Phosphorylation and nuclear translocation of a regulator of G protein signaling (RGS10).

Heterotrimeric G proteins are involved in the transduction of hormonal and sensory signals across plasma membranes of eukaryotic cells. Hence, they are a critical point of control for a variety of agents that modulate cellular function. Activation of these proteins is dependent on GTP binding to their alpha (Galpha) subunits. Regulators of G protein signaling (RGS) bind specifically to activated Galpha proteins, potentiating the intrinsic GTPase activity of the Galpha proteins and thus expediting the termination of Galpha signaling. Although there are several points in most G protein controlled signaling pathways that are affected by reversible covalent modification, little evidence has been shown addressing whether or not the functions of RGS proteins are themselves regulated by such modifications. We report in this study the acute functional regulation of RGS10 thru the specific and inducible phosphorylation of RGS10 protein at serine 168 by cAMP-dependent kinase A. This phosphorylation nullifies the RGS10 activity at the plasma membrane, which controls the G protein-dependent activation of the inwardly rectifying potassium channel. Surprisingly, the phosphorylation-mediated attenuation of RGS10 activity was not manifested in an alteration of its ability to accelerate GTPase activity of Galpha. Rather, the phosphorylation event correlates with translocation of RGS10 from the plasma membrane and cytosol into the nucleus.

Inhibition of a Gi-activated potassium channel (GIRK1/4) by the Gq-coupled m1 muscarinic acetylcholine receptor.

The G protein-coupled inwardly rectifying K+ channel, GIRK1/GIRK4, can be activated by receptors coupled to the Galpha(i) subunit. An opposing role for Galpha(q) receptor signaling in GIRK regulation has only recently begun to be established. We have studied the effects of m1 muscarinic acetylcholine receptor (mAChR) stimulation, which is known to mobilize calcium and activate protein kinase C (PKC) by a Galpha(q)-dependent mechanism, on whole cell GIRK1/4 currents in Xenopus oocytes. We found that stimulation of the m1 mAChR suppresses both basal and dopamine 2 receptor-activated GIRK 1/4 currents. Overexpression of Gbetagamma subunits attenuates this effect, suggesting that increased binding of Gbetagamma to the GIRK channel can effectively compete with the G(q)-mediated inhibitory signal. This G(q) signal requires the use of second messenger molecules; pharmacology implicates a role for PKC and Ca2+ responses as m1 mAChR-mediated inhibition of GIRK channels is mimicked by PMA and Ca2+ ionophore. We have analyzed a series of mutant and chimeric channels suggesting that the GIRK4 subunit is capable of responding to G(q) signals and that the resulting current inhibition does not occur via phosphorylation of a canonical PKC site on the channel itself.

Single channel studies of inward rectifier potassium channel regulation by muscarinic acetylcholine receptors.

Negative regulation of the heartbeat rate involves the activation of an inwardly rectifying potassium current (I(KACh)) by G protein-coupled receptors such as the m2 muscarinic acetylcholine receptor. Recent studies have shown that this process involves the direct binding of G(betagamma) subunits to the NH(2)- and COOH-terminal cytoplasmic domains of the proteins termed GIRK1 and GIRK4 (Kir3.1 and Kir3.4/CIR), which mediate I(KACh). Because of the very low basal activity of native I(KACh), it has been difficult to determine the single channel effect of G(betagamma) subunit binding on I(KACh) activity. Through analysis of a novel G protein-activated chimeric inward rectifier channel that displays increased basal activity relative to I(KACh), we find that single channel activation can be explained by a G protein-dependent shift in the equilibrium of open channel transitions in favor of a bursting state of channel activity over a long-lived closed state.

Stimulus-specific assembly of enhancer complexes on the tumor necrosis factor alpha gene promoter.

The human tumor necrosis factor alpha (TNF-alpha) gene is rapidly activated in response to multiple signals of stress and inflammation. We have identified transcription factors present in the TNF-alpha enhancer complex in vivo following ionophore stimulation (ATF-2/Jun and NFAT) and virus infection (ATF-2/Jun, NFAT, and Sp1), demonstrating a novel role for NFAT and Sp1 in virus induction of gene expression. We show that virus infection results in calcium flux and calcineurin-dependent NFAT dephosphorylation; however, relatively lower levels of NFAT are present in the nucleus following virus infection as compared to ionophore stimulation. Strikingly, Sp1 functionally synergizes with NFAT and ATF-2/c-jun in the activation of TNF-alpha gene transcription and selectively associates with the TNF-alpha promoter upon virus infection but not upon ionophore stimulation in vivo. We conclude that the specificity of TNF-alpha transcriptional activation is achieved through the assembly of stimulus-specific enhancer complexes and through synergistic interactions among the distinct activators within these enhancer complexes.

Tyrosine kinase Pyk2 mediates G-protein-coupled receptor regulation of the Ewing sarcoma RNA-binding protein EWS.

Ewing family tumors result from the effects of chromosomal translocations that fuse the Ewing sarcoma (EWS) gene to various genes encoding transcription factors. The resulting chimeric EWS fusion proteins are transcriptional activators with transforming potential that have received much study. By contrast, the cellular function of somatic EWS remains obscure. EWS belongs to a family of RNA-binding proteins thought to play role in RNA synthesis or processing. Here, we show that EWS interacts with Pyk2, a protein tyrosine kinase implicated in a variety of signal transduction processes. G-protein-coupled receptor signaling and other stimuli of Pyk2 kinase activity significantly block the interaction between EWS and Pyk2. Furthermore, as assessed by sucrose gradient centrifugation, EWS partitions with dense ribosome-containing fractions in a manner that is enhanced by signaling from the G-protein-coupled m1 muscarinic acetylcholine receptor (mAChR). We conclude that extranuclear EWS is a previously unrecognized target of G-protein-coupled receptor regulation.

Receptor protein tyrosine phosphatase alpha participates in the m1 muscarinic acetylcholine receptor-dependent regulation of Kv1.2 channel activity.

The phosphorylation state of a given tyrosine residue is determined by both protein tyrosine kinase (PTK) and protein tyrosine phosphatase (PTP) activities. However, little is known about the functional interaction of these opposing activities at the level of an identified effector molecule. G protein-coupled receptors (GPCRs), including the m1 muscarinic acetylcholine receptor (mAChR), regulate a tyrosine kinase activity that phosphorylates and suppresses current generated by the Kv1.2 potassium channel. We examined the possibility that PTPs also participate in this signaling pathway since the tyrosine phosphatase inhibitor vanadate increases the extent of both Kv1.2 phosphorylation and suppression. We show that an endogenous transmembrane tyrosine phosphatase, receptor tyrosine phosphatase alpha (RPTPalpha), becomes tyrosine phosphorylated and co-immunoprecipitates with Kv1.2 in a manner dependent on m1 receptor activation. The N- and C-termini of Kv1.2 are shown to bind RPTPalpha in vitro. Overexpression of RPTPalpha in Xenopus oocytes increases resting Kv1.2 current. Biochemical and electrophysiological analysis reveals that recruiting RPTPalpha to Kv1.2 functionally reverses the tyrosine kinase-induced phosphorylation and suppression of Kv1.2 current in mammalian cells. Taken together, these results identify RPTPalpha as a new target of m1 mAChR signaling and reveal a novel regulatory mechanism whereby GPCR-mediated suppression of a potassium channel depends on the coordinate and parallel regulation of PTK and PTP activities.

The small GTP-binding protein RhoA regulates a delayed rectifier potassium channel.

Tyrosine kinases activated by G protein-coupled receptors can phosphorylate and thereby suppress the activity of the delayed rectifier potassium channel Kv1.2. Using a yeast two-hybrid screen, we identified the small GTP-binding protein RhoA as a necessary component in this process. Coimmunoprecipitation experiments confirmed that RhoA associates with Kv1.2. Electrophysiological analyses revealed that overexpression of RhoA markedly reduced the basal current generated by Kv1.2 expressed in Xenopus oocytes. Furthermore, in 293 cells expressing Kv1.2 and ml muscarinic acetylcholine receptors, inactivating RhoA using C3 exoenzyme blocked the ability of ml receptors to suppress Kv1.2 current. Therefore, these results demonstrate that RhoA regulates Kv1.2 activity and is a central component in the mechanism of receptor-mediated tyrosine kinase-dependent suppression of Kv1.2.

The m3 muscarinic acetylcholine receptor differentially regulates calcium influx and release through modulation of monovalent cation channels.

Several types of transmembrane receptors regulate cellular responses through the activation of phospholipase C-mediated Ca2+ release from intracellular stores. In non-excitable cells, the initial Ca2+ release is typically followed by a prolonged Ca2+ influx phase that is important for the regulation of several Ca2+-sensitive responses. Here we describe an agonist concentration-dependent mechanism by which m3 muscarinic acetylcholine receptors (mAChRs) differentially regulate the magnitude of the release and influx components of a Ca2+ response. In transfected Chinese hamster ovary cells expressing m3 mAChRs, doses of the muscarinic agonist carbachol ranging from 100 nM to 1 mM evoked Ca2+ release responses of increasing magnitude; maximal Ca2+ release was elicited by the highest carbachol concentration. In contrast, Ca2+ influx was maximal when m3 mAChRs were activated by moderate doses (1-10 microM) of carbachol, but substantially reduced at higher agonist concentrations. Manipulation of the membrane potential revealed that the carbachol-induced Ca2+ influx phase was diminished at depolarized potentials. Importantly, carbachol doses above 10 microM were found to couple m3 mAChRs to the activation of an inward, monovalent cation current resulting in depolarization of the cell membrane and a selective decrease in the influx, but not release, component of the Ca2+ response. These studies demonstrate, in one experimental system, a mechanism by which a single subtype of G-protein-coupled receptor can utilize the information encoded in the concentration of an agonist to generate distinct intracellular Ca2+ signals.

Activation of protein tyrosine kinase PYK2 by the m1 muscarinic acetylcholine receptor.

Several G protein-coupled receptors are known to direct the tyrosine phosphorylation, and in some cases the activation, of diverse tyrosine kinases. Using a stable cell line approach, we characterize the activation of PYK2, a tyrosine kinase structurally related to focal adhesion kinase, by the G protein-coupled m1 muscarinic acetylcholine receptor. We find that PYK2 tyrosine kinase activity is critical for the m1 receptor-stimulated tyrosine phosphorylation of PYK2. Furthermore, we identify two tyrosine residues that are subject to phosphorylation in response to muscarinic signaling and show that this phosphorylation induces two cytosolic proteins, c-Src and Grb2, to bind to PYK2. This is the first demonstration of the significance played by distinct PYK2 tyrosine residues in G protein-coupled signaling to this kinase. By comparison, though m1 receptors induce the tyrosine phosphorylation of the cytoskeletal protein paxillin, the association of paxillin with PYK2 is unaffected by muscarinic signaling. We also provide evidence that PYK2 specifically phosphorylates the carboxyl-terminal cytosolic portion of the potassium channel Kv1.2 in a manner regulated by the m1 receptor. These results delineate molecular events attending the m1 muscarinic receptor stimulation of this tyrosine kinase and establish PYK2 as an effector of the m1 muscarinic receptor in the regulation of multiple cell functions.

The m1 muscarinic acetylcholine receptor transactivates the EGF receptor to modulate ion channel activity.

Intracellular tyrosine kinases link the G protein-coupled m1 muscarinic acetylcholine receptor (mAChR) to multiple cellular responses. However, the mechanisms by which m1 mAChRs stimulate tyrosine kinase activity and the identity of the kinases within particular signaling pathways remain largely unknown. We show that the epidermal growth factor receptor (EGFR), a single transmembrane receptor tyrosine kinase, becomes catalytically active and dimerized through an m1 mAChR-regulated pathway that requires protein kinase C, but is independent of EGF. Finally, we demonstrate that transactivation of the EGFR plays a major role in a pathway linking m1 mAChRs to modulation of the Kv1.2 potassium channel. These results demonstrate a ligand-independent mechanism of EGFR transactivation by m1 mAChRs and reveal a novel role for these growth factor receptors in the regulation of ion channels by G protein-coupled receptors.

RGS10 is a selective activator of G alpha i GTPase activity.

Polypeptides that define a protein family termed RGS (for regulators of G-protein signalling) are encoded by the SST2 gene of the yeast Saccharomyces cerevisiae, the EGL-10 gene of the nematode Caenorhabdatis elegans, and several related mammalian genes. Genetic studies in invertebrates and mammalian cell-transfection experiments indicate that RGS proteins negatively regulate signalling pathways involving seven transmembrane receptors and heterotrimeric G proteins. However, the biochemical mechanism by which RGS proteins control these pathways is unknown. Here we report the characterization of human RGS10, a member of this protein family. Co-immunoprecipitation studies demonstrate that RGS10 associates specifically with the activated forms of two related G-protein subunits, G alphai3, and G alphaz, but fails to interact with the structurally and functionally distinct G alphas subunit. In vitro assays with purified proteins indicate that RGS10 increases potently and selectively the GTP hydrolytic activity of several members of the G alphai family, including G alphai3, G alphaz, and G alpha0. These results demonstrate that RGS proteins can attenuate signalling pathways involving heterotrimeric G proteins by serving as GTPase-activating proteins for specific types of G alpha subunits.

Identification of domains conferring G protein regulation on inward rectifier potassium channels.

Cardiac m2 muscarinic acetylcholine receptors reduce heart rate by coupling to heterotrimeric (alpha beta gamma) guanine nucleotide-binding (G) proteins that activate IKACh, an inward rectifier K+ channel (IRK). Activation of the GIRK subunit of IKACh requires G beta gamma subunits; however, the structural basis of channel regulation is unknown. To determine which sequences confer G beta gamma regulation upon IRKs, we generated chimeric proteins composed of GIRK and RB-IRK2, a related, G protein-insensitive channel. Importantly, a chimeric channel containing the hydrophobic pore region of RB-IRK2 joined to the amino and carboxyl termini of GIRK exhibited voltage- and receptor-dependent activation in Xenopus oocytes. Furthermore, carboxy-terminal sequences specific to this chimera and GIRK bound G beta gamma subunits in vitro. Thus, G beta gamma may regulate IRKs by interacting with sequences adjacent to the putative channel pore.

Coincidence detection at the level of phospholipase C activation mediated by the m4 muscarinic acetylcholine receptor.

BACKGROUND: One of the principal mechanisms by which G-protein-coupled receptors evoke cellular responses is through the activation of phospholipase C (PLC) and the subsequent release of Ca2+ from intracellular stores. Receptors that couple to pertussis toxin (PTX)-insensitive G proteins typically evoke large increases in PLC activity and intracellular Ca2+ release. In contrast, receptors that use only PTX-sensitive G proteins usually generate weak PLC-dependent responses, but efficiently regulate a second effector enzyme, adenylyl cyclase. For example, in many cell types, agonist binding by the m4 muscarinic acetylcholine receptor (m4 receptor) results in a strong inhibition of adenylyl cyclase and very little stimulation of PLC activity or release of intracellular Ca2+. We have investigated whether the weak, PTX-sensitive stimulation of PLC activity by the m4 receptor can play a significant role in the generation of cellular responses. RESULTS: We report here that PTX-sensitive Ca2+ release mediated by the m4 receptor in transfected Chinese hamster ovary cells is greatly enhanced when endogenous purinergic receptors simultaneously activate a PTX-insensitive signaling pathway. Furthermore, m4-receptor-induced transcription of the c-fos gene (a Ca(2+)-sensitive response) is similarly potentiated when purinergic receptors are coactivated. These enhanced m4-receptor-dependent Ca2+ responses do not require an influx of external Ca2+, and occur in the absence of detectable purinergic-receptor-stimulated Ca2+ release; they apparently require the activation of both PTX-sensitive and PTX-insensitive G-protein pathways. Measurements of phosphoinositide hydrolysis indicate that the enhancement of m4-receptor-mediated Ca2+ signaling by purinergic receptors is due to a synergistic increase in agonist-stimulated PLC activity. CONCLUSIONS: These studies demonstrate that the potency of m4-receptor-mediated PLC signaling is highly dependent upon the presence or absence of other PLC-activating agonists. The ability of the m4 receptor to evoke a strong, but conditional, activation of PLC may allow this type of receptor to participate in a coincidence-detection system that amplifies simultaneous PLC-activating signals through a mechanism involving crosstalk between PTX-sensitive and PTX-insensitive G-protein pathways.

Dual modulation of a potassium channel by the m1 muscarinic and beta2-adrenergic receptors.

Neurotransmitter receptors alter membrane excitability and synaptic efficacy by generating intracellular signals that ultimately change the properties of ion channels. Given their critical role in controlling cell membrane potential, potassium channels are frequently the targets of modulatory signals from many different G protein-coupled receptors. However, due to the heterogeneity of potassium channel expression in vivo, it has been difficult to determine the molecular mechanisms governing the regulation of molecularly defined potassium channels. Through expression studies in Xenopus oocytes and mammalian cells, we found that the m1 muscarinic acetylcholine receptor (mAChR) potently suppresses a cloned delayed rectifier potassium channel, termed RAK, through a pathway involving phospholipase C activation and direct tyrosine phosphorylation of the RAK protein. In contrast, we found that RAK channel activity is strongly enhanced following agonist activation of beta2-adrenergic receptors; this effect requires a single PKA consensus phosphorylation site located near the amino terminus of the channel protein. These results demonstrate that a specific type of potassium channel that is widely expressed in the mammalian brain and heart is subject to both positive and negative regulation by G protein-dependent pathways.

Heterotrimeric G proteins containing G alpha i3 regulate multiple effector enzymes in the same cell. Activation of phospholipases C and A2 and inhibition of adenylyl cyclase.

Previous studies have shown that a single type of transmembrane receptor is able to regulate multiple effectors through the activation of heterotrimeric G proteins. For example, the m2 muscarinic acetylcholine receptor (mAChR) expressed in Chinese hamster ovary (CHO) cells inhibits adenylyl cyclase, stimulates phospholipase C-dependent intracellular Ca2+ release, and activates phospholipase A2 through pertussis toxin-sensitive G proteins. However, it is unclear whether multiple effector enzymes can be regulated by one type of heterotrimeric G protein within a single cell. To investigate this question, we constructed a derivative of G alpha i3 (termed G alpha i3 C > S) in which the carboxyl-terminal cysteine residue, the site for pertussis toxin modification, was changed to a serine. Following pertussis toxin treatment of transfected CHO cells expressing the m2 mAChR, we found that the G alpha i3 C > S protein underwent guanine nucleotide exchange in response to the muscarinic agonist carbachol, while the m2 mAChR failed to activate the endogenous G alpha i2 and G alpha i3 proteins. Moreover, coupling of heterotrimeric G proteins containing G alpha i3 C > S to the m2 mAChR resulted in pertussis toxin-resistant inhibition of adenylyl cyclase, stimulation of phospholipase C-induced intracellular Ca2+ release, and phospholipase A2-mediated arachidonic acid release. Therefore, these studies provide conclusive evidence that heterotrimeric G proteins containing just G alpha i3 can regulate multiple effector enzymes within the same cell type.

Molecular basis of cardiac potassium channel stimulation by protein kinase A.

Cardiac beta-adrenergic receptors accelerate heart rate by modulating ionic currents through a pathway involving cyclic AMP-dependent protein kinase A (PKA). Previous studies have focused on the regulation of Ca2+ channels by PKA; however, due to the heterogeneity of K+ channels expressed within the heart, little is known about the mechanism by which PKA modulates individual K+ channels. Here we report that PKA strongly enhanced the activity of a cloned delayed rectifier K+ channel that is normally expressed in cardiac atria. This effect required a single PKA consensus phosphorylation site located near the amino terminus of the channel protein. Furthermore, patch clamp analysis revealed that PKA phosphorylation increased the open time that single channels spend in higher conductance states. These studies provide evidence that hormonal modulation of a cardiac K+ channel involves direct phosphorylation by PKA.

Tyrosine kinase-dependent suppression of a potassium channel by the G protein-coupled m1 muscarinic acetylcholine receptor.

Neurotransmitter receptors alter membrane excitability and synaptic efficacy by generating intracellular signals that ultimately change the properties of ion channels. Through expression studies in Xenopus oocytes and mammalian cells, we found that the G protein-coupled m1 muscarinic acetylcholine receptor potently suppresses a cloned delayed rectifier K+ channel through a pathway involving phospholipase C activation and direct tyrosine phosphorylation of the K+ channel. Furthermore, analysis of neuroblastoma cells revealed that a similar tyrosine kinase-dependent pathway links endogenous G protein-coupled receptors to suppression of the native RAK channel. These results suggest a novel mechanism by which neurotransmitters and hormones may regulate a specific type of K+ channel that is widely expressed in the mammalian brain and heart.

Charged amino acids required for signal transduction by the m3 muscarinic acetylcholine receptor.

The five muscarinic acetylcholine receptor (mAChR) subtypes, termed m1-m5, transduce agonist signals across the plasma membrane by activating guanine nucleotide binding (G) proteins. The large cytoplasmic domain joining the fifth and sixth transmembrane segments of mAChRs plays a critical role in controlling the specificity of G protein coupling. In this study, we determined which sequences within this domain are required for activation of signaling by the m3 mAChR. By measuring the ability of normal and mutant m3 mAChRs to couple to the G protein pathway leading to activation of phospholipase C and Ca(2+)-dependent chloride currents in RNA-injected Xenopus oocytes, we found that two clusters of charged residues near the fifth and sixth transmembrane segments were required for normal signaling; furthermore, the position of these sequences was critical for their function. Finally, analysis of deletion mutant m3 mAChRs confirmed the importance of these sequences; receptors containing as few as 22 out of 239 amino acids of the cytoplasmic domain were fully active in signaling if they included the critical charged residues. Sequence comparisons suggest that similar charged sequences may be required for signal transduction by many G protein-coupled receptors.

Transfected m2 muscarinic acetylcholine receptors couple to G alpha i2 and G alpha i3 in Chinese hamster ovary cells. Activation and desensitization of the phospholipase C signaling pathway.

Muscarinic acetylcholine receptor subtypes (m1-m5) differentially regulate phosphoinositide-specific phospholipase (PLC) through the activation of distinct guanine nucleotide-binding (G) proteins which can be distinguished on the basis of their sensitivity to inhibition by pertussis toxin (PTX). In transfected Chinese hamster ovary cells, the m2 receptor subtype regulates the stimulation of PLC and inhibition of adenylyl cyclase (AC) through PTX-sensitive G proteins. In this study, we utilized the ability of cholera toxin (CTX) to ADP-ribosylate PTX-sensitive alpha subunits as part of the ternary complex formed by heterotrimeric G proteins and agonist-bound receptors to detect and characterize the interactions between transfected m2 receptors and endogenous G proteins in Chinese hamster ovary cells. In membranes derived from cells expressing the m2, but not the m3 receptor, the cholinergic agonist carbachol stimulated CTX modification of a 40-kDa species (G alpha 40). Importantly, similar carbachol dose dependence values and PTX dose sensitivities were observed for m2 receptor-mediated PLC signaling and G alpha 40-CTX modification. High resolution urea-SDS-polyacrylamide gel electrophoresis analysis revealed that G alpha i2 (40 kDa) and G alpha i3 (41 kDa) were components of the G alpha 40 identified by m2 receptor-dependent CTX modification. Furthermore, the sensitivities of G alpha i2 and G alpha i3 to PTX modification were determined to be the same as those for PTX inhibition of G alpha 40 labeling by CTX and m2 receptor-mediated PLC signaling. Similarly, agonist-induced desensitization of m2 receptor-G protein signaling required doses of agonist associated with stimulation of PLC. Desensitization involved receptor sequestration and down-regulation of G alpha i3; however, only the reduction of G alpha i3 required prior activation PLC signaling. Finally, desensitization of m2-G protein coupling could be partially mimicked by treatment with thapsigargin, an inducer of intracellular Ca2+ release, without altering the number of cell surface receptors or G protein levels. These results demonstrate that m2 receptors couple to both G alpha i2 and G alpha i3 in vivo and that this interaction is integral to both positive and negative regulatory pathways leading to activation of PLC and desensitization of receptor signaling.

Agrobacterium tumefaciens transfers extremely long T-DNAs by a unidirectional mechanism.

During crown gall tumorigenesis, part of the Agrobacterium tumefaciens tumor-inducing (Ti) plasmid, the T-DNA, integrates into plant DNA. Direct repeats define the left and right ends of the T-DNA, but tumorigenesis requires only the right-hand repeat. Virulence (vir) genes act in trans to mobilize the T-DNA into plant cells. Transfer of T-DNA begins when the VirD endonuclease cleaves within the right-hand border repeat. Although the T-DNA right-border repeat promotes T-DNA transmission best in its normal orientation, an inverted right border exhibits reduced but significant activity. Two models may account for this diminished tumorigenesis. The right border may function bidirectionally, with strong activity only in its wild-type orientation, or it may promote T-DNA transfer in a unidirectional manner such that, with an inverted right border, transfer proceeds around the entire Ti plasmid before reaching the T-DNA. To determine whether a substantial portion of the Ti plasmid is transferred to plant cells, as predicted by the unidirectional-transfer hypothesis, we examined T-DNAs in tumors induced by strains containing a Ti plasmid with a right border inverted with respect to the T-DNA oncogenes. These tumors contained extremely long T-DNAs corresponding to most or all of the Ti plasmid. To test whether the right border can function bidirectionally, we inserted T-DNAs with either a properly oriented or an inverted right border into a specific site in the A. tumefaciens chromosome. A border situated to transfer the oncogenes first directed T-DNA transfer even from the bacterial chromosome, whereas a border in the opposite (inverted) orientation did not transfer the oncogenes to plant cells. Our results indicate that the right-border repeat functions in a unidirectional manner.