2.Marker Assisted Selection and the Future of Agricultural Biotechnology
NOTE: The articles this is responding to are available at
1.Subject: Re GMW: Plant breeding without the madness
From: Doug Gurian-Sherman
Although very encouraging, there are a few important cautions about MAB (or MAS is it is more commonly called) that need to be recognized. It is also important to understand the difference between marker assisted selection (MAS, or MAB) and conventional breeding, which was included in the following piece along with MAB.
MAS uses relatively new (about 20 year old) genetic technology that 'maps' the genes of a plant by identifying unique genetic 'fingerprints' (the markers of MAS)throughout the genome of the crop (the entire complement of DNA in the plant's cells). These fingerprints can be associated with particular genes or traits that are useful, such as drought tolerance. The advantage of MAS is that the many genetically complex traits in crops (such as drought tolerance or yield), are more likely to be successfully utilized than with conventional breeding. Also, more simple genetic traits can often be bred more quickly with MAS than conventional methods.
Conventional types of breeding, including making hybrids, select for the crop's visible (or otherwise measurable) characteristics to improve the crop, rather than using genetic markers. This can be very difficult for traits that are complex genetically (controlled by many genes), respond to environmental conditions (e.g. weather changes) or that are very rare in the crop. Despite these limitations, however, conventional breeding has been very successful, and will continue to be. Conventional breeding can also be done with a crop's wild relatives - this does not require MAS. So many potentially useful traits from wild species that can breed with a crop are not limited to the use of MAS. MAS, however, can often facilitate breeding with crop wild relatives or allow breeding for traits that conventional breeding is not good at using (like complex traits).
What are the potential concerns about MAS? First, in some (perhaps many) countries, patents may be given for MAS. Intellectual property protection is of course one of the big problems with genetic engineering technology, in that it gives the company undue control over seeds (such as not allowing farmers to save seed). In fact, the big genetic engineering companies are using MAS frequently now, and not just as an alternative to GE, but to facilitate moving GE traits into many new varieties of a crop, or combining MAS traits with GE. Second, similar to genetic engineering, this technology takes some sophisticated (and expensive) equipment and training to carry out, so it cannot be done by low-tech breeders or farmer-breeders. It also requires good genetic maps of the crops involved, something not available yet for some crops - especially many important to developing countries.
Finally, contrary to the statement [in the article posted - see NOTE above], MAS does have risks for food safety. Breeding crops with their wild relatives has led to instances of introducing toxic substances into crops. It must be done with caution. Plants produce many substances to protect themselves from insects, diseases, and herbivores, and some of these are toxic to us as well (some are also useful as medicines, etc.). Crops tend to have lost many of these, or they are removed by cooking or other processing. But wild crop relatives may produce others. This is also one reason why indigenous knowledge of wild plants can be so incredibly important. However, many scientists feel that GE goes considerably beyond these risks by introducing entirely new genes to food that could never have gotten there otherwise. By contrast, the risks from wild relatives comprises a much narrower span of possibilities, and are more likely to be similar to risks already recognized in the crop.
The main point is that although MAS can be very useful, and can be held up as an alternative to GE in many cases, we should be aware of its limitations as well, and not push it as a panacea.
Doug Gurian-Sherman, Ph.D.
Food and Environment
Union of Concerned Scientists
2.Is Genetic Engineering Obsolete?
BY DOUG GURIAN-SHERMAN
Marker Assisted Selection and the Future of Agricultural Biotechnology [originally published in the magazine Gene Watch in 2006 http://www.gene-watch.org/pages/genewatch.html ]
Increasing attention has been recently focused on a technology for breeding plants and animals that employs rapidly accumulating data from genomic studies, but does not involve genetic engineering (GE). This has led to the hope that this sophisticated hybrid between molecular genetics and traditional breeding, called Marker Assisted Selection (MAS), may meet some of the unmet promises of genetic engineering, and perhaps even replace it. This hope is strengthened by the biotech industry's continuing failure to develop transgenic crops with substantial benefits, the increasing use of MAS by industry leaders such as Monsanto and Dupont, and the continuing rejection of gene-spliced crops by citizens in Europe, Japan, and elsewhere.
With so much at stake, it is a good time to ask: What is MAS? Will it succeed in those areas where genetic engineering has failed? If so, will it eliminate concerns about unintended consequences of our use of biotechnology, or is it merely another technological fix that looks promising at first glance, but that reveals serious drawbacks upon closer examination?
HOW MAS WORKS
MAS uses information from the DNA sequence of an organism, such as rice or tomatoes, to improve the speed and efficiency of breeding. Certain unique DNA sequences from each crop or domestic animal can be used to track specific useful traits during the breeding process. The use of these genetic sequences, called markers, to breed for useful traits is in contrast to traditional breeding, which uses the expressed trait itself (or phenotype) to select desired progeny. MAS uses the same breeding 'material' as traditional breeding; the diversity of qualities exhibited by livestock, crops, and the related wild species that can breed with them.
By improving the availability and breadth of this diversity, MAS may expand the range of traits that can be bred into crops and animals without using genes from other species (transgenes), as is the case with GE. By improving efficiency, MAS can also greatly accelerate the breeding process.
The markers used in MAS are arranged along an organism's chromosomes, along with the genes associated with useful traits. Markers close to a gene of interest, such as one associated with disease resistance or higher yield, can be used to track that trait during breeding, because the same progeny acquire both the marker and the trait from their parents.
Although a trait could be tracked by using the sequence of the associated genes, the breeder usually does not know the genesÃ¢ exact locations or sequences. It is often easier to find genetic markers that are close to the trait rather than to find the exact position of the gene.
Tracking is needed because only a fraction of the progeny from breeding experiments have the desired trait; so they must be identified and separated from the other progeny. Because some traits can be difficult to identify during breeding, it is sometimes easier to track the genetic marker rather than the trait.
For example, a particular variety of disease-resistant potatoes may be bred to provide resistance traits for other varieties of potatoes. However, determining which of the progeny of a breeding experiment carry the resistance trait may be difficult, especially if the disease is sporadic. A marker associated with the resistance, however, could be used to determine if the trait is present. In other words, the marker is always present and does not depend on external conditions, such as the presence of a pathogen.
Perhaps the biggest promise of MAS involves the breeding of quantitative trait loci, or QTL. QTL are chromosomal regions that have a partial effect on a trait, rather than an all-or-nothing effect, and are usually controlled by several genes at various locations along the chromosomes. Most agronomic traits, such as yield size, drought resistance, and many types of pest resistance, are controlled by QTL. In the disease resistance example above, each of four QTL may control between 15% and 35% of the potato's overall resistance to a certain disease. To achieve a major effect, several of these QTL must be brought together in a single variety of crop or animal.
Because QTL have minor effects individually, their expression can be hard to track using traditional trait-based breeding. QTL also tend to be sensitive to environmental effects and interactions with other genes in the organism, and their effects may therefore be masked or altered under some circumstances.
Finally, because different QTL are in different places in the genome, they are each independently distributed to different progeny during breeding. Therefore, the probability that all or most of the desired QTL will appear in a single given offspring can be small.
In theory, by using the markers associated with each QTL, breeding of these traits might be accomplished without relying on the vagaries of selecting for the traits. Furthermore, the breeder does not have to wait for the crop or animal to reach the stage development when the quality is expressed; for example, at harvest in the case of grain yield. A small sample of tissue from a seedling can be used to assay for the genetic markers. This can save a lot of time, space, crop material, and expense, especially for organisms such as timber or orchard tree species that are large and have very long developmental periods.
Cost and availability should also be considered when evaluating MAS. The identification of numerous markers needed to carry out MAS requires technology similar to that used for GE. To have reasonable assurance of success for many types of MAS, at least several hundred markers are required, spread over the genome of the target organism. Among other things, MAS requires DNA synthesizers to make the probes needed to locate the markers in parents and progeny, and the molecular technology to perform the detection and mapping of markers on the genome. Traditional breeding facilities are also needed to perform MAS. Therefore, the cost of MAS can substantially exceed that of traditional breeding, and requires advanced technologies that are not always widely available.
MAS IN PRACTICE
Although the most anticipated use of MAS has been for the breeding of QTL, this application has been the least successful. Breeding for QTL that are heavily affected by the environment or other genes of an organism is equally difficult using either MAS or traditional breeding methods. To be successful, the progeny in QTL breeding experiments must be frequently checked in several different environments to ensure expected performance. The need to continue to use traditional phenotypic trait selection in conjunction with MAS raises the cost, and thereby often makes it impractical. Where environmental factors or the other genes of the progeny least affect desired traits, traditional breeding may be more economical due to the higher costs of MAS technology. Therefore, unless technological barriers are overcome, MAS may be effective mainly for QTL that fall between these extremes.
So far, MAS is showing most promise in accelerating the breeding of traits that have strong phenotypes that are not easily observable, such as with some forms of disease resistance. In addition, MAS can facilitate the combination of several of such traits, a process that would be impractical or very time consuming using traditional methods. In these cases, MAS can often cut the breeding time in half. This can be of great importance, because traditional breeding may otherwise require a decade or more.
MAS can also facilitate the improvement of traits and the genetic diversity of crops through breeding with sexually compatible wild relatives. For example, tomatoes and potatoes have several wild relatives in South America that can provide many useful traits not already found in the domesticated crops. However, one drawback to using wild relatives is that wild genes associated with the desired trait are often linked to unacceptable properties as well, such as reduced yields. This 'linkage drag' can often be addressed by crossing of progeny with the crop over several generations, while selecting for the desired trait. This eliminates most of the unwanted genes contributed by the wild plants. Undesirable genes that are physically close to the ones associated with the desired trait are more difficult to eliminate, because the vast majority of the progeny will contain both the desired trait and the deleterious one. Use of MAS, especially if markers are very close to the trait, can accelerate the elimination of linkage drag.
So far, MAS has had only limited success. Several years ago, an informal survey of the MAS literature revealed limited production of new crop or livestock varieties.1 Repeating a similar survey for this article continues to reveal few new varieties. Instead, most papers report the development of markers that may be used in MAS, theoretical assessments of how to best use MAS, or whether it can be as cost-effective as other approaches to breeding.
MAS VS. GENETIC ENGINEERING
One reason that MAS has attracted attention is as a possible replacement for GE. However, there are several reasons why this is unlikely, in addition to the limitations discussed above.
Perhaps the most obvious reason is that, even though there is a lot of untapped potential in crops and their wild relatives, the kinds of traits that can be naturally found among these organisms are limited. For example, GE may be able to provide genes that exceed the range of drought tolerance available in any wheat variety or wild relative. However, as noted in an early seminal paper on MAS: 'Single genes with major effect that are amenable to genetic engineering usually have deleterious pleiotropic [multiple, usually unexpected] effects.'2
Therefore, even where exotic genes are found that initially appear to exceed the traits of the crop, deleterious unintended properties may prevent commercialization. Perhaps this is part of the reason why, despite hundreds of different genes in over 11,000 field-trial approvals regulated by USDA, only a few herbicide resistant, insect resistant, and virus resistant GE crops have been successfully commercialized.3 But because the pipeline for GE research is subsidized by academia, GE companies may continue to produce GE crops, in part because they can cherry-pick the most promising applications. That pipeline may continue to find occasional commercial successes that could not be accomplished by MAS.
In other cases, MAS or other advanced breeding techniques may provide a reasonable alternative to genetic engineering. For example, so-called 'allergy free' soybeans were recently developed by genetic engineering, but shortly afterward, non-GE allergy free soybeans were also developed.4 The latter offer the same benefit as the former, although neither have been commercialized.
Similarly, although the alteration of oil content in several crops has been achieved experimentally using genetic engineering, the recently touted low-linolenic-acid soybeans, which have reduced trans fat, were developed without genetic engineering.5 However, the developer of one of these types of soybean, Monsanto, has crossed it with soybean varieties containing genetically engineered herbicide resistance. So now farmers who want modified-oil soybeans from Monsanto also get GE glyphosate resistance in the bargain.
This example illustrates the coexistence of GE and MAS, and how MAS can be used in conjunction with genetic engineering. Genetic engineering would be practically useless without traditional breeding, which has always developed varieties of crops adapted to different environments that transgenes are then bred into. Genetic engineering developed despite the successes of traditional breeding, and it is likely to continue even if MAS proves successful.
MAS is even used to accelerate the movement of transgenes, such as ones from Bacillus thuringiensis (Bt), which convey pest resistance, just as it is used for naturally occurring traits. In this way, MAS is actually facilitating genetic engineering. Finally, much of the equipment and infrastructure needed for GE can also be used for MAS, and vice versa, so there is no need for a company to choose one over the other due to technological limitations.
POTENTIAL DRAWBACKS OF MAS
Other than the absence of foreign genes, MAS and GE have much in common. Both require capital-intensive infrastructure to perform genetic assays and to analyze the resulting data, as well as specialized, highly trained staff. This may slant its use toward large institutions, such as corporations and research universities, and make it less accessible to smaller breeders and developing countries.
Another concern is that, in the U.S. and elsewhere, the methods for performing MAS and the resulting organisms may be patented, as is the case with GE. The patenting of seeds and genes may impede development of new varieties, increase corporate control of the seed supply, and reduce competition.
The inability of farmers to save seed, due to corporate ownership under patent law, fundamentally changes the nature of ownership in farming. The ability to control the use of MAS and genetic engineering afforded by patenting, as well as their capital-intensive and centralized attributes, may also partly explain the interest of large corporations and their academic partners in these technologies. By contrast, agroecological methods that contribute to productivity, biodiversity, and sustainability, are less susceptible to corporate capture. Therefore, the development of technologies like GE or MAS may come at the expense of research on more decentralized technologies.
Some argue that MAS patenting problems will be minimized as researchers pursue 'open sourcing,' similar to what is occurring in computer software (See Open Source Biotechnology, GeneWatch 18-1), with anticipated reductions in the high cost of MAS techniques facilitating a move to less restrictive control over intellectual property. However, although some inroads may be expected on both fronts, it is unlikely that the analogy between MAS and computer software should be taken very far. MAS will never be used by more than a small fraction of the population compared to software innovations. The restriction of MAS to a specialized community that is largely permeated by corporate and academic participants that favors intellectual property protection, probably will limit both the amount of 'open source' MAS and reduction in the cost of its infrastructure.
Although the seminal paper on MAS was published in 1990, and some MAS applications go back even further, progress has been slow. The sequencing of genomes over the past 10 years should help, as well as studies examining the best and most efficient ways to apply the technology. But in some cases, the inherent limitations of genetics may prevent certain applications from being achieved. In other words, many features of agriculture, including crop quality, may be determined only to a limited degree by genes, and will always primarily depend on other environmental factors, such as cultivation practices. It is important to remember that of the substantial advances in crop yields over the past 100 years or so, half or more have been due to factors other than breeding.6 In fact, the emphasis on genetic engineering over agroecological approaches in the past 20 years represents a substantial imbalance in agricultural research and development.
Although MAS may contribute substantially to crop improvement without resorting to genetic engineering, it is unlikely to replace GE. MAS can and should be pursued in the context of agroecology as another example of our ability to advance agriculture without the use of genetic engineering. But as with more traditional types of breeding before it, which have contributed greatly to crop improvement, corporations will continue to attempt to use MAS in conjunction with GE to further their continuing monopolization of the seed industry.
Doug Gurian-Sherman is Senior Scientist at the Center for Food Safety.
1. Dekkers JCM and Hospital F, 2002, The use of molecular genetics in the improvement of agricultural populations, Nature Rev. Genet. 3: 22-32
2. Lande R and Thompson R, 1990, Efficiency of marker assisted selection in the improvement of quantitative traits, Genetics 124: 743-756
3. Field Test Releases in the US, Information Systems for Biotechnology, http://www.isb.vt.edu/cfdocs/fieldtests1.cfm
4. Paul J, 'Researchers Find Allergy-Free Soybean,' 29 Sep 2004, http://www.phillyburbs.com/pb-dyn/news/247-09292004-374273.html , Associated Press
5. CHS To Process Monsanto's VistiveÃº Low-Linolenic Soybeans; Soybean Variety Provides a Trans Fat Solution to the Food Industry, from a press release, Aug. 9, 2005, www.monsanto.com
6. Lande R, op. cit.; and Fehr WR, (ed.), 1984, Genetic Contributions to Yield Gains of Five Major Crop Plants, Special Pub. No. 7, Crop Science Society of America, Madison, WI