Several years ago, the Cassava Biotechnology Network (CBN) decided that availability of efficient reproducible transformation and regeneration protocols would be crucial to develop improved cassava varieties. Recently, four research groups have been able to develop mature transgenic cassava plants.

In order to develop a reliable system for plant genetic modification three issues have to be addressed. Firstly, a system that allows the insertion of target genes into the plant cell has to be established. Secondly, selection markers must be used to distinguish transformed cells from non­transformed cells. Thirdly, a culture system is needed in which the individual modified cells are efficiently converted into plants.
The major problem for genetic engineering of cassava was to combine the available regeneration methods with the existing transformation systems. Four different research groups have overcome this barrier and developed transgenic cassava plants.

Centro Internacional de Agricultura Tropical
The Centro Internacional de Agricultura Tropical (CIAT), Colombia, was the first laboratory that claimed success. CIAT researchers obtained cassava embryos through somatic embryogenesis. These embryos were cultivated together with Agrobacterium tumefaciens to achieve gene transfer. Somatic embryogenesis is a method by which 'non­reproductive' plant cells are stimulated to form embryos identical to the mother plants. Subsequently these embryos are developed into mature plants. The embryos originate from multiple cells located in the internal plant tissue. Agrobacterium tumefaciens is a bacteria that is able to transfer genes to plant cells. By cultivating the embryos with the bacteria, genes from the bacteria are transferred to the cassava embryos.
The developed technology, however, has several flaws. Firstly, CIAT uses a natural type of Agrobacterium tumefaciens, which transfers not only the desired genes, but also its disease­causing genes. Consequently, the regenerated plants are not suitable for cultivation. Secondly, the system is highly genotype specific. Thirdly, since the embryo is formed out of several cells from the deeper tissues of the plant, the method is not efficient. Due to their position these cells are a difficult target for transformation. In addition, since the embryos originate from multiple cells, most embryos are chimeric, consisting of transformed and non­transformed tissue.
The problems of this specific method of transformation and regeneration of cassava were known to CIAT. Nevertheless, the rationale behind this choice was that other methods had failed to transform cassava. Therefore, to establish whether it was at all possible to transform cassava, this rather drastic method was applied.

Swiss Federal Institute of Technology
A method developed at the Swiss Federal Institute of Technology (ETH­Zentrum), Switzerland, circumvents parts of these problems. They use a special type of Agrobacterium tumefaciens, which does not transfer disease causing genes to the plant cells.
Besides, they regenerate the cells by organogenesis. This is a regeneration method by which formed embryos originate from cells at, or very close to, the cut surface. Therefore, these cells are an easier target for transformation. However, also in this method genotype specificity remains a problem.

Wageningen
The methods that are used by Wageningen Agricultural University (WAU), the Netherlands, and the International Laboratory for Tropical Agricultural Biotechnology (ILTAB), of the Scripps Research Institute, USA, are less prone to problems such as genotype specificity and chimeric plants. Both use an embryogenic suspension culture (ESC) method developed by the University of Bath, UK. In this system clusters of embryogenic cells are suspended in liquid medium. These clusters can be regenerated into mature plants relatively easy. The fact that the cells are loosely organized in a suspension improves accessibility for insertion of new genes. At the same time selection procedures are easier for transformed cells than for embryos. This makes the ESC method very suitable for genetic transformation protocols.
WAU and ILTAB combine the ESC method with particle bombardment (i.e. DNA coated microparticles are shot into plant cells) to transform the cassava plant cells. A difference between their work is the use of different marker genes to distinguish transformed tissue from non­transformed tissue.
Although this method is quite efficient and less genotype specific, the system is very labour intensive and time consuming. It also requires expensive equipment and advanced laboratories. This makes the technique less suitable for transfer to laboratories in developing countries.

Coordination of research priorities
Although the developed methods are promising, much work still has to be done to arrive at a standardized ready­to­use protocol for the transformation and regeneration of cassava. The different methods are still in an exploratory phase and their specific problems have to be addressed.
The original breakthrough of producing transgenic cassava plants was achieved with marker genes only. Now that the initial protocols are available, the research groups have started with the transfer of functional genes that encode for specific, desired traits in the cassava plant. In this context ILTAB is working on virus resistance with specific emphasis on the African Cassava Mosaic Virus (ACMV). ETH is active on a broader basis. They are currently working on virus resistance, insect resistance, manipulation of physiological characteristics and root quality aspects.
The choice by the individual laboratories on which specific traits to work appears to be based on a combination of several aspects. Firstly, the laboratories individually consulted with local cassava breeders and scientists at CIAT and the International Institute of Tropical Agriculture (IITA). These research centres of the CGIAR hold the mandate to stimulate cassava research in South­America and Africa respectively.
Secondly, the priority list by the Cassava Biotechnology Network (CBN) has been taken into account. When the Dutch Directorate General International Cooperation (DGIS) started funding the CBN coordination in 1992, DGIS stimulated the network to integrate farmers' perspectives in its priorities. This initially caused divergence between the priorities of CBN and the CGIAR centres. However, debate has brought them closer again.
A third related point is the fact that the donors funding cassava research have their influence on the research direction. They all have their own goals and expectations of the research that is funded by them. Part of the research budgets of the groups involved in cassava transformation and regeneration comes from national research budgets. These public funds are assigned with an emphasis on the scientific value of the funded research. Other funds are derived from foreign and international development organizations. These organizations are more interested in the applicability of the results in their own specific priority areas. WAU, for example, is funded by DGIS while the Rockefeller Foundation, USA, provides funding for ILTAB and supports several other research projects in cassava. At the same time the Overseas Development Agency (ODA), UK, and the Swiss Development Cooperation are involved in the funding of cassava research. These donors use different criteria in funding research projects. However, the bottom line is that the projects have to be directed towards food security for small­scale cassava farmers and users. Therefore, the traits that the different laboratories work on fit in this framework.
Fourthly, interest of the private industry for cassava research is increasing. For example, part of the research at the WAU is funded by AVEBE, a major Dutch producer of starch products. WAU's research on transgenesis for starch characteristics is of interest for both CBN and AVEBE. CBN puts a high priority on enhanced starch content and quality for small­scale cassava farmers. AVEBE is interested in cassava starch as a possible raw material for the company's products.

Cooperation between research groups
Cooperation and information exchange between the different cassava research groups has resulted in the fast development of transformation and regeneration protocols. This shared interest was an important rationale behind CBN, which provided funds to facilitate international cooperation. In recent years the cooperation has resulted in several joint publications on important achievements in cassava research.
Now that the initial protocols for transformation and regeneration are available, the CBN coordinated research groups are currently discussing how future cooperation can be structured. The aim is to develop further and standardize the available protocols to a point where transgenic methods could be easily used across a range of agriculturally important genotypes.
There are different ways to achieve this goal. One option is for a small group of advanced laboratories to join forces and to deliver a generally applicable technique in the shortest possible time. For example, ILTAB and WAU could work together with one or two other laboratories to elaborate the particle bombardment/ESC method. Although this option would probably mean fast delivery, it is very focused on a particular technique.
Participation of international research institutes located in cassava growing countries, such as CIAT and IITA, could mean input from the users side. Inclusion of the users' point of view may deliver cassava transformation and regeneration protocols which are better attuned to local situations and wishes.
At the same time, researchers from the national institutes in developing countries could already be trained in the new techniques. However, the involvement of a number of research groups, including less advanced laboratories, will take more time and money before a ready­to­use transformation and regeneration method for cassava will be available. Inadequate funding for cassava research should not be overlooked.
The cooperation of a broad range of research groups also means that a broad range of institutional interests have to be taken into account. For example, intellectual property rights (IPR) need to be harmonized.

Intellectual property rights
The fact that cassava is an important subsistence crop in many developing countries makes the issue of free access to cassava science and technology an important issue. Different strategies can be followed in order to secure accessibility. CBN policy on IPR is based on a preference for publication and early disclosure rather than on IPR protection. However, CBN recognizes that IPR can offer protection against misappropriation.
Research groups such as ETH and the University of Bath have chosen not to seek patents. They want to ensure availability and free exchange of information for developing countries. Patents, however, have been requested for the techniques that are developed at WAU and ILTAB. They believe that patents can also offer protection or can be used as a medium of exchange of patents between patentees.
In a funding agreement, DGIS has a contract with WAU which states that DGIS will be actively involved in IPR issues. In developing countries, separate cases of commercial use of the specific patented technology will have to be submitted to DGIS and will be tested by DGIS policy. In this way, DGIS wants to secure free access to cassava science and technology for developing countries.
The IPR policy of the Rockefeller Foundation states that materials and technology resulting from supported research will be available at zero royalty for use in developing countries. They also want to be involved in any agreement that will be in conflict with this obligation. The Rockefeller Foundation is funding research at ILTAB, and ILTAB has a policy of free access by cassava growing developing countries.
The requests for patenting specific cassava transformation and regeneration protocols are currently pending. Whether these patents will be granted remains to be seen. Different aspects of the technology used are already protected by previous patents, such as the technique of particle bombardment, owned by Dupont. Another patent concerning kanamycin resistance, which is used for selection, is held by Monsanto. Marketing of transgenic cassava plants that are developed using these different techniques will necessarily involve negotiations with the patent holders.
Anne van Dijk

Free Lance Journalist, Schoolmeesterstraat 102, 1053 MC Amsterdam, the Netherlands. Fax (+31) 161 220822

Sources
N.J. Taylor et al (1996), "Development of friable Embryogenic Callus and Embryogenic Suspension Culture Systems in Cassava." Nature Biotechnology, Vol 14, pp. 726­730.

C. Schöpke et al (1996), "Regeneration of Transgenic Cassava Plants from Microbombarded Embryogenic Suspension Cultures." Nature Biotechnology, Vol 14, pp. 731­735.

H.­Q. Li et al (1996), "Genetic Transformation of Cassava." Nature Biotechnology, Vol. 14, pp. 736­740.

Personal communications with A.M. Thro (Cassava Biotechnology Network), H. Wessels (DGIS), R. Visser (WAU), G. Henshaw (University of Bath), J. Puonti­Kaerlas (ETH), and G. Toenniessen (Rockefeller Foundation).