Baked cod consumption delayed the development of kidney and liver dysfunction and affected plasma amino acid concentrations, but did not affect blood pressure, blood glucose or liver triacylglycerol concentrations in obese fa/fa Zucker rats.

Availableonlineatwww.sciencedirect.com

journalhomepage:www.elsevier.com/locate/NTR

Original Research

Baked cod consumption delayed the development of kidney and liver dysfunction and affected

plasma amino acid concentrations, but did not affect blood pressure, blood glucose or liver

triacylglycerol concentrations in obese fa/fa Zucker rats. ^✩

Linn A. Vikøren

, Aslaug Drotningsvik

, Øivind Midttun

, Adrian McCann

,

Marthe T. Bergseth

, Maren H. Austgulen

, Gunnar Mellgren

, Per Magne Ueland

, Oddrun A. Gudbrandsen

aDietaryProteinResearchGroup,DepartmentofClinicalMedicine,UniversityofBergen,5021Bergen,Norway

bDepartmentofClinicalScience,UniversityofBergen,5021Bergen,Norway

cBevitalAS,5021Bergen,Norway

dMohnNutritionResearchLaboratory,DepartmentofClinicalScience,UniversityofBergen,HaukelandUniversityHospital,5020Bergen, Norway

eHormoneLaboratory,DepartmentofMedicalBiochemistryandPharmacology,HaukelandUniversityHospital,5021Bergen,Norway.

a r t i c l e i n f o

Articlehistory:

Received26November2020 Revised13May2021 Accepted23May2021

Keywords:

Aminoacids Metabolicsyndrome Cod

Fishfillet ObeseZuckerrat

a b st r a c t

Obesityisassociatedwithchangesinaminoacidmetabolism,andstudiesshowthatinges- tionoffishproteinsinfluenceaminoacidcompositioninplasmaandurine,inadditionto affectingriskfactorsformetabolicsyndrome.Sincethemajorityoffishproteinsconsumed byhumansareasfishfillet,itisofinteresttoinvestigateifcodfilletintakeaffectsaminoacid compositionandmetabolicdisorders.WehypothesizedthatamodifiedAIN-93Gdietcon- tainingcodfilletwouldaffectaminoacidcompositionsinplasmaandurineinobeserats, andalsoaffectriskfactorsformetabolicsyndromewhencomparedtoratsfedaregularAIN- 93Gdietwithcaseinastheproteinsource.ObeseZuckerfa/farats,aratmodelofmetabolic syndrome,receiveddietscontaining25%proteinfromlyophilizedbakedcodfilletand75%

proteinfromcasein(Bakedcoddiet),oraControldietwithcaseinforfourweeks.TheBaked coddietaffectedtheaminoacidcompositioninplasma,withe.g.,lowerglycine,histidine, homoarginine,homocysteine,methionine,prolineandtyrosineconcentrations,butdidnot affectaminoacidconcentrationsinurine.Theconcentrationsofmarkersforkidneyand liverdysfunctionwerelowerintheBakedcodgroup,howeverbloodpressuredevelopment,

✩ Abbreviations:ELISA,enzyme-linkedimmunosorbentassay;HDL,highdensitylipoprotein;LDL,lowdensitylipoprotein;TIM-1,Tcell ImmunoglobulinMucin-1;TMAO,trimethylamineN-oxide

∗ Correspondingauthorat:UniversityofBergen,5021Bergen,NorwayTel.:+4755975553 E-mailaddress:nkjgu@uib.no(O.A.Gudbrandsen).

https://doi.org/10.1016/j.nutres.2021.05.009

0271-5317/© 2021 The Author(s).Published by Elsevier Inc.This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/)

fastingandpostprandialglucose,andhepatictriacylglycerolconcentrationsweresimilar totheControlgroup.Toconclude,substituting25%ofdietaryproteinwithbakedcodfillet affectedconcentrationsofsomeaminoacidsinplasmaanddelayeddevelopmentofkidney andliverdysfunction,butdidnotaffectbloodpressure,glucoseconcentrationorfattyliver.

© 2021 The Author(s).Published by Elsevier Inc.

ThisisanopenaccessarticleundertheCCBYlicense (http://creativecommons.org/licenses/by/4.0/)

1. Introduction

Theprevalenceofobesitycontinuestoriseglobally,andthisis ofmajorconcernsinceobesityisstronglyassociatedwithco- morbiditiessuchashypertension,hypercholesterolemia,hy- perlipidemia,fatty liver,type2diabetes, renaldisease, and cardiovasculardiseases[1-4].Thecriteriaformetabolicsyn- drome include abdominal adiposity, reduced high density lipoprotein(HDL)cholesterol,elevatedbloodpressureandel- evatedfastingbloodglucose.Inaddition,severalfactorssuch asatherogenicdyslipidemia(elevatedlowdensitylipoprotein (LDL)cholesterol,reducedHDLcholesterol),elevatedliverfat, fastinginsulin,glucoseafteroralglucosetolerancetest,and chronicrenaldiseaseare listedasimportanttargets forfu- tureresearchonmetabolicsyndrome[5].Fishconsumption hasbeenassociatedwithreducedriskofcoronaryheartdis- easeandheartfailure[6-9]andprevalenceoftype2diabetes andmetabolicsyndromearelowinpopulationswithhighfish intake[10-13].Ofthese,onlyKarlssonetal.[13]distinguished betweentheeffectsofintakeofleanandfattyfish.Thus,more researchisneededtoidentifymetaboliceffectsofconsump- tionofleanandfattyfishseparately.

Thekidneysandtheliverarecentralorgansintheregu- lationofaminoacidmetabolism,anddysfunctionsinthese organsmayaffect theconcentrationsofaminoacidsincir- culation and in urine. Circulating concentrations ofamino acidshavebeenshowntobedifferentinpatientswithobe- sity,metabolicsyndromeand/orinsulinresistancewhencom- paredtohealthypersons.Specifically,elevatedconcentrations ofbranched-chain aminoacids (isoleucine, leucine,valine), andreducedglycine(aliphatic)concentrationaremarkersfor prediabetes,insulinresistance,andfuturetype2diabetesin humans[14,15].Associationsofbranchedchainaminoacids andaromaticaminoacids(phenylalanine,tryptophan,tyro- sine)withcardiovasculardiseaseandobesityhavealsobeen demonstrated[15],andevidencesuggeststhatthekynurenine pathwayoftryptophandegradationisupregulatedinobesity [16].

Theeffectsoffishintakeonaminoacidsconcentrations havenotbeen extensively explored,howeverafew clinical studies haveshownthatintakeoffish mayaffect thecon- centrationsofaminoacid and theirmetabolites incircula- tionandinurine.Fishintakewasnotastrongdeterminantof theplasmaconcentrationsoftryptophanmetabolitesinthe kynureninepathwayinpatientswithcoronaryarterydisease, howeverleanfishintakehadsomeinfluenceonthekynure- ninepathwayinasubgroupofpatientsthatalsopresented diabetes[17].Innon-diabeticadultswithoverweightorobe- sity, serum concentrationsofasymmetric dimethylarginine

andlysineweresignificantlyincreasedaftercodintake[18], butdidnotaffectserumconcentrationsofkynurenines,how- everurineconcentrationofkynurenine(relativetocreatinine) wasincreased[18].Thecytokineinterferon-γ stimulateboth theconversionoftryptophantokynureninesandthebiosyn- thesisofneopterin,andthelower neopterinconcentrations observedinhumansafterintakeofleanfish[17,19]suggests thatthekynureninemetabolismpathwaymaybeaffectedby consumptionofleanfishsuchascod.

TheobeseZuckerfa/faratisawidelyusedmodelofge- neticobesityandpresentsvisibleobesityalready3-4weeks afterbirth[20].Thisratspontaneouslydevelopsabnormali- tiesresemblinghumanmetabolicsyndromesuchasdyslipi- demia,insulinresistance,mildglucoseintoleranceandhyper- insulinemia[20].Inaddition,theobeseZucker ratdevelops proteinuria[21]andhighbloodpressure[22,23]beforetheage of10weeks,withindicationsofdecreasedrenalfunctionfrom around12weeksofage[24].

Werecentlypresentedfindingsthatinclusionofbakedfil- letfromAtlanticcod(Gadusmorhua)aspartofaregulardiet resultedinlowerserumcholesterolconcentrationsinobese Zuckerfa/farats,mostlikelythroughdown-regulationofen- dogenoushepatic cholesterolsynthesis[25].Littleisknown abouttheeffectsofcodfilletconsumptiononaminoacidcom- position,kidneyfunction,vitaminstatus,bloodpressure,and concentrationsofbloodglucoseandlivertriacylglycerol.The mainobjectiveofthepresentstudywastoinvestigatetheef- fectsofbakedcodfillet intakeon concentrationsofamino acidsandrelatedmetabolitesinplasmaandurine,withthe secondaryaimstoinvestigateeffectsonkidneyandliverfunc- tion,vitaminstatus,thedevelopmentofhighbloodpressure, bloodglucoseandlivertriacylglycerolconcentrationsinobese Zuckerfa/farats.Wehypothesizedthatcodfilletintakewould affectaminoacidconcentrationsinplasmaandurineinthis obeseratmodel.Totestthishypothesis,obese Zuckerfa/fa rats were fedmodifiedAIN-93G dietswith25% ofproteins from baked cod fillet and 75% ofprotein from casein (the Bakedcodgroup),orwithcaseinasthesole proteinsource (theControlgroup)for4weeks.Aminoacidswerequantified inplasmaandurine,andkidneyandliverfunctionmarkers, bloodpressureandconcentrationsofbloodglucoseandliver triacylglycerolswereassessed.

2. Methods and materials

TwentymaleZuckerfa/farats (HsdHlr:ZUCKER-Leprfa,from HarlanLaboratories,Indianapolis, IN,USA)were housed in

Table1– Compositionoftheexperimentaldiets

g/kg Bakedcod Control

diet diet diet

Casein¹ 162.16 216.22

Freezedriedbakedcod² 62.50 -

Cornstarch 503.22 511.67

Sucrose 90.00 90.00

Cellulose 50.00 50.00

SoybeanOil 70.00 70.00

t-Butylhydroquinone 0.014 0.014

MineralMix(AIN-93-MX) 35.00 35.00

VitaminMix(AIN-93-VX) 10.00 10.00

L-Methionine 1.60 1.60

L-Cystine 3.00 3.00

CholineBitartrate³ 2.50 2.50

GrowthandMaintenanceSupplement⁴ 10.00 10.00

1 Contains92.5%crudeprotein

2 Contains80%crudeprotein

3 Contains41%choline

4 ContainsvitaminB12(40mg/kg)andvitaminK1(25mg/kg) mixedwithsucrose(995g/kg)anddextrose(5g/kg)

pairsinMakrolonIVcages,andwerekeptata12hlight/dark cycleat20–23°C.Afteracclimatizationundertheseconditions for aminimum ofseven days,the rats were randomly as- signedthetwoexperimentalgroupswithcomparablemean body weight.Theinterventionstartedwhenthe rats’body- weight were355 ±10 g(mean± SD),i.e.approximately 8–

9weeksold.Theratswerefedmodifiedexperimentaldiets inaccordancewiththeAmericanInstituteofNutrition’srec- ommendationforgrowinglaboratoryrodents(AIN-93G)[26], added1.6gmethionine/kgdietand1wt%Growthandmain- tenancesupplementcontainingvitaminB12andvitaminK1 as recommendedby Reeves[27].Thedietsdiffered onlyin their proteinsources(Table 1). Since the rats would be in thegrowthphasethroughouttheinterventionperiod(based on growth chartsforZuckerrats from HarlanLaboratories, https://www.envigo.com),andsinceobeseZuckerratshavean impairedproteinmetabolismleadingtopoorerproteinutiliza- tionandrequiresagreaterproteinintaketomaintainamax- imalrateofproteingainduringgrowth[28],weusedtheAIN- 93Gdietwith20%(w/w)wt%proteininsteadoftheAIN-93M dietformaintenancecontaining15wt%protein.TheBaked coddietcontained5wt%proteinsfromcodfillet(Atlanticcod (Gadusmorhua)providedbyLerøySeafoodGroup,Hordaland, Norway)and15wt%proteinsfromcasein.Theadditionof25%

ofdietary proteinsfrombaked codfilletishigher thanthe US andNorwegianrecommendationsof300-450g fish/week [29,30],amountingto9%-13%ofthetotalproteinintakebased onadailyaveragetotalproteinintakeof91-96g[31,32],but correspondswellwithapreviousclinicalinterventionstudy whereadultshad amedianintakeof22% ofproteinintake asfishproteins[33].Skinfreecodfilletswerebakedinoven (180°Cfor20min)withnoaddedfat.Bakedcodfilletswere mincedusingahandblender,andwerethereafterfreezedried andground.TheControldietcontained20wt%proteinsfrom casein.Dietswerefrozenimmediatelyafterpreparation.Ca- seinwaspurchasedfromSigma-Aldrich(Munich,Germany), the other feed ingredients were purchased from DyetsInc.

Table2– Aminoacidsinthediets

g/kgdiet Bakedcoddiet Controldiet

Alanine 6.4 5.2

Arginine 7.1 6.4

Asparticacid+asparagine 14 12

Glutamicacid+glutamine 36 38

Glycine 4.2 3.6

Histidine 4.8 5.1

Isoleucine 8.9 9.2

Leucine 16 16

Lysine 14 14

Methionine 6.8 6.4

Phenylalanine 8.8 9.1

Proline 17 19

Serine 9.6 9.9

Taurine ND 0.3

Threonine 7.3 7.2

Tryptophan 2.1 2.5

Tyrosine 7.4 7.8

Valine 12 12

Valuesaremeansoftwomeasurements,deviation<5%between parallels

ND;notdetected

Thecontentsofaminoacids andtaurineinthe dietswere analyzedbyNofimaBioLab(Hordaland,Norway),andarepre- sentedinTable2.

2.2. Ethicalapproval

TheNationalAnimalResearchAuthority(Norway)approved the study protocol inaccordance with the Animal Welfare ActandtheRegulationofanimalexperiments(approvalno 2014/6979).Allapplicableinternational,nationalandinstitu- tional guidelinesforthe careand use ofanimalswere fol- lowed.

2.3. Design

Ratshad freeaccesstotheexperimentaldietsandtapwa- terfor4weeks.Feedwasgivenasapowderformulaandwas containedinceramicbowlsthatweretooheavyfortherats toknockover,andratswereprovidedwithnewlythawedfeed everydayexceptSundays(ratsweregivendoubleportionson Saturdays).Ratsalwayshadaccesstowoodenblocksforchew- ingandplasticcageinsertsforhousing.Theratswereweighed everyseventhdayduringtheinterventionperiod.

Theratswerehousedseparatelyincageswithgridsfor24h onday18oftheinterventionperiodformeasurementsofwa- terintakeandcollectionofurine,withoutfastinginadvance.

Urinewasstoredat−80°Cuntilanalysis.

Systolic anddiastolicblood pressureswere measuredin consciousratsatbaseline(Day0,beforeratswereintroduced totheexperimentaldiets)and3daysbeforeendpoint(Day 26)usingthetail-cuffmethod(CODA-6,KentScientificCorpo- ration,Torrington,CT,USA).Ratswerepre-warmedinaheat- ingcabinetfor30minat32°Cbeforebloodpressuremeasure- ments.

A meal tolerancetest witha standardizedhigh-sucrose mealwasconductedonday22oftheinterventionperiod,with

thefollowingingredientsperkilodiet:400gsucrose,217gca- sein,70gofsoybeanoil,390gofsucrose,210gofmaizestarch, 50gofcellulose,35gofmineralmix,10gofvitaminmix,3g ofL-cystine,1.6gofL-methionine,2.5gofcholinebitartrate, 10gofgrowthandmaintenancesupplement,and0.014gof tert-butylhydroquinone.Theratswerehousedseparatelyun- derfastingconditionsfrom20.00to08.00hourswithfreeac- cesstotapwater,beforetheyreceivedamealcorrespondingto 2gofsucrose/kgbodyweight.Allratswerereceivedthesame dietforthemealtolerancetest,andtheywereallowedamax- imumof15mintofinishthemeal.Thedorsaltailveinwas punctuatedandbloodglucosewasmeasuredusingtheCon- tourbloodglucosemeasuringdevice(BayerConsumerCare AG)inthefastingconditionand60and120minaftertherats hadfinishedeatingthemeal.

Therats were euthanizedafter four weeksintervention whileunderanesthesia withisoflurane(Isoba vet,Intervet, Schering-PloughAnimalHealth,Boxmeer,TheNetherlands) mixedwithoxygenandnitrousoxide,aftera12hfastwith free access totap water. Blood was drawn from the heart andwascollectedinVacuetteZSerumClotActivatorTubes (GreinerBio-one)andVacuetteK2EDTA(GreinerBio-One)for isolationofserum andplasma, respectively.Theheart was removed immediately after collection of blood. Liver, kid- neysandepididymalwhiteadiposetissueweredissectedand weighed.Serum,plasmaandorganswerefrozeninliquidni- trogenandwerestoredat−80°Cuntilanalysis.

Assessorsresponsibleforfeeding,weighing,generaldaily animalcare,euthanasiaandanalysesofsampleswereblinded todietgroups.

2.4. Analysesofserum,plasma,urineandliver

Serumconcentrationsofcreatinine,totalprotein,carbamide and uric acid were analyzed by accredited methods using theCobasc702system(RocheDiagnosticsGmbH,Mannheim, Germany) by the Department ofMedical Biochemistry and Pharmacology, HaukelandUniversity Hospital(Bergen, Nor- way).Serumalaninetransaminase andaspartatetransami- nase(measuredwithpyridoxalphosphateactivation),plasma ammonium,andurineconcentrationsofcreatinine,totalpro- tein,carbamideanduricacidwereanalyzedontheCobasc111 system(RocheDiagnostics)usingtheALTL(Alanineamino- transferase acc. IFCC), ASTL (Aspartate aminotransferase), NH3L (Ammonia), CREP2 (Creatinine plus ver.2), TP2 (Total ProteinGen.2monochromatic),UREAL(Urea/BUN),andUA2 (Uric Acid ver.2) kits from Roche Diagnostics. Lipids were extracted from liverby themethod ofBlighand Dyer [34], evaporated undernitrogen and re-dissolved inisopropanol beforetriacylglycerolwasquantified usingtheTRIGLkiton Cobasc111(RocheDiagnostics).Theseanalyteswereanalyzed spectrophotometrically,accordingtothefollowingtestprinci- ples:creatinine,enzymaticcolorimetricmethod;totalprotein, colorimetric assay; carbamide,kinetictestwithurease and glutamatedehydrogenase;uricacid,enzymaticcolorimetric method;thecatalyticactivitiesofalaninetransaminaseand aspartametransaminaseweredeterminedbymeasurements oftherateofoxidationofnicotinamideadeninedinucleotide;

ammonium,enzymaticmethodwithglutamatedehydroge- nase;andtriacylglycerol,enzymaticcolorimetricmethod.

Urine concentration of T cell immunoglobulin mucin-1 (TIM-1) was quantified using the Rat TIM-1/KIM-1/HAVCR Quantikineenzyme-linkedimmunosorbentassay(ELISA)kit (RKM100)from R&DSystems, Bio-Techne,Minneapolis,MN, US.

SerumvitaminDwasquantifiedusingthe25-OHVitamin DTotal(Rat)ELISAkit(EIA-553;DRGInstrumentsGmbH).

Plasmaforglutathionemeasurementswasaddedice-cold 5%metaphosphoricacid(1:4),mixedandkeptonicefor15 minbeforecentrifugation,and the supernatantwas stored at−80°Cuntilanalysis.Totalglutathionewasanalyzedusing theglutathionedetectionkit(ADI-900-160)fromEnzoLifeSci- encesAG.

Aminoacids,relatedmetabolitesandpotentialbiomark- ers of fish intake (trimethylamine N-oxide (TMAO), 1- methylhistidine (π-methylhistidine), 3-methylhistidine (τ- methylhistidine) and creatinine) were measured in plasma and urine by Bevital AS (http://www.bevital.no) using gas chromatographyandhigh-performanceliquidchromatogra- phywithtandemmassspectrometry,aspreviouslydescribed [35,36].Arginine,asparticacidandglutamineweremeasured inplasma but could not bequantified in urine. Since sar- cosineis present in the VacuetteK2EDTA tubes, sarcosine couldnotbemeasuredinEDTA-plasma butwasquantified inurine.Otherwise,thesamecompoundswereanalyzedin plasmaandurine.Thiamine(vitaminB1),riboflavin(vitamin B2),nicotinamide(vitaminB3),pyridoxal5’-phosphate(vita- minB6)andrespectivevitamersaswellaskynurenine[36]and othermetabolitesinthekynureninepathway[37]wereana- lyzedinplasmabyBevitalAS.

2.5. AnalysesofmRNAgeneexpressioninliver,kidney andadiposetissue

WepurifiedtotalRNAfromliver,kidneyandepididymalwhite adipose tissue with the RNeasy Mini Kit (Qiagen, Hilden, Germany)followingthemanufacturer’s protocol.TheQIAx- pert System (Qiagen, Hilden, Germany) was used to mea- sureRNAconcentration andquality. cDNAfrom 150ngto- tal RNA per sample was synthesized by using the High- Capacity cDNA Reverse Transcription Kit (Applied Biosys- tems,CA,USA),andwasdilutedwithPCR-gradewater (1:5) before qPCR was performed (in triplicate) using the Light- Cycler480 rapid thermal cycler system (Roche Diagnostics GmbH, Basel, Switzerland) with the LightCycler 480 SYBR Green I Master (Roche, Basel, Switzerland). We used the following primer pairs: Angiotensinogen; forward primer 3’

agcacgacttcctgacttgga’5 and reverse primer 3’ttgtaggatcccc- gaatttcc’5,Renin; forwardprimer3’caaaggtttcctcagccaagat’5 andreverseprimer3’ctcggtgacctctccaaagg’5(Sigma-Aldrich).

Wetestedthreeprimerpairsasreferencegenes:60Sriboso- malproteinL32(RPL32);forwardprimer3’gtggctgccatctgttttg’5 and reverse primer 3’ttcttggtcctctttttgacg’5 (Sigma-Aldrich), 60Sacidicribosomalprotein P0(RPLP0); forwardprimer 3’gat- gcccagggaagacag’5 and reverse primer 3’gaagcattttgggtagt- catcc’5 (Sigma-Aldrich), and 18S ribosomal RNA (18S); for- wardprimer3’agtccctgccctttgtac’5andreverseprimer3’gatcc- gagggcctca’5(Eurogentec,Seraing,Belgium).Ofthese,18Shad leastvariationinreadingsandwasleastaffectedbytheinter- ventiondietsinliversamplesintriplets,whereasRPL32had

leastvariationinreadingsandwasleastaffectedbytheinter- ventiondietsofkidneyandadiposetissuesamplesintriplets.

Therefore,hepaticmRNAconcentrationsarecalculatedrela- tiveto18SrRNA,andkidneyandadiposetissuemRNAcon- centrationsarecalculatedrelativetoRPL32.

2.6. Outcomes

Theprimaryoutcomeofthisstudywastoassesstheeffectof dietaryintakeofcodfilletonconcentrationsofaminoacids andrelatedmetabolitesinplasmaandurineinobeseZucker fa/farats.Thesecondaryoutcomeswereeffectsofcodfillet intakeonmarkersofkidneyandliverfunction,thedevelop- mentofhighbloodpressure,vitaminstatus,bloodglucoseand livertriacylglycerolconcentrations.

2.7. Samplesize

Thisisthefirststudytoinvestigatetheeffectsofbakedcod filletaspartofanormaldietforgrowingrodents(AIN-93G) onconcentrationsofaminoacidsandrelatedmetabolitesin plasma and urine in obese Zucker fa/fa rats,thereforethe presentstudyisconsideredtobeapilotstudy.Weincluded10 ratsintheControlgroupand10ratsintheBakedcodgroup, basedonpreviousstudieswereweusedsalmonfilletaspart oftheAIN-93Gdietwith6ratsineachgroup[38]andrecent clinicaltrialsshowingthatcodfilletinaffectedbiochemical parameterstoalesserextentwhencomparedtosalmonfillet [18,39-41].

2.8. Statisticalanalysis

TheexperimentalgroupswerecomparedusingIndependent Samples T Test since the data were mainly normally dis- tributed(Shapiro–Wilktest).Changesinbloodpressurefrom baselinetoendpointwithineachgroupweretestedusingthe pairedsampledT-test.Thecut-offlevelforstatisticalsignifi- cancewastakenataprobabilityof0.05.SPSSStatisticsversion 25(SPSS,Inc.,IBMCompany)wereusedforallstatisticalanal- yses.Ratsfedacasein-baseddietservedascontrols.Dataare presentedasmean±standarddeviation.OneratintheCon- trolgrouphadtobeeuthanizedduetoawoundthatdidnot healandisnotincludedintheresults,thereforeweanalyzed n=10intheBakedcodgroupandn=9intheControlgroup.

3. Results

3.1. Bodyweightandgrowth

ThestartingweightwassimilarintheBakedcodgroupand theControlgroup,andbodyweightmeasuredafter1,2,3and 4weeksweresimilarbetweenthegroups(Fig.1,Pvalues0.80, 0.87,0.20,0.55and0.17,respectively).

3.2. Aminoacidsandvitaminsinplasmaandurine

Plasmaconcentrationsoftheessentialaminoacidsmethion- ineandhistidine,thenon-essentialaminoacidsalanine,as- paragine,asparticacid,glycine,prolineandtyrosine,andof

Fig.1– Bodyweightmeasuredatbaseline(week0),and after1,2,3and4weeksofinterventionforratsfedBaked coddiet(withdottedline,n=10)orControldiet(◦with solidline,n9).Valuesaremeanandstandarddeviations.P

<0.05wereconsideredstatisticallysignificant.No differenceswereobservedbetweenthegroupsatanytime duringtheinterventionperiod.Groupsarecomparedusing IndependentSamplesTTestassumingequalvariances

homoarginine,totalhomocysteine,α-ketoglutaricacid,nicoti- namideandmethylmalonicacidwerelowerintheBakedcod groupcomparedtotheControlgroup(Table3).Theplasma concentrationsofthepotentialbiomarkersoffishintake,i.e., 1-methylhistidine,3-methylhistidine,creatineandTMAO[18], werehigherintheBakedcodgroupcomparedtotheControl group.Thegroupsweresimilarwithregardtocirculatingcon- centrationsofallmeasuredfat-andwater-solublevitamins (Table4).

Inurine,the concentrations(relativetocreatinine) of1- methylhistidine,3-methylhistidine,creatineandTMAOwere higherintheBakedcodgroupcomparedtotheControlgroup, butotherwisenodifferenceswereseenbetweenthegroups forconcentrationsofaminoacidsandrelatedmetabolitesin urine(Table5).

3.3. Organweightsandwaterintake

WehavepreviouslypublishedthatBakedcodgroupandthe Control groupwere similar with regard to body weight-to- squarebodylengthandweightsofselectedwhiteadiposetis- suedepots,liverandthighmusclerelativetobodyweightat endpoint,andthattheaveragedailyenergyintakewassimilar inthetwogroups[25].Nowweshowthatalsokidneyweight andwaterintakerelativetobodyweightweresimilarbetween thegroups(Table6).

3.4. Bloodpressure,bloodglucoseandlivertriacylglycerol

Systolicanddiastolicbloodpressuresweresimilarbetween the groupsatbaseline (Table 6).Blood pressures increased in both groups from baselineto 4 weeks; P values for in- crease in systolic blood pressure were 0.0062and 0.015 in

Table3– Plasmaconcentrationsofaminoacidsandre- latedmetabolitesatendpoint

μmol/l Bakedcod Control p

group group

Alanine 320±43 389±82 0.032

Arginine 21.8±25.2 17.2±21.1 0.68

Asparagine 50.8±3.3 60.5±8.8 0.0046 Asparticacid 16.1±3.9 22.5±7.1 0.026

Betaine 80.5±10.9 72.1±7.3 0.070

Choline 17.6±2.4 18.0±3.9 0.78

Creatine 192±39 116±47 0.0012

Cystathionine 0.37±0.03 0.42±0.11 0.19 Cysteine(total) 201±27 188±22 0.26 Asymmetric

dimethylarginine

0.69±0.06 0.65±0.05 0.11

Symmetric dimethylarginine

0.37±0.04 0.36±0.03 0.53

Dimethylglycine 9.10±1.17 8.38±1.95 0.34

Glutamine 488±48 555±102 0.078

Glutamicacid 99.5±13.9 119.1±36.9 0.14

α-Ketoglutaricacid 37.6±6.1 45.2±8.9 0.042

Glutathione(total) 34.7±11.5 42.7±17.0 0.24

Glycine 108±10 127±17 0.011

Histidine 53.6±4.6 61.6±5.1 0.0024

1-Methylhistidine 15.2±4.3 5.6±0.8 5.2×10^-6 3-Methylhistidine 4.09±0.47 2.76±0.17 3.9×10⁻⁷ Homoarginine 0.90±0.27 1.18±0.20 0.019 Homocysteine(total) 2.13±0.21 2.56±0.38 0.0065 3-Hydroxyisobutyrate 20.2±4.8 23.2±5.2 0.22 Isoleucine 83.2±11.1 87.2±11.0 0.44

Leucine 134±20 142±22 0.44

Lysine 290±42 302±30 0.50

Methionine 41.8±5.8 48.5±5.9 0.024

Methioninesulfoxide 1.18±0.54 1.58±0.58 0.14 Methylmalonicacid 0.29±0.04 0.36±0.05 0.0061

Ornithine 125±30 142±28 0.24

Phenylalanine 67.2±4.2 71.8±9.0 0.16

Proline 105±18 130±27 0.029

Sarcosine ND ND (-)

Serine 181±12 191±20 0.19

Threonine 183±26 206±29 0.090

TMAO 13.23±16.24 1.56±0.44 0.00017

Trimethyllysine 1.18±0.09 1.15±0.12 0.61 Tryptophan 94.4±12.7 100.5±11.1 0.28

Tyrosine 56.3±8.3 83.9±17.2 0.00029

Valine 192±34 208±28 0.28

Dataarepresentedasmeans±standarddeviation;n=10ratsin BakedCodGroupandn=9ratsinControlGroup.P<0.05werecon- sideredsignificant.GroupsarecomparedusingIndependentSam- plesTTestassumingequalvariances.TMAO;TrimethylamineN- oxide

theBakedcodgroupand Controlgroup,respectively,andP values forincrease indiastolic blood pressure were 0.0046 and 0.0060 in the Bakedcod group and Control group, re- spectively (datanotpresented).Thewithin-groups changes in systolic and diastolic blood pressures from baseline to end pointwere notsignificantlydifferentwhen thegroups werecompared(Pvalues0.91and0.93,respectively,datanot presented), and no differences were seen between groups for systolic and diastolic blood pressures at end point (Table6).

Table4 – Circulatingconcentrations offat-soluble vita- mins,Bvitamins andkynureninepathwaymetabolites atendpoint

Bakedcod Control P

group group

Fatsolublevitamins

all-transRetinol,μmol/l 1.68±0.30 1.91±0.27 0.10 25-OHVitaminD(total),

ng/mL

24.0±6.4 26.1±5.7 0.47

α-Tocopherol,μmol/l 94.0±10.1 93.3±9.7 0.88

gamma-Tocopherol,μmol/l 3.58±1.36 3.80±0.94 0.69 Bvitamins

Thiamine,nmol/l 165±22 192±52 0.15

Thiaminemonophosphate, nmol/l

677±146 672±98 0.94

Riboflavin,nmol/l 77.4±13.2 92.4±21.0 0.077 Flavinmononucleotide,

nmol/l

56.6±13.6 53.5±11.6 0.60

Nicotinicacid,nmol/l <LOD <LOD

Nicotinamide,nmol/l 3743±463 4416±851 0.044 N¹-Methylnicotinamide,

nmol/l

383±119 417±215 0.67

Pyridoxal5-phosphate, nmol/l

1222±115 1090±268 0.17

Pyridoxal,nmol/l 437±79 434±76 0.94 4-Pyridoxicacid,nmol/l 66.5±15.7 67.8±14.7 0.86 Pyridoxine,nmol/l <LOD <LOD

Kynureninepathway metabolites

Kynurenine,μmol/l 1.72±0.24 1.64±0.39 0.59 Kynurenicacid,nmol/l 82.9±13.5 84.9±18.5 0.79 Anthranilicacid,nmol/l 84.3±22.0 99.8±23.2 0.15 3-Hydroxykynurenine,

nmol/l

5.42±2.98 4.95±1.92 0.69

Xanthurenicacid,nmol/l 14.9±3.3 13.9±2.6 0.47 3-Hydroxyanthranilicacid,

nmol/l

3.83±0.95 4.55±0.97 0.12

Picolinicacid,nmol/l 113±26 110±15 0.74 Quinolinicacid,nmol/l 279±79 302±119 0.62

Dataarepresentedasmeans±standarddeviation;n=10ratsin BakedCodGroupandn=9ratsinControlGroup.P<0.05werecon- sideredsignificant.GroupsarecomparedusingIndependentSam- plesTTestassumingequalvariances.

LOD,levelofdetection(LODswere1nmol/Lforpyridoxineand20 nmol/Lfornicotinicacid)

Geneexpressionsofangiotensinogeninliverandofreninin kidneyandepididymalwhiteadiposetissueweresimilarbe- tweenthedietarygroups(Pvalues=0.19,0.37and0.75,re- spectively,datanotpresented).SincemRNAexpressionofan- giotensinogenisreduced byfastinginwhite adiposetissue (butnotinliver)[42],angiotensinogenmRNAexpressioninadi- posetissuewasnotmeasuredinourratsastheywerefasted overnight.

After22days,thefastingbloodglucoseconcentrationwas similarbetweenthegroups(Table6).Thepostprandialblood glucose concentrations after intake ofa high-sucrose feed werealsosimilarbetweenthegroups,measuredafter60and 120minutes(Pvalues=0.35and0.74,respectively).Livertria- cylglycerolconcentrationwassimilarintheBakedcodgroup andtheControlgroup.

Table5– Urineconcentrationsofaminoacidsandrelated metabolitesatendpoint

μmolpermmol Bakedcod Control P

creatinine group group

Alanine 56.5±14.9 63.3±44.9 0.66

Asparagine 8.13±5.13 9.81±3.59 0.43

Betaine 92.7±27.0 83.1±48.5 0.60

Choline 46.8±36.2 31.8±18.6 0.28

Creatine 1560±652 31.5±33.5 2.1×10⁻⁶ Cystathionine 0.33±0.11 0.50±0.40 0.22 Cysteine(total) 65.5±30.3 55.4±35.4 0.51 Asymmetric

dimethylarginine

0.12±0.10 0.23±0.49 0.50

Symmetric dimethylarginine

4.28±0.97 4.30±3.11 0.99

Dimethylglycine 67.1±22.7 64.1±37.8 0.83 Glutamicacid 26.4±10.9 30.2±19.0 0.60

α-Ketoglutaricacid 487±176 354±337 0.29

Glycine 57.5±11.1 87.7±96.8 0.34

Histidine 16.4±4.0 19.2±10.0 0.43

1-Methylhistidine 57.02±38.13 7.64±7.48 1.4×10⁻³ 3-Methylhistidine 36.60±16.63 9.97±8.62 4.8×10⁻⁴ Homoarginine 22.9±7.2 33.3±19.0 0.13 Homocysteine(total) 2.74±1.16 2.79±2.70 0.96 3-Hydroxyisobutyrate 7.3±2.6 8.8±8.9 0.62 Isoleucine 8.44±5.39 8.27±3.25 0.94 Kynurenine 0.09±0.04 0.23±0.36 0.26

Leucine 14.9±7.6 19.3±6.8 0.20

Lysine 41.3±8.5 44.2±12.0 0.55

Methionine 25.0±25.4 16.8±8.2 0.37 Methioninesulfoxide 2.43±1.41 2.90±1.73 0.52 Methylmalonicacid 10.97±2.34 13.44±8.37 0.38 Ornithine 5.30±1.72 7.15±5.07 0.29 Phenylalanine 11.2±4.3 15.9±13.2 0.30

Proline 33.1±10.0 48.6±55.1 0.39

Sarcosine 3.81±1.38 3.39±1.90 0.59

Serine 19.9±11.1 21.9±10.3 0.68

Threonine 48.4±55.1 45.3±17.6 0.65

TMAO 2534±1106 69.8±52.6 4.0×10⁻⁶

Trimethyllysine 8.23±3.47 6.61±9.09 0.61 Tryptophan 2.89±1.86 4.66±4.80 0.29

Tyrosine 10.9±3.8 18.8±19.0 0.22

Valine 20.3±7.0 23.9±9.5 0.36

Dataarepresentedasmeans±standarddeviation;n=10ratsin BakedCodGroupandn=9ratsinControlGroup.p<0.05werecon- sideredsignificant.GroupsarecomparedusingIndependentSam- plesTTestassumingequalvariances.TMAO;TrimethylamineN- oxide

3.5. Markersoforgandamageandkidneyfunction,and nitrogen-containingcompounds

Rats fedbaked codhad significantlylower circulating con- centrationsofalaninetransaminase,aspartatetransaminase creatinine,carbamide,uricacidandammoniumwhencom- paredtotheControlgroup(Table7).Theurineconcentration (relativetocreatinine)ofTIM-1wassignificantlylowerinthe Bakedcodgroup,whereastheurinecreatinineconcentration andconcentrations(relativetocreatinine)oftotalprotein,car- bamide,uricacidandtotalamountofproteinogenicamino acids(argininewasnotmeasured)weresimilarbetweenthe twogroups.

Table6– Kidneyweight,waterintake,bloodglucose,liver triacylglycerolandbloodpressure

Bakedcod Control P

group group

Baselinemeasurements Systolicbloodpressure,

mmHg

131±14 136±15 0.49

Diastolicbloodpressure, mmHg

92±14 97±12 0.45

Endpointmeasurements Kidneys,g/kg

bodyweight

4.7±0.5 5.1±0.6 0.21

Waterintake,g/kg bodyweight/24h

47.2±13.5 42.6±13.1 0.46

Fastingbloodglucose, mmol/l

4.8±0.7 4.9±1.3 0.82

Livertriacylglycerols, μmol/gliver

162±39 170±55 0.70

Systolicbloodpressure, mmHg

149±9 150±12 0.75

Diastolicbloodpressure, mmHg

108±5 110±10 0.55

Dataarepresentedasmeans±standarddeviation;n=10ratsin BakedCodGroupandn=9ratsinControlGroup.p<0.05werecon- sideredsignificant.GroupsarecomparedusingIndependentSam- plesTTestassumingequalvariances.

4. Discussion

Hereweshowforthefirsttimethatfeedingratsadietcon- tainingbaked codfillet resultedinlower concentrationsof severalaminoacidsandrelatedmetabolitesinplasmawhen comparedtoacontrolgroupfedafish-freediet,butdidnot affectaminoacidcompositioninurine.Theconcentrations of1-methylhistidine,3-methylhistidine,creatineandTMAO, whichareproposedaspotentialbiomarkersoffishintake[18], were higherinbothplasmaand urine fromrats fedbaked cod.TheBakedcoddietcontainedmore(thatis,>1g/kgdiet difference)alanineandasparticacid+asparagine,andless glutamicacid+glutamineandproline,allofwhicharenon- essentialaminoacids,whencomparedwiththeControldiet, with no differences between the diets foressential amino acids.DifferencesintheaminoacidcompositionoftheBaked coddietandtheControldietwere,ingeneral,notreflectedin theplasmaandurineconcentrationsofaminoacidsinrats fedthesediets.Concentrationsofmarkersofkidneydysfunc- tion andliverdamagemeasured inserumorurine suggest thatthefunctionofbothorganswerebetterintheBakedcod group.Vitaminstatusisaffectedbykidneystatus[43],butno differenceswereseenbetweenthegroupsforcirculatingcon- centrationsofwater-andfat-solublevitamins.Groupswere similarwithregardtogrowth,blood pressuredevelopment, fastingandpostprandialglucoseconcentrations,andlivertri- acylglycerolconcentration.

Patients withchronickidney failure are advised tocon- sumeadietwithmodestproteinrestrictioninordertolimit the development of toxic nitrogenous metabolites, uremic symptomsand othermetaboliccomplications [44,45].How- ever,informationislackinginregardtowhetherdifferentdi- etaryproteinsmay havedissimilarimpact onkidney func-

Table7– Markersoforgandamageandkidneyfunction, andnitrogen-containingcompoundsmeasuredinserum, plasmaorurineatendpoint

Bakedcod Control P

group group

Serumcreatinine,μmol/l 12.3±0.7 13.8±1.2 0.0038 Serumalanine

transaminase,U/l

99±29 147±59 0.034

Serumaspartate transaminase,U/l

185±54 301±149 0.034

Serumprotein,g/l 60.1±2.6 61.8±3.3 0.23 Serumcarbamide,mmol/l 5.9±1.3 7.3±1.5 0.047 Serumuricacid,μmol/l 73.2±25.6 106.6±26.9 0.013 Plasmaammoniumμmol/l 89.7±21.5 114.8±16.0 0.011 Urinecreatinine,mmol/l 5.4±2.7 5.7±2.7 0.83 UrineTIM-1,ng/mmol

creatinine

192±24 266±72 0.0066

Urinetotalprotein,g/mmol creatinine

1.9±0.6 1.8±1.4 0.90

Urinecarbamide, mmol/mmolcreatinine

378±126 446±56 0.15

Urineuricacid,μmol/mmol creatinine

368±110 335±116 0.54

Urinetotalproteinogenic aminoacids,μmol/mmol creatinine

467±127 533±295 0.52

Dataarepresentedasmeans±standarddeviation;n=10ratsin BakedCodGroupandn=9ratsinControlGroup.P<0.05werecon- sideredsignificant.GroupsarecomparedusingIndependentSam- plesTTestassumingequalvariances.TIM-1;Tcellimmunoglobulin mucin-1

tion,anditisofinterestthatintakeoffishhasbeenassoci- atedwithreduced riskofdevelopingkidneydisease[46,47]. WeusedobeseZuckerfa/faratsaged8–9weeksatthestart oftheintervention(ratswereeuthanizedatage12-13weeks), sincetheydevelophighbloodpressureandproteinuriabefore oraround10weeksofage[21-24].Theoccurrenceofproteins and aminoacids inurineare amongtheearliestindicators ofrenaldysfunctioninbothhumansandanimals[48,49].Pro- teinsandfreeproteinogenicaminoacidswereobservedinthe urineofallrats,withnodifferenceinconcentrationsbetween theBakedcodgroupandtheControlgroup,thusindicatingre- naldysfunctioninbothgroups.Therelativeurineproteincon- centrationinthepresentstudywasaroundthreetimeshigher thanpreviouslyobservedinyounghealthymaleWistarrats [50].AbetterurinemarkerfortubularinjuryisTIM-1,which isexpressedontheproximaltubuleapicalmembraneinre- sponsetorenalinjurybutisnotdetectableinurinewhenkid- neysarehealthy[51].ThepresenceofTIM-1inurinefromall ratsalsoindicatestheoccurrenceoftubularinjuryinbothex- perimentalgroups;however,thelowerTIM-1concentrationin urinefromratsintheBakedcodgroupsuggestthatconsump- tionofcodfilletmaysomehowprotectagainstordelaythede- velopmentofkidneyinjuryinobeseZuckerfa/farats.Other findingssupportthisassumption;thelower serumconcen- trationsofcreatinineandcarbamideintheBakedcodgroup aresuggestiveofalessimpairedkidneyfunction[51,52]com- paredtotheControlgroup.Thisisinlinewithourprevious reportontheeffectsofcodproteinintakeonurinarymarkers

ofrenaldysfunctioninobeseZuckerfa/farats[53],indicating aprotectiveeffectofcodproteinsonkidneyfunction.

RatsfedtheBakedcoddiethadlowerglycineplasmacon- centrationwhencomparedtotheControlgroup,howeverno differenceswere seen betweenthe groupsforplasmacon- centrationsofbranchedchainaminoacidsandkynurenines, andofthearomaticaminoacidsonlytyrosineconcentration waslowerwhencomparedtotheControlgroup.Thelower plasmaconcentrationsofglycineand tyrosineareofinter- est asthese are associatedwith prediabetes, insulinresis- tance,cardiovascular disease and futuretype2diabetes in humans[14,15].Plasmaconcentrationsofthenon-essential aminoacids alanine,asparagine, asparticacid and proline, andofα-ketoglutaricacid(theketonederivateofglutaricacid) werelowerinratsfedtheBakedcoddiet.Alaninetransami- naseandaspartatetransaminasecatalyzetheconversionbe- tweenα-ketoglutaricacidandeitheralanineorasparticacid, respectively, and glutamic acid and either pyruvateor ox- aloacetate,andthelowerconcentrationsofalanine,aspartic acidandα-ketoglutaricacidintheBakedcodgroupmaybea consequenceofdifferentactivityratesofthesetransaminases betweenthegroups.Thelowerhistidineandprolineconcen- trationsmaybecausedbyincreasedfluxintothetricarboxylic cycle,and inadditionthe lower prolineconcentrationmay suggestthatitsformationfromglutamateisdownregulated.

Inthis contextitisofinterestthatfattyliverdisease isas- sociatedwithhigherplasmaconcentrationsofseveralamino acidsincluding alanine,histidine,methionine,tyrosineand proline[54].

ThelowertotalhomocysteineconcentrationintheBaked codgroupmaybebeneficialsinceelevatedhomocysteinecon- centrationisassociatedwithincreasedriskfordevelopingcar- diovasculardiseaseinhumans[55].Inaddition,elevatedho- mocysteineconcentrationisprevalentinpatientswithacute andchronicrenaldisease, underscoringthe keyrole ofthe kidneysinremovinghomocysteinefromthecirculation[56]. Thelowertotalhomocysteineconcentrationtogetherwiththe lowermethionineconcentrationinplasmamayindicatethat one-carbonmetabolismwasaffectedbytheBakedcoddiet,as plasmaconcentrationsofhomocysteineandmethionineare showntobepositivelycorrelated[57].Plasmaconcentrations ofother metabolitesrelatedtotheone-carbonmetabolism, suchas the methyl donorsbetaine, choline and dimethyl- glycine,werenotdifferentbetweenthedietarygroups.Thus, thelowertotalhomocysteineconcentrationmaybeaconse- quenceoflowerplasmaconcentrationoftheprecursorme- thionine,despitesimilarmethioninecontentintheBakedcod dietandintheControldiet.

The plasma concentrations of arginine, asymmetric dimethylarginine, symmetric dimethylarginine and ho- moarginineare ofinterestinthe contextofbloodpressure regulation.Nitricoxideisapotentvasodilator,andissynthe- sized from L-arginine byendothelial nitric oxide synthase.

Asymmetric dimethylarginine (produced by methylation ofarginineresidues ofthe intracellularproteins) is anen- dogenouscompetitiveinhibitorofnitricoxidesynthase[58], symmetricdimethylarginineinhibitsthetransportofarginine [59]andhomoarginineisacompetitivenitricoxidesynthase substrate[60],thusallofthesecanacttodecreasenitricoxide productionfrom L-arginine.Ofthesefour compounds,only

homoarginineconcentrationwasaffectedbyBakedcoddiet, with alower plasmaconcentration whencompared tothe Controlgroup.However,bloodpressureisalsoinfluencedby therenin-angiotensinsystemwhereangiotensinogen(mainly producedbytheliver)iscleavedbyrenin(primarilyproduced bykidneys)tothebiologicallyinactiveangiotensin I,which isfurtherconvertedtothevasoconstrictorangiotensinIIby the angiotensin-converting enzyme[61], with renin as the rate determining enzyme [62]. Adipose tissues in humans androdentscontainallcomponentoftherenin-angiotensin system,andtheactivityofthissystemisincreasedinobesity [61].Therenin-angiotensinsystemisimportantfortheblood pressureincreaseinobeseZuckerfa/farats,possiblythrough increasedsensitivitytowardsangiotensinII[63].Weobserved nodifferencesbetweentheBakedcodgroupandtheControl group with regard to blood pressure development or gene expressionsofangiotensinogeninliverorofrenininkidneyand inadiposetissue,andthekidneyweightwassimilarbetween the groups.These resultsareincoherence withourrecent findings where a diet containing cod fillet did not affect bloodpressuredevelopmentornitrite+nitrateconcentration in serum from obese Zucker fa/fa rats [64]. From this we conclude thatbakedcodfillet didnothavethepotentialto preventor delaythetypicalblood pressureincreasethat is expectedinobeseZuckerfa/farats.

Methylhistidines from dietary anserine and cod muscle proteinsare notre-utilizedforproteinsynthesisor metab- olized but are excreted in the urine, and the concentra- tions of methylhistidines in urine have been proposed as a useful biomarker of meat intake [65]. Also, the major- ity of TMAO from diet is excreted unchanged by the kid- neys[66].Thehigherconcentrationsof1-methylhistidine,3- methylhistidine,creatineandTMAOinplasmaandurineof ratsfedtheBakedcoddietwereexpectedsincecodmuscle containsrelativelyhighlevelsofthesefourcompounds[18], andisinlinewithourrecentreportshowinghighurinecon- centrationsofthesecompoundsinobeseZuckerfa/faratsfed codprotein[53].

Consumptionofbakedcoddidnotaffecttheconcentra- tionsofindividualaminoacidsandthetotalamountofpro- teinogenicaminoacidsinurine.Thiswassomewhatsurpris- ing,sincewepreviouslyhaveshownthatZuckerfa/faratsfed codproteinhadamuchlowerurineconcentrationofalmost all measuredaminoacidsandalowertotalamountofpro- teinogenic aminoacidswhencomparedtothose fedacon- troldietdevoidoffishprotein[53].Inaddition,circulatingand urinemarkersindicatedabetterkidneyfunctioninratsfed codprotein[53],thusthelowerurinesecretioninthoserats maybeaconsequenceofbetterkidneyfunction.Resultsfrom the presentstudysuggestthatalthoughtheBakedcoddiet seemstooffersomeprotectionofthekidneys,thiswasnot sufficienttoaffecttheamountofaminoacidsthatwasex- cretedinurine.

Thenutritionaladvicefornon-alcoholicfattyliverdisease patientsistoachieveweightlossthroughahypocaloricdiet but does notexplicitly address proteinintake [67],and ef- fectoffishintakeonfattyliverhasnotbeendescribed.Ala- ninetransaminaseisacirculatingmarkerforliverfunction, whereasaspartatetransaminaseisregardedasamoregeneral markerfororgandamage.Thelowerconcentrationofserum

alanine transaminase in rats fedbaked codindicates that theliverfunctionwasbettercomparedtotheControlgroup, althoughthehepatic triacylglycerolcontentwassimilarbe- tweenthegroups.Also,thelowerserumaspartatetransami- naseconcentrationmaysupportthelowerconcentrationsof markersforkidneydysfunctionintheBakedcodgroup.These findingsindicatesaprotectiveeffectofcodfilletonliverfunc- tion,andareinlinewiththeobservationthatobeseZucker fa/faratsfedsalmonfillethadlowerserumconcentrationsof bothtransaminaseswhencomparedtoratsfedafish-freediet butdidnotaffecthepatictriacylglycerolconcentration[68].

TheratsintheBakedcodgroupwere feddietscontain- ing25%ofproteinsfrombakedcodfilletand75%ofproteins fromcasein.Thisintakeoffishproteinsishigherthantherec- ommendationsfromtheAmericanHeartAssociationandthe Norwegiangovernment,whorecommendafishintakeof300- 450g/week(2-3dinnermealsperweek)forthegeneralpub- lic[29,30].Thedailyaveragetotalproteinintakeforadultsin theUSandinNorwayisestimatedtobe91-96g[31,32],thus theserecommendationscorrespondstoapproximately60-90g offishproteinsweekly,amountingto9-13%ofthetotalpro- teinintake.Thecodproteinintakeinthepresentstudycorre- spondswellwithapreviousclinicalstudywhereadultswith overweight/obesityconsumed750gperweekofcodfilletor salmonfillet;participantshadanestimatedmedianintakeof 96gtotalprotein/daywith22%ofproteinintakeasfishpro- teins[33].Byusing75%ofproteinsfromcaseinintheBaked coddiet,and100%ofproteinsfromcaseinintheControldiet, withextraaddedmethionineandcystine(1.6and3.0g/kgdiet, respectively),weensure thatthe intake ofessentialamino acidsisadequateandavoidthatapotentiallackofessential aminoacidsinfluencetheeffectsofthedietsonbiochemical andphysiologicalparameters.

Thepresentstudyhassomemethodologicalstrengthsand limitations. Strengthsinclude simultaneous measurements ofmanyaminoacidsandrelatedmetabolites.Sampleswere treatedaccordingtoastrictprotocolforpre-analyticalsample handling,andsampleswerethawedforthefirsttimeforthese analyses. The measurements of blood pressure were con- ductedinconscious(unanesthetized)ratsusingthetail-cuff method(volume-pressurerecording),whichisanon-invasive andinexpensivemethodthatdoesnotrequiresurgery.Rats werehand-tameandtrainedtobeintheconstrainerbefore thestartoftheintervention.Limitationstothestudyincludes theratmodelused;althoughtheobeseZuckerfa/faratisa relevantmodelofhumanobesitythatspontaneouslydevel- opscomorbiditiesofobesity[20],thetranslationofthepresent findingstohumansmustbefurtherinvestigatedsincethere arestillimportantdifferencesbetweenhumansandrats,and thefindingsinthecurrentstudycouldbespecificforobese Zuckerfa/farat.

Wehypothesizedthatintakeofbakedcodfilletwouldaf- fectaminoacidconcentrationsinplasmaandurineinobese Zucker fa/fa rats. Partially inline withour hypothesis, we foundthatratsintheBakedcodgrouphadlowerplasmacon- centrationsoftotalhomocysteine,homoarginine,theessen- tialaminoacidsmethionineandhistidine,thenon-essential aminoacidsalanine,asparagine,asparticacid,glycine,pro- lineandtyrosine.Regardingparametersassociatedwiththe metabolic syndrome, we found that markers of renal and

liverfunctionwerelessimpairedinratsfedbakedcod,while wefoundnodifferencesbetweenthegroupsforbloodpres- suredevelopment,fastingandpostprandialglucoseandliver triacylglycerolconcentrations.Toconclude,substituting25%

ofdietaryproteinswithlyophilizedbakedcodfilletaffected plasmaconcentrationsofseveralaminoacidsanddelayedthe developmentofkidneydysfunctionandliverdamage,butdid notaffecturineaminoacidcomposition,bloodpressuredevel- opment,vitaminstatus,bloodglucoseandlivertriacylglycerol concentrationswhencomparedtoadietwith100%ofproteins fromcasein.

Author statement

LAV,AD,GMandOAGformulatedtheresearchquestionand designedthestudy.LAV,AD,MTBandOAGconductedthean- imalstudy.LAV,AD,ØM,AM,MHA,PMUandOAGanalyzed thedata.OAGperformedstatisticalanalyses,draftedthepa- perandhadprimaryresponsibilityforthefinalcontent.All authorshavecontributedtothewritingandapprovedthefi- nalmanuscript.

Acknowledgments

None.

Sources of Support

ThisresearchwassupportedbyTheNorwegianSeafoodRe- searchFund(FHF,grantnumber900842).Thesponsorwasnot involvedinthedesignofthestudy,datacollection,analysis andinterpretationofdata,writingofthearticleorinthede- cisiontosubmitthearticleforpublication.

Author Declarations

Theauthorsdeclarenoconflictsofinterest.

r e f e r e n c e s

[1] GreggEW, ShawJE.Globalhealtheffectsofoverweightand obesity.NEnglJMed2017;377:80–1.

[2] MokdadAH, FordES, BowmanBA, DietzWH, VinicorF, BalesVS, etal. Prevalenceofobesity,diabetes,and

obesity-relatedhealthriskfactors,2001.JAMA2003;289:76–9. [3] Globalburdenofmetabolicriskfactorsforchronicdiseases

collaboration.cardiovasculardisease,chronickidneydisease, anddiabetesmortalityburdenofcardiometabolicrisk factorsfrom1980to2010:acomparativeriskassessment.

thelancet.Diabetes&endocrinology2014;2:634–47.

[4] CarrMC, BrunzellJD.Abdominalobesityanddyslipidemiain themetabolicsyndrome:importanceoftype2diabetesand familialcombinedhyperlipidemiaincoronaryarterydisease risk.JClinEndocrinolMetab2004;89:2601–7.

[5] GrundySM, CleemanJI, DanielsSR, DonatoKA, EckelRH, FranklinBA, etal. Diagnosisandmanagementofthe metabolicsyndrome:anAmericanHeart

Association/NationalHeart,lung,andbloodinstitute scientificstatement.Circulation2005;112:2735–52. [6]VirtanenJK, MozaffarianD, ChiuveSE, RimmEB.Fish

consumptionandriskofmajorchronicdiseaseinmen.AmJ ClinNutr2008;88:1618–25.

[7]ZhengJ, HuangT, YuY, HuX, YangB, LiD.Fishconsumption andCHDmortality:anupdatedmeta-analysisofseventeen cohortstudies.PublicHealthNutr2012;15:725–37.

[8]DjousseL, AkinkuolieAO, WuJH, DingEL, GazianoJM.Fish consumption,omega-3fattyacidsandriskofheartfailure:a meta-analysis.ClinNutr2012;31:846–53.

[9]LeungYinkoSS, StarkKD, ThanassoulisG, PiloteL.Fish consumptionandacutecoronarysyndrome:a

meta-analysis.AmJMed2014;127:848–57.

[10]KromannN, GreenA.Epidemiologicalstudiesinthe Upernavikdistrict,Greenland.Incidenceofsomechronic diseases1950-1974.ActaMedScand1980;208:401–6. [11]FeskensEJ, BowlesCH, KromhoutD.Inverseassociation

betweenfishintakeandriskofglucoseintolerancein normoglycemicelderlymenandwomen.DiabetesCare 1991;14:935–41.

[12]NkondjockA, ReceveurO.Fish-seafoodconsumption, obesity,andriskoftype2diabetes:anecologicalstudy.

DiabetesMetab2003;29:635–42.

[13]KarlssonT, Rosendahl-RiiseH, DierkesJ, DrevonCA, TellGS, NygardO.Associationsbetweenfishintakeandthe metabolicsyndromeanditscomponentsamong

middle-agedmenandwomen:theHordalandHealthStudy.

Foodnutrires2017;61:1347479.

[14]GarC, RottenkolberM, PrehnC, AdamskiJ, SeisslerJ, LechnerA.Serumandplasmaaminoacidsasmarkersof prediabetes,insulinresistance,andincidentdiabetes.Crit revclinlabsci2018;55:21–32.

[15]NagaoK, YamakadoM.Theroleofaminoacidprofilesin diabetesriskassessment.CurrOpinClinNutrMetabCare 2016;19:328–35.

[16]FavennecM, HennartB, CaiazzoR, LeloireA, YengoL, VerbanckM, etal. Thekynureninepathwayisactivatedin humanobesityandshiftedtowardkynurenine

monooxygenaseactivation.Obesity(SilverSpring) 2015;23:2066–74.

[17]KarlssonT, StrandE, DierkesJ, DrevonCA, OyenJ, MidttunO, etal. Associationsbetweenintakeoffishandn-3long-chain polyunsaturatedfattyacidsandplasmametabolitesrelated tothekynureninepathwayinpatientswithcoronaryartery disease.EurJNutr2017;56:261–72.

[18]HagenIV, HellandA, BratlieM, MidttunO, McCannA, SveierH, etal. TMAO,creatineand1-methylhistidinein serumandurinearepotentialbiomarkersofcodandsalmon intake:arandomisedclinicaltrialinadultswithoverweight orobesity.EurJNutr2020;59:2249–59.

[19]HellandA, BratlieM, HagenIV, MidttunO, SveierH, MellgrenG, etal. Effectofhighintakeofcodorsalmonon serumtotalneopterinconcentration:arandomisedclinical trial.EurJNutr2021.

[20]deArtinanoAA, CastroMM.Experimentalratmodelsto studythemetabolicsyndrome.BrJNutr2009;102:1246–53. [21]StevensonFT, WheeldonCM, GadesMD, vanGoorH,

SternJS.Hyperphagiaasamediatorofrenaldisease initiationinobeseZuckerrats.Obesres2001;9:492–9. [22]KurtzTW, MorrisRC, PershadsinghHA.TheZuckerfattyrat

asageneticmodelofobesityandhypertension.

Hypertension1989;13:896–901.

[23]LuoH, WangX, ChenC, WangJ, ZouX, LiC, etal. Oxidative stresscausesimbalanceofrenalreninangiotensinsystem (RAS)componentsandhypertensioninobeseZuckerrats.J AmHeartAssociat2015;4:e001559.

[24]LapingNJ, OlsonBA, DayJR, BricksonBM, ContinoLC,

ShortBG, etal. Theage-relatedincreaseinrenalclusterin mRNAisacceleratedinobeseZuckerrats.JAmSocNephrol 1998;9:38–45.

[25]VikorenLA, DrotningsvikA, BergsethMT, MjosSA, AustgulenMH, MellgrenG, etal. Intakeofbakedcodfillet resultedinlowerserumcholesterolandhigherlongchain n-3pufaconcentrationsinserumandtissuesin

hypercholesterolemicobeseZuckerfa/faRats.Nutrients 2018;10:840–55.

[26]ReevesPG, NielsenFH, FaheyGCJr.AIN-93purifieddietsfor laboratoryrodents:finalreportoftheAmericanInstituteof Nutritionadhocwritingcommitteeonthereformulationof theAIN-76Arodentdiet.JNutr1993;123:1939–51.

[27]ReevesPG.AIN-93purifieddietsforthestudyoftrace elementmetabolisminrodents.In:TraceElementsin LaboratoryRodents.BocaRaton,FL,USA:CRCPressInc.;

1996.p.3–37.R.R.Watson,editor.

[28]DunnMA, HartsookEW.Comparativeaminoacidand proteinmetabolisminobeseandnon-obeseZuckerrats.J Nutr1980;110:1865–79.

[29]Kris-EthertonPM, HarrisWS, AppelLJ.Fishconsumption, fishoil,omega-3fattyacids,andcardiovasculardisease.

Circulation2002;106:2747–57.

[30]TheNorwegianDirectorateofHealth(2014)

Recommendationsregardingdiet,nutritionandphysical activity[Anbefalingeromkosthold,ernæringogfysisk aktivitet].

[31]FulgoniVL,3rd.CurrentproteinintakeinAmerica:analysis oftheNationalHealthandNutritionExaminationSurvey, 2003-2004.AmJClinNutr2008;87:1554S-7S.

[32]TheNorwegianDirectorateofHealth(2012)Norkost3, NationaldietarysurveyconductedamongadultsinNorway aged18-70years[Enlandsomfattende

kostholdsundersøkelseblantmennogkvinneriNorgei alderen18-70år,2010-11].

[33]BratlieM,HagenIV,HellandA,ErchingerF,MidttunO, UelandPM,etal.Effectsofhighintakeofcodorsalmonon gutmicrobiotaprofile,faecaloutputandserum

concentrationsoflipidsandbileacidsinoverweightadults:

arandomisedclinicaltrial.EurJNutr2020inpress.

doi:10.1007/s00394-020-02417-8.

[34]BlighEG, DyerWJ.Arapidmethodoftotallipidextraction andpurification.Canjbiochemphysiol1959;37:911–17. [35]MidttunO, KvalheimG, UelandPM.High-throughput,

low-volume,multianalytequantificationofplasma metabolitesrelatedtoone-carbonmetabolismusing HPLC-MS/MS.Analbioanalchem2013;405:2009–17. [36]MidttunO, McCannA, AarsethO, KrokeideM, KvalheimG,

MeyerK, etal. Combinedmeasurementof6fat-soluble vitaminsand26water-solublefunctionalvitaminmarkers andaminoacidsin50muLofserumorplasmaby high-throughputmassspectrometry.AnalChem 2016;88:10427–36.

[37]MidttunO, HustadS, UelandPM.Quantitativeprofilingof biomarkersrelatedtoB-vitaminstatus,tryptophan metabolismandinflammationinhumanplasmabyliquid chromatography/tandemmassspectrometry.Rapid CommunMassSpectrom2009;23:1371–9.

[38]DrotningsvikA, MidttunO, VikorenLA, McCannA, UelandPM, MellgrenG, etal. Urineandplasma

concentrationsofaminoacidsandplasmavitaminstatus differ,andaredifferentlyaffectedbysalmonintake,inobese Zuckerfa/faratswithimpairedkidneyfunctionandin Long-Evansratswithhealthykidneys.BrJNutr 2019;122:262–73.

[39]HagenIV, HellandA, BratlieM, BrokstadKA, RosenlundG, SveierH, etal. Highintakeoffattyfish,butnotofleanfish, affectsserumconcentrationsofTAGandHDL-cholesterolin

healthy,normal-weightadults:arandomisedtrial.BrJNutr 2016;116:648–57.

[40]HellandA, BratlieM, HagenIV, MjosSA, SornesS, Ingvar HalstensenA, etal. Highintakeoffattyfish,butnotoflean fish,improvedpostprandialglucoseregulationand increasedthen-3PUFAcontentintheleucocytemembrane inhealthyoverweightadults:arandomisedtrial.BrJNutr 2017;117:1368–78.

[41]BratlieM, HagenIV, HellandA, MidttunO, UlvikA, RosenlundG, etal. Fivesalmondinnersperweekwerenot sufficienttopreventthereductioninserumvitaminDin autumnat60degreesnorthlatitude:arandomisedtrial.BrJ Nutr2020;123:419–27.

[42]FrederichRCJr, KahnBB, PeachMJ, FlierJS.Tissue-specific nutritionalregulationofangiotensinogeninadiposetissue.

Hypertension1992;19:339–44.

[43]SteiberAL, KoppleJD.Vitaminstatusandneedsforpeople withstages3-5chronickidneydisease.Jrennutr

2011;21:355–68.

[44]KoppleJD.NationalkidneyfoundationK/DOQIclinical practiceguidelinesfornutritioninchronicrenalfailure.Amj kidneydis2001;37:S66–70.

[45].KDOQIClinicalpracticeguidelinefornutritioninchildren withckd:2008update.executivesummary.In:Amjkidney dis,53;2009.p.S11–104.

[46]GopinathB, HarrisDC, FloodVM, BurlutskyG, MitchellP. Consumptionoflong-chainn-3PUFA,alpha-linolenicacid andfishisassociatedwiththeprevalenceofchronickidney disease.BrJNutr2011;105:1361–8.

[47]MollstenAV, DahlquistGG, StattinEL, RudbergS.Higher intakesoffishproteinarerelatedtoalowerriskof

microalbuminuriainyoungSwedishtype1diabeticpatients.

DiabetesCare2001;24:805–10.

[48]GrossJL, deAzevedoMJ, SilveiroSP, CananiLH, CaramoriML, ZelmanovitzT.Diabeticnephropathy:diagnosis,prevention, andtreatment.DiabetesCare2005;28:164–76.

[49]MacphersonNA, MoscarelloMA, GoldbergDM.

Aminoaciduriaisanearlierindexofrenaltubulardamage thanconventionalrenaldiseasemarkersinthe

gentamicin-ratmodelofacuterenalfailure.Clinicaland investigativemedicine.Medecinecliniqueetexperimentale 1991;14:101–10.

[50]RonnevikAK, GudbrandsenOA.Substitutionofhigh-dose sucrosewithfructoseinhigh-fatdietsresultedinhigher plasmaconcentrationsofasparticacid,cystine,glutamic acid,ornithineandphenylalanine,andhigherurine concentrationsofarginineandcitrulline.NutrRes 2020;79:100–10.

[51]HanWK, WaikarSS, JohnsonA, BetenskyRA, DentCL, DevarajanP, etal. Urinarybiomarkersintheearlydiagnosis ofacutekidneyinjury.Kidneyint2008;73:863–9.

[52]FergusonMA, WaikarSS.Establishedandemergingmarkers ofkidneyfunction.Clinchem2012;58:680–9.

[53]DrotningsvikA, MidttunO, McCannA, UelandPM, HogoyI, GudbrandsenOA.Dietaryintakeofcodproteinbeneficially affectsconcentrationsofurinarymarkersofkidneyfunction andresultsinlowerurinarylossofaminoacidsinobese Zuckerfa/farats.BrJNutr2018;120:740–50.

[54]YamakadoM, TanakaT, NagaoK, ImaizumiA, KomatsuM, DaimonT, etal. Plasmaaminoacidprofileassociatedwith fattyliverdiseaseandco-occurrenceofmetabolicrisk factors.Scirep2017;7:14485.

[55]NygardO, VollsetSE, RefsumH, BrattstromL, UelandPM. Totalhomocysteineandcardiovasculardisease.Jinternmed 1999;246:425–54.

[56]KarminO, SiowYL.Metabolicimbalanceofhomocysteine andhydrogensulfideinkidneydisease.Currentmedchem 2018;25:367–77.

[57]RefsumH, GuttormsenAB, FiskerstrandT, UelandPM. Hyperhomocysteinemiaintermsofsteady-statekinetics.

Eurojpediatri1998;157(Suppl2):S45–9.

[58]TousoulisD, GeorgakisMK, OikonomouE, PapageorgiouN, ZaromitidouM, LatsiosG, etal. Asymmetric

Dimethylarginine:ClinicalSignificanceandNovel TherapeuticApproaches.Currentmedchem 2015;22:2871–901.

[59]TsikasD, BogerRH, SandmannJ, Bode-BogerSM, FrolichJC. Endogenousnitricoxidesynthaseinhibitorsareresponsible fortheL-arginineparadox.FEBSLett2000;478:1–3.

[60]MoaliC, BoucherJL, SariMA, StuehrDJ, MansuyD.Substrate specificityofNOsynthases:detailedcomparisonof L-arginine,homo-L-arginine,theirNomega-hydroxy derivatives,andNomega-hydroxynor-L-arginine.

Biochemistry1998;37:10453–60.

[61]FrigoletME, TorresN, TovarAR.Therenin-angiotensin systeminadiposetissueanditsmetabolicconsequences duringobesity.JNutrBiochem2013;24:2003–15.

[62]ZamanMA, OparilS, CalhounDA.Drugstargetingthe renin-angiotensin-aldosteronesystem.Narev.Drugdiscov 2002;1:621–36.

[63]Alonso-GaliciaM, BrandsMW, ZappeDH, HallJE.

HypertensioninobeseZuckerrats.RoleofangiotensinIIand adrenergicactivity.Hypertension1996;28:1047–54.

[64]VildmyrenI, DrotningsvikA, OterhalsA, WeenO, HalstensenA, GudbrandsenOA.CodResidualProtein PreventedBloodPressureIncreaseinZuckerfa/faRats, PossiblybyInhibitingActivitiesofAngiotensin-Converting EnzymeandRenin.Nutrients2018;10:1820.

[65]Altorf-vanderKuilW, BrinkEJ, BoetjeM, SiebelinkE, BijlsmaS, EngberinkMF, etal. Identificationofbiomarkers forintakeofproteinfrommeat,dairyproductsandgrains:a controlleddietaryinterventionstudy.BrJNutr

2013;110:810–22.

[66]Al-WaizM, MitchellSC, IdleJR, SmithRL.Themetabolismof 14C-labelledtrimethylamineanditsN-oxideinman.

Xenobiotica1987;17:551–8.

[67]ChalasaniN, YounossiZ, LavineJE, DiehlAM, BruntEM, CusiK, etal. Thediagnosisandmanagementof

non-alcoholicfattyliverdisease:practiceGuidelinebythe AmericanAssociationfortheStudyofLiverDiseases, AmericanCollegeofGastroenterology,andtheAmerican GastroenterologicalAssociation.Hepatology2012;55:2005–23. [68]VikorenLA, DrotningsvikA, MwakimongaA, LehS,

MellgrenG, GudbrandsenOA.Dietscontainingsalmonfillet delaydevelopmentofhighbloodpressureandhyperfusion damageinkidneysinobeseZuckerfa/farats.JAmSoc Hypertens2018;12:294–302.