Agrobacterium T-DNA integration: molecules and models Tzvi Tzfira, Jianxiong Li, Benoıˆt Lacroix and Vitaly Citovsky Department of Biochemistry and Cell Biology, State University of New York, Stony Brook, NY 11794-5215, USA Genetic transformation mediated by Agrobacterium involves the transfer of a DNA molecule (T-DNA) from the bacterium to the eukaryotic host cell, and its inte- gration into the host genome. Whereas extensive work has revealed the biological mechanisms governing the production, Agrobacterium-to-plant cell transport and nuclear import of the Agrobacterium T-DNA, the inte- gration step remains largely unexplored, although several different T-DNA integration mechanisms have been suggested. Recent genetic and functional studies have revealed the importance of host proteins involved in DNA repair and maintenance for T-DNA integration. In this article, we review our understanding of the specific function of these proteins and propose a detailed model for integration. Agrobacterium is well known for its natural capability of trans-kingdom DNA transfer [1]. Although used mainly for plant genetic engineering [2], Agrobacterium can transform virtually any living cell, from other prokaryotes [3] to yeast [4] and fungi [5,6] to human cells [7]. The molecular basis for the ability of Agrobacterium to genetically transform its hosts has therefore been the subject of numerous studies over the past several decades [8–12]. In nature, Agrobacterium causes crown-gall disease, a neoplastic growth that results from the transfer of a transferred DNA (T-DNA) segment from the bacterial tumor-inducing (Ti) plasmid to the host cell, its integration into the host genome and the expression of its encoded genes [8–12]. The molecular machinery needed for T-DNA generation and transport into the host cell comprises proteins that are encoded by the bacterial chromosomal virulence (chv) genes and encoded by the T-plasmid virulence (vir) genes [8–12]. A great deal is known about the biological functions of most of the Agrobacterium Vir proteins, and about their function in T-DNA production and processing inside the bacterial cells. Less is known, however, about the mechanism by which the T-DNA travels into the host cell, and we have only recently begun exploring the function of host proteins in the transformation process, especially in its last two stages: T-DNA nuclear import and integration [2,12]. The genetic transformation process The Ti plasmid carries two components needed for genetic transformation: the vir and T-DNA regions (Figure 1a). Although the vir region encodes most of the bacterial proteins necessary for processing, transport and nuclear import of the T-DNA, the T-DNA carries no specific targeting signal nor does it encode any transport or integration functions. In fact, the T-DNA borders, which comprise two 25-bp direct repeats (Figure 1b), are the only required cis elements that delineate and thereby deter- mine the T-DNA region on the Ti plasmid [13], enabling replacement of the native T-DNA sequences between the borders with DNA of interest for genetic engineering [2]. Plant genetic transformation initiates with the induc- tion of the Agrobacterium vir region by specific host signals, usually small phenolic and certain monosacchar- ide molecules [14]. The induced VirD1 and VirD2 proteins act together as a site-specific nuclease [15], which cuts the bottom strand at the T-DNA borders (Figure 1b) and releases a single-stranded (ss) T-DNA molecule (T-strand). The T-strand (Figure 1c), together with several Vir proteins, is then exported to the host cell through a channel that is formed by the VirB and VirD4 proteins [16]. Once inside the host cell cytoplasm, the T-strand pre- sumably exists as a nucleoprotein complex (T-complex) with a single VirD2 molecule covalently attached to its 5 0 -end, and numerous VirE2 molecules covering its entire length. The T-complex is imported into the nucleus of the host cell with the assistance of VirD2, VirE2 and their cellular interactors [10,17]. VirD2 and VirE2 might also be involved in the final stage of the transformation, T-DNA integration [18–21], although it is generally recognized that plant factors, and not just the bacterial Vir proteins, have a crucial role in this process [8,22–25]. Although T-DNA integration is the crucial step of the transformation process, our understanding of its mechanism is still poor. In the following sections, we present the current knowl- edge of the function of bacterial and plant proteins in T-DNA integration and we discuss possible mechanisms for this process. From junction sequences to illegitimate recombination models Early studies concentrated on modeling the T-DNA integration mechanism by analyzing the junctions and structures of the integrated T-DNA (Figure 1d–f). This showed that T-DNA integration is not site-specific, and that integrated T-DNAs appeared to be distributed Corresponding author: Tzvi Tzfira (ttzfira@ms.cc.sunysb.edu). Review TRENDS in Genetics Vol.20 No.8 August 2004 www.sciencedirect.com 0168-9525/$ - see front matter q 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.tig.2004.06.004