Cassava is the staple food for nearly a billion people in 105 countries . Because of its resilience and capacity to grow on marginal lands, the importance of cassava cultivation in farming systems affected by climate change is expected to increase in the future . The use of cassava as energy crop also contributes to its increasing production acreage in tropical countries [3, 4]. However, the lack of resistance genes in the available germplasm, high heterozygosity, allopolyploidy, low fertility, and unsynchronized flowering make cassava improvement by conventional breeding a long and tedious process [5, 6]. Therefore cassava genetic transformation has emerged as a valuable alternative and complementary approach to improve cassava [7, 8]. Several protocols using either cotyledons or embryogenic cultures as target tissues and particle bombardment or Agrobacterium-mediated transformation procedures have been reported in the literature . However the use of embryogenic tissues (i.e. friable embryogenic callus, FEC) in combination with Agrobacterium-mediated transformation has become the favoured method because of its higher efficiency compared to the cotyledon-based protocol [9–13]. Despite this progress cassava remains difficult to transform partly as the result of low transformation and regeneration frequencies. The Agrobacterium-FEC system is also unstable and in some conditions produces highly variable numbers of transgenic events . As a consequence, cassava transformation requires well-trained tissue culture specialists, substantial amounts of plant material and repeated transformation cycles to generate a sufficient number of independent transgenic lines for research and product development. The instability of the transformation system renders the establishment of cassava transformation technology under less favourable conditions more challenging [14–16]. Recent progress in the optimization of the transformation protocol has substantially increased efficiency and robustness [17, 18]. The improved transformation protocol was subsequently established in laboratories located in Africa based on hands-on workshops and training of local scientists [15, 16].
The majority of transgenic cassava reports have been based on the transformation of the model cultivar 60444 (previously referred to as TMS 60444) . While proof-of-concept is possible with the model cultivar, the importance of transforming farmer- and industry-preferred cassava cultivars is essential for the adoption of transgenic cassava [7, 15, 16]. Because transgenic strategies to improve cassava are now being evaluated in the field [9, 12, 13] it is also important to assess the technology in cassava genotypes adapted to the respective field environments. Locally adapted cultivars and landraces have often been selected and adopted by farmers because of particular improved traits . Production of transgenic events in those selected genotypes offer the possibility to rapidly stack improved traits. Improvement of farmer-preferred genotypes using Agrobacterium-mediated transformation of FEC, however, has been limited by the difficulty of generating plant tissue suitable for transformation, the low regeneration efficiency of FEC, and the time necessary for embryo maturation following their co-cultivation with Agrobacterium. In particular, FEC initiation and time to regeneration are genotype dependent [20, 21].
Here we describe a modified and efficient method for transformation of three farmer-preferred cassava landraces that were selected based on their virus resistance [19, 22] as well as preferential and extensive use in Africa. Production of tissues suitable for transformation were generated and tested for regeneration. Because their multiplication and regeneration dynamics differed from cv. 60444, a modified transformation protocol was developed.