Cellulose binding domains (CBDs) represent a subset of carbohydrate binding modules (CBMs). CBDs are found associated with the majority of saprophyte-encoded cellulolytic enzymes, whereas they are generally not found associated with plant pathogen-encoded or with plant-encoded cellulolytic enzymes [1, 2]. The principle function of the CBD is to mediate adherence to the carbohydrate substrate [3, 4], while an additional function can be found in the ability to of some CBDs to directly dissociate cellulose microfibrils [5]. Within the CBD families, there are a few examples of CBDs that are present as independent proteins, often associated with the cell wall. An example is CBD 1 from the potato late blight disease inducing phytopathogen Phytophthora infestans [6]. The P. infestans CBD 1 protein is found tightly associated with the cellulose-containing hyphal cell wall, and may play a role in assembly and/or integrity of the cell wall. This led us to the idea that expression of this protein in potato plants may alter the host: pathogen interactions through binding to the host cellulose molecules, proving a potential shield from pathogen endoglucanases, or conversely, binding to Phytophthora hyphae, providing a shield from host endoglucanases. We report the unexpected outcome of P. infestans CBD 1 expression in transgenic Bintje potato plants, restoration of female fertility. We demonstrate the usefulness of the fertility through crosses resulting in late blight resistant plants with expanded tuber colors.

Our original hypothesis was that introducing CBD 1 into transgenic potato might alter the host: pathogen interaction. One way this could happen would be if CBD 1 acted in a manner similar to Cladosporium fulvum AVR4, where the chitin binding AVR4 protein protects the hyphae from plant chitinases, preventing release of elicitor fragments [9, 10]. In the soybean: P. sojae interaction, soybean produces glucanases that attack the cellulosic P. sojae hyphae, releasing elicitor fragments. CBD 1 protein could be one mechanism for protecting the hyphae, along with the reported glucanase inhibitor proteins [11]. This would lead to greater host susceptibility, but this was not seen in our studies.

An alternative possibility would be that CBD 1 interacts with the host cell wall. Interactions with the host cell wall could increase susceptibility if the CBD acted to loosen cellulose microfibrils, or reduce susceptibility if the CBD bound to and shielded substrates susceptible to pathogen enzymes. While there was no evidence that these mechanisms were active, based on indistinguishable differences in susceptibility between control and transgenic CBD plants, there was an obvious effect on gross plant morphology.

There is evidence to suggest that CBDs can interact with the plant cell wall. This is seen in the use of binding modules as molecular probes of cell wall structure [3, 4]. It is also suggested in the limited reports of their use in transgenic plants. In one case, a family 3 CBD protein from the saprophytic bacterium Clostridium cellulovorans [12], was found to increase early growth of transgenic Desiree potato plants [13, 14]. At higher copy numbers there was an inhibition of growth. Interestingly, our use of a family 1 CBD resulted in greater growth only at later stages of plant development, and allowed for ovary development. The use of a Family 22 xylan binding module in transgenic tobacco failed to show any apparent effects on morphology or growth rates [15], while another study showed a marked reduction in growth of tobacco when a tandem CBM was expressed in transgenic materials [16]. Given the variable effects of CBDs in transgenic plants, there can be no assumptions made about the effect of any carbohydrate binding modules until tested.

The unusual feature of the Phytophthora CBD 1 expressing transgenic Bintje is the ovary formation. Even in the absence of fertilization the transgenic plants are able to produce berries. This would be similar to parthenocarpic fruit development, where fruit develops in the absence of seed development. This has been engineered into Solanaceous crops through manipulation of tissue-targeted auxin production [17].

Auxin has a well known role in causing cell wall loosening and cell expansion, through induction of various plant-encoded cell wall carbohydrate modifying enzymes [18]. After successful fertilization, the Bintje seed berries are larger than those produced without fertilization, indicating a normal interaction between seed development and ovary growth. The impact of CBD 1 on ovary development is unique as it has never been seen in other transgenic Bintjes, including the Phytophthora CBD4 (FJ524852) which encodes a double, or tandem, cellulose binding region (unpublished data). This discovery provides a new model for understanding ovary development in relation to cell wall proteins. It is of interest that Bintje can act as a female parent, once the block to ovary development is overcome. Further investigations may provide a genetic basis for this lack of ovary formation within the Bintje genome. Another use for this information may be in the development of tomato fruits. The developing fruit of tomato is the equivalent of the potato seed berry, thus there may be methods for producing fruits without pollination, or producing modified fruit shapes through use of CBD 1.

The mechanism of CBD-mediated ovary development remains to be characterized, however, CBD expression can be a useful tool for developing the Bintje gene pool, one that hasn’t changed since 1904 [19]. The cultivar Bintje is one of the most widely grown European cultivars (http://www.europotato.org) due to strong yields, growth under variable conditions, and excellent flavor under different cooking methods. Cultivar weakness is seen in a lack of resistance to many major pathogens of potato. We demonstrate the ability to integrate multigenic late blight resistance, as well as nutrition-relevant color changes, into the Bintje gene pool. Selection can be made for progeny of transgenic Bintje that lack the CBD 1 transgene. The primary value of CBD 1 expression is in breeding. The progeny segregating for loss of the CBD 1 revert to the sterile form of the original Bintje cultivar. This will allow for improved Bintje without carryover of transgenes, for those markets where that is desired. Initial experiments indicate that the CBD 1 gene can also be used in other potatoes lacking female fertility, such as Russet Burbank, although initial progeny have not been as vigorous (unpublished).

Expression of cellulose binding domain protein resulted in development of ovaries in the sterile potato cultivar Bintje. Ovaries formed in the absence of fertilization, while larger ovaries were formed after outcrossing with fertile pollen and subsequent seed production. This opens the possibility for restoring ovary formation to other sterile germplasm, as well as the possibility of modifying the production of consumable ovaries such as tomatoes.