Increase In Interstitial Adenosine Research Articles

Traditional Acupuncture Triggers a Local Increase in Adenosine in Human Subjects

Acupuncture is a form of Eastern medicine that has been practiced for centuries. Despite its long history and worldwide application, the biological mechanisms of acupuncture in relieving pain have been poorly defined. Recent studies in mice, however, demonstrate that acupuncture triggers increases in interstitial adenosine, which reduces the severity of chronic pain through adenosine A1 receptors, suggesting that adenosine-mediated antinociception contributes to the clinical benefits of acupuncture. We asked here whether acupuncture in human subjects is also linked to a local increase in interstitial adenosine concentration. We collected microdialysis samples of interstitial fluid before, during, and after delivering 30 minutes of conventional acupuncture in the Zusanli point in human subjects. The interstitial adenosine concentration increased significantly during acupuncture and remained elevated for 30 minutes after the acupuncture. Acupuncture-mediated adenosine release was not observed if acupuncture was not delivered in the Zusanli point or if the acupuncture needle was inserted, but not rotated. This study strengthens the role of adenosine in acupuncture-mediated antinociception by directly providing such evidence in humans. PerspectiveThis article presents further evidence of the role of adenosine in acupuncture-mediated antinociception by demonstrating that local adenosine concentrations increase in the acupoint in human subjects receiving traditional acupuncture.

Cardioprotection following adenosine kinase inhibition in rat hearts

Adenosine kinase phosphorylates adenosine to AMP, the primary pathway for adenosine metabolism under basal conditions. Inhibition of adenosine kinase results in a site-specific increase in interstitial adenosine. Using a rat model of myocardial infarction, we examined the protective effects of adenosine kinase inhibition. Male Sprague-Dawley rats underwent 30 min regional occlusion followed by 90 min reperfusion. Infarct size, expressed as a percent of the area-at-risk, IS/AAR(%), was 58.0 +/- 2.1 % in untreated rats. Pretreatment with the adenosine kinase inhibitor, 5-iodotubercidin (1 mg/kg), limited infarct development to 37.5+/-3.7% (P < 0.001). The A(1) adenosine receptor (A(1)AR) antagonist, DPCPX (100 microg/kg), abolished the infarct-sparing effect of 5-iodotubercidin (IS, 62.8 +/- 1.3%). Similarly, the A(3) adenosine receptor (A(3)AR) antagonist, MRS-1523 (2 mg/kg), and the delta-opioid receptor (DOR) antagonist, BNTX, (1 mg/kg) abolished the reduction of IS produced by iodotubercidin. Pretreatment with the ROS scavenger, 2-MPG (20 mg/kg), or the PKC-delta antagonist, rottlerin (0.3 mg/kg) also abolished iodotubercidin-mediated cardioprotection. Furthermore, pretreatment with 5-HD, a mitochondrial K(ATP) (mitoK(ATP)) channel inhibitor, but not the sarcolemmal K(ATP) channel blocker, HMR-1098, abrogated the beneficial effects of adenosine kinase inhibition (IS, 59.5 +/- 3.8%). These data suggest that inhibition of adenosine kinase is effective in reducing infarct development via A(1)AR, A(3)AR and DOR activation. Data also suggest that this protection is mediated via ROS, PKC-delta and mitoK(ATP) channels.

Regulation of flow and wall shear stress in arteriolar networks of the hamster cheek pouch.

Our purpose was to define arteriolar network hemodynamics during moderate increases in interstitial adenosine or nitric oxide in the hamster (n = 34, pentobarbital sodium 70 mg/kg) cheek pouch tissue. The network consists of a feed arteriole (approximately 12-microm diameter, approximately 800-microm length) with three to six branches. Observations of diameter, red blood cell flux, and velocity were obtained at the feed before the branch and within the branch. A comparison of baseline with suffused adenosine or sodium nitroprusside (SNP) 10(-9) to 10(-5) M showed the following. First, diameter change was heterogeneous by agonist, did not reflect the expected dilatory response, and was related to location within the network. With adenosine, upstream branch points constricted and those downstream dilated, even at 10(-5) M. With SNP, upstream branch points dilated, whereas those downstream constricted. Second, with adenosine, changes in diameter, flux, and velocity together resulted in no change in wall shear stress until 10(-5) M. Wall shear stress was not maintained at a constant level with Nomega-nitro-L-arginine (10(-5) M), suggesting a role for flow-dependent diameter changes with adenosine. With SNP, diameter change correlated with the baseline (before SNP) shear stress conditions.

Effects of dynamic exercise on mean blood velocity and muscle interstitial metabolite responses in humans.

We examined the effects of dynamic one-legged knee extension exercise on mean blood velocity (MBV) and muscle interstitial metabolite concentrations in healthy young subjects (n = 7). Femoral MBV (Doppler), mean arterial pressure (MAP) and muscle interstitial metabolite (adenosine, lactate, phosphate, K(+), pH, and H(+); by microdialysis) concentrations were measured during 5 min of exercise at 30 and 60% of maximal work capacity (W(max)). MAP increased (P < 0.05) to a similar extent during the two exercise bouts, whereas the increase in MBV was greater (P < 0.05) during exercise at 60% (77.00 +/- 6.77 cm/s) compared with 30% W(max) (43.71 +/- 3.71 cm/s). The increase in interstitial adenosine from rest to exercise was greater (P < 0.05) during the 60% (0.80 +/- 0.10 microM) compared with the 30% W(max) bout (0.57 +/- 0.10 microM). During exercise at 60% W(max), interstitial K(+) rose at a greater rate than during exercise at 30% W(max) (P < 0.05). However, pH increased (H(+) decreased) at similar rates for the two exercise intensities. During exercise, interstitial lactate and phosphate increased (P < 0.05) with no difference observed between the two intensities. After 5 min of recovery, MBV decreased to baseline levels after exercise at 30% W(max) (4.12 +/- 1.10 cm/s), whereas MBV remained above baseline levels after exercise at 60% W(max) (Delta19.46 +/- 2.61 cm/s; P < 0.05). MAP and interstitial adenosine, K(+), pH, and H(+) returned toward baseline levels. However, interstitial lactate and phosphate continued to increase during the recovery period. Thus an increase in exercise intensity resulted in concomitant changes in MBV and muscle interstitial adenosine and K(+), whereas similar changes were not observed for MAP or muscle interstitial pH, lactate, or phosphate. These data suggest that K(+) and/or adenosine may play an active role in the regulation of skeletal muscle blood flow during exercise.

INHIBITION OF CYCLOOXYGENASE INCREASES INTERSTITIAL ADENOSINE IN SKELETAL MUSCLE OF MAN

Locally formed vasoactive compounds are likely to interact to produce the fine-tuned regulation of skeletal muscle blood flow. In the present study we examined whether inhibition of cyclooxygenase (COX), inhibition of nitric oxide synthase (NOS) and a combined inhibition of NO synthase and cyclooxygenase (NOS/COX) would be compensated for by an increase in interstitial adenosine in human skeletal muscle. Microdialysis probes were placed in the vastus lateralis muscle of 8 healthy male subjects. After an 80 min equilibration period dialysate was collected regularly in 20 min periods over the next 3 h. After 40 min of base-line collection the NOS inhibitor NG nitro L-arginine (L-NA) was administered through 11 of the probes. No administration took place through the remaining 11 probes (COX). At 100 min acetylsalicylic acid (500 mg) was orally administered to all of the experimental subjects. As control, in three of the subjects (7 probes) microdialysate was regularly collected for the three hour period without administration of inhibitors. All dialysate samples were analysed for adenosine. The baseline concentration of interstitial adenosine was 0.14 ± 0.03 (± SEM) μmol/l and the concentration remained unaltered (P > 0.05) with NOS inhibition (0.14 ± 0.04 μmol/l). After administration of acetylsalicylic acid the interstitial adenosine concentration rose (P < 0.05) similarly in NOS/COX and COX and reached a level of 0.30 ± 0.04 and 0.37 ± 0.05 μmol/l, respectively, at 90 min. Control experiments showed that the increase in adenosine was not an effect of time. The results demonstrate that adenosine formation in human skeletal muscle in vivo is enhanced by inhibition of cyclooxygenase and that additional inhibition of NOS has no further effect. Supported by Danish National Research Foundation (504–14).

Adenosine and preconditioning revisited.

1. Myocardial tolerance against infarction is substantially increased by exposing myocytes to 3-10 min transient ischaemia. In this phenomenon, termed 'preconditioning', the adenosine receptor is one of the redundant triggers and the best characterized factor in the cardioprotective mechanism. 2. An increase in interstitial adenosine during preconditioning is thought to be derived primarily from hydrolysis of 5'-AMP in the myocyte by cytosolic 5'-nucleotidase, although a contribution of ectosolic 5'-nucleotidase remains controversial. Adenosine production during ischaemia is substantially suppressed in the preconditioned myocardium, probably due to a decrease in ATP utilization. 3. The adenosine receptor needs to be activated not only at the time of preconditioning ischemia, but also during ischaemic insult for the preconditioning to be cardioprotective. However, the extent of cardioprotection afforded by preconditioning is primarily determined by the interstitial adenosine level achieved during preconditioning ischaemia, not by the level during sustained ischaemia. These data suggest that a post-receptor mechanism downstream of the adenosine receptor may be up-regulated after preconditioning. 4. Studies in vitro suggest that the subtypes of adenosine receptor relevant to preconditioning against infarction are A1 and A3, the activation of which appears to provide additive protection. The functional interrelationship between these subtypes in vivo remains unknown. 5. An important step downstream of adenosine receptor activation is protein kinase C (PKC), which facilitates opening of ATP-sensitive potassium (KATP) channels, probably leading to enhancement of myocardial tolerance. However, activation of other protein kinases, such as tyrosine kinase, may also be important in preconditioning, depending on the animal species and preconditioning protocols. The PKC isoform and location of KATP channels (i.e. sarcolemmal vs mitochondrial KATP) that induce anti-infarct tolerance in myocytes remain to be identified.

Myocardial Preconditioning Produced by Ischemia, Hypoxia, and a KATPChannel Opener: Effects on Interstitial Adenosine in Dogs

Previous research has demonstrated that a transient increase in interstitial adenosine and subsequent activation of ATP-sensitive K+(KATP) channels are involved in triggering ischemic preconditioning (PC), however, the role of adenosine in mediating the cardioprotection of hypoxic PC and that produced by KATPchannel openers is less clear. Thus, the aim of the present study was to determine the role of adenosine in mediating the cardioprotective effects of PC produced by 5 min of ischemia, hypoxia, or by a 5-min intracoronary (i.c.) infusion of the KATPchannel opener bimakalim (1μg/min). A single microdialysis probe was implanted into the midwall of the ischemic area for sampling of interstitial fluid adenosine and its breakdown products during the PC stimulus, prolonged occlusion (60 min) and during the first 30 min of the reperfusion (3 h) period. Ischemic, hypoxic and bimakalim pretreatment significantly reduced infarct size, 5.3±1.5; 8.9±2.5; 11.4±3.2, respectively, as compared to control: 27.3±6.5. Both ischemic and hypoxic PC produced similar and significant increases (0.56±0.13μmol/l to 1.12±0.12μmol/l and 1.32μmol, control, ischemic and hypoxic PC, respectively) in dialysate adenosine concentration which persisted during the brief 10-min reperfusion period following PC. However, i.c. bimakalim resulted in a significant decrease in dialysate adenosine (0.56±0.13μmol/l to 0.22±0.04μmol/l) which persisted during the 10-min drug-free period. All three PC protocols resulted in similar decreases in dialysate adenosine, inosine and uric acid concentrations throughout the prolonged ischemic period as compared to control animals. In conclusion (1); PC produced by ischemia or hypoxia results in an increase in interstitial adenosine prior to a prolonged occlusion period; (2) the KATPchannel agonist, bimakalim, significantly decreased interstitial adenosine prior to a prolonged occlusion period; (3) ischemic PC, hypoxic PC, and bimakalim pretreatment produced a similar reduction in interstitial adenosine and its breakdown products during the prolonged ischemic period. These results suggest that an increase in interstitial adenosine may be necessary for the initiation of the protective effect of ischemic and hypoxic PC but an increase in adenosine is not necessary for the cardioprotective effect of a direct opener of the KATPchannel.

Myocardial stunning: A therapeutic conundrum

Dobutamine and pyruvate are two inotropic agents with different mechanisms of action. Although both agents alter postischemic myocardial dysfunction, their potential metabolic effects in the setting of in vivo myocardial stunning have not been addressed. In this study, the effects of dobutamine and pyruvate on systolic wall thickening, myocardial phosphorylation potential index, interstitial fluid adenosine level, and myocardial oxygen consumption in in vivo stunned porcine myocardium were assessed. Stunning was induced with a 10-minute occlusion of the left anterior descending coronary artery. After 30 minutes of reperfusion, pigs were treated with either intravenous dobutamine (10 μg/kg per minute) or intracoronary pyruvate (1 ml/min, 150 mmol/L solution, pH 7.4). Infusion of both agents resulted in a marked improvement in regional systolic wall thickening. The dobutamine effect, however, produced a marked increase in myocardial oxygen consumption and was associated with an increase in interstitial adenosine caused by myocardial de-energization, because the myocardial phosphorylation potential index ratio decreased from 0.17 ± 0.02 to 0.09 ± 0.02 (p <0.05). In contrast, pyruvate enhanced myocardial energy status, because the myocardial phosphorylation potential index ratio increased from 0.20 ± 0.03 to 0.55 ± 0.08 (p <0.01). These experimental findings suggest that under certain circumstances the use of β-receptor agonists to treat myocardial stunning may be suboptimal, if not undesirable. Further investigation is warranted to determine the optimum therapy for the stunned heart. (J T HORAC C ARDIOVASC S URG 1995;110:1391-1401)

Effects of anesthesia and K+ ATP channel blockade on interstitial adenosine accumulation in ischemic rabbit myocardium.

Glibenclamide, a K+ ATP channel antagonist, blocks the anti-infarct effect of ischemic preconditioning in rabbits, but only when the latter are anesthetized with ketamine-xylazine. Furthermore, the protection triggered by pinacidil, a K+ ATP channel opener, can be aborted by treatment with the adenosine antagonist 8-(P-sulfophenyl)theophylline. This study tests whether either the anesthetic regimen or glibenclamide affects infarct size by modulating interstitial adenosine levels. Interstitial adenosine and total purine concentrations were assessed in open-chest rabbits by the microdialysis technique. Dialysis fibers were inserted into myocardium served by a coronary artery branch surrounded by a snare. All animals sustained a 30-min coronary occlusion and then 120-min reperfusion. Rabbits were anesthetized with either sodium pentobarbital or a ketamine-xylazine mixture. Half of the latter animals also received glibenclamide. The control levels of adenosine in the dialysate were comparable in the three groups, as were those of total purines, and the infusion of glibenclamide caused no change. Ischemia led to 10- to 20-fold increases in interstitial adenosine and 10- to 40-fold rises in total purine concentrations. These increases were equivalent in all groups. Further-more, infarct size as a percentage of the myocardium at risk was also comparable in the three groups. Neither the anesthetic agent nor glibenclamide appears to modulate interstitial adenosine release from ischemic tissue.

Effects of xanthine amine congener on hypoxic coronary resistance and venous and epicardial adenosine concentrations

The aim was to define the contributions of interstitial and vascular adenosine in regulating coronary vascular resistance during hypoxia. To help in the assessment of adenosine in the vasodilator response, a potent adenosine receptor antagonist, xanthine amine congener (XAC), was used to block adenosine receptors. Seven isolated guinea pig hearts were perfused at constant flow with Krebs buffer. Coronary vascular resistance was determined during normoxia (95% O2) and mild hypoxia (60% O2) in the absence or presence of 200 or 400 nM XAC. Interstitial fluid was sampled by the epicardial disc technique and the interstitial concentration of XAC (ISF[XAC]) was determined directly by a radioreceptor assay or as tritiated XAC. Venous and epicardial concentrations of adenosine were determined by high performance liquid chromatography. In six additional experiments, the vasodilator effect of 1 microM intracoronary adenosine was measured in the absence or presence of 100 or 200 nM XAC. Mild hypoxia decreased coronary resistance by 37 (SEM 4)% in the absence of XAC and 26(5)% or 17(4)% in the presence of 200 or 400 nM XAC, respectively. ISF[XAC] rapidly equilibrated with [XAC] in the arterial perfusate or venous effluent. XAC 400 nM markedly increased (p < 0.05) the hypoxic levels of venous and epicardial fluid adenosine from 49(19) and 251(42) nM to 75(11) and 495(48) nM, respectively. XAC 100-200 nM almost completely prevented the vasodilatation induced by 1 microM intracoronary adenosine. Adenosine mediates at least 54% of hypoxic vasodilatation. XAC rapidly equilibrates within the myocardial interstitial space and, as a result of blocking adenosine receptors, increases interstitial and venous adenosine concentrations. Increases in interstitial adenosine may partially overcome the adenosine receptor blockade by XAC, thereby reducing the effectiveness of XAC in attenuating the hypoxic vasodilatation. XAC attenuates intracoronary adenosine induced vasodilatation (mediated by endothelial adenosine receptors) much more effectively than it attenuates hypoxic vasodilatation, underscoring the minimal role played by the endothelial receptors in hypoxic vasodilatation.

Increases in interstitial adenosine and cerebral blood flow with inhibition of adenosine kinase and adenosine deaminase.

The purpose of this study was to determine the changes in interstitial fluid (ISF) adenosine and cerebral blood flow (CBF) during inhibition of adenosine kinase or adenosine deaminase. Brain microdialysis was used to (a) measure CBF (H2 clearance), (b) sample cerebral ISF, and (c) deliver drugs locally to the brain. Microdialysis probes were implanted bilaterally in the caudate nucleus of halothane-anesthetized rats (n = 11). One probe was perfused with artificial cerebrospinal fluid (CSF) containing iodotubercidin (IODO), an adenosine kinase inhibitor, while the other probe was perfused with erythro-2-(2-hydroxy-3-nonyl)adenine (EHNA), an adenosine deaminase inhibitor. Both probes were subsequently perfused with EHNA+IODO. Finally, 8-(p-sulfophenyl)theophylline (SPT), an adenosine receptor antagonist, was added to EHNA + IODO in one probe, while the other probe continued to receive only EHNA + IODO. CBF and dialysate adenosine levels increased with either EHNA or IODO; however, the increases were greater with IODO. EHNA + IODO further increased CBF and dialysate adenosine. The hyperemia observed with EHNA + IODO was abolished by adenosine receptor blockade. These data suggest that basal adenosine levels are influenced to a greater extent by adenosine kinase than by adenosine deaminase. In addition, the increased CBF observed with inhibition of adenosine metabolism and the attenuation of this vasodilatory response with adenosine receptor blockade support a role for adenosine in CBF regulation.

Hypoxia enhances isoproterenol-induced increase in heart interstitial adenosine, depressing beta-adrenergic contractile responses.

Endogenous interstitial adenosine may protect the hypoxic heart by attenuating beta-adrenergic-induced contractile and metabolic responses, thereby reducing energy utilization. Constant-flow perfused rat hearts were used to study: 1) the effect of hypoxia on isoproterenol (ISO)-induced increase in interstitial adenosine, as estimated with epicardial surface transudates, and 2) the role of endogenous adenosine in hypoxic depression of ISO-induced contractile responses. ISO (1 nM for 10 minutes) in the normoxic heart increased transudate adenosine 114% from a pre-ISO normoxic value of 343 pmol/ml. ISO administered to the hypoxic heart increased transudate adenosine 357% from a pre-ISO hypoxic value of 797 pmol/ml. The absolute magnitude of the ISO-induced increase in transudate adenosine was 625% greater during hypoxia than during normoxia. This was associated with a reduction in the ISO-induced contractile response during hypoxia. In other experiments, with normoxia ISO (10 nM for 10 seconds) increased developed left ventricular pressure by 140 mm Hg, and the maximum rates of left ventricular pressure development and relaxation by 5,860 and 2,771 mm Hg/sec, respectively, above control values of 90 mm Hg, 2,250 mm Hg/sec, and 1,875 mm Hg/sec. Hypoxia reduced the three ISO-induced contractile responses by 50%, 56%, and 36%. However, 1,3-dipropyl-8-cyclopentylxanthine (5 x 10(-7) M), an adenosine A1-receptor antagonist, added to the hypoxic hearts resulted in only a 31%, 39%, and 9% reduction in the ISO-induced responses in developed left ventricular pressure and the maximum rates of left ventricular pressure development and relaxation, respectively.(ABSTRACT TRUNCATED AT 250 WORDS)

Sciatic nerve stimulation does not increase endogenous adenosine production in sensory-motor cortex.

Adenosine participates in the coupling of cerebral blood flow to oxygen consumption in the brain during such stimuli as hypoxia, ischemia, and seizures. It has been suggested that it also participates in the regulation of cerebral blood flow during somatosensory stimulation, a condition during which cerebral blood flow and oxygen consumption appear to be uncoupled. Interstitial adenosine was estimated by the microdialysis technique and cerebral blood flow was measured by hydrogen clearance in the hindlimb sensory-motor cortex during sciatic nerve stimulation. Cerebral blood flow increased from 102 to 188 ml min-1 100 g-1 (p less than 0.001) in the cortex contralateral to the stimulated leg without an associated increase in interstitial adenosine (baseline 0.624 microM, stimulation 0.583 microM). Infusion of the adenosine antagonist 8-sulfophenyltheophylline failed to block an increase in cerebral blood flow during central sciatic nerve stimulation, but decreased basal cerebral blood flow (69 ml min-1 100 g-1). These results suggest that adenosine does not mediate changes in cerebral blood flow during somatosensory stimulation, but may participate in the regulation of cerebral blood flow in the basal state.

Myocardial adenosine formation during hypoxia: Effects of ecto-5′-nucleotidase inhibition

Release of adenosine and AMP into epicardial fluid and coronary venous effluent of isovolumic guinea-pig hearts was examined during normoxic (95% O 2) and hypoxic (30% O 2) perfusion with and without the ecto-5′-nucleotidase inhibitor α,β-methylene adenosine diphosphate (AOPCP) ∗ ∗ Abbreviations: AOPCP is α,β-methylene-adenosine diphosphate. . Normoxic epicardial and venous adenosine levels were 221 ± 27 and 67 ± 11 n m, respectively, in untreated hearts. During 15 min of hypoxia, epicardial and venous adenosine levels increased in a phasic manner, reaching maximal values of 498 ± 32 and 441 ± 43 n m, respectively, during the initial 5 min of hypoxia. Epicardial and venous adenosine levels then declined slightly during the subsequent 10 min to 332 ± 33 and 224 ± 34 n m, respectively. Infusion of 50 μ m AOPCP significantly reduced venous adenosine levels during normoxia (< 50% of control), but was without effect on normoxic epicardial adenosine. Epicardial and venous adenosine levels increased during hypoxia with AOPCP but the increases were lower than those for untreated hypoxic hearts. Epicardial and venous adenosine levels recovered to baseline levels following 30 min of reoxygenation in both groups. Epicardial and venous AMP levels were elevated by AOPCP treatment during normoxia and hypoxia. Coronary vascular resistance decreased during hypoxia but the decline in resistance was less in AOPCP treated hearts. It is concluded that whereas basal interstitial adenosine levels appear to be independent of ecto-5′-nucleotidase activity, the hypoxic increase in interstitial adenosine is partially derived from an AOPCP sensitive ecto-5′-nucleotidase. Venous adenosine appears to be significantly dependent on ecto-5′-nucleotidase activity during normoxia and hypoxia. The data are also consistent with a role for endogenous adenosine in regulating coronary vascular resistance during hypoxia.

Adenosine as a mediator of postcontraction hyperemia in dog gracilis muscle.

The role of adenosine in postcontraction hyperemia (PCH) following sustained, maximal isometric contractions was studied in free-flowing dog gracilis muscles. The hemodynamic responses to contraction were examined in the presence and absence of dipyridamole (an adenosine transport inhibitor), erythro-9-(2-hydroxy-3-nonyl)adenine (EHNA, an adenosine deaminase inhibitor), or alpha, beta-methyleneadenosine-5'-diphosphate (AOPCP, an inhibitor of 5'-nucleotidase). Each muscle was stimulated to contract for 1, 3, 5, and 10 s during saline and drug infusions. For each contraction, the tension-time integral (TT), excess flow (EQ), and excess oxygen consumption (EVo2) were computed. Linear regression analyses were then performed on EQ vs. TT, EVo2 vs. TT, and EQ vs. EVo2. An alteration of the PCH response by the drug was determined as any significant change from the saline control in the slope of the linear regression of EQ vs. EVo2. Dipyridamole and EHNA caused increases of 73 and 48%, respectively, in the slope of EQ vs. EVo2, whereas AOPCP decreased the slope by 41%. The changes in the PCH produced by these drugs are consistent with the hypothesis that an increase in interstitial adenosine during muscular contraction contributes to PCH.