is not found in existing, regionally adapted cultivars but it may be found in cultivars from other climate conditions, in wild relatives of the cultivar, or in other, similar species. Common examples of such characteristics are disease resistance and the more extreme forms of stresstolerance. Breeders often come up against the limitations of crossing when wild plants or other species are used. Basically, there is a limit to the ability of one plant to fertilise a plant of another species. There are several biological explanations for this which we will not go into here.

Breeders have developed a number of methods to get around these cross limitations. Below, we describe methods which involve the plant as a whole and which are carried out in the greenhouse. These methods usually have mediocre success rates.

1.2.3 Bridge crosses

How it works
In effect, a bridge cross is an interim cross which is made if a cultivar cannot be crossed directly with its wild relative. The wild plant is first crossed with another (wild) species, its progeny is selected for the desired characteristic and these are then crossed with the cultivar. Many combinations are possible with bridging crosses, all of which should lead to the incorporation of the desired characteristic in the cultivar.

Breeders must have access to a large collection of wild relatives and must have considerable knowledge of how they respond to intercrossing. The desired characteristic must be relatively easy to select.

Making bridge crosses is a time-consuming process. Once the characteristic has been incorporated in the cultivar, a number of backcrosses are needed to rogue out as many undesirable 'wild characters' as possible.

1.2.4 Temperature treatment of the style

How it works
Some barriers in the style may be broken down by exposing the plant or the style to higher temperatures for a certain period of time. After the temperature treatment, the pollen may succeed in working its way down the style to the ovary.

A special temperature-controlled space is needed for temperature treatment.

A previously impossible cross is made possible without a majorintervention.

1.2.5 Cut style

How it works
This method is used when the pollen does germinate on the stigma, but cannot work its way far enough into the style, so that the ovary is never fertilised. Sometimes fertilisation can be achieved by cutting (part of) the style of the female plant and mixing the pollen with stigma juice exuding from the cut. The pollen tube now has only a short distance to grow, increasing the chance of fertilisation.

No special facilities are needed for this method. It can be carried out in the greenhouse. This method is suitable for only some plants, such as ornamentals with a long style. It is likely that seed is formed only if the two species are closely related.

This method is particularly suitable for crossing short-style flowers with long-style flowers. Pollen is helped partly down the long road to the ovary. Subsequent fertilisation and seed formation proceed naturally.

1.2.6 Grafting on the style

How it works
This method can be used when pollen fails to germinate on the stigma of the female plant. Germination is a positive response reaction. When pollen and stigma fail to 'recognise' each other, pollen will not germinate. Grafting was developed for plants with long styles (such as lilies). Pollen is first applied to the stigma of a plant of the same species, so that it germinates effectively and the pollen grows some way into the style. The style is then cut off, just below the point reached by the pollen germ tubes, and grafted on a cut style of the female plant. When the two styles have joined, the pollen germ tubes continue down the style of the female plant and fertilise the ovary.

These are the same as for the cut style method. For practical reasons, the method is only feasible for plants with a fairly long style.

This is a more drastic method than cutting the style, since theresponse mechanism is bypassed. However, fertilisation and seed formation then proceed naturally.

1.2.7 Mentor pollen technique

How it works
The mentor pollen technique can be used to solve recognition and growth problems. To this end, pollen from the chosen male parent is mixed with pollen from the same species as the maternal plant. The latter's pollen has been partially inactivated by irradiation; it still germinates but does not fertilise. The mentor pollen germ tubes 'guide' the pollen germ tubes from the male parent to the ovary, which is fertilised by pollen from the male parent. Mentor pollen which has not been irradiated, and which has thus retained its vigour, may also be used but is less efficient.

Pollen radiation must be outsourced. Otherwise, no special facilities are needed.

For this technique, the species' own pollen is first inactivated by means of irradiation. Fertilisation and seed formation proceed naturally. When untreated mentor pollen is used, many more crosses are needed to achieve the desired cross. The numerous seedlings resulting from such crosses must all be

assessed, a daunting task. Selection with DNA markers might be an option here (see section 3.4).

1.2.8 Repeated backcrossing

How it works
Species crosses may yield progeny that have so many exotic or wild characteristics as to make direct selection for valuable genotypes impossible. Sometimes, the genetic distance between cross parents results in instability in progeny, for example through polyploidy, aneuploidy, abortion of seeds, fertility problems. In these cases, repeated backcrossing with a well-adapted cross-parent may eliminate some of the wild and/or exotic traits, finally producing a genotype that is highly similar to the well-adapted parent but with the additional desired characteristic. A breeder must usually make three or four backcrosses before being able to switch to pedigree selection. Selection is focused on the presence of the desired characteristic and minimal 'wild' characteristics.

Repeated backcrossing is time-consuming. Seed formation andfertility problems may yet crop up in later backcrosses.

Repeated backcrossing is most suitable when a desirable new characteristic must be incorporated in an existing variety. The new character should not alter the genotype's suitability for cultivation. Molecular markers can help speed up the process of selecting a new characteristic while at the same time testing for the absence of other genetic donor material (marker assisted backcrossing, see also 1.7 and 3.4 indirect selection). Gene constructs can also be transferred into an existing variety (see 3.1 and 3.2). However, traditional backcrossing is by no means outdated, having already proven its worth for a great many varieties which have gained a high tolerance to biotic and abiotic stress.

1.2.9 Hybrid breeding

How it works
Selecting new cultivars is particularly difficult in cross-pollinators as a lot of intra-varietal variation is inevitable. Inbreeding becomes a risk when propagation between selection cycles is limited to the selected lines only. Moreover, both cross-pollinators and self-pollinators show greater vigour when there are two different alleles for a gene (heterozygous), rather than if the plant is homozygous for that allele. A plant that is heterozygous for a single-factor pair is a genetic hybrid and its increase in vigour is called hybrid vigour, or heterosis. Hybrids are produced by crossing two homozygous (inbred) lines. This breeding method is particularly popular for cross-pollinators. Parent lines must be artificially inbred to obtain sufficiently homozygous lines. This can reduce the vigour of the parents. Breeders select for the most vigorous inbred lines so as to produce the most vigorous F1.

Since the F1 progeny are the new cultivar, large quantities of F1 seed must be produced. Emasculating and pollinating by hand has only limited feasibility. In monoecious plants, where the male and female inflorescences are distinct, it pays to manually emasculate the male inflorescence. An example of this is maize. Dioecious plants and self-incompatibility (S-alleles, which prevent selfing, such as in cabbage), too, allow for the production of F1 hybrid seed. In many crops, naturally occurring cytoplasmic male sterility (cms) is incorporated in the maternal line by crossing or protoplast fusion (see 2.3.3). So-called male-fertile B-restorer lines which have the same genetic make-up as the maternal line are used to maintain the female line. In seed-forming plants, these restorer genes are also found in paternal lines, so that the resulting F1 can produce fertile pollen and form seed.

As yet, few hybrids have been produced from self-pollinators, although hybrid vigour has been shown. Developing and maintaining inbred lines is not a major problem in self-pollinating crops, since inbreeding is a natural consequence of selfing. It is seed production which forms the crux of the problem here. Deactivating the pollen of maternal plants is no easy matter; besides, the structure of the flower is rarely conducive to large-scale pollination by paternal plants with the aid of wind and/or insects. In some crops, such as tomatoes, female lines are emasculated and subsequently pollinated manually. Each pollination produces large quantities of seed which sells for a good price, so that this labour pays off.

Hybrid breeding requires a sound system for large-scale production of F1 seed. The usual system is based on cytoplasmic male sterility and the presence of restorer genes. In addition, during seed production, plants must be protected against pollination by stray pollen to produce pure hybrid seed. Next to manual emasculation and male-sterility, use is made of gametocides and gene constructs to regulate the production of fertile or sterile pollen.

Hybrid vigour usually yields a uniform and highly productive plant. However, it is not usually possible to produce progeny since the characteristics of the hybrid, an F1, will be split up. In effect, this is the hybrids' natural protection against propagation and it is one of the reasons why they are so popular with breeders, especially in countries without breeders' rights. As a result, the breeding of seed-forming varieties has often been neglected.

In cross-pollinators, the inbred parent lines of hybrids have reduced vigour. Some lines are maintained through tissue culture techniques. If cms is used, maternal lines - and in seed forming plants, also the F1 generation - can be maintained with a restorer line. Thus, some specialisation in reproduction has occurred at the plant level. When no restorer line is available, the maternal line is literally the end of the line as far as breeding goes.

In theory, breeders keep their own inbred lines for themselves. They are not available for other breeders or other breeding programmes. This contributes to the genetic erosion of the cultivar.

The gametocides and gene constructs mentioned above are secrets of industry and/or are protected by patents, which restricts their use to a small number of licenced breeding firms.

1.2.10 Parthenocarpy and cucumber

How it works
Parthenocarpy is the ability of certain plants to form fruit without prior fertilisation. This quality is strongly familial and thus fairly easily incorporated in cross-breeding. Parthenocarpy is used in cucumber.

Cucumber was originally a monoecious plant with male and female inflorescences. However, some varieties had hermaphrodite inflorescences or a combination of hermaphrodite and male inflorescences. In short, flowers of all types occur in cucumber.

Today, cucumber breeding is aimed at the production of true female hybrids, combining the female character with the character for parthenocarpy. Breeding has taken this route since fertilisation leads to small, round cucumbers full of seeds. These are not the cucumbers we want to eat. Naturally, seed formation remains necessary for the production of hybrid seed and maintaining parent lines. When the young plants of true maternal lines are treated with silver nitrate or silver thiosulfate, they grow male flowers. The resulting hybrid is female, so that there is no risk of ending up with seed balls which reduce production.

The tendency to form only female flowers is strongly determined by the surroundings. A climate-controlled greenhouse is necessary for this purpose.

Highly toxic silver nitrate is applied to force a female line to produce male flowers for seed production.

Female lines must also be treated in other ways to produce male inflorescence. The plants are subjected to stress, for example by placing them too close to each other, by denying them food and/or water or by growing them in poor light. Favourable results (number of male flowers) are by no means certain. The method might technically form a feasible alternative for organic farming, however it is not an economically sound solution.

Alternatively, a hermaphrodite line could be used as the paternal line, but at the moment this is still a theoretical solution since existing hermaphrodite lines are impossible material. Because they form seed, their shape cannot be assessed and the lines have many undesirable characteristics. These could be incorporated in a hybrid if a hermaphrodite line were to be used as paternal line. Many more years of breeding might possibly result in this approach becoming a feasible alternative to the silver nitrate method.

1.2.11 Synthetic varieties

How it works
The term synthetic variety is often used for cross-pollinating varieties for which components have been individually selected for good combining ability with pollination in the open. In other words, the components have a good general combining ability, or GCA. Synthetic varieties are a good solution for crops with a strong inbreeding depression.

The term may also refer to so-called self-pollinating 'multi-lines'. Repeated backcrossing may yield purebred lines of varietal quality which differ from each other in only one gene, for example a resistance gene. By combining these lines in one variety, a mix of characteristics is obtained which meets breeders' uniformity requirements while at the same time the variation in resistance genes creates a barrier for pathogens.

Before synthetic varieties can be bred, research must be carried out on the GCA of the components through series of pair crosses and testing sample combinations. Such activities require considerable investments of time and effort. Multilines are obtained through repeated backcrossing with the basic parent line of the multiline variety.

Synthetic varieties often have heterosis. The progeny of a synthetic variety may have a poorer quality, although this effect is not as marked as with hybrids. Breeding according to the multiline concept requires long-term investment. The basic variety is selected first and only then can component lines be obtained through repeated backcrossing followed by selfing. The risk of multilines is that the cultivar might be outdated by the time it is introduced on the market. Moreover, it would appear that each distinct component line needs to be registered separately for breeders' rights; merely registering the multiline offers inadequate protection. This, too, is a costly affair.

1.2.12 Mutation breeding

How it works
New alleles might be considered desirable, in addition to existing alleles for a gene. New alleles often arise through mutations, or changes in DNA. Mutations may occur spontaneously, for example by exposure to sunlight, cold or heat, radicals, and sub-optimal conditions during cell division. Mutations are an interesting and valuable source of genetic variation and may be provoked by exposing plants to irradiation or chemical mutagens. Most mutations are recessive and as such are not directly visible in the treated plant. However, after selfing, the mutated genotypes may be identified in progeny. Some examples of genetic variationthrough induced mutations are: flower colour, trueness to type and the elimination of undesirable secondary metabolites.

There is a large selection of tried and tested mutagens on the market. Suppliers of irradiation are also fairly easy to find. Mutagens and irradiation should be applied by skilled personnel and safety guidelines should be complied with.

The plants are usually safe after treatment, so that the selection methods described above can be carried out.

Induced mutations occur at random; it should be kept in mind that only a handful of mutations contribute positively to the expression of the desired characteristic compared to the original seed stock.

1.3 Propagation and maintenance

How it works
When a new variety has been selected out of the new genetic variation, it must be propagated and maintained as a purebred. In self-pollinating and cross-pollinating plants, the selected genotypes flower in isolation. Positive mass selection ensures that the elite parent lines are of high quality and meet all the requirements of a variety. Diseases in parent plants are not usually passed on to seed. Most seed can be stored for long periods of time.

In asexually propagated plants, such as beets, bulbs or grafted plants, extra care must be taken to avoid deterioration in the health of the seed or plant stock. Stock often does not keep well, thus these lines are maintained through a continuous process of propagation. However, the advantage of vegetative propagation is that the progeny has exactly the same genotype as the parent, whether heterozygous or homozygous. Generative propagation through seed demands a much greater degree of homozygosity.

Seed can also be propagated asexually, however. This phenomenon is called apomixis and it occurs in cultivars (oranges, tropical plants) and wild plants alike. Seed is formed from the tissues of the maternal plant, often the embryo sac or the ovule. There are obligate apomicts, whose seed contains apomictic embryos, and facultative apomicts, whose seed forms both sexual and apomictic embryos.

Apomixis has caught breeders' interest because it combines the advantages of seed propagation - health and keeping quality - to the identical reproduction of the maternal genotype. However, we still have a limited understanding of this process. The fact that apomixis is a deviation of sexual reproduction makes a study ofthe genetic principles of this occurrence difficult. Moreover, apomixis occurs only in polyploids. It would appear that apomixis is a dominant trait expressed by only a small number of genes.

The propagation and maintenance of cultivars requires great accuracy and attention to health aspects. Isolated propagation plots and positive mass selection of elite parent plants are often essential. This applies even more strongly to vegetatively propagated plants.

Apomictic propagation appears a promising method to maintain varieties. Unfortunately the technique is potentially suitable for only a limited number of species.

Maintaining varieties requires considerable knowledge of plants and technical expertise. The use of tissue culture techniques is almost inevitable to ensure the health of stock material in the propagation of vegetative plants. Apomixis is an interesting alternative, but, at the moment, it is suitable for only a limited number of species. Moreover, the advantages gained in propagating and maintaining a variety through apomixis are almost entirely undone by the problems that such varieties typically give in cross-breeding (which is sometimes made impossible), recombination and selection.

2 Tissue culture techniques

Tissue culture techniques take advantage of the fact that each individual cell in a plant harbours the potential to develop into a complete, new plant. This is called regeneration. The process must be carried out entirely in sterile conditions. In other words, plant stock (from the greenhouse or the open field) must undergo a superficial disinfection with alcohol and chlorine before being placed on a sterile culture medium in sterile (glass) culture pots. All operations are carried out in a laminar flow cabinet. The main ingredients of culture media are water, minerals, sugars and vitamins. A coagulant such as agar or gelrite may be added to obtain a solid substrate. The process of regeneration may be orchestrated by adding synthetic plant hormones (auxins and cytokinines) which regulate cell division and differentiation. Plants are cultured in a climate-controlled room; light intensity, length of day and daytime/nighttime temperature are controlled.

2.1 Propagation

In vitro (in glass) propagation is an asexual propagation method. It is an efficient means of obtaining a homogeneous population of a successful line or maintaining and/or propagating varieties in disease-free conditions. The method is expensive, but economically viable as it produces considerable plant material in a short space of time. The breeding cycle may be reduced, so that new varieties can be introduced on the market sooner. Various in vitro propagation methods exist, which we discuss below.

2.1.1 In vitro propagation

How it works
Depending on the plant species, some part of a plant - most commonly, a piece of stem with an axillary bud, or part of a leaf or bulb scale - is cultured in vitro. These sections of plant grow out to form shoots, which can in turn be cut and propagated. This may be repeated several times, so that the number of plants - and thus the amount of labour involved - increases with time. When enough plants are made, these are grown until their roots develop, they are then hardened off and transferred to normal greenhouse and/or field conditions.

Micropropagation is a collective term for various in vitro methods. A distinction can be made between axillary and adventitious propagation. As the name suggests, axillary buds form the basis of axillary propagation. All axillary buds contain meristems. By cutting a shoot into sections, each with a leaf and an axillary bud, the axillary buds are activated and produceshoots (not unlike the process of cuttings). Since the parent plant's own meristems are used, there is little chance that plants will deviate from the parent plant. Adventitious propagation takes advantage of the fact that each cell in a plant is totipotent, that is, has the potential to develop into a complete new plant (regeneration). In this method, sections without axillary buds are used, for example a piece of a leaf or a disc section of stem. Plant hormones in the culture medium may stimulate these so-called explantates to produce adventitious shoots which are propagated according to the axillary method. With this method, there is a greater risk of undesirable deviations in progeny because the tissues are exposed to artificial stimuli (plant hormones) for a longer period of time.

A laboratory setting is vital, since all operations must be carried out in sterile conditions. Basic laboratory equipment includes a laminar flow cabinet, an autoclave and an incubator. Most propagation techniques have not been patented, so that the techniques may be used freely.

Micropropagation in vitro is used in a breeding programme once a breeder is certain that a new genotype (a future variety) will meet with success on the market. Through in vitro propagation, large quantities of stock to be sold on the market are obtained in a short space of time. Depending on the type of plant, the breeding cycle can be cut back by as much as a few years.

In vitro propagation is usually faster than propagation in the greenhouse or the open field. It is carried out in sterile conditions, so that the propagated material is guaranteed free from disease. It is an asexual propagation method, meaning that the progeny of one maternal plant are genetically identical, ie. they are clones of each other.

Micropropagation is also used to create transgenic varieties. Regeneration must follow the transformation and this is done fastest through in vitro propagation. Some species are nearly always propagated in vitro. Often, these species contract diseases too easily when propagated in the greenhouse or in the field, or else their rate of propagation in normal conditions is too slow.

In vitro propagation is a fairly labour-intensive method. Much of this work is carried out in countries where labour is cheap (Poland, eastern Europe, India). Recently, however, some of this work is coming back to the Netherlands, since better assurances of quality can be made here. This has given a new impulse to another propagation method using somatic embryos.

Breeders have had some success in using this method to maintain inbred lines for hybrid seed production. The most common example of this are lines for leek hybrids.

2.1.2 Meristem culture

How it works
Meristem culture provides virus-free plant material. Viruses can be especially persistent in asexually propagated plants, because they are passed on to the following year's stock (cauliflower, garlic, potatoes etc.). Viruses proliferate and spread in the plant, but they never quite catch up with the rapidly dividing cells in the plant meristem. These cells divide so fast that the virus has no time to infect the newly formed cells. In meristem culture, uninfected cells from the meristem are isolated and cultured on substrate. The resulting plants are screened for the virus with the ELISA method. Uninfected plants are then propagated in the manner described above.

The same preconditions apply as for micropropagation. Meristems are next to impossible to see with the naked eye, so visual magnification is a must. Meristem culture must be carried out by trained, experienced staff.

This method yields virus-free plants when virus-free stock of a certain variety is no longer available. The method helps prolong a variety's 'working life', in other words it stimulates sustainable use.

2.1.3 Somatic embryogenesis

How it works
Somatic embryogenesis is an adventitious propagation method. Somatic embryos are polar constructs with a shoot pole and a root pole. Somatic embryos are created from the somatic cells of explantates and may or may not first have undergone a callus phase. Callus may be described as a proliferation of tissue consisting of non-differentiated cells. Callus, which is commonly propagated in liquid medium, is placed on top of a vortex, so that the tissue falls apart and forms a cell suspension. Plant hormones are added to the explantate cells, the callus or the cell suspension to stimulate the cells' development into somatic embryos. Often, these embryos are in turn capable of forming secondary embryos, thus leading to a continuous propagation process.

The advantages of this method compared to other micropropagation methods are the far greater propagation factor and much reduced labour intensiveness. The method lends itself well to large-scale or even automated production. However, as with all adventitiouspropagation methods, there is a greater risk of deviant plants.

There are two options once an appropriatenumber of somatic embryos has been obtained. First, somatic embryos may be placed on a substrate and grow to mature plants, be hardened off and transferred to the open field or the greenhouse. Second, somatic embryos can be dried and coated, so that they resemble seed. Like seed, they can then be sown in the field or the greenhouse.

If carried out on a large scale, this method requires more special equipment than the other micro-propagation methods. Naturally, the large scale application of this method also requires a greater capital investment. On the other hand, the method is less labour intensive and (parts of) the method lend themselves well to automation, so that somatic embryos may be produced profitably in Europe as well. The labour required and the scale of production strongly depend on the plant species. What may be an optimal propagation method for one species might be less efficient for another.

Propagation using somatic embryos is likely to produce a number of deviant plants. Propagation factors are high and propagation is usually fast. Because deviant somatic embryos can also produce secondary somatic embryos, the number of deviants can increasequickly. Of course selection must be part of any micro-propagation method, but it is particularly crucial for propagation with somatic embryos.

2.2 Variation and selection

Plant breeding is all about variation, combination and selection. Some combinations of characteristics are rare (such as tolerance to aridity and high crop yield). A variety with such a combination of characteristics would normally require the testing of countless plants on trial plots. In vitro selection reduces the size of the trial plot to a number of rows of petri dishes in incubators or climate controlled rooms. New concepts can be tested for their viability and for the desired characteristics using biochemical models. Plants which appear promising according to these models are selected. Selections may be made for tolerance to heavy metals, to salt, to extreme temperatures, to variation in metabolic production, and so on. Using a tried and tested product line, a new variety is created in only one generation. Below, we discuss techniques for inducing variation or identifying existing variation, after which we will present techniques by which variants can be combined.

2.2.1 In vitro selection and variation

How it works
Somaclonal variation is the occurrence of mutation in tissue culture, resulting in deviant plants (see the sections on adventitious propagation). It is highly undesirable in micro-propagation and every possible measure is taken to prevent it. However, in other situations, somaclonal variation is stimulated, for example by applying high concentrations of plant hormones. In vitro selection uses somaclonal variation to induce and select plants with the desired characteristics. For example, to obtain plants with a high tolerance to salt, selection occurs in a saline environment. Plants which do not have the desired characteristic will die.

Selection for resistance to diseases and pests can also be carried out in vitro. The pathogen is applied to the plants in controlled conditions. Sometimes, resistance may be identified by testing with a filtrate in which the pathogen was cultured, or with a toxin produced by the pathogen.

The same preconditions apply as for micropropagation. Commercial success depends on a high correlation between the expression of the characteristics in vitro and in the greenhouse or in the open field. Plants that have been obtained using in vitro selection are therefore always tested in field circumstances.

Variation is induced, that is, mutations are provoked deliberately, followed by selection. Inducing mutations is a random and undirected process. In contrast, selection is directed at a desired characteristic and because the goal of selection is clear, it can be carried out quickly and efficiently among a large population within the confines of the laboratory. Plants resulting from this selection process may be marketed directly as a new variety (after in vitro propagation) or they may be used as a cross-parent in further breeding.

In vitro selection is narrowly focused on one desirable characteristic. It is quite possible that plants with considerable potential in other respects are discarded. Thus, this method is only suited for specific characteristic-oriented breeding programmes, rather than programmes aimed at overall improvement.

2.2.2 Polyploidisation

How it works
A cell is polyploid if it has at least twice the normal number ofchromosomes. Polyploidy can be induced using chemicals such as colchicine. Normally, during cell division the spindle in the cell ensures that each half of the set of chromosomes is reorganised into two new cells during cell division. Colchicine dissolves the spindle, so that the chromosomes remain in one cell. When small meristems or seed are treated with colchicine, they are likely to grow into a completely doubled plant. This method often follows anther or microspore culture, but it can also be applied after species crosses - by different techniques - to restore fertility in progeny.

In addition to all the general laboratory facilities, a fume cupboard is needed to work with thehighly toxic chemical.

Doubling the number of chromosomes is often necessary to restore the fertility of plants obtained through species crosses or haploidisation. However, polyploidy is also applied for another reason entirely. Plants with a double set of chromosomes are generally larger or more robust than plants with the normal number of chromosomes. They may therefore be more profitable or have a higher ornamental value. Colchicine is normally applied to larger parts of a plant (seed, meristems, even the whole plant) which increases the chance of chimeras. Depending on the species or plant, chimeras may impede performance and have to be eliminated. Through regeneration and micropropagation a completely doubled plant can be obtained from a small piece of doubled plant tissue. However, this adds another step to the breeding cycle, thus extending its duration.

2.2.3 Anther culture

How it works
Anther culture is applied to obtain homozygous plants in a short space of time. In this method, immature anthers from, for example, an F2 (differentiating population) are grown on substrate. Normally, a grain of pollen divides into a vegetative and a generative cell (asymmetric division) and the generative cell later divides again during sexual reproduction. An embryo develops from the fusion of one of the generative cells with an ovary; the other generative cell fuses with the central cell in the embryo sac thus producing endosperm. In the culture of anthers, plant hormones are applied to prevent the asymmetric division and to force a symmetric division which produces two identical cells. Further division and differentiation of these cells will result in embryos. The embryos are haploid (of each chromosome only one is present in the cell) and will grow to form haploid plants.However, haploid plants usually lack vigour and what is more, they are usually sterile. An interim step is therefore included to double the chromosomes using colchicine in the manner described above. The most common method is to apply a solution containing 0.5% colchicine applied to the plant at the roots or on the axillary buds. In the case of the latter, the existing leaves die and new axillary shoots will have the double set of chromosomes. These shoots are homozygous. This doubling may occur spontaneously in vitro.

Roughly the same preconditions apply as for micropropagation. Anther culture protocols do not yet exist for all plant types. In some cases, considerable research is necessary to develop sound protocols. To make matters more difficult, different varieties of a crop often require a different culture protocol.

Anther culture is usually carried out at the beginning of a breeding programme. Once a new population of homozygous plants is available, new varieties are developed by mass selection and pedigree selection. A feature of homozygous plants is that recessive traits are also expressed, ie. are visible in the phenotype. This simplifies the breeder's task of selection.

2.2.4 Microspore culture

How it works
Culture of microspores is actually a further elaboration of anther culture. Rather than using whole anthers, only the microspores are cultured. The method is generally regarded as more complex than anther culture particularly with respect to keeping the microspores alive and stimulating symmetric division. The advantage of this culture method is that there is no question of plants identical to the maternal plant growing from the anther wall, as is possible in anther culture.

The culture of microspores is far more complex than that of anthers. Much more specialised equipment is necessary, such as a centrifuge to isolate the microspores and a fluorescence microscope. A lot of research also goes into protocol development.

From a genetic point of view, plants obtained through the culture of anthers or microspores are identical to inbred lines. The variation in microspores is brought to expression as for anther culture (see above).

2.3 Recombination

Variation, the basis of all breeding, can be induced in a plant or plant tissue by means of, for example, somaclonal variation or mutations. The most common method, however, is to create new combinations by crossing different genotypes. Sometimes the desired combination of characteristics cannot be obtained through crossing; in that case, we can speak of a cross barrier. Breeders often meet cross barriers in making species crosses. Some tissue culture techniques avoid or break through these barriers. They constitute a technical solution which enables 'traditional' crossing, even though the 'natural' species authenticity has been violated.

2.3.1 In vitro pollination

How it works
In species crosses, for example between a cultivar and its wild relative, fertilisation sometimes fails, either because the pollen won't germinate or because it can't get to the ovary. Sometimes pollination and fertilisation can be achieved in vitro. Ovaries or ovules are removed from the plants and fertilised in vitro with pollen. Cross barriers between the stigma and the ovary are thus avoided.

The cut style and grafted style methods are also used in vitro. However, rather than being carried out on the whole plant they are carried out on isolated ovules cultured in a test tube or petri dish.

The cross barrier must be located in the stigma or the style. Collected pollen should be sterile (ie. disease-free) and germinate well in vitro.

More genotype and species combinations can be made than was previously possible. Species authenticity can be violated. The method has only a meagre chance of success, although this does depend strongly on the species used and the combinations made. A slightly different protocol may be needed for each combination. The technique is usually applied at the beginning of a breeding programme which is aimed specifically at introducing a certain character, for example a resistance gene from a wild plant, in a cultivar. Undesired characteristics are eliminated through repeated backcrossing.

2.3.2 Ovary and embryo culture

How it works
Despite successful in vitro fertilisation, an embryo may not develop or be rejected prematurely. There are several reasons for this. The endosperm might be underdeveloped or the embryo and the endosperm may somehow be mismatched. The purpose of ovary and embryo culture is to transfer the embryo to an artificial nutritive substrate at an early stage, so that it need no longer depend on the endosperm. In ovary (plaque) culture, the entire ovary or slices of the ovary (plaques) are placed on the substrate. The ovules swell and at a certain point are removed from the ovaries and cultured independently. By this time, the ovules have in fact become seeds on a substrate, where they may germinate.

In embryo culture, embryos are placed on a nutritive substrate to germinate. After fertilisation, ovary culture can be applied sooner than can embryo culture.

For the success of ovary and embryo culture, fertilisation must have taken place in vitro or on the plant. The first steps of embryogenesis must have been made.

More combinations can be achieved than with standard crosses, since the first steps of embryogenesis have been made on the plant and rejection is avoided by growing the embryo on substrate. As with in vitro pollination, new combinations are obtained through sexual (natural) reproduction, contrary to protoplast fusion, described below.

2.3.3 Protoplast fusion

How it works
Protoplasts are cells without a cell wall. These are obtained by treating fragments of leaves with cell wall dissolving enzymes.The protoplasts then grow a cell wall again and divide, resulting in a callus from which plants can regenerate. Protoplasts of different plant species can be fused with chemical or electric stimuli (somatic hybridisation). The resulting tetraploid has all the characteristics of both parent species. During regeneration the chromosomes of both parents may be mixed, so that many combinations are produced.

In such a fusion, the organelles of both plants are combined (in crossing, only the maternal organelles are passed on to progeny.) Cell organelles (chloroplasts and mitochondria) contain unique genetic information which contribute to the new phenotype. Incombining this information, new combinations are made. An example is male sterility, which is determined by organelles. This familial character can be influenced by cybridisation. The plant which is allowed to supply organelles is used for the production of cytoplasts. Cytoplasts are protoplasts without a nucleus, or cells with fragmented chromosomes (through gamma irradiation). These cells cannot regenerate but are fused with the protoplast of another plant. After fusion, selection is directed at the cytoplasmic characteristic. This method is also used to insert chromosome fragments in a new genetic base.

The method requires more specialised equipment and knowledge than micropropagation. Culturing protoplasts is a tricky matter. Much research must precede it.

This method produces new combinations of hereditary material, allowing unusual combinations of species. Somatic cells are fused and the products of this fusion have double ploidy. In potato breeding this feature is incorporated by breeding at the diploid level and then raising the plants to the tetraploid level by protoplast fusion (cultivated potatoes are tetraploid, wild potatoes are diploid).

3 Molecular techniques

With molecular breeding techniques, a single gene from a plant, animal or micro-organism can be integrated in the plant genome. A single characteristic can thus be incorporated in the plant.

Several steps are necessary to produce a transgenic plant. Most of these steps are common, regardless of the exact transformation method. The steps are named below. The various transformation methods will be discussed further on in this chapter.

How it works
Transgenic plants are obtained by following these steps:

1. Collect, isolate and clone the DNA fragment (from plant, animal or micro-organism) with the genetic code for the desired characteristic.

2. Determine the gene's DNA sequence, necessary for a patent application.

3. Construct a DNA chimera with a promotor, the gene and a terminator to guarantee sufficient expression in the plant if this is not possible with the plant's own means.

4. Clone in a vector suited for transfer to the plant and for the selection of transgenic plants (eg. antibiotic resistance, herbicide resistance).

5. Carry out transformation using one of:

* 'natural genetic transfer'

Agrobacterium tumefaciens, Agrobacterium rhizogenes, viral vectors

* direct DNA transfer

electroporation, PEG mediated, particle bombardment

6. Regenerate and select transgenic plants using tissue culture methods and antibiotics/herbicides (that is why resistance needs to be incorporated in step 4).

7. Check transgenic plants for ploidy level and phenotype using molecular methods and practical selection criteria.

8. Introduce the transgenic line into the cross-breeding programme.

Modern laboratories are necessary for research into and application of these techniques. Setting up and running such laboratories is expensive. In addition, licences are usually required for the commercial application of these techniques since many of them are patented. Patents may apply to any or all of the following steps:

1. a gene's DNA sequence;

2. the use of a transformation method;

3. the use of an Agrobacterium vector;

4. the use of a plant promotor;

5. the use of an antibiotic/herbicide resistance gene.

To produce a transgenic plant, these molecular techniques are applied with tissue culture techniques such as protoplast culture and/or regeneration. This means that suitable techniques must be available for the plant concerned.

Contrary to the situation with crosses and some tissue culture methods in which genetic material is combined, DNA fragments are incorporated randomly in a plant. This will always affect the structure of a plant's DNA and may result in a visible mutation. Such mutations will be expressed after selfing, in the S1 generation (homozygous transgenic plants). The chance of such mutations varies between 5 and 50 per cent, depending on species, transformation method, tissue culture method and the DNA construct used. Many more mutations without any visible effect may have occurred. It is assumed that further selection of transgenic plants in a crossing programme will eliminate all undesirable mutations.

Since DNA is a universal code for genetic character, DNA from any organism can be used to incorporate new qualities in a plant, such as bacterial herbicide and insecticide resistance. As molecular techniques become more advanced, plants are increasingly treated as bioreactors. For example, the processing industry is able to set specific requirements for plants, which are achieved by changing the plant's genome so that it will produce foreign substances.

3.1 'Natural' gene transfer

Vectors are micro organisms which can transfer fragments of their own DNA plasmid into a plant's DNA. The term 'natural' gene transfer refers to the use of a naturally occurring vector in genetic transformation. Agrobacterium tumefaciens and Agrobacterium rhizogenes are the bacteria most commonly used. Vectors may also be viruses. Molecular scientists home in on this ability and replace the DNA fragment usually transferred by the vector with the DNA fragment containing the desired gene. The vectors do the rest of the work.

3.1.1 Transformation using Agrobacterium tumefaciens

How it works
Agrobacterium tumefaciens causes root collar gall. It is the most common vector for genetically manipulating dicotyledonous plants. The scientist uses a shuttle vector (DNA plasmid) suitable for both laboratory strains of Escherichia coli and Agrobacterium. DNA is manipulated (cut and joined) in E. coli and the new geneconstruct is conjugated in Agrobacterium. The gene construct and the accompanying selection marker (antibiotic or herbicide resistance) is transferred to a plant cell, preferably in young leaf tissue (cotyls or hypocotyls) which is bathed in an Agrobacterium solution a day after being cut. It is left there for a number of days, during which Agrobacterium must infect the plant. Subsequently, various antibiotics are added: penicillin derivates to kill Agrobacterium (otherwise it will outgrow the plant material) and kanamycin or a herbicide to select the transformed and now growing cells from the wild tissue at the cut. The first shoots appear after 4 to 8 weeks and are transferred to a medium which stimulates root growth. After a further 4 weeks growth, the plant can be moved to the greenhouse. At this stage, plants with deviating ploidy levels, visible mutations and sterile plants are eliminated. The progeny of the transgenic plants are assessed; lines bearing a mutation are eliminated as are lines where the selection marker (antibiotic/herbicide resitance) does not have the desired 3:1 ratio. Only then are plants with one transgene selected using DNA analysis methods and the genotype tested. Since this method requires relatively brief tissue culture times (2-4 months), there is only minimal somaclonal variation.

This method is applicable only with plants that allow themselves to be transformed by Agrobacterium. In practice, the method is rarely used for monocotyledenous plants (such as cereals, grasses and bulbs), since these plants resist infection with Agrobacterium. Recently, rice has been successfully transformed by Agrobacterium. The method requires a laboratory that meets stringent safety standards (containment) to prevent transformed organisms escaping. Greenhouses where transgenic plants are planted must also meet these containment standards.

Transgenic plants resulting from Agrobacterium transformation do not contain bacterial DNA. Next to the desired gene, selectable resistance genes for antibiotics and herbicides are also incorporated. These characteristics are used during the selection procedure to distinguish transformed plants from non-transformed plants. However, they are not usually removed after having served in this process and thus form part of the final product. One question is what consequences this may have for the development of antibiotic resistant bacteria through horizontal gene transfer.

Scientists first believed they could incorporate a single characteristic into a variety in this way. However, in practice, not all variety characteristics are retained with this method. Thus transgenic plants are often used as cross parents in a breeding programme.

3.1.2 Viral vectors

How it works
The technique of making transgenic plants with plant viruses has never been seriously developed, since most plant viruses are RNA viruses which do not incorporate well in the genome. The most common DNA viruses, on the other hand, cannot carry enough foreign DNA and are therefore unsuitable for gene transfer. Recently, viral vectors have come back into the spotlight because of their ability to only temporarily express the virus (transience). This ability is interesting for two applications: 1) inducing anti-sense effects (see anti-sense technology) by introducing a recombinant virus with the plant's gene. The recombinant virus spreads through the plant and defuses the endogen (the plant's own gene). 2) in the production of pharmaceutical peptides and proteins. Tobacco plants are used for this purpose. Just before they mature they are infected with the recombinant virus. Within two weeks, the plants produce large quantities of viral biomass, which are filtered and purified to obtain the pharmaceutical ingredient.

Almost the same facilities are needed as for Agrobacterium.

The use of recombinant viruses is limited to infected plants. It cannot pass to seed. Thus the plant can be manipulated, but regulations pertaining to transgenic plants do not apply. The question is how recombinant viruses will behave and whether horizontal gene transfer problems will arise.

3.2 Direct DNA transfer

This method does not use natural vectors such as bacteria or viruses. DNA is inserted in the cell by chemical or ballistic means.

3.2.1 PEG-mediated transformation, electroporation and micro-injection

How it works
PEG (polyethyleneglycol) mediated transformation and electroporation are done on protoplasts (see tissue culture techniques). Protoplasts are brought together in a DNA plasmid solution with the desired gene and the resistance marker together. PEG and electrical shock both make the plasma membrane of the protoplast porous so that the plasmid can enter the protoplast.After that, the DNA must enter the nucleus of the cell and become incorporated with the plant's DNA. The success of this method depends strongly on luck. The use of large numbers of protoplasts (107) may ensure a reasonable number of transformed cells at the end. However, many transformants are eliminated during the long tissue culture process because they have a deviating ploidy level or an undesirable mutation. Often, several DNA fragments are incorporated in the genome as a head-to-tail chain, resulting in highly unstable phenotypes (with respect to the introduced gene). In micro-injection, DNA is injected into the cell nucleus, such as an ovary or (somatic) embryos or other cell clusters.

These methods can only be applied when a protocol is available to culture protoplasts. More equipment is needed than with transformation using Agrobacterium.

The method is still used exclusively on monocotyledonous plants, since these plants do not transform well with the Agrobacterium method. In using bacterial plasmids for vectors, there is a possibility that they will incorporate in the genome as a whole. This yields a plant with DNA that might prove useful for prokaryote micro-organisms. Theoretically, this raises the chance of horizontal gene transfer.

3.2.2 Particle gun transformation

How it works
Originally, the method was developed to 'shoot' DNA into the chloroplasts of single-cell algae. As with other direct DNA transfer techniques, plasmid DNA vectors are used. Tungsten or gold particles are coated with DNA and these particles are shot into the cell with a 'gun' using gunpowder or helium. The DNA detaches from the particles and may enter the nucleus and become incorporated there. The method can be used with cells, cell clusters, leaf tissue, (somatic) embryos, pollen grains or even whole plants. It served as a breakthrough in obtaining transgenic monocotyledonous plants

All the usual facilities and a particle gun are needed.

In addition to the consequences of using plasmid DNA, discussed above, other consequences must be considered. The method uses high concentrations of DNA. Complex integration patterns of foreign DNA are thus formed, resulting in high numbers of useless plants.Many-celled tissue often forms the target, increasing the likelihood of chimeras, plants where one part is transformed and contains the desired gene and the other part is not transformed. Completely transformed plants may as yet be obtained by regenerating those parts of the plant that are transformed, however this is definitely a detour.

3.3 Anti-sense technology

How it works
Anti-sense technology was developed to influence the expression of a plant's own genes. A gene is forcibly expressed in the anti-sense (reverse) direction in a transgenic plant. The result of anti-sense RNA production is that homologous genes are eliminated in these transgenic plants, resulting in a null-phenotype. An example is the Flavr-Savr tomato in which maturation is delayed. For some genes, the anti-sense effect is less stable than was initially assumed.

The same preconditions apply for the anti-sense method as for the methods described above.

The anti-sense effect is not always stable and the null-phenotype may be absent or only partially present in progeny, with large interplant variation. The fact that identical phenomena (anti-sense effects) occur with 'normal' gene constructs reveals that more interaction takes place between the transgene and a plant's own genes than we are as yet able to observe.

3.4 DNA marker selection

How it works
The variation in the DNA sequence around a gene or genes determining the desired characteristic can be revealed by isolating the DNA and making parts of the DNA base sequence visible by gel electrophoresis. DNA is fragmented by restriction enzymes and placed on an electrically charged gel. The DNA fragments move through the gel due to the magnetic field created by the charge. The length of a DNA fragment determines its speed through the gel. The resulting banded pattern of slow-moving and fast-moving DNA fragments is used to find those bands which are linked to the desired characteristic. Some of the markers used are RFLP, AFLP, RAPD and micro satellites.

In identifying molecular markers for indirect selection, it often occurs that several bands are linked to one characteristic. Thisshows that several genes are involved in the expression of a characteristic. These results help increase breeders' understanding of a desired characteristic and also lead to more efficient selection, especially when genes respond differently to, for example, environmental factors.

A prerequisite for indirect selection is that a marker is found for a desired characteristic. This requires additional genetic research before a breeding programme can get underway.

Good, reliable facilities are required when working with isozymes and molecular markers. A considerable investment in instruments and staff is needed. However, many organisations are currently developing simple, robust tests (dot-blot, dipstick, similar to blood glucose and pregnancy tests) which would be a feasible option for smaller businesses.

If genetic research shows that a marker is strongly linked to a certain characteristic which is difficult to observe directly, indirect selection may help advance a breeding programme. Patent applications have been made for molecular marker techniques especially. It is currently unclear whether licences will be granted to use the techniques if patents are given. Nor will the use of the simple, robust tests named above be free, but they will probably be made available to all interested and potential users.

Next section of "Sustainable Organic Plant Breeding."

Back to List of Contents

To Ifgene Home Page