Evaluation of elevated dietary corn fiber from corn germ mealin growing female pigs'

T. E. Weber, 2 S. L. Trabue, C. J. Ziemer, and B. J. Kerr

USDA. AR.S. Ames. IA 50011-3310

ABSTRACT: To evaluate the effects of dietary hemi-cellulose from corn on growth and metabolic measures,female pigs (ii = 48: initial BW 30.8 kg) were fed di-ets containing 0 to 38.6% solvent-extracted corn germInca! for 28 d. Increasing the hemicellulose level had noimpact on ADG or ADFT, but resulted in a quadraticresponse (P < 0.03) oil To investigate physiologi-cal changes that occur with increased dietary heniiccl-lulose, blood, colon contents, and tissue samples fromthe liver and intestine were obtained from a subset (n= 16: 8 pigs/treatment) of pigs fed the least and great-est hermucellulose levels. The abundance of phospho-adenosiire monophosphate-activated protein kinase(AMPK) and the mitociionclrial respiratory protein.cytochrorne C oxidase II (COXTI) were determined inliver, jejunum, ileum, and colon by Western blotting.The HIRNA expression levels of AMPKct1, AMPKo2.PPAR. coactivator 1& (POd-a), PPAR.2, and sirtuin1 (Sirtl) were determined in liver arid intestinal tis-sues. When compared with pigs fed the control diet.pigs fed the high hemiceilulose diet had increased (F< 0.02) plasma triglycerides but there was no differ-

ence in plasma cholesterol, glucose, or insulin. Absoluteand relative liver weights were decreased (P < 0.03) inpigs consuming the Ingli liemicellulose diet. The high-fiber diet led to a tendency (P < 0.12) for decreasedliver triglyceride content. In pigs fed the high liemicel-lulose diet, ile.al inucosal alkaline phosphatase activitywas increased (P < 0.08) and sucrase activit y tended(P < 0.12) to he increased. The high henricellulose diethad no effect oil AMPK mRNA, orcolonic VFA. but in pigs consuming the bight fiber dietthere was a greater (P < 0.03) abundance of COXIIill tissue. The expression of PGC1-a, PPARr'. orSirti mRNA was not altered by dietary fiber illjejunum. or ileumtissue. In colon tissue from pigs fedthe high fiber diet there was an increase (P < 0.09) inSirti niRNA and a trend (P < 0.12) toward increasedof PGC1-a mRNA. These data suggest that alterationsin metabolism involved ill adaptation to a diet hUll inhcmicellulose are associated with increased colonic SirtimRNA and COXII expression, indicating an increasedpropensity for oxidative metabolism by the intestine.

Key words: corn germ. dietary fiber, energy metabolism heinicellulose, Pi-

02010 American Society of Animal Science. All riglif. ,esf-'rved. Anim. Sci. 2010. 88:192 201doi:10.2327/Jas.2009-1896

INTRODUCTION

Due to the recent growth of the corn ethanol industry,the availability of coproducts, which tend to he high inhemicellulose, is likely to increase. To date, there arefew data available oil amount of fiber from cornthat can he tolerated by growing pigs without nega-tively affecting growth performance. Increasing dietaryfiber content decreases dietary lipid (Graham et al.,

i\ lent ion of a trade lialile. proprietary product. or specific eqiup-nient does not constitute a guarantee or warranty by the USDA anddoes not imply approval to the exclusion of other products that maybe suitable. The authors thank Jennifer Cook. Kerrie Frauzen. ShariSteadhiaiti. and Kellie Winter of the USDA-ABS for assistance withtissue sample collection and laboratory analyses.

2 Corresponding author: toin.weherttars.iis(la.govReceived February 16. 2069.Accepted September 21. 2009.

1986) and energy (Owusu-Asiedu et al., 2006: Wilfartet al., 2007) absorption. Intestinal mass (Pond et nil..1988: Anugwa et al., 1989) and inucosal cmix Inc andprotein synthesis (Piel et al.. 2005; Heclemanu et al..2006) are increased Lv high fiber diets. These changesin the intestinal tract coupled with decreased dietaryenergy absorption likely lead to an increased energydemand by the intestinal tract and repartitioning ofenergy metabolites. There is little information as tothe metabolic adaptations or mechanisms regulatingthe adaptations that occur when pigs are fed diets con-taining high levels of hemicellulose from corn. It wasrecently observed that feeding fructo-ohigosacchariclesincreases colonic mitochondrial gene expression (Roden-burg et al.. 2008). suggesting an increased capacity foroxidative metabolism. The effects of other fiber sourceson mitochoncirial genes have not been reported. Theadenosine mouopiuosphate-activated protein kinase

192

\aI,,e

GE, kcal/ha 42.19cc', 2 1.1)7AA, Sf

Ala 1.26Arg 1.49Asp 1.51)Gin 0.33(3ly 2.87His 1.17He 0.61Len 0.75Lys 1.7))Met 1.04Phe (1.37Pro 0.91Ser 1.1)7Thr 0.89Tip 1)78lyr 11.18Vat

0.63Crude fat, Sf

2.12Ash, Sf

2.41Ca, Si

(1.03P, %

5.79Starch. fl/,

11.20Crude blaT. Si

9.53Total dietary fiber, Si

42.57

NDF. Si

.5441ADF, Sf

11.13Henucellulose, V

43.28Cellulose, 5<

1(1.44Lignin. 5<

1.149

Anal ysis by Universit y of Missouri C'o!i,iubi, Expiruaeiit St ,it jellChemical Laboratories.

(AMPK), a sensor of cellular energy status (Hardieet al., 2006), may mediate metabolic adaptations to di-etary fiber, but the effects of fiber on AMPK activationhave not been studied. The objectives of the study wereto determine the effect of increasing dietary corn fiberon feed intake and growth of growing pigs. The effectsof (lietmy heinicellulose on liver and plasma energy me-tabolites and intestinal enz yme activity were evaluated.To investigate possible meclianisnis involved in adapt

-ing to a, high fiber diet, the expression of genes involvedin initochoncirial energy metabolism and AMPK acti-vation in intestinal and liver tissue were determined.

MATERIALS AND METHODS

All animal care and handling procedures used in thisstudy were reviewed and approved by the Iowa StateUi iiversitv Animal Care and Use Committee.

Animals, Experimental Design,a71.d Sample Collection

Crossbred female pigs (n = 48: C22 x 337. Pig un-Provemnent Company. Franklin, KY; initial BW 30.8 ±

( 'on 1 fi )er ill at gro\ving i"g 193Table 1. Composition of solvent-extracted corn geniimeal (as-fed basis)'

Table 2. Coinpositioti of experimental diets (as-fed ha.-

IIeiiiici'lI,i]oe level. V

Item 6.2 1040 14.0 IS.))

Iiigred alit. V

Corn 71.05 62.60 53.75 1 1.9 1Soybean meal 25.70 21-65 17.50 13.30Corn germ meal' - 12.55 25.60 38.69Dicalciuni phosphate 1.61) 1.40 1.19 (1.98Limestone 0.55 0.70 0.85 1.00Sodium chloride 0.35 0. 1 3 5 0.35 035

III mix 2 0.36 0.36 0.36 0.36

L-LvsHCI 0.26 0.29 0.32 0.35i,-Thr 0.07 0.06 0.05 0.03Di,-Met 0.06 0.04 0.03 -

Caicuiati-'d analysis. V.CF 17.8 18.7 19.7 20.7Lys 1.05 1.03 1.02 1.00Met. 0.34 0.33 0.33 0.32rhr 0.64 0.63 0.62 0.61Ca 0.69 0.69 0.69 0.69P 0.35 035 0.35 0.35NDF 9.6v 11.57 19.67 24.79AIW 3.50 1.50 5.67 0.78Hemicellulose 6.17 10.01 14.00 18.00Starch 14.98 41.34 37.61 33.89Crude fat 2.53 2.18 2.12 2.38ML. Meal/kg 3.31 :1.26 :i.i :1.16

'Soh-en)-extracted corn germ macal.2The vilanmill and trace m,ueral premix provided I he lolloivimig per

kg of diet): vitamin A. 11.023 IC; vitamin fl,. 2.7511 1t: vita win [.55I U: vitamin B,,, 55.0 pg: riboflavin, 16.535 tog: pantotiieiiic acid. 44.1tug: niacin. 82.7 ing: Zn. 150 log: Fe. 175 lag: Mn, 60 log: (",u. 17.5 tag;I. 2 mug: and Se. 0.3 mg.

0.9 kg) were randomly assigned to 1 of 4 treatmentsof increasing dietary henncellulose levels (6.2 to 18%Iiemieellulose). The dietar y hetnicellulose level was in-creased by the addition of solvent-extracted corn germmeal (Cargill. Eddyville. IA), which is relatively highin heinicellulose and NDF content.. The nutrient analy-sis (Table 1) of the corn germ meal used in the studywas conducted by the University of Missouri ColumbiaExperiment Station Chemical Laboratories. The di-ets (Table 2) were formulated to meet or exceed NRC(1998) requirements for growing swine for all nutrients.Pigs were reared in individual pens within air environ-mentally controlled facility, and there were 12 pigs perdietary treatment. All pigs were allowed ad libitum ac-cess to feed and water. and BW and feed intake weremeasured every week of the 4-wk experiment.

Nonfasting blood samples were collected front each ofthe pigs fed the least. (6.2%) and greatest levels (18%) ofhenriceflulose on d 7. 14, and 28 between 0700 and 0800h on each sampling day. Blood (10 mL) was collectedfrom a jugular vein into vacuum containers contain-ing sodium heparin (Becton Dickinson. Franklin Lakes,NJ), and the resulting plasma was stored at -20°C mi-til analyzed for metabolites and insulin. Plasma washarvested by centrifugation at 3,500 x g for 20 min at4°C. On d 28, at the completion of the growth stud y. a

kkkk

194 \Teber et al.

random subset (n 16: 8 pigs/treatment) of pigs fedthe least and greatest liemicellulose levels were killedby penetrating captive bolt. The liver was immediatelyexcised and weighed, and liver samples were snap frozenin liquid nitrogen and stored at —80°C until analyzedfor glycogen, triglycerides, and protein. Samples of livertissue were also placed into tubes containing RNAlaterRNA stabilization solution (Ambion Inc., Austin, TX).Tissue samples from the jejunum were collected fromthe middle third of the small intestine, and tissue sam-pies from the ileuni were collected from 15 cm cranialto the ileal-cecal junction. Colon tissue samples andcolon contents were taken from a location 15 cm distalto the cecum. Colon contents were immediately placedon ice and stored at —20°C until they were analyzed forVFA. After dissection, intestinal tissue samples wererinsed with ice-cold PBS and immediately placed intotubes containing RNAlater or snap-frozen in liquid N.Snap-frozen tissue samples were stored at —80°C un-til they were analyzed for proteins. In addition. 10-cmsections of jejunum and ileum were collected from thesame location detailed above, rinsed with ice-cold PBS,placed on dry ice, and stored at —20°C until analyzedfor sucrase and alkaline phosphatase enzyme activity.

Hepatic Glycogen and Triglyceride Content

Hepatic glycogen content was determined using de-scribed previously methods (Benevenga et al.. 1989)with slight modifications. Frozen liver tissue was pow-dered with a mortar and pestle, and 0.5 g of the pow-dered sample was homogenized in 5 mL of ice-cold 0.15N perchloric acid. The homogenate was centrifuged at800 x q for 20 min at 4°C. and the supernatant wasdiluted 1:6 with ice-cold sterile water. The release ofglucose from glycogen was determined by adding 25[LL of the diluted homogenate to 230 1wL of an ace-tate buffer (0.2 M; pH 5.0) containing 78.3 units/mLof amyloglucosidase (Sigma Chemical Co., St. Louis,MO) and incubating the mixture for 1 ii at 38°C. Theglucose liberated from glycogen was determined using ahexokinase-based glucose assay kit (GAHK20, Sigma).The tissue glycogen content was corrected for free glu-cose by incubating the tissue homogenate sample at thesame conditions but without amyloglucosiclase.

Analysis of liver triglyceride content was performedby tissue saponification in ethanolic potassium hydrox-ide (Salmon and Flatt. 1985: Norris et al.. 2003). Brief-ly, powdered liver tissue (0.1 to 0.3 g) was digested byincubation overnight in 350 1iL of ethanolic potassiumhydroxide [2 parts ethanol:1 part 30% (wt/vol) potassi-um hydroxide] at 55°C. After the overnight incubation,the volume of the digested tissue mixture was adjustedto 1 mL with 50% ethanol and centrifuged for 5 mmat 10,000 x g at room temperature.temperature. The resulting su-pernatant was diluted to 1.2 mL with 50% ethanol andvortexed. A portion (200 1iL) of the diluted supernatantwas combined with 215 1iL of 1 M magnesium chlorideand placed oil for 10 nun. The mixture was centri-

fuged at 10,000 x g for 5 inin at room temperature, andtime supernatant was analyzed for glycerol content usingan enzymatic triglyceride kit (T7531, Pointe ScientificInc., Lincoln Park, MI).

Small Intestinal Mucosal Sucraseand Alkaline Phosphatase Activity

The intestinal segments were thawed on ice and cutlongitudinally, and the mucosa of each segment wasscraped from the underlying layers using a glass micro-scope slide. Mucosal sucrase and alkaline pliosphataseactivities were determined using previously detailedmethods (Tang et al.. 1999) with modifications. Mu-cosal samples (0.2 g) were homogenized in 1 mL of ice-cold deionized water and centrifuged at 2.200 x g for 30min at 4°C. The supernatant was diluted in deionizedwater 1:25 for sucrase activity and 1:6 for alkaline phos-phatase activity. Protein concentrations of the dilutedmucosal hoiiiogenates were determined using the Bicin-choninic Acid Assay (Pierce. Rockford, IL) method us-lug BSA as a standard. Sucrase activity was determinedusing the sucrose concentration and incubation condi-tions described b y Kidder and Mariners (1980). Theglucose liberated froni sucrose was measured using ahexokinase- based glucose assay kit (GAHK20. Sigma).Alkaline phosphatase activity was determined using acommercially available kit (A7505, Pointe Scientific:)per the manufacturer's instructions.

Determination of VFA in Colon Contents

Approximately 4 g of the mixed colon contents wasplaced into a. previously tared 15-mL polypropylenecentrifuge tube. The tubes containing the colon con-tents were then centrifuged at 21,000 x g for 23 minat 4°C to remove sample debris. The supernat.alit wasplaced into a new tube and o-phosphoric acid was add-ed to obtain a pH of 2.0 to 2.5. Then 1 niL of samplewas added to a gas chromatography (GC) vial (20-mLCC headspace vial with a septum cap. Agileut Tech-nologies, Wilmington, DE) along with 0.3 g of NaCl.A solid phase microextract.ion (SPME) fiber holderfor manual injection andi fused silica fiber coated with70-111 Carbowax/DivinylhenZene were obtained fromSupelco (Bellefonte, PA). The SPME fiber was condi-tioned for 30 min at 300°C in a. helium atmosphere inthe injection port of the GC instrument before use. Thesample was magnetically stirred at 70°C for 15 mimi on aMPS2 multipurpose sampler (Cerstel Inc., Linthicum.MD). Sampling was then performed by exposing theSPME fiber to the headispace of the vial for 5 mm. Thefiber was then retracted into the syringe and quicklytransferred to the CC inlet liner iii the injection port.The fiber was manually pushed out of the syringe at230°C for 300 s to allow time extracted a.mma.lytes to hethermally desorbed from the fiber as the CC te i miperil-

ture prograni was initiated.

Corn fiber and growing pigs 195The samples were aiialvzed on an Agilent 7890A CC

flunit equipped with a flame ionization detector and HP-FFAP (011111th (30 in x 0.25 mm x 0.25 nn AgilentTechnologies). The CC parameters were as follows:splitless mode: inlet temperature, 230 C C: inlet pressure.24.56 psi: septiun purge flow, 30 mnL/min; constant col-umu flow 1 rnL/mniri (helium); and detector I eiiipera-ture. 300°C. The GC oven temperature program wasinitial temperature, 100°C, 2 miii hold: ramp of 10°C/min to the final temperature of 240°C. hold for 2 min.

Plasma Metabolites and Insulin

Plasma glucose concentrations were determined usingall enzymatic kit (GAHK20. Sigma Chemical) based onhiexokinase activity. Plasma cholesterol and triglycer-ides were quantified using enzvimiat ic kits (C7510 andT7531, respectively. Pointe Scientific). Time intra- andintera.ssav CV for the cholesterol assay were 0.8 and1.1%, respectively, and the iutra- and interassay CV forthe triglyceride assay were 1.0 and 2.9 0/c, respectively.Serum insulin concentrations were determined using aPorcine-specific insulin ELISA kit (10-1129-01, ALP-CO. \Vindliam. NH). The insulin ELISA has a range ofdetection of 0.02 to 1.5 ng/mnL and intra- and interns-say CV less than 100/c.

Isolation of Total RNA and Real-TimeRT-PCR

Total HNA was isolated from liver and intestinal tis-sue samples using Trizol (Invitrogen Inc.. Carlsbad,CA) reagent according to the manufacturer's protocol.and the RNA pellets were resuspended in nuclease-freewater. To eliminate possible genonnc DNA contamina-tion, the RNA samples were treated with a DNase I kit(DNA-free, Ambion Inc.). Total RNA was quantified bymeasuring the absorbance at 260 nin using a NanoDropND-100 spectrophotometer (NanoDrop Technologies.Rockland, DE), and the purit y was assessed by deter-mining the ratio of the absorbance at 260 and 280 nm.All samples had 260/280 nm ratios above 1.8. Addition-ally, the integrity of the RNA preparations was verifiedby visualization of the 18S and 28S ribosomal bandsstained with ethicliuin bromide after electrophoresis on1.2% agarose gels (E-gel: Invitrogen Inc.). Total RNA(1 Lg) was reverse transcribed using a commerciallyavailable cDNA synthesis kit (iScript, Bio-Rad Labo-ratories, Hercules, CA). The iScript kit used a blendof oligo (, (IT) and random hexamer primers for eDNAsynthesis, and the reverse transcriptase is RNase H' toensure removal of the RNA template.

Real-time PCR detection of the mRNA was conduct-ed utilizing the SYBR Green assa y. Primers used forreal-time PCR are presented in Table 3. Amplificationwas carried out in a total volume of 25 nL containing1X iQ SYBR Green Supermix (Bio-Rad Laboratories),forward and reverse primers (0.1 ig/tL), and 1 1jL ofthe 20-1iL eDNA reaction. After an initial 5-nun de-

naturation step at 95°C, the reactions were cycled 40times under the following parameters: .95°C for 30 s.60°C for 30 s, and 72°C for 30 s. Optical detection wascarried out at 72°C. At the end of the PCR, melt curveanalysis was conducted to validate the specificity ofthe primers. Thermal cycling conditions and real-timedetection were conducted using an iQ5 multi-color real-time PCR detection system (Bio-Rad Laboratories). Anontemplate control was included with every assa y, andall deterirunat ions were performed in duplicate. Thepresence of a single PCR product of the correct sizefor each prinier set was verified by visualizing the PCRproducts via electrophoresis on 1% agarose gels stainedwith ethidiuni bromide. Samples of PCR products werealso sequenced to confirm the identity of each gene.The niRNA expression values for each sample werenormalized to cyclophilimi A according to themethod (Liva.k and Schmit.tgen 2001). The amplifica-tion efficiencies of cvclophihin A and the genes of in-terest were similar as determined by the generation ofstandard curves using diluted plasmicls containing therespective eDNA. The ruRNA expression of eyclophihinA was not affected by dietary fiber level in any of thetissues analyzed.

Western Blot Analyses

Snap-frozen samples of jejunum, ileum, colon, andliver tissue were homogenized in ice-cold T-PER Main-mnaiia.n Protein Extraction Reagent (Pierce) supple-niented with protease (Calbiochemn. San Diego, CA) andphosphatase (Sigma) inhibitors. Protein concentrationswere determined using the Bicinclioninic Acid Assay(Pierce) method. Equal amounts of protein extracts (50pg) were fractionated by SDS-PAGE and transferredto 0.45- p.m nitrocellulose membranes. The niembraneswere stained with Ponceau S to ensure uniform loadingloadingand transfer. Blocking of the membranes was achievedby incubating for I h at room temperature with 5%(wt/vol) nonfat dry milk in Tris-buffered saline con-taining 0.10/ Tween-20. Then the membranes wereincubated overnight at 4°C with the phospho-AMPK(Thr172) primar y antibody (Cell Signaling Technol-ogy, Beverly. MA) at a 1:5,000 dilution. After wash-ing with Tris-buffered saline containing 0.1% Tween20, the membrane was incubated with a horseradishperoxidase-labeled secondary antibody (Zymned, SouthSan Francisco, CA). The reaction complexes were visu-alized using a chemilumuimiescent detection kit (Pierce),and images were acquired using a Kodak Image Pro4000 min imaging system (Rochester. NY). The blotswere then stripped using Restore Western Blot Strip-ping Buffer (Pierce) for 15 nun at 37°C. The abundanceof AMPKII was detected using a rabbit anti-AMPKaantibody (Cell Signaling Technology) diluted at 1:5.000in blocking solution. The AMPKQ. reaction complexeswere visualized as described above. Band densit y forphosphorylated-AMPK was normalized against thedensity for AMPK. The abundance of cytochrome C

.Aflipli(( insize, bp

105

115

93

424

147

144

160

2CC('.1&)Ii mmii ar

EU 22639

NM_2 14266

NM_21 3963

AJ006757

AY242 124

EUO3()283

AYO(I)8846

196 Weber et al.

Table 3. Porcine-specific prilliers used for real-time PCR

Primer sequences (5' -. 3 (2

(S) GCATAGTTC4GGTGAGCCACA(AS) CCTG(TTCA1'CCACACA1(\(S) C'.GGTCAAGCTTGC;ATTC'TGT(AS) TCAGACTGGCACACATTTCG(S) GATGTGTCGCCTTCTTGTTC(AS) CATC'CTF'I'GGGCTCTTTGAG(S) TTCAAACACATCAC'C'CCCC'TGC(AS) GCTTCACATTC'AGCAAA(CTGGGC(5) ACTCACAGGGCACGTTTACC(AS) TCC'CTTCAGCA'I'GTCTCTCA(S) CTGGAACAGCTTGCAGGAAT(AS) C'CTAGGACATCGAGCAAC'CJ\(S) GCGTC'I'CCTTCGAGCTGTI(AS) CcATIATGGCGTCTGAAG'1C

(en&

AMPF1oJ

AMPKn2

PG C-10

PPAR2

Proglucagon

Slit!

Cvclophilin A

1 AMPK = adenosine n ionophospliate-activatcd protein kmuae. PE4('- j = PPAII lIar tim ,Our 1mm. PPAR52

= PPAR52 isoforin. sirti = sirtun I.= sense pruner, AS antisense primer.

oxidase subumt IT (COXII) was (icterinined using theWestern blotting procedures described for AMPKa.The mouse anti-COXII monoclonal antibody (Mito-Sciences, Eugene, OR) was diluted 1:1000 in blockingsolution, and after an overnight incubation and wash-ing, the membrane was incubated with the horseradishperoxidase-labeled secondary antibody. The reactioncomplexes were visualized as described above.

Statistical Analysis

The data were anal yzed as a randomized completeblock design by using the GLM procedure (SAS Inst.Inc.. Cary, NC). To determine effects of increasing lev-els of dietary hemicellulose oil performance andfeed intake (11 = 12 pigs/treatment group), the datawere analyzed by single-factor ANOVA with the GLMprocedure of SAS. The model included dietary treat-inent, and the residual mean square error was used asthe error term. Linear and quadratic contrasts were alsoperformed to determine the nature of the dose-responseto dietary henucellulose level. The data obtained fromtissue samples collected from the subset aininals killedat the completion of the study (the least and great-est treatment groups) were analyzed as a single factorANOVA. For these data, the individual pig (ml = 8 pigs/treatment group) served as the experimental unit. Dif-ferences were considered statistically significant whenP < 0.10, and treatment trends were discussed when0.10 < P < 0.13.

RESULTS

Growth Performance

Effects of increasing dietary hcmicellulose froni cornoil performance in growing pigs are presentedin Table 4. During wk 1 of the study, there was a linearand quadratic decrease (P < 0.05) in ADG. ADFI. andG:F as the level of heimncellulose was increased. The ad-

dition of hemiceilulose resulted in a linear increase (P< 0.08) in ADFI during wk 4. Over the entire courseof the 4-wk study, dietary hemicellulose level had noeffect oil or ADFJ. but altered (P < 0.03) G:F ina quadratic fashion.

Hepatic and Plasma Metabolite Measurementsand Intestinal Enzyme Activities

At the completion of the 28-d experiment, the a.b-solute and relative liver weights were decreased (P <0.05) in pigs consUIllilIg the diet high in hemicelhilose(Table 5). Consumption of the high hemicellulose diettended to decrease the abundance of glycogen (P <0.13) and triglycerides (P < 0.12) in liver tissue. Plas-nia cholesterol or insulin concentrations were not al-tered by dietary heiniceilulose oil 7 or 28, but bothtended (P < 0.13) to he decreased oil 14 of the study.Plasma glucose concentrations did not differ betweentIme 2 treatment groups at an y time point, but plasmatriglyceride concentrations were increased (P < 0.05)in pigs fed the high hemnicellulose diet oil 7 and 28 ofthe study.

There was no difference in jejunal mucosal alkalinephosphatase or sucrase activity between the low andhigh hemnicellulose treatnient groups (Table 6). How-ever, there was greater (P < 0.08) mucosal alkalinephosphatase activity and a tendency (P < 0.12) forgreater inucosal sucrase activity in the ileuni of pigsfed the high hemmucellulose diet. The total protein con-tent of jejlinumn. ileuni, or colon tissue was not affectedby dietary treatment. Acetate, propionate. or butyrateconcentrations in colon contents were not different inpigs fed the low or high hemicellulose diets.

Intestinal and Hepatic mRNA and Proteins

The expression of AMPKu1. AMPKe2, or PPAR2mRNA in jejummm, ileum. colon, or liver tissue were not

Corn fiber and growing pigs 197

Table 4. Growth perforitiatice of pigs fed increasing amounts of corn fiber'

1-Jeiu ieelluloat'. ¶4 Cl 06 mat I-'- vo in

mm

6. 10.0 14.0 18.0 SF Moth'! P-value 1,iiteam Quadratic

Tuitial B'tV. kg 30.5 30.8 31.2 30.5 (LU 0.94 0.92 0.1S.5Final BIN'. kg 5.1.4 51.8 5-1.7 54.5 1.4 0.89 0.63 0 74ADO. kgWk 1 084 0.79 0.66 0.59 0.04 <0.001 <0.001 <0.001Wk 2 0.79 0.90 0.81 088 0.05 0.36 0.48 0.71Wk 3 0.89 0.86 (1.01 0.86 0.06 0.92 0.82 ((.96Wk 4 0.90 0.9-1 1.00 0.96 0.05 0.5)) 024 036Wk 1 to 4 0.84 0.88 0.81 0.82 0102 0.51 0.35 0.40

ADFT. kgWk 1 1.69 1.55 1.48 1.11 0.06 0.007 <0.001 0.002Wk 2 1.70 1.84 1.68 1.77 0.08 0.51 0.86 0.94Wk 3 2.09 2.0.1 2.12 2.06 007 0.89 0.05 ((.99\V1, 4 2.28 2.2.1 2.32 2.4:3 0.07 0.23 0.08 0.12Wk 11.04 1.92 1.89 1.88 1.91 0.05 0.06 0.89 0.85

Ci:F. kg/kgWk 1 0.492 0.510 0.-150 0415 0.019 0.005 0.002 0.003Wk 2 0.467 0. 196 0. 162 0.496 0.024 0.62 0.61 0.87Wk 3 0.118 0.422 0.430 0.412 01026 0.97 0.90 091Wk -1 0.396 0-118 0.430 0.397 01016 ((.37 0.90 0.23\Vk I to 1 0.141 0.164 0.446 0.130 0.008 003 0.17 04)3

l)i('t (try Item jeellulose levels were increased via the addition of solveiit -cx) racted torn gem, 111(01: ii - 12 individually fed pigs/treatntenl.

affected by dietary liemicellulose level (Table 7). Like-wise, a lugh dietary level of heiriicellulose had no effecton the expression of proglucagon niR,NA in jejuntun.ilcuni, or ('01011 tissue. The expression levels of PGC-laor Sirt 1 mR NA were not altered by dietary hemicellu-lose in jejunum, ileuni, or liver tissue, but pigs fed thediet with high heniicellulose had a greater (P < 0.09)

Sirt. 1 mRNA expression and t.eiidcd (P < 0.12) to havegreater PGC-1o, mRNA expression in colon tissue.

The activation of AMPIK hi intestinal or liver tissuewas not affected by dietary hemicellulose level (Figure1). Levels of the rnitocliondria.l respiratory chain proteinCOXII were increased (P < 0.05) in the colon tissue ofpigs fed the high liemicellulose diet., but there was no

Table 5. Effect of elevated dietary coru fiber on liver weight., liver glvcogeu, liver trig-lvcericles. plasma iiieta,hohites. and insulin'

iummirim

(tent 6.2 1"'M SF P-value

BW. kg 55.6 53.7 1.4 11.33Liver weight. kg 1.52 1.35 0.05 0.03Relative liver weight, ¶4 of BW 2.7:3 2.51 0.06 0.03Liver glycogen. ing/g 27.72 18.36 4.43 0.13Liver trigl ycerides. nig/g 6.22 5.50 0.31) 0.12Plasma metabolitesCholesterol, ,ng/dLd 7 82.3 8:3.0 3.4 0.80d 14 87.3 80.0 3.3 0.13d 28 83.6 82.2 3.9 0.81

Glucose. mg/d L(I 7 128.2 111.7 8.5 0.28d14 125.8 133.3 10.1 0.61d 28 130.8 137.0 10.3 0.67

lriglvcerides. ing/dL(I 7 '21.9 33.2 2.5 0.005d14 24.6 28.0 2.7 0.39d 28 32.8 4-1.8 3.4 0.02

lilsi thu. ug/mLml 7 0.12 0.1:3 0.01 056d 1 0.1-1 0.12 0.01d 2N 0.17 0.17 0.02 0.88

I leuineelltilose level was iulc'rease I by adding solvent-ext racted con 1 genIi nicol to the diet. n = 8 pigs/treat-ment for liver nleasi,re',ients. and n = 12 pigs/treatment (hr plasma metabolites and insi thin. 4

198

Weber et al.

Table 6. Effect of elevated dietary corn fiber on small intestinal mucosal alkaline phos-phatase and sucrase activity, tissue proteill content, and colonic VFA'

Item

Alkaline pliospl i at abe. mm i,l ii in nig of proteinJejunumIleum

Sucrase activity. Lmolmirm •g of protein -

JejunmimIleum

T1SSUC protein. nig/gJej unionIleurnColon

Colonic VFA, inmnol/LAcetatePropionateButvrate

Heniicellulose. /1

6.2 18.0 SE P-value

11.9

9.1

1.8

0.3-1

4.4

7.2

1.0

1)08

149.5

108.7

24.2

0.2.5

74.6

131.1

23.9

0.12

43.9

41.1

4.6

0.68

46.1

42.0

2.7

0.29

36.7

35.8

3.9

0.87

192.7

178.9

15.7

0.54

88.0

90.6

4.2

0.67

33.1

36.9

2.5

((.31

'Dietary henucellulose level was increased b y the addition of solvent-extracted corn germ meal. mm = 8 pigs/treatment.

effect of dietary hernicellulose on COXII protein abun-(lance in jejunum, ileum, or liver tissue (Figure 2).

DISCUSSION

A major objective of this experiment was to deter-mine the level of hemicellulose provided by corn germmeal that could be tolerated by growing pigs when MEwas not equilibrated across treatments. Except for thedepression in ADFI and ADG during the first weekof the study, increasing the level of hemicellulose hadlittle impact on overall growth performance during the4-wk study. In fact, at decreased levels it appears asif increasing dietary henucellulose by increasing corngerm meal increased G:F. This finding agrees with re-cent work in pigs where it was observed that feedingwheat mill run (another feedstuff that is relatively highin NDF and hemicellulose content) improves G:F (PorkCRC, 2007). Our findings are also in agreement witha recent study (Harbach et al., 2007) in which dietarycorn germ meal, when fed up to 40% of the diet, had nodeleterious effect on growth performance when the di-ets were formulated to he isocaloric. These findings alsoagree with what has been observed for growing pigsfed dried distillers grains containing solubles (DDGS),another corn coproduct that is high in henilcellulose(25%). A recent review of studies examining the useof DDGS in swine diets (Stein and Shurson. 2009) in-dicates that DDGS up to a dietary concentration of30% has no deleterious effects on growth performance.Taken together, these data suggest that corn coprod-ucts that are high in hemicellulose appear to be wellutilized by growing pigs.

It is interesting that the abundance of phospliory-lated AMPK was not increased in any of the tissuesthat were measured in pigs fed the diet with corn germmeal. Given the decreased ME content of the high corngerm meal diet vs. the control diet and the well-accept-

Table 7. Effect of elevated dietary corn fiber on therelative abundance of selected mRNA for genes involvedin energy metabolism in intestinal and liver tissue

llemnicellulose. 2 V

Tissue nilINA'

6.2 18 SE P-value

Jejummimimi mnRNAAMPKs 1

0.383 0.479 0.187

((.71AMPKo2

0.091 0.197 0.083

(1.38PGC- 1

0.473 0.980 0.380

0.36PPAi(2

0.232 0.629 0.324

(1.40Proglucagon 0.239 0.373 0.121

0.44

Sirti

0.614 0.791 0.355

0.73Ileuum niB NA

AMPKO 1

0.032 0.093 0.044

0.34AMPKc2

0.072 0.12)) 0.049

(1.49PGC-lo 0.053 0.066 0.033

0.79

PPAB52

0.036 0.058 0.024

(1.48Proglucagon

0.221 0.385 0.105

0.27Sirtl

0.125 0.065 0.037

(1.26Colon mRNAAMPKo1

0.147 0.140 0.029

0.86AMPKc2

0.003 0.003 0.001

(1.80PGC-lo, 0.052 0.114 0.027

0.12

PPAR52

0.128 0.097 0.028

(1.42Proglucagon

0.127 0.151 0.036

0.64Sirti

0.203 0.366 0.066

0.09Liver moB NAAMPKu I

0.388 0.413 0.136

0.90AMPNo 2

0.622 0.368 0.177

0.31PGC- 11, 0.432 0.395 0.137

(1.85

PPAB52

0.059 0.038 0.015

0.39Sirti

0.426 0.339 0.094

(1.52

'AI\IPI< = adenosine muonophosphate-activated protein kinase:PGC-lo = PPAR coactivator 10 PPAB,2 = PPAR52 isoform; sirti= sirtuin 1.

2Pigs (n = 8/treatment) were fed control diets or diets containingelevated fiber as hemicelhmlose (room corn by the addition of solvent-extracted corn germ meal. The mHNA af ,unil,imiee is normalized to theniB NA expression of cvlophilimi A.

Corn fiber and growing pigs 199

. - .,p-AMPK

I —

AMPK—* -

Cl) 1

I.-

06.a0.4

00.2

a o

SE = 0.05

Jejunum

U 6.2% HCD18% HC

SE = 0.05SE 0.03

Ileum Colon

SE = 0.17

Liver

Tissue 1

Figure 1. Effect of high iIietarv fiber as lieinicelhilose (HC) from corn Oilo the al iiiiidance of pliospliorylate d adenosine nionophospliat e-activat -ed protein kiiiase (Thr 172; p-AMPK) iii various intestinal segments and liver tissue. The values for pA-MPR were normalized to tile abundanceof AMPKo. The values (n = 8 pigs/treatmeiit) are presented as arbitrary cleusitonietric units. There was no significant difference between thetreatment groups in any of the tissues.

ed notion that dietary hemicellulose reduces dietaryNE content (Noblet et al., 1994), it was hypothesizedthat feeding the diet with corn germ meal would in-crease the activation of AMPK. The trend toward de-creased hepatic glycogen and triglycerides may be dueto a reduced NE content. Also, the tendency towarddecreased liver glycogen may be due to the reducedstarch (25%) content of the diet with corn germ meal.Previous studies in pigs have reported a reduction in

—COXIL

muscle glycogen content when dietary fiber levels wereincreased (Rosenvold et al.. 2001). However, furtherstudies are needed to determine whether the reductionin LM glycogen content observed by Rosenvold et al.(2001) and the trend toward decreased liver glycogenin the current study are due to a reduction in dietarystarch content or due to an overall reduction in dietaryNE. The reduction in liver weight in pigs fed the dietcontaining corn germ meal agrees with previous work

.,26 kDa

3,000U 6.2% HC SE = 156 SE = 171

2,500 D18% HC b

2,000 ar-

1,500

SE = 204

.: 1,000 SE = 75

0

0 _ilftl

Jejunum

Ileum

Colon

LiverPP Tissue

Figure 2. Effect of high diet arv hieuncelh ili se (11C) from corli on the abundance of the mitochondrial protein c ytochironie c oxidase 11 (COXIIin different intestinal tissue segments and liver tissue. The values (n = S pigs/treatment) are presented as arbitrary densitometric units. \leansWithin all location with different letters are different (P < 0.05).

200 Weber et al.

where it was observed a high fiber diet containing al-falfa meal decreases liver weight (Anugwa et al., 1989).Whether reductions in liver weight are entirely due toreductions in hepatic energy stores remains undeter-nuned, but these data do suggest there is a repartition-ing of nutrients from the liver when lugh fiber diets arcfed.

Although AMPK activation has been shown to in-crease mit.ochondrial protein expression (Jorgensen etal. 2007) and mitochonclrial biogenesis (Bergeron etal., 2001; Zong et al., 2002), we did not find increasedAMPK activation or expression to coincide with theincrease in colonic COXIT in the current stud y. Thissuggests that increasing dietary hemicellulose increasescolonic nutochondrial protein expression independentof ail ill activation or expression. Al-though AMPK was not altered, the expression of an-other regulator of mitochondrial function, Sirti, wasincreased, indeed, it was recently reported (Feige et al.,2008) that the activation of Sirti call oxidativemetabolism of lipids, without the direct activation ofAMPK. Likewise, the expression of PGC-1, a masterregulator of mitochondrial biogenesis (Wu et al., 1999),tended to he increased in colon tissue in pigs fed thehigh fiber diet. The activation of Sirti is associated withincreased PGC-la expression (Feige et al., 2008). Fur-thermore, Sirtl directly stimulates PGC-1a. activity viaLys deacetylation (Ncmoto et al. 2005), indicating analternative pathway ill dietary fiber may regulatePGC-la activity in addition to PGC-la expression. Ithas been well established that the expression of Sirt.1is increased by caloric restriction (Cohen et al. 2004;Chen et al.. 2008) but this is the first report suggestingthat dietary fiber regulates colonic Sirtl expression.

The increase in COXII expression may indicate in-creased ability for oxidative metabolismin the colonof pigs fed a high-fiber diet. This agrees with studiesin rats (Marsman and McBurney. 1995) where colono-cytes isolated from rats fed a high fiber diet had greaterrates of glucose and VFA oxidation than rats fed thecontrol diet. Additionally, the organs drained by theportal vein in pigs fed a diet with increased fiber con-sun me more oxygen than pigs fed typical corn-soy-baseddiets (Yen et al., 2004). Mechanistically, ailin dietary fiber levels may lead to increased oxidativemetabolism in the colon. in part through increasing theexpression of Sirti and the subsequent effects of Sirtion initoehondrial metabolism.

Although the increased COXII in colon tissue ob-served in pigs fed the high fiber diet agrees with thefindings of Rodenburg et al. (2008), where feeding fume-to-oligosaccharides increased the expression of COXTI,an increase in colonic VFA was not observed in thepresent study. Rodenhurg et al. (2008) proposed thatan increase in colon VFA production and a concomitantdecrease in intestinal harrier function were factors re-sponsible for increased initochondrial gene and proteinexpression. Given the increase in ileal alkaline phos-phatase activity and trend for increased ilea,l sucrase

activity in the present study, it is unlikely that pigs fedthe high fiber diet had decreased intestinal harrier func-tion. Other studies have reported a positive relation-ship between intestinal harrier function and mucosalenzyme activity (Luk et al., 1981; Gomez-Conde et al.,2007). Even though all increase in colonic VFA was notobserved in pigs fed the high hemicellulose diet, it ispossible that VFA were increased in other sections ofthe intestine. Kass et al. (1980) observed that VFA con-centration was increased in the cecuni, but not in thecolon when pigs are fed increased levels of alfalfa meal.Nonetheless, our data indicate that other niechanisms,in addition to or aside frorn ail in colonic VFAproduction, are responsible for the increased mitochon-drial protein expression in colon tissue of animals fed(bets high in fiber.

It appears as if another metabolic adaptation in pigsfed the diet high in corn fiber was ail in plasmatriglyceride concentrations oil 7 and 14. This is in-teresting because increasing the fiber level in pig dietsby adding wheat bran decreased dietary lipid absorp-tion (Wilfart et al.. 2007). Indeed, studies conducted inrodents (Artiss et al.. 2006; Galisteo et al.. 2008) havereported that some types of dietary fiber decreased cir-culating triglycerides, attributed to the decreased Up-take of dietary lipid. Therefore, it is unlikely that theincreased plasma triglycerides observed ill pigs fed corngerm meal in the present study originated from the dietand may reflect repartitioning of energy stores withinthe body as represented by the tendencies for decreasedliver triglycerides. This repartitioning of lipid may bean adaptation to the reduced starch content of the diethigh in corn fiber as the pigs may be oxidizing more lip-ids to spare glucose. In addition, increased circulatinglipids may provide substrate for the increase in oxida-tive metabolism occurring in the intestine of pigs fed adiet high) in fiber.

In conclusion, the data presented herein demonstratethat growing pigs can tolerate relatively high levels ofhenncehhuhose from corn germ meal without depressinggrowth performance over a 4-wk period. These resultsalso suggest that manipulation of dietary hiemicelluloseby feeding corn germ meal may regulate feed efficiency.The metabolic adaptations to corn fiber include de-creased liver mass, trends for decreased liver energystores, and increased plasma triglycerides. Alterationsin colonic mitochondrial COXII content, associatedwith time increase in colonic Sirti expression, suggestthat colonic mitochondrial protein expression may beaffected by dietary fiber in a manlier independent ofAMPK activation. These results support the notion ofan increased capacity for energy utilization by the in-testiiie in pigs fed high fiber diets and suggest a repar-titioninig of nutrients toward the intestinal tract.

LITERATURE CITED

Ajiugva. F. 0. 1., V. H. Varti .1. S. Dickson. W. C. Pond. and L. P.krook. 1989. Effects of dietary fiber 0.11(1 protein COlICclItratlall

Corn fiber and growing pigs 201on growth. Iced efficienc y, visceral organ wei ghts and large iii-testme microbial populations of swine INut r. 119:879 886.

Artiss, J. D.. K. Brogan, M. Brucal. M. Niogliaddititi. and K. L.Jen.2006 The effects of a new soluble diet ar fi sr on Nvejgljt gainand selected Wood parameters iii rats. Mc I a I olisi u 55:193-202.

Beitevenga. N......I K. Steiiiniaii-Coldsworthv. I. B. Crenshaw.aid J. Odle. 1989. Utilization of medium-chain trigl ycerides 1 >yituntiat al piglets I. Effects oil consumption amid bode fuelntiln'aiion. J. Anita. Sci. 67:3331 3339.

Bergeron, HI. M. Ben. K. S. Cadnian. T. N. Moore. P. Perret.NI. Pvpaert. L. II. Young. C. F. Senienloa'ich. and C. I. Simid-man. 2001. Chronic activation of ANtI > kinase results iii NBF-I activation and iii itoehoialrial biogenesis Anti. Phvsiol.281:E1340 E134t1.

Chen. D.. J. Bruno, N. Easlon. S. J. Lilt. H. L. Cheng. F. NV. Alt.and L. Citarente. 2008. Tissue-specific regulation of SIIITI bycalorie restriction. Genes Uev. 22:4753 1757.

Cohen. II. Y.. C. Miller. N. J. Bitterinan. N. H. \Vall. B. Hekking. B.Kessler. N. T. H(witz. M. Gorospe. R. de Cabo. and D. A. Sin-clair. 2001. Calorie restriction promotes mann italian cell surviv-al by inducing flit SIRT I tleaeetvlase. Science 365:390 :392.

Feige. J. N.. M. Lagonge. C. Canto. A. Strehle. S. M. Houten. J. C.Nlilne, P. D. Lamberi . C. Nlataki. P.3. Elliot. and 3. Auwerx.2008. Specific SJR1'1 activation niinues low energy levels andprotects against diet-induced metabolic disorders by enhancingfat oxidation. Cell Metab. 8:347 358.

Galisteo......I. Duarte, and A. Zarauelo. 2008. Effects of dlietar\fibers on disturbances clustered in the metabolic syndrome. J.Nutr. Biocbein. 19:71 81.

Coinez-Conde, NI. S.. J. Garcia. S. Cliamnorro. P. Eiras. P. C. HcTa)l-lar, A. Perez de Rozas. I. Ba(liola, C. ne Blus. and R. ('nra-I)ano. 2007. Neutral detergent-soluble fiber improves gut barrierfunction in twent y-five-clay-old weaned rabbits. J. Anim. SO,85:3313-3321.

Graham. H.. K. Hesselmnan. and P. Anian. 1986. The influence ofwheat bran and sugar-beet pull) oil digestibility of dietarycomponents in a cereal-based pig diet. J. Nutr. 116:242 251.

Harbach. A. P. II.. M. C. R. cia Costa, A. L. Soares. A. M. Bridi, M.Shimokoniaki, C. A. cia Silva, and N. 1. Ida. 2007. Dietary corngerm containing plivtic acid prevents pork meat lipid oxidationwhile maintaining non normal animal growth performance. FoodChem. 100:1630-1633.

Hardie, I). C., S. A. Hawle y. and .1. W. Scott. 2006, AMP-activatedprotein kit use Development, of the energy sensor concept. J.Phvsiol. .57 1:7 IS.

Hedemann. M. S.. M. Eskildsen, H. N. Laerke. C. Pedersen, J. E.Lindberg. P. Lnurinen. atal K. E. Knudsen. 2006. Intestinalmorphology and enzymatic activit y in newly weaned pigs fedrunt rusting fiber concentrations and fiber propertie s . ....'dmuin.Sci. 81:1375 1386.

Jorgensen. S. B.. J. T. Treebak. B. \iollet. P. Sclijerling. S. Van-lont. J. F. P. \Vojtaszewski, and N. A. Richter. 2007. Role ofANIPKc2 ill training-. and AICA1I-induced GLUT4.Icexokinase ii. and nntochoudrial protein expression in mousemuscle. Am. J. Ph ysiol. 292:E331 E339.

Kate,. M. L.. P. J. Van Soest, and \V. G. Pond. 1980. Utilization ofdiet are fiber front alfalfa by growing swine. 11. Volatile fatt vacid cot cclii rat u tics in and disappearance from time gastrointes-tinal tract.... .Animic. Sci. 50:192-197.

Kidder, B. E.. and M. .1. Manners. 1980. Time level and distributionof carbohvdrascs ill lie small intestine mucosa of pigs from 3weeks of age to uatimrit y. Br. J. Niitr. 43:141-153.

Lfvak. K. J., and T. D. Schnuttgen. 2001. Analysis of relative geneexpression data using real-time quantitative PCR and the2 ' method. Methods 25:402 -408.

Luk, C. D.. NV. 1 > . Vaughan. P. J. Burke. and S. B. Baylin. 1981.Diannne oxidase as a pltisnia mmmarker of rat intestinal maim-rosal imijtmrv and regeneration After administration of 1-13-u-arabinofuranosylcvtosimie, Cancer Has. 41:2334 23:37.

Nlarsmncui. N. F.. and Ni. 1. McBuriie' y. 1995. Dietary filter icicreases

oxidative n ietaholismim iii colonoctes but not in distal small in-testinal enterocytes isolated from rats. J. Ncmtr. 125:273 282.

Nemicot.o. S., M. M. Fergnmsson. and I. Finkel. 2005. 511111 f'uuiction-all interacts with the met a) ohic regulator and transcriptionald'oa.c'tivator I CC- ii> .3. Biol. CIani. 280:16156 16460.

Nolilet. .....II. Fortune. X . S. Sin, andI S. Dubois. 1994. Predictionof net energy vahie of feeds for growing pigs. ....Nit i ma. SO72:344 35-1.

Norris, A. NV.. L. ('lien. S. J. Fisher. I. Szanto. M. Ristow. A. C..Jnzsi, M. F. Hirshmmian. F. D. Rosen. L. .1. Good year. F. Gon-zalez. B. Al. Spiegel mnui. (01(1 C. R. Kalimi . 2003. Muscle-specificPPAR-.-deficient mice develop increased adiposity and insulinresistance hut respond to t lua,.olidimied iota's .1. Clin. Invest112:608 618.

NBC. 1998. Nutrient Requirements of Swine. 10th ed. N till. Acad.Press. Washington. DC.

Owu.stm-Asiedtm.....I. F. Patience. B. Laarvehl, A. C. Vum TKesset, P.IT. Siinnnns. and R. T. Zijlstra. 2006. Effects of guar guni andm 'ebb mbose on digesta passage mate. ileal microbial populations,eia'rgy and protein digestibility. and Performance of growerpigs. J. Anim. Sci. 84:843 852.

Piel, C.. L. Nbontagne. B. Seve and J. P. Ladles. 2005. Increasing di-gesta viscosit y using carboxvmmicthivlcellumlose in weaned pigletsstimulates den I goblet, cell numbers and maturation. J. N lit m'.1:35:86-91.

Pond. NV. C.. H. C. Jimg. and V. H. Varel. 1988. Effect of dietaryfiber cat young adult geimetically leamm, obese, and contemporarypigs: Body weight, carcass measurements. organ weights, and(Iiges ta content. J. Anim. Sci. 66:699 706.

Pork CRC. 2007. l'lfect of dietary NDF content oil rate andvefficiemie of finisher pigs. lmttp://www.porkcrc.com.amm/2B_Ittl_

Final Report -Thie_c'ffect of dietar y NDF.1mdf Accessed Jim.26. 2009.

Hoc lemihuu'g. IV., J. IKeijer. N. Kramer, C. Vink. H. vail den Meer. acidI. M. J. Bovee-Oumdenhoven. 2008. Impaired barrier function In'dietary fruct.o-oligosaccharictes (FOS) in rats is accompanied byincreased colonic nuitochomahial gene expression . 135 IC Gemmon c-ics 9:144-159.

Rosenvold. K., J. S. Peterson. II. N. Laerke. S. K. .Jensen, M.'I'Iierkililsemi ..A, H. Karlssoa. II. S. Moller, and H. J. Alider-sell. 2001. Muscle glycogen stores mccl meat piabitv as affectedby strategic limushimig feeding of slaughter pigs. J. Anim. Sci.79:382-391.

Salniomi. D. M. NV.. and J. I'. Flatt. 1985. Effect of dietar y fat con-lent oil incidence of itT esi lv aimmomig ad libi faint fed nice. liiiJ. Ohes. 9:113 449.

Steimi. H. 1-1.. (111(1 C. C. Shurson. 2009. Board-Iimvited Review: T laumse and application of distillers dried grains with sol ubles illswine diets. J. Amummi. Sci. 87:1292 1303.

Tang, NI., B. Laarvehl, A. C. Vail Kessel. D. L. Hamilton. A. Estra-da. and J. F. Patience. 1999. Effect of segregated earlweaningoh postweai ung small latest imial development in pigs. J. A mmmi.Sd. 77:3191 3200.

NVilfart, A.. L. Nlomitagne, P. H. Simnnuns. J. van Nlilgen. and J. No-blet. 2007. Sites of nutrient digestion in growing pigs: Effect ofdietary fiber. ....Aimimn. Sci. 85:976 983.

Wu. Z.. P. Puigserver, U. Aimderssomi. C. Zhamig. G. Adelimiamit. V.Nlootha, A. Tn)y, S. Cint.i. B. Lowell. R. C. Scam'pulla. andB. M. Spiegelnmami. 1999. NIcd'hmamnsnis controlling mnitochondrialbiogenesis and respiration through the thd'rniogenic coactivatorPGC-1. Cell 98:115-124.

Yen. J. 'F.. V. Ii. Varel, and .1. A. Nicmialier. 2001.- Metabolic andmicrobial responses in westcrmi crossbred and Nleishan growingpigs fed a high-fiber diet.]. Animn. Sci. 82:1740 1755.

Zomig. H., J. M. Ben. L. II. Young, M. Pypaert .3. Mu, M. J. [hrmm-baumii. and C. I. Shiil uiami. 2002. AMP kimiase is required for mmmi-toehondm'ial biogeimesis ill skeletal muscle iii response to chim'omimeenergy deprivation. Proc. Natl. Acad. Sci. USA 99:1598315987.