The calmodulin hypothesis of neurotransmission

Ca²⁺ plays a major role in neurotransmission and synaptic modulation. Evidence is presented to support the calmodulin hypothesis of neurotransmission developed in this laboratory stating that calmodulin, a major Ca²⁺ binding protein in brain, mediates the effects of Ca²⁺ on neurotransmission. Calmodulin was isolated from highly enriched preparations of synaptic vesicles and nerve terminal cytoplasm. Ca²⁺ and calmodulin were shown to regulate several synaptic processes in isolated and intact preparations, including endogenous synaptic Ca²⁺-calmodulin protein kinase activity, neurotransmitter release, and synaptic vesicle and synaptic membrane interactions. Ca²⁺ and calmodulin were shown to activate a synaptic tubulin kinase system which was shown to be a distinct enzyme system from the cyclic AMP protein kinase. Ca²⁺ and calmodulin stimulated phosphorylation of tubulin altered the properties of tubulin, forming insoluble tubulin fibrils. Evidence for the role of Ca²⁺-calmodulin kinase activity, especially the calmodulin-tubulin kinase, in neurotransmission are presented. The effects of several neuroactive drugs on the synaptic calmodulin system are presented. The results support the hypothesis that calmodulin mediates many of calcium's actions at the synapse, and that the effects of Ca²⁺ on synaptic protein phosphorylation, especially synaptic tubulin, may provide a biochemical mechanism for converting the Ca²⁺ signal into a motor force in the process of neurotransmission.     

Interaction of Arabidopsis kinesin-like calmodulin-binding protein with tubulin subunits: modulation by Ca2+-calmodulin

K inesin-like c almodulin-b inding p rotein (KCBP) is a recently identified novel kinesin-like protein that appears to be unique to and ubiquitous in plants. KCBP is distinct from all other known KLPs in having a calmodulin-binding domain adjacent to its motor domain. We have used different regions of KCBP to study its interaction with tubulin subunits and the regulation of this interaction by Ca²⁺-calmodulin. The results show that the carboxy-terminal part of the KCBP, with or without calmodulin-binding domain, binds to tubulin subunits and this binding is sensitive to nucleotides. In the presence of Ca²⁺-calmodulin the motor with calmodulin-binding domain does not bind to tubulin. This Ca²⁺-calmodulin modulation is abolished in the presence of antibodies specific to the calmodulin-binding domain of KCBP. Similar binding studies with the carboxy-terminal part of KCBP lacking the calmodulinbinding domain show no effect of Ca²⁺-calmodulin. These results indicate that Ca²⁺-calmodulin modulates the interaction of KCBP with tubulin subunits and this modulation is due to the calmodulin-binding domain in the KCBP. Calcium-dependent calmodulin modulation of KCBP interaction with tubulin suggests regulation of KCBP function by calcium, the first such regulation of a kinesin heavy chain among all the known kinesin-like proteins.     

Ca2+ calmodulin-dependent protein kinase activity in the ascomycetes Neurospora crassa

DEAE-cellulose column chromatography of Neurospora crassa soluble mycelial extracts leads to the resolution of three major protein kinase activity peaks designated PKI, PKII, and PKIII.PKII activity is stimulated by Ca²⁺ and Neurospora or brain calmodulin. Maximal stimulation was observed at 2 µM-free Ca²⁺ and 1 µg/ml of the modulator. The stimulatory effect of the Ca²⁺-calmodulin complex was blocked by EGTA and by some calmodulin antagonists such as phenothiazine drugs or compound 48/80.PKII phosphorylates different proteins, among which histone II-A at a low concentration and CDPKS, the synthetic peptide specific for Ca²⁺-calmodulin dependent protein kinases, are the best substrates. Some phosphorylation can be detected in the absence of any exogenous acceptor. PKII activity assayed in the presence of histone II-A or in the absence of exogenous phosphate acceptor (autophosphorylation) co-elute in a DEAE-cellulose column at 0.28 M NaCl. As result of the autophosphorylation reaction of the purified enzyme a main phosphorylated component of 70 kDa was resolved by SDS-polyacrylamide gel electrophoresis. It is possible that this component is an active part of this enzyme.     

The role of calmodulin in opioid-induced changes in the phosphorylation of rat striatal synaptic membrane proteins

Chronic morphine treatment of rats decreased the level of phosphorylation of synaptic membrane proteins of the striatum assayed in vitro. Although the patterns of phosphorylated proteins separated on SDS-gel electrophoresis from morphine-tolerant rats resembled patterns produced by lowering Ca²⁺ levels in the assay, supplementation of the protein kinase assay with Ca²⁺ and its binding protein, calmodulin, did not restore full kinase activity. The addition of methadone or etorphine to the protein kinase in vitro however, was able to block the Ca²⁺-calmodulin stimulation of phosphorylation in both synaptic membranes and intact synaptosomes. These data suggest that opioids produce an irreversible (or slowly reversible) defect in the Ca²⁺-dependent protein kinase system of striatal membranes.This paper is dedicated to Dr. Derek Richter on his seventy-fifth birthday.     

Ca2+ and Calmodulin-Dependent Phosphorylation of Endogenous Synaptic Vesicle Tubulin by a Vesicle-Bound Calmodulin Kinase System

Endogenous synaptic vesicle alpha- and beta-tubulin were shown to be the major substrates for a Ca2+-calmodulin-regulated protein kinase system in enriched synaptic vesicle preparations from rat cortex as determined by two-dimensional gel electrophoresis and peptide mapping. The activation of this endogenous tubulin kinase system was dependent on Ca2+ and the Ca2+ binding protein, calmodulin. Under maximally stimulated conditions, approximately 40% of the tubulin present in enriched synaptic vesicles was phosphorylated within less than 50 s by the vesicle Ca2+-calmodulin kinase. Evidence is presented indicating that the Ca2+-calmodulin tubulin kinase is an enzyme system distinct from previously described cyclic AMP protein kinases. alpha-Tubulin and beta-tubulin were identified as major components of previously designated vesicle phosphorylation bands DPH-L and DPH-M. The Ca2+-calmodulin tubulin kinase is very labile and specialized isolation procedures were necessary to retain activity. Ca2+-activated synaptic vesicle tubulin phosphorylation correlated with vesicle neurotransmitter release. Depolarization-dependent Ca2+ uptake in intact synaptosomes simultaneously stimulated the release of neurotransmitters and the phosphorylation of synaptic vesicle alpha- and beta-tubulin. The results indicate that regulation of the synaptic vesicle tubulin kinase by Ca2+ and calmodulin may play a role in the functional utilization of synaptic vesicle tubulin and may mediate some of the effects of Ca2+ on vesicle function and neurosecretion.     

Studies on the kinetics of cation-associated fluorescence changes in chloroplast membranes

A soluble Ca²⁺- and Ca²⁺—calmodulin-activated protein kinase was partially purified from wheat germ. The phosphorylation of histones and casein catalyzed by this enzyme is largely Ca²⁺-dependent. After repeated gel filtration of the protein kinase in the presence of 1 mM EGTA, the phosphorylation of casein and histones by the enzyme is activated 3-fold and up to 16-fold, respectively, by added calmodulin (12.5 μM). Such activation of the protein kinase by calmodulin is Ca²⁺-dependent. The protein kinase binds to calmodulin—Sepharose 4B in a Ca²⁺-dependent fashion. This type of Ca²⁺-activated protein kinase may be involved in stimulus—response coupling in plants.     

Ca2+/calmodulin-dependent protein kinase in Saccharomyces cerevisiae

Ca²⁺-dependent chromatography of soluble cytosolic proteins on calmodulin-Sepharose gave a fraction that exhibited Ca²⁺- and calmodulin-dependent phosphorylation of several polypeptides, including 60, 56 and 45 kDa species. At 0.2 μM beef calmodulin the phosphorylation was optimal at 3 μM free Ca²⁺, and at 80 μM free Ca²⁺ it was half-maximal at about 0.1 μM beef calmodulin. It is concluded that the fraction contains calmodulin-dependent protein kinase(s) which is (are) autophosphorylated or associated with substrates.     

Stimulation of heart sarcolemmal calcium pump by calmodulin

The sarcolemmal membranes obtained from rat heart by sucrose-density gradient method were found to exhibit Ca²⁺ stimulated Mg²⁺ dependent ATPase and ATP-dependent Ca²⁺ binding activities. The Ca²⁺ stimulated ATPase activity was increased by calmodulin; maximal effect was seen at 1 to 5 μg/ml concentrations of calmodulin. The observed activation of the enzyme was associated with an increase in Vmax value from 3.45 to 5.26 μmol Pi/mg protein/hr and a decrease in Ka value from 2.78 to 0.84 μM Ca²⁺. Calmodulin was also found to increase ATP-dependent Ca²⁺ binding by 1.6 to 2.2 fold. These results suggest that the activity of Ca²⁺ pump mechanism in heart sarcolemma is regulated by calmodulin.     

Phosphorylation of Caldesmon at Sites between Residues 627 and 642 Attenuates Inhibitory Activity and Contributes to a Reduction in Ca-Calmodulin Affinity

Caldesmon is an actin- and myosin-binding protein found in smooth muscle that inhibits actin activation of myosin ATPase activity. The activity of caldesmon is controlled by phosphorylation and by binding to Ca²⁺-calmodulin. We investigated the effects of phosphorylation by p²¹-activated kinase 3 (PAK) and calmodulin on the 22 kDa C-terminal fragment of caldesmon (CaD22). We substituted the major PAK sites, Ser-672 and Ser-702, with either alanine or aspartic acid to mimic nonphosphorylated and constitutively phosphorylated states of caldesmon, respectively. The aspartic acid mutation of CaD22 weakened Ca²⁺-calmodulin binding but had no effect on inhibition of ATPase activity. Phosphorylation of the aspartic acid mutant with PAK resulted in the slow phosphorylation of Thr-627, Ser-631, Ser-635, and Ser-642. Phosphorylation at these sites weakened Ca²⁺-calmodulin binding further and reduced the inhibitory activity of CaD22 in the absence of Ca²⁺-calmodulin. Phosphorylation of these sites of the alanine mutant of CaD22 had no effect on Ca²⁺-calmodulin binding but did reduce inhibition of ATPase activity. Thus, the region between residues 627 and 642 may contribute to the overall regulation of caldesmon's activity.     

Apparent variations in the activation characteristics of human erythrocyte membrane Ca2+-pump ATPase may be caused by variable membrane permeability

Published studies of the Ca²⁺-pump ATPase of the human erythrocyte membrane record a variety of patterns of activation by Ca²⁺ and calmodulin and also suggest that activation by Ca²⁺-calmodulin is slow rather than immediate. We have re-analysed these points in various types of human erythrocyte membrane preparation of widely different permeability characteristics, both in the intact state and after being rendered fully permeable by saponin. The various membrane preparations initially showed very different patterns of activation, but when permeabilised with saponin they all exhibited identical characteristics: these included highly cooperative activation by Ca²⁺ with maximum activity at ～ 1 μM-Ca²⁺ and high sensitivity to calmodulin. Activation of Ca²⁺-ATPase by Ca²⁺-calmodulin in freely permeable ghosts was immediate. We therefore conclude that the Ca²⁺-pump ATPase exhibits high sensitivity to Ca²⁺ and calmodulin and responds rapidly to Ca²⁺-calmodulin. Apparent evidence to the contrary seems likely to have been a result of misinter-pretation of data derived from studies of partially sealed erythrocyte ghosts in which the added activators, Ca²⁺ and calmodulin, did not have free access to the appropriate sites on the ATPase.     

Stimulation by phspholipid of calcium-dependent phosphorylation of endogenous proteins from mammalian tissues

The occurrence of endogenous substrate proteins for calcium-dependent protein kinase, augmented by either phospholipid or calmodulin, was examined in extracts of several tissues. Pancreas, vas deferens, adrenal and liver were found to contain substrate proteins for phospholipid-sensitive protein kinase. Phosphorylation of pancreatic substrate protein for phospholipid-senstivie protein kinase was rapid and highly sensitive to Ca²⁺, being detectable within 15 s a following exposure to Ca²⁺ and phosphatidylserine and at concentrations of Ca²⁺ as low as 0.5 μM. These findings suggest that the phospholipid-sensitive protien kinase system may serve to mediate some effects of Ca²⁺ in a variety of mammalian cell types.     

Calmodulin-binding proteins in brain

It is now widely accepted that actions of intracellular Ca²⁺ are mediated by a four-domain Ca²⁺-binding protein, calmodulin. Brain is especially rich in calmodulin, containing about 400 mg (24 μmol) of EGTA-extractable calmodulin per kg of brain. However, only a fraction of the above amount is required for the calmodulin-activated enzymes and most of the rest may be assigned to calmodulin-binding proteins, proteins which are apparently devoid of enzyme activities but undergo Ca²⁺-dependent associations with calmodulin. Several of such proteins have been recently discovered in brain. These include a heat-labile 80 K phosphodiesterase inhibitor protein (calcineurin), a heat-stable 70 K phosphodiesterase inhibitor protein, a 50 K protein, myelin basic protein, tubulin, microtubule τ (tau) factor, a spectrin-like doublet protein (240 plus 235 K) (calspectin; fodrin) and a particle-associated 155 K protein.Functions of these calmodulin-binding proteins have not been fully elucidated yet. Some proteins may be calmodulin-regulated enzymes catalyzing yet unknown biochemical reactions, e.g. a protein phosphatase activity was found for calcineurin. Some proteins may interact with contractile elements or cytoskeleton of the cell, e.g. τ factor and calspectin interacted with tubulin and F-actin, respectively and tubulin itself is a calmodulin-binding protein. So, interesting possibilities are the regulation of the functions of cytoskeleton by calmodulin through these calmodulin-binding proteins. Regulation of microtubule assembly by Ca²⁺-dependent binding of calmodulin to tubulin and/or τ factor and possible involvement of calspectin in the mechanism regulating axonal transport of neuronal proteins have been suggested. Thus, the exploration of the regulating functions of Ca²⁺/calmodulin in brain depends largely upon the further study of the properties of these calmodulin-binding proteins.     

The binding of phosphorylase kinase to immobilized calmodulin

The binding of phosphorylase kinase to calmodulin-Sepharose 4B was studied by column and batch methods. It was found that the Ca²⁺ dependence of the interaction strongly depended strongly depended on the degree of substitution of agarose with calmodulin. Equilibrium adsorption isotherms (i.e., bulk ligand binding functions and lattice site binding functions) of phosphorylase kinase were measured on calmodulin-Sepharose. Sigmoidal bulk ligand binding functions (bulk adsorption coefficients: 1.5–5.8) were found which indicate intermolecular attraction during binding. Hyperbolic lattice site binding functions (lattice adsorption coefficients: 1.0) were obtained thus excluding the existence of a critical surface concentration of immobilized calmodulin and indicating single independent binding sites on the gel surface and on phosphorylase kinase. These findings were combined to optimize the adsorption of phosphorylase kinase on calmodulin-Sepharose, for purification procedures at low Ca²⁺ concentrations (5–10 μM ) minimizing proteolysis by calpains. With this novel method phosphorylase kinase from rabbit and frog skeletal muscle could be purified ca 100- and 200-fold, respectively, in two steps.     

In Vitro Stimulation of Protein Kinase C by Melatonin

It has been shown that melatonin through binding to calmodulin acts both in vitro and in vivo as a potent calmodulin antagonist. It is known that calmodulin antagonists both bind to the hydrophobic domain of Ca²⁺ activated calmodulin, and inhibit protein kinase C activity. In this work we explored the effects of melatonin on Ca²⁺ dependent protein kinase C activity in vitro using both a pure commercial rat brain protein kinase C, and a partially purified enzyme from MDCK and N1E-115 cell homogenates. The results showed that melatonin directly activated protein kinase C with a half stimulatory concentration of 1 nM. In addition the hormone augmented by 30% the phorbol ester stimulated protein kinase C activity and increased [³H] PDBu binding to the kinase. In contrast, calmodulin antagonists (500 M) and protein kinase C inhibitors (100 M) abolished the enzyme activity. Melatonin analogs tested were ineffective in increasing either protein kinase C activity or [³H] PDBu binding. Moreover, the hormone stimulated protein kinase C autophosphorylation directly and in the presence of phorbol ester and phosphatidylserine. The results show that besides the melatonin binding to calmodulin, the hormone also interacts with protein kinase C only in the presence of Ca²⁺. They also suggest that the melatonin mechanism of action may involve interactions with other intracellular hydrophobic and Ca²⁺ dependent proteins.     

Crystallization and preliminary diffraction analysis of Ca2+-calmodulin-drug and apocalmodulin-drug complexes

Ca²⁺-calmodulin is crystallized with two new and potent drugs: a bisindol derivative (KAR-2, 3”-(β-chloroethyl)-2”,4”-dioxo-3,5”-spiro-oxazolidino-4-deacetoxy-vinblastine) with antitumor activity and an arylalkylamine fendiline analogue (N-(3,3-diphenylpropyl)-N'-[1-(3,4-di-n-butoxy-phenyl)-ethyl]-1,3-diaminopropane) with anticalmodulin activity. The crystals diffract beyond 2.8 Å and differ in unit cell parameters from each other as well as from crystals of Ca²⁺-calmodulin or Ca²⁺-calmodulin-ligand complexes, as reported thus far. Attempts to crystallize Ca²⁺-free calmodulin without drugs failed, in consonance with earlier results; however, single Ca²⁺-free calmodulin crystals diffracting beyond 2.5 Å resolution were grown in the presence of KAR-2. Results indicate that binding of the two drugs to apocalmodulin or Ca²⁺-calmodulin may induce unique novel protein conformers, targets of further detailed X-ray studies. © 1997 Wiley-Liss Inc.     

Calmodulin binding to platelet plasma membranes

Calmodulin copurifies with platelet plasma membranes isolated by glycerol-induced lysis and density gradient centrifugation. These membranes also bind ¹²⁵I-labeled calmodulin in vitro in the presence of Ca²⁺. Binding is largely reduced by replacing Ca²⁺ by Mg²⁺ or by addition of an excess unlabeled calmodulin. The specific component of binding is saturable, with an apparent K_d of 27 nM and a maximum of 15.9 pmol binding sites per mg of membrane protein. This is equivalent to approx. 4100 binding sites per platelet. Binding was inhibited by addition of phenothiazines, a group of calmodulin antagonists. Half-maximal inhibition was attained with approx. 20 μM trifluoperazine or 50 μM chlorpromazine. In contrast, chlorpromazine-sulfoxide which is inactive towards calmodulin, did not affect the binding. Calmodulin binding polypeptides of the plasma membrane were identified by a gel-overlay technique. A major calmodulin-binding component of molecular weight 149 000 was detected. Binding to this band was Ca²⁺-dependent and inhibited by chlorpromazine. The molecular weight of this polypeptide is similar to that of glycoprotein I and also that of the red cell (Ca²⁺ + Mg²⁺)-stimulated ATPase, which is known to bind calmodulin. The possible role of calmodulin in platelet activation is analysed.     

Effects of Ca2+, calmodulin and cyclic AMP on the phosphorylation of endogenous proteins by homogenates of rat islets of Langerhans

Activation of Ca²⁺-calmodulin- and cyclic AMP-dependent protein kinases has been suggested to be involved in stimulus-secretion coupling in the pancreatic β-cell. To study the properties of such kinases and their endogenous protein substrates homogenates of rat islets of Langerhans were incubated with [γ-³²P]ATP. Phosphorylated proteins were separated by sodium dodecyl sulphate polyacrylamide gel electrophoresis and detected by autoradiography. The phosphorylation of certain proteins could be enhanced by Ca²⁺ plus calmodulin or by cyclic AMP. The major effect of Ca²⁺ and calmodulin was to stimulate the phosphorylation of a protein (P53) of molecular weight 53 100±500 (n = 15). Maximum phosphorylation of protein P53 occurred within 2 min with 2 μM free Ca²⁺ and 0.7 μM calmodulin. Incorporation of label into protein P53 was inhibited by trifluoperazine or W7 but not by cyclic AMP-dependent protein kinase inhibitor. Phosphorylation of a protein of similar molecular weight could be enhanced to a lesser extent in the absence of Ca²⁺ but in the presence of cyclic AMP and 3-isobutylmethylxanthine: this phosphorylation was blocked by cyclic AMP-dependent protein kinase inhibitor. Cyclic AMP also stimulated incorporation of label into polypeptides of molecular weights 55 000 and 70–80 000. The results are consistent with the hypothesis that protein phosphorylation mechanisms may play a role in the regulation of insulin secretion.     

Inhibitory effect of calcium-binding protein regucalcin on protein kinase activity in the nuclei of regenerating rat liver

The effect of Ca²⁺-binding protein regucalcin on protein kinase activity in the nuclei of normal and regenerating rat livers was investigated. Protein kinase activity in the nuclei isolated from normal rat liver was significantly increased by addition of Ca²⁺ (500 μM) and calmodulin (10 μg/ml) in the enzyme reaction mixture. Nuclear protein kinase activity was significantly decreased in the presence of EGTA (1.0 mM), trifluoperazine (TFP; 20 μM), dibucaine (10⁻⁴ M), or staurosporine (10⁻⁷ M), indicating that Ca²⁺-dependent protein kinases are present in the nuclei. Protein kinase activity was significantly elevated in the liver nuclei obtained at 6 to 48 h after a partial hepatectomy. Hepatectomy-increased nuclear protein kinase activity was significantly decreased in the presence of EGTA (1.0 mM), TFP (20 μM), or staurosporine (10⁻⁷ M) in the enzyme reaction mixture. The presence of regucalcin (0.1–0.5 μM) caused a significant decrease in protein kinase activity in the nuclei obtained from normal and regenerating rat livers. Meanwhile, the nuclear protein kinase activity from normal and regenerating livers was significantly elevated in the presence of anti-regucalcin monoclonal antibody (50–200 ng/ml). The present study suggests that regucalcin plays a role in the regulation of protein kinase activity in the nuclei of proliferative liver cells. J. Cell. Biochem. 71:569–576, 1998. © 1998 Wiley-Liss, Inc.     

Inhibition of Ca2+/calmodulin-dependent phosphatase activity by regucalcin in rat liver cytosol: Involvement of calmodulin binding

The regulatory effect of regucalcin on Ca²⁺/calmodulin-dependent phosphatase activity and the binding of regucalcin to calmodulin was investigated. Phosphatase activity toward phosphotyrosine, phosphoserine, and phosphothreonine in rat liver cytosol was significantly increased by the addition of Ca²⁺ (100 μM) and calmodulin (0.30 μM). Thess increases were clearly inhibited by the addition of regucalcin (0.50–1.0 μM) into the enzyme reaction mixture. The cytosolic phosphoamino acid phosphatase activity was significantly elevated by the presence of anti-regucalcin monoclonal antibody (0.2 μg/ml), suggesting that endogenous regucalcin in the cytosol has an inhibitory effect on the enzyme activity. This elevation was prevented by the addition of regucalcin (0.50 μM). Purified calcineurin phosphatase activity was significantly increased by the addition of calmodulin (0.12 μM) in the presence of Ca²⁺ (1 and 10 μM). This increase was completely inhibited by the presence of regucalcin (0.12 μM). The inhibitory effect of regucalcin was reversed by the addition of calmodulin with the higher concentration (0.36 μM). Regucalcin has been demonstrated to bind on calmodulin-agarose beads by analysis with sodium dodecyl sulfate–polyacrylamide gel electrophoresis. The present study demonstrates that regucalcin inhibits Ca²⁺/calmodulin-dependent protein phosphatase activity in rat liver cytosol, and that regucalcin can bind to calmodulin. J. Cell. Biochem. 71:140–148, 1998. © 1998 Wiley-Liss, Inc.     

Inhibitory effect of regucalcin on Ca2+/calmodulin-dependent protein kinase activity in rat renal cortex cytosol

The effect of regucalcin on Ca²⁺/calmodulin-dependent protein kinase activity in the cytosol of rat renal cortex was investigated. Regucalcin is a calcium-binding protein which exists in rat liver and renal cortex. Protein kinase activity in renal cortex cytosol was markedly increased by the addition of CaCl₂ (0.5 mM) plus calmodulin (10 µg/ml) in the enzyme reaction mixture. This increase was completely prevented by the addition of trifluoperazine (25 µM), an antagonist of calmodulin. The cytosolic Ca²⁺/calmodulin- dependent protein kinase activity was clearly inhibited by the addition of regucalcin; an appreciable effect of regucalcin was seen at 0.01 µM. The cytosolic Ca²⁺/calmodulin-dependent protein kinase activity was fairly increased by increasing concentrations of added Ca²⁺ (100-1000 µM). This increase was markedly blocked by the presence of regucalcin (0.1 µM). The inhibitory effect of regucalcin on the protein kinase activity was also seen with varying concentrations of calmodulin (2-20 µg/ml). These results demonstrate that regucalcin can regulate Ca²⁺/calmodulin-dependent protein kinase activity in renal cortex cells.