Polylysine binding to unphosphorylated smooth muscle myosin enhances formation and stabilizes myosin filaments in vitro

Previously, we demonstrated that positively charged polylysine, our model for biological polyamines, activates the Mg2+ ATPase activity of unphosphorylated smooth muscle myosin and shifts the myosin conformation from the folded 10S to linear 6S form. These effects of polylysine were reversed by the oppositely charged heparin (Szymanski et al. (1993) Am J Physiol 265, C379). In the present report, we provide further information on polylysine binding to smooth muscle myosin, and test the hypothesis that polylysine binding to unphosphorylated myosin involves filament formation. To relate the effects of polylysine on contractility in smooth muscle to physiologically relevant material, we investigated the ability of naturally occurring positively charged polyamines, histones, cadaverine, putrescine and spermidine to activate the Mg2+ ATPase activity of unphosphorylated smooth muscle myosin. Our data show that polylysine binding to individual unphosphorylated myosin molecules stimulates formation of myosin filaments. Polylysine also interacts with myosin filaments, causing enhancement of their size and the numbers, and this could be reversed by heparin. Polylysine binding to myosin filaments made them more resistant to disassembly by high salt concentrations (KCl) or ATP. Naturally occurring polyamines in millimolar concentrations activate the Mg2+ ATPase activity of unphosphorylated smooth muscle myosin. We suggest that the electrostatic interactions between naturally occurring positively charged polyamines and unphosphorylated smooth muscle myosin may play a role in stabilization of thick filament structurein situ.

Calponin binds to the 20-kilodalton regulatory light chain of myosin

Calponin (CaP) is a 34 kDa smooth muscle-specific protein that has been implicated in regulation of smooth muscle contractility. Two CaP binding sites on smooth muscle myosin rod have been recently described [Szymanski and Tao (1997) J.Biol.Chem. 272, 11142-11146]. We used a combination of cosedimentation, overlay, and fluorescence assays to determine the interaction between CaP and both subfragment 1 of myosin and isolated 20 kDa regulatory light chain of myosin (RLC). Subfragment 1, which was generated by cleavage of myosin with Staphylococcus aureus protease (myosin S1SA) inhibits cosedimentation of CaP with myosin filaments. Fluorescence assay showed that CaP labeled with fluorescent label (DAN-CaP) interacts with myosin S1SA in solution via a single class of binding sites. The binding constant (kaff) of this interaction at 50 mM NaCl is (2. 1 +/- 0.2) x 10(6) M-1 (n = 3). The interaction between DAN-CaP and myosin S1SA depends on ionic strength, and the EC50 of inhibition of this interaction occurs at about 130 mM NaCl. In contrast, the subfragment 1 that was generated by papain digestion (myosin S1PA), which cleaves RLC 4 kDa away from the NH2-terminal end of the molecule, does not interact with DAN-CaP. Overlay and fluorescent assay in solution showed that CaP binds to isolated RLC, suggesting that the interaction between CaP and subfragment 1 of myosin is due to a direct binding of CaP to RLC. CaP binding to myosin S1SA is stronger than to subfragment 2 in physiological salt concentrations. CaP binding to myosin head strengthened upon phosphorylation of RLC by Ca2+/calmodulin-dependent myosin light chain kinase. We suggest that CaP binds to subfragment 1 of myosin, exclusively via the NH2-terminal end of RLC, and this interaction could play a role in regulation of the actin-myosin interaction in smooth muscle contractility.

Differences in contractile protein content and isoforms in phasic and tonic smooth muscles

The basis of tonic vs. phasic contractile phenotypes of visceral smooth muscles is poorly understood. We used gel electrophoresis and quantitative scanning densitometry to measure the content and isoform composition of contractile proteins in opossum lower esophageal sphincter (LES), to represent tonic muscle, and circular muscle of the esophageal body (EB), to represent phasic smooth muscle. The amount of protein in these two types of muscles is similar: approximately 27 mg/g of frozen tissue. There is no difference in the relative proportion of myosin, actin, calponin, and tropomyosin in the two muscle types. However, the EB contains approximately 2.4-times more caldesmon than the LES. The relative ratios of alpha- to gamma-contractile isoforms of actin are 0.9 in the LES and 0.3 in EB. The ratio between acidic (LC17a) and basic (LC17b) isoforms of the 17-kDa essential light chain of myosin is 0.7:1 in the LES, compared with 2.7:1 in the EB. There is no significant difference in the ratios of smooth muscle myosin SM1 and SM2 isoforms in the two muscle types. The level of the myosin heavy chain isoform, which contains the seven-amino acid insert in the myosin head, is about threefold higher in the EB compared with LES. In conclusion, the esophageal phasic muscle in contrast to the tonic LES contains proportionally more caldesmon, LC17a, and seven-amino acid-inserted myosin and proportionally less alpha-actin. These differences may provide a basis for functional differences between tonic and phasic smooth muscles.

Hybrid interior point training of modular neural networks

Modular neural networks use a single gating neuron to select the outputs of a collection of agent neurons. Expectation-maximization (EM) algorithms provide one way of training modular neural networks to approximate non-linear functionals. This paper introduces a hybrid interior-point (HIP) algorithm for training modular networks. The HIP algorithm combines an interior-point linear programming (LP) algorithm with a Newton-Raphson iteration in such a way that the computational efficiency of the interior point LP methods is preserved. The algorithm is formally proven to converge asymptotically to locally optimal networks with a total computational cost that scales in a polynomial manner with problem size. Simulation experiments show that the HIP algorithm produces networks whose average approximation error is better than that of EM-trained networks. These results also demonstrate that the computational cost of the HIP algorithm scales at a slower rate than the EM-procedure and that, for small-size networks, the total computational costs of both methods are comparable.

Tyrosine phosphorylation in contraction of opossum esophageal longitudinal muscle in response to SNP

Sodium nitroprusside (SNP) has been shown to elicit a guanosine 3',5'-cyclic monophosphate (cGMP)-mediated, indomethacin-sensitive contraction of the opossum esophageal longitudinal muscle. We examined the role of tyrosine phosphorylation in the signal transduction pathway of contractions induced by SNP and cGMP in longitudinal muscle strips in vitro. Force of isometric contractions was expressed as the percentage of responses to KCl (73 mM). SNP (100 microM)-induced contractions were 75 +/- 5% before and 3 +/- 2% after 50 microM genistein (P < 0.005) and 86 +/- 16% before and 0 +/- 0% after 50 microM tyrphostin B46. Contractions in response to 8-bromo-cGMP (8-BrcGMP; 1 mM) were 74 +/- 15% before and 3 +/- 2% after genistein (P < 0.01) and 63 +/- 15% before and 18 +/- 4% after tyrphostin B46 (P < 0.05). In contrast, KCl-induced contractions were 82 +/- 8% and 96 +/- 9% of the control value after genistein and tyrphostin B46 treatments, respectively (P > 0.05 for both). Carbachol contractions were partially suppressed by genistein (106 +/- 8% vs. 79 +/- 8%; P < 0.05) but unaffected by tyrphostin B46 (114 +/- 10% vs. 107 +/- 12%; P > 0.05). Western blot analysis revealed a 116-kDa phosphotyrosine protein in the control muscle strips. The level of this protein was increased to 206 +/- 15% of control after SNP treatment. Both genistein and tyrphostin B46 blocked this increase. These studies show that contractions of the esophageal longitudinal muscle induced by SNP and cGMP utilize a signal transduction pathway different from that used by the depolarizing agent KCl and the muscarinic agonist carbachol. Contractions induced by SNP and cGMP involve tyrosine phosphorylation of a protein, possibly identified as a 116-kDa protein, as a key step in the signaling pathway.

Electrostatic effects of smooth muscle calponin on actin assembly

The contribution of electrostatic interactions to the effects of chicken gizzard calponin on the kinetics of actin polymerization and the bundling of F-actin were characterized by a combination of fluorescence, light-scattering, co-sedimentation, and electron-microscopic methods. Stoichiometric amounts of calponin accelerate actin polymerization in low-ionic-strength solutions, but this effect is diminished at [KCI] = 150 mM. At low ionic strengths, micromolar concentrations of calponin induce the formation of large bundles of actin filaments, and lower concentrations of calponin quench the fluorescence of pyrene-labeled F-actin. The latter effect is related to binding of calponin to F-actin rather than to bundling of the filaments. The concentration of calponin required to bundle a fixed concentration of actin filaments increases with increasing ionic strength, as the average diameter of the bundles decreases. Millimolar concentrations of ATP, GTP or ITP are equally efficient at dispersing actin bundles to single filaments or smaller aggregates, even though a significant fraction of calponin remains bound to F-actin. Our findings show that the binding of calponin to actin is determined at least in part by electrostatic interactions, and that the polycationic nature of calponin is primarily responsible for the formation of F-actin bundles via its ability to reduce the electrostatic repulsion between the negatively charged actin filaments.

Localization of protein regions involved in the interaction between calponin and myosin

Calponin is a 33-kDa smooth muscle-specific protein that has been suggested to play a role in muscle contractility. It has previously been shown to interact with actin, tropomyosin, and calmodulin. More recently we showed that calponin also interacts with myosin (Szymanski, P. T., and Tao, T. (1993) FEBS Lett. 331, 256-259). In the present study we used a combination of co-sedimentation and fluorescence assays to localize the regions in myosin and calponin that are involved in the interaction between these two proteins. We found that recombinant chicken gizzard alpha-calponin co-sediments with myosin rod and, to a lesser extent, with light meromyosin. Fluorescently labeled recombinant calponin shows interaction with heavy meromyosin and myosin subfragment 2 but not subfragment 1. A fragment comprising residues 7-182 and a synthetic peptide spanning residues 146-176 of calponin co-sediment with myosin, but fragments comprising residues 7-144 and 183-292 do not. Our results indicate that there are calponin binding sites in the subfragment 2 and light meromyosin regions of myosin, and that the region comprising residues 145-182 of calponin mediates its interaction with myosin.

Correlation between calponin and myosin subfragment 1 binding to F-actin and ATPase inhibition

Calponin is a thin-filament-associated protein that has been implicated in the regulation of smooth-muscle contractility. It binds to F-actin and inhibits the MgATPase activity of actomyosin. In the present work we have examined the effect of recombinant chicken gizzard alpha-calponin (R alpha CaP) on the binding of rabbit skeletal-muscle myosin subfragment 1 (S1) to F-actin and on the inhibition of its actin-activated MgATPase. We have found that binding of one R alpha CaP molecule to every three to four actin monomers is sufficient for maximal inhibition of acto-S1 ATPase. At this R alpha CaP/actin ratio R alpha CaP does not interfere with S1 binding to F-actin. At higher concentrations, R alpha CaP displaces S1 from F-actin and a 1:1 R alpha CaP-actin monomer complex is formed. R alpha CaP is also able to displace troponin I from its complex with F-actin which may reflect the amino acid sequence similarity between R alpha CaP and troponin I in their actin-binding regions.

Effects of calponin on isometric force and shortening velocity in permeabilized taenia coli smooth muscle

Calponin, a thin filament-associated protein, inhibits actomyosin adenosinetriphosphatase in solution and has been suggested to modulate smooth muscle contractility. We used permeabilized guinea pig taenia coli smooth muscle to investigate whether calponin can modulate actin-myosin interaction in a more organized contractile system. Fibers were permeabilized with Triton X-100 and glycerol, which permit access of large macromolecules to the contractile apparatus. For contractures elicited by Ca2+ (6.6 microM + 0.1 microM calmodulin), the recombinant alpha-isoform of chicken gizzard calponin (CaP) decreased isometric force (Fo) and unloaded shortening velocity (Vus) in a dose-dependent manner; 1 microM CaP had minimal effects on force (< 10%) but reduced Vus by approximately 50% and 10 microM CaP reduced Fo to 27% of control and Vus to near zero levels. To eliminate any effects of the binding of calmodulin by CaP and consequent inhibition of myosin light chain kinase activity, we also studied fibers activated by thiophosphorylation of the myosin regulatory light chain. Fo was only moderately inhibited, remaining at approximately 75% of control in the presence of CaP (10 microM), whereas Vus was reduced to 32% of control. A similar inhibition was obtained with a mutant (CaPcys175) that retains the ability to bind to actin. CaP phosphorylated by protein kinase C and CaPcys175 mutant labeled with 1,5-IAEDANS, which bind actin poorly, were not effective inhibitors. Our results indicate that 1) CaP more strongly inhibits Vus (approximately cross-bridge cycle rate) than Fo (approximately number of activated cross bridges) and 2) the effects of CaP are related to its binding to actin. Thus the function of CaP in regulation of smooth muscle contractility may be more strongly related to its function as a modulator of velocity, as related to the "latch state," than as an "on-off" switch.

Interaction between calponin and smooth muscle myosin

Calponin is a thin filament-associated protein in smooth muscle that has been shown to bind actin, tropomyosin and calmodulin, and has been implicated to play a role in regulation of smooth muscle contractility. Using a centrifugation assay we found that calponin interacts with unphosphorylated filamentous smooth muscle myosin. We found that this calponin-myosin interaction is reversed by Ca(2+)-CaM, and depends on ionic strength. At 50 mM NaCl the binding constant and the stoichiometry of this interaction were estimated to be 2 x 10(6) M-1, and 1.2-2.4 calponin per myosin, respectively. We suggest that the calponin-myosin interaction could be involved in regulation of smooth muscle contractility by anchoring myosin to actin.

Polylysine activates smooth muscle myosin ATPase activity via induction of a 10S to 6S transition

Polylysine (10-13 kDa) stimulates contraction in smooth muscle skinned fibers and activates actomyosin adenosinetriphosphatase (ATPase) activity in the absence of myosin light chain phosphorylation [P. T. Szymanski and R. J. Paul. Adv. Exp. Med. 304: 363-368, 1991; P. T. Szymanski, J. D. Strauss, G. Doerman, J. DiSalvo, and R. J. Paul. Am J. Physiol. 262 (Cell Physiol. 31): C1445-C1455, 1992]. To provide further information on the mechanism of polylysine action on contractility in smooth muscle, we investigated its effect on ATPase activity and conformation of purified gizzard myosin. We report here that polylysine directly stimulates myosin ATPase activity in a concentration-dependent manner. This stimulation could be completely abolished with the addition of heparin, a negatively charged heteropolysaccharide. Polylysine (10 microM) increases myosin ATPase activity to a level similar to that of myosin phosphorylation. Addition of 10 microM polylysine to phosphorylated myosin [with myosin light chain kinase and adenosine 5'-O-(3-thiotriphosphate) (ATP gamma S), to approximately 1.9 mol P/mol myosin], however, did not further stimulate ATPase activity. At 0.2 M KCl (the salt concentration at which myosin exists primary in the 10S form), the addition of polylysine increases myosin ATPase activity to a level comparable to that of untreated myosin in 0.3 M KCl. These changes parallel the increase in solution viscosity elicited by polylysine. These results suggest that polylysine induces a transition in myosin conformation from the 10S to the 6S form, and this was confirmed by electron microscopy.(ABSTRACT TRUNCATED AT 250 WORDS)

Polylysine activates smooth muscle actin-myosin interaction without LC20 phosphorylation

Phosphorylation/dephosphorylation of the 20-kDa light chain of smooth muscle myosin is a major regulator of actin-myosin interaction. Phosphatase inhibitors have thus been shown to enhance contraction in smooth muscle. The activity of type II phosphatase against phosphorylated myosin light chains is inhibited by polylysine. Thus we studied the effects of polylysine (10-13 kDa) on actin-myosin interaction in permeabilized guinea pig taenia coli fibers and in bovine aortic actomyosin. Addition of polylysine (10-20 microM) to Ca-ethylene glycol-bis(beta-aminoethyl ether)-N,N,N',N'-tetraacetic acid buffered solution ([Ca2+] less than 0.01 microM) elicited a contraction in fibers of 40 +/- 8% (n = 6) of maximally stimulated contractions ([Ca2+] congruent to 1.5 microM). Untreated fibers did not generate any significant force in parallel control experiments. Similarly, polylysine stimulated the ATPase activity both in fibers and actomyosin in a dose-dependent manner. This stimulation could be completely inhibited and abolished upon addition of heparin, a negatively charged heteropolysaccharide. In actomyosin previously phosphorylated with ATP gamma S, polylysine in a concentration range of 2-13 microM did not further stimulate enzyme activity. These increases in activity were not connected with significant changes in the phosphorylation of 20-kDa myosin light chain nor could any incorporation of 32P associated with polylysine stimulation be detected in both skinned fibers and actomyosin by autoradiography of SDS gels. Our data indicate that polylysine increases actin-myosin interaction in both smooth muscle model systems by directly influencing contractile proteins. As such, polylysine may be a useful probe for the mechanism of activation of smooth muscle.

Human retinal vascular cells differ from umbilical cells in synthetic functions and their response to glucose

Cell culture systems have commonly been used to study mechanisms implicated in the pathogenesis of diabetic retinopathy, but the great majority of cell preparations used have been either of nonhuman retinal origin or nonretinal human origin. Because of questions of species and organ specificity in the function of cells of vascular origin, in this study, cultured microvascular endothelial cells (HREC), pericytes (HRPC), and pigment epithelial cells from the postmortem human retina, and endothelial cells from human umbilical vein (HUVEC) were evaluated with respect to cell proliferation, and secretory products potentially important in diabetic retinopathy, i.e., prostaglandins (PG) and plasminogen activators (PA), normalized to DNA content/well, under both basal (5 mM) and high (25 mM) glucose conditions. Glucose (25 mM) reduced DNA content similarly in both types of endothelial cells, had a lesser effect on HRPC, and did not significantly alter the proliferation of pigment epithelial cells. Basal secretion of PGI2 (measured as 6-keto-PGF1 alpha) was in the order HRPC much greater than HREC greater than HUVEC, whereas PGE2 secretion was in the order HREC much greater than HRPC greater than HUVEC. Glucose (25 mM) stimulated PGI2 secretion by HRPC, but not by either type of endothelial cell, and enhanced PGE2 secretion by HREC, but not by HUVEC or HRPC. Release of plasminogen activator activity differed between HUVEC and HREC under basal conditions and addition of 25 mM glucose stimulated release only from HREC. Glucose (25 mM) stimulated PA secretion by HREC, but not by HUVEC. These findings provide evidence that human retinal pericytes are an important source of prostacyclin, and that there are differences between HREC and HUVEC with respect to secretory functions and their modulation by glucose, indicating regional specificity of these functions. Extrapolation to human retinal vascular cells from experiments using cells from heterologous vascular beds to draw inferences about the pathophysiology of diabetic retinopathy are not valid for these cellular functions.