Bacterial cell structures are macromolecular features found among all types bacteria. They often serve essential or specialized functions, and are important contributors to the widespread diversity and success of bacteria. I have focused my research on the formation and localization of two of these structures in the plant pathogen Agrobacterium tumefaciens: the type IV secretion system (T4SS) and the bacterial cell envelope.

A. tumefaciens is an Alphaproteobacterium uniquely capable of genetically transforming plant host cells during its infection process. DNA and protein substrates are transferred into plant cells via the T4SS, which forms multiple complexes around the bacterial circumference. To understand the mechanisms driving T4SS positioning, I assessed their spatial distribution by quantitative analysis and modeling. My findings indicate that these secretion complexes localize in a non-random periodic pattern along the cell perimeter. These results, along with additional plant infection studies, suggest this spatial organization serves to promote efficient delivery of substrates into host cells

A potential mechanism behind this periodic arrangement is the unipolar cell growth recently described for A. tumefaciens. To better understand this unusual growth mechanism, I carefully scrutinized growing cells, nascent peptidoglycan, and peptidoglycan synthesis components by electron and fluorescence microscopy. My results indicate that the unipolar growth process is surprisingly dynamic, multi-dimensional, and involves both novel and familiar components. These findings thus represent a significant departure from the canonical growth mechanism of E. coli and other well-studied bacilli.

The interaction between Agrobacterium tumefaciens and its plant host is a classic model system for investigating host pathogen interactions. Significant progress has been made in various aspects of A. tumefaciens mediated genetic transformation of plants, which leads to the formation of crown gall tumors. It is known that wounded plant cells release phenolic compounds induces the expression of the transformation products, the T-strand and virulence (Vir) proteins, which are encoded on the tumor-inducing plasmid (pTi). The Vir proteins assemble a translocation channel called the virulence-induced type IV secretion system (vir-T4SS). The T-strand is a single stranded segment of DNA that gets integrated into the host plant cell and the expression of the genes on the T-strand causes crown gall tumors.

The vir-T4SS is composed of 11 VirB proteins and VirD4, and all 12 components are essential for maximal DNA and protein transport. These components can be divided into 3 major groups. The first group consists of the T-pilus and its associated components (VirB1, VirB2, VirB3, and VirB5). VirB1 has homology to lytic transglycolases and is likely to cleave the peptidoglycan to facilitate assembly of the vir-T4SS. VirB2 is the major component of T-pilus, and VirB5 is localized at the T-pilus tip. VirB1 and VirB3 are required for T-pilus assembly. The second group consists of the core transport complex that spans the inner and outer membranes and periplasm (VirB6-VirB10). The third group consists of the energetic components (VirB4, VirB11, and VirD4). These components have ATPase homology and ATP-binding motifs, and may energize the assembly and transfer of DNA and proteins through the vir-T4SS. VirD4 is also the coupling protein that brings DNA and vir-T4SS substrate proteins to the vir-T4SS. The vir-T4SS transports the T-strand, and at least 4 Vir protein substrates (VirD2, VirE2, VirE3, and VirF) into the host cell. It is estimated that 50 T-strands and thousands of VirE2, VirE3, and VirF proteins are all transported through the vir-T4SS. However, it still remains unclear how many vir -T4SS are localized on the bacterial cell and how the vir -T4SS interacts with the host cell to transport the T-strand and the vir-T4SS protein substrates.

It was proposed that most Vir proteins were localized at the pole of A. tumefaciens and that A. tumefaciens attached to host cells via its poles. However, early studies from the Zambryski lab suggested that at least one component of the T4SS, VirB8, localized around the bacterial perimeter in transverse-section and in multiple sites in longitudinal section of vir-induced cells. Other studies suggested that not all T4SS components localize to the same position perhaps suggesting subassemblies. Therefore, to clear these inconsistencies and to identify how many vir-T4SS are localized on the bacterial cell and to better understand how the vir-T4SS interact with the host cell to transport the T-strand and the vir-T4SS protein substrates, I analyzed the localization of the vir-T4SS components and substrates in the presence and absence of its host.

I cloned the structural components of the vir-T4SS and the vir-T4SS substrates as GFP fusion vectors for induced expression in the two most common lab strains of A. tumefaciens. After induction, GFP fusions were imaged by high-resolution deconvolution microscopy. Tumorigenesis assays were used to test the functionality of fusion proteins. I found that those that carried a cytoplasmically localized GFP retained their functions, while those with predicted periplasmic GFP and GFP fusions to VirB10 had a dominant-negative effect on tumor formation. Thus, suggesting the addition of GFP to these proteins may interfere with the correct assembly and proper function of the vir-T4SS. Three-dimensional reconstruction of the deconvolved images revealed that most of the proteins were found in multiple foci around the cell periphery. I confirmed these findings by examining endogenous protein patterns by immunofluorescence microscopy. Extensive quantitative analyses also verified that the Vir proteins of the vir-T4SS are localized to multiple foci around the periphery of the Agrobacterial cell. In support, I found multiple T-pili on bacterial cells. Interestingly, the localization of the vir-T4SS components resembled a helical pattern. Other proteins that form helical patterns are bacterial cytoskeletal components, such as MinD. In fact, VirB8 and MinD colocalize suggesting that cytoskeletal components may provide an existing scaffold for vir-T4SS assembly. Furthermore, I found that Agrobacteria attaches predominantly laterally to host cells and few cells attach at their poles. Also, the localization of the vir-T4SS does not change upon contact to host cells suggesting that perhaps multiple vir-T4SS are utilized.

In summary, my results revealed that the vir-T4SS is localized around the perimeter of the cell resembling a helical pattern, and that A. tumefaciens attached to host cells predominantly along their sides. My work challenges the existing paradigm of polar localization and orientation of attachment and has led to a novel model with multiple vir-T4SS around the bacteria. This would maximize the possibility for effective attachment along the bacterial length, which may allow efficient transfer of numerous amounts of T-strands and thousands of protein substrates into the host cell.

Agrobacterium tumefaciens is a rod-shaped gram-negative bacterium, which elongates by unipolar addition of new cell envelope material. Approaching cell division, the growth pole transitions to a non-growing old pole, and the division site creates new growth poles in sibling cells. This thesis describes experiments concerning cell division and polar growth in A. tumefaciens. The cell division proteins FtsZ-GFP and FtsA-GFP localize to the growth pole in addition to the Z-ring at the mid-cell. The A. tumefaciens homolog of the Caulobacter crescentus polar organizing protein PopZCc (PopZAt) localizes specifically to growth poles. In contrast, the A. tumefaciens homolog of the C. crescentus polar organelle development protein PodJCc (PodJAt) localizes to the old pole early in the cell cycle and accumulates at the growth pole as the cell cycle proceeds. FtsA and FtsZ also localize to the growth pole for most of the cell cycle prior to Z-ring formation. To further characterize the function of polar localizing proteins, I created a deletion of podJAt . ΔpodJAt cells display ectopic growth poles (branching), growth poles that fail to transition to an old pole, and elongated cells that fail to divide. In ΔpodJAt cells, PopZAt -GFP persists at non-transitioning growth poles post division and also localizes to ectopic growth poles, as expected for a growth pole specific factor. Even though GFP-PodJAt does not localize to the midcell in wild type, deletion of podJAt impacts localization, stability, and function of Z-rings as assayed by localization of FtsA-GFP and FtsZ-GFP. Z-ring defects are further evidenced by minicell production. Together these data indicate that PodJAt is a critical factor for polar growth, and ΔpodJAt displays a cell division phenotype. These results reveal intimate connections between polar growth and cell division in a model system displaying a novel form of cell growth.

Plant cells transport molecules through small channels called plasmodesmata (PD); these nanoscopic channels are ~30-50 nm in diameter and cross through cellulosic walls to connect neighboring cells. PD regulate the movement of key developmental factors (e.g. transcription factors, small RNAs, and hormones) from cell to cell, which influence the formation of new organs, cell types, embryos, and the meristem. Mutants that affect PD transport were identified indirectly, in screens for defective developmental processes (e.g. stomata formation or root development), and in screens directed at finding PD transport mutants.

The genes identified in mutant screens for altered PD transport encode proteins that do not localize to PD. These genes that affect PD transport include INCREASED SIZE EXCLUSION LIMIT (ISE) 1 and ISE2, which encode RNA helicase proteins that localize to mitochondria and plastids, respectively; GFP ARRESTED TRAFFICKING 1 (GAT1), which encodes a plastid-localized thioredoxin; CHAPERONIN CONTAINING TCP1 8 (CCT8), which encodes a cytoplasmic chaperonin subunit; and DECREASED SIZE EXCLUSION LIMIT 1 (DSE1), which encodes a WD40-repeat protein found in the cytoplasm and nucleus.

Loss of DSE1 causes decreased PD transport both in Arabidopsis thaliana embryos and in the leaves of a distantly related eudicot species Nicotiana benthamiana. Through transcriptome analysis, we demonstrate here that DSE1 affects the expression of seed maturation and abscisic acid (ABA) response genes and that ABA INSENSITIVE 5 (ABI5), a key ABA transcription factor, reduces cell-cell transport. Further, dse1 embryos have reduced PD frequency, suggesting that the PD transport phenotype observed in this mutant is due to fewer PD in the cell wall.

We characterize a new PD mutant ise3 that was identified in a screen for mutants with increased PD transport in A. thaliana embryos. ISE3 encodes a SEL1-like repeat mitochondrial protein and we show that this protein interacts with an EMBRYO DEFECTIVE-mitochondrial-PPR-domain-containing protein in Bimolecular Fluorescence Complementation. Loss of the N. benthamiana orthologs of ISE3 or its interactor, ISE3 INTERACTING PROTEIN 1 (IPR1), cause increased PD transport in leaves. We also show that ISE3, like ISE1 affects the expression of genes involved in reactive oxygen species (ROS) regulation and show that loss of ISE3 or IPR1 cause increased H2O2 production. These results support findings that ROS can regulate PD transport, and further suggest that the mitochondria and the plastids signal to the nucleus (through ‘retrograde signaling’) to coordinate nuclear gene expression and PD transport, in a pathway called ‘organelle-nucleus-PD signaling’, or ONPS. This thesis supports the growing body of evidence that intercellular transport is driven by critical intracellular signals and processes.

Retrograde signalling pathways require signal exchange among the organelles and the nucleus, and a putative pathway for this exchange is through “stromules,” which are stroma-filled tubular extensions of plastids. Herein we describe the first identified signal transduction pathways initiated by the plastid that are linked to stromule formation. We show that stromule frequency is affected by light cues as well as by ROS and the redox status of the chloroplast. Finally, we demonstrate that isolated chloroplasts can make stromules, indicating that the plastid drives the formation of these tubule structures, which may be involved in intracellular signalling.

Unlabelled

Importance