Differential stabilization of two hydrophobic cores in the transition state of the villin 14T folding reaction.

We report the distribution of hydrophobic core contacts during the folding reaction transition state for villin 14T, a small 126-residue protein domain. The solution structure of villin 14T contains a central beta-sheet with two flanking hydrophobic cores; transition states for this protein topology have not been previously studied. Villin 14T has no disulfide bonds or cis-proline residues in its native state; it folds reversibly, and in an apparently two-state manner under some conditions. To map the hydrophobic core contacts in the transition state, 27 point mutations were generated at positions spread throughout the two hydrophobic cores. After each point mutation, comparison of the change in folding kinetics with the equilibrium destabilization indicates whether the site of mutation is stabilized in the transition state. The results show that the folding nucleus, or the sub-region with the strongest transition state contacts, is located in one of the two hydrophobic cores (the predominantly aliphatic core). The other hydrophobic core, which is mostly aromatic, makes much weaker contacts in the transition state. This work is the first transition state mapping for a protein with multiple major hydrophobic cores in a single folding unit; the hydrophobic cores cannot be separated into individual folding subdomains. The stabilization of only one hydrophobic core in the transition state illustrates that hydrophobic core formation is not intrinsically capable of nucleating folding, but must also involve the right specific interactions or topological factors in order to be kinetically important.

Folding kinetics of villin 14T, a protein domain with a central beta-sheet and two hydrophobic cores.

The thermodynamics and kinetics of folding are characterized for villin 14T, a 126-residue protein domain. Equilibrium fluorescence measurements reveal that villin 14T unfolds and refolds reversibly. The folding kinetics was monitored using stopped-flow with fluorescence and quenched-flow with NMR and mass spectrometry. Unfolding occurs in a single-exponential phase in the stopped-flow experiments, and about 75% of the total amplitude is recovered in the fast phase of refolding. The remaining 25% of the amplitude probably represents trapping in cis-trans proline isomerization pathways. At 25 degreesC, the stability estimate obtained by extrapolation from the transition region of the stopped-flow chevron matches the stability value from equilibrium urea titrations (DeltaG = 9.7 kcal/mol, m value = 2.2 kcal mol-1 M-1). At low final urea concentrations, however, the refolding kinetics deviates from the two-state model, indicating the formation of an intermediate. Under these conditions, quenched-flow followed by NMR and mass spectrometry show no detectable hydrogen-bonded intermediate in the fast refolding phase. In contrast, agreement is observed between the equilibrium and kinetic estimates of stability at 37 degreesC (DeltaG = 6.0 kcal/mol, m value = 1.6 kcal mol-1 M-1), at all observed urea concentrations, demonstrating apparent two-state folding at this temperature. This result shows that the two-state folding model, previously applied to small domains with single, central hydrophobic cores, can also describe the folding of a larger domain with multiple core structures.

The amino acid sequence of the light chain of Acanthamoeba myosin IC.

The amino acid sequence of the light chain of Acanthamoeba myosin IC deduced from the cDNA sequence comprises 149 amino acids with a calculated molecular weight of 16,739. All but the 3 N-terminal residues were also determined by amino acid sequencing of the purified protein, which also showed the N-terminus to be blocked. Phylogenetic analysis shows Acanthamoeba myosin IC light chain to be more similar to the calmodulin subfamily of EF-hand calcium-modulated proteins than to the myosin II essential light chain or regulatory light chain subfamilies. In pairwise comparisons, the myosin IC light chain sequence is most similar to sequences of calmodulins (approximately 50% identical) and a squid calcium-binding protein (approximately 43% identical); the sequence is approximately 37% identical to the calcium-binding essential light chain of Physarum myosin II and approximately 30% identical to the essential light chain of Acanthamoeba myosin II, and the essential light chain and regulatory light chain of Dictyostelium myosin II. The sequence predicts four helix-loop-helix domains with possible calcium-binding sites in domains I and III, suggesting that calcium may affect the activity of this unconventional myosin. This is the first report of the sequence of an unconventional myosin light chain other than calmodulin.

NMR structure of the 35-residue villin headpiece subdomain.

The NMR structure of an autonomously folding subdomain from villin headpiece is reported. It forms a novel three helix structure with the actin-binding residues arrayed on the C-terminal helix.

Identification of an actin binding region and a protein kinase C phosphorylation site on human fascin.

Fascin is a 55-58-kDa actin-bundling protein, the actin binding of which is regulated by phosphorylation (Yamakita, Y., Ono, S., Matsumura, F., and Yamashiro, S. (1996) J. Biol. Chem. 271, 12632-12638). To understand the mechanism of fascin-actin interactions, we dissected the actin binding region and its regulatory site by phosphorylation of human fascin. First, we found that the C-terminal half constitutes an actin binding domain. Partial digestion of human recombinant fascin with trypsin yielded the C-terminal fragment with molecular masses of 32, 30, and 27 kDa. The 32- and 27-kDa fragments purified as a mixture formed a dimer and bound to F-actin at a saturation ratio of 1 dimer:11 actin molecules with an affinity of 1.4 x 10(6) M-1. Second, we identified the phosphorylation site of fascin as Ser-39 by sequencing a tryptic phosphopeptide purified by chelating column chromatography followed by C-18 reverse phase high performance liquid chromatography. Peptide map analyses revealed that the purified peptide represented the major phosphorylation site of in vivo as well as in vitro phosphorylated fascin. The mutation replacing Ser-39 with Ala eliminated the phosphorylation-dependent regulation of actin binding of fascin, indicating that phosphorylation at this site regulates the actin binding ability of fascin.

A thermostable 35-residue subdomain within villin headpiece.

The actin-bundling protein villin contains, at its extreme C terminus, a compact f-actin binding domain called "headpiece". This 76-amino acid domain from chicken is highly thermostable. Here, we show that the stable folded structure in headpiece is localized to a subdomain formed by the C-terminal 35 residues. The subdomain, denoted HP-35, is monomeric and retains high thermostability, with a Tm of 70( +/- 1) degree C at PH 7.0. There are no cysteine residues in HP-35 and its folding is not dependent on the binding of metals or other ligands. HP-35 is not a molten globule, but instead, has properties expected for a fully folded protein with a unique structure. In particular, the slowly exchanging amide protons in HP-35 have protection factors that are slightly larger than those predicted if exchange occurred only from globally unfolded molecules. NMR studies indicate that the headpiece subdomain contains three short alpha-helices, and that these same helices are present in the corresponding regions of intact headpiece. HP-35 is the smallest monomeric polypeptide characterized consisting of only naturally occurring amino acids that autonomously folds into a unique and thermostable structure without disulfide bonds or ligand binding.

Recombinant expression of the brush border myosin I heavy chain.

Although the specific functions of myosin I motors are not known, their localization to membrane structures suggests a function in membrane motility. Different myosin I isoforms in the same cell or in different cells can possess different localizations. To determine if the localization and biochemical activity of the best-characterized mammalian myosin I, chicken intestinal epithelium brush border myosin I, was dependent on determinants of the membrane or actin cytoskeleton specific to epithelial cells, we transfected the cDNA for the heavy chain of this myosin into COS cells. Transient transfection of COS cells with the chicken brush border myosin heavy chain resulted in the production of recombinant myosin I. Recombinant brush border myosin I localized to protrusions of the plasma membrane, particularly at spreading edges, and also to unknown cytoplasmic structures. Some cells expressing particularly high levels of brush border myosin I possessed a highly irregular surface. Recombinant brush border myosin I purified from COS cells bound to actin filaments in an ATP-dependent manner and decorated actin filaments to form a characteristic appearance. The recombinant myosin also catalyzed calcium-sensitive, actin-activated MgATPase activity similar to that of the native enzyme. Thus, any cellular factor required for the general membrane localization or biochemical activity of brush border myosin I is present in COS cells as well as intestinal epithelium.

The COOH terminus of the c-Abl tyrosine kinase contains distinct F- and G-actin binding domains with bundling activity.

The myristoylated form of c-Abl protein, as well as the P210bcr/abl protein, have been shown by indirect immunofluorescence to associate with F-actin stress fibers in fibroblasts. Analysis of deletion mutants of c-Abl stably expressed in fibroblasts maps the domain responsible for this interaction to the extreme COOH-terminus of Abl. This domain mediates the association of a heterologous protein with F-actin filaments after microinjection into NIH 3T3 cells, and directly binds to F-actin in a cosedimentation assay. Microinjection and cosedimentation assays localize the actin-binding domain to a 58 amino acid region, including a charged motif at the extreme COOH-terminus that is important for efficient binding. F-actin binding by Abl is calcium independent, and Abl competes with gelsolin for binding to F-actin. In addition to the F-actin binding domain, the COOH-terminus of Abl contains a proline-rich region that mediates binding and sequestration of G-actin, and the Abl F- and G-actin binding domains cooperate to bundle F-actin filaments in vitro. The COOH terminus of Abl thus confers several novel localizing functions upon the protein, including actin binding, nuclear localization, and DNA binding. Abl may modify and receive signals from the F-actin cytoskeleton in vivo, and is an ideal candidate to mediate signal transduction from the cell surface and cytoskeleton to the nucleus.

Characterisation of the F-actin binding domains of villin: classification of F-actin binding proteins into two groups according to their binding sites on actin.

The F-actin binding properties of chicken villin, its headpiece and domains 2-3 (V2-3) have been analysed to identify sites involved in bundle formation. Headpiece and V2-3 bind actin with Kd values of approximately 7 microM and approximately 0.3 microM, respectively, at low ionic strength. V2-3 binding, like that of villin, is weakened with increasing salt concentration; headpiece binding is not. Competition experiments show that headpiece and V2-3 bind to different sites on actin, forming the two cross-linking sites of villin. Headpiece does not compete with the F-actin binding domains of gelsolin or alpha-actinin, but it dissociates actin depolymerizing factor. We suggest that the F-actin binding domains of actin severing, crosslinking and capping proteins can be organized into two classes.

Relative distribution of actin, myosin I, and myosin II during the wound healing response of fibroblasts.

Myosin I is present in Swiss 3T3 fibroblasts and its localization reflects a possible involvement in the extension and/or retraction of protrusions at the leading edge of locomoting cells and the transport of vesicles, but not in the contraction of stress fibers or transverse fibers. An affinity-purified polyclonal antibody to brush border myosin I colocalizes with a polypeptide of 120 kD in fibroblast extracts. Within initial protrusions of polarized, migrating fibroblasts, myosin I exhibits a punctate distribution, whereas actin is diffuse and myosin II is absent. Myosin I also exists in linear arrays parallel to the direction of migration in filopodia and microspikes, established protrusions, and within the leading lamellae of migrating cells. Myosin II and actin colocalize along transverse fibers in the lamellae of migrating cells, while myosin I displays no definitive organization along these fibers. During contractions of actin-based fibers, myosin II is concentrated in the center of the cell, while the distribution of myosin I does not change. Thus, myosin I is found at the correct location and time to be involved in the extension and/or retraction of protrusions and the transport of vesicles. Myosin II-based contractions in more posterior cellular regions could generate forces to separate cells, maintain a polarized cell shape, maintain the direction of locomotion, maximize the rate of locomotion, and/or aid in the delivery of cytoskeletal/contractile subunits to the leading edge.

Expression and localization of villin, fimbrin, and myosin I in differentiating mouse F9 teratocarcinoma cells.

F9 embryonic carcinoma cells are a multipotent cell line which can be induced to differentiate into cells resembling the visceral endoderm, an extraembryonic absorptive epithelium characterized by apical microvilli. We have examined the role of villin, fimbrin, and myosin I, the major actin-binding proteins in the intestinal and visceral yolk sac microvilli, in the development of epithelial polarity and the assembly of the microvillus cytoskeleton in differentiating F9 cells. By immunoblot analysis villin was first detected at 4 days of differentiation. Confocal microscopy localized villin at Day 4 to the apical surface and by Day 6 to the basolateral surfaces as well. In comparison, fimbrin and myosin I were both present in undifferentiated F9 cells and became associated with the apical surface after villin during differentiation to visceral endoderm. The accumulation of villin, fimbrin, and myosin I at the apical surface in differentiating F9 cells correlated with the appearance of microvilli containing organized actin filament bundles. Two mouse villin cDNAs were isolated and characterized to examine villin expression during F9 differentiation. Mouse villin was encoded by two transcripts (3.8 and 3.4 kb) which differ in their 3'-noncoding region. Both villin mRNAs were first detected by Day 4 of differentiation and their appearance coincided with expression of the visceral endoderm marker alpha-fetoprotein. The pattern of expression and order of accumulation of villin, fimbrin, and myosin I in differentiating F9 cells are common to developing gut and yolk sac epithelium. This suggests that microvillus assembly is directed by a sequence of temporally and spatially regulated localizations of these actin-binding proteins.

Phosphoinositide-binding peptides derived from the sequences of gelsolin and villin.

The polyphosphoinositides phosphatidylinositol 4-monophosphate (PIP) and phosphatidylinositol 4,5-bisphosphate (PIP2) inactivate the actin filament-severing proteins villin and gelsolin and dissociate them from monomeric and polymeric actin. A potential polyphosphoinositide- (PPI) binding site of human plasma gelsolin regulating filament severing has been localized to the region between residues 150-169 and to the corresponding region in villin which occurs in the second of six homologous domains present in both proteins. Synthetic peptides based on these sequences bind tightly to both PIP and PIP2, in either micelles or bilayer vesicles, compete with gelsolin for binding to PPIs, and dissociate gelsolin-PIP2 complexes, restoring severing activity to the protein. These peptides also bind with moderate affinity to F-actin, suggesting that inactivation of the severing function of the intact proteins by PPIs results from competition between actin and PPIs for a critical binding site on gelsolin-villin. The PPI-binding peptides contain numerous basic amino acids, but their effects on PPIs are far greater than those of Arg or Lys oligomers, a highly basic peptide derived from the calmodulin-binding site of myristoylated, alanine-rich kinase C substrate protein, or the 5-kDa actin-binding protein thymosin beta-4, suggesting that specific aspects of the primary and secondary structure of these basic peptides are important for their interaction with the acidic headgroups of PPIs. In addition to elucidating the structure of PIP2-binding sites in gelsolin, the results describe a sensitive assay for phosphoinositide-binding molecules based on their ability to prevent inhibition of gelsolin function.

Coordinate expression of lactase-phlorizin hydrolase mRNA and enzyme levels in rat intestine during development.

The development of rat intestinal lactase-specific activity displays a well-known post-weaning decline. In contrast, total lactase activity increases to reach maximal levels around weaning, and remains high subsequently. In order to elucidate the molecular basis for these patterns, a rat lactase cDNA was isolated and characterized, and used in the quantification of lactase mRNA during development. This lactase cDNA uniquely hybridized to a 6.8-kilobase mRNA in the small intestine. To assess the amount of lactase mRNA encoding for lactase enzyme activity in the small intestine, total intestinal RNA was isolated and analyzed by Northern and dot-blot hybridization. The pattern of total lactase mRNA during development followed that of total lactase activity, suggesting that over this time span the level of lactase activity is primarily controlled at the transcriptional level. However, the magnitude of increase of total lactase activity during lactation compared to that of total lactase mRNA suggests that additional mechanisms are involved in regulating lactase levels. Analysis of the regional distribution of lactase mRNA along the small intestine at 14 days revealed that mRNA was high in the proximal three regions, but was dramatically lower in the distal regions. Total lactase activity, in contrast, displayed maximum activity in the mid-intestine with decreased levels both proximally and distally. Thus, lactase activity in the intestine appears to be regulated during development predominantly by transcriptional mechanisms, while alterations during lactation, and along the proximal to distal gradient, are the result of other control mechanisms.

Differential localization of villin and fimbrin during development of the mouse visceral endoderm and intestinal epithelium.

The apical surface of transporting epithelia is specially modified to absorb nutrients efficiently by amplifying its surface area as microvilli. Each microvillus is supported by an underlying core of bundled actin filaments. Villin and fimbrin are two actin-binding proteins that bundle actin filaments in the intestine and kidney brush border epithelium. To better understand their function in the assembly of the cytoskeleton during epithelial differentiation, we examined the pattern of villin and fimbrin expression in the developing mouse using immunofluorescence and immunoelectron microscopy. Villin is first detected at day 5 in the primitive endoderm of the postimplantation embryo and is later restricted to the visceral endoderm. By day 8.5, villin becomes redistributed to the apical surface in the visceral endoderm, appearing in the gut at day 10 and concentrating in the apical cytoplasm of the differentiating intestinal epithelium 2-3 days later. In contrast, fimbrin is found in the oocyte and in all tissues of the early embryo. In both the visceral endoderm and gut epithelium, fimbrin concentrates at the apical surface 2-3 days after villin; this redistribution occurs when the visceral endoderm microvilli first contain organized microfilament bundles and when microvilli first begin to appear in the gut. These results suggest a common mechanism of assembly of the absorptive surface of two different tissues in the embryo and identify villin as a useful marker for the visceral endoderm.

Functional comparison of villin and gelsolin. Effects of Ca2+, KCl, and polyphosphoinositides.

A family of homologous actin-binding proteins sever and cap actin filaments and accelerate actin filament assembly. The functions of two of these proteins, villin and gelsolin, and of their proteolytically derived actin binding domains were compared directly by measuring their effects, under various ionic conditions, on the rates and extents of polymerization of pyrene-labeled actin. In 1 mM Ca2+ and 150 mM KCl, villin and gelsolin have similar severing and polymerization-accelerating properties. Decreasing [Ca2+] to 25 microM greatly reduces severing by villin but not gelsolin. Decreasing [KCl] from 150 to 10 mM at 25 microM Ca2+ increases severing by villin, but not gelsolin, over 10-fold. The C-terminal half domains of both proteins have Ca2+-sensitive actin monomer-binding properties, but neither severs filaments nor accelerates polymerization. The N-terminal halves of villin and gelsolin contain all the filament-severing activity of the intact proteins. Severing by gelsolin's N-terminal half is Ca2+-independent, but that of villin has the same Ca2+ requirement as intact villin. The difference in Ca2+ sensitivity extends to 14-kDa N-terminal fragments which bind actin monomers and filament ends, requiring Ca2+ in the case of villin but not gelsolin. Severing of filaments by villin and its N-terminal half is shown to be inhibited by phosphatidylinositol 4,5-bisphosphate, as shown previously for gelsolin (Janmey, P.A., and Stossel, T.P. (1987) Nature 325, 362-364). The functional similarities of villin and gelsolin correlate with known structural features, and the greater functional dependence of villin on Ca2+ compared to gelsolin is traced to differences in their N-terminal domains.

Structural and functional relationship between the membrane and the cytoskeleton in brush border microvilli.

Electron microscopic and biochemical studies have described the organization and composition of microvilli from chicken intestinal brush borders. An actin-based cytoskeleton, composed of a paracrystalline core of bundled microfilaments, maintains the finger-like shape of the membrane through a helical array of membrane-microfilament linkages. Two proteins, fimbrin and villin, are components of the core bundle in situ and can independently bundle the actin filaments in vitro. Structural studies comparing microvillar core bundles with villin bundles and fimbrin bundles suggest that fimbrin, and not villin, is the major actin-filament-bundling protein in the microvillus core. These points, together with the capability of villin to sever actin filaments when activated by Ca2+, raise questions about villin's function in the microvillus. One possible explanation is that villin induces vesiculation of the membrane by disassembling the underlying cytoskeleton.

Partial reconstruction of the microvillus core bundle: characterization of villin as a Ca++-dependent, actin-bundling/depolymerizing protein.

The brush border, isolated from chicken intestine epithelial cells, contains the 95,000 relative molecular mass (M(r)) polypeptide, villin. This report describes the purification and characterization of villin as a Ca(++)-dependent, actin bundling/depolymerizing protein. Then 100,000 g supernatant from a Ca(++) extract of isolated brush borders is composed of three polypeptides of 95,000 (villin), 68,000 (fimbrin), and 42,000 M(r) (actin). Villin, following purification from this extract by differential ammonium sulfate precipitation and ion-exchange chromatography, was mixed with skeletal muscle F-actin. Electron microscopy of negatively stained preparations of these villin-actin mixtures showed that filament bundles were present. This viscosity, sedimentability, and ultrastructural morphology of filament bundles are dependent on the villin:actin molar ratio, the pH, and the free Ca(++) concentration in solution. At low free Ca(++) (less than 10(-6) M), the amount of protein in bundles, when measured by sedimentation, increased as the villin: actin molar ratio increased and reached a plateau at approximately a 4:10 ratio. This behavior correlates with the conversion of single actin filaments into filament bundles as detected in the electron microscope. At high free Ca(++) (more than 10(-6) M), there was a decrease in the apparent viscosity in the villin-actin mixtures to a level measured for the buffer. Furthermore, these Ca(++) effects were correlated with the loss of protein sedimented, the disappearance of filament bundles, and the appearance of short fragments of filaments. Bundle formation is also pH-sensitive, being favored at mildly acidic pH. A decrease in the pH from 7.6 to 6.6 results in an increase in sedimentable protein and also a transformation of loosly associated actin filaments into compact actin bundles. These results are consistent with the suggestions that villin is a bundling protein in the microvillus and is responsible for the Ca(++)-sensitive disassembly of the microvillar cytoskeleton. Thus villin may function in the cytoplasm as a major cytoskeletal element regulating microvillar shape.

Organization of the cross-filaments in intestinal microvilli.

We studied the arrangement of the cross-filaments in intestinal microvilli to understand how microfilaments interact with the membrane. Observations on thin-sectioned or negatively stained microvilli with the electron microscope demonstrate that the cross-filaments on the core bundle lie opposite to one another and are spaced 32.5 nm apart. In sections grazing through the membranes, the cross-filaments appear as transverse stripes in a barber-polelike arrangement. The cross-filaments point away from the microvillus tip. This subfragments S1 or HMM. The cross filaments are associated not only with the microfilaments but also with electron-dense patches on the inside surface of the membrane. These results suggest the cross-filaments are arranged as a double helix around the core bundle. Furthermore, the cross-filaments can serve as in situ markers for microvillar polarity. Lastly, the cross-filaments interact not only with specific portions on the actin filaments but also with dense patches on the membrane. These observations are summarized in a model of the microvillus cytoskeleton.

Identification and organization of the components in the isolated microvillus cytoskeleton.

We have examined the effects of ATP and deoxycholate (DOC) on the cytoskeletal organization of Triton-demembranated microvilli (MV) isolated from chicken intestine brush borders. Isolated MV are composed of a core of tightly bundled microfilaments from which arms project laterally to the plasma membrane with a 33-nm periodicity. These lateral arms spiral around the core microfilaments as a helix with a 25 degrees pitch. Demembranated MV consist of four polypeptides with mol wt of 110,000, 95,000, 68,000, and 42,000, present in molar ratios of 1.1:1.6:1.3:10.0. After addition of 50 microM ATP and 0.1 mM Mg++, the cytoskeletons are organized as a tight bundle of microfilaments from which lateral arms are missing. In these ATP-treated cytoskeletons, the 110-kdalton polypeptide is reduced in amount and the 95,000, 68,000, and 42,000 polypeptides are present in a 1.3:1.2:10.0 ratio. In contrast, after incubation with 0.5% DOC, the core microfilaments are no longer tightly bundled yet the lateral arms remain attached with a distinct 33-nm periodicity. These DOC-treated cytoskeletons are depleted of the 95,000 and 68,000 polypeptides and are composed of the 110,000 and 42,000 polypeptides in a 2:10 molar ratio. These results suggest that the microfilaments are associated into a core bundle by the 95- and 68-kdalton polypeptides and from this core bundle project the lateral arms composed of the 110-kdalton polypeptide.