nutrition and cancer evidence for a novel serum factor ......body fat before losing protein mass and...

PLEASE SCROLL DOWN FOR ARTICLE

This article was downloaded by: [LSU Medical Center]On: 11 May 2010Access details: Access Details: [subscription number 917399047]Publisher RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Nutrition and CancerPublication details, including instructions for authors and subscription information:http://www.informaworld.com/smpp/title~content=t775653687

Evidence for a Novel Serum Factor Distinct From Zinc Alpha-2Glycoprotein That Promotes Body Fat Loss Early in the Development ofCachexiaLauri O. Byerley abc; Sang Ho Lee a; Steve Redmann d; Cathy Culberson a; Mark Clemens a;Mark O.Lively e

a University of North Carolina at Charlotte, Charlotte, North Carolina, USA b Pennington BiomedicalResearch Center, Baton Rouge, Louisiana, USA c Louisiana State University Health Science Center,New Orleans, Louisiana, USA d Redmann Consulting, St. Gabriel, Louisiana, USA e Wake ForestUniversity School of Medicine, Winston-Salem, North Carolina, USA

Online publication date: 27 April 2010

To cite this Article Byerley, Lauri O. , Lee, Sang Ho , Redmann, Steve , Culberson, Cathy , Clemens, Mark andLively, MarkO.(2010) 'Evidence for a Novel Serum Factor Distinct From Zinc Alpha-2 Glycoprotein That Promotes Body Fat LossEarly in the Development of Cachexia', Nutrition and Cancer, 62: 4, 484 — 494To link to this Article: DOI: 10.1080/01635580903441220URL: http://dx.doi.org/10.1080/01635580903441220

Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf

This article may be used for research, teaching and private study purposes. Any substantial orsystematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply ordistribution in any form to anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contentswill be complete or accurate or up to date. The accuracy of any instructions, formulae and drug dosesshould be independently verified with primary sources. The publisher shall not be liable for any loss,actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directlyor indirectly in connection with or arising out of the use of this material.

http://www.informaworld.com/smpp/title~content=t775653687

http://dx.doi.org/10.1080/01635580903441220

http://www.informaworld.com/terms-and-conditions-of-access.pdf

Nutrition and Cancer, 62(4), 484–494Copyright C© 2010, Taylor & Francis Group, LLCISSN: 0163-5581 print / 1532-7914 onlineDOI: 10.1080/01635580903441220

Evidence for a Novel Serum Factor Distinct From ZincAlpha-2 Glycoprotein That Promotes Body Fat Loss Early inthe Development of Cachexia

Lauri O. ByerleyUniversity of North Carolina at Charlotte, Charlotte, North Carolina; Pennington Biomedical ResearchCenter, Baton Rouge, Louisiana; and Louisiana State University Health Science Center, New Orleans,Louisiana, USA

Sang Ho LeeUniversity of North Carolina at Charlotte, Charlotte, North Carolina, USA

Steve RedmannRedmann Consulting, St. Gabriel, Louisiana, USA

Cathy Culberson and Mark ClemensUniversity of North Carolina at Charlotte, Charlotte, North Carolina, USA

Mark O. LivelyWake Forest University School of Medicine, Winston-Salem, North Carolina, USA

We provide evidence that a factor other than the previouslyidentified lipid mobilizing factor, zinc alpha-2 glycoprotein, pro-motes lipolysis in the MCA-induced sarcoma-bearing cachexiamodel. Cachexia is characterized by progressive loss of adiposetissue and skeletal muscle without a concurrent increase in foodintake to restore lost tissue stores. We compared tumor-bearing adlib fed (TB) animals to nontumor bearing ad lib fed (NTB) ani-mals or nontumor-bearing pair-fed (PF) animals at various timepoints throughout development of tumor derived cachexia. Priorto cachexia, the TB animals lost more than 10 ± 0.7% of theirbody fat before losing protein mass and decreasing their food in-take. Fat loss occurred because adipocyte size, not number, wasreduced. Increased turnover of palmitate and significantly higherserum triglyceride levels prior to cachexia were further indicatorsof an early loss of lipid from the adipocytes. Yet, circulating levelsof norepinephrine, epinephrine, TNF-α, and zinc α-2 glycoproteinwere not increased prior to the loss of fat mass. We provide evi-dence for a serum factor(s), other than zinc alpha-2 glycoprotein,that stimulates release of glycerol from 3T3-L1 adipocytes and pro-motes the loss of stored adipose lipid prior to the loss of lean bodymass in this model.

Submitted 26 March 2009; accepted in final form 13 August 2009.Address correspondence to L. O. Byerley, Physiology Department,

Louisiana State University Health Science Center, 1901 Perdido St.,New Orleans, LA 70112. Phone: 504-568-6170. Fax: 504-568-6158.E-mail: [email protected]

INTRODUCTION

Several decades ago, DeWys and colleagues (1,2) publishedtheir seminal papers showing weight loss was common amongcancer patients and that that weight loss significantly impactedprognosis. This weight loss, termed cachexia for the Greek wordmeaning bad condition, includes both lean and fat body mass.Although there is no concise definition for this condition (3),generally a cancer patient is considered cachectic if they haveunintentionally lost more than 10% of their usual body weightin a year without concurrent anorexia (4).

Are lean and fat mass lost simultaneously or is one lost beforethe other? Several studies, have reported on body compositionalchanges in human cancer patients using a variety of techniquesincluding anthropometry (5–7); whole body potassium deter-mination (8–11); neutron activation (12,13); computerized to-mography (14); bioelectric impedance (15,16); and the goldstandard, dual-energy, x-ray absorptiometry (DXA or DEXA)(5,17–20). Most studies have been cross-sectional, but a fewwere longitudinal. None of the studies have followed the can-cer patient before and after the onset of cachexia. Despite thelimitations of study designs and measurement techniques, moststudies have indicated body fat declines first when a patientsuffers from a progressive disease (6,21).

A recent study evaluated time course of changes in bodycomposition, measured using DEXA, in cancer patients who

484

Downloaded By: [LSU Medical Center] At: 14:34 11 May 2010

CACHEXIA AND LIPOLYSIS 485

received palliative care (systemic anti-inflammatory treatment,recombinant erythropoietin treatment, and/or specialized nutri-tional support care) during a 4- to 62-mo follow-up of theirdisease (22). At the time of admission to the study, on average,the cohort had already lost 8 ± 9% of their body weight, and noeffective tumor treatment was available. During the follow-upperiod, further loss of body fat mass was observed and not leanbody mass, which was maintained or increased slightly. Inter-estingly, the amount of whole body fat was also a significantpredictor of survival.

Halabi et al. (23) investigated the correlation between bodymass index (BMI) and clinical outcome of patients withmetastatic, castration-recurrent prostate cancer. A BMI greaterthan 25 (defined as overweight by the Centers for Disease Con-trol and Prevention) had a protective effect against overall mor-tality and prostate cancer-specific mortality. These patients livedan average 5 mo longer.

Our knowledge and understanding of adipocyte biologyand function has changed dramatically in the last decade.Adipose tissue is no longer considered just an energy store-house but a dynamic endocrine organ that controls a com-plex network of important biological functions (24). These in-clude energy homeostasis, immune function, cell signaling, andreproduction.

We propose that body fat loss occurs early, before loss of leanbody mass and before onset of cachexia is clinically observed.Ideally, we would like to test this hypothesis in humans; how-ever, ethical reasons make this difficult, so we have used an ani-mal tumor-bearing model. We selected the methylcholanthrene(MCA)-induced sarcoma model. The tumor does not metasta-size and has been used in more than 100 published studies in-cluding the original parabiosis studies (25), which demonstratedblood from a tumor-bearing animal promotes weight loss in theparabiosed nontumor-bearing pair.

The precise mechanism behind the loss of body tissue in thetumor-bearing host is still unknown. Our working hypothesis ismalignant growth promotes changes in circulating factors thatalter adipocyte biology prior to the clinical manifestations ofcachexia. A number of hormones (including insulin, glucocorti-coids [corticosterone for the rat], thyroid stimulating hormone,ACTH, norepinephrine and epinephrine), natriuretic peptides,and cytokines are known to affect lipolysis. The circulating gly-coprotein, zinc alpha-2 glycoprotein (ZAG), has lipolytic activ-ity (26,27). The specific aims of this study were twofold: 1) todetermine if fat mass is lost prior to lean body mass and if thatloss occurs prior to the clinical manifestation of cachexia; and 2)to determine if known circulating lipolytic factors are involvedin promoting this early fat loss. This study is unique becausewe followed changes before and after the onset of cachexia.We report here that an unidentified serum factor, most likely aprotein, distinct from zinc alpha-2 glycoprotien, promotes lossof lipid stored in adipocytes.

MATERIALS AND METHODS

Animals and DietsMale Fischer 344 rats, ranging in weight from 240 to 300

g, were purchased from Charles River (Wilmington, MA) andhoused in the vivarium at Pennington Biomedical Research Cen-ter (Baton Rouge, LA) or the University of North Carolina atCharlotte. This protocol was approved by the Institutional An-imal Care and Use Committee at Pennington Biomedical Re-search Center and the University of North Carolina at Charlotte.

All animals were acclimated to individual housing for 7 to10 days prior to the start of any experiment. Then each ratwas weighed and randomly assigned to one of the followingad libitum fed groups: tumor-bearing (TB) or nontumor-bearing(NTB). The remaining animals were weight-matched to a tumor-bearing animal. Weight-matching was important to avoid pairinga tumor-bearing animal with a pair-fed control that weighedmore. This group, called pair fed (PF), received the same amountof food that their matched tumor-bearing pair consumed theprevious 24 h. Pair feeding started on Day 10 since food intakedid not differ between the TB and NTB until after this day.All animals were fed rat chow (Purina; St. Louis, MO). The PFanimals were fed twice a day, around 8 AM and 5 PM, to preventgorging. Food intake and body weight were recorded daily.

On Day 1, the animals were anesthetized with isofluraneand the MCA-induced sarcoma implanted subcutaneously inthe hind flank of the TB animals. The NTB animals, both adlibitum fed and pair-fed, were sham operated. Briefly, the hindflank was shaved and disinfected. A small skin piercing incisionwas made and using blunt surgical technique, a short tunnelmade under the skin. For the TB animals, a 2 mm3 piece of thetumor was placed in the tunnel whereas nothing was placed forthe sham-operated animals. Then the wound was closed withwound clips.

At specific time points after tumor implantation, animalsfrom the 3 groups were sacrificed. On Day 10, animals fromthe TB and NTB groups were sacrificed. On Day 15 and 21,animals from all 3 groups—TB, NTB, and PF—were sacrificed.All food was removed the evening prior to the day of sacrifice.This experiment has been repeated 3 times with similar resultseach time.

Tumor dimensions (length, width, height) were measureddaily by the same investigator starting on Day 10 when the tu-mor was first palpable. From this measurement, the daily cubicvolume was calculated. At the time of sacrifice, the tumor wascarefully excised and weighed. Daily tumor weight was calcu-lated from the ratio of the daily cubic volume to the final cubicvolume as described by Morrison (28).

Analysis of Body CompositionProtein and fat content of Day 15 and Day 21 TB and PF

animal’s carcasses were determined by chemical analysis. The


486 L. O. BYERLEY ET AL.

PF group, and not the NTB group, was included to account fordifferences as a result of changes in food intake. For the TBanimals, the tumor was removed and its fat and protein con-tent determined separately from the host’s carcass. Otherwise,the carcass comprised the whole body of the animal except forapproximately 10 ml of blood and a 1 gram sample of subcuta-neous, omental, and epididymal fat.

A known amount of water was combined with the animal’scarcass and then homogenized. A weighed aliquot of the ho-mogenate was completely dried at 70◦C. Body water contentwas calculated as the difference in weight before and after dry-ing minus the portion of water added. The fat content of thesample was determined using the solvent extraction technique,a Soxhlet apparatus, and ethyl ether as the solvent. The remain-ing defatted sample was finely ground into a highly homoge-nous powder and the carcass protein content determined by themicro-Kjeldahl method (29). The percentage of fat and proteinwas calculated in relation to carcass mass.

Adipose Cell Size and CountAdipocyte size and number were determined for a subcu-

taneous fat sample. Fat cells were fixed with osmium and thencounted on a Beckman Multisizer III Coulter Counter (BeckmanCoulter, Fullerton, CA).

Energy ExpenditureDaily energy expenditure was measured using the Oxymax

system (Columbus Instruments, Columbus, OH). The animalswere acclimated to individual housing and metabolic cage forat least 1 wk prior to implanting the tumor. The Oxymax systemwas kept closed except for the brief time each day the animaland its food were weighed. Pair-feeding started 10 days aftertumor implantation.

Free Fatty Acid TurnoverFree fatty acid turnover was quantified using a primed, con-

tinuous infusion of 1–14C palmitate solubilized with albumin.On Day 17, 10 animals in the TB and NTB group were anes-thetized (pentobarbital, 6 mg/100 g body weight) and catheterssurgically implanted in the jugular vein and the carotid artery. Apriming dose of 1–14C palmitate was slowly administered intothe jugular catheter. This was followed by a 2.5 h continuousinfusion (0.09 µCi 14C palmitate/min). Between 2 and 2.5 hafter the start of the infusion, a blood sample was taken and theserum separated and stored at –80◦C for analysis later.

Total free fatty acids were quantified using a modified methoddescribed by Falholt et al. (30). The amount of 14C labelingpresent in palmitate was quantified after the lipids were ex-tracted from plasma (30) and further separated by thin layerchromatography. The plate was stained with Rhodamine-B dyeand the palmitate region scraped into a liquid scintillation vialfilled with liquid scintillation cocktail. Radioactivity (cpm) wasmeasured by a scintillation counter. The samples were countedin triplicate and then averaged and corrected for counting effi-ciency (dpm).

Host free fatty acid turnover was calculated using the follow-ing equation:

infusion rate (dmp/ min)

specific activity (dpm/µmol)

where the infusion rate was calculated by multiplying the in-fusate activity (dpm/ml) by the rate of infusion (ml/min). Thespecific activity (dpm/µmol) was calculated by dividing thepalmitate activity of the plasma sample (dpm/ml) by the totalplasma free fatty acid concentration (µmol/ml). Turnover wasnormalized to the weight of the animal (kg).

Blood Collection and AnalysesAt specific time points after tumor implantation, animals

from each group were fasted overnight and then anesthetized andblood collected by cardiac puncture and appropriately processedfor the collection of serum or plasma. Blood for quantitation ofcatecholamines (epinephrine and norepinephrine) was collectedfollowing decapitation.

Serum triglycerides were quantified enzymatically (TR0100;Sigma Aldrich, St. Louis, MO). Insulin (sensitive rat insulinRIA, SRI-13K), glucagon (GL-32K), and leptin (RL-83K) werequantified using kits purchased from Linco Research (Milli-pore, Billerica, MA). Adiponectin (44-ADPRT-E01) and ACTH(021-SDX018) were quantified using ELISA kits purchasedfrom ALPCO Diagnostics (Windham, NH). Corticosterone wasquantified using an RIA specific for rats purchased from ICNBiomedicals, Inc. (07–120102, Costa Mesa, CA). The rat cy-tokines, rIL-1β, rIL-6, rIL-10, rGRO/KC and rrTNF-α werequantified using a multiplex kit and Luminex technology (LincoResearch, Millipore, Billerica, MA). Plasma norepinephrineand epinephrine were quantified using the ESA plasma cate-cholamine methodology and ESA Coulochem LC/EC system(45–0161, Chelmsford, MA).

Serum zinc-alpha 2 glycoprotein (ZAG) was quantified byWestern blotting. Antibody raised against a peptide mappingnear the C-terminus of rat ZAG was purchased from SantaCruz Biotechnologies (Santa Cruz, CA; ZAG (G-20):sc-11245).Serum was combined with Pro-Prep solution (Intron Biotech,Inc., Sungnam, Kyungkido), which included protease inhibitors(1 mM phenylmethylsulfonyl fluoride, 0.7 µg/ml pepstatin A,0.5 µg/ml leupeptin, and 2 µg/ml aprotinin) and then cen-trifuged at 13,000 rpm for 10 min to remove the insoluble frac-tion. Total protein concentration for each sample was measuredusing the Micro BCA protein assay kit (Pierce, Rockford, IL). A10 µg protein sample was boiled for 5 min in Laemmli loadingbuffer and then the protein separated by SDS-PAGE on a 10%acrylamide gel. Nitrocellulose membranes were used to elec-troblot the proteins, which were then stained with Ponceau Sto confirm equal protein loading and transfer. Membranes werewashed in TBE with 0.1% vol/vol Tris-Buffered Saline Tween20 (TBS-T), blocked in 5% wt/vol nonfat milk in TBS-T bufferfor 1 h, and then incubated with goat anti-ZAG (1:1000, Santa



Cruz Biotechnology) antibodies at 4◦C overnight. Membraneswere then washed repeatedly in TBS-T and incubated for 1 hat room temperature with a horseradish peroxidase-conjugateddonkey antigoat IgG antibody (Santa Cruz Biotechnology). Im-munoreactive bands were detected using the enhanced chemi-luminescence reagent (GE Healthcare, Piscataway, NJ) and ex-posed to CL-X Posure film (Pierce, Rockford, IL). The quan-tification of band intensity was performed using a densitometricanalysis program (Quantity One, Bio-Rad Laboratories, Her-cules, CA).

Lipolysis AssayFor this assay, 3T3-L1 cells were grown to confluency in

96-well plates with 10% calf-serum in DMEM culture mediasupplemented with L-glutamine, sodium pyruvate, nonessentialamino acids, and penicillin/streptomycin added. The cells werecontinued 2 days beyond confluency and then stimulated to dif-ferentiate with 0.5 mM 1-methyl-3-isobutylxanthine (IBMX), 1µM dexamethasone, and 86 nM insulin for 3 days. Then the cellswere exposed to 86 nM insulin for 3 days in 10% fetal bovineserum in DMEM culture media. Three days later, the media wasswitched to 10% fetal bovine serum in DMEM in which theadipocytes were maintained. We used the cells within 7 days ofreaching maturity. For all experiments, the media was carefullyremoved from the cells, the cells were washed with sterile PBStwice, and then the appropriate mixture was applied.

First, the following groups were tested for their lipolytic po-tential: Day 10 TB, Day 10 NTB, Day 15 TB, Day 15 NTB,Day 15 PF, Day 21 TB, Day 21 NTB, and Day 21 PF. Fifty µlof serum was mixed with 150 µl of DMEM culture media withpenicillin/streptomycin added. Twenty-four hours later, the me-dia was removed and the glycerol content immediately quanti-fied using an enzymatic assay (Sigma, St. Louis, MO or ZenBio,Research Triangle Park, NC). The amount of glycerol releasedwas determined from the difference in glycerol content beforeand after 24 h on the adipocytes.

The lipolytic potential of the lipid and delipidated portionof the serum was tested to determine if the factor was a lipid.Serum lipid was isolated using the method described by Blighand Dryer (31). The serum was delipidated using fumed silica.Each fraction was mixed with DMEM media (40% vol/vol) andthen placed on the 3T3 adipocytes. Twenty-four hours later, themedia was removed and the glycerol released determined aspreviously described. Tumor serum, from which the sampleswere made, was also tested.

A lipolysis time course was determined. Serum from Day 21TB and Day 21 PF animals was pooled and combined with 3T3adipocyte maintenance media devoid of serum. This was placedon 3T3-L1 adipocytes and 2, 4, 6, 8, and 24 h later, a smallsample of the media was removed and analyzed for glycerolcontent. From this, glycerol release was calculated.

Serum (2 ml) from Day 21 TB animals was dialyzedovernight against 50 mM tris buffer, pH 7.5, and then mixedwith 0.5g DEAE cellulose (DE52, Whatman, Florham Park,

NJ) equilibrated to 50 mM tris pH 7.5. The supernatant (labeledsupernatant from DEAE), which contains unbound protein, wasremoved, and protein bound to the DEAE was eluted with anincreasing NaCl solution (50 mM NaCl, 100 mM NaCl, and150 mM NaCl). These 4 fractions (DEAE supernatant plus 3salt fractions) were dialyzed against PBS overnight (4 changes)and concentrated to the original serum volume (2 ml) usingglycerol-free, 5,000 kDa molecular weight, cutoff filters (Milli-pore, Billerica, MA). These samples were mixed with DMEMmedia (40%, vol/vol) and placed on 3T3 adipocytes. After 24 h,the media was removed and the glycerol content quantified en-zymatically. The ZAG content of these samples was determinedas described earlier (see Blood Collection and Analyses). Tu-mor serum, from which the samples were made, was also tested.

Tumor CellsTumor cells were harvested from an MCA-induced sarcoma

at the time of implant. Briefly, the cells were released by treat-ment with collagenase and then sterile filtered and plated. Thecells were maintained in RPMI 1640 media containing fetalbovine serum (10%), penicillin, and streptomycin. At 80% con-fluency, the media was replaced with fresh media. This me-dia was collected 24 h later, combined with the 3T3 media(40:60 vol/vol), and placed on 3T3-L1 adipocytes. Twenty-fourhours later, this media was removed, quickly cooled, and glyc-erol content quantified. Also, we tested the glycerol-releasingcapacity of media not previous placed on the cultured MCA-induced sarcoma cells. Also, we tested the 3T3-L1 adipocytemaintenance media before and 24 h after it had been on the3T3-L1 adipocytes.

Statistical AnalysisLongitudinal observations of body weight, net body weight

(body weight net of tumor burden), food intake, and energymeasures were modeled statistically as repeated measurements.These statistical models included an additional random effectparameter to account for covariance among observations onmatched animals. The fixed effects portion of the models wascomposed of a factorial arrangement of treatment level and studyday.

Analysis of body composition, serum lipolysis, blood con-stituents, and metabolites (Tables 1 and 2) preceded using mixedeffects models with a factorial arrangement of treatment leveland study day upon which the animals were sacrificed as fixedeffects, and included a random effect to account for the effects ofthe initial matching of animals. Statistical significance for com-parisons between treatment groups on a particular study daywas derived from tests of differences between marginal means;in all cases, significance was declared relative to a 5% Type 1error rate, unadjusted for multiple comparisons.

The serum lipid and lipid-free fractions were comparedwith tumor serum using 1-way analysis of variance. The sameanalysis was used to compare the 5 groups generated with DEAEseparation.



TABLE 1Blood concentration of circulating metabolites, lipolytic hormones, and selected adipokines 15 and 21 days after tumor

implantation or sham operation

Day 15 Day 21

Blood Constituent TB PF NTB TB PF NTB

Serum Triglycerides (mg/dl) 38.9 ± 3.1a 27.1 ± 2.7b 18.6 ± 3.2c 99.4 ± 7.7c 25.7 ± 3.5b 26.2 ± 4.3b

Norepinephrine (ng/ml) 11 ± 3 15 ± 5 15 ± 5 14 ± 3 21 ± 3 15 ± 5Epinephrine (ng/ml) 11 ± 2 19 ± 8 21 ± 6 22 ± 9 40 ± 6 28 ± 9Corticosterone (ng/ml) 711 ± 40 870 ± 50 919 ± 55 791 ± 87a 869 ± 65a,b 1079 ± 47b

ACTH(pg/ml) 128 ± 13a 206 ± 38b 208 ± 23b 68 ± 10a 268 ± 38b 244 ± 34b

Adiponectin (ng/ml) 4.6 ± 0.2a 6.2 ± 0.3b 6.2 ± 0.3b 2.8 ± 0.4a 8.8 ± 0.4b 7.2 ± 0.4b

Leptin (ng/ml) 2.5 ± 0.3 2.7 ± 0.4 2.9 ± 0.3 1.5 ± 0.2a 3.7 ± 0.4b 2.6 ± 0.3b

Insulin (ng/ml) 0.56 ± 0.04a 0.71 ± 0.07b 0.79 ± 0.03b 0.38 ± 0.08a 0.81 ± 0.03b 0.75 ± 0.08b

Glucagon (pg/ml) 131 ± 9 141 ± 13 132 ± 11 144 ± 13 118 ± 18 144 ± 8Insulin to Glucagon ratio 4.4 ± 0.3a,b,d 5.4 ± 0.6a 6.3 ± 0.5a,c,d 2.8 ± 0.2b 7.7 ± 1.3c 5.3 ± 0.3d

ZAG (arbitrary units/µgprotein)

72 ± 13 72 ± 12 63 ± 16 72 ± 13 75 ± 14 104 ± 12

Note. Abbreviations are as follows: TB, tumor bearing; PF, pair fed; NTB, nontumor bearing; ACTH, adrenocorticotropic hormone; ZAG,zinc alpha-2 glycoprotein.Values are means ± SE. Within each day, groups with different subscripts are significantly different, P < 0.01. n =8/group.

All data are presented in the tables and figures as mean± SE.

RESULTS

Changes in Body Weight and Food Intake With TumorGrowth

Changes in body weight and food intake as the tumor greware shown in Fig. 1. Throughout the entire 24-day period bodyweight measurements were collected, the total body weight ofthe TB animals did not differ significantly from the NTB ani-mals. However, the host body weight of the TB animals, that

is, the weight of the TB animal minus the weight of the tumor,was significantly less than the NTB and PF animals 15 daysafter tumor implantation (Day 15). This difference continuedand became significantly greater with each subsequent day.

Food intake for the TB and NTB animals is shown in bottomgraph of Fig. 1. Food intake did not differ significantly amongthe groups until Day 18 when the TB animals consumed sig-nificantly less food than the NTB. This difference was furtheraccentuated as tumor size increased.

The growth of the tumor is shown in the middle panel ofFig. 1. The tumor is less than 2.5% of body weight on Day 15when the first significant host weight loss in the TB animals is

TABLE 2Blood concentration of cytokines 15 and 21 days after tumor implantation or sham operationa

Day 15 Day 21

Cytokine NTB TB PF TB PF

IL-1β(pg/ml) 18 ± 12(5/12) 39 ± 33(3/5) 12 ± 9(3/5) 14 ± 10(2/5) 0 (0/5)rGRO/KC(pg/ml) 76 ± 3(12/12) 131 ± 66(5/5) 114 ± 45(5/5) 115 ± 49(5/5) 40 ± 15(5/5)rIL-10(pg/ml) 19.5 ± 16.9(2/12) 1.6 ± 1.0(3/5) 1.9 ± 1.6(3/5) 0 (0/5) 0 (0/5)rIL-6(pg/ml) 18 ± 21(2/12) 30 ± 27(4/5) 30 ± 49(2/5) 0 (0/5) 4 ± 10(1/5)rTNF α(pg/ml) 18 ± 21(2/12) 23 ± 11(4/5) 11 ± 5(5/5) 0 (0/5) 6 ± 7(2/5)

aAbbreviations are as follows: NTB NTB, nontumor bearing; TB, tumor bearing; PF, pair fed; IL, interleukin; TNF, tumor necrosis factor.Values are mean ± SE. IL-1β, rGRO/KC, rIL-10, rIL-6, and rTNFα were quantitated using Luminex technology. For IL-1β, rGRO/KC,rIL-10, rIL-6, and rTNFα, the first number in parenthesis is the number of samples that were quantifiable; the second number is the numberof samples measured. To calculate the mean and SD, a non detectable value was assigned zero. The day 15 and 21 NTB values were averagedtogether.



FIG. 1. Changes in body weight and food intake with tumor growth. The topgraph shows the changes in body weight among the 3 groups (TB, PF, andNTB) and the tumor bearing animal’s weight minus the weight of the tumor.The middle graph shows the rate of tumor growth. The bottom graph shows foodintake for the TB and NTB groups. The PF are not included since their foodintake was the same as the TB except delayed by 24 h. Each arrow indicates theday when body weight or food intake was first significantly different (P<0.05).Each subsequent day to the right of the arrows was significantly different amongthe groups. We have repeated this experiment 3 times with similar results eachtime. Values are means ± SE (n = 8 per group).

observed. The observed 2% weight loss is slightly less than thetumor burden. There was no difference in water intake amongthe three groups at any time point measured (data not shown).

Three days passed between the start of the decline in thehost’s body weight and a drop in food intake. Clearly, the de-crease in body weight occurs first and is not the result of adecrease in food intake. The inclusion of the PF group to con-trol for food intake clearly demonstrates the portion of weightloss due to a reduction of food intake or due to a growing tumor(Fig. 1). The TB animals lost more weight than can be accountedfor solely from a reduction in food intake alone.

Changes in Body CompositionComparisons of the fat content of the TB and PF animals are

shown in Fig. 2. The fat content of the TB animals comparedto the NTB and PF animals was significantly less on Day 15

FIG. 2. Changes in body lipid content with tumor growth. A: Lipid content ofthe TB, NTB, and PF host’s body on day 15 and 21. NTB and PF were pooledbecause the lipid content was not significantly different between day 15 andday 21. Values are means ± SE (Day 15, n = 13 rats/group; Day 21, n = 11rats/group). B: Change in the total lipid content of the tumor bearing animal’sbody as tumor size increased. C: Change in subcutaneous adipocyte size. Solidsquare, day 10 tumor-bearing (n = 8); solid circle, day 15 tumor-bearing (n =13); solid triangle, day 21 tumor-bearing (n = 11).

when the first difference in TB host body weight was observed(Fig. 2A). This difference remained at Day 21. No significantdifference in protein content was observed on Day 15 or Day21. By Day 28, the protein content of the TB animals was less(Day 15: NTB 57 ± 3 g, PF 49 ± 4 g, TB 50 ± 4 g; Day 21:NTB 46 ± 4 g, PF 44 ± 3 g, TB 41 ± 2 g; Day 28: NTB 48 ±4 g, PF 44 ± 4 g, NTB 35 ± 6 g).

The lipid and protein content of the tumor did not changefrom Day 15 to Day 21 (not shown), so this does not account



for the difference in fat content of carcasses observed in the TBanimals.

There was a negative correlation between host lipid content(Fig. 2B) and tumor size; the larger the tumor, the less host lipid.This was not true for host protein content (data not shown; y= –0.09x ± 49.9, P < 0.03, R2 = 0.09), which was lost muchmore slowly. The rate of lipid loss (–0.25 g carcass lipid/g tumorweight) was 2.5 times more rapid than the rate of protein loss(–0.09 g carcass protein/g tumor weight).

The loss of lipid by the TB animals was the result of theadipocytes shrinking in size (Fig. 2C). The total number ofadipocytes (3,400 ± 1,200 cells/mg tissue) did not change be-tween Days 15 and 21. Instead, the size of the adipocytes shifteddramatically by Day 21 from a size of 50 to 90 um to <10 µm.

Energy ExpenditureFigure 3 reports the energy expenditure (kcal/h) during the

dark and light cycle for the 3 groups. The energy expended dailydid not differ significantly among the groups regardless of thetime of day.

The respiratory exchange ratio (RER) is reported as a cumu-lative frequency for the TB animals on the lower graph in Fig.3. As the tumor grew and fat mass was lost, more fat was usedas a fuel. This is demonstrated by a shift in RER from 0.85 onDay 10 to <0.8 on Day 20.

Blood Metabolites and ConstituentsWe further verified the early changes in total body lipid con-

tent by measuring free fatty acid turnover on Day 17. Free fattyacid turnover was significantly higher in the TB animals (10.9 ±0.9 mmol/day, n = 10, P < 0.05) compared to the NTB animals(9.3 ± 0.3 mmol/day, n = 10). This increased rate correlatedwith a significant elevation in serum triglycerides for the TBanimals on Day 15 (Table 1). This trend continued, and by Day21, serum triglyceride levels were more than fourfold higherthan the PF and NTB groups.

A number of hormones and cytokines have been identifiedthat promote lipolysis. We examined selected representatives todetermine if any may be involved in promoting the body fat losswe observed. No significant differences in serum levels of nore-pinephrine, epinephrine, and corticosterone were found early(Day 15) when the first decline in body weight was observed(Table 1). ACTH was significantly lower in the TB animals onDay 15 and not elevated. Adiponectin was significantly loweron Day 15, but a decrease in leptin was not observed until Day21.

Although glucagon levels did not differ among the groups,insulin levels were significantly lower in the TB on Day 15and 21, making the glucagon to insulin ratio lower. However,this trend was not significant on Day 15 when the first changein body fat mass was observed and prior to cachexia. Sinceinsulin levels decreased as cachexia progressed, we calculatedthe insulin to glucagon ratio. This ratio changed significantlywhen the animals were cachectic but not before.

FIG. 3. Energy expenditure and respiratory exchange ratio (RER). Graph Aand B are energy expenditure during the dark and light cycle for the TB (darkcircle), PF (open circle) and NTB (dark triangle). The bottom graph (C) showsthe cumulative RER frequency for the TB animals. The solid squares are Day10; solid triangles, Day 15; solid diamonds, Day 18; and solid hexagonal, Day20.There are 5 animals per group and the experiment has been repeated 3 timeswith similar results. Values are means ± SE.

ZAG has been shown to have lipolytic activity, so we deter-mined the levels of this protein at different time points. We foundno difference in the amount of ZAG present in the sera amongthe 3 groups on Day 15 and Day 21 (Table 1). In addition toZAG, we measured the serum concentrations for 5 different



cytokines (Table 2). There were no significant differencesamong the groups on any day measured.

Serum-Induced LipolysisSera from Days 10, 15, and 21 for TB, NTB, and PF animals

was added to 3T3 adipocytes to quantify the lipolytic potential asmeasured by the release of glycerol (Fig. 4). Increased glycerolrelease was observed on Days 15 and 21 for the TB animalscompared to the PF and NTB. Two lines of evidence suggestthis release was not due to a reduction in food intake. First, onDay 15, when food intake had not yet decreased, the TB serumstimulated significantly more glycerol release than sera from PFand NTB animals. Second, the amount of glycerol released onDay 21 was almost twofold greater for the TB group comparedto the PF group. Our time course study indicates it takes between8 and 24 h for the TB serum to stimulate glycerol release (Fig.4). Also, we tested the lipolytic potential of media that had beenon cultured MCA-induced sarcoma cells for 24 h. The resultsof this experiment are shown in Fig. 4. The media exposed tothe MCA-induced sarcoma cells stimulated a greater release ofglycerol than naı̈ve media that had not been exposed to anycells and media that had been exposed to 3T3-L1 adipocytes for24 h.

Unknown Lipolytic Factor or FactorsFirst, we determined if the factor or factors responsible for

stimulating lipolysis was a lipid or was present in the lipid-freefraction of the serum. As shown in Fig. 5, delipidated serumstimulated a greater release of glycerol than the lipid fraction.The delipidated fraction had 71% of the activity of the tumorserum. Then we determined the stability properties of the lipoly-tic activity in the lipid-free serum fraction. This fraction washeated at 90◦C for 10 min, insoluble material was removed bycentrifugation, and the supernatant was tested in the lipolysisassay. The amount of glycerol released was sixfold less than theamount of glycerol released by the serum and the delipidatedfraction and similar to that observed from the lipid portion ofthe serum.

Because ZAG has lipolytic activity (27), we measured theZAG content of our active fractions to determine whether theobserved lipolysis could be attributed to ZAG (Fig. 5). The frac-tion with the greatest lipolytic activity contained no measurableZAG; thus, we conclude that the lipolytic activity observed inthe serum of tumor-bearing animals was not due to the presenceof ZAG.

DISCUSSIONUsing the MCA-induced sarcoma model, we investigated

changes before and after the development of cachexia as thetumor grew. We report that a loss of body fat occurs early, beforeloss of lean tissue mass, and prior to a reduction in food intake.Our interest is in identifying the promoters of this fat loss. In thismodel, there were 3 days between the reduction in fat mass andthe decrease in food intake, suggesting something other than the

FIG. 4. A: Serum lipolytic potential. The amount of glycerol released from3T3-L1 adipocytes was quantitated following 24-h incubation with media con-taining serum from TB, NTB, and PF animals. Animals were fasted overnight.Within each day, TB are the solid bar, PF are the open bar, and NTB are the crosshatch bar. ∗ P<0.05 TB vs. PF or NTB. Values are mean + SE (n = 8/group). B:Time course. Glycerol release was quantitated 2, 4, 6, 8, and 24 h after pooledserum from day 21 TB and day 21 PF animals was placed on 3T3 adipocytes.The amount of glycerol released was not significantly different until the 24-h time point. C: Lypolytic potential of MCA-induced sarcoma media. Media(RPMI-1640) used to maintain cultured MCA-induced sarcoma cells for 24 hwas tested against naı̈ve media (DMEM media control and RPMI-1640 mediacontrol) and media used to maintain 3T3-L1 adipocytes for 24 h (DMEM). Themedia from the cultured MCA-induced sarcoma cells stimulated a significantlygreater release of glycerol than any of the other 3 treatments.



FIG. 5. Serum lipolytic factor. A: Lipid isolated from the serum, serum withthe lipid removed (delipidated), and heat inactivated serum were tested forlipolytic activity. All 3 groups are significantly different (P< 0.001) from tumorserum. There was no significant difference between the lipid and heat inactivatedserum fractions. B: DEAE fractions were prepared as described in the Materialsand Methods section. C: The DEAE fractions shown in B were tested for thepresence of the ZAG protein using the Western blot technique. Values are means± SE. Experiments were repeated twice with the same result.

reduction in food promotes the loss of body fat. We hypothesizethat the promoters of cancer cachexia are present well beforecachexia is observed clinically.

The reduction in body fat prior to loss of lean mass observedin our model agrees with the most recent human study publishedby Fouladiun and associates (22). They found fat mass was lostfirst and more rapidly than lean mass. In our model, as tumorsize increased, more lipid was lost from the host’s carcass thanprotein, adipocyte size decreased, and more lipid was present inthe blood. Lean mass was lost more slowly.

During the last several decades, a number of theories havebeen proposed to explain cancer cachexia. These have included1) increased hypermetabolism, 2) decreased food intake, or 3)

circulating lipid mobilizing factors. We have examined each ofthese to determine if they played a role in the increased fat massloss we observed in this model.

Hypermetabolism is often used to refer to a resting metabolicrate greater than predicted for healthy individuals. This has beenstudied using a variety of different methods and approaches, andsome studies have reported hypermetabolism in some cachecticcancer patients (32,33). However, most studies have reportedhypo or normal metabolism for cachectic cancer patients (34).In our model, energy expenditure was not different among the3 groups prior to or after the development of cachexia throughDay 21. We report energy expenditure as kcal/h. Changes inactive cell mass can mask changes in energy expenditure sinceenergy expenditure is determined by lean mass. We did not finda difference in lean body mass in the TB animals from Day 15to 21 when cachexia developed. The loss in body weight wasprimarily fat mass. So we did not find daily energy expenditurechanged or explained the early decrease in body fat mass.

Decreased food intake does promote weight loss and couldcontribute to cachexia. In our model, we had a 3-day periodbetween the first changes in body weight and body compositionand a decrease in food intake. Certainly a decrease in food intakepromotes some of the weight loss reported from Day 18 to 21.The PF group controls for this and shows the amount of weightloss that is expected for the decreased amount of food the TBanimals ate. So the early changes in body weight and fat masscan not be explained by a decrease in food intake alone.

Since changes in energy expenditure and a decrease in foodintake did not adequately explain the mobilization of fat and leantissue in animal tumor models, this led to the conclusion thata circulating lipid-mobilizing factor or factors was responsiblefor this effect. This is supported by a previously published para-biosis study (25) in the same tumor model we used and manyin vivo and in vitro models including the 3T3 adipocyte model.Since the parabiosis study (25), there has been a long-standinginterest in identifying factors that promote lipolysis in cachec-tic cancer patients, and a number have been identified includ-ing toxohormone-L (35), azaftig (36), and zinc-α2-glycoprotein(37).

Tisdale and colleagues (26,27) isolated a lipid mobilizingfactor from the urine of cachectic cancer patients and identi-fied it as ZAG. This is a ubiquitously expressed protein (38,39),and mRNA expression is upregulated in the adipose tissue ofcachectic mice (40). ZAG upregulates G protein expression,which stimulates production of cAMP and the release of glyc-erol (41). Glucocorticoids stimulate lipolysis by increasing ZAG(40).

In the MCA model, increased lipolysis does not appear to bedependent on ZAG. First, serum levels of ZAG are not increased;and second, ZAG-depleted serum still promotes lipolysis. Re-cently, two studies have been published suggesting the role ofZAG is not simply that of a lipid mobilizing factor. First, an im-munohistochemical analysis of murine xenografts did not findZAG involved in cachexia (42). Second, Rolli and coworkers



(43) published the phenotype for mice with both ZAG alle-les inactivated. These mice had increased body weight, mostlikely the result of reduced lipolysis. This does suggest ZAGis involved in lipid mobilization; however, the reduced lipoly-sis did not occur via the adrenoreceptor pathway as previouslyreported (44), nor were differences between +/+ and –/– miceobserved for a number of lipid metabolism molecules, includ-ing hormone sensitive lipase, lipoprotein lipase, or fatty acidsynthase.

Other hormones and cytokines circulate and are known topromote lipolysis including catecholamines, cortisol, natriureticpeptide, and insulin. We did not measure natriuretic peptides be-cause the lipolytic effect of these peptides appears confined toprimates (45). We do report levels for catecholamines and cor-tisol and did not find these adequately account for the lipolysiswe observed.

Insulin provides a block to lipolysis. Reducing circulatinglevels of insulin can promote lipolysis, and administration ofinsulin has been suggested as an anabolic treatment for cancercachexia. Londos et al. (46) showed that decreasing concentra-tions of insulin increases lipolysis, and this effect occurs withinminutes. We show the serum must be present for more than 8 hfor lipolysis to occur, suggesting insulin is not the primary factoror only factor involved in our model. The long time for induc-tion of the effect suggests that the mechanism of the factor(s)described here may require gene transcription.

A number of cytokines have been associated with cancercachexia, including TNF-α, interleukin-1 (IL-1), interleukin 6(IL-6), interferon-gamma (INF-γ ), and ciliary neurotrophic fac-tor (CNTF). Administration of these cytokines to animals aloneor in combination leads to anorexia, weight loss, protein and fatbreakdown, increased cortisol and glucagon, decreased insulin,and elevated energy expenditure. However, in weight losing can-cer patients, blood levels of these cytokines are rarely elevated.We found the same in our model, suggesting these particularcytokines do not circulate and promote lipid mobilization. Theymay act locally within adipose tissue.

In summary, as is observed in human cancer patients, fat massis lost first and more rapidly than lean tissue in the MCA-inducedtumor model. The adipocytes shrink, and blood levels of triglyc-eride increase. We investigated the known lipolytic hormonesand cytokines to determine if these are involved in promotingthe loss of lipid from the adipocyte and the early developmentof cachexia. None of the circulating factors we investigatedpromoted this loss, including the most recently identified lipidmobilizing factor, ZAG. We conclude that a previously uniden-tified factor, most likely a protein, promotes this loss of lipidfrom the adipocyte. These studies will form the basis for furtherresearch to isolate and characterize this lipolytic protein.

ACKNOWLEDGMENTSWe thank Sadie Hebert, Jun Zhou, Judy Wiles, Suzanne

Trunick, and J. Mark Morris for their scientific contributions

to this research. We graciously thank Laura Hale for supply-ing her mouse zinc alpha-2 glycoprotein antibodies and LuisSanchez for supplying a human zinc alpha-2 glycoprotein stan-dard to test ZAG antibodies as well as his expert comments andsuggestions. This research was generously supported by a grantfrom the National Cancer Institute (R29CA060137), Ameri-can Institute for Cancer Research, and institutional grants fromPennington Biomedical Research Center and the University ofNorth Carolina at Charlotte.

REFERENCES1. DeWys WD: Anorexia in cancer patients. Cancer Res 37: 2354–2358,

1977.2. DeWys D, Begg C, Lavin PT, Band PR, Bennett JM, et al.: Prognostic

effect of weight loss prior to chemotherapy in cancer patients. Am J Med69, 491–499, 1980.

3. Kulstad R and Schoeller DA: The energetics of wasting diseases. Curr OpinClin Nutr Metab Care 10, 488–493, 2007.

4. Thomas DR: Loss of skeletal muscle mass in aging: examining the rela-tionship of starvation, sarcopenia and cachexia. Clin Nutr 26, 389–399,2007.

5. Maturo G, Vespasiani G, Mohamed EI, Maiolo C, Finazzi AE, et al.: Eval-uating body composition of Italian prostate cancer patients without metas-tases. Acta Diabetol 40(1 Suppl.), S168–S170, 2003.

6. Segal RJ, Reid RD, Courneya KS, Malone SC, Parliament MB, et al.: Resis-tance exercise in men receiving androgen deprivation therapy for prostatecancer. J Clin Oncol 21, 1653–1659, 2003.

7. Sarhill N, Mahmoud F, Walsh D, Nelson KA, Komurcu S, et al.: Evaluationof nutritional status in advanced metastatic cancer. Support Care Cancer11, 652–659, 2003.

8. Hyltander A, Drott C, Unsgaard B, Tolli J, Korner U, et al.: The effect onbody composition and exercise performance of home parenteral nutritionwhen given as adjunct to chemotherapy of testicular carcinoma. Eur J ClinInvest 21, 413–420, 1991.

9. McMIllan DC, Preston T, Watson WS, Simpson JM, Fearon KC, et al.:Relationship between weight loss, reduction of body cell mass and inflam-matory response in patients with cancer. Br J Surg 81, 1011–1014, 1994.

10. Moley JF, Aamodt R, Rumble W, Kaye W, and Norton JA: Body cell massin cancer-bearing and anorexic patients. JPEN J Parenter Enteral Nutr 11,219–222, 1987.

11. Burke M, Bryson EI, and Kark AE: Dietary intakes, resting metabolic rates,and body composition in benign and malignant gastrointestinal disease. BrMed J 280, 211–215, 1980.

12. Hansen RD, Raja C, and Allen BJ: Total body protein in chronic diseasesand in aging. Ann NY Acad Sci 904, 345–352, 2000.

13. MacFie J and Burkinshaw L: Body composition in malignant disease.Metabolism 36, 290–294, 1987.

14. Smith MR, Finkelstein JS, McGovern FJ, Zietman AL, Fallon MA, et al.:Changes in body composition during androgen deprivation therapy forprostate cancer. J Clin Endocrinol Metab 87, 599–603, 2002.

15. Simons JP, Schols AM, Westerterp KR, ten Velde GP, and WoutersEF. The use of bioelectrical impedance analysis to predict total bodywater in patients with cancer cachexia. Am J Clin Nutr 61, 741–745,1995.

16. Fredrix EW, Saris WH, Soeters PB, Wouters EF, Kester AD, et al.: Esti-mation of body composition by bioelectrical impedance in cancer patients.Eur J Clin Nutr 44, 749–752, 1990.

17. Koch J: The role of body composition measurements in wasting syndromes.Semin Oncol 25, 12–19, 1998.

18. Heymsfield SB and McManus CB: Tissue components of weight loss incancer patients. A new method of study and preliminary observations.Cancer 55, 238–249, 1985.



19. Cohn SH, Ellis KJ, Vartsky D, Sawitsky A, Gartenhaus W, et al.: Com-parison of methods of estimating body fat in normal subjects and cancerpatients. Am J Clin Nutr 34, 2839–2847, 1981.

20. Ali PA, al Ghorabie FH, Evans CJ, el Sharkawi AM, and Hancock DA: Bodycomposition measurements using DXA and other techniques in tamoxifen-treated patients. Appl Radiat Isot 49, 643–645, 1998.

21. Heymsfield SB, McManus C, Stevens V, and Smith J: Muscle mass: reliableindicator of protein-energy malnutrition severity and outcome. Am J ClinNutr 35, 1192–1199, 1982.

22. Fouladiun M, Korner U, Bosaeus I, Daneryd P, Hyltander A et al. : Bodycomposition and time course changes in regional distribution of fat andlean tissue in unselected cancer patients on palliative care—correlationswith food intake, metabolism, exercise capacity, and hormones. Cancer103, 2189–2198, 2005.

23. Halabi S, Ou SS, Vogelzang NJ, and Small EJ: Inverse correlation betweenbody mass index and clinical outcomes in men with advanced castration-recurrent prostate cancer. Cancer 110, 1478–1484, 2007.

24. Mora S and Pessin JE: An adipocentric view of signaling and intracellulartrafficking. Diabetes Metab Res Rev 18, 345–356, 2002.

25. Norton JA, Moley JF, Green MV, Carson RE, and Morrison SD: Para-biotic transfer of cancer anorexia/cachexia in male rats. Cancer Res 45,5547–5552, 1985.

26. Hirai K, Hussey HJ, Barber MD, Price SA, and Tisdale MJ: Biologicalevaluation of a lipid-mobilizing factor isolated from the urine of cancerpatients. Cancer Res 58, 2359–2365, 1998.

27. Todorov PT, McDevitt TM, Meyer DJ, Ueyama H, Ohkubo I et al.: Purifi-cation and characterization of a tumor-lipid-mobilizing factor. Cancer Res58, 2353–2358, 1998.

28. Morrison SD: In vivo estimation of size of experimental tumors. JNCI 71,407–408, 1983.

29. Albanese AA and Orto LA: Protein and amino acids. In: Newer Methods ofNutritional Biochemistry. New York: Academic, 1963, p. 84.

30. Falholt K, Lund B, and Falholt W: An easy colorimetric micromethod forroutine determination of free fatty acids in plasma. Clin Chim Acta 46,105–111, 1973.

31. Bligh EG and Dyer WJ: A rapid method of total lipid extraction and purifi-cation. Can J Biochem Physiol 37, 911–917, 1959.

32. Bosaeus I, Daneryd P, and Lundholm K: Dietary intake, resting energyexpenditure, weight loss, and survival in cancer patients. J Nutr 132,3465S–3466S, 2002.

33. Jatoi A, Daly BD, Hughes VA, Dallal GE, Kehayias J, et al.: Dopatients with nonmetastatic non-small cell lung cancer demonstrate

altered resting energy expenditure? Ann Thorac Surg 72, 348–351,2001.

34. Knox LS, Crosby LO, Feurer ID, Buzby GP, Miller CL, et al.: Energyexpenditure in malnourished cancer patients. Ann Surg 197, 152–162,1983.

35. Busquets S, Alvarez B, van Royen M, Carbo N, Lopez-Soriano FJ, et al.:Lack of effect of the cytokine supressive agent FR 167653 on tumor growthand cachexia in rats bearing the Yoshida AH-130 ascites hepatoma. CancerLett 157, 99–103, 2000.

36. Figueroa J, Vijayagopal P, Debata C, Prasad A, and Prasad C: Azaftig,a urinary proteoglycan from a cachectic cancer patient, causes profoundweight loss in mice. Life Sci 64, 1339–1347, 1999.

37. Lorite MJ, Cariuk P, and Tisdale MJ: Induction of muscle protein degreda-tion by a tumour factor. Br J Cancer 76, 1035–1040, 1997.

38. Tada T, Ohkubo I, Niwa M, Sasaki M, Tateyama H, et al.: Immunohisto-chemical localization of Zn-alpha2-glycoprotein in normal human tissues.J Histochem Cytochem 39, 1221–1226, 1991.

39. Rolli V, Radosavljevic M, Astier V, Macquin C, Castan-Laurell I, et al.:Lipolysis is altered in MHC class I zinc-alpha(2)-glycoprotein deficientmice. FEBS Lett 581, 394–400, 2007.

40. Russell ST and Tisdale MJ: The role of glucocorticoids in the induction ofzinc-alpha2-glycoprotein expression in adipose tissue in cancer cachexia.Br J Cancer 92, 876–881, 2005.

41. Islam-Ali B, Khan S, Price SA, and Tisdale MJ: Modulation of adipocyteG-protein expression in cancer cachexia by a lipid-mobilizing factor (LMF).Br J Cancer 85, 758–763, 2001.

42. Kamoshida S, Watanabe K, Suzuki M, Mizutani Y, Sakamoto K, et al.:Expression of cancer cachexia-related factors in human cancer xenografts:an immunohistochemical analysis. Biomed Res 27, 275–281, 2006.

43. Rolli V, Radosavljevic M, Astier V, Macquin C, Castan-Laurell I, et al.:Lipolysis is altered in MHC class I zinc-alpha(2)-glycoprotein deficientmice. FEBS Lett 581, 394–400, 2007.

44. Russell ST, Hirai K, and Tisdale MJ: Role of beta3-adrenergic receptors inthe action of a tumour lipid mobilizing factor. Br J Cancer 86, 424–428,2002.

45. Sengenes C, Zakaroff-Girard A, Moulin A, Berlan M, Bouloumie A, et al.:Natriuretic peptide-dependent lipolysis in fat cells is a primate specificity.Am J Physiol Regul Integr Comp Physiol 283, R257–R265, 2002.

46. Londos C, Honnor RC, and Dhillon GS: cAMP-dependent protein kinaseand lipolysis in rat adipocytes. III. Multiple modes of insulin regulation oflipolysis and regulation of insulin responses by adenylate cyclase regulators.J Biol Chem 260, 15139–15145, 1985.


nutrition and cancer evidence for a novel serum factor ......body fat before losing protein mass and...

Documents