meal-feeding rodents and toxicology research

Meal-Feeding Rodents and Toxicology ResearchGale B. Carey* and Lisa C. Merrill

Department of Molecular, Cellular & Biomedical Sciences, University of New Hampshire, Durham, New Hampshire 03824, UnitedStates

ABSTRACT: Most laboratory rodents used for toxicology studies are fed adlibitum, with unlimited access to food. As a result, ad libitum-fed rodents tend toovereat. Research demonstrates that ad libitum-fed rodents are physiologicallyand metabolically different from rodents fed controlled amounts of food atscheduled times (meal-fed). Ad libitum-fed rodents can develop hyper-triglyceridemia, hypercholesterolemia, diet-induced obesity, nephropathy, car-diomyopathy, and pituitary, pancreatic, adrenal, and thyroid tumors, conditionslikely to affect the results of toxicology research studies. In contrast, meal-feedingsynchronizes biological rhythms and leads to a longer life span, lower bodyweight, lower body temperature, hypertrophy of the small intestine, andsynchronization of hepatic and digestive enzymes. The circadian rhythms presentin nearly all living organisms are entrained by light intensity and food intake, andperipheral clocks in all organs of the body, especially the GI tract and liver, are particularly sensitive to food intake. Feedingschedule has been demonstrated to alter the toxicity and metabolism of drugs including sodium valproate, chloral hydrate,acetaminophen, gentamicin, and methotrexate. Feeding schedule alters the expression of genes that code for Phase I, II, and IIIproteins, thereby altering the rate and amplitude of drug disposition. Rhythms of plasma insulin and glucagon that fluctuate withfood ingestion are also altered by feeding schedule; ad libitum feeding promotes hyperinsulinemia which is a precursor fordeveloping diabetes. The emerging field of chronopharmacology, the interaction of biological rhythms and drugs, will lead tooptimizing the design and delivery of drugs in a manner that matches biological rhythms, but it is wise for toxicology researchersto consider feeding schedule when designing these experiments. It has been 10 years since the Society for Toxicologic Pathologyvoiced its position that feeding schedule is an important variable that should be controlled in toxicology experiments, andresearch continues to underscore this position.

■ CONTENTS

Introduction 1545Circadian Clocks, Meal Feeding, and Chronophar-macology 1546Meal Feeding and Drug Metabolism 1546

Pharmacokinetics 1546Drug-Metabolizing Enzymes 1547

Meal Feeding and Metabolic Homeostasis 1547Conclusion and Implications for Toxicology Re-search 1548Author Information 1549Acknowledgments 1549Abbreviations 1549References 1549

■ INTRODUCTIONThe nutritional status of laboratory rodents used for toxicologystudies is rarely controlled. In contrast to canines, pigs, andnonhuman primates, who are fed controlled amounts of food atscheduled times, laboratory rodents are typically fed ad libitum,with constant and unlimited access to food. Given this freedom,rodents will eat intermittently throughout the dark cycle andcontinue to consume small amounts of food during the lightcycle. As a result, ad libitum-fed rodents tend to overeat.

The physiologic, metabolic, and biochemical interiore milleuof ad libitum-fed rodents dramatically alters their sensitivity toxenobiotics, compared to rodents fed restricted levels of food.1

This suggests that the meal-fed rodent is physiologically andmetabolically different from its ad libitum-fed counterpart. It isprecisely these differences that led Keenan to describe adlibitum feeding as the “most poorly controlled variable [inrodent bioassays] and it adversely affects every physiologicalprocess and anatomical structure to the molecular level.”2

Keenan has observed hypertriglyceridemia, hypercholesterole-mia, and dietary-induced obesity in ad libitum-fed ratscompared to that in rats fed a restricted diet. He has observedthe development of spontaneous nephropathy, cardiomyop-athy, and degenerative changes in multiple organs, as well asearly development and progression of pituitary, pancreas,adrenal, and thyroid tumors in ad libitum-fed compared torestricted-fed rats. One outcome of this work is the suggestionthat the increasing study-to-study variability associated withrodent bioassays since 1999, particularly in toxicity andcarcinogenicity studies, could be explained primarily by thepractice of ad libitum feeding.2−5 Indeed, unlimited access tohigh-energy food, along with a trend toward the genetic

Received: March 12, 2012Published: May 29, 2012

Perspective

pubs.acs.org/crt

© 2012 American Chemical Society 1545 dx.doi.org/10.1021/tx300109x | Chem. Res. Toxicol. 2012, 25, 1545−1550

pubs.acs.org/crt

selection of animals that grow faster and exhibit increasedfecundity, has resulted in increased rodent weights.1 Martin etal.,6 in their excellent review on the standard control rodent asmetabolically morbid, note that “the beneficial effects of somedrugs in animal models might result from their effects onprocesses associated with an unhealthy lifestyle rather than aspecific effect of the drug on the disease process. This is acritical issue that should be addressed in future studies.”One way to control the food intake of rodents in a way that

mimics that in the wild, and the focus of this review, is mealfeeding. Meal feeding, also referred to as time-limited feeding,time-restricted feeding, or timed feeding, grants animalsunlimited access to food during a defined time span. Themeal time can vary from 2 to 4 h every 24 h for smaller animalsto 8 h every 48 h for larger animals.7 This method of feedinglaboratory rodents, predominantly rats, has been employedsince the 1950s in nutrition research.Meal feeding is an important synchronizer of many behaviors

and biological rhythms, including food anticipatory activity andorganization of metabolism.8−16 Meal feeding restricts weightgain in experimental rodents compared to that in ad libitum-fedrodents: ad libitum feeding causes rapid weight gain, a trendthat continues throughout the life of the animal.17−19 Otherphysiologic and metabolic adaptations that occur in the meal-fed versus the ad libitum-fed rodent are a longer life span, lowerbody temperature, decreased mitotic activity, hypertrophy ofthe stomach and small intestines, delayed gastric emptying,alteration of circadian rhythms, and synchronization ofnumerous hepatic and digestive enzymes.8−10,20−28

Does the feeding protocol of rodents used in toxicologicalstudies influence research findings? Findings from studiesconducted over the past 40 years would suggest that it does.This perspective will examine many of these findings, withparticular attention to how meal feeding affects chronopharma-cology, drug metabolism, and metabolic homeostasis. The goalof this perspective is to stimulate the thinking of toxicologists toconsider feeding protocol when designing experiments.

■ CIRCADIAN CLOCKS, MEAL FEEDING, ANDCHRONOPHARMACOLOGY

The master circadian clock in mammals resides in thesuprachiasmatic nucleus (SCN) located in the ventral part ofthe anterior hypothalamus.29,30 The SCN is entrained primarilyby the 24-h variation in light intensity31 and is relativelyresistant to the effects of food intake. The SCN signalssecondary clocks in the form of clock genes, including Clock,Bmal1, Per1−3, and Cry1−2, and their protein products in boththe SCN and brain as well as peripheral tissues such as the liver,heart, muscle, kidney, pancreas, adipose tissue, leukocytes,smooth muscle, and lung.32,33 While light synchronizes themaster circadian clock, the feeding−fasting cycle can shift clockgene oscillations in the periphery.32,34

Using transgenic mice in which a luciferase reporter waslinked to the Per1 gene, Stokkan et al.34 demonstrated thatmeal-feeding animals for a 4 h-period each day during the lightcycle caused a 12-h shift in hepatic clock gene expression afteronly two days. The response of the lung tissue was slower andless pronounced, shifting six hours after seven days of mealfeeding. There was no change in clock expression in the SCN,which remained phase-locked to the light−dark cycle. The liverand lung shifts did not respond to the corticosterone injection,suggesting that other feeding signals such as taste of food,stomach distension, metabolite, or insulin levels may be

responsible. These findings were confirmed by Damiola etal.,32 by measuring mRNA expression of several clock genes andtranscription factors in tissues of mice under time-restrictedfeeding conditions. Their findings were consistent with those ofStokkan et al.34 and were extended to demonstrate that theresponse of the liver was most abrupt, followed by the kidney,heart, and pancreas and that even in mice kept under constantdarkness, restricted feeding uncoupled the response ofperipheral clock genes from the SCN, confirming that feedingtime does not entrain SCN neurons. Damiola et al. proposethat the role of circadian gene expression in the liver is torespond to the processing of food.32

How does this relate to toxicology? It has long been knownthat the toxicity, efficacy, metabolism, and elimination of manydrugs change over a 24-h period.35 For example, Kim and Leereported that a single 400 mg/kg i.p. dose of acetaminophenwas hepatotoxic to mice at 20:00 h but not at 8:00 or 14:00.36

More recent knowledge about the clock genes and othermolecular mechanisms that underlie these circadian changeshas given rise to the field of chronopharmacology, the study ofthe interaction of biological rhythms and drugs, andchronopharmaceutics, the design and evaluation of deliveringdrugs in a time-controlled manner and rhythm that matchesbiological requirements.30 The inherent circadian rhythms ofmany diseases, including asthma, arthritis, cancer, diabetes,hypertension, and hypercholesterolemia, make chronopharma-ceutics a promising field for the health care industry.30

■ MEAL FEEDING AND DRUG METABOLISM

Given, then, that drug metabolism is subject to circadianrhythm and that meal feeding can entrain peripheral circadianrhythms, a logical question is this: does the pattern of foodintake alter the metabolism of drugs? Nakano et al.37 were thefirst to report that feeding schedule could influence thecircadian rhythm of drug action and kinetics. Mice fed for 8 hduring the light period demonstrated a 12-h shift in their toxicresponse to phenobarbital, compared to that by ad libitum-fedmice. Song et al. found that mice injected with 100 mg/kg i.p.methotrexate38 or 180 mg/kg s.c. gentamicin39 and fed for 8 hduring the light period had a higher mortality, higher plasmadrug concentration, and lower clearance of the drug during thedark (nonfeeding) period compared to that in mice fed adlibitum. Feeding schedule has been demonstrated to alter thetoxicity and metabolism of sodium valproate,40 chloralhydrate,41 and acetaminophen.42 It is clear, then, thatcontrolled food intake, in comparison to the usual practice ofad libitum feeding, has striking effects on the toxicity, efficacy,and processing of drugs.

Pharmacokinetics. Song et al. examined the effect offeeding schedule on the pharmacokinetics of gentamicin, anantibiotic used to treat gram-negative bacterial infections,39 andmethotrexate, a potent anticancer agent.40 Six-week old ICRmice were meal-fed for 8 h during the light cycle, or fed adlibitum. After 14 days, mice were injected with 180 mg ofgentamicin/kg s.c. or 100 mg of methotrexate/kg i.p. Theplasma gentamicin concentration 30 min after dosing and totalbody clearance of gentamicin was shifted by 12 h as a result ofthe meal feeding, demonstrating that feeding schedule modifiesgentamicin pharmacokinetics. Methotrexate was even moresensitive to meal feeding, in which plasma levels at critical timesof the day were from 50% to 130% higher compared to that inad libitum-fed mice.

Chemical Research in Toxicology Perspective

dx.doi.org/10.1021/tx300109x | Chem. Res. Toxicol. 2012, 25, 1545−15501546

Ohdo et al.40 examined the influence of meal feeding (8 hduring the light phase of 0700−1900) of 5-week old male ICRmice on the pharmacokinetics and efficacy of valproic acid(VA), an antiepileptic drug. When VA was administered as a600 mg/kg oral bolus at 0900, the plasma level of VA 30 minlater was the highest in meal-fed mice (1550 ug/mL) butlowest in ad libitum-fed mice (500 ug/mL), with the curvesinverse to each other. When VA was constantly infused at1072.6 ug/h at 0900, plasma VA was identical for ad libitum-and meal-fed mice at 0900 (approximately 33 ug/mL).However, the levels peaked at 55 ug/mL for ad libitum-fedmice at 1700, then declined to 25 ug/mL at 0500, while levelsremained low at 25 ug/mL for meal-fed mice at 1700, then roseto 55 ug/mL at 0500. The authors conclude that VA action andkinetics are altered by feeding schedule regardless of the light−dark cycle, suggesting feeding may be a stronger synchronizerof VA kinetics and response. Moreover, short-term manipu-lation of feeding, that is, fasting the ad libitum-fed mice for 12 hbefore the experiment, did not overcome the rhythm created bythe feeding schedule. The authors speculate that regional bloodflow, which is altered by food intake, may be a controlling factorby which food intake affects drug kinetics and activity.Moreover, the authors note that short-term manipulations offeeding do not overcome metabolic shifts induced by mealfeeding.Drug-Metabolizing Enzymes. The metabolism of drugs is

conducted by three groups of phase I, II, and III proteins. PhaseI proteins are the microsomal cytochrome P450 superfamily ofenzymes with oxidase, reductase, and hydrolase activities thatcan inhibit or activate drugs, while phase II proteins contain thesulfotransferases, UDP-glucuronotransferases, glutathione-S-transferases, and N-acetyltransferases that can conjugate drugsto deactivate them or alter their hydrophilicity. Phase IIItransporters contain the solute carrier and ATP-bindingcassette proteins that transport drugs across biologicalmembranes, into or from the cells of the body.The genes responsible for expressing phase I, II, and III

proteins are subject to circadian rhythm. Using male C57BL/6mice on a 12 h light−dark cycle (lights on 5 a.m.) and fed adlibitum, Zhang et al.43 demonstrated that phase I enzymes aremore highly expressed during the dark phase, while the phase IIenzyme expression pattern suggests more glutathione con-jugation in the early light phase, glucuronidation in the late-light phase, and sulfation in the early dark phase. A keyquestion that arises from this work is as follows: Is the circadianexpression of drug-metabolizing enzymes, which can beeliminated in SCN-lesioned rats, influenced by peripheralclocks that respond to food intake?Hirao et al.44 addressed this question by examining the effect

of meal feeding on total hepatic phase I enzyme activity. MaleF344/DuCrj rats were divided into two groups: ad libitum fedor meal fed for 8 h during the light period. After an 8-dayadaptation, the left lateral lobes of the livers were removed andassayed for hepatic 7-alkoxycoumarin O-dealkylase activitywhich reflects global phase I activity. Meal feeding during thelight cycle inverted the circadian rhythm of hepaticalkoxycoumarin O-dealkylase activity, demonstrating that it iscircadian factors in peripheral, not central, organs that drivehepatic phase I enzyme rhythms.In mice, the commonly used analgesic acetaminophen is

bioactivated to N-acetyl-p-benzoquinone imine (NAPQI),which can deplete hepatic glutathione and cause acetamino-phen toxicity. This bioactivation occurs via the cytochrome

P450 enzymes CYP 1A2 and CYP2E1, while detoxificationoccurs by conjugation to glucuronic acid and sulfate. Given thatglutathione levels demonstrate circadian rhythm and thatfasting can alter the 24-h rhythm of acetaminophen toxicity,Matsunaga et al.42 sought to determine the 24-h rhythm of CYP2E1 activity under meal-feeding conditions.Male 5-wk old ICR mice that were meal fed for 8 h during

the light cycle (0700−1900) or ad libitum fed were injecteddaily with 0, 300, or 600 mg/kg acetaminophen i.p.42 After twoweeks of treatment, hepatic CYP2E1 activity was measured at4-h intervals over a 24-h period. In the absence of drugtreatment, CYP2E1 activity in the livers of ad libitum-fed micepeaked at 9 a.m., at a level that was nearly 50% of the peak inlivers of meal-fed mice, which peaked at 9 p.m. Moreover, meal-fed mice had a striking circadian rhythm in CYP 2E1 activitythat was absent in the ad libitum mice. The 24-h rhythm inhepatic glutathione levels in meal-fed mice was inverse and hadgreater amplitude than that in ad libitum-fed mice. The authorsconclude that the 24-h rhythm of glutathione levels andCYP2E1 activity, which are altered by feeding regimen, maydictate the hepatotoxicity of acetominophen.

■ MEAL FEEDING AND METABOLIC HOMEOSTASISJust as meal feeding entrains circadian rhythms associated withthe metabolism of drugs, it can entrain rhythms associated withthe metabolism of food. This entrainment can improve thehealth and well being of the animal as well as the accuracy ofexperimental data; both of these are important considerationsfor toxicology researchers.The major dietary sources of energy in animals are

carbohydrates and lipids, whose primary constituents areglucose and fatty acids, respectively. Once ingested, theseconstituents are either oxidized immediately for energy orstored as glycogen (liver and muscle) and triacylglycerol(adipose tissue) for mobilization at a later time when energy isneeded. The master conductors of this metabolic dancebetween “using it now” vs “storing and using it later” areinsulin and glucagon. The release of these two hormones ishard-wired into the metabolic machinery of all animals in orderto maintain normal blood glucose levels, essential for brainfunctioning and thus life. When dietary carbohydrates aredigested and absorbed, the rise in blood glucose levels and theresulting release of insulin from the β cells of the pancreaspromotes the uptake and conversion of glucose to glycogen andfatty acids by the liver. Glucose not removed by the liver passesinto the peripheral blood where insulin stimulates its uptakeinto skeletal muscle and adipose tissue for oxidation, conversionto fatty acids, and storage as triacylglycerol, or storage asglycogen. When dietary lipids are ingested, fatty acids are eitheroxidized directly by most tissues or esterified and stored astriacylglycerol in adipose tissue for later use.In contrast, when an animal does not eat, the metabolic

dance shifts to mobilizing stored energy. A drop in bloodglucose and the resulting release of glucagon from the α cells ofthe pancreas, promotes liver glycogen breakdown and the denovo synthesis of glucose from endogenous substrates. Thishepatic response provides glucose to the blood to maintainnormal blood glucose levels. A rise in glucagon, along with adrop in insulin, promotes the mobilization of triacylglycerolfrom adipose tissue, releasing fatty acids and glycerol intoperipheral blood. Fatty acids can then serve as an energysource, and glycerol can serve as a gluconeogenic precursor.The liver circumvents the inability of free fatty acids to



penetrate the blood−brain barrier by converting fatty acids toketone bodies, which are soluble and penetrate the blood−brain barrier. Ketone body synthesis reduces the brain’s relianceon glucose and lessens the demand on the liver to produceglucose.Thus, the meal-fed animal is metabolically distinct from its ad

libitum-fed counterpart. When food availability is restricted, themeal-fed animal must store fuel quickly and mobilize it slowlythroughout the day. For example, when Escobar et al.45

restricted the feeding of rats to a daily two-hour period forthree weeks, hepatic glycogen content in meal-fed rats wasnearly depleted one hour prior to the scheduled feeding time,and free fatty acid and ketone body levels were significantlyelevated compared to ad libitum-fed rats, whose hepaticglycogen was still present and free fatty acids, and ketonebodies were not elevated. An example of tightened metabolichomeostasis with meal feeding is that throughout a day, meal-fed animals maintain plasma glucose concentrations approx-imately 17 mg/dL below their ad libitum-fed counterparts (p <0.001).46 The only time this is not the case is after the meal-fedanimals are provided with food, at which time, their plasmaglucose levels approach those of the ad libitum-fed rats. Indeed,the rise in plasma glucose following glucose ingestion isdampened, and the values return to basal more quickly, in meal-fed rats compared to that in the ad libitum-fed rats.47−49 Thissuggests a greater insulin sensitivity of adipose and muscletissues.The suggestion that insulin sensitivity of adipose tissue in

meal-fed rats was greater than that in ad libitum-fed rats wastested in our laboratory.50 Young, male Wistar rats were eitherfed ab libitum or maintained on a two-hour meal-feedingschedule for approximately three weeks. Experiments onepididymal fat were performed immediately after feeding, 10h after feeding, and 20 h after feeding, using the method ofRodbell51 in which adipocytes were isolated and incubated with14C-glucose in the absence and presence of insulin. Whenadipocytes from ad libitum-fed rats were incubated with 14C-glucose plus 10−7 M insulin, glucose oxidation increased 2-fold,while oxidation in adipocytes from meal-fed rats in the presenceof insulin increased by 3- to 4- fold (Figure 1). Moreover, thestandard deviation and coefficient of variation for glucoseoxidation in adipocytes isolated from ad libitum-fed rats are

nearly double those of adipocytes from meal-fed animals. Bythe 20-h time point, the absolute variability in both groups ofanimals had decreased to values that were well below thosemeasured immediately after feeding, suggesting that time fromfood ingestion impacts experimental results. These findingsconfirm that food intake is a major source of variation in themeasurement of adipocyte glucose oxidation and that insulinsensitivity of adipose tissue from meal-fed rats is greater thanthat of ad libitum-fed rats.The heightened insulin sensitivity of the meal-fed animal and

thus a reduced need for insulin was demonstrated by Sugden etal.52 Four weeks of meal feeding resulted in a plasma insulinconcentration of 15 μU/mL before food intake and a rise to 30μU/mL 2 h after the provision of food (p < 0.5); the insulinlevel returned to baseline 4 h after the removal of food. Incontrast, peak plasma insulin concentration in ad libitum-fedrats was nearly 47% higher than that of meal-fed rats, at 44 μU/mL. Twelve weeks of meal feeding demonstrated that thebaseline serum insulin level in meal-fed rats was 65% that of adlibitum-fed rats (18.1 versus 28.0 μU/mL, p < 0.05) 12 to 14 hafter food removal.19

Two separate studies, varying in length of feeding time,confirmed these findings. The first, a 3-week study conductedon adult, male Wistar rats by Diaz-Munoz et al.,53 showed thatprior to feeding a 2-h meal, the serum insulin level was 50% ofthe ad libitum-fed control group (p < 0.01). After feeding,insulin levels increased to above control, remaining higher thanthe control 4 h postprandial, then gradually returning to 50% ofad libitum-fed levels. The second study, lasting 20 months,revealed plasma insulin levels that were consistently andsignificantly lower in meal-fed rats than in ad libitum-fed rats (p< 0.05) throughout the entire study at all measured timepoints.46 Thus, data supports the notion that meal feedingpromotes insulin sensitivity and that ad libitum feedingpromotes hyperinsulinemia, a precursor for the developmentof diabetes.In summary, the physiological, metabolic, enzymatic, and

genetic responses of rodents to a meal-feeding regimen canhave dramatic effects on the disposition of drugs and animalhealth. What should be done with this information?

■ CONCLUSION AND IMPLICATIONS FORTOXICOLOGY RESEARCH

Meal-fed rodents adapt, both physiologically and metabolically,to having limited access to food. They live longer, are lessdisease-prone, and are more insulin-sensitive than their adlibitum-fed counterparts. The enhanced insulin sensitivity ofthe meal-fed rodent suggests that toxicology researchers usingan ad libitum model are unknowingly using a diabetic-typemodel.Meal feeding is associated with less experimental variation

than that observed with ad libitum feeding: by controlling thetiming of food intake, toxicologists can more accurately andreproducibly determine an animal’s temporal and metabolichandling of the drug, compared to the ad libitum-fed animal, inwhich time from eating is unknown.From a practical standpoint, timed feeding is attainable.

Toxicology experiments could be conducted under a reverselight−dark cycle (lights on 7 a.m.−7 p.m.) in which rodents arefed between 9 a.m. and 5 p.m. Feeding can be done manually orvia a timed-feeding device. An alternative approach to reduceanimal morbidity has been used by National ToxicologyProgram studies, the NTP-2000 diet. This diet replaces NIH-

Figure 1. Effect of time of day on fold-increase of insulin-stimulatedglucose oxidation over basal in adipocytes from ad libitum (AD, blue)-and 2-hr meal-fed (MF, red) rats. Results are expressed as means ±SD, n = 4−6. Coefficient of variation around the mean at 11 a.m., 9p.m., and 7 a.m. is 47%, 61%, and 21% for ad libitum-fed rats(unpublished data), and 27%, 27%, and 17% for meal-fed rats,50

respectively.



07, as certain components of NIH-07 such as protein content,calcium-to-phosphorus ratio, and vitamin D concentration,were thought to contribute to age-related illnesses. Rodents fedNTP-2000 showed significantly increased survival time andreduced nephropathy and cardiomyopathy.54

It has been 10 years since the publication of the Society forToxicologic Pathology position paper noting feeding as animportant variable that should be controlled in toxicologyexperiments.5 Data support the notion that circadian clocks andtheir gene products including CYPs, respond to food intake.Therefore, it is crucial that feeding regimen be carefullyconsidered in designing toxicology experiments.

■ AUTHOR INFORMATION

Corresponding Author*Phone: (603) 862-4628. Fax: (603) 862-1148. E-mail: [email protected].

NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTS

We thank Lauren Hufnagle for conducting experiments on adlibitum-fed rats and the NH Agricultural Experiment Station forfinancial support, contributing to Figure 1.

■ ABBREVIATIONSCYP, cytochrome P450; SCN, suprachiasmatic nucleus; VA,valproic acid

■ REFERENCES(1) Hart, R. W., Newmann, D. A., and Robertson, R. T. Eds. (1995)Dietary Restriction: Implications for the Design and Interpretation ofToxicity and Carcinogenicity Studies, ILSI Press, Washington, DC.(2) Keenan, K. P., Ballam, G. C., Soper, K. A., Laroque, P., Coleman,J. B., and Dixit, R. (1999) Diet, caloric restriction, and the rodentbioassay. Toxicol. Sci. 52, S24−34.(3) Keenan, K. P., Ballam, G. C., Dixit, R., Soper, K. A., Laroque, P.,Mattson, B. A., Adams, S. P., and Coleman, K. B. (1997) The effects ofdiet, overfeeding and moderate dietary restriction on Sprague-Dawleyrat survival, disease and toxicology. J. Nutr. 127, 851S−856S.(4) Keenan, K. P., Hoe, C., Mixson, L., McCoy, C. L., Coleman, J. B.,Mattson, B. A., Ballam, G. A., Gumprecht, L. A., and Soper, K. A.(2005) Diabesity: a polygenic model of dietary-induced obesity fromad libitum overfeeding of Sprague-dawley rats and its modulation bymoderate and marked dietary restriction. Toxicol. Pathol. 33, 650−674.(5) Nold, J. B., Keenan, K. P., Nyska, A., and Cartwright, M. E.(2001) Society of Toxicologic Pathology Position Paper: Diet as avariable in rodent toxicology and carcinogenicity studies. Toxicol.Pathol. 29, 585−586.(6) Martin, B., Sunggoan, J., Maudsley, S., and Mattson, M. P. (2010)Control” laboratory rodents are metabolically morbid: Why it matters.Proc. Natl. Acad. Sci. U.S.A. 107, 6127−6133.(7) Carey, G. B., Cheung, C., Cohen, N. S., Brusilow, S., andRaijman, L. (1993) Regulation of urea and citrulline synthesis underphysiological conditions. Biochem. J. 292, 241−247.(8) Cohn, C., and Joseph, D. (1960) Role of rate of ingestion of dieton regulation of intermediary metabolism (“meal eating” vs“nibbling”). Metabolism 9, 492−500.(9) Boulos, Z., and Terman, M. (1980) Food availability and dailybiological rhythms. Neurosci. Biobehav. Rev. 4, 119−131.(10) Phillipens, K. M. H. (1980) Synchronization of Rhythms toMeal Timing, in Chronobiology: Principles and Applications to Shifts inSchedules (Scheving, L. E., and Halberg, F., Eds.) Sithoff andNoordhoff, Alpehn an den Rijn, The Netherlands.

(11) Ahlers, I., Ahlersova, E., and Smajda, B. (1983) Circadianrhythm of serum and tissue lipids in the rat: the effect of limited accessto food. Physiol. Bohemoslov. 32, 64−72.(12) Ahlers, I., Ahlersova, E., Datelinka, I., and Toropila, M. (1984)Effect of food synchronization on the circadian rhythm of bonemarrow and thymus lipids in rats. Physiol. Bohemoslov. 33, 511−514.(13) Comperatore, C. A., and Stephan, F. K. (1987) Entrainment ofduodenal activity to periodic feeding. J. Biol. Rhythms 2, 227−242.(14) White, W., and Timberlake, W. (1995) Two meals promoteentrainment of rat food-anticipatory and rest-activity rhythms. Physiol.Behav. 57, 1067−1074.(15) Stokkan, K., Yamasaki, S., Tei, H., Sakaki, Y., and Menaker, M.(2001) Entrainment of the circadian clock in the liver by feeding.Science 291, 490−493.(16) Satoh, Y., Kawai, H., Kudo, N., Kawashima, Y., and Mitsumoto,A. (2006) Time-restricted feeding entrains daily rhythms of energymetabolism in mice. Am. J. Physiol. Regul. Integr. Comp. Physiol. 290,R1276−1283.(17) DiGirolamo, M., and Rudman, D. (1968) Variations in glucosemetabolism and sensitivity to insulin of the rat’s adipose tissue, inrelation to age and body weight. Endocrinology 82, 1133−1141.(18) Salans, L. B., and Dougherty, J. W. (1971) The effect of insulinupon glucose metabolism by adipose cells of different size. J. Clin.Invest. 50, 1399−1410.(19) Reiser, S., and Hallfrisch, J. (1977) Insulin sensitivity andadipose tissue weight of rats fed starch or sucrose diets ad libitum or inmeals. J. Nutr. 107, 147−155.(20) Tepperman, J., and Tepperman, H. M. (1958) Effects ofantecedent food intake pattern on hepatic lipogenesis. Am. J. Physiol.193, 55−64.(21) Holeckova, E., and Fabry, P. (1959) Hyperplasia and gastrichypertrophy in rats adapted to intermittent starvation. Br. J. Nutr. 13,260−266.(22) Hollifield, G., and Parson, W. (1962) Metabolic adaptations to a“stuff and starve” feeding program. I. Studies of adipose tissue and liverglycogen in rats limited to a short daily feeding period. J. Clin. Invest.41, 245−249.(23) Bolles, R. C., Duncan, P. M., Grossen, N. E., and Matter, C. F.(1968) Relationship between activity level and body temperature inthe rat. Psychol. Rep. 23, 991−994.(24) Saito, M., Murakami, E., and Suda, M. (1976) Circadianrhythms in disaccharidases of rat small intestine and its relation to foodintake. Biochim. Biophys. Acta 421, 177−179.(25) Lanza-Jacoby, S., Stevenson, N. R., and Kaplan, M. L. (1986)Circadian changes in serum and lipid metabolites and liver lipogenicenzymes in ad libitum-and meal-fed, lean and obese Zucker rats. J.Nutr. 116, 1798−1809.(26) Nelson, W., and Halberg, F. (1986) Meal-timing, circadianrhythms and life span of mice. J. Nutr. 116, 2244−2253.(27) Kohsaka, A., and Bass, J. (2006) A sense of time: how molecularclocks organize metabolism. Trends Endocrinol. Metab. 18, 4−11.(28) Varady, K. A., Roohk, D. J., and Hellerstein, M. E. (2007) Doseeffects of modified alternate-day fasting regimens on in vivo cellproliferation and plasma insulin-like growth factor-1 in mice. J. Appl.Physiol. 103, 547−551.(29) Mendoza, J. (2007) Circadian clocks: setting time by food. J.Neuroendocrinol. 19, 127−137.(30) Youan, B. B. (2004) Chronopharmaceutics: gimmick orclinically relevant approach to drug delivery? J. Controlled Release 98,337−353.(31) Ohdo, S., Koyanagi, S., Matsunaga, N., and Hamdan, A. (2011)Molecular basis of chronopharmaceutics. J. Pharm. Sci. 100, 3560−3576.(32) Damiola, F., Minh, N. L., Preitner, N., Kornmann, B., Fleury-Olela, F., and Schibler, U. (2000) Restricted feeding uncouplescircadian oscillators in peripheral tissues from the central pacemaker inthe suprachiasmatic nucleus. Genes Dev. 14, 2950−2961.(33) Balsalobre, A. (2002) Clock genes in mammalian peripheraltissues. Cell Tissue Res. 309, 193−199.



mailto:[email protected]

mailto:[email protected]

(34) Stokkan, K., Yamazaki, S., Tei, H., Sakaki, Y., and Menaker, M.(2001) Entrainment of the circadian clock in the liver by feeding.Science 291, 490−493.(35) Gachon, F., and Firsov, D. (2011) The role of circadian timingsystem on drug metabolism and detoxification. Expert Opin. DrugMetab. Toxicol. 7, 147−158.(36) Kim, Y. C., and Lee, S. J. (1998) Temporal variation inhepatotoxicity and metabolism of acetaminophen in mice. Toxicology128, 53−61.(37) Nakano, S., Nagai, K., and Ogawa, N. (1984) Role of feedingschedule and pharmacokinetics in chronopharmacology of drugs actingon the central nervous system. Ann. Rev. Chronopharm. 1 (1), 263−266.(38) Song, J. G., Nakano, S., Ohdo, S., and Ogawa, N. (1993)Chronotoxicity and chronopharmacokinetics of methotrexate in mice:modification by feeding schedule. Japan. J. Pharmacol. 62, 373−378.(39) Song, J., Ohdo, S., Ogawa, N., and Nakano, S. (1993) Influenceof feeding schedule on chronopharmacological aspects of gentamicinin mice. Chronobiol. Int. 10, 338−348.(40) Ohdo, S., Ogawa, N., Nakano, S., and Higuchi, S. (1996)Influence of feeding schedule on the chronopharmacological aspects ofsodium valproate in mice. J. Pharm. Exp. Ther. 278, 74−81.(41) Seng, J. E., Agrawal, N., Horsley, E. T. M., Leakey, T. I., Scherer,E. M., Xia, S., Allaben, W. T., and Leakey, J. E. A. (2003)Toxicokinetics of chloral hydrate in ad libitum-fed, dietary controlled,and calorically restricted male B6C3F1 mice following short-termexposure. Toxicol. Appl. Pharmacol. 193, 281−292.(42) Matsunaga, N., Nakamura, N., Yoneda, N., Qin, T., Terazono,H., To, H., Higuchi, S, and Ohdo, S. (2004) Influence of feedingschedule on 24-h rhythm of hepatotoxicity induced by acetaminophenin mice. J. Pharm. Exp. Ther. 311, 594−600.(43) Zhang, Y. J., Yeager, R. L., and Klaassen, C. D. (2009) Circadianexpression profiles of drug-processing genes and transcription factorsin mouse liver. Drug Metab. Dispos. 37, 106−115.(44) Hirao, J., Arakawa, S., Watanabe, K., Ito, K., and Furukawa, T.(2006) Effects of restricted feeding on daily fluctuations of hepaticfunctions including P450 monooxygenase activities in rats. J. Biol.Chem. 281, 3165−3171.(45) Escobar, C., Diaz-Munoz, M., Encinas, F., and Aguilar-Roblero,R. (1998) Persistence of metabolic rhythmicity during fasting and itsentrainment by restricted feeding schedules in rats. Am. J. Physiol. 274,R1309−R16.(46) Masoro, E. J., McCarter, R. J. M., Katz, M. S., and McMahan, C.A. (1992) Dietary restriction alters characteristics of glucose fuel use. J.Gerontol. 47, B202−B208.(47) Leveille, G. A., and Chakrabarty, K. (1968) Absorption andutilization of glucose by meal-fed and nibbling rats. J. Nutr. 96, 69−75.(48) Wiley, J. H., and Leveille, G. A. (1970) Significance of insulin inthe metabolic adaptation of rats to meal ingestion. J. Nutr. 100, 1073−1080.(49) Romsos, D. R., and Leveille, G. A. (1974) Effect of mealfrequency and diet composition on glucose tolerance in the rat. J. Nutr.104, 1503−1512.(50) Merrill, L. (2008) Factors Affecting Insulin-Stimulated in VitroGlucose Oxidation in Rat Epididymal Adipocytes, Master’s Thesis,University of New Hampshire, Durham, NH.(51) Rodbell, M. (1964) Metabolism of isolated fat cells. J. Biol.Chem. 239, 375−380.(52) Sugden, M. C., Grimshaw, R. M., and Holness, M. J. (1999)Caloric restriction leads to regional specialization of adipocyte functionin the rat. Biochim. Biophys. Acta 1437, 202−213.(53) Diaz-Munoz, M., Vazquez-Martinez, O., and Aguilar-Roblero, R.(2000) Anticipatory changes in liver metabolism and entrainment ofinsulin, glucagon, and corticosterone in food-restricted rats. Am. J.Physiol. 279, R2048−R2056.(54) Rao, G. N., Morris, R. W., and Seely, J. C. (2001) Beneficialeffects of NTP-2000 diets on growth, survival, and kidney and heartdiseases of Fischer 344 rats in chronic studies. Toxicol. Sci. 63, 245−255.



meal-feeding rodents and toxicology research

Documents