Physiological state and the sensitivity of liver mitochondrial outer membrane carnitine palmitoyltransferase to malonyl-CoA. Correlations with assay temperature, salt concentration and membrane lipid composition.

1. Liver mitochondrial outer membranes were pre-exposed to media of low (20 mM phosphate) or high salt concentration (20 mM phosphate + 0.3 M KCl) before assay of carnitine palmitoyltransferase (CPT) at 25 degrees C. 2. With membranes from fed rats, exposure to high salt decreased sensitivity of CPT to malonyl-CoA whereas high salt increased sensitivity of CPT to malonyl-CoA in membranes from 48 hr-fasted rats. These changes were paralleled by alterations in the KD for high affinity binding of [14C]malonyl-CoA to outer membranes. 3. Decreasing the CPT assay temperatures from 25 to 10 degrees C caused qualitatively similar changes to those seen on exposure to high salt. 4. The relative content of sphingomyelin was increased 2-fold and 4-fold in liver mitochondrial outer membranes from fasted and diabetic rats respectively. Fasting had no effect on the content of cholesterol whereas diabetes decreased this by a third.     

Intertissue differences in the hysteretic behaviour of carnitine palmitoyltransferase in the presence of malonyl-CoA.

In the presence of malonyl-CoA, the overt form of carnitine palmitoyltransferase (CPT1) in mitochondria from rat liver, kidney cortex, heart, skeletal muscle and brown adipose tissue shows non-linear time courses, suggesting hysteretic behaviour. The pattern of this hysteresis is similar in heart, skeletal muscle and brown adipose tissue, but the hysteretic behaviour of the enzyme in these three tissues differs markedly from that seen in liver and kidney.     

Binding of malonyl-CoA to isolated mitochondria. Evidence for high- and low-affinity sites in liver and heart and relationship to inhibition of carnitine palmitoyltransferase activity.

[14C]Malonyl-CoA bound to intact mitochondria isolated from rat liver and heart in a manner consistent with the presence of two independent classes of binding sites in each tissue. The binding characteristics for mitochondria obtained from fed male rats were: for heart, KD(1) = 11-18nM, KD(2) = 30 microM, N1 = 7pmol/mg of protein, N2 = approx. 660pmol/mg of protein; for liver, KD(1) = 0.1 microM, KD(2) = 5.6 microM, N1 = 11pmol/mg of protein, N2 = 165pmol/mg of protein. In the presence of 40 microM-palmitoyl-CoA the characteristics of binding at the high-affinity sites were changed, so that for heart KD(1) = 0.26 microM, with no change in N1 and for liver KD(1) = approx. 2 microM, with N1 increased to approx. 40pmol/mg of protein. Differences between the two tissues in tightness of malonyl-CoA binding at the high-affinity sites explains the considerably greater sensitivity of heart CPT1 (overt form of carnitine palmitoyltransferase) to inhibition by malonyl-CoA [Saggerson & Carpenter, (1981) FEBS Lett. 129, 229-232; McGarry, Mills, Long & Foster (1983) Biochem. J. 214, 21-28]. Starvation (24h) did not change the characteristics of [14C]malonyl-CoA binding to liver mitochondria and did not alter the I50 (concentration giving 50% inhibition) for displacement of [14C]malonyl-CoA by palmitoyl-CoA. Therefore the decreased sensitivity of liver CPT1 to inhibition by malonyl-CoA in starvation [Saggerson & Carpenter (1981) FEBS Lett. 129, 225-228; Bremer (1981) Biochim. Biophys. Acta 665, 628-631] is not explained by differences in malonyl-CoA binding. Percentage occupancy of the high-affinity sites in heart mitochondria by malonyl-CoA correlated closely with percentage inhibition of CPT1 measured under similar conditions. This finding supports the proposal that the high-affinity binding sites are the functional sites mediating inhibition of CPT1 by malonyl-CoA. Similar experiments with liver mitochondria also suggested that the occupancy of high-affinity sites by malonyl-CoA regulates CPT1 activity. 5,5'-Dithiobis-(2-nitrobenzoic acid), which decreased the sensitivity of heart or liver CPT1 to inhibition by malonyl-CoA [Saggerson & Carpenter (1982) FEBS Lett. 137, 124-128], also decreased [14C]malonyl-CoA binding to the high-affinity sites of heart mitochondria. N1 values for [14C]malonyl-CoA binding to high-affinity sites in liver mitochondria were determined in various physiological states which encompassed a 7-fold range of CPT1 maximal activity (fed, starved, pregnant, hypothyroid, foetal). The N1 value did not change in these states.(ABSTRACT TRUNCATED AT 400 WORDS)     

Carnitine palmitoyltransferase in liver and five extrahepatic tissues in the rat. Inhibition by DL-2-bromopalmitoyl-CoA and effect of hypothyroidism.

Mitochondria were isolated from rat adult liver, foetal liver, kidney cortex, heart, skeletal muscle and interscapular brown adipose tissue. DL-2-Bromopalmitoyl-CoA inhibited the overt form of carnitine palmitoyltransferase (CPT1) in heart, skeletal muscle and brown adipose tissue, with an IC50 value (concentration giving 50% inhibition) of 1.3-1.6 microM. By contrast, the IC50 value for inhibition of the kidney or adult liver enzyme was 0.08-0.1 microM. CPT1 in near-term foetal liver differed from that in adult liver in that the IC50 for inhibition by 2-bromopalmitoyl-CoA was 0.57 microM. It is suggested that there may be tissue-specific forms of the catalytic entity of CPT1 and that foetal liver may contain a mixture of adult liver- and muscle-type enzymes. In rats made hypothyroid by administration of propylthiouracil and an iodine-deficient diet, hepatic CPT1 activity was decreased by 83%. However, CPT1 activity in extrahepatic tissues showed no adaptive decrease in hypothyroidism.     

Re-evaluation of the interaction of malonyl-CoA with the rat liver mitochondrial carnitine palmitoyltransferase system by using purified outer membranes.

1. The interaction of malonyl-CoA with the outer carnitine palmitoyltransferase (CPT) system of rat liver mitochondria was re-evaluated by using preparations of highly purified outer membranes, in the light of observations that other subcellular structures that normally contaminate crude mitochondrial preparations also contain malonyl-CoA-sensitive CPT activity. 2. In outer-membrane preparations, which were purified about 200-fold with respect to the inner-membrane-matrix fraction, malonyl-CoA binding was largely accounted for by a single high-affinity component (KD = 0.03 microM), in contrast with the dual site (low- and high-affinity) previously found with intact mitochondria. 3. There was no evidence that the decreased sensitivity of CPT to malonyl-CoA inhibition observed in outer membranes obtained from 48 h-starved rats (compared with those from fed animals) was due to a decreased ratio of malonyl-CoA binding to CPT catalytic moieties. Thus CPT specific activity and maximal high-affinity [14C]malonyl-CoA binding (expressed per mg of protein) were increased 2.2- and 2.0-fold respectively in outer membranes from 48 h-starved rats. 4. Palmitoyl-CoA at a concentration that was saturating for CPT activity (5 microM) decreased the affinity of malonyl-CoA binding by an order of magnitude, but did not alter the maximal binding of [14C]malonyl-CoA. 5. Preincubation of membranes with either tetradecylglycidyl-CoA or 2-bromopalmitoyl-CoA plus carnitine resulted in marked (greater than 80%) inhibition of high-affinity binding, concurrently with greater than 95% inhibition of CPT activity. These treatments also unmasked an effect of subsequent treatment with palmitoyl-CoA to increase low-affinity [14C]malonyl-CoA binding. 6. These data are discussed in relation to the possible mechanism of interaction between the malonyl-CoA-binding site and the active site of the enzyme.     

Regulation of carnitine palmitoyltransferase (CPT) I during fasting in rainbow trout (Oncorhynchus mykiss) promotes increased mitochondrial fatty acid oxidation

Periods of fasting, in most animals, are fueled principally by fatty acids, and changes in the regulation of fatty acid oxidation must exist to meet this change in metabolic substrate use. We examined the regulation of carnitine palmitoyltransferase (CPT) I, to help explain changes in mitochondrial fatty acid oxidation with fasting. After fasting rainbow trout (Oncorhynchus mykiss) for 5 wk, the mitochondria were isolated from red muscle and liver to determine (1) mitochondrial fatty acid oxidation rate, (2) CPT I activity and the concentration of malonyl-CoA needed to inhibit this activity by 50% (IC(50)), (3) mitochondrial membrane fluidity, and (4) CPT I (all five known isoforms) and peroxisome proliferator-activated receptor (PPARα and PPARβ) mRNA levels. Fatty acid oxidation in isolated mitochondria increased during fasting by 2.5- and 1.75-fold in liver and red muscle, respectively. Fasting also decreased sensitivity of CPT I to malonyl-CoA (increased IC(50)), by two and eight times in red muscle and liver, respectively, suggesting it facilitates the rate of fatty acid oxidation. In the liver, there was also a significant increase CPT I activity per milligram mitochondrial protein and in whole-tissue PPARα and PPARβ mRNA levels. However, there were no changes in mitochondrial membrane fluidity in either tissue, indicating that the decrease in CPT I sensitivity to malonyl-CoA is not due to bulk fluidity changes in the membrane. However, there were significant differences in CPT I mRNA levels during fasting. Overall, these data indicate some important changes in the regulation of CPT I that promote the increased mitochondrial fatty acid oxidation that occurs during fasting in trout.     

Effects of chronic cold exposure on the activities of cytochrome c oxidase, glutathione peroxidase and glutathione reductase in rat tissues (Rattus norvegicus)

The effects of cold acclimation on the activity levels of cytochrome c oxidase, glutathione peroxidase and glutathione reductase in various tissues of the rat (Rattus norvegicus) were investigated. One group was individually housed at 4 +/- 1 degrees C and the other at 24 +/- 1 degrees C for 6 months. Chronic cold acclimation resulted in significantly (P < 0.05) increased cytochrome c oxidase activity levels in liver, kidney, heart, interscapular brown adipose tissue and gastrocnemius muscle. The activity of glutathione peroxidase was significantly (P < 0.05) elevated in liver, interscapular brown adipose tissue, lung and muscle, whereas glutathione reductase was only significantly (P < 0.05) elevated in interscapular brown adipose tissue as a result of chronic cold exposure. The results obtained are possibly indicative of a positive compensatory response against the increased production of oxygen derived radicals as a result of chronic cold exposure.     

Target size analysis by radiation inactivation of carnitine palmitoyltransferase activity and malonyl-CoA binding in outer membranes from rat liver mitochondria.

The functional molecular sizes of the protein(s) mediating the carnitine palmitoyltransferase I (CPT I) activity and the [14C]malonyl-CoA binding in purified outer-membrane preparations from rat liver mitochondria were determined by radiation-inactivation analysis. In all preparations tested the dose-dependent decay in [14C]malonyl-CoA binding was less steep than that for CPT I activity, suggesting that the protein involved in malonyl-CoA binding may be smaller than that catalysing the CPT I activity. The respective sizes computed from simultaneous analysis for molecular-size standards exposed under identical conditions were 60,000 and 83,000 DA for malonyl-CoA binding and CPT I activity respectively. In irradiated membranes the sensitivity of CPT activity to malonyl-CoA inhibition was increased, as judged by malonyl-CoA inhibition curves for the activity in control and in irradiated membranes that had received 20 Mrad radiation and in which CPT activity had decayed by 60%. Possible correlations between these data and other recent observations on the CPT system are discussed.     

Evidence that the sensitivity of carnitine palmitoyltransferase I to inhibition by malonyl-CoA is an important site of regulation of hepatic fatty acid oxidation in the fetal and newborn rabbit. Perinatal development and effects of pancreatic hormones in cultured rabbit hepatocytes.

The temporal changes in oleate oxidation, lipogenesis, malonyl-CoA concentration and sensitivity of carnitine palmitoyltransferase I (CPT 1) to malonyl-CoA inhibition were studied in isolated rabbit hepatocytes and mitochondria as a function of time after birth of the animal or time in culture after exposure to glucagon, cyclic AMP or insulin. (1) Oleate oxidation was very low during the first 6 h after birth, whereas lipogenesis rate and malonyl-CoA concentration decreased rapidly during this period to reach levels as low as those found in 24-h-old newborns that show active oleate oxidation. (2) The changes in the activity of CPT I and the IC50 (concn. causing 50% inhibition) for malonyl-CoA paralleled those of oleate oxidation. (3) In cultured fetal hepatocytes, the addition of glucagon or cyclic AMP reproduced the changes that occur spontaneously after birth. A 12 h exposure to glucagon or cyclic AMP was sufficient to inhibit lipogenesis totally and to cause a decrease in malonyl-CoA concentration, but a 24 h exposure was required to induce oleate oxidation. (4) The induction of oleate oxidation by glucagon or cyclic AMP is triggered by the fall in the malonyl-CoA sensitivity of CPT I. (5) In cultured hepatocytes from 24 h-old newborns, the addition of insulin inhibits no more than 30% of the high oleate oxidation, whereas it stimulates lipogenesis and increases malonyl-CoA concentration by 4-fold more than in fetal cells (no oleate oxidation). This poor effect of insulin on oleate oxidation seems to be due to the inability of the hormone to increase the sensitivity of CPT I sufficiently. Altogether, these results suggest that the malonyl-CoA sensitivity of CPT I is the major site of regulation during the induction of fatty acid oxidation in the fetal rabbit liver.     

Changes in the ability of malonyl-CoA to inhibit carnitine palmitoyltransferase I activity and to bind to rat liver mitochondria during incubation in vitro. Differences in binding at 0 degree C and 37 degrees C with a fixed concentration of malonyl-CoA.

Time courses for inhibition of carnitine palmitoyltransferase (CPT) I activity in, and [14C]malonyl-CoA binding to, liver mitochondria from fed or 48 h-starved rats were obtained at 37 degrees C by using identical incubation conditions and a fixed concentration of malonyl-CoA (3.5 microM), which represents the middle of the physiological range observed previously [Zammit (1981) Biochem. J. 198, 75-83] Incubation of mitochondria in the absence of malonyl-CoA resulted in a time-dependent decrease in the ability of the metabolite instantaneously to inhibit CPT I and to bind to the mitochondria. Both degree of inhibition and binding were restored in parallel over a period of 6-8 min on subsequent addition of malonyl-CoA to the incubation medium. However, the increased inhibition of CPT I activity on addition of mitochondria directly to malonyl-CoA-containing medium was not accompanied by an increase in the amount of [14C]malonyl-CoA bound to mitochondria at 37 degrees C. Time courses for binding of [14C]malonyl-CoA performed at 0 degree C were different from those obtained at 37 degrees C. There was little loss of ability of [14C]malonyl-CoA to bind to mitochondria on incubation in the absence of the metabolite, but there was a time-dependent increase in binding on addition of mitochondria to malonyl-CoA-containing medium. It is suggested that these temperature-dependent differences between the time courses obtained may be due to the occurrence of different changes at 37 degrees C and at 0 degree C in the relative contributions of different components (with different affinities) to the binding observed at 3.5 microM-malonyl-CoA. Evidence for multi-component binding was obtained in the form of strongly curvilinear Scatchard plots for instantaneous (5s) binding of malonyl-CoA to mitochondria. Such multi-component binding would be expected from previous results on the different affinities of CPT I for malonyl-CoA with respect to inhibition [Zammit (1984) Biochem. J. 218, 379-386]. Mitochondria obtained from starved rats showed qualitatively the same time courses as those described above, with notable quantitative differences with respect both to the absolute extents of CPT I inhibition and [14C]malonyl-CoA binding achieved as well as to the time taken to attain them.     

Regulation of fatty acid synthesis and malonyl-CoA content in mouse brown adipose tissue in response to cold-exposure, starvation or re-feeding.

1. The effect of nutritional status on fatty acid synthesis in brown adipose tissue was compared with the effect of cold-exposure. Fatty acid synthesis was measured in vivo by 3H2O incorporation into tissue lipids. The activities of acetyl-CoA carboxylase and fatty acid synthetase and the tissue concentrations of malonyl-CoA and citrate were assayed. 2. In brown adipose tissue of control mice, the tissue content of malonyl-CoA was 13 nmol/g wet wt., higher than values reported in other tissues. From the total tissue water content, the minimum possible concentration was estimated to be 30 microM 3. There were parallel changes in fatty acid synthesis, malonyl-CoA content and acetyl-CoA carboxylase activity in response to starvation and re-feeding. 4. There was no correlation between measured rates of fatty acid synthesis and malonyl-CoA content and acetyl-CoA carboxylase activity in acute cold-exposure. The results suggest there is simultaneous fatty acid synthesis and oxidation in brown adipose tissue of cold-exposed mice. This is probably effected not by decreases in the malonyl-CoA content, but by increases in the concentration of free long-chain fatty acyl-CoA or enhanced peroxisomal oxidation, allowing shorter-chain fatty acids to enter the mitochondria independent of carnitine acyltransferase (overt form) activity.     

Glycerokinase activity and lipolysis regulation in brown adipose tissue of cold acclimated rats.

Glycerol release by brown adipocytes from constant cold adapted rats was not stimulated by norepinephrine. On the contrary, the release was stimulated in rats adapted to a nycthemeral fluctuatiing temperature from 5 degrees to 28 degrees C. Glycerokinase activity was greatly increased in brown adipose tissue by cold adptation ; there was no change in the liver. However this increased activity cannot entirely explain the lack of norepinephrine stimulation of glycerol release in the brown adipose tissue of cold adapted rats.     

Interaction of malonyl-CoA and related compounds with mitochondria from different rat tissues. Relationship between ligand binding and inhibition of carnitine palmitoyltransferase I.

The sensitivity of carnitine palmitoyltransferase I (CPT I; EC 2.3.1.21) to inhibition by malonyl-CoA and related compounds was examined in isolated mitochondria from liver, heart and skeletal muscle of the rat. In all three tissues the same order of inhibitory potency emerged: malonyl-CoA much greater than succinyl-CoA greater than methylmalonyl-CoA much greater than propionyl-CoA greater than acetyl-CoA. For any given agent, suppression of CPT I activity was much greater in skeletal muscle than in liver, with the heart enzyme having intermediate sensitivity. With skeletal-muscle mitochondria a high-affinity binding site for [14C]malonyl-CoA was readily demonstrable (Kd approx. 25 nM). The ability of other CoA esters to compete with [14C]malonyl-CoA for binding to the membrane paralleled their capacity to inhibit CPT I. Palmitoyl-CoA also competitively inhibited [14C]malonyl-CoA binding, in keeping with its known ability to overcome malonyl-CoA suppression of CPT I. For reasons not yet clear, free CoA displayed anomalous behaviour in that its competition for [14C]malonyl-CoA binding was disproportionately greater than its inhibition of CPT I. Three major conclusions are drawn. First, malonyl-CoA is not the only physiological compound capable of suppressing CPT I, since chemically related compounds, known to exist in cells, also share this property, particularly in tissues where the enzyme shows the greatest sensitivity to malonyl-CoA. Second, malonyl-CoA and its analogues appear to interact with the same site on the mitochondrial membrane, as may palmitoyl-CoA. Third, the degree of site occupancy by inhibitors governs the activity of CPT I.     

The acute regulation of mitochondrial proton conductance in cells and mitochondria from the brown fat of cold-adapted and warm-adapted guinea pigs

Cells and mitochondria were prepared from the brown adipose tissue of adult guinea-pigs adapted to either 4-7 degrees C or 22-25 degrees C. The cold-adapted cells displayed noradrenaline-stimulated, propranolol-sensitive respiration, but noradrenaline failed to increase the respiration of the warm-adapted cells. Purine-nucleotide-sensitive proton conductance was greater in cold-adapted mitochondria than in warm-adapted controls. At the same time cold-adapted mitochondria were extremely sensitive to the uncoupling effect of endogenous and infused fatty acids, and resembled the mitochondria from the brown adipose tissue of cold-adapted hamsters. Warm-adapted mitochondria were ninefold less sensitive, and resembled liver mitochondria. With cold-adapted, but not warm-adapted mitochondria, respiration increased proportionately to the rate of fatty acid infusion. It is concluded that the presence of the 32000-Mr proton conductance pathway is necessary for the expression of a high sensitivity to fatty acid uncoupling, suggesting that the fatty acids interact directly with this protein to modulate the proton conductance during the acute regulation of thermogenesis.     

Carnitine acyltransferase activities in rat brain mitochondria. Bimodal distribution, kinetic constants, regulation by malonyl-CoA and developmental pattern.

Carnitine palmitoyltransferase and carnitine octanoyltransferase activities in brain mitochondrial fractions were approx. 3-4-fold lower than activities in liver. Estimated Km values of CPT1 and CPT2 (the overt and latent forms respectively of carnitine palmitoyltransferase) for L-carnitine were 80 microM and 326 microM, respectively, and K0.5 values for palmitoyl-CoA were 18.5 microM and 12 microM respectively. CPT1 activity was strongly inhibited by malonyl-CoA, with I50 values (concn. giving 50% of maximum inhibition) of approx. 1.5 microM. In the absence of other ligands, [2-14C]malonyl-CoA bound to intact brain mitochondria in a manner consistent with the presence of two independent classes of binding sites. Estimated values for KD(1), KD(2), N1 and N2 were 18 nM, 27 microM, 1.3 pmol/mg of protein and 168 pmol/mg of protein respectively. Neither CPT1 activity, nor its sensitivity towards malonyl-CoA, was affected by 72 h starvation. Rates of oxidation of palmitoyl-CoA (in the presence of L-carnitine) or of palmitoylcarnitine by non-synaptic mitochondria were extremely low, indicating that neither CPT1 nor CPT2 was likely to be rate-limiting for beta-oxidation in brain. CPT1 activity relative to mitochondrial protein increased slightly from birth to weaning (20 days) and thereafter decreased by approx. 50%.     

Effect of temperature on lipoprotein lipase and lipogenic enzyme activities in brown adipose tissue of hypophysectomized rats

It has been shown that the same modifications on the composition of brown adipose tissue (BAT) which are normally induced following cold stimulation are also observed in hypophysectomized rats acclimated either at 28 degrees C or 15 degrees C. To test the possibility of BAT stimulation in hypophysectomized rats, we have determined some enzymatic activities known to modulate the energy supply to that organ. Seven week old Long-Evans rats were hypophysectomized. Three weeks later, they were exposed to either 28 degrees C or 15 degrees C ambient temperature for five or six weeks. Hypophysectomized rats were compared to age matched or weight matched controls. Total lipoprotein lipase activity (LPL) (triglyceride uptake) was enhanced in BAT of 28 degrees C hypophysectomized rats compared to controls. Cold acclimation led to a large increased activity. Total LPL activity was comparable in BAT of hypophysectomized and control rats. Total malic enzyme and glucose-6-phosphate dehydrogenase activities (in situ lipogenesis) were doubled in BAT of 28 degrees C hypophysectomized compared to controls. A large enhancement was observed in BAT of either 15 degrees C control or 15 degrees C hypophysectomized rats. Among the studied organs (liver, white adipose tissue, heart, BAT) hypophysectomy promotes the three enzyme activities only in BAT. These variations were discussed with relation to the effect of hypophysectomy on brown adipose tissue at 15 degrees C and 28 degrees C.     

Differential response of rat brown and white adipose tissue to environmental or nutritional stress.

1. In vivo fatty acid synthesis by brown adipose tissue was enhanced in rats exposed to cold (5 degrees C) or altitude (4300 m) for 7 days but was unaltered in rats exposed to heat (35 degrees C) for an equivalent period. In vivo fatty acid synthesis by white adipose tissue was depressed by cold exposure while altitude and heat exposure had no effect. 2. In vitro, CO2 production and lipid synthesis were elevated in brown adipose tissue from rats fasted for 4 days. Refeeding (4 days) such rats reversed these effects, leading to depressed values relative to those of control rats. In contrast, these metabolic events in white adipose tissue were decreased by fasting and increased compared to controls during subsequent refeeding.     

Number of mice per cage influences uncoupling protein content of brown adipose tissue.

The effect of housing density of mice on the thermogenic state and capacity of their brown adipose tissue was studied. Mice were housed one, two, or six per cage at 28 degrees C for 15 days. Increased housing density suppressed the thermogenic capacity of brown adipose tissue (decreased the total amount of uncoupling protein) and decreased the thermogenic state of brown adipose tissue mitochondria (decreased GDP binding). A density of six mice per cage had a greater effect than a density of two mice per cage. The size of brown adipose tissue (wet weight and protein content), the content of mitochondria in it (cytochrome oxidase content), and the total activity of thyroxine 5'-deiodinase were not altered by housing density. We conclude that even at a temperature close to thermoneutrality (29-33 degrees C for the mouse), the occurrence of social thermoregulation (huddling) reduces the requirement for brown adipose tissue thermogenesis and results in a reduction in its thermogenic capacity. It is clearly of importance that the design of studies of mouse brown adipose tissue take into account not only the temperature at which the mice are housed, but also the number of mice housed per cage.