Immunoglobulin treatment reduces atherosclerosis in apo E knockout mice.

Atherosclerosis is associated with immune activation. T cells and macrophages infiltrate atherosclerotic plaques and disease progression is associated with formation of autoantibodies to oxidized lipoproteins. In the apo E knockout mouse, a genetic model of cholesterol-induced atherosclerosis, congenital deficiency of macrophages, lymphocytes, or interferon-gamma receptors result in reduced lesion formation. We have now evaluated whether immune modulation in the adult animal affects disease development. Injections of 7-wk-old male apo E knockout mice with polyclonal immunoglobulin preparations (ivIg) during a 5-d period reduced fatty streak formation over a 2-mo period on cholesterol diet by 35%. Fibrofatty lesions induced by diet treatment for 4 mo were reduced by 50% in mice receiving ivIg after 2 mo on the diet. ivIg treatment also reduced IgM antibodies to oxidized LDL and led to inactivation of spleen and lymph node T cells. These data indicate that ivIg inhibits atherosclerosis, that it is effective both during the fatty streak and plaque phases, and that it may act by modulating T cell activity and/or antibody production. Therefore, immunomodulation may be an effective way to prevent and/or treat atherosclerosis.

Immunohistochemical study of the human glomerular C3b receptor in normal kidney and in seventy-five cases of renal diseases: loss of C3b receptor antigen in focal hyalinosis and in proliferative nephritis of systemic lupus erythematosus.

The presence and distribution of C3b receptors in normal human kidneys and in biopsies from 75 patients with renal disease were investigated by immunohistochemical techniques using monospecific rabbit antibody to the 205,000-mol wt glycoprotein that is the C3b receptor of human peripheral blood cells. Anti-C3b receptor bound exclusively to podocytes in normal renal cortex, and was homogeneously distributed on the plasma membrane of these cells. Biosynthesis of the receptor by the podocyte was suggested by the presence of antigenic activity in the Golgi apparatus. Although occupancy of receptor sites following the interaction of kidney sections with aggregated IgG preincubated with normal serum inhibited binding to glomeruli of C3b coated cells, the C3b receptor remained accessible to anti-C3b receptor antibody. No staining of podocytes was found in extra-capillary proliferating cells in rapidly progressive glomerulonephritis (GN). Segmental loss of staining was found in focal hyalinosis, nodular diabetic glomerulosclerosis, and amyloidosis while no detectable C3b receptor antigen was found in severe proliferative nephritis of systemic lupus erythematosus (SLE). Normal staining of podocytes was found in other nephropathies with endocapillary proliferation such as acute GN and mesangial GN and in renal diseases associated with immune deposits containing C3 such as mesangial proliferative and membranous SLE nephritis, idiopathic membranous GN, membranoproliferative GN types I and II, mesangial GN with IgA or C3 deposition and Henoch Schönlein's purpura. Loss of C3b receptor antigen in the diffuse proliferative nephritis of SLE distinguishes it both from nonproliferative lupus nephritis and other immunologically mediated proliferative GN.

Peritubular cells are the site of erythropoietin synthesis in the murine hypoxic kidney.

Erythropoietin (Epo)-producing cells were identified in the murine hypoxic kidney by in situ hybridization. Profound anemia was induced in order to greatly increase Epo production. This resulted in high levels of Epo mRNA in the kidney. 35S-labeled DNA fragments of the murine Epo gene were used as probes for in situ hybridization. Control experiments conducted in parallel included kidneys of nonanemic mice, RNase-treated hypoxic kidney sections, and 35S-labeled non-Epo-related DNA. The Epo probe gave a specific hybridization signal in the hypoxic kidney in the cortex and to a lesser extent in the outer medulla. Glomerular and tubular cells were not labeled. All positive cells were identified as peritubular cells. Using immunofluorescence, we showed that cells with the same topography contained Factor VIII-related antigen. These data demonstrated that peritubular cells, most likely endothelial cells, constitute the major site of Epo production in the murine hypoxic kidney.

In vivo downregulation of T helper cell 1 immune responses reduces atherogenesis in apolipoprotein E-knockout mice.

BACKGROUND: A chronic immune response involving proinflammatory T helper cell 1 (Th1) lymphocyte activation occurs in the atherosclerotic lesion, but whether this activation is protective or deleterious remains unclear. Methods and Results-- We modulated the immune response of the atherosclerosis-prone apolipoprotein E-deficient (apoE(-/-)) mouse. Eight-week-old apoE(-/-) mice were treated daily with pentoxifylline (PTX), a known inhibitor of the Th1 differentiation pathway, or PBS (control) for 4 weeks or 12 weeks. Twelve-week PTX treatment reduced atherosclerotic lesion size by 60% (P<0.01). PTX-treated mice developed lesions that were limited to the degree of fatty streaks. In contrast, control mice developed mature fibrofatty atherosclerotic lesions. In parallel, the proportion of interferon (IFN)-gamma-producing Th1 splenic lymphocytes was significantly reduced by PTX, and lesion size was correlated to the proportion of IFN-gamma(+) T cells. In vitro addition of PTX to cultured spleen cells did not modify the production of IFN-gamma but increased the production of IL-10 by T cells, indicating that PTX does not suppress IFN-gamma production but rather blocks Th1 polarization while promoting Th2 polarization. CONCLUSIONS: Thus, PTX protected mice from atherosclerosis by reducing the Th1 polarization of T helper lymphocytes. This study demonstrates that the Th1 immune response associated with atherosclerosis is deleterious and that a modulation of the Th1 differentiation pathway may provide a new pharmacological tool to treat this disease.

Early glomerular macrophage recruitment in streptozotocin-induced diabetic rats.

Diabetic glomerulosclerosis is defined by increased glomerular extracellular matrix (ECM) that is mainly synthesized by mesangial cells that underwent an activation mediated by cytokines and growth factors from various cellular origins. In this study, we tested whether macrophages could infiltrate the glomeruli and influence ECM synthesis in experimental diabetes. To test our hypothesis, we initially studied the dynamics of glomerular macrophage recruitment in streptozotocin-induced diabetic rats at days 1, 2, 4, 8, 15, and 30 by using quantitative reverse transcriptase-polymerase chain reaction (RT-PCR) on isolated glomeruli and immunohistochemistry and morphometry. We then assessed the role of macrophages on the basis of the pharmacological modulation of their recruitment by insulin or ACE inhibitor treatments and by X-irradiation-induced macrophage depletion at days 8 and 30. Macrophages were recruited within the glomeruli at the very early phase of hyperglycemia by using RT-PCR CD14 detection from day 2 and by using ED1 immunohistochemistry from day 8. This glomerular macrophage infiltration was associated with an increase in alpha1-chain type IV collagen mRNA. In parallel, the diabetic glomeruli became hypertrophic with an increase in the mesangial area. Macrophage recruitment was preceded by or associated with an increased glomerular expression of vascular cell adhesion molecule 1, intracellular adhesion molecule 1, and monocyte chemoattractant protein 1, which contributes to monocyte diapedesis. Glomerular interleukin-1beta mRNA synthesis was also enhanced as early as day 1 and could be involved in the increase in ECM and adhesion molecule gene expressions. Insulin treatment and irradiation-induced macrophage depletion completely prevented the glomerular macrophage recruitment and decreased alpha1-chain type IV collagen mRNA and mesangial area in diabetic rats, whereas ACE inhibitor treatment had an incomplete effect. It can be concluded that in the streptozotocin model, hyperglycemia is followed by an early macrophage recruitment that contributes to the molecular and structural events that could lead to glomerulosclerosis. Therefore, besides direct stimulation of mesangial cells by hyperglycemia, macrophages recruited in the glomeruli during the early phase of hyperglycemia could secondarily act on mesangial cells.

Nephrotic syndrome, linear glomerular IgG deposits, and minimal glomerular changes. Report of a case.

A young adult patient had an unusual acute idiopathic nephrotic syndrome. This nephrotic syndrome was remarkable for (1) association with acute renal failure and hypertension, (2) finding of minimal glomerular changes with a linear fixation of the anti-human IgG conjugate along the glomerular capillary wall without demonstrable antiglomerular basement membrane antibodies, and (3) complete recovery, including disappearance of the linear staining, after treatment with prednisone, cyclophosphamide, and plasmapheresis.

Long-term blood pressure changes in renal homotransplantation.

Long-term blood pressure changes were studied in 50 patients who had undergone renal homotransplantation. Excluded were those subjects with arterial stenosis of the transplanted kidney, acute or rapidly progressive rejection, or recurrent glomerulonephritis, as well as those retaining their own diseased kidney(s). The blood pressure after the end of the first year was stable and, therefore, was utilized as the reference blood pressure for this study. One year after transplantation, hypertension was observed in 20% of the patients. Mean blood pressure was positively correlated with age (P less than .01), body weight (P less than .001), and serum creatinine level (P less than .001), and negatively correlated with maintenance dose of prednisone (P less than .01). A higher incidence of hypertension was observed in cadaver kidney transplantation than in living related-donor transplantation. The study minimizes the role of glucocorticoids and emphasizes the role of renal factors in the mechanism of the long-term blood pressure changes.

Mediators of perivascular inflammation in the left ventricle of renovascular hypertensive rats.

OBJECTIVE: Inflammatory cells invade the fibrotic myocardium of spontaneously hypertensive rats at the same sites as where fibroblasts are produced. The role of these inflammatory cells in myocardial fibrogenesis was studied in the present work. METHODS: The production and distribution of proteins that may be implicated in inflammation was examined by immunohistochemistry of sections of left ventricles from 1-month and 4-month renovascular hypertensive and age-matched control rats using antibodies against ICAM-1, LFA-1, TGF beta 1, PDGF-A, T and H kininogens, IgG, IgM, C3, and C5b-9. Infiltrating inflammatory cells were phenotyped by immunohistochemistry. The TGF beta 1 and PDGF-A mRNA levels were checked by RT-PCR. RESULTS: Infiltrating cells were mainly T helper lymphocytes and macrophages, and there were more inflammatory cells in hypertensive rats than in control rats, localized especially around coronary arteries and in microscars. There were more ICAM-1 and LFA-1 in the ventricles of hypertensive than in control rats at 1 month, but the ICAM-1 expressions in hypertensive and control rats were similar at 4 months. TGF beta 1 and PDGF-A mRNA steady states increased in 4-month hypertensive rats, but there was no labeling for TGF beta or PDGF by immunohistochemistry. There was only faint labeling for T and H kininogens, and it was not increased in hypertensive rats. There were deposits of IgM and C5b-9 only in hypertensive rats. CONCLUSION: Thus, inflammatory cells infiltrate the cardiac tissue of renovascular hypertensive rats as in the case of spontaneously hypertensive rats and these cells may use the ICAM-1/LFA-1 system to infiltrate, but neither TGF beta 1 and PDGF-A, nor the kininogen system seem to be associated with cardiac fibrogenesis. Otherwise, the complement system could act as arteriosclerotic and/or leukocyte mobilizing factors.

Inflammatory cells and myocardial fibrosis: spatial and temporal distribution in renovascular hypertensive rats.

OBJECTIVE: The fibroblasts producing collagen are co-localized with inflammatory cells in myocardial fibrosis areas of spontaneously hypertensive rats, suggesting that collagen overproduction in this model may be modulated by inflammatory cells. The present study extends these observations to the Goldblatt model of hypertension in which the renin-angiotensin system is activated. METHODS: Inflammatory cells were identified with monoclonal antibodies directed against macrophages (ED1+), T helper (CD4+) and cytotoxic lymphocytes (CD8+), and MHC class II-expressing cells (Ia+). The alkaline phosphatase-anti-alkaline phosphatase (APAAP) immuno-staining technique was used. A new computer-assisted morphometric method was utilized to quantify the inflammatory infiltrate in each cardiac compartment with polarized-light microscopy. Cells responsible for the collagen synthesis were identified by in situ hybridization. The collagen content was estimated by morphometry on left ventricle sections stained with Sirius red, and by biochemical quantification of the hydroxyproline concentration. RESULTS: Computer-assisted morphometry under polarized light was well suited to quantify inflammatory cells labeled by the APAAP technique. Inflammatory cells were co-localized with collagen-synthesizing fibroblasts. The main inflammatory cells were CD4+ lymphocytes > Ia+ > ED1+ > CD8+ cells. These cell densities were increased in hypertensive rats in all cardiac areas compared to control rats except for IA+ cells which were concentrated in microscars. Macrophage density was correlated with plasma renin activity. The inflammatory cell density which best correlated with fibrosis was macrophage density, and which best correlated with systolic blood pressure was macrophage and T helper lymphocyte densities. CONCLUSIONS: One can speculate that the correlation between macrophage density and blood pressure as well as with plasma renin activity may indicate that angiotensins and/or elevation of blood pressure could participate in the initial signalling which may mobilize inflammatory cells. These inflammatory cells could promote fibrosis by releasing mediators such as growth factors or cytokines which act upon fibroblasts.

Left ventricular fibrosis in renovascular hypertensive rats. Effect of losartan and spironolactone.

Myocardial fibrosis resulting from arterial hypertension alters myocardial structure and function. Myocardial fibrosis is characterized by a pathological accumulation of types I and III collagens. We used an aldosterone antagonist (spironolactone) and an angiotensin II antagonist (losartan) to elucidate the respective role of these hormones and hypertension in the development of myocardial fibrosis in the Goldblatt model of two-kidney, one clip hypertension in the rat. Fibrosis was assessed by computer-assisted morphometry in the interstitial space, around coronary arteries, in microscar areas, and on left ventricular sections stained with Sirius red and by biochemical techniques. Morphometry was performed with both standard light and polarization microscopy; this latter method was used to quantify yellow-red and green collagen fibers. Concurrently, type I and type III collagen mRNAs were evaluated by a semiquantitative polymerase chain reaction method. The collagen content of the untreated two-kidney, one clip hypertensive rats increased mainly around the coronary arteries; the number and surface area of microscars also increased in chronic hypertension. Losartan treatment decreased systolic pressure and yellow-red collagen fiber content in all areas, whereas spironolactone treatment decreased green collagen fiber content without decreasing systolic pressure. mRNA levels for types I and III collagens showed profiles similar to those of yellow-red and green collagen fiber contents, respectively, suggesting that yellow-red collagen fibers are mainly type I collagen fibers and green collagen fibers are mainly type III collagen fibers. These results suggest that angiotensin II, possibly together with hypertension, and aldosterone, independently of hypertension, have a major influence on myocardial fibrosis, inducing type I and type III collagen deposits, respectively, mainly around coronary arteries.

Glomerular capillary network of cortical nephrons is reduced in male but not in female aging rats.

The gender differences in the age-related changes of glomerular structures were determined in 10- and 30-month-old rats. In adult animals, glomerular volume, urinary space, capillary lumen area and mesangial domains of deep and superficial nephrons were larger in males than in females. Glomerular hypertrophy was evidenced with age in both males and females. This hypertrophy was greater in female (+70%) than in male (+20%) rats. Age-related hypertrophy concerned equally the urinary space and the glomerular tuft. The mesangial domain, however, increased more markedly than glomerular volume (+400%). As a result, the ratio of mesangial domain to glomerular section area was more than doubled between 10 and 30 months. In females, the age-related renal hypertrophy was associated with a constant total capillary lumen area in cortical nephrons. In contrast, the total capillary lumen area of male rats was reduced by 20% in superficial glomeruli and by 36% in deep glomeruli between 10 and 30 months. These morphological changes are in good agreement with the maintained glomerular filtration rate reported in old female rats and the decrease in renal blood flow and filtration rate reported in male rats. They suggest that the aging process does not similarly affect the vascular system of the kidney of male and female rats, although their mean blood pressure was comparable.

Suppression of HLA-specific alloantibodies by high-dose intravenous immunoglobulins (IVIg). A potential tool for transplantation of immunized patients.

Renal transplantation in patients presenting end-stage renal failure can be hampered by the presence of alloantibodies against HLA antigens. In 4 out of 5 patients with HLA-specific alloantibodies waiting for a renal allograft, treatment with high-dose i.v. Ig resulted in a prolonged suppression (over 3 months) of most of the panel-reactive anti-HLA antibodies (PRA). Intravenous polyclonal human Ig (IVIg) and F(ab')2 fragments from IVIg inhibited the binding of patients' plasma and IgG fractions to peripheral blood lymphocytes from normal donors as well as their cytotoxicity, suggesting that the in vivo effect of IVIg was mediated by the presence, in the IVIg preparation, of anti-idiotypes directed against idiotypes borne on the anti-HLA antibodies. Thus, treatment with IVIg can be a valuable tool toward the transplantation of immunized patients.

Presensitization accelerates allograft arteriosclerosis.

Transplant arteriosclerosis is the major factor influencing allograft survival after the first year posttransplantation. The host's immunologic response is one of the principal effectors responsible for the constitution of this vascular wall lesion, but the effector pathway and the factors influencing the immune injury are not clear. In a rat abdominal aortic allograft model, we used a skin priming method to study the influence of sensitization on the occurrence of vascular wall lesions. Primed rats developed transplant arteriosclerosis lesions involving medial decellularization and intimal proliferation before the 21st day, whereas naive animals had the same lesions at 2 months posttransplantation. A significant difference between primed and naive rats was found for medial thickness (48.00 +/- 2.85 microm versus 79.34 +/- 2.55 microm, P<0.001) and smooth muscle cell content (160 +/- 28 cell/mm versus 466 +/- 19 cell/mm, P<0.001) at 21 days posttransplantation, and intimal hyperplasia was seen in primed animals at that time, whereas it was not observed in naive rats until the 60th day. The immune profile in naive and primed animals was different. The immune cells infiltrating the arterial wall in naive rats, were principally macrophages and CD8+ T-lymphocytes. No Ig or complement deposition was detected. IgG and complement activated fraction were present in the media of primed animals as early as the fifth day posttransplantation and CD4+ T lymphocytes were the dominant immune cell population. In conclusion, sensitization influences the immune mechanisms responsible for the development of transplant arteriosclerosis and alters the rate of its evolution.

Sequential immunological targeting of chronic experimental arterial allograft.

Arterial wall is the main site involved in the chronic rejection process. The rat aortic allograft model was used here to characterize and describe the sequential evolution of the different targets and effectors of arterial wall immunological injury and response during arterial allograft rejection. Rat abdominal aortae were isografted or allografted from Brown-Norway to Lewis rats. Endothelial and smooth muscle cell injury and humoral and cellular immunological effectors were characterized from 0 to 60 days after transplantation using a battery of specific antibodies. The intimal proliferative response was also characterized over this time. Isografted Brown-Norway aorta adventitia had very few cellular components, which suggests that donor adventitia would be poorly antigenic in allografts. In contrast, allograft adventitia was the site of a major inflammatory cell invasion in which the expression of an adhesion molecule by colonizing capillary endothelial cells could play a main role. This adventitial infiltration continued as long as medial smooth muscle persisted. The luminal endothelial cells disappeared early, probably associated with macrophage margination. In contrast, medial smooth muscle cell disappearance occurred later and was specifically targeted by immunoglobulins. Intimal proliferation was the most delayed phenomenon, involving both inflammatory cell infiltration at an early stage and later myofibroblastic proliferation, and could be related to the specific expression of growth factors in this layer. The rat aortic allograft model appeared useful for characterizing specific targets and effectors of chronic arterial graft rejection, demonstrating an early stage of endothelial injury and the presence of immunoglobulins involved in chronic medial smooth muscle cell injury.

A comparative ultrastructural localization of concanavalin A, wheat germ and Ricinus communis on glomeruli of normal rat kidney.

Three lectins, concanavalin A, peroxidase labeled wheat germ and peroxidase labeled Ricinus communis have been utilized to determine the localization of various saccharide determinants in the glomerulus of the normal rat kidney. In order to obtain a homogeneous penetration of the lectins throughout the whole section, various technical parameters were studied. Only with concanavalin A, a diffuse labeling of the endoplasmic reticulum was obtained. With the three lectins, the basement membrane, mesangial matrix and the cell coat of the three cell types in the glomerulus were labeled. Differences in the labeling of the capillary wall were also noted.

Podocytes in aminonucleoside glomerulonephritis investigated ultrastructurally with concanavalin A.

Peroxidase labeled concanavalin A (Con A) permits the detection of some saccharide determinants. This histochemical technique permits the visualization of cellular pathological modifications not observed with other methods. Its use in an ultrastructural study of nephritis induced by a single injection of aminonucleoside demonstrated the following in podocytes. The Con A positive endoplasmic reticulum (ER), essentially the rough ER, lost its normal linear and network appearance to take on a dot dash pattern. ER contents but not attached ribosomes and membranes were Con A positive. The dot dash pattern, due to a fragmentation of the ER, appeared prior to the onset of proteinuria and was attenuated before the disappearance of proteinuria. These changes of the ER were not observed in other proteinuric states. This suggests that aminonucleoside can damage the synthesis apparatus of podocytes, revealed by the Con A method.

Use of a specific antiserum for renin detection in human kidney.

The use of anti-human renin antibodies made possible the intrarenal localization of renin in human kidney by immunofluorescence. In normal kidney, only some juxtaglomerular apparatus (JGA) were fluorescent. In these JGA, granular or diffuse fluorescence was only seen in afferent arterioles and was not present in all cells. In the ischemic areas of partially infarcted kidney, fluorescence was seen in all JGA and in interlobular arteries. In these arteries the most eccentric cells were often the most positive. In the nonischemic areas of the same kidneys, fluorescence was not seen in JGA, but was observed in proximal tubular cells, suggesting the reabsorption of filtered renin at this site.

Freeze-fracture cytochemistry of rat glomerular capillary tuft. Determination of wheat germ agglutinin binding sites and localization of anionic charges.

We propose here the use of freeze-fracture to gain access and to label in vitro glomerular components and locate WGA receptors and anionic sites. Tissues are frozen, fractured under liquid nitrogen, and thawed. Freeze-fracture rendered all glomerular structures directly accessible to the reagents. This made possible study of the nature and topology of cationized ferritin and WGA binding sites. WGA-gold complexes were observed over plasma membranes of podocytes and of endothelial and mesangial cells. Labeling of podocytes and endothelial cells was similar in the mesangial area and in the peripheral part of the capillary loop. Cross-fractures of extracellular matrices showed that WGA bound uniformly to the glomerular basement membrane (GBM) as well as to mesangial matrix. In fractured specimens treated with neuraminidase, WGA was no longer observed over podocytes but it consistently labeled the surface of endothelial and mesangial cells. Whereas in GBM cross-sections WGA binding was greatly reduced or even abolished, it remained unmodified in the mesangium. This shows that only NeuNAc (sialic acid) might account for the binding of WGA to podocytes, whereas GlcNAcs appear to be the main WGA binding sites on endothelial and mesangial cells and in the mesangial matrix. Both NeuNAc and GLcNAc residues are probably associated in GBM. With cationized ferritin (pI 8.3) at pH 7.4, intense, continuous labeling was seen all over the different plasma membranes, denser in podocytes than in endothelial cells. CF was also observed in cross-fractured profiles of extracellular matrices and never appeared agglutinated in discrete sites.

Histochemical and ultrastructural characterization of subendothelial glycoprotein microfibrils interacting with platelets.

The interaction of human blood platelets with collagenase-treated rabbit subendothelium was studied by histochemical ultrastructural methods and by morphometric semi-quantitative analysis. Aortas were deendothelialized and incubated: 1) with a highly purified bacterial collagenase whose specificity was controlled; and 2) with the same collagenase followed by chymotrypsin. For histochemical studies, tannic acid, ruthenium red, and peroxidase-labeled Ricinus communis and concanavalin A were used. Electron microscopy showed that after digestion of fibrillar collagen by collagenase, adherent and aggregated platelets were observed on Ricinus communis-, concanavalin A-, and ruthenium red-positive glycoprotein microfibrils. After successive incubation with collagenase and chymotrypsin, the microfibrils disappeared. No platelets were observed on the remnant amorphous elastin. Morphometric analysis confirmed the interaction of platelets with collagenase-treated subendothelium. In addition, glycoproteins were extracted from collagenase-treated rabbit aortas using 5 M guanidine. Using an in vitro quantitative test, significant platelet adhesion to these glycoproteins was observed. Our results show an interaction between platelets and noncollagenic glycoprotein microfibrils.

Effects of captopril on prostaglandin and natriuresis in patients with essential hypertension.

The antihypertensive, renal and hormonal effects of captopril were studied in 10 patients with essential hypertension. Captopril significantly decreased arterial blood pressure with a concomitant increase in glomerular filtration rate, natriuresis and kaliuresis and a significant selective increase in urinary (renal) prostaglandin E2; other plasma and urinary prostaglandin (F2 alpha, 6-keto-prostaglandin F1 alpha; thromboxane B2) were not significantly changed. The urinary prostaglandin E2 increase was observed even in patients with pretreatment subnormal prostaglandin E2 excretion. Increases in urinary prostaglandin E2 were significantly positively correlated with increases in urinary sodium concentration. It is concluded that the antihypertensive effect of captopril is mediated, at least partially, by prostaglandin E2 release from renal and extrarenal tissues. Captopril enhances natriuresis at a lower perfusion pressure.