Intracellular surfactant removal from phagocytized minerals: development of a fluorescent method using a BODIPY-labeled phospholipid.

Lung surfactant serves as a protective coating when adsorbed on particle surfaces, so its removal or rate of removal in vivo may affect expression of mineral cytotoxicity. Removal of phospholipid surfactant components from the surface of mineral particles ingested by alveolar macrophages (AM) was measured using fluorescence microscopy. Dipalmitoylphosphatidylcholine with a fluorescent label (BODIPY(trade mark)) substituted for C1-C4 on the second acyl chain (DPPC*), was mixed with dioleoylphosphatidylcholine (DOPC) to coat respirable quartz and kaolin particles. Fluorescence from quartz or kaolin particles of 3-4, 5-6 and 8-9 microm size decreased in intensity with increasing ratios of DOPC/DPPC* for the same DOPC concentration of 0.4 mg/ml. There was a direct correlation between fluorescence and residual phospholipid surfactant remaining on particles using phospholipase A2 (PLA(2)) digestion in a cell-free system, indicating that the presence of the fluorophore on DPPC did not hinder enzymatic recognition. Lavaged primary AM obtained from male Fischer rats were challenged in vitro with DOPC/DPPC* (10:1 mol:mol) coated particles at 50 microg particles/10(6) cells. In contrast to the biexponential response seen in cell-free experiments, the rate of fluorescence decay from ingested coated quartz or kaolin particles over 7 days was monoexponential, with the same t(1/2) (41 h) for each dust. This study suggests that the rate of phagolysosomal digestion and removal of the adsorbed surfactant is not a determinant of the different mineral-specific pathogenicities or toxicities of quartz and kaolin, although residual fluorescence remained on particles even after 7-8 days.

Role for nitric oxide but not prostaglandins in acetylcholine-induced relaxation of rat cremaster third-order arterioles in 5-hour ischemia-reperfusion control rats.

Intravital videomicroscopy was used for 1 hr in anesthetized 4- to 5-week-old rats, while mean femoral arterial blood pressure and suffusate Po2 were continuously monitored. The total duration for experimentation was 5 hr in order to mimic the controls used previously for a 4-hr ischemia and 1-hr reperfusion model. The specific aim was to examine further the efficacy of this model by (a) assessing the potential role(s) of nitric oxide (NO) and/or prostaglandins (PG) in acetylcholine (ACh)-induced relaxation, and (b) determining if inherently low vasomotor tone (VT) and wall shear stress (WSS) mask latent NO- and/or PG-mediated responses to ACh. Reactivity to 10(-4) M ACh or 10(-6) M sodium nitroprusside (NP) were determined in resting third-order arterioles (3A) or in those preconstricted with norepinephrine (10(-6) M NE) at physiological suffusate Po2 (25-30 mm Hg). Repeated and randomized topical administrations of ACh, NP, arachidonic acid (10(-5) M AA), NE, or 10(-5) M atropine (ATR), NG-nitro-L-arginine methyl ester (L-NAME), or ibuprofen (Ib) alone or in combination to the surface of exteriorized cremaster flaps, provoked no alteration in mean systemic arterial blood pressure. ACh and NP were equipotent evoking relaxations on the same order of magnitude and duration as reported previously for arterioles with spontaneous or NE-enhanced VT. ATR or L-NAME decreased resting internal diameter by 12 to 14% and reversed relaxation of resting or preconstricted arterioles to ACh but not to NP. Ib failed to elicit blockade. However, administration of AA demonstrated Ib-inhibitable increases of 20 and 52% in resting and NE-preconstricted arterioles, respectively, implicating NO but not PG in the regulation of resting (relaxant) tone and in ACh-induced dilation following activation of 3A muscarinic receptors. The absence of PG-mediated responses appears unrelated to low initial WSS at physiologic suffusate Po2, since NE-induced elevations of VT and centerline cell velocity also did not cause Ib-inhibitable relaxation. These and our previous findings suggest that the impaired relaxant function of 3A arterioles is caused in part by a paucity in spontaneous tone inherent to this model and by depressed vasoreactivity arising from disruption of NO biosynthesis.

Phospholipid surfactant adsorption by respirable quartz and in vitro expression of cytotoxicity and DNA damage.

Respirable-sized quartz was treated with a saline dispersion of dipalmitoyl phosphatidylcholine (DPPC), a primary component of pulmonary surfactant, to model the adsorption of phospholipid surfactant onto quartz dust following particle deposition in the bronchoalveolar region of the lung. Control and surfactant-treated dusts were used to challenge lavaged rat pulmonary macrophages in vitro over a 1-week period, to determine the effects of adsorbed surfactant on the expression of quartz cytotoxicity and genotoxicity. DNA damage was determined by the single cell gel electrophoresis 'comet' assay. Untreated quartz induced DNA damage, increasing with dose and with time of incubation of dust with macrophages over a 5 day period. DPPC treatment of quartz suppressed DNA damage through 1 day of macrophage challenge. DNA damage then increased over a 5 day period, to approximately half the positive control (untreated quartz) values. Cytotoxicity was measured by trypan blue dye exclusion and by the Live-Dead fluorescence assay for cell viability. Cytotoxicity of surfactant-treated quartz measured one day after challenge of lavaged macrophages was suppressed to values near those of the negative controls, and then increased over a 1 week incubation period to levels near those expressed by native quartz positive controls. Quartz similarly treated with dioleoyl phosphatidylcholine mixed with DPPC substituted in one acyl group with a boron-containing fluorescent chromophore was used with confocal microscopy to measure particle-associated fluorescent surfactant in cells. Approximately half of the fluorescence intensity was lost over a 1 week period following challenge of lavaged macrophage. Results are discussed in terms of a model of restoration of quartz particle surface toxicity as prophylactic surfactant is removed from particle surface by cellular enzymatic digestion processes.

The effects of two levels of vasomotor tone at physiologic suffusate PO2 on acetylcholine- and sodium nitroprusside-induced relaxation of cremaster third-order arterioles in 5-hour ischemia-reperfusion control rats.

Intravital microscopy was used for 1 h in the cremaster skeletal muscles of normotensive 4- to 5-week-old rats. The total duration for experimentation was 5 h in order to mimic the controls used previously for a 4-hour ischemia and 1-hour reperfusion model which was equilibrated with room air. Responsiveness of third-order (3A) arterioles with resting vasomotor tone (VT) was assessed to topically applied 10(-2) to 10(-6) M acetylcholine (ACh) or sodium nitroprusside (NP) using a suffusate PO2 of 25-30 mm Hg. ACh (10(-4) M) or NP (10(-6) M) were retested at this Po2 in 3A arterioles with norepinephrine (NE) (10(-6) M)-enhanced VT. Results were compared against those using room air to increase VT. No dose-response relationships were demonstrated for ACh or NP in resting conditions. Moreover, our current and former responses were maximal and of a lesser magnitude than those reported by others using room air. All doses except 10(-4) or 10(-6) M ACh, or 10(-6) M NP, also severely depressed systemic arterial blood pressure. Enhancement of VT by 29% resulted in a 3-fold greater dilation to 10(-4) M ACh or 10(-6) M NP. The relative increase in volumetric blood flow (Q) to ACh or NP was 5.3-5.7 times greater than with resting VT. However, there were no differences in the absolute maximal values attained for internal diameter (D) or Q between drugs. The times to peak response and recovery were accelerated for Q but not D in NE-preconstricted arterioles, and the relative increases in D were less than reported by others after equilibration with room air. At both levels of tone, topical administration of NE at the end of each experiment caused similar decreases in D and Q, while mean centerline cellular velocity, wall shear rate (WSR), and VT were found to increase. These results suggested that the relative increases in D and Q were due to NE-induced decreases in absolute predrug baselines, while both the depressed VT and peak responses in D were caused by a fall in vasoreactivity. Po2 did not appear to be a factor attenuating endothelium-dependent responses, since ACh was equipotent to NP in resting or NE-preconstricted arterioles at physiological suffusate Po2 or in cremaster flaps equilibrated with room air. Therefore, either NE at physiological suffusate Po2 or room air appears acceptable for elevating initial VT/WSR when examining the endothelium-dependent and endothelium-independent mechanisms regulating physiological (dilator) tone and perfusion in 3A arterioles.

Hepatic microvascular regulatory mechanisms. XII. Effects of 5-HT2-receptor blockade on serotonin-induced intralobular hypoperfusion.

A 5-HT2-receptor antagonist (LY53857) was evaluated in the livers of male Sprague-Dawley rats receiving a dose of serotonin (5-HT) producing systemic (arterial) hypotension and low flow. The specific aim of this investigation was to determine the cross-blocking potential of a 5-HT2 blocker having low affinity for alpha-adrenergic and histamine H 1-receptors. This was a follow-up study to one which characterized the effects of normo- and hypotensive doses of 5-HT on intralobular perfusion and volumetric rates of blood flow at the inlet of periportal and outlet of centrivenous sinusoids. Twenty rats were injected intraperitoneally (i.p.) with 0.05 mg per g b.w. pentobarbital five min following an i.p. injection of 0.1 mg per 100 g b.w. LY53857. The left lobes of the livers from these rats were exteriorized and examined with intravital videomicroscopic and electro-optical methods following surgical implantation of a catheter into the ileocecal vein. This venous catheter served as the route for endoportal infusion of hypotensive dose of 5-HT (10 micrograms per 100 g b.w.) in 10 rats, while the remaining 10 rats were given an equivalent volume of its carrier as a control (0.1 ml per 100 g b.w. Ringer's solution). Injection of LY53857 completely antagonized 5-HT-elicited low flow at the inlet of periportal and outlet of centrivenous sinusoids. In addition, no change in sinusoidal internal diameter was observed following blockade of 5-HT2 receptors. These results, in the concert with those from previous studies characterizing 5-HT vasoresponsiveness in the liver, suggest that: a) constrictor 5-HT2 receptors are localized on hepatic sinusoids, and b) 5-HT-provoked hypoperfusion is mediated by activation of the 5-HT2-receptor subtype.

Heterogeneity in superoxide production as measured by nitro blue tetrazolium reduction from individual PAM.

Heterogeneity in superoxide (O2-.) production (as determined by the reduction of nitro blue tetrazolium to a diformazan precipitate), cell volume, and cell spreading were measured from single rat pulmonary alveolar macrophages (PAM). Of the 330 cells that produced O2-. due to stimulation by adherence, the maximum diformazan produced (MAX) varied between 1.8 and 47.2 fmol and the maximum (initial) rate of diformazan production (R) varied between 2.2 and 81.7 X 10(-3) fmol/s. Only six PAM of 336 analyzed failed to produce O2-. suggesting that O2-. production is not a specific function of a small subpopulation of PAM but rather that all PAM are capable of producing O2-.. Importantly, O2-. production by single cells showed extensive heterogeneity that did not correlate (r2 less than 0.12) with individual cell volumes, suggesting that the observed heterogeneity may not be due to differences in maturation and/or differentiation. However, when the single cell data were grouped into cell volume subfractions small but increasing linear trends (r2 greater than 0.81) in MAX and R were found. This discrepancy suggests that PAM heterogeneity in O2-. fraction cannot be estimated adequately by measurements on subpopulations of cells.

Hepatic microvascular regulatory mechanisms. XI. Effects of serotonin on intralobular perfusion and volumetric flowrates at the inlet of periportal and outlet of centrivenous sinusoids.

A normotensive (1.0 microgram per 100 g b.w.) or hypotensive (10.0 micrograms per 100 g b.w.) dose of serotonin (5-HT) was administered endoportally while changes in microcirculation at the inlet and outlet regions of hepatic lobules were measured on-line using quantitative in vivo microscopy. The number of sinusoids with decreased (cellular) flow also was counted to index intralobular perfusion in video recordings of microvasculature examined off-line. The normotensive and hypotensive doses of 5-HT elicited decreases in intralobular perfusion within periportal and centrivenous sinusoids. Hypoperfusion was accompanied by a transient decrease in volumetric flowrate (Q) at the outlet of centrivenous sinusoids in 40% of normotensive and in 100% of hypotensive rats. At the inlet of periportal sinusoids, Q was depressed in 75% of hypotensive and in 27% of normotensive rats. The remainder of these segments had either an increase or no change in Q at the inlet and outlet. These results suggested that during conditions of 5-HT induced (lobular) hypoperfusion: (a) Q at the inlet is maintained in 73% of normotensive rats by redistribution of intralobular blood flow, and decreased in all but 25% of hypotensive rats as a function of transient reductions in total hepatic (arterial) and/or portal (venous) blood flow(s), and (b) Q at the outlet is depressed in 40% of normotensive rats by apparent increases in flow redistribution and resistance to flow generated during sinusoidal constriction, whereas in all hypotensive rats this mechanism is aggravated by decreased total hepatic (arterial) and/or portal (venous) inflow(s). Therefore, although the initial time course for microvascular responses tended to be similar for normo- and hypo-tensive doses of 5-HT, quantitative differences in regional flow distribution and Q emphasize (a) the importance of intra- and extra-hepatic determinants in the regulation of blood flow within hepatic (unit) lobules, and (b) the presence of microvascular heterogeneity within these lobular units.

A kinetic model of superoxide production from single pulmonary alveolar macrophages.

A kinetic model was developed to describe the production of superoxide (O2-) by single pulmonary alveolar macrophages (PAM). Model predictions were compared with experimental results obtained from single rat PAM. The O2- was quantified by measuring the reduction of nitro blue tetrazolium (NBT) to a diformazan precipitate (NBTH2) from video-recorded images of individual cells. The kinetic model considered three reactions: 1) the production of extracellular O2- from the reduction of oxygen by NADPH oxidase using intracellular NADPH as the substrate, 2) the subsequent dismutation of O2- to form H2O2, and 3) the reaction of O2- and NBT to form diformazan. NBT specificity for O2- was analyzed by comparing results in the presence and absence of superoxide dismutase (SOD) that catalyzes the dismutation of O2- to H2O2. Measured PAM heterogeneity was accounted for by varying the concentration of intracellular NADPH, its rate of depletion, and the concentration of intracellular NADPH oxidase in the kinetic model. Model predictions compared favorably with experimental results except when SOD was present. This discrepancy may be due to diffusional limitations because NBT is a relatively small molecule (818 mol wt) compared with SOD (34,000 mol wt). In addition, the cell surface is both ruffled and negatively charged, which may introduce steric hindrances and/or electrostatic effects, since SOD is also negatively charged, whereas NBT is positively charged.

A quantitative study of fluorescein isothiocyanate-dextran transport in the microcirculation of the isolated perfused rat liver.

Hepatic extraction of solutes depends on microvascular angioarchitecture, hemodynamics and solute concentrations. These factors may contribute to the heterogeneity observed in solute transport and uptake in the hepatic lobules. However, predictions of liver extraction based on black-box models require assumptions about these factors and the microvascular transport mechanisms involved. Consequently, the purpose of this study was to investigate solute transport and uptake by hepatocytes. Livers from male Sprague-Dawley rats were perfused at physiological flowrates and portal pressures on the stage of an in vivo microscope using a low-hematocrit Ringer solution. A bolus of fluorescein isothiocyanate-dextrans (17,900, 39,000, 65,600 or 156,900 MW), which are considered inert fluid-phase markers, was injected into the portal vein. Fluorescein isothiocyanate fluorescence, as a measure of solute concentration, was video recorded in periportal or centrivenular regions of the lobules. Spatial and temporal fluorescence data, measured in sinusoids and hepatocytes, were fit to one-dimensional transport models to determine estimates for an intracellular effective diffusion coefficient and for hepatocyte permeability. The calculated effective diffusion coefficients were 2.5 times larger for dextrans less than 66,000 MW, but were not different between the periportal and centrivenular regions. Also, the values did not show the inverse log-log molecular weight dependency for dextrans seen in other microvascular tissues.(ABSTRACT TRUNCATED AT 250 WORDS)

Hepatic microvascular regulatory mechanisms. X. Effects of alpha-one or -two adrenoceptor blockade on glucoregulation in normotensive endotoxic rats with optimal perfusion and flowrates.

Circulating-blood glucose, hepatic glycogen distribution, and the glycogen contents of liver and skeletal muscle, were determined for 60 min in 31 fed and anesthetized Sprague-Dawley rats. These rats received an endoportal infusion of 15 mg per kg b.w. E. coli endotoxin (026:B6) or of sterile saline solution as a control. Either substance was given intravenously at 9:30 a.m. following an intraperitoneal injection at 9:00 a.m. of 0.1 mg per kg b.w. prazosin or 0.3 mg per kg b.w. yohimbine or of the carrier, distilled water. Infused endotoxin elevated blood glucose without affecting hepatic glycogen distribution and total glycogen contents of liver and skeletal muscle when compared to control. Prazosin inhibited endotoxin-induced hyperglycemia, and prazosin plus endotoxin provoked centrilobular glycogen depletion and decreased total hepatic glycogen content. However, no significant alteration in the glycogen content of skeletal muscle accompanied blockade of glucogenesis. Prazosin administered by itself produced no changes in hepatic and muscle glycogen. Although yohimbine blocked endotoxin-induced hyperglycemia, yohimbine, or yohimbine plus endotoxin, produced no significant change in the glycogen contents of liver and skeletal muscle. Blockade in the latter case was associated with some depletion of glycogen in hepatocytes dispersed randomly throughout the unit lobule and in cells located centrivenously. These results suggested that endotoxin-induced hyperglycemia is evoked by activation of alpha-1 and -2 adrenergic receptors. Since no detectible change in hepatic glycogen distribution and in the contents of liver and muscle glycogen accompanied glucogenesis, glycogen catabolism and deposition are postulated to proceed simultaneously and at equivalent rates by 60 min following the experimental induction of endotoxemia. Blockade of alpha (one or two) adrenoceptors is hypothesized to inhibit endotoxin-induced hyperglycemia by facilitating glucose utilization and not by stimulating glycogenesis or by antagonizing glycogenolysis in the liver or skeletal muscle.

Measurement of superoxide release from single pulmonary alveolar macrophages.

An electrooptical method was developed to quantify superoxide (O2-) release from single rat pulmonary alveolar macrophages (PAM) during adherence to the bottom of a culture dish. This was done by measuring the reduction of nitro blue tetrazolium (NBT) to a diformazan precipitate at 550 nm from videorecorded images of individual cells. Temporal changes in cell optical density, which are proportional to the mass of diformazan produced, were calculated from videophotometric measurements of the change in light intensity over individual cells. Total diformazan produced increased 78 and 126% with an increase in NBT from 0.5 to 1.0 and 2.0 mg/ml, respectively. Total diformazan produced and maximum rate of production among individual PAM varied two- to threefold providing strong evidence for heterogeneity in O2- production. Specific inhibition of O2- production by superoxide dismutase, iodoacetate, and chlorpromazine significantly reduced the total diformazan produced and maximum rate of diformazan production. Hydrogen peroxide was not involved in NBT reduction, since catalase alone did not significantly change diformazan production. This novel method to quantify O2- release from single PAM should be valuable in analyzing heterogeneity and single cell kinetics of O2- production, in assessing the effects of exposure of cells to particulates on O2- release, and in relating release to electrophysiological measurements.

Hepatic microvascular regulatory mechanisms. V. Effects of lodoxamide tromethamine or phentolamine-HCl on vascular responses elicited by serotonin.

Initial changes in hepatic microvasculature and carotid (arterial) blood pressure were measured in anesthetized Sprague-Dawley rats receiving an endoportal injection or topical application of serotonin alone, or in combination with selected doses of lodoxamide tromethamine or phentolamine-HCl. Although infusion of 1.0 microgram per 100 g b.w. serotonin and/or 10 micrograms per 100 g b.w. lodoxamide produced no change in arterial blood pressure, 10 or 20 micrograms per 100 g b.w. serotonin evoked hypotension within 40 sec. During this period, in vivo microscopy of transilluminated livers revealed that 1.0 or 10 micrograms per 100 b.w. serotonin elicited constriction of portal venules and sinusoids, and a significant increase in the percentage of these microvessels containing decreased cellular flow. None of these responses was antagonized by 10 micrograms per 100 g b.w. lodoxamide. Topical application of 1.0, 10, or 20 micrograms serotonin provoked constriction of sinusoids and decreased sinusoidal and central venous perfusion, but no systemic hypotension. There was no significant difference in the magnitude of responses to these doses of serotonin alone or with 0.1 microgram lodoxamide; however, 100 micrograms phentolamine potentiated responses to 1.0 microgram serotonin. Given these results, and those of previous studies indicating that equivalent doses of these antagonists block mast cell degranulation or catecholamine-induced activation of alpha receptors, it is suggested that the vascular responses are mediated directly by serotonin and not indirectly by serotonin-evoked mast cell degranulation or alpha-adrenergic stimulation. The vascular receptor(s) triggered remain(s) to be elucidated.

A compartmental analysis of the splenic circulation in rat.

Isolated, acellular washout experiments of 125I-labeled bovine serum albumin (BSA) from control, anemic, and polycythemic rat spleens were used to develop a model of the splenic plasma circulation. The results indicated that the plasma circulation can be described adequately by two compartments. As in red blood cell (RBC) washouts [Am. J. Physiol. 239 (Heart Circ. Physiol. 8): H272-H277, 1980] the fast compartment represents intrasplenic vessels that bypass the red pulp, whereas the intermediate/slow compartment represents plasma flow through the red pulp (filter). The combined plasma and RBC parameters suggest the rat spleen is not an RBC reservoir and that splenic RBC filtration capacity decreases during polycythemia and anemia. The ratio of fast compartment to systemic hematocrit indicates hemodilution occurs, supporting the concept of plasma skimming. A small plasma holdup occurs in the red pulp of anemic and polycythemic spleens probably due to RBC congestion. This congestion, in turn, might be due to reticulocyte sequestration and/or erythropoiesis in anemic spleens and RBC sequestration and/or destruction in polycythemic spleens. There is plasma redistribution in polycythemic spleens possibly to meet the increased metabolic demand.

Hepatic microvascular regulatory mechanisms. III. Aminergic mechanisms as related to mast cells.

The responses of the hepatic microvasculature to aminergic stimulation were evaluated in Sprague-Dawley rats anesthetized with urethane or pentobarbital. Various concentrations (10(-10)-10(-4) g/ml) of serotonin, histamine or compound 48/80 (a mast cell degranulator) alone or in combination with appropriate blockers were administered topically to the livers of these rats while changes in the microvasculature were measured for a period of 15 min using in vivo microscopic methods. The influence of compound 48/80 on hepatic mast cells was verified by histochemical methods. Histamine induced dilatation of portal venules which was not antagonized by diphenhydramine (H1-blocker) or metiamide (H2-blocker). Constriction was produced by serotonin in sinusoids and by 48/80 in sinusoids and central venules, but not in other parts of the microvasculature. Vasoconstriction was accompanied by enlargement and bulging of the nuclear region of cells lining sinusoids. Concomitantly, platelets and leukocytes adhered to the endothelium of sinusoids, and central and sublobular venules. None of these responses was antagonized by methysergide. Given these results and histochemical demonstration of the release of mast cell constituents following 48/80-induced degranulation, it is hypothesized that the responses in sinusoids and hepatic venules are mediated by serotonin and other mast cell constituents (e.g., prostaglandins, PAF, heparin), but at sites other than histamine and serotonin-specific receptors. These results support the involvement of mast cell constituents in hepatic microvascular regulatory mechanisms.

Hepatic microvascular regulatory mechanisms. II. Cholinergic mechanisms.

Several cholinergic agonists and their antagonists were administered topically at various concentrations (10(-10) to 10(-4) gm per ml) to the livers of anesthetized Sprague-Dawley rats. Changes in the microvasculature were measured for a period of 15 min using in vivo microscopic methods. The influence of cholinergic agonists on hepatic mast cells was determined by histochemical methods. Cholinergic agonists elicited constriction of portal venules, sinusoids, and terminal hepatic (central) venules accompanied within 5 to 20 sec by the adherence of platelets and leukocytes to the endothelial lining of these vessels. These responses were not observed in hepatic arterioles or hepatic (sublobular) venules. However, the responses drastically altered cellular flow through all segments of the microvasculature. The drugs were effective in the following ranking: bethanechol greater than pilocarpine greater than carbachol. Vascular responses were not antagonized by atropine; phentolamine modified the constrictor response of portal venules but of no other microvascular segment. These results suggest that the responses were not mediated by cholinergic (muscarinic) receptors and that the constriction of portal venules was caused by the chemical activation of alpha-adrenergic receptors. Histochemical methods revealed that pilocarpine and bethanechol also elicited a significant decrease in the number of visible mast cells suggestive of degranulation and the release of serotonin, histamine, and other polyanions (e.g., heparin) from these cells. Release occurred within the same period of time as the microvascular alterations (15 min.). These data provide additional evidence that mast cell constituents and adrenergic mechanisms are involved in hepatic cholinergic mechanisms.

Compartmental analysis of circulation of erythrocytes through the rat spleen.

Control, polycythemic, and anemic spleens of Sprague-Dawley rats surgically removed under pentobarbital sodium anesthesia (0.06 mg/g ip) were perfused in vitro by methods similar to those reported by Song and Groom (Am. J. Physiol. 220: 779-784, 1971). Red Blood cell (RBC) washout curves obtained for each group could be described adequately by a series of three exponentials--fast, intermediate, and slow--representing functional compartments in the spleens. Best-fit estimates of the parameters were used to characterize the flow distributions, cellular capacities, and t1/2 for washout of these components. It is thought that the fast compartment represents flow through intrasplenic vessels (including shunts) that bypass the red pulp, the intermediate compartment represents RBCs that flow through less impeded passageways in the red pulp, and the slow compartment consists of more tortuous, convoluted pathways through the red pulp. During polycythemia and anemia, inflow to the intermediate component was decreased, which suggested decreased filtration. In the slow compartment, the shutdown during polycythemia suggested RBC storage whereas during anemia control levels of inflow but increased t1/2 for washout suggested sequestration or erythropoiesis.