Regulation of Plasma Ferritin by the Isolated Perfused Rat Liver

S ummary . The isolated perfused rat liver has been used to investigate the regulation of plasma ferritin in normal, iron-deficient and iron-overloaded states. Both ¹²⁵I-labelled and non-labelled rat liver ferritins were rapidly cleared from the perfusion circuit with a half-life of approximately 30 min. Perfusion of livers from normal, iron-loaded or iron-deficient rats with blood obtained from normal, iron-loaded or iron-deficient rats showed that the liver takes up plasma ferritin, and releases ferritin into the perfusate to achieve a perfusate ferritin level appropriate to the iron stores of the animal from which the liver was taken. It is concluded that the liver is capable of both uptake and release of ferritin and that within the liver there resides a mechanism which maintains circulating ferritin concentrations at a level appropriate to body iron stores.     

The Uptake of Fibrin and Fibrin-Degradation Products by the Isolated Perfused Rat Liver

The uptake of Se⁷⁵-labeled rat fibrin and a specific preparation of fibrindegradation products by an isolated, perfused liver was studied. The rapidphase T for fibrin in the perfusate was approximately 9 minutes. This iscomparable to that found for aggregated I¹³¹ B.S.A. and colloidal gold in thesame system. The loss of activity from the blood closely paralleled the increasedactivity recorded in the liver. No radioactivity was found in the bile samplesor the T.C.A.-soluble fraction of plasma. Microradioautographs of washedliver tissue following the labeled, fibrin perfusion suggest the presence of thelabel on or in the Kupffer cells.

Labeled rat fibrin preparations, digested by the action of "cold" rat plasminto a distinct endpoint, were found to contain 3 distinct fractions on paper electrophoresis. Addition of these fractions to the perfusate resulted in approximately 25 per cent clearance of radioactivity over a 3-hour period. Hence,clearance of this specific F.D.P. preparation occurs at a much slower ratethan that of fibrin. The T as found to be in excess of 12 hours. Thesefindings appear to confirm Lee’s observation that cells of the reticuloendothelialsystem are involved in the elimination of free circulating fibrin and its metabolic products.

Submitted on February 18, 1966 Accepted on October 8, 1966     

Brief Report: Bilirubin Formation in the Liver from Nonhemoglobin Sources: Experiments with Isolated, Perfused Rat Liver

Rapid formation of C¹⁴-bilirubin from glycine-2-C¹⁴ or -aminolevulinicacid-4-C¹⁴ has been demonstrated in isolated, perfused rat liver. The kineticsof labeled bilirubin production in vitro were similar to those of the "earlylabeling" pigment fraction observed in intact rats. The magnitude of bilepigment formation from C¹⁴-ALA by isolated liver was comparable to that inintact animals. Derivation of the C¹⁴-bilirubin from hepatic degradation oflabeled hemoglobin was excluded by perfusion of the liver with blood poor inreticulocytes, or with plasma. The findings identify the liver as an importantsource of the "early-labeling" pigment fraction.

Submitted on April 15, 1965 Accepted on June 18, 1965     

Kinetics of Insulin Release from the Perfused Rat Pancreas Caused by Glucose, Glucosamine, and Galactose

Under appropriate conditions, not only glucose but also glucosamine and galactose can serve as potent stimulants for insulin release from the isolated, perfused rat pancreas. Since galactose and, probably, glucosamine are not metabolized in the islets, and since these three compounds have in all likelihood common sites of action, it is postulated that a glucoreceptor of broad specificity is involved in the mechanism of insulin release, and that metabolism of glucose is not an essential part of the releasing action of this sugar.     

Effect of Carbachol on Vascular and Luminal Release of Immunoreactive Gastrin from Isolated Perfused Rat Duodenum

The release of intestinal gastrin was examinedusing the isolated vascularly and luminally perfused ratduodenum. Results show that duodenal gastrin wasreleased into both the intestinal lumen and circulation. The basal release of immunoreactive gastrin(IRG) into the lumen was 23.3 ± 1.8 pg/min andthat into the vasculature was 70.1 ± 8.2 pg/min.The administration of 1 M carbachol into thevascular perfusate resulted in a marked reduction of luminalrelease of IRG; however, vascular release of IRG was notaffected. The carbacholinduced inhibition of luminalrelease of IRG was blocked by atropine but not by hexamethonium. These data suggest that only theluminal release of IRG from the rat duodenum, but notvascular release of IRG, is under the inhibitory controlof the cholinergic muscarinic mechanism.     

Chronic Ethanol Administration Impairs Degradation of Formaldehyde-Treated Albumin by the Perfused Rat Liver

Nonparenchymal cells of the liver appear to be important in the pathogenesis of various liver diseases, including that caused by ethanol. It is known that chronic ethanol administration impairs the process of receptor-mediated endocytosis in hepatocytes. Liver endothelial cells are also actively endocytic cells, playing a prominent role in the clearance from the circulation of a variety of macro-molecules. In this study, we assessed the effect of ethanol administration on this "scavenger" function of liver endothelial cells by measuring the degradation of formaldehyde-treated albumin in isolated, perfused livers of ethanol-fed rats. Rats were pair-fed for 1 or 4 weeks with a liquid diet containing either ethanol as 36% of total calories or an isocaloric amount of carbohydrate. Chronic ethanol administration in this manner for 1 or 4 weeks significantly impaired the degradation of this endothelial cell iigand (by 60 ± 9% and 37 ± 9%, respectively). Liver perfusions were also performed on rats that had been administered ethanol acutely or in which ethanol was added to the perfusate. No acute effect of ethanol on the degradation of this Iigand was seen. These results demonstrate that chronic ethanol ingestion impairs receptor-mediated endocytosis of formaldehyde-treated albumin by liver endothelial cells, indicating that the adverse effects of ethanol on protein trafficking within the liver are not limited to the hepatocytes.     

Ethanol Alters the Metabolic Response of Isolated, Perfused Rat Liver to a Phagocytic Stimulus

Intercellular communication in the liver is a potentially important mechanism for the regulation of hepatic metabolism. Since alcohol (ethanol, ETOH) can interact with both parenchymal and nonparenchymal cells, the present study was performed to assess the possible effects of ETOH on the nonparenchymal cell-to-hepatocyte signal traffic by studying the glycogenolytic and glycolytic response of the perfused rat liver to colloidal carbon, a phagocytic stimulus for Kupffer and sinusoidal endothelial cells. Livers from fed rats were perfused with hemoglobin-free Krebs Ringer bicarbonate buffer containing ETOH (20 mm) or acetaldehyde (1 mm). Twenty minutes after initiating the infusion of ETOH or acetaldehyde, colloidal carbon was infused and the rate of carbon uptake, glucose, lactate and pyruvate output, and oxygen consumption were determined. In control livers, carbon stimulated the output of glucose (60%), lactate (25%), and pyruvate (53%), without affecting the lactate/pyruvate ratio. ETOH, but not acetaldehyde, enhanced the carbon effect on glucose output (38%), but suppressed the increased lactate and pyruvate output (48% and 91% respectively) resulting in a dramatic 10-fold increase in the lactate/pyruvate ratio. By using inhibitors of cyclooxygenase or alcohol dehydrogenase (indomethacin and 4-methylpyrazole, respectively) in the presence of carbon and/or ETOH, we determined that: (1) following carbon stimulation prostaglandins are the likely mediators secreted by nonparenchymal cells that increase carbohydrate output; and (2) the ETOH-induced enhancement of carbonstimulated glycogenolysis is also mediated by prostaglandins and is not dependent on the oxidative metabolism of ETOH.     

Effect of S -Adenosyl-L-Methionine and Dilinoleoylphosphatidylcholine on Liver Lipid Composition and Ethanol Hepatotoxicity in Isolated Perfused Rat Liver

We investigated whether S -adenosyl-L-methionine(SAMe), dilinoleoylphosphatidylcholine (DLPC), or SAMe+ DLPC influence liver lipid composition as well asacute ethanol hepatotoxicity in the isolated perfused rat liver (IPRL). SAMe (25 mg/kgintramuscularly three times a day) was administered forfive consecutive days, while DLPC was administeredintraperitoneally for five days. The liver was thenisolated, perfused with taurocholate to stabilize bilesecretion, and exposed to 0.5% ethanol for 70 min. SAMe,without changing total phospholipid (PL) content,induced an increase in the phosphatidylcholine/phosphatidylethanolamine (PC/PE) molar ratio in both liver homogenateand microsomes and a significant enrichment of 16:0-20:4and 18:0-20:4 PC molecular species. DLPC induced asignificant enrichment of PL in liver homogenate and microsomes due to a contemporary increasein PC and PE. The PC enrichment specifically involved16:0-20:4 and 18:0-20:4 PC molecular species besides theHPLC peak containing the administered 18:2- 18:2 PC species. DLPC + SAMe increased theconcentration of PC in liver homogenate and microsomesdue to a specific enrichment of 16:0-22:6, 16:0-20:4,and 18:0-20:4 PC molecular species, and the HPLC peakcontaining the administered 18:2-18:2 PC species. Ethanolacute exposure in the control IPRLs for 70 min induceda depletion of cholesterol in both liver homogenate andmicrosomes without significant changes in the composition of PL classes and PC molecularspecies. SAMe, DLPC, or SAMe + DLPC counteracted thecholesterol depletion induced by ethanol, indicatingthat phospholipid changes promoted by these treatments all induce a major resistance of livermembranes to the effect of ethanol. Ethanoladministration in control IPRLs induced a fivefoldincrease of AST and LDH release in the perfusate,depletion of glutathione in homogenates and mitochondria, decreasedoxygen liver consumption, and inhibition of bile flow.These effects of ethanol were significantly antagonizedby SAMe. In contrast, DLPC alone only minimally attenuated enzyme release in the perfusate andthe inhibitory effect of ethanol on bile flow, but itfailed to influence the depletion of total andmitochondrial glutathione or the depressed oxygenconsumption induced by ethanol. DLPC, administered togetherwith SAMe, added nothing to the protective effect ofSAMe against ethanol hepatotoxicity and cholestasis. Inconclusion, this study demonstrates that both SAMe and DLPC induced marked modifications inthe lipid composition of liver membranes with a similarenrichment of polyunsaturated PC molecular species. OnlySAMe, however, significantly protected against the hepatotoxic and cholestatic effect of acuteethanol administration, an effect associated withmaintained normal glutathione mitochondrial levels andoxygen liver consumption. This indicates that the protective effect of SAMe against ethanoltoxicity is linked to multiple mechanisms, themaintenance of glutathione levels probably being one ofthe most important.     

Effects of Glucose on Glycogen Synthetase, Phosphorylase, and Glycogen Deposition in the Perfused Rat Liver

An increase in the perfusate glucose concentration from near zero to about 11 mM increased glycogen synthesis in the perfused, isolated rat liver from zero to a value about half the maximum seen in the intact animal. Increased synthesis appeard to due not only to provision of substrate but also to conversion of glycogen synthetase to the active form and of glycogen phosphorylase to the inactive form. These glucose effects, which are apparently independent of changes in levels of hormones or adenosine 3':5'-cyclic phosphate, may be physiologically significant for control of the blood glucose level.     

Iron Release from Isolated Hepatocytes

S ummary . The isolated hepatocyte suspension was evaluated as an experimental procedure for investigating liver iron metabolism. Following prelabelling in vivo with transferrin-⁵⁹Fe, isolated hepatocytes released radioactive iron in vitro by a temperature dependent process, without change in cell viability. Iron mobilization was increased by serum, apotransferrin and a range of iron chelators, of which the most effective were citrate, desferrioxamine and the ionophore A 23187. The rate of iron release was inversely related to oxygen levels, indicating that a ferric—ferrous reduction was involved in iron mobilization. The uncoupler TTFB, DTPA, and hypercapnia caused a reduction in iron release, but the metabolites cysteine, NADH and ascorbic acid had no effect. It was concluded that isolated hepatocytes are a useful experimental model for studying iron metabolism and for further evaluation of iron chelators.     

Effect of Pharmacological Modulation of Liver P-Glycoproteins on Cyclosporin A Biliary Excretion and Cholestasis A Study in Isolated Perfused Rat Liver

In different cell types P-glycoproteins (P-gp)are involved in the transport of cyclosporin A (CyA).The aim of this study was to evaluate the effect of thepharmacological modulation of the hepatic P-gp on biliary secretion of CyA and on cholestasisinduced by acute administration of CyA in the isolatedperfused rat liver (IPRL). Verapamil was used as a P-gpspecific inhibitor and acetylaminofluorene (AAF) as a P-gp inducer. CyA biliary excretion wasdetermined by administering in the IPRL a tracer dose of[³H]CyA with or without verapamil or AAF. Theeffect on bile flow was evaluated by administering increasing doses of CyA (2.8, 8, and 20 mg/kgbody wt) in the IPRL. Morphological evidence of damagewas evaluated by optical and electron microscopy in theliver as well as in primary culture of rat hepatocytes exposed to CyA ± verapamil. Verapamilsignificantly inhibited the biliary excretion of atracer dose of [³H]CyA (0.15 ± 0.04 vs0.33 ± 0.07%; P < 0.05). In contrast,pretreatment with AAF significantly increased the biliaryexcretion of [³H]CyA, (0.61 ± 0.10 vs0.33 ± 0.07%; P < 0.05). CyA induced adose-dependent inhibition of bile flow with a maximaleffect at 20 mg/kg CyA (–49.3 ± 4.5% decreaseof basal bile flow). CyA cholestasis was significantlyworsened by the P-gp inhibitor, verapamil (–75.5± 7.5%; P < 0.05), but it was unaffected byinduction of P-gp via AAF pretreatment (–44.9 ±1.7%). During CyA cholestasis, the cumulative biliaryexcretion of [³H]CyA was lower than in theabsence of cholestasis (0.22 ± 0.05 vs 0.33± 0.07%; P < 0.05), was inhibited by verapamil (0.08± 0.01%; P < 0.05), but was unaffected by AAF(0.23 ± 0.05%). No morphological evidence ofdamage was observed in the liver, and no evidence ofcytoskeleton derangement was seen in primary cultures of rathepatocytes exposed to CyA ± verapamil. Wedemonstrated that pharmacological modulation of P-gp mayinfluence the biliary excretion of CyA. The acutecholestatic effect of CyA is worsened by P-gp inhibitors,while it is unaffected by P-gp inducers. This indicatesCyA should not be given with other P-gp substrates orinhibitors.     

Hepatic Saturation Mechanism of Ethanol: Application of Mathematical Models to Ethanol Outflow Profiles in the Perfused Rat Liver

The saturation mechanism of hepatic ethanol (EtOH) elimination was studied in the perfused rat liver. EtOH outflow profiles after the instantaneous administration of 3 (mg/ml) × 0.4 (ml), 12 × 0.1, 24 × 0.1, and 3 × 0.1 mg (as a dose concentration × a volume) through the portal vein were analyzed by the statistical moment analysis and mathematical models (i.e., dispersion models). Results for 3 × 0.1 and 12 × 0.1 mg doses by moment analysis were similar. This demonstrated that the elimination exhibits linear kinetics. Recovery ratio and hepatic volume of distribution for 3 × 0.4 and 24 × 0.1 mg were larger than those for 3 × 0.1 and 12 × 0.1 mg doses and were similar. Kinetics after administration of 3 × 0.4 and 24 × 0.1 mg may be nonlinear. A difference in the relative dispersion ( CV ²) obtained by moment analysis between 3 × 0.4 and 24 × 0.1 mg doses indicated different properties of the nonlinear elimination kinetics. There were no differences in all the parameters in the one-compartment dispersion model between 3 × 0.4 and 24 × 0.1 mg doses. In the two-compartment dispersion model, there were differences in the blood volume ( V _B ) and the forward partition rate constant ( K ₁₂) between 3 × 0.4 and 24 × 0.1 mg ( p < 0.05), whereas the elimination rate constant ( k _e) and the dispersion number values for these doses were similar. These findings demonstrated that there is difference in the no-equilibrium process between 3 × 0.4 and 24 × 0.1 mg doses. Therefore, we suggest that the continuous EtOH input into the liver causes the saturation of enzyme pathways and the change of the nonequilibrium process.     

The Uptake of Ferric Iron by Rat Liver Ferritin in Vivo and in Vitro

S ummary . The incorporation of ferric iron labelled with ⁵⁹Fe into rat liver ferritin has been studied in whole animals, into liver homogenate and into purified protein. Uptake of Fe³⁺ into purified rat liver ferritin followed a pattern similar to that with horse spleen ferritin given Fe²⁺ and an oxidant. The distributions of iron incorporated as a function of molecular iron content obtained in vivo resembled the in vitro patterns for iron contents above 500 Fe atoms/molecule and times after injection of up to 12 h. At larger intervals a maximum label moved to molecules of highest iron content as the molecules accumulated more iron. The apparently reduced uptake of injected ⁵⁹Fe into molecules of low iron content might be due either to the chase of cold iron through an existing iron pool or the presence of functionally different ferritins at more than one anatomic site within the cell.     

Iron Deficiency: Impaired Liver Growth and DNA Synthesis in the Rat

S ummary . This study deals with liver growth and DNA synthesis in young, rapidly growing rats weaned to an iron deficient diet. Shortly after the development of anaemia there was a depression in the rate of liver growth that was more profound than that of body growth. The rate of increase in total liver DNA was similarly depressed, in comparison with iron-supplemented animals. Paradoxically, in the deficient rats, two normally sensitive indices of DNA synthesis, incorporation of [³H]thymidine into DNA and activity of thymidine kinase, were unaffected. This was also the case in animals acclimated to a 4 hr/day restricted feeding regimen in order to control for possible variations in feeding patterns. Although [³H]thymidine incorporation was normal throughout a 24 hr period, incorporation of ³²P into liver DNA was depressed in the iron deficient rats. The results indicate changes in liver DNA synthesis due to iron deficiency that are distinct from those seen in starvation and protein-calorie malnutrition. Probably DNA synthesis in iron deficiency is modified through a depression of de novo synthesis, in the presence of normal salvage pathway activity.     

The Deuterium Isotope Effect on Ethanol Metabolism in Perfused Rat Liver: Effect of Reversed Perfusion on Ethanol and Oxygen Uptake

Livers from rats fasted for 24 hr were subjected to nonrecirculating perfusion with Krebs-Ringer bicarbonate solution containing 10 mm ethanol. The deuterium isotope effect was measured using (1- R )-[1-¹⁴C, 1-²H]ethanol. A value of 2.57 ± 0.09 (SO) was obtained independent of the direction of perfusion. Oxygen uptake and ethanol metab-olism in contrast were significantly increased when reverse perfusion (i.e., from vena cava to vena portae) was used. The magnitude of the isotope effect indicates that contribution from microsomal ethanol-oxidizing system if this is the only supplementary system is 9.8% under the experimental conditions. At high ethanol concentrations, the contribution would approach 18%. Equal activities of microsomal ethanol-oxidizing system and catalase under the experimental conditions would mean that both contribute 7.3% of the total ethanol metabolism. At high ethanol concentrations (80 mm), how-ever, catalase will be 6.8% and microsomal ethanol-oxidizing system is calculated to 13.3%. Preliminary experiments with rats pretreated with phenobarbital showed no change in the isotope effect or in the rate of ethanol metabolism, but a 40–50% increase in oxygen consumption. The acetaldehyde concentration in the effluent medium was below 1 μM.     

Gastrin Produces an Immediate and Dose-Dependent Histamine Release Preceding Acid Secretion in the Totally Isolated,Vascularly Perfused Rat Stomach

Increasing doses of gastrin 1–17 (Gl-17) were administered to totally isolated, vascularly perfused rat stomachs prestimulated with the phosphodiesterase inhibitor isobutyl methylxanthine (IMX). Vascular and luminal histamine outputs and luminal acid output were monitored at short intervals. (31–17 induced an immediate histamine release to the vascular perfusate, preceding the increase in acid secretion by approximately 10 min. Vascular histamine output increased from a base line (IMX only) of 4.0 ± 0.4 to a maximum of 34.5 ± 7.3 nmol/60 min (mean ± SEM) after 1040 pM G1–17, and acid output from 8.0 ± 2.8 to 61.5 ± 7.0 μmo1/60 min after 520 pM GI-17. Acid output was correlated to vascular histamine release (r = 0.64. p < 0,001). Gastrin produced a histamine release giving gastric venous concentrations of the same magnitude as the concentration of histamine necessary to induce a comparable acid response. Histamine release to the lumen, on the other hand, paralleled the acid secretion in time, suggesting it to be a passive phenomenon secondary to acid secretion. Thus, the present study for the first time shows that gastrin induces vascular histamine release of such a magnitude that this substance could be the mediator of the gastrin effect on acid secretion.     

Synthesis of Procoagulant Factor VIII, Factor VIII Related Antigen and other Coagulation Factors by the Isolated Perfused Rat Liver

The synthesis of factor VIII and other coagulation factors has been studied using an isolated, perfused rat liver. Synthetic function of the liver was validated by adding [35S]L-methionine to the perfusion medium and performing two-dimensional crossed immunoelectrophoresis and autoradiography on samples obtained during perfusion. Progressive incorporation of radioactivity into plasma proteins was demonstrated. This was inhibited by cycloheximide. Coagulation factor assays demonstrated synthesis of factors II, IX and X and of factor V and procoagulant factor VIII (VIIIC). Synthesis of factor VIII related antigen (VIIIRAg), measured in an immunoradiometric assay, was not significantly demonstrated. Addition of warfarin to the perfusion medium inhibited the synthesis of factors II, IX and X but not of factors V and VIII. Cycloheximide completely inhibited synthesis of all coagulation factors but actinomycin acted only after a latent period. Reticuloendothelial cell blockade was attempted by adding ethionine to the perfusion medium or by administration of Indian ink to the donor animals prior to removal of the livers. In these instances synthesis of factor V and factor VIIIC was inhibited but not that of factors II, IX and X. The results confirmed the functional capacity of the isolated liver for synthesizing proteins and the vitamin K dependent coagulation factors, and suggested similar kinetic features for the synthesis of factors V and VIIIC. Failure to detect significant VIIIRAg synthesis in these experiments is consistent with the hypothesis that this protein is released by vascular endothelial cells throughout the body and is activated or joined to VIIIC or stimulates its production in the liver.     

Evaluation of Regional Liver Damage by Magnetic Resonance Imaging with Superparamagnetic Iron Oxide in Rat Liver

The purpose of this study was to investigate whether regional liver damage could be detected by means of enhanced MR imaging with a superparamagnetic iron oxide (SH U 555A) in an ischemia–reperfusion model of rat liver. Ischemic liver damage was induced in the right lobe by vascular clamping for 0 (sham), 30 (I-30), 60 (I-60), and 90 minutes (I-90). There was no significant difference in relative enhancement (RE) between the ischemic and nonischemic lobes in the sham, I-30 and I-60 groups, while RE of the ischemic lobe was significantly lower than that of its nonischemic counterpart in the I-90 group as seen on SH U 555A enhanced proton density spin echo images (P < 0.05). Histological examination revealed that iron deposits were significantly smaller in the ischemic than the nonischemic lobe in the I-90 group (P < 0.05), although there was no significant difference in the number of Kupffer cells. Our results indicate that severe regional liver damage can be evaluated by MR imaging with SH U 555A.     

The Influence of Flow and Hematocrit on the Laser Doppler Flux Signal from the Surface of the Perfused Pig Liver

Objective: We tested the hypothesis that the measurement volume of the laser Doppler flowmeter (LDF) is too small to provide reliable quantitative estimates of total liver blood flow of large mammals, such as the pig. Methods: In a perfused pig liver, the influence of changing (i) hepatic arterial (HA) and portal venous flows individually (n = 9), (ii) HA flow at fixed portal venous flow (50%, 70%, and 100% expected total liver blood flow), and (iii) hematocrit (0–30%) at fixed total liver blood flow on LDF flux was tested (n = 8). Results: Linearity of LDF with hepatic arterial flow and portal venous flow was confirmed; however, the slope of the regression lines was higher for hepatic artery [1.92 ± 0.60 (SD)] than portal vein perfused livers (0.66 ± 0.34; P < 0.001). With portal venous flow at 50% and 70% total liver blood flow, changing hepatic arterial flow produced linear LDF versus flow responses, but at 100% total liver blood flow, linearity was achieved in only 6/9 livers. The coefficient of variation for the slopes of regression lines was always > 30%. At constant total liver blood flow (100 ml/min per 100 g), LDF response decreased linearly by a factor of about 2 on changing the hematocrit from 30% to 5% and markedly fell as the hematocrit was further decreased to zero. Conclusions: These results suggest that (i) the LDF flux signal from the liver surface provides a poor measure of hepatic microcirculatory blood flow during changes in total liver blood flow as the LDF responds with about three times greater sensitivity to changes in hepatic arterial than in portal venous flow, and (ii) when hematocrit is falling, LDF may underestimate hepatic perfusion to a significant extent. In addition, due to high measurement variability, the LDF flux signal cannot be quantified in absolute perfusion units.