Identification and characterization of a Ca2+ -sensitive interaction of the vanilloid receptor TRPV1 with tubulin

The vanilloid receptor TRPV1 plays a well-established functional role in the detection of a range of chemical and thermal noxious stimuli, such as those associated with tissue inflammation and the resulting pain. TRPV1 activation results in membrane depolarization, but may also trigger intracellular Ca2+ -signalling events. In a proteomic screen for proteins associated with the C-terminal sequence of TRPV1, we identified beta-tubulin as a specific TRPV1-interacting protein. We demonstrate that the TRPV1 C-terminal tail is capable of binding tubulin dimers, as well as of binding polymerized microtubules. The interaction is Ca2+ -sensitive, and affects microtubule properties, such as microtubule sensitivity towards low temperatures and nocodazole. Our data thus provide compelling evidence for the interaction of TRPV1 with the cytoskeleton. The Ca2+ -sensitivity of this interaction suggests that the microtubule cytoskeleton at the cell membrane may be a downstream effector of TRPV1 activation.     

Importance of Non-Selective Cation Channel TRPV4 Interaction with Cytoskeleton and Their Reciprocal Regulations in Cultured Cells

Background

TRPV4 and the cellular cytoskeleton have each been reported to influence cellular mechanosensitive processes as well as the development of mechanical hyperalgesia. If and how TRPV4 interacts with the microtubule and actin cytoskeleton at a molecular and functional level is not known.

Methodology and Principal Findings

We investigated the interaction of TRPV4 with cytoskeletal components biochemically, cell biologically by observing morphological changes of DRG-neurons and DRG-neuron-derived F-11 cells, as well as functionally with calcium imaging. We find that TRPV4 physically interacts with tubulin, actin and neurofilament proteins as well as the nociceptive molecules PKCε and CamKII. The C-terminus of TRPV4 is sufficient for the direct interaction with tubulin and actin, both with their soluble and their polymeric forms. Actin and tubulin compete for binding. The interaction with TRPV4 stabilizes microtubules even under depolymerizing conditions in vitro. Accordingly, in cellular systems TRPV4 colocalizes with actin and microtubules enriched structures at submembranous regions. Both expression and activation of TRPV4 induces striking morphological changes affecting lamellipodial, filopodial, growth cone, and neurite structures in non-neuronal cells, in DRG-neuron derived F11 cells, and also in IB4-positive DRG neurons. The functional interaction of TRPV4 and the cytoskeleton is mutual as Taxol, a microtubule stabilizer, reduces the Ca²⁺-influx via TRPV4.

Conclusions and Significance

TRPV4 acts as a regulator for both, the microtubule and the actin. In turn, we describe that microtubule dynamics are an important regulator of TRPV4 activity. TRPV4 forms a supra-molecular complex containing cytoskeletal proteins and regulatory kinases. Thereby it can integrate signaling of various intracellular second messengers and signaling cascades, as well as cytoskeletal dynamics. This study points out the existence of cross-talks between non-selective cation channels and cytoskeleton at multiple levels. These cross talks may help us to understand the molecular basis of the Taxol-induced neuropathic pain development commonly observed in cancer patients.     

Intact microtubules preserve transient receptor potential vanilloid 1 (TRPV1) functionality through receptor binding

The transient receptor potential cation channel subfamily V member 1 (TRPV1) is a protein currently under scrutiny as a pharmacological target for pain management therapies. Recently, the role of TRPV1-microtubule interaction in transducing nociception stimuli to cells by cytoskeletal rearrangement was proposed. In this work, we investigate TRPV1-microtubule interaction in living cells under the resting or activated state of TRPV1, as well as in presence of structurally intact or depolymerized cytoskeletal microtubules. We combined a toolbox of high resolution/high sensitivity fluorescence imaging techniques (such as FRET, correlation spectroscopy, and fluorescence anisotropy) to monitor TRPV1 aggregation status, membrane mobility, and interaction with microtubules. We found that TRPV1 is a dimeric membrane protein characterized by two populations with different diffusion properties in basal condition. After stimulation with resiniferatoxin, TRPV1 dimers tetramerize. The tetramers and the slower population of TRPV1 dimers bind dynamically to intact microtubules but not to tubulin dimers. Upon microtubule disassembly, the interaction with TRPV1 is lost thereby inducing receptor self-aggregation with partial loss of functionality. Intact microtubules play an essential role in maintaining TRPV1 functionality toward activation stimuli. This previously undisclosed property mirrors the recently reported role of TRPV1 in modulating microtubule assembly/disassembly and suggests the participation of these two players in a feedback cycle linking nociception and cytoskeletal remodeling.     

Structural and functional differences between porcine brain and budding yeast microtubules

The cytoskeleton of eukaryotic cells relies on microtubules to perform many essential functions. We have previously shown that, in spite of the overall conservation in sequence and structure of tubulin subunits across species, there are differences between mammalian and budding yeast microtubules with likely functional consequences for the cell. Here we expand our structural and function comparison of yeast and porcine microtubules to show different distribution of protofilament number in microtubules assembled in vitro from these two species. The different geometry at lateral contacts between protofilaments is likely due to a more polar interface in yeast. We also find that yeast tubulin forms longer and less curved oligomers in solution, suggesting stronger tubulin:tubulin interactions along the protofilament. Finally, we observed species-specific plus-end tracking activity for EB proteins: yeast Bim1 tracked yeast but not mammalian MTs, and human EB1 tracked mammalian but not yeast MTs. These findings further demonstrate that subtle sequence differences in tubulin sequence can have significant structural and functional consequences in microtubule structure and behavior.     

TRPV1 at nerve endings regulates growth cone morphology and movement through cytoskeleton reorganization

While the importance of Ca(2+) channel activity in axonal path finding is established, the underlying mechanisms are not clear. Here, we show that transient receptor potential vanilloid receptor 1 (TRPV1), a member of the TRP superfamily of nonspecific ion channels, is physically and functionally present at dynamic neuronal extensions, including growth cones. These nonselective cation channels sense exogenous ligands, such as resenifera toxin, and endogenous ligands, such as N-arachidonoyl-dopamine (NADA), and affect the integrity of microtubule cytoskeleton. Using TRPV1-transiently transfected F11 cells and embryonic dorsal root ganglia explants, we show that activation of TRPV1 results in growth cone retraction, and collapse and formation of varicosities along neurites. These changes were due to TRPV1-activation-mediated disassembly of microtubules and are partly Ca(2+)-independent. Prolonged activation with very low doses (1 nM) of NADA results in shortening of neurites in the majority of isolectin B4-positive dorsal root ganglia neurones. We postulate that TRPV1 activation plays an inhibitory role in sensory neuronal extension and motility by regulating the disassembly of microtubules. This might have a role in the chronification of pain.     

TRPV1-tubulin complex: involvement of membrane tubulin in the regulation of chemotherapy-induced peripheral neuropathy

Existence of microtubule cytoskeleton at the membrane and submembranous regions, referred as 'membrane tubulin' has remained controversial for a long time. Since we reported physical and functional interaction of Transient Receptor Potential Vanilloid Sub Type 1 (TRPV1) with microtubules and linked the importance of TRPV1-tubulin complex in the context of chemotherapy-induced peripheral neuropathy, a few more reports have characterized this interaction in in vitro and in in vivo condition. However, the cross-talk between TRPs with microtubule cytoskeleton, and the complex feedback regulations are not well understood. Sequence analysis suggests that other than TRPV1, few TRPs can potentially interact with microtubules. The microtubule interaction with TRPs has evolutionary origin and has a functional significance. Biochemical evidence, Fluorescence Resonance Energy Transfer analysis along with correlation spectroscopy and fluorescence anisotropy measurements have confirmed that TRPV1 interacts with microtubules in live cell and this interaction has regulatory roles. Apart from the transport of TRPs and maintaining the cellular structure, microtubules regulate signaling and functionality of TRPs at the single channel level. Thus, TRPV1-tubulin interaction sets a stage where concept and parameters of 'membrane tubulin' can be tested in more details. In this review, I critically analyze the advancements made in biochemical, pharmacological, behavioral as well as cell-biological observations and summarize the limitations that need to be overcome in the future.     

Identification and characterisation of novel tubulin-binding motifs located within the C-terminus of TRPV1

Previously, we reported that TRPV1, the vanilloid receptor, interacts with soluble alphabeta-tubulin dimers as well as microtubules via its C-terminal cytoplasmic domain. The interacting region of TRPV1, however, has not been defined. We found that the TRPV1 C-terminus preferably interacts with beta-tubulin and less with alpha-tubulin. Using a systematic deletion approach and biotinylated-peptides we identified two tubulin-binding sites present in TRPV1. These two sequence stretches are highly conserved in all known mammalian TRPV1 orthologues and partially conserved in some of the TRPV1 homologues. As these sequence stretches are not similar to any known tubulin-binding sequences, we conclude that TRPV1 interacts with tubulin and microtubule through two novel tubulin-binding motifs.     

The parkinsonism producing neurotoxin MPP+ affects microtubule dynamics by acting as a destabilising factor

Dysfunction of the microtubule system is emerging as a contributing factor in a number of neurodegenerative diseases. Looking for the potential role played by the microtubule cytoskeleton in neuron degeneration underlying Parkinson's disease (PD), we investigate the influence of the parkinsonism producing neurotoxin 1-methyl-4-phenylpyridinium (MPP+) on microtubule dynamics. We find that it acts as a strong catastrophe promoter causing a decrease of the average length of microtubules assembled from purified tubulin. We also find that it reduces the number of microtubules nucleated from purified centrosomes. Finally, binding assays demonstrate that the neurotoxin binds specifically to tubulin in the microtubule lattice in a close to stoichiometric manner. This paper provides the first evidence that dynamic instability of microtubules is specifically affected by MPP+ and suggests that it could play a role in neuronal cell death underlying PD.     

Estrogen destabilizes microtubules through an ion-conductivity-independent TRPV1 pathway

Recently, we described estrogen and agonists of the G-protein coupled estrogen receptor GPR30 to induce protein kinase C (PKC)ε-dependent pain sensitization. PKCε phosphorylates the ion channel transient receptor potential, vanilloid subclass I (TRPV1) close to a novel microtubule-TRPV1 binding site. We now modeled the binding of tubulin to the TRPV1 C-terminus. The model suggests PKCε phosphorylation of TRPV1-S800 to abolish the tubulin-TRPV1 interaction. Indeed, in vitro PKCε phosphorylation of TRPV1 hindered tubulin-binding to TRPV1. In vivo, treatment of sensory neurons and F-11 cells with estrogen and the GPR30 agonist, G-1, resulted in microtubule destabilization and retraction of microtubules from filopodial structures. We found estrogen and G-1 to regulate the stability of the microtubular network via PKC phosphorylation of the PKCε-phosphorylation site TRPV1-S800. Microtubule disassembly was not, however, dependent on TRPV1 ion conductivity. TRPV1 knock-down in rats inverted the effect of the microtubule-modulating drugs, Taxol and Nocodazole, on estrogen-induced and PKCε-dependent mechanical pain sensitization. Thus, we suggest the C-terminus of TRPV1 to be a signaling intermediate downstream of estrogen and PKCε, regulating microtubule-stability and microtubule-dependent pain sensitization.     

The phosphorylation of p25/TPPP by LIM kinase 1 inhibits its ability to assemble microtubules

LIM kinase 1 (LIMK1) is a key regulator of actin dynamics as it phosphorylates and inactivates cofilin, an actin-depolymerizing factor. LIMK1 activity is also required for microtubule disassembly in endothelial cells. A search for LIMK1-interacting proteins identified p25alpha, a phosphoprotein that promotes tubulin polymerization. We found that p25 is phosphorylated by LIMK1 on serine residues in vitro and in cells. Immunoblotting analysis revealed that p25 is not a brain specific protein as previously reported, but is expressed in all mouse tissues. Immunofluorescence analysis demonstrated that endogenous p25 is co-localized with microtubules and is also found in the nucleus. Down-regulation of p25 by siRNA decreased microtubule levels while its overexpression in stable NIH-3T3 cell lines increased cell size and levels of stable tubulin. Bacterially expressed unphosphorylated p25 promotes microtubule assembly in vitro; however, when phosphorylated in cells, p25 lost its ability to assemble microtubule. Our results represent a surprising connection between the tubulin and the actin cytoskeleton mediated by LIMK1. We propose that the LIMK1 phosphorylation of p25 blocks p25 activity, thus promoting microtubule disassembly.     

Rotavirus Infection of Cells in Culture Induces Activation of RhoA and Changes in the Actin and Tubulin Cytoskeleton

Rotavirus infection induces an increase in [Ca²⁺]_cyto, which in turn may affect the distribution of the cytoskeleton proteins in the infected cell. Changes in microfilaments, including the formation of stress fibers, were observed starting at 0.5 h.p.i. using fluorescent phalloidin. Western blot analysis indicated that RhoA is activated between 0.5 and 1 h.p.i. Neither the phosphorylation of RhoA nor the formation of stress fibers were observed in cells infected with virions pre-treated with an anti-VP5* non-neutralizing mAb, suggesting that RhoA activation is stimulated by the interaction of the virus with integrins forming the cell receptor complex. In addition, the structure of the tubulin cytoskeleton was also studied. Alterations of the microtubules were evident starting at 3 h.p.i. and by 7 h.p.i. when microtubules were markedly displaced toward the periphery of the cell cytoplasm. Loading of rotavirus-infected cells with either a Ca²⁺ chelator (BAPTA) or transfection with siRNAs to silence NSP4, reversed the changes observed in both the microfilaments and microtubules distribution, but not the appearance of stress fibers. These results indicate that alterations in the distribution of actin microfilaments are initiated early during infection by the activation of RhoA, and that latter changes in the Ca²⁺ homeostasis promoted by NSP4 during infection may be responsible for other alterations in the actin and tubulin cytoskeleton.     

During Drosophila spermatogenesis beta 1, beta 2 and beta 3 tubulin isotypes are cell-type specifically expressed but have the potential to coassemble into the axoneme of transgenic flies

alpha and beta Tubulins exist in a number of different isotypes with distinct expression patterns during development. We have shown by immunofluorescent staining that beta 1, beta 2 and beta 3 tubulins are distributed very specifically in the testes of Drosophila. beta 3 Tubulin is present exclusively in cytoplasmic microtubules of cells somatic in origin, while the beta 1 isotype is localized in the somatic cells and in early germ cells of both the microtubules of the cytoskeleton as well as in the mitotic spindle. In contrast, beta 2 tubulin is present in all microtubular arrays (cytoskeleton, meiotic spindles, axoneme) of germ cells from meiotic prophase onward, though not detectable in somatic cells. Thus, a switch of beta tubulin isotypes from beta 1 to beta 2 occurs during male germ cell differentiation. This switch is also observed in the distantly related species Drosophila hydei. By fusing beta 1 or beta 3 amino acid coding regions to the control region of the beta 2 tubulin gene and performing germ line transformation experiments, we have examined the copolymerization properties of the different tubulin isotypes. Neither beta 1 nor beta 3 are detectable in the axoneme in the wild-type situation. Analysis of transgenic flies carrying beta 2-beta 1 fusion genes or beta 2-beta 3 fusion genes revealed that both beta 1 and beta 3 tubulin isotypes have the potential to co-incorporate with beta 2 tubulin into microtubules of the sperm axoneme. Male flies homozygous for the fusion genes (beta 2-beta 1 or beta 2-beta 3) remain fertile, despite the mixture of beta tubulin isotypes in the axoneme.     

A fluorescent GTP analog as a specific, high-precision label of microtubules

Fluorescent imaging of cytoskeletal structures permits studies of both organization within the cell and dynamic reorganization of the cytoskeleton itself. Traditional fluorescent labels of microtubules, part of the cytoskeleton, have been used to study microtubule localization, structure, and dynamics, both in vivo and in vitro. However, shortcomings of existing labels make imaging of microtubules with high precision light microscopy difficult. In this paper, we report a new fluorescent labeling technique for microtubules, which involves a GTP analog modified with a bright, organic fluorophore (TAMRA, Cy3, or Cy5). This fluorescent GTP binds to a specific site, the exchangeable site, on tubulin in solution with a dissociation constant of 1.0±0.4 μM. Furthermore, the label becomes permanently incorporated into the microtubule lattice once tubulin polymerizes. We show that this label is usable as a single molecule fluorescence probe with nanometer precision and expect it to be useful for modern subdiffraction optical microscopy of microtubules and the cytoskeleton.     

Formin-1 protein associates with microtubules through a peptide domain encoded by exon-2

Formin family proteins coordinate actin filaments and microtubules. The mechanisms by which formins bind and regulate the actin cytoskeleton have recently been well defined. However, the molecular mechanism by which formins coordinate actin filaments and microtubules remains poorly understood. We demonstrate here that Isoform-Ib of the Formin-1 protein (Fmn1-Ib) binds to microtubules via a protein domain that is physically separated from the known actin-binding domains. When expressed at low levels in NIH3T3 fibroblasts, Fmn1-Ib protein localizes to cytoplasmic filaments that nocodazole disruption confirmed as interphase microtubules. A series of progressive mutants of Fmn1-Ib demonstrated that deletion of exon-2 caused dissociation from microtubules and a stronger association with actin membrane ruffles. The exon-2-encoded peptide binds purified tubulin in vitro and is also sufficient to localize GFP to microtubules. Exon-2 does not contain any known formin homology domains. Deletion of exon 5, 7, 8, the FH1 domain or FH2 domain did not affect microtubule binding. Thus, our results indicate that exon-2 of Fmn1-Ib encodes a novel microtubule-binding peptide. Since formin proteins associate with actin filaments through the FH1 and FH2 domains, binding to interphase microtubules through this exon-2-encoded domain provides a novel mechanism by which Fmn1-Ib could coordinate actin filaments and microtubules.     

Computational predictions of volatile anesthetic interactions with the microtubule cytoskeleton: implications for side effects of general anesthesia

The cytoskeleton is essential to cell morphology, cargo trafficking, and cell division. As the neuronal cytoskeleton is extremely complex, it is no wonder that a startling number of neurodegenerative disorders (including but not limited to Alzheimer's disease, Parkinson's disease and Huntington's disease) share the common feature of a dysfunctional neuronal cytoskeleton. Recently, concern has been raised about a possible link between anesthesia, post-operative cognitive dysfunction, and the exacerbation of neurodegenerative disorders. Experimental investigations suggest that anesthetics bind to and affect cytoskeletal microtubules, and that anesthesia-related cognitive dysfunction involves microtubule instability, hyper-phosphorylation of the microtubule-associated protein tau, and tau separation from microtubules. However, exact mechanisms are yet to be identified. In this paper the interaction of anesthetics with the microtubule subunit protein tubulin is investigated using computer-modeling methods. Homology modeling, molecular dynamics simulations and surface geometry techniques were used to determine putative binding sites for volatile anesthetics on tubulin. This was followed by free energy based docking calculations for halothane (2-bromo-2-chloro-1,1,1-trifluoroethane) on the tubulin body, and C-terminal regions for specific tubulin isotypes. Locations of the putative binding sites, halothane binding energies and the relation to cytoskeleton function are reported in this paper.     

Interplay between exchange protein directly activated by cAMP (Epac) and microtubule cytoskeleton

"Exchange protein directly activated by cAMP" (Epac) is a newly discovered cAMP receptor that mediates the intracellular cAMP actions in addition to the classic cAMP-dependent protein kinase system. In this study, we show that Epac interacts directly with tubulin, co-purifies with cellular microtubules, and co-localizes with the mitotic spindle assembly. Association with microtubules suppresses Epac-mediated Rap1 activation, while the binding of Epac promotes microtubule formation. These results demonstrate that Epac plays an important role in connecting the microtubule cytoskeleton network and intracellular cAMP-signalling.     

XKCM1 acts on a single protofilament and requires the C terminus of tubulin

The stability of microtubules during the cell-cycle is regulated by a number of cellular factors, some of which stabilize microtubules and others that promote breakdown. XKCM1 is a kinesin-like protein that induces microtubule depolymerization and is required for mitotic spindle assembly. We have examined the binding and depolymerization effects of XKCM1 on different tubulin polymers in order to learn about its mechanism of action. Zinc-induced tubulin polymers, characterized by an anti-parallel protofilament arrangement, are depolymerized by XKCM1, indicating that this enzyme acts on a single protofilament. GDP-tubulin rings, which correspond to the low-energy state of tubulin, are stable only under conditions that inhibit XKCM1 depolymerizing activity, but can be stabilized by XKCM1 bound to AMPPNP. Tubulin polymers made of subtilisin-treated tubulin (lacking the tubulin C-terminal tail) are resistant to XKCM1-induced depolymerization, suggesting that the interaction of the acidic tail of tubulin with basic residues in XKCM1 unique to Kin I proteins is required for depolymerization.     

Subpellicular and flagellar microtubules of Trypanosoma brucei brucei contain the same alpha-tubulin isoforms

The cytoskeleton of the parasitic hemoflagellate Trypanosoma brucei brucei essentially consists of two microtubule-based structures: a subpellicular layer of singlet microtubules, which are in close contact with the cell membrane, and the flagellar axoneme. In addition, the cells contain a small pool of soluble tubulin. Two-dimensional gel electrophoretic analysis of the tubulins present in these subcellular compartments revealed two distinct electrophoretic isoforms of alpha-tubulin, termed alpha 1 and alpha 3. alpha 1-Tubulin most likely represents the primary translation product, while alpha 3-tubulin is a posttranslationally acetylated derivative of alpha 1-tubulin. In the pool of soluble cytoplasmic tubulin, alpha 1 is the predominant species, while the very stable flagellar microtubules contain almost exclusively the alpha 3-tubulin isoform. The subpellicular microtubules contain both isoforms. Neither of the two alpha-tubulin isoforms is organelle specific, but the alpha 3 isoform is predominantly located in stable microtubules.     

Distinct roles of α‐ and β‐tubulin polyglutamylation in controlling axonal transport and in neurodegeneration

Tubulin polyglutamylation is a post‐translational modification of the microtubule cytoskeleton, which is generated by a variety of enzymes with different specificities. The “tubulin code” hypothesis predicts that modifications generated by specific enzymes selectively control microtubule functions. Our recent finding that excessive accumulation of polyglutamylation in neurons causes their degeneration and perturbs axonal transport provides an opportunity for testing this hypothesis. By developing novel mouse models and a new glutamylation‐specific antibody, we demonstrate here that the glutamylases TTLL1 and TTLL7 generate unique and distinct glutamylation patterns on neuronal microtubules. We find that under physiological conditions, TTLL1 polyglutamylates α‐tubulin, while TTLL7 modifies β‐tubulin. TTLL1, but not TTLL7, catalyses the excessive hyperglutamylation found in mice lacking the deglutamylase CCP1. Consequently, deletion of TTLL1, but not of TTLL7, prevents degeneration of Purkinje cells and of myelinated axons in peripheral nerves in these mice. Moreover, loss of TTLL1 leads to increased mitochondria motility in neurons, while loss of TTLL7 has no such effect. By revealing how specific patterns of tubulin glutamylation, generated by distinct enzymes, translate into specific physiological and pathological readouts, we demonstrate the relevance of the tubulin code for homeostasis.     

KT5720 and U-98017 inhibit MAPK and alter the cytoskeleton and cell morphology

We previously identified KT5720 and U-98017 as agents that had paclitaxel (taxol)-like activity in a Chinese hamster ovary (CHO) paclitaxel-dependent cell screen for paclitaxel mimetics. In vitro polymerization of purified brain tubulin is not affected substantially by these compounds, suggesting that, unlike paclitaxel, these agents do not directly affect tubulin. However, these compounds cause profound rearrangements of the cytoskeleton in intact cells, including an apparent alteration of microtubule length, overlapping of cells, and an increase in cell size. We show that KT5720 and U-98017 effectively inhibit mitogen-activated protein kinase (MAPK) activity in vitro. Staurosporine, a poor inhibitor of MAPK but a potent inhibitor of cAMP-dependent protein kinase A (PKA) activity, phospholipid/Ca⁺⁺-dependent kinase (PKC), and cdc2, does not cause similar changes. In addition, paclitaxel-dependent cells grown in U-98017 have substantially decreased levels of stimulated MAPK. In correlation with these results, we have confirmed the presence of MAPK in isolated tubulin and microtubules in cells. We have examined the hypothesis that these compounds are working through inhibition of MAPK to alter microtubules by inhibiting the phosphorylation of microtubule-associated proteins. A MAPKK dominant negative mutation transfected in CHO cells inhibits activation of MAPK. Transfectants carrying this dominant mutant have impaired activation of MAPK and an altered cell morphology, similar in some respects to that seen with KT5720 and U-98017. These results support a role for MAPK family members in the control of microtubule dynamics and suggest that in intact cells U-98017 and KT5720 achieve their effects of altering cytoskeleton and supporting partial growth of paclitaxel-dependent cells through inhibition of kinases such as MAPK. J. Cell. Physiol. 176:525–536, 1998. © 1998 Wiley-Liss, Inc.