Human Glomerular Mesangial Cells Research Articles

PDGF-mediated activation of phosphatidylinositol 3 kinase in human mesangial cells

Platelet-derived growth factor (PDGF) stimulates mitogenesis and exerts other biologic activities in glomerular mesangial cells. The precise mechanism of PDGF-induced mitogenesis in these cells is not clear. The activation of a signal transducing enzyme, phosphatidylinositol 3 kinase (PI 3 kinase) is associated with mitogenesis. Activation of PI 3 kinase results from stimulation of tyrosine kinase and G-protein-coupled classes of receptors. The synthesis of D3 phosphorylated inositides, the products of this enzymatic reaction, in non-nucleated cells such as blood platelets is dependent upon protein kinase C activation and G-proteins. We studied the activation of PI 3 kinase in response to PDGF in human glomerular mesangial cells. Using a PI 3 kinase 85 kD subunit specific antibody, we detected mesangial cell PI 3 kinase protein as 110 and 85 kD heterodimer. PDGF stimulated PI 3 kinase activity in antiphosphotyrosine immunoprecipitates in a dose-dependent manner showing maximum activation at 12 ng/ml. The antiphosphotyrosine associated PI 3 kinase activity showed biphasic kinetics with a fast peak within two minutes followed by a second peak at 10 minutes. Antiphosphotyrosine and PI 3 kinase immunoprecipitation studies indicated the association of the 85 kD PI 3 kinase subunit with PDGFR. Direct immunoprecipitation with PDGFR beta antibody showed the association of PI 3 kinase activity with the PDGF-receptor. The isoquinoline sulfonyl piperazine compound H7 at concentrations that inhibit PDGF-stimulated PKC activity had no effect on PDGF-stimulated PI 3 kinase activity in antiphospotyrosine immunoprecipitates. These data indicate that PI3 kinase activation is insensitive to PKC. Treatment of mesangial cells with pertussis toxin at concentrations that partially inhibited PDGF-induced DNA synthesis in human mesangial cells did not inhibit PDGF-induced PI 3 kinase activation. These data indicate that PDGF activates PI 3 kinase in mesangial cells and that pertussis toxin-sensitive G-proteins are not involved in PI 3 kinase activation. The data further dissociate activation of PI 3 kinase from mitogenesis in human mesangial cells.

Cytokine-Mediated Induction of HLA Antigen Expression on Human Glomerular Mesangial Cells

Well-characterized human glomerular mesangial cells were stimulated in vitro with interferon-γ (IFN-γ), interleukin-1, and tumor necrosis factor α (TNF-α) to determine the influence of these cytokines on the expression of class I and II HLA antigens. IFN-γ induced a dose-dependent increase in expression of both class I and class II HLA antigens. TNF-α induced only class I antigens when used alone, but acted synergistically with IFN-γ in the expression of class I and II antigens. Interleukin-1 had little effect on HLA antigen expression. These results raise the possibility that stimulated mesangial cells may act as antigen-presenting cells and may mediate local immune reactions in the glomerulus.

Effect of dopamine on ecto-5'-nucleotidase expression in human glomerular mesangial cells.

Ecto-5'-nucleotidase (5'-ribonucleotide phosphohydrolase, EC 3.1.3.5) of mesangial cells may be the main source of adenosine within the glomerulus, and thus essential in the regulation of glomerular microcirculation. c-AMP and c-AMP-stimulating agents were found to induce ecto-5'-nucleotidase of mesangial cells. Dopamine is a catecholamine known to increase c-AMP levels in mesangial cells. We have studied the effect of dopamine on ecto-5'-nucleotidase expression and DNA synthesis of glomerular mesangial cells in culture. Human mesangial cells were exposed to dopamine in the concentration range from 0.1 microM- to 1 mM, for 6-72 h. Ecto-5'-nucleotidase activity of human mesangial cells increased from 118.6 +/- 7.7 to 171 +/- 12 nmol/min/mg in a 72 h culture. This effect was time- and dose-dependent. Cycloheximide, an inhibitor of protein synthesis did not modify basal 5'-nucleotidase activity but it suppressed the stimulatory effect of 10 microM dopamine. DNA synthesis of human mesangial cells, studied after exposure of these cells to the same concentrations of dopamine used in the 5'-nucleotidase stimulation, was inhibited, being also dose dependent. These results indicate that dopamine induces ecto-5'-nucleotidase and inhibits DNA synthesis of cultured human mesangial cells. This action of dopamine on glomerular mesangial cells may be important in the regulation of glomerular hemodynamics.

Simvastatin inhibits PDGF-induced DNA synthesis in human glomerular mesangial cells

Inhibitors of 3-hydroxy-3-methylglutaryl coenzyme A-(HMGCoA) reductase ameliorate glomerular pathology and renal dysfunction in different models of glomerular disease. This effect has generally been attributed to a decrease in the circulating levels of cholesterol. Focal or diffuse mesangial cell proliferation is a common feature of glomerular pathology. There is now evidence from studies in vitro and in vivo that platelet-derived growth factor (PDGF) is an important mediator of glomerular hypercellularity. The activity of HMGCoA reductase has previously been shown to be a requirement for cell growth. In the present study, we examined the effect of simvastatin, and HMGCoA reductase inhibitor, on PDGF-induced DNA synthesis and PDGF B chain gene expression in human glomerular mesangial cells. In addition, we investigated the effect of simvastatin on phospholipase C (PLC) and protein kinase C (PKC) activation stimulated by PDGF. We demonstrate that treatment of the cells with simvastatin completely inhibits PDGF-induced DNA synthesis. This inhibition is reversed by mevalonate but not by cholesterol or farnesol, two major metabolites of the mevalonate pathway. On the other hand inhibition of HMGCoA reductase does not influence PDGF-induced activation of PLC and PKC, or PDGF B chain gene expression. These data suggest that simvastatin acts at a late step in the PDGF mitogenic pathway without interfering with other early cellular responses elicited by this growth factor. These studies also raise the possibility that the ameliorative effect of HMGCoA reductase inhibitors on glomerular pathology may be mediated, at least in part, by a direct cellular effect.

Porins and lipopolysaccharide stimulate platelet activating factor synthesis by human mesangial cells

Porins, a family of hydrophobic proteins located in the outer membrane of the cell wall of gram-negative bacteria and lipopolysaccharide (LPS), were shown to stimulate the synthesis of platelet activating factor (PAF), a phospholipid mediator of inflammation and endotoxic shock, by cultured human glomerular mesangial cells (MC). The synthesis of PAF induced by porins was rapid (peak at 20 min) and independent either from contamination by LPS or from generation of an endotoxin-induced cytokine such as tumor necrosis factor (TNF) since it was not prevented by cycloheximide, an inhibitor of protein synthesis or anti-TNF blocking antibodies. LPS also stimulated PAF synthesis by MC. However, the kinetic of PAF synthesis induced by LPS was biphasic with an early and transient peak at 10 minutes and a second and sustained peak at three to six hours. This second peak required an intact protein synthesis and was prevented by anti-TNF antibodies, suggesting the dependency on LPS-induced synthesis of TNF. Experiments with labeled precursors demonstrated that in MC, either after stimulation with porins or LPS, PAF was synthesized via the remodeling pathway that involves acetylation of 1-0-alkyl-sn-glyceryl-3-phosphorylcholine (2-lyso-PAF) generated from 1-0-alkyl-2-acyl-sn-glyceryl-3-phosphorylcholine by phospholipase A2 (PLA2) activity. Porins and LPS, indeed, induced PLA2-dependent mobilization of [14C]-arachidonic acid that was inhibited by p-bromodiphenacylbromide (PBDB). PBDB, an inhibitor of PLA2, also blocked PAF synthesis by preventing the mobilization of 2-lyso-PAF, the substrate for PAF-specific acetyltransferase.(ABSTRACT TRUNCATED AT 250 WORDS)

HIV infects glomerular endothelial and mesangial but not epithelial cells in vitro

The infectivity of human immunodeficiency virus (HIV-1) in human glomerular cells was evaluated by exposing homogeneous cultures of human glomerular capillary endothelial, mesangial and epithelial cells to HIV in vitro. Infectivity and HIV expression was assessed by: 1) the measurement of p24 antigen production from culture supernatants; 2) the presence of p24 antigen intracellularly by immunofluorescence; 3) levels of P24 antigen production or syncytia formation following the cocultivation of glomerular cells exposed to HIV with normal human peripheral blood mononuclear cells or MT-2 lymphocytes; and 4) the presence of intracellular HIV DNA by polymerase chain reaction. The results indicate that HIV can infect and replicate in glomerular capillary endothelial cells and in a small percentage of mesangial cells, but not in human glomerular epithelial cells in vitro.

Mesangial cell proteoglycans: synthesis and metabolism.

In cultures of human adult glomerular mesangial cells, large chondroitin sulfate proteoglycans (CSPG) and small dermatan sulfate proteoglycans (DSPG) are synthesized. The large CSPG has a core protein, M(r) of 400,000 (major) and M(r) of 500,000 (minor), and binds to hyaluronic acid to form large aggregates. The two small DSPGs (Mr of approximately 350,000 and M(r) of approximately 200,000) were related to biglycan and decorin, respectively. The majority of these proteoglycans were located in the culture medium, but a hydrophobic form of the CSPG was extracted from the cell layer. Mesangial cells in the growing phase synthesized and secreted all three types of proteoglycans, but in cells arrested in G0 by serum deprivation the incorporation of (35S)sulfate in CSPG was drastically reduced. In the same cells stimulated to proliferate by replacing the medium with one containing serum, the synthesis of CSPG dramatically enhanced. The synthesis of CSPG and DSPG was also elevated in cells cocultured with cytokines but in contrast was significantly reduced when cultured in medium containing hyperglycemic levels of glucose. Finally, preliminary experiments are reported that indicate that CSPG and DSPG bind to low-density lipoproteins in vitro. These observations suggest a possible specialized function for proteoglycans in cellular processes characteristic of glomerular disease.

Specialized collagen mRNA and secreted collagens in human glomerular epithelial, mesangial, and tubular cells.

Isolated glomeruli and cultured cells were examined under nonmitogenic conditions by Northern hybridization of steady-state mRNA levels for procollagens alpha 1(I), alpha 1(III), alpha 1(IV), alpha 2(IV), beta-actin, and fibronectin. Procollagens concurrently secreted from these cells were characterized after limited pepsin digestion. Poly(A)-mRNA from freshly isolated whole porcine glomeruli was primarily type IV. For cultured glomerular and tubular epithelial cells, the collagen mRNA species were almost exclusively alpha 1(IV) and alpha 2(IV). Correspondingly, the secreted collagen was almost entirely type IV. The mRNA signals for collagens in glomerular mesangial cells included alpha 1(I), alpha 1(IV), alpha 2(IV), and less alpha 1(III). The secreted collagens were also types I and IV, with less types III and V. There were similar patterns of mRNA signal levels for the two type IV collagens and similar patterns of expression of alpha 1(I) and alpha 1(III). In situ hybridization showed the fibroblast and epithelial cell populations homogeneous in expressing the same mRNA signals seen by Northern hybridization. These findings establish the correlation of collagen mRNA and protein expression of collagens in differentiated glomerular cells in culture, under resting nonmitotic conditions.

Functional endothelin 1 receptors on human glomerular podocytes and mesangial cells.

To determine if endothelin 1 (Et1) receptors are present in human glomeruli, and which glomerular cells possess these receptors, 125I Et1 binding to isolated glomeruli and cultured glomerular mesangial and epithelial cells was studied. The latter were identified as podocytes. We demonstrated that Et1 binds specifically and reversibly to isolated human glomeruli and to cultured glomerular mesangial and epithelial cells. Scatchard analysis of competitive inhibition of 125I Et1 binding gave the following results (m +/- SEM, n = 3): isolated glomeruli, Kd = 4.2 +/- 2.1 x 10(-10) M, Bmax = 8.1 +/- 1.2 x 10(10) sites/mg protein; mesangial cells, Kd = 5.2 +/- 1.5 x 10(-10) M, Bmax = 1.87 +/- 0.49 x 10(4) sites/cell; epithelial cells, Kd = 7.2 +/- 1.5 x 10(-10) M, Bmax = 2.46 +/- 0.15 x 10(4) sites/cell. These receptors seem to be functional, since in both mesangial and epithelial cells Et1 induces a rapid and transient increase in intracellular [Ca2+]i. All these results indicate that Et1 may regulate glomerular filtration rate through an autocrine-paracrine pathway on mesangial cells and on podocytes.

Characterization and Control of Expression of Cell Surface Aminopeptidase N Activity in Human Mesangial Glomerular Cells

Intact human glomerular mesangial cells in culture hydrolyzed synthetic substrates of aminopeptidase N. Mesangial cell aminopeptidase N is an ectoenzyme as demonstrated by the extracellular location of the product, the inhibition of enzyme activity by the diazonium salt of sulfanilic acid, a nonpenetrating agent, the similar activities measured with intact and sonicated cells and the localization of the immunoreactive antigen by immunofluorescence on the cell surface. A broad substrate specificity was documented with various <i>p</i>-nitroanilide substrates in the following order of affinity: Ala, Leu > Lys, Arg > Pro, Gly and Val. Using Ala <i>p</i>-nitroanilide as substrate, apparent K_m and V_max values of 0.86 ± 0.70 m<i>M</i> and 3.54 ± 0.13 nmol min^–1 mg^–1, respectively were obtained. Enzyme activity was inhibited by bestatin and 1, 10-phenanthroline. Inhibition by 1 10-phenanthroline was noncompetitive. Cell surface aminopeptidase N activity increased in the presence of phorbolmyristate acetate (PMA). This effect was time-dependent, requiring a lag time of 12 h. It was also concentration-dependent with a threshold at 1 ng/ml and a maximum stimulation (2.2 times basal value) at 10 ng/ml. The effect of PMA was prevented by cycloheximide, an inhibitor of protein synthesis, and actinomycin D, an inhibitor of RNA synthesis. It was also suppressed by H7, an inhibitor of protein kinase C activity. The presence of aminopeptidase N in cultured glomerular mesangial cells and control of its expression by mitogens suggest that this enzyme may play a role in glomerular cell proliferation.

100. Hormonal regulation of sodium-calcium countertransport in cultured human glomerular mesangial cells

100. Hormonal regulation of sodium-calcium countertransport in cultured human glomerular mesangial cells

Functional expression of NADPH oxidase components (alpha- and beta-subunits of cytochrome b558 and 45-kDa flavoprotein) by intrinsic human glomerular mesangial cells.

Recently, we have shown that human glomerular mesangial cells (HMCs) release oxygen radicals from the plasma membrane in response to cytokines. Now we have used diphenylene iodonium, a covalent binding inhibitor of activated 45-kDa flavoprotein, in neutrophils radiolabeled with 125I and could identify a 45-kDa protein band in a separated HMC plasma membrane fraction. Low temperature difference spectroscopy showed a peak absorbance at 428 and 558 nm. Direct potentiometry of HMC membranes (-340 to -160 mV) showed the presence of a low potential cytochrome (76 pmol/mg to HMC membrane protein) identified as cytochrome b558. In slot blots, mouse monoclonal antibody (mAb) 7D5, specific for the extracellular domain of the alpha-subunit, showed a positive reaction with HMCs. In Western blots, mAb 449, directed against the cytoplasmic epitope of the alpha-subunit, identified a 23-kDa protein; and mAb 48, raised against the large (beta) subunit of cytochrome b558 of human neutrophils (Verhoeven, A. J., Bolscher, B. G. J. M., Meerhof, L. J., van Zwieten, R., Keijer, J., Weening, R. S., and Roos, D. (1989) Blood 73, 1686-1694), detected a smear between 75 and 100 kDa in denatured HMC membrane protein. These data determined with HMCs, suggest for the first time the expression of three essential components of NADPH:O2- oxidoreductase in mesenchymal cells.

Cytomegalovirus infection of human kidney cells in vitro

To study which structures of a kidney allograft are the main targets for cytomegalovirus (CMV), human glomerular epithelial and mesangial cells, as well as tubular epithelial and endothelial cells were isolated by steel meshes of different pore sizes and enzymatic treatments. The various cultured cell types were characterized by morphology and specific antibodies. Human CMV was inoculated onto cell monolayers using two different culture methods: conventional tissue culture and rapid shell vial culture. To analyze whether CMV had a direct effect on the immunologic properties of kidney parenchymal cells, MHC class I and class II antigen expression was estimated before and after the infection. CMV infected all kidney cells identically. All cells expressed class I strongly after the infection, but they were class I positive prior to infection. Class II antigens were not expressed on the cell surface either before or after the infection. In conclusion, human kidney cells of glomerular, tubular and vascular origin were all infected by CMV without any difference. CMV had no significant direct effects on the antigenic properties of the cells.

Thromboxane stimulates synthesis of extracellular matrix proteins in vitro

The vasoconstrictor eicosanoid thromboxane plays an important role in the pathogenesis of several renal diseases. As an autacoid, its local release alters blood flow and induces platelet aggregation. We report a direct stimulatory effect of thromboxane on extracellular matrix protein production and gene expression in vitro. Treatment of two cell types, differentiated mouse teratocarcinoma cells (F9+) and human glomerular mesangial cells, with two different thromboxane analogues resulted in increased production of components of the extracellular matrix including fibronectin and the basement membrane proteins laminin and type IV collagen. These responses to thromboxane were not the result of a mitogenic effect of thromboxane nor the result of an increase in total cellular protein. The increased production of extracellular matrix proteins was, at least in part, due to an increase in the steady-state level of mRNA for these genes. Furthermore, the effect of thromboxane was markedly inhibited by cotreatment with a thromboxane-receptor antagonist. These results suggest a new potential role for thromboxane as a mediator of the sclerotic and fibrotic responses to injury.

Characterization of proteoglycans synthesized by human adult glomerular mesangial cells in culture

1. The newly synthesized proteoglycans from human adult glomerular mesangial cells labelled in vitro for 24 h with [35S]sulphate have been characterized using biochemical and immunological techniques. 2. The following proteoglycans were identified (% of total synthesized). (i) A large chondroitin sulphate proteoglycan, CSPG-I, Mr approximately 1 x 10(6) (10.6%). This proteoglycan consisted of a protein core of Mr approximately 4 x 10(5) and glycosaminoglycan chains of Mr 2.5 x 10(4), and was present in both the cell layer and the culture medium. (ii) A major small dermatan sulphate proteoglycan, DSPG-I, Mr 3.5 x 10(5) (46%), which was mainly located in the culture medium. (iii) A second minor small dermatan sulphate, DSPG-II, Mr approximately 2 x 10(5) (9.8%). This molecule was exclusively located in the culture medium. (iv) A large heparan sulphate proteoglycan, HSPG-I, Mr 8 x 10(5) (3.3%). (v) A second large heparan sulphate proteoglycan HSPG-II, Mr approximately 6 x 10(5) (23%). HSPG-I and HSPG-II were extracted from both the culture medium and the cell layer. 3. Western blot analysis of the core proteins released by chondroitin ABC lyase treatment of DSPG-I and DSPG-II identified these dermatan sulphate proteoglycans as biglycan and decorin respectively. Both DSPG-I and DSPG-II had core proteins of Mr 45,000. 4. The cell-layer-associated forms of CSPG-I, HSPG-I and HSPG-II were accessible to limited trypsin treatment, bound to octyl-Sepharose and could be inserted into liposomes, indicating a possible cell membrane location. 5. Pulse-chase experiments indicated that the cell-layer-associated [35S]proteoglycans undergo limited metabolism to inorganic [35S]sulphate, the majority of which is accounted for by the degradation of HSPG-II and to a lesser extent DSPG-I.

Monoclonal antibodies to cultured human glomerular mesangial cells. I. Reactivity with haematopoietic cells and normal kidney sections

The aim of this study was to produce monoclonal antibodies to cultured human glomerular mesangial cells in order to obtain specific markers for these cells and to aid the study of their function. Using standard monoclonal antibody techniques, 29 hybridomas producing antibodies directed to cultured mesangial cells were obtained. Most of these antibodies were not reactive with normal or neoplastic haematopoietic cell lines by flow cytometry. Fourteen of the 29 culture supernatants bound to various components of normal human kidney sections stained by the alkaline phosphatase/anti-alkaline phosphatase (APAAP) method. Ten of these supernatants reacted with components within the glomerulus, with six binding to the mesangium. These studies suggest that (1) mesangial cells in culture may show significant de-differentiation, because most supernatants which reacted with mesangial cells in culture did not do so in tissue sections; (2) antibodies reactive with haematopoietic cells may not detect the majority of immunogenic surface antigens on cells in tissues; and (3) some of the antibodies which we have produced may prove to be useful markers for mesangial cells in glomerular disease.

Induction of Mediator Release from Human Glomerular Mesangial Cells by the Terminal Complement Components C5b-9

Exposure of cultured human glomerular mesangial cells (GMC) to normal human serum and an activator of the complement system results in rapid uptake of the terminal complement proteins C5b-9 by the cells. This 'innocent bystander' complement attack, however, does not result in cell killing, but in the stimulation of the GMC to release prostaglandin E (PGE), interleukin 1 (Il-1) and tumor necrosis factor (TNF). Endogenously synthesized Il-1 in turn activates PGE release, indicating that the C5b-9 attack initiates an autocrine feedback stimulation. Together with the fact that C5b-9 is found in many forms of glomerulonephritis, the data point to a role of the terminal complement proteins in the initiation and perpetuation of an inflammatory response.

100. Hormonal regulation of sodium-calcium countertransport in cultured human glomerular mesangial cells

100. Hormonal regulation of sodium-calcium countertransport in cultured human glomerular mesangial cells

Identification and characteristics of a Na+/Ca2+ exchanger in cultured human mesangial cells

Excitable cells express Na+/Ca2+ exchange activity among other mechanisms modulating rapid fluctuations of cytosolic free Ca2+ ([Ca2+]i). We studied functions and regulation of a Na+/Ca2+ exchanger in cultured human glomerular mesangial cells. Fura-2-loaded confluent monolayers reacted to removal of ambient Na+ with an immediate, transient elevation of [Ca2+]i, assessed by single/dual wavelength fluorometry. Peak [Ca2+]i was inversely correlated with the extracellular Na+ concentration. Ca2+ influx was the sole mechanism implicated, as the [Ca2+]i rise was prevented by EGTA. The process was inhibited by 1 mM amiloride, but not by blockers of voltage-operated Ca2+ channels. Re-addition of Na+ resulted in a rapid decrease of [Ca2+]i, indicating bimodal operation of the exchanger. Na(+)-loading the cells with monensin and ouabain enhanced Ca2+ uptake. Prior stimulation of [Ca2+]i with the thromboxane A2 mimetic, U-46619, or angiotensin II also increased Ca2+ uptake upon subsequent Na+ removal, suggesting induction of the exchanger by vasoconstrictors. Moreover, the magnitude of agonist-induced [Ca2+]i transients was amplified by Na+ removal, indicating that the exchanger modulates the effects of vasoconstrictors. These results demonstrate that an inducible Na+/Ca2+ antiporter is operative in resting and stimulated human mesangial cells, further confirming their smooth muscle origin and potential regulatory role on glomerular hemodynamics.

Characterization of atrial natriuretic factor receptors in human glomerular epithelial and mesangial cells

To evaluate the distribution and functions of receptors of atrial natriuretic factor (ANF) in human glomeruli, we studied the binding sites of ANF-(1-28) in homogeneous populations of human glomerular epithelial cells or mesangial cells. 125I-labeled ANF bound specifically to both cell types. Equilibrium saturation binding curves suggested one group of receptor sites in mesangial cells (Kd = 99 +/- 32 pmol/l, Bmax = 15.3 +/- 3.5 fmol/mg) but multiple groups in glomerular epithelial cells. Binding was greater at 37 than at 4 degrees C in mesangial cells. The reverse was observed in glomerular epithelial cells due to marked degradation of the tracer at 37 degrees C. The fractions of undisplaceable tracer in a hypertonic acid medium after 60 min incubation were 45 and 16% at 37 degrees C for glomerular epithelial and mesangial cells, respectively. ANF-(1-28) and C-ANF-(4-23), a specific ligand of clearance receptors, similarly inhibited 125I-ANF binding to mesangial cells, whereas [Ala7-Ala23]-ANF, a linear analogue, was slightly less potent. In epithelial cells, C-ANF-(4-23) competitively inhibited 125I-ANF binding but with a lower potency than ANF, whereas linear ANF at low concentrations (10-100 pmol/l) stimulated 125I-ANF binding. In addition, linear ANF markedly inhibited the degradation of 125I-ANF in the incubation medium of epithelial and mesangial cells, whereas thiorphan, an inhibitor of enkephalinase, was inactive. ANF-(1-28) stimulated cGMP production in glomerular epithelial cells but not in mesangial cells. Both analogues were inactive in both cell types and did not modify ANF-(1-28)-dependent cGMP synthesis.(ABSTRACT TRUNCATED AT 250 WORDS)