High-altitude ancestry and hypoxia acclimation have distinct effects on exercise capacity and ...

13 Pages·2015·1.11 MB·English

AmJPhysiolRegulIntegrCompPhysiol308:R779–R791,2015.
FirstpublishedFebruary18,2015;doi:10.1152/ajpregu.00362.2014.
High-altitude ancestry and hypoxia acclimation have distinct effects on
exercise capacity and muscle phenotype in deer mice
Mikaela A. Lui,1 Sajeni Mahalingam,1 Paras Patel,1 Alex D. Connaty,1 Catherine M. Ivy,1
Zachary A. Cheviron,2 Jay F. Storz,3 Grant B. McClelland,1 and Graham R. Scott1
1DepartmentofBiology,McMasterUniversity,Hamilton,Ontario,Canada;2SchoolofIntegrativeBiology,University
ofIllinois,Urbana,Illinois;and3SchoolofBiologicalSciences,UniversityofNebraska,Lincoln,Nebraska
Submitted26August2014;acceptedinfinalform15February2015
Lui MA, Mahalingam S, Patel P, Connaty AD, Ivy CM, Che- rements in aerobic capacity (i.e., maximal rate of oxygen
viron ZA, Storz JF, McClelland GB, Scott GR. High-altitude consumption, V˙O2max).
ancestry and hypoxia acclimation have distinct effects on exercise Both acclimatization and genotypic (evolutionary) adapta-
capacity and muscle phenotype in deer mice. Am J Physiol Regul tion to high altitude can potentially enhance V˙O2max of high-
IntegrCompPhysiol308:R779–R791,2015.FirstpublishedFebru-
land natives in hypoxia (54, 63). The hypoxia acclimation
ary 18, 2015; doi:10.1152/ajpregu.00362.2014.—The hypoxic and
response that is typical of lowlanders can involve adaptive
cold environment at high altitudes requires that small mammals
physiologicaladjustmentsthatimproveoxygentransportrates,
sustain high rates of O transport for exercise and thermogenesis
2 butitcanalsoincludemaladaptiveresponsesthatarecounter-
while facing a diminished O availability. We used laboratory-born
2 productive to oxygen transport over the course of prolonged
and -raised deer mice (Peromyscus maniculatus) from highland and
hypoxia exposure (47, 63, 72). The unique genotypic adapta-
lowland populations to determine the interactive effects of ancestry
andhypoxiaacclimationonexerciseperformance.MaximalO con- tions of highlanders generally improve oxygen transport in
2
sumption(V˙O2max)duringexerciseinhypoxiaincreasedafterhypoxia hypoxia (54, 63) and could, in principle, involve adaptive
acclimation(equivalenttothehypoxiaat(cid:2)4,300melevationfor6–8 modifications in the ancestral acclimatization response to en-
wk) and was consistently greater in highlanders than in lowlanders. vironmental hypoxia. Although there is good evidence that
V˙O2maxduringexerciseinnormoxiawasnotaffectedbyancestryor genotypicadaptationtohighaltitudesoftenbluntsmaladaptive
acclimation. Highlanders also had consistently greater capillarity, hypoxia responses (18, 59), relatively few studies have as-
oxidative fiber density, and maximal activities of oxidative enzymes sessedtheinteractiveeffectsofhighlandancestryandhypoxia
(cytochrome c oxidase and citrate synthase) in the gastrocnemius acclimation on V˙O2max and its physiological basis in highland
muscle, lower lactate dehydrogenase activity in the gastrocnemius,
natives.
andgreatercytochromecoxidaseactivityinthediaphragm.Hypoxia
The capacity for oxygen diffusion in active skeletal muscle
agcacsltirmocanteiomniudsidphneontotayfpfeectofanhyigholfantdheerssewmaussacslesotcriaaittesd. Twhitehuhnigiqhueer isoneofthemostimportantdeterminantsofV˙O2maxinhypoxia
(8, 56, 69). Although some evidence suggests that acclimati-
mRNA and protein abundances of peroxisome proliferator-activated
receptor (cid:3) (PPAR(cid:3)). Vascular endothelial growth factor (VEGFA) zation to high-altitude hypoxia increases muscle capillarity, it
transcriptabundancewaslowerinhighlanders,andhypoxiaacclima- ismorecommonforcapillaritytobeunalteredbyhigh-altitude
tion reduced the expression of numerous genes that regulate angio- hypoxia in mammals (43). This finding is counter-intuitive
genesisandenergymetabolism,incontrasttotheobservedpopulation given the potential benefit of muscle capillarity for oxygen
differences in muscle phenotype. Lowlanders exhibited greater in- transport in hypoxia, and the general expectation is that tissue
creasesinbloodhemoglobincontent,hematocrit,andwetlungmass oxygen limitation is a primary signal for angiogenesis (22).
(but not dry lung mass) than highlanders after hypoxia acclimation. Some highland birds are characterized by increased muscle
Genotypic adaptation to high altitude, therefore, improves exercise capillarity (33, 44, 55), but the potential interactive effects of
performance in hypoxia by mechanisms that are at least partially highlandancestryandhypoxiaacclimationonmusclecapillar-
distinctfromthoseunderlyinghypoxiaacclimation.
ity are poorly understood.
high-altitude adaptation; oxygen transport; capillarity; oxidative ca- The deer mouse (Peromyscus maniculatus) is an emerging
pacity;PPAR(cid:3) modelspeciesforstudyinghigh-altitudephysiology.Thisspe-
cies has the broadest altitudinal distribution of any North
American mammal, stretching from below sea level in Death
HIGH ALTITUDE IS ONE OF THE harshest terrestrial environments Valley, California, to more than 4,300 m above sea level in
inhabited by endothermic vertebrates. The decline in oxygen
numerous mountain ranges (30). Highland deer mice are sub-
tension with elevation can reduce aerobic scope dramatically jecttostrongdirectionalselectionforincreasedaerobiccapac-
bylimitingtissueoxygensupplyandmitochondrialrespiration ity in hypoxia (25), presumably due to the high premium on
(73).High-altitudehypoxiaisaparticularlyseverechallengeto thermogenesis and aerobic exercise performance under condi-
small endotherms, because the cold temperatures at elevation tionsofhypoxiaandcoldstress.Highlandpopulationsofdeer
require that they concurrently sustain high rates of oxygen micearegeneticallydistinctfromlowlandpopulations,andthe
transporttofuelthermogenesis(24).Thisrequiressmallhigh- inferred levels of gene flow are very low across the Great
land natives to somehow overcome the hypoxia-induced dec- Plains-Colorado Front Range transition (60). Deer mice have
proven to be extremely useful for understanding the genetic,
molecular, and structural adaptations to high altitude that
Address for reprint requests and other correspondence: G. R. Scott: Life
increase hemoglobin-oxygen affinity and that likely improve
SciencesBldg.,1280MainSt.West,Hamilton,ON,L8S4K1,Canada(e-mail:
[email protected]). blood-oxygentransportinhypoxia(10,48,61,62).Integrative
http://www.ajpregu.org 0363-6119/15Copyright©2015theAmericanPhysiologicalSociety R779
R780 HIGH-ALTITUDEADAPTATIONINDEERMICE
physiologicalstudiesofdeermice,therefore,holdgreatprom- from a certified compressed gas mixture (balance N ). Excurrent
2
iseforelucidatingthemechanisticbasisofhigh-altitudeadap- chamber air was subsampled at 200 ml/min, dried using prebaked
tation and acclimatization across the oxygen transport path- drierite, and passed through O2 and CO2 analyzers (Sable Systems).
The treadmill (kept at a constant angle of 10°) was initially set at 4
way.
m/min, and the speed was increased by 3 m/min every 2 min until
The objective of this study was to determine how highland
exhaustion.Thiswasdefinedaswhenanytwoofthefollowingthree
ancestry(whichmayreflectgenotypicadaptation)andhypoxia
conditionsweremet:1)themousecouldnolongermaintainposition
acclimation affect V˙O2max, as well as muscle capillarity and on the treadmill belt, 2) there is no increase in O consumption rate
2
oxidative phenotype in deer mice. To overcome the inherent with speed, and/or 3) respiratory exchange ratio (quotient of CO
2
limitationsofstudyingindividualsintheirnativeenvironment production and O consumption rates) was greater than one. Data
2
for disentangling the effects of ancestry and environment, we were acquired every 1 s with Expedata data acquisition software
established captive breeding colonies from wild-caught mice (SableSystems),andV˙O2maxwascalculatedaspreviouslydescribed
nativetohighandlowaltitudes.Ourresultsonfirstgeneration (37). Normoxic and hypoxic exercise trials were conducted (cid:2)1 wk
(F )lab-raisedmicesuggestthatbothgenotypicadaptationand apartforeachmouse,andtheorderinwhichthetrialswereconducted
1
hypoxiaacclimationcanimproveoxygentransportcapacityin wasrandomized.
hypoxia, but that adaptation alone affects muscle capillarity Restingratesofoxygenconsumption(V˙O2rest)weremeasuredina
subsetofunfastedmicefrombothpopulationsthatwereacclimatedto
and oxidative phenotype.
normoxia. Measurements were made at room temperature as de-
MATERIALS AND METHODS scribed for treadmill exercise with the following exceptions. Trials
were conducted in a small ((cid:2)500 ml) metabolic chamber that was
Populationsofhighlandandlowlanddeermice.Captivebreeding coveredtominimizestress,andairwassuppliedataflowrateof650
populationsofdeermicewerecreated,aspreviouslydescribed(12). ml/min. Normoxic air was first supplied to the chamber for 1 h,
Adults were live trapped at high altitude on the summit of Mount followed by a 1-h exposure to hypoxic air (12% O fraction). The
2
Evans (Clear Creek County, CO, at 39°35=18==N, 105°38=38==W, V˙O2restdatareportedherearetheminimumratesofoxygenconsump-
4,350 m above sea level) (P. m. rufinus) and at low altitude in the tionmeasuredforonecontinuousminuteinnormoxiaandinhypoxia.
Great Plains (Nine Mile Prairie, Lancaster County, NE, at Musclehistology.Onegastrocnemiusmusclefromeachmousewas
40°52=12==N, 96°48=20.3==W, 430 m above sea level) (P. m. ne- coatedinembeddingmedium,frozeninliquidN -cooledisopentane,
2
bracensis).MicewerethentransportedtotheUniversityofNebraska and stored at (cid:5)80°C. Tissue was sectioned (10 (cid:6)m) transverse to
(elevation360m)andhousedincommon-gardenlabconditionstobe musclefiberlengthinacryostatat(cid:5)20°C.Capillarieswereidentified
usedasaparentalstocktoproduceF progeny.TheF progenywere bystainingforalkalinephosphataseactivity(assaybufferconcentra-
1 1
bornandraisedinacommonnormoxicenvironmentandthenshipped tions in mM: 1.0 nitroblue tetrazolium, 0.5 5-bromo-4-chloro-3-
to McMaster University (elevation 50 m) for all experiments in this indoxylphosphate,28NaBO ,and7MgSO ;pH9.3)for1hatroom
2 4
study. Mice were held at McMaster University in standard holding temperature.Oxidativemusclefibers(bothslowandfast)wereiden-
conditionsat24–25°Cwithunlimitedaccesstochowandwaterforat tified by staining for succinate dehydrogenase activity (assay buffer
least 1 mo before experimentation (12:12-h light-dark photoperiod). concentrations in mM: 0.6 nitroblue tetrazolium, 2.0 KH PO , 15.4
2 4
AllanimalprotocolsfollowedguidelinesestablishedbytheCanadian Na HPO , and 16.7 sodium succinate) for 1 h at room temperature.
2 4
Council on Animal Care and were approved by the McMaster Uni- Slowoxidativemusclefiberswereidentifiedbyslowmyosinimmu-
versityAnimalResearchEthicsBoard. noreactivityusingamouseIgAprimaryantibody(S58;Developmen-
Treatments. Mice from each population were split into two accli- talStudiesHybridomaBank,IowaCity,IA)asfollows.Sectionswere
mation groups: 1) normobaria in standard normoxic holding condi- fixedinacetonefor10min,blockedin10%normalgoatserum[made
tions, or 2) hypobaric hypoxia (barometric pressure of 60 kPa) upofPBS(0.15mol/l,pH7.4)containing1%TritonX-100and1.5%
equivalent to that at an elevation of 4,300 m. Mice in the hypoxia BSA(PBS/TX/BSA)]for1h,andincubatedovernightat2°CinS58
acclimationgroupwereplacedinspeciallydesignedhypobariccham- antibodysolution(1:10dilutioninPBS/TX/BSA).Thenextmorning,
bers,andpressurewasdecreasedgraduallyoverthefirstfewdays,as sections were treated with peroxidase blocking reagent (Dako, Burl-
described previously (45). Mice were weighed every 3 days during ington,ON,Canada)for10min,incubatedinsecondaryantibody(rat
cagecleaning,whichrequiredthatthehypobaricgroupsbereturnedto anti-mousebiotinIgA;SouthernBiotech,Birmingham,AL)solution
normobaria for a brief period ((cid:4)1 h). Mice were subjected to respi- (1:20 dilution in PBS/TX/BSA) for 1 h, incubated in ExtrAvidin-
rometrymeasurements(seebelow)after6–8wkofacclimation.After Peroxidase(1:50dilutioninPBS/TX/BSA;Sigma-Aldrich,Oakville,
these measurements were completed, blood was collected from the ON,Canada)for30min,andfinallydeveloped(0.4mg/ml3-amino-
maxillary vein using a lancet, and blood hemoglobin concentration 9-ethyl-carbazoleand0.02%H O in0.05Msodiumacetatebuffer,
2 2
wasmeasuredusingDrabkin’sreagent(Sigma-Aldrich,Oakville,ON, pH5.0)for(cid:2)5min.SectionswerewellrinsedinPBSbetweeneach
Canada), according to the manufacturer’s instructions. Hematocrit oftheabovesteps.
wasalsomeasuredonasubsetofmice.Miceweretheneuthanizedby Stereologicalmethodswereusedtomakeunbiasedmeasurements
cervical dislocation, lungs were weighed (i.e., lung wet mass) and ofseveralhistologicalvariablesfortheabovesections,aspreviously
frozen,andtissuesweresampledasdescribedbelow.Lungswerelater described (17, 71). Images were collected systematically using light
driedtoconstantmassat60°Cinanoventodeterminelungdrymass. microscopy, such that there was equal representation of images
Respirometry measurements. Maximal rates of oxygen consump- analyzedfromacrosstheentiregastrocnemius.Sufficientimageswere
tion (V˙O2max) were measured in each mouse in both normoxia and analyzedforeachsampletoaccountforheterogeneity,determinedby
normobaric hypoxia (12% O fraction, equivalent to 4,300 m eleva- the number of replicates necessary to yield a stable mean value
2
tion) using open-flow respirometry in a rodent treadmill (working (roughly half of the entire section was imaged). The density of fast
chambervolumeof(cid:2)800ml)thathasbeendescribedpreviously(53). oxidativefiberswascalculatedasthedifferencebetweenthedensities
Micewerefirstplacedonthetreadmillandallowedtorestfor10min ofoxidativefibersandslowoxidativefibers.
tobecomeaccustomedtothechamber.Incurrentairwasdeliveredat Fluorescenceimmunohistochemistrywasalsousedtoconcurrently
aflowrateof2,000ml/minusingamass-flowmeterandpump(Sable identifycapillariesandslowoxidativemusclefibers,usingaratIgG
Systems, Las Vegas, NV). Incurrent normoxic air was pulled from primary antibody against CD31 protein (BD Biosciences, Missis-
outside of the building and scrubbed free of water and CO using sauga, ON, Canada) and the S58 antibody, respectively, to qualita-
2
drierite, soda lime, and ascarite. Incurrent hypoxic air was delivered tively confirm the measurements that were made using light micros-
AJP-RegulIntegrCompPhysiol•doi:10.1152/ajpregu.00362.2014•www.ajpregu.org
HIGH-ALTITUDEADAPTATIONINDEERMICE R781
copy.Sectionswereblockedin10%normalgoatseruminPBS/TX/ and DNase-free water. Each reaction was run in duplicate using a
BSAfor1h,andincubatedovernightat2°CinPBS/TX/BSAsolution CFX Connect real-time system (Bio-Rad), beginning with a 30-s
containingtheS58(1:10dilution)andCD31(1:400dilution)primary initialdenaturationat95°Cfollowedby40cyclesof95°Cfor5sand
antibodies. The next morning, sections were incubated in secondary 60°Cfor30s.Acontrolrunwithnotemplatewasrunforeachprimer
antibody diluted in PBS/TX/BSA (1:400, Alexa Fluor 488 goat settocheckforthepresenceofcontaminants,andadissociationcurve
anti-rat IgG; 1:200, Alexa Fluor 594 goat anti-mouse IgG; Life wasconstructedafteramplificationforeachsampletoensurethatonly
Technologies, Burlington, ON, Canada) for 2 h. Sections were well
asingleproductwasamplified.Agarosegelelectrophoresisofrepre-
rinsedinPBSbetweeneachoftheabovesteps.Stainedsectionswere
sentativeproductswasusedtoverifythepresenceofasingleampli-
imagedusingaconfocalfluorescencemicroscope.
conoftheappropriatesize.Astandardcurveforeachprimersetwas
Enzyme activity measurements. Gastrocnemius and diaphragm
generated for one sample, and all other samples were expressed
muscles were frozen in liquid N , powdered under liquid N , and
2 2 relative to this standard. All results were normalized to the cDNA
storedat(cid:5)80°C.Samplesweighing(cid:2)20mgwerehomogenizedin20
levelof12SrRNAgene.ThecDNAlevelof12S,aswellas(cid:7)-actin
volumes of homogenization buffer (concentrations in mM: 100
(actb),didnotvarybetweengroups.
KH PO ,5ethylenediaminetetraaceticacid,and0.1%Triton-X-100,
2 4 Westernblotanalysis.Peroxisomeproliferator-activatedreceptor(cid:3)
atpH7.2).Cytochromecoxidase(COX)activitywasassayedshortly
afterhomogenization,andcitratesynthase(CS)andlactatedehydro- (PPAR(cid:3))proteinwasmeasuredingastrocnemiusmusclebyWestern
genase(LDH)activitiesweremeasuredafterstorageofhomogenateat blot analysis. The powdered muscle samples were briefly homoge-
(cid:5)80°C. Activity was assayed at 37°C by measuring the change in nized with a motorized homogenizer in ice-cold buffer (150 mM
absorbance over time (CS, 412 nm; COX, 550 nm; LDH, 340 nm) NaCl,50mMTris·HCl,1.0%TritonX-100,0.5%deoxycholicacid,
underthefollowingconditions(inmMunlessotherwisestated):CS, and0.1%SDS,atpH8.0).Theproteinisolatesweredenaturedfor5
40Tris,0.01oxaloacetate,0.23acetyl-coA,0.1DTNB,pH8.0;COX, minat100°CinLaemmlisamplebuffer(Bio-Rad)containing(cid:7)-mer-
100KH PO ,0.1reducedcytochromec,pH7.0;LDH,40Tris,0.28 captoethanol. Denatured proteins were separated in precast 12%
2 4
NADH,2.4pyruvate,andatpH7.4.Preliminaryexperimentsverified sodiumdodecylsulfate-polyacrylamidegels(Bio-Rad)for45minat
thatsubstrateconcentrationsweresaturating.Allenzymeassayswere 120Vandthenfor15minat150VinaMini-ProteinTetraSystem
runintriplicate.Enzymeactivitiesweredeterminedbysubtractingthe (Bio-Rad). Separated proteins were transferred to polyvinylidene
background rate (measured in control reactions that contained no difluoridemembranes(Bio-Rad)at25Vfor10minusingaTransBlot
substrate) from the rates measured in the presence of substrate. Turbo Transfer System (Bio-Rad). Membranes were incubated in
Extinctioncoefficientsweredeterminedempiricallyforeachassayin blocking solution overnight [5% skim milk powder in PBS-Tween
theappropriateassayconditions.
(PBST)buffer:1.5mMNaH PO ·H O,8.1mMNa HPO ,145.5mM
RNA extraction and quantitative real-time PCR. Total RNA was 2 4 2 2 4
NaCl,0.05%Tween-20,atpH7.4].Membraneswerethenwashedfor
isolated from 30–40 mg of the powdered muscle samples using
5mininPBSTandincubatedfor1hwithprimaryantibody(arabbit
TRIzol reagent (Life Technologies), following the manufacturer’s polyclonal antibody raised against human PPAR(cid:3) that recognizes
instructions.cDNAwasthensynthesizedusingSuperScriptIIRNase
H(cid:5)reversetranscriptase(Invitrogen,Burlington,ON,Canada)from1 both (cid:3)1 and (cid:3)2 human isoforms and detects a band at (cid:2)75 kDa;
(cid:6)g of total RNA (treated with DNase I), according to the manufac- ab19481;Abcam,Cambridge,MA)diluted1:500inPBSTcontaining
1%BSA(PBST-BSA).Membraneswerethenincubatedfor1hwith
turer’sinstructionsusingamixtureofrandomhexamerandoligo-dT
donkey anti-rabbit secondary antibody (horseradish peroxidase con-
primers. The resulting cDNA was diluted 50-fold using RNase and
DNase-freewater(Invitrogen). jugated;SantaCruzBiotechnology,SantaCruz,CA)atadilutionof
Transcript sequences were mined from deer mouse sequences in 1:10,000inPBST-BSA.MembranesweredevelopedinECLClarity
the NCBI sequence read archive (accession no. SRA051883, Solution (Bio-Rad), and band intensity was detected by chemilumi-
SRA091630)(11,13),andwereusedasatemplateonwhichtodesign nescence using a ChemiDoc MP Imaging System (Bio-Rad). A
primers for quantitative real-time PCR (qPCR) using Primer 3 soft- common sample was included on each gel to control for transfer
ware (32, 68) (Table 1). Each qPCR reaction contained 4 (cid:6)l of the efficiency,andbandswerequantifiedusingthegelanalysisfeatureof
five-fold diluted cDNA, 5 (cid:6)l of Sso Fast Eva Green supermix ImageLabsoftware(Bio-Rad).Thebandintensityofeachsamplewas
(Bio-Rad,Mississauga,ON,Canada),0.4(cid:6)lofeachofforwardand normalizedtotheamountofloadedprotein(determinedbyCoomas-
reverse primer (final concentration of 10 (cid:6)M), and 0.2 (cid:6)l of RNase siebluestaining)andtothenormalizedbandintensityofthecommon
sample and is expressed here relative to lowland mice acclimated to
normoxia.
aTnaabllyes1is.oPfrgimeneersexupseredssfoiornquantitative real-time PCR Calculationsandstatistics.V˙O2maxdatawerecorrectedforbody
mass (M ) using the residuals from allometric regressions. Least-
b
Target squares regressions to the equation V˙O2max (cid:8) a Mbb were carried
Transcript ForwardPrimer ReversePrimer out using GraphPad Prism software (La Jolla, CA). The residual
from the regression was calculated for each individual, and these
Angpt1 TCAGTGGCTGCAAAAACTTG TCAGTTTTCGGGTCTGCTCT
residuals were used for statistical comparisons (see below). Data
Angpt2 GATCTTCCTCCAGCCCCTAC TGCACCACATTCTGTTGGAT
Flt1 AGCGGGACTTCTTCTTCCTC CTGCAATCCTCGTCTTCCTC are reported graphically as residuals as well as the sum of the
Kdr CCACCTCACCTGTTTCCTGT TCTGTCTGGCTGTCATCTGG residual and the expected V˙O2max value for an average-sized 22-g
Nrd1d1 GCTCCACTTGTCTCCCTCAG GAGTCAGGGACTGGAAGCTG mouse (i.e., V˙O2max corrected to a body mass of 22 g). A similar
Ppara CTCGTGCAGGTCATCAAGAA ACCTAGGCTCAGCCCTCTTC residualapproachwasusedtoexaminevariationincapillaritythat
Ppard CCCAGAATTCCTCTCCTTCC ACAGCTCCGGTCACACTTCT
was independent of muscle oxidative phenotype, but in this case,
Pparg CCGTGCAAGAGATCACAGAA GGGCTCCATAAAGTCACCAA
we used a linear regression.
Ppargc1a CAAGCTGTTTTTGACGACGA AGAGCAAGAAGGCGACACAT
Ppargc1b AGAGGAACTGGTCAGCCTCA CGTCAAGGACTCCTCAAAGC Data are generally reported as means (cid:9) SE (except when data
Pprc1 TTCCCTCATCTCCCTCATTG CTCTCCCGAGGAGACACAAG pointsfromindividualsamplesareshown).Two-factorANOVAand
Tek TTGGCCATGTGACAGTTTGT CAAATGCCACACATCACTCC Bonferroni multiple-comparisons tests were used as appropriate to
Vegfa GTACCTCCACCATGCCAAGT ACCTCATTAGGGGCACACAC assess the main effects and interactions of ancestral altitude and
12S CTGGCCATCGCTTAAAACTC TTGCTTCCCACCTCATAAGC acclimationenvironment.AsignificancelevelofP(cid:4)0.05wasused
Actb GTCGTACCACTGGCATTGTG AGGGCAACATAGCACAGCTT
throughout.
AJP-RegulIntegrCompPhysiol•doi:10.1152/ajpregu.00362.2014•www.ajpregu.org
R782 HIGH-ALTITUDEADAPTATIONINDEERMICE
RESULTS ml/minat12%O ),suggestingthatmicecanincreaseratesof
2
aerobic metabolism by approximately four-fold when exercis-
Respirometry. Both high-altitude ancestry and hypoxia ac-
ing in normoxia and approximately five- to six-fold when
climation improved aerobic exercise capacity in hypoxia. The
exercisinginhypoxiaandthattheobservedeffectsofhighland
(mmaexaismuareldraoten oaftroexaydgmenillcaotns1u2m%ptiionnspi(rVe˙dO2mOax)frianctihoynp)oxinia- ancestryonV˙O2maxshouldresultincomparableimprovements
2 in aerobic scope.
creased by (cid:2)13% in both populations after 6–8 wk of accli-
Lung and blood measurements. There were interactive ef-
mationtohypobarichypoxia(60kPatotalpressure,equivalent
fectsofhigh-altitudeancestryandhypoxiaacclimationonlung
to that at an elevation of 4,300 m), based on comparisons
size. Lung wet mass increased by (cid:2)23% in lowland mice in
betweenhypoxiaandnormoxiaacclimationgroupswithineach
population (Fig. 1A). Hypoxic V˙O2max was (cid:2)8% higher in rtoespoonnlyse(cid:2)to8%hypionxtihaeachcilgimhlaatnidonm,bicuett(hFisigr.es2pAo)n.sTehweasefbfleucntteodf
highlanddeermicethaninlowlanddeermicewhencompared
ancestryonthemagnitudeoftheresponsewasnotattributable
ineithernormoxicorhypoxicacclimationtreatments(Fig.1A).
to differences in lung dry mass, which underwent a similar
There was no significant interaction between ancestry and
(cid:2)20–30%increaseinbothgroups.Instead,itisthedifference
acclimationtreatment,indicatingthattheancestralacclimation
response of V˙O2max to hypoxia has been retained in high- i0n.6l%un)gfrwoamtehricgohnlatenndtmthiacted(i7st9i.n2gu(cid:9)is0h.e6s%lo)walfatenrdhmypicoexi(a81a.c0cl(cid:9)i-
altitudemicethataresubjecttochronichypoxiaintheirnative
habitat. Because we observed allometric rather than isometric mation.
scaling of V˙O2max to body mass, we used a residual-based Thereweresimilarinteractiveeffectsofhigh-altitudeances-
approach to correct for body mass before making these com- tryandhypoxiaacclimationonbloodhemoglobincontentand
parisons (Fig. 1B). hematocrit. Hemoglobin content was similar between popula-
Aerobic exercise capacity in normoxia was not affected by tions in normoxia, but it increased by a much greater magni-
either high-altitude ancestry or hypoxia acclimation. Hypoxia tude in lowland mice ((cid:2)38%) than in highland mice ((cid:2)23%)
acclimation led to only a slight ((cid:2)4%), but statistically non- after hypoxia acclimation (Fig. 2B). The blunted response in
significant,increaseinV˙O2maxinnormoxia,andtherewereno thehighlandersresultedinasignificantancestry(cid:10)acclimation
differences between highlanders and lowlanders (Fig. 1C). environment interaction for blood hemoglobin content. Blood
Rates of oxygen consumption measured in normoxia-accli- hematocritalsoincreasedbyagreatermagnitudeinlowlanders
matedmiceatrestweresimilarinhighlandmice(0.96(cid:9)0.10 (46.2 (cid:9) 0.9% in normoxia, 59.1 (cid:9) 2.1% in hypoxia; n (cid:8) 7)
ml/min at 21% O , 0.41 (cid:9) 0.07 ml/min at 12% O ) and thaninhighlanders(42.6(cid:9)1.4%innormoxia,51.5(cid:9)2.0%in
2 2
lowland mice (1.11 (cid:9) 0.09 ml/min at 21% O , 0.45 (cid:9) 0.08 hypoxia;n(cid:8)10),astherewasasignificantmaineffectofboth
2
A B
MpF(rbdec(tmhViMflaye˙icifeogremOebc0ixr.rcs;2e.arai61omtecgVl5tms˙.cnsiar)oOuVixeMsi˙od)ncsl2(fwOausemDaiti2anoxheaiom)lnxtrnii.snhhmgeag(cid:8)xayhoAhm(atnplCfhtyla0oaoehp)nar.idx6ayonnbdreift7epxdoaeosfaoildsMrociownxCdyrowcitbueV:cfhne˙m0rals.eoaeOoV4ll˙tasner9x2rgsOmr)fdym(ysetr2A(agoegmBax(d)aaexrVna)˙eetxnfixw.decoesOirStscsrreo2Vhi˙oirimaomynemsbnOnanhpos2xihiaoluaomdcaios(cid:8)lwmxeygarlsfxo.ihespcammw0slnotTasi.taaomeio5nhhesctrtn3dhdees---- -1VHypoxic (ml min)O max2 22333.....68024 HLoigwhllaannddee*rr --00000...024..42 -1ual of hypoxic V(ml min)O max 2 -1VHypoxic (ml min)O max 2 22334.....05050
hybpoxicandnormoxicV˙O2maxforanaverage- -0.6 esid 1.5
sized 22-g mouse, calculated for each group R
byaddingtheresidualtotheV˙O2maxpredicted Normoxia Hypoxia 10 15 20 25 30 35
at22gbytheregression.Therewasasignif- Acclimation Environment Body mass (g)
tircyan(t*eFffect(cid:8)on5h.0y2p;oxPic(cid:8)V˙O0.20m3a0x)oafnbdotahccalnicmeas-- C D
tion (†F1,14,343 (cid:8) 12.20; P (cid:8) 0.001), with no 0.4 -1)n 5.5 Nor-low
significant interaction between each factor mi Hyp-low
Fs0lrlclmv(oooeFie.e1g0pwmwos1r,n40rttx,ll43reeei1aai3yfidtscnn;(cid:8)rec(cid:8)iFddPncatheer1tn2ii0(cid:8)rr,rga4t.ss9e.23hn0gv570(lg1((cid:8)ra%aN;.Hle9r9nesoiP29ydsa0scr;)9peti.-o,(cid:8)io1Prl)-onos.ln8nnofiw(cid:8)B,80w(d;(cid:8)ia.,Nan1e,0nPŒnn3o.dno)8dc19r(cid:8),o-9e4;)nDhrn,1imii0:n)(cid:8)fgnn(cid:8),o.tdoht6ebreax,69r(cid:8)ru1siahv7cdht0cfya;aetotV1lphifr˙adsrook4oecOngrrnoxc2ergflofimhFaiocrrwmyaytarm1hxph,yaa4leoonsii(t3gniaxxoiaonh(cid:4)enniirlnoccs----- -1VNormoxic (ml min)O max2 3344....6802 --0000..02..42 ual of normoxic VOmax (ml 2 -1VNormoxic (ml min)O max2 33445.....05050 HNyopr--hhiigghh
3.4 d
landers (Hyp-high, light gray inverted trian- -0.6 si 2.5
e
gles). Normoxia Hypoxia R 10 15 20 25 30 35
Acclimation Environment Body mass (g)
AJP-RegulIntegrCompPhysiol•doi:10.1152/ajpregu.00362.2014•www.ajpregu.org
HIGH-ALTITUDEADAPTATIONINDEERMICE R783
A try on capillary surface density than on capillary density may
11 Lowlander
* havebeenassociatedwithanincreaseinvesseltortuosityinthe
-1g g) 10 Highlander hbiyghalanladrgmeicdeirfefelaretinvceetointhethloewplaanttdermnicoef.Tchaipsilwlaarsysusgtagiensintegd,
m observedusingbothalkalinephosphatasehistochemistry(Fig.
ss ( 9 3, A and B) and immunohistochemistry against CD31 protein
a
m (Fig. 3, G and H): staining wrapped the muscle fibers much
et more often in highland mice (probably indicating that the
w 8
g capillarywassectionedlongitudinallytovessellength)thanin
n
u lowland mice (where staining was more discrete, indicating
L
7
that the capillary was sectioned transversely).
-1g g) 2.0 oxHidiagthiv-ealtpithuedneotyanpceesotfrythweagsasatrlsoocnaesmsoiucsiamteduscwlei.thHaighmlaonrde
m
s ( 1.8
s
a
m
dry 1.6
g
n
u
L 1.4
Normoxia Hypoxia
Acclimation Environment
B
24
*
-1dl) 22
g
n] ( 20
bi
glo 18
o
m
e 16
h
od [ 14
o
Bl
12
Normoxia Hypoxia
Acclimation Environment
Fig.2.Someresponsestohypoxiaacclimationarebluntedinhighlanddeer
mice.A:hypoxiaacclimationincreasedbothwetlungmass(F (cid:8)9.35;P(cid:8)
1,72
0.003)anddrylungmass(F (cid:8)24.32;P(cid:4)0.001)(bothrelativetobody
1,70
mass).Theeffectofaltitudeofancestryonwetlungmassnearedsignificance
(F (cid:8) 3.42; P (cid:8) 0.069), and there was a significant pairwise difference
1,72
betweenhighlandersandlowlandersinhypoxia(*P(cid:4)0.001),buttherewere
noeffectsofancestryondrylungmass(F (cid:8)0.261;P(cid:8)0.611).Normoxic
1,70
lowlanders, n (cid:8) 16 and 14, respectively; all other groups, n (cid:8) 20. B:
hemoglobincontentofthebloodwasaffectedbyhypoxiaacclimation(F (cid:8)
1,65
76.37P(cid:4)0.001),ancestry(F (cid:8)12.62;P(cid:8)0.001),andtheirinteraction
1,65
(F (cid:8)5.18;P(cid:8)0.026).Normoxiclowlanders,n(cid:8)14;hypoxiclowlanders,
1,65
n(cid:8)18;normoxichighlanders,n(cid:8)18;andhypoxichighlanders,n(cid:8)19.
altitude of ancestry (F (cid:8) 10.28; P (cid:8) 0.003) and hypoxia
1,30
acclimation (F (cid:8) 38.80; P (cid:4) 0.001) and a significant
1,30
pairwise difference between highlanders and lowlanders in
hypoxia. Fig.3.Histologicalanalysisofcapillarityandfibertypeinthegastrocnemius
muscleofdeermice.Quantificationusingstereologicalmethodswascarried
Muscle phenotype. High-altitude ancestry, but not hypoxia
outforcapillariesthatwereidentifiedusingalkalinephosphataseactivity(A
acclimation, was associated with elevated capillarity in the andB),oxidativemusclefibersthatwereidentifiedusingsuccinatedehydro-
locomotory (gastrocnemius) muscle. Several indices of capil- genase(SDH)activity(CandD),andslowoxidativemusclefibersthatwere
larity were greater in highland mice than in lowland mice, identified by staining slow myosin ATPase using peroxidase immunohisto-
including capillary surface density ((cid:2)35–41% higher), capil- chemistry (E and F). Capillaries (green) and slow myosin (red) were also
identifiedconcurrentlyusingfluorescenceimmunohistochemistry(GandH).
lary-to-fiber ratio ((cid:2)20–30% higher), and capillary density
Therewerecleardifferencesinstainingintensitybetweenhighland(A,C,E,
((cid:2)9–18%higher)(Figs.3and4).Thegreatereffectofances- G)andlowland(B,D,F,H)deermice.
AJP-RegulIntegrCompPhysiol•doi:10.1152/ajpregu.00362.2014•www.ajpregu.org
R784 HIGH-ALTITUDEADAPTATIONINDEERMICE
A Lowlander We used a regression approach to assess whether the vari-
ationincapillaritywasmerelyaresultofthepotentiallygreater
Highlander
0.05 * oxygen demands associated with having a more oxidative
e *
ac1) muscle phenotype. There is generally a strong relationship
y surfμ-y (m 0.04 bcaeptwilelaernytshuerfmaciteocahvoanildarbilaelofoxrygoexnygdeenmsaunpdpslyof(a27ti)s.sCueonasnidsttehnet
pillarensit 0.03 wbeittwhetehnisceaxppilelacrtaytiosunr,fathceeredewnsaistyaasntdronthgelianreeaarl cdoernrseiltaytioonf
ad oxidative fibers (Fig. 6A). However, residuals of capillary
C
surface density from this regression to the areal density of
0.02 oxidativefibersweresignificantlyhigherinhighlanddeermice
B
thaninlowlanddeermice(Fig.6B).Thisobservationsuggests
2.2
* that capillarity is enhanced in highlanders by a greater magni-
r 2.0 * tudethanwouldbeexpectedsolelyfromtheeffectsofhaving
peer a greater abundance of oxidative muscle fibers.
aries cle fib 1.8 phTenhoetyhpiestwoleorgeicaaslsodciifafteerdenwceitsh cinomcmapeinllsaurritayteadnidffeorexnidceastivine
pillus 1.6 the maximal in vitro activity of oxidative enzymes. Highland
Cam deermicehad(cid:2)52–71%higheractivitiesofCOX,(cid:2)29–35%
1.4 higheractivitiesofCS,and(cid:2)20–30%loweractivitiesofLDH
in the gastrocnemius muscle (Fig. 7). Differences were not
1.2
C restricted to the locomotory muscle, as there was a similar
1200 differenceinCOXactivityinthediaphragm(Fig.7).Interest-
ingly,theratioofCOX/CSactivitiesinthegastrocnemiuswas
sity 1100 significantly greater in highland deer mice (0.75 (cid:9) 0.03 in
n normoxia and 0.64 (cid:9) 0.04 in hypoxia) than in lowland deer
de2) mice (0.56 (cid:9) 0.06 in normoxia and 0.56 (cid:9) 0.09 in hypoxia)
ary -mm 1000 (F1,30 (cid:8) 6.52; P (cid:8) 0.016). Comparable differences in the
pill( COX/CS ratios in the diaphragm were also observed between
a 900 highland (0.83 (cid:9) 0.09 in normoxia and 0.83 (cid:9) 0.06 in
C
hypoxia) and lowland (0.46 (cid:9) 0.06 in normoxia and 0.61 (cid:9)
800 0.06 in hypoxia) mice (F (cid:8) 15.71; P (cid:4) 0.001).
1,30
Normoxia Hypoxia Expressionofcandidategenesinthemuscle.Theexpression
of 13 candidate genes that are important for regulating angio-
Acclimation Environment
genesis and energy metabolism (including mitochondrial bio-
Fig.4.Capillarityinthegastrocnemiusmuscleisgreaterinhighlanddeermice genesis) was compared between populations and acclimation
thaninlowlanddeermice,butitisnotaffectedbyhypoxiaacclimation.There environments, and two of these candidate genes were differ-
were significant effects of altitude of ancestry on capillary surface density
entially expressed between highland and lowland deer mice
(CSD;(cid:6)mofcapillarysurfaceper(cid:6)m2oftransversemusclearea)(F (cid:8)
25.06;P(cid:4)0.001)(A),thenumberofcapillariespermusclefiber(C:F)(F (cid:8)21,333.64; (Fig. 8, Table 3). PPAR(cid:3) transcript (Pparg) abundance was
1,33
P (cid:4) 0.001) (B), and the density of capillaries (CD; capillaries per mm2 of approximately twofold higher in highlanders than in lowland-
transversemusclearea)(F1,33(cid:8)5.22;P(cid:8)0.029)(C).Theeffectsofhypoxia ers (Fig. 8B). PPAR(cid:3) protein abundance was elevated in
acclimationwerenotsignificant(CSDF1,33(cid:8)0.319;P(cid:8)0.576;C:FF1,33(cid:8) highlandersbyacomparablemagnitude,butitsabundancealso
0.242;P(cid:8)0.626;andCDF (cid:8)0.504;P(cid:8)0.483).*Significantpairwise
1,33 increased with hypoxia acclimation (Fig. 9). The level of
differencebetweenhighlandersandlowlanderswithinanacclimationenviron-
ment(P(cid:4)0.05).Normoxiclowlanders,n(cid:8)8;hypoxiclowlanders,n(cid:8)9; transcriptabundanceofVEGFA(Vegfa),detectedusingqPCR
normoxichighlanders,n(cid:8)10;andhypoxichighlanders,n(cid:8)10. primers (Table 1) that were designed to recognize all known
Vegfa splice variants, was only 35–55% in highlanders com-
pared with the abundance in lowlanders.
deer mice had a greater proportion of oxidative fibers in the The expression of several candidate genes decreased in the
muscle, (cid:2)25–30% as a proportion of the total transverse area muscle in response to hypoxia acclimation (Fig. 8, Table 3).
of the muscle and (cid:2)7–17% as a proportion of total fiber Vegfa expression decreased in response to hypoxia in both
number (Figs. 3 and 5). The main cause of these differences populations to levels of transcript abundance that were 42–
was a 60–70% higher abundance of slow oxidative (type I) 66% of those in normoxia, suggesting that both highland
muscle fibers by area (Figs. 3 and 5), but there were also ancestry and hypoxia acclimation altered Vegfa expression in
smaller nonsignificant differences in the abundance of fast the same direction. The expression of several metabolic regu-
oxidative(typeIIa)musclefibers(Table2).Thedifferencesin lators also decreased in response to hypoxia acclimation (to
areal density between populations were generally greater than levelsrangingfrom25to56%ofnormoxiccontrols),including
thedifferencesinnumericaldensitybecausebothslowandfast PPAR(cid:11) (Ppara), PPAR(cid:12) (Ppard), PPAR(cid:3) coactivator (PGC)
oxidativefiberswerelargerinhighlanddeermice(by(cid:2)20%or 1(cid:7) (Ppargc1b), PGC-related protein 1 (Pprc1), and nuclear
more) with no difference in the size of fast glycolytic fibers receptor subfamily 1 group D member 1 (“Rev-ErbA (cid:11)”;
(Table 2). In contrast, hypoxia acclimation had no significant Nrd1d1). The expression of the remaining candidate genes—
effectsonfiber-typecompositioninthegastrocnemiusmuscle. angiopoietin 1 (Angpt1), angiopoietin 2 (Angpt2), VEGF re-
AJP-RegulIntegrCompPhysiol•doi:10.1152/ajpregu.00362.2014•www.ajpregu.org
HIGH-ALTITUDEADAPTATIONINDEERMICE R785
Lowlander Fig.5.Highlanddeermicehaveamoreoxida-
nsity e fibers 0.60 H*ighlander * density fiberse 00..6605 * tTainvhceeerseptrhywenoeornetythpseeiganirniefiactlhadneetngseaifstfytercootcfsnoeoxmfidiuaatslitvitmeudufiesbceloresf.
Areal deof oxidativ 00..4500 Numerical of oxidativ 00..5505 [fit1ohA9xbee.iA5drt(s9aoot;trxiaev,Pllmaettirfi)(cid:4)va,benaest0rroves.0eat[rh0Nseo1eNf)t]o(a,ootraxextlhai,dmneoua)ftmn,ivutnbhemueermefimorbbficeuefiarrssblcoelfrderesoel(nx(aFFsit1diiv,ta3ye4tiv(cid:8)o(cid:8)toef
1,34
0.30 0.45 8.44; P (cid:8) 0.006)], and the areal [AA(sox,m);
F (cid:8) 28.18; P (cid:4) 0.001], and numerical
1,33
Areal density of w oxidative fibers 0000....11220505 * * merical density ofw oxidative fibers 00..2205 [*l00oaFaNc.S.f13n9c,Nis03d9lgl46(ie4onms;r;wo(cid:8)isfiPaxNtco,ai(cid:8)m0aNonxn.n(di1)std;02oa.pFl25wxtoai8;1,wveim,43rPre;wl3e)aAfi(cid:8)in(cid:8)FsbdAne1ee(o8,0sr3rdt.os3s.2i7.xf4(cid:8)2s,Tfwm;ei9hgiPr);teen0hnFNi.(cid:8)ie2ficn1fN7c,ef30(0ae3ao.bn;cn0(cid:4)xett0Ps,tma7wo0[c](cid:8)A).efc0delAh0Fien0m1y(1no.;p,h6sa3xoPi4i0tt,gixmi7o(cid:8)eh(cid:8)i]nas)-.
o uo
sl Nsl 0.15 environment(P(cid:4)0.05).Normoxiclowlanders,
0.05 n(cid:8)9/9;hypoxiclowlanders,n(cid:8)10/10;nor-
Normoxia Hypoxia Normoxia Hypoxia moxichighlanders,n(cid:8)10/9;hypoxichighland-
ers, n (cid:8) 9/9 (oxidative fibers/slow oxidative
Acclimation Environment Acclimation Environment fibers,respectively).
ceptor 1 (Flt1), VEGF receptor 2 (Kdr), angiopoietin receptor at least partially underpinned by distinct physiological and
(Tek), and PGC-1(cid:11) (Ppargc1a)—were unaffected by both molecular mechanisms.
ancestry and acclimation environment. There were no candi- Oxygen transport at high altitudes. Our results show that
date genes whose transcripts increased in abundance in re- both genotypic adaptation and phenotypic plasticity can en-
sponse to hypoxia acclimation. hance V˙O2max in hypoxia in deer mice. Although previous
findings have suggested that both ancestry and environment
DISCUSSION cancontributetoV˙O2maxinhigh-altitudedeermice(9,10,12,
Deer mice at high altitudes sustain high metabolic rates to 52), we are the first to use F1 generation mice to clearly
support locomotion and thermogenesis (24) and are under distinguishtheirrelativeeffects.Therewasnostatisticalinter-
strongdirectionalselectionforanincreasedaerobiccapacityin action between the effects of highland ancestry and hypoxia
hypoxia (25). Our results show that high-altitude adaptation acclimation on V˙O2max in hypoxia (Fig. 1), suggesting that
has promoted an enhanced rate of maximal oxygen consump- genotypic adaptation has not changed the overall functional
htiiognhla(Vn˙dO2amnacxe)stirny.hHyyppooxxiaiaianccFli1mgaetinoenraatlisoonendheearncmedicVe˙Ow2mitahx oaluttictuodmeehyopfoxthiae.aLnikceestthraelinaccrceliamseatiinzaV˙tiOo2nmarxe,stphoendserytmoahsisgohf-
in hypoxia by a similar magnitude in both highland and the lungs increased in parallel after hypoxia acclimation in
lowland mice and should have, thus, facilitated the overall highland and lowland deer mice (Fig. 2), suggesting that each
process of high-altitude adaptation in this species. The effects population increased the amount of lung tissue available for
of highland ancestry on aerobic capacity were associated with gas exchange to enhance oxygen uptake in hypoxia (23, 31).
an increase in the capillarity and oxidative capacity of the Genotypic adaptation to high altitudes appears to have
locomotorymuscle,possiblydue,inpart,toanincreaseinthe bluntedtheincreasesinlungwater,bloodhemoglobincontent,
expression of PPAR(cid:3). Both genotypic adaptation and pheno- andhematocritthatoccurwithhypoxiaacclimationinlowland
typic plasticity, therefore, contribute to aerobic performance deermice(Fig.2).Theformermayindicatethathighlanddeer
(and presumably fitness) at high altitudes, but each process is mice are less prone to developing pulmonary edema in hyp-
Table 2. Fiber types in the gastrocnemius muscle of deer mice
Variable Env. Lowlanders Highlanders MainEffectVariable F P
A (IIa,m) Norm 0.288(cid:9)0.034 0.341(cid:9)0.032 Anc. 2.524 0.122
A
Hyp 0.285(cid:9)0.018 0.314(cid:9)0.009 Env. 0.340 0.564
N (IIa,m) Norm 0.340(cid:9)0.031 0.390(cid:9)0.028 Anc. 0.389 0.537
N
Hyp 0.366(cid:9)0.023 0.347(cid:9)0.012 Env. 0.119 0.733
TypeIarea Norm 1025(cid:9)58 1523(cid:9)90* Anc. 23.61 (cid:4)0.001
Hyp 1141(cid:9)77 1362(cid:9)64 Env. 0.087 0.770
TypeIIaarea Norm 1393(cid:9)86 1691(cid:9)80 Anc. 7.067 0.012
Hyp 1350(cid:9)95 1548(cid:9)108 Env. 0.992 0.327
TypeIIbarea Norm 2185(cid:9)169 2483(cid:9)124 Anc. 0.157 0.695
Hyp 2325(cid:9)137 2145(cid:9)168 Env. 0.442 0.511
Transverseareaofeachfibertype(slowoxidative,typeI;fastoxidative,typeIIa;fastglycolytic,typeIIb)arereportedinsquaremicrometers.A (IIa,m),areal
A
densityoffastoxidativefibers;N (IIa,m),numericaldensityoffastoxidativefibers;Norm,normoxia;Hyp,hypoxia.Onedegreeoffreedomforeachmaineffect
N
variable(Anc.,altitudeofancestry;Env.,acclimationenvironment)and33fortheresidual.*Significantpairwisedifferencebetweenhighlandersandlowlanders
withinanacclimationenvironment.
AJP-RegulIntegrCompPhysiol•doi:10.1152/ajpregu.00362.2014•www.ajpregu.org
R786 HIGH-ALTITUDEADAPTATIONINDEERMICE
3and4).Themagnitudeofthisincreaseincapillarityappears
tobegreaterthantheincreaseinabundanceofoxidativefibers
(Fig. 6). This would suggest that oxygen diffusion capacity is
furtherenhancedtoimproveoxygentransportinhypoxiaifthe
variation in oxidative fiber density is reflective of similar
variation in muscle oxidative capacity. The comparable popu-
lationdifferencesinthearealdensityofoxidativefibersandthe
activityofCS(butnotCOX)inthegastrocnemiusmuscle(Fig.
7)suggestthatthismight,indeed,bethecase.Surprisingly,the
increase in capillarity was not associated with highland deer
micehavingahigherV˙O2maxinnormoxia(Fig.1),despitethe
strong theoretical advantage that a high oxygen diffusion
capacity should impart for V˙O2max across a wide range of
environmental oxygen tensions (8, 56, 69). This distinction
underscores the complexity and interactions between the var-
ioussystems-leveladjustments,potentiallyfromacrosstheO
2
cascade (i.e., ventilation, pulmonary O diffusion, circulation,
2
tissue O diffusion, O utilization), which could improve
2 2
V˙O2max in hypoxia. The best understood high-altitude adapta-
tionindeermice(andinmanyotherspecies)isanincreasein
inherenthemoglobin-oxygenaffinity(10,48,61,62),whichis
generally presumed to be adaptive under severe hypoxia be-
causeithelpspreservearterialoxygensaturationinspiteofthe
reduced alveolar and arterial PO2, thereby minimizing decre-
ments in peripheral oxygen delivery. In normoxia, however,
when arterial blood is fully saturated, a high hemoglobin-
oxygen affinity will reduce the mean capillary O tension (the
2
drivingforceforO2diffusion)andcouldacttoreduceV˙O2max
(4,42,66,67).Therefore,thecombinationofahighhemoglo-
bin-oxygenaffinityandanincreaseinoxygendiffusioncapac-
ity in the muscle would act synergistically to improve V˙O2max
in hypoxia but could act in opposition and have no effect on
Fvaigri.a6ti.oHniignhmlaunsdcldeeeorximdaictievehapvheenaogtyrepaet.eAr:cathpeilrlearwitaystahasntroenxgpelcinteedarfrcoomrretlhae- V˙O2max in normoxia.
High-altitudeadaptationandmusclephenotype.Thehighly
tionbetweencapillarysurfacedensity(CSD)andthearealdensityofoxidative
fibers[A (ox,m)]inthegastrocnemiusmuscle[CSD(cid:8)0.0699A (ox,m)(cid:13) oxidative phenotype of skeletal muscle in highland deer
A A
0.00591; P (cid:4) 0.001]. Symbols are as described in Fig. 1, and dashed lines mice is consistent with previous findings in some high-
represent the 95% confidence interval of the regression. B: there was a altitude bird species (33, 44, 55) and could have multiple
significant effect of altitude of ancestry on the residual CSD from the
potential benefits at high altitudes. Highland animals must
regression in A (F (cid:8) 4.78; P (cid:8) 0.036), reflected by a greater overall
1,32
residualCSDinhighlandersthaninlowlanders(*),buttherewasnosignificant generally cope with colder temperatures than lowland ani-
effectofhypoxiaacclimation(F (cid:8)1.91;P(cid:8)0.177). mals, so a more oxidative muscle phenotype could have
1,32
evolved to enhance the capacity for shivering and possibly
oxia, which is counterproductive to gas exchange and can nonshiveringthermogenesis(46).Ahighlyoxidativemuscle
resultfromhypoxicpulmonaryvasoconstrictionandhyperten- phenotype may also be a strategy for increasing the total
sion in lowland species that sojourn to high elevation (6, 41). mitochondrialO fluxofanentiremusclewhenintracellular
2
Thiswouldbeconsistentwithsomeotherhighlandmammals, O tensions fall (29, 57). Mitochondrial respiration can
2
in which the degree of hypoxic pulmonary hypertension is become O limited during intracellular hypoxia (20), which
2
modest compared with lowlanders (18, 39). Alternatively, the would limit the maximum attainable respiration of individ-
blunted increase in lung water content in highlanders may ual muscle fibers. The greater abundance of oxidative fibers
indicatethattheirlungscontainproportionallylessblood.The in highland deer mice could, therefore, represent a surplus
attenuated increase in blood hemoglobin content and hemato- aerobic capacity in normoxia, but an important counterbal-
crit after hypoxia acclimation in highlanders is also consistent ance for the inhibitory effects of hypoxia in individual
withfindingsinmanyotherhighlandtaxa(5,47,59)andmay muscle fibers.
havearisenbecauseastrongerythropoieticresponsecanover- The oxidative muscle phenotype of highland deer mice
shoot the optimal blood hemoglobin content at elevation, occurs in conjunction with an increase in the activity and
increase blood viscosity, and lead to cardiovascular complica- expression of enzymes involved in oxidative phosphorylation,
tions. These patterns are consistent with countergradient vari- the tricarboxylic acid cycle, and fatty acid oxidation (11, 13).
ation, in which genotypic adaptation opposes the ancestral The magnitude of the differences between populations was
acclimation response, and have likely arisen because the an- comparable ((cid:2)25%) for the areal density of oxidative fibers
cestral responses are maladaptive at high altitudes (63). (Fig. 5), citrate synthase activity, and lactate dehydrogenase
The higher hypoxic V˙O2max in highland deer mice is asso- activity (Fig. 7). The change in fiber-type composition in the
ciated with greater capillarity in the locomotory muscle (Figs. gastrocnemius muscle of highland mice may, therefore, drive
AJP-RegulIntegrCompPhysiol•doi:10.1152/ajpregu.00362.2014•www.ajpregu.org
HIGH-ALTITUDEADAPTATIONINDEERMICE R787
60 Lowlander
120
Highlander
y
astrocnemius COX activitμ-1-1(mol g tissue min) 23450000 * * Diaphragm COX activityμ-1-1(mol g tissue min) 10246800000 * * Fig.7.Oxidativecapacityinthegastrocnemius
G
muscleanddiaphragm.Therewasasignificant
effectofaltitudeofancestryontheactivitiesof
10 0
cytochromecoxidase(COX;F (cid:8)16.52;P(cid:4)
1,30
70 * 120 0.001),citratesynthase(CS;F1,30(cid:8)15.88;P(cid:4)
y 0.001),andlactatedehydrogenase(LDH;F1,30(cid:8)
astrocnemius CS activitμ -1-1(mol g tissue min) 456000 * Diaphragm CS activityμ -1-1(mol g tissue min) 1101890000 Ldao201b0hiuf..3.yDa102st.ppi4056HCgnoh;129onxOr;;)aFPii,efigXPaP1famc,f(cid:8)3an(cid:4)e(cid:8)a0,(ccdnFbtc(cid:8)0t00u1tlo.hi.e.,1tf03m92fe53n0h8fr.a3oe3y1e0t(cid:8)ct)9p)i;wtooo;oiCnnnxoPer1Sifra8tt(Lehh(cid:8)CaF.ae9DenncO14c0aogc,H3;Xcel.a0i1tssmsiP(3it(cid:8)vtFFgrra3iytont1i(cid:4))i1e,cio.3ofis.nn3T9nceo06h(cid:8)((cid:8)amftC.;enh0CPriOte0u31eS1eXs..(cid:8)a12wf)c(mf86FFate0i;;iu1s1cnv.,,tPsP313ias30ct7tlilhs(cid:8)o(cid:8)e(cid:8)(cid:4)2eoesf;,
G 0.084;CSF (cid:8)2.44;P(cid:8)0.11,2338;LDHF (cid:8)
1,33 1,33
30 70 0.578;P(cid:8)0.452).*Significantpairwisediffer-
ence between highlanders and lowlanders
withinanacclimationenvironment(P(cid:4)0.05).
180
Normoxic lowlanders, n (cid:8) 7/7; hypoxic low-
H activity -1min) 600 * activity -1min) 160 l9na/en1md0ei;urhssy,/dpnioax(cid:8)pihcr7ah/g1igm0h;,lanrenosdrpmeerocs,xtiinvce(cid:8)lhyi)g1.h1l/a1n0d(egras,stnro(cid:8)c-
strocnemius LDμ-1(mol g tissue 450000 Diaphragm LDH μ-1(mol g tissue 112400
a
G
300 100
Normoxia Hypoxia Normoxia Hypoxia
Acclimation Environment Acclimation Environment
an increased capacity for substrate oxidation and a reduced protein abundance of PPAR(cid:3) (Figs. 8 and 9), consistent with
capacityforanaerobicglycolysis.Incontrast,themagnitudeof recent findings that the pparg gene is a strong target of
the increase in cytochrome c oxidase activity in highlanders selectioninhigh-altitudeTibetanandMongolianpeoples(74).
wasgreaterthantheincreaseinoxidativefiberdensity,andthe AlthoughPPAR(cid:3)expressionisgenerallylowerinmusclethan
ratioofCOX/CSactivitywasalsogreaterinhighlanders.This in adipose tissue, activation of muscle-specific PPAR(cid:3) can
excesscapacityofCOXshouldhaveimportantimplicationsfor promote insulin sensitivity, shift the skeletal muscle toward a
mitochondrialfunctioninhypoxiabecauseitwouldalloweach moreoxidativephenotype,andincreaseCOXexpressioninthe
COXenzymetooperateatalowerturnoverrate,whichwould muscle (2, 28). This could result from PPAR(cid:3)-induced pro-
helpsustainmitochondrialrespirationatloweroxygentensions duction of adiponectin and a corresponding increase in mito-
(i.e., it would increase mitochondrial oxygen affinity) (21). chondrial biogenesis (1, 2, 50). PPAR(cid:3) activation has also
Unlike the appreciable differences between populations, beenshowntoincreasecapillarityinadiposetissue(19),sothe
hypoxiaacclimationhadonlymodestandstatisticallyinsignif- upregulation of PPAR(cid:3) that we observed in the locomotory
icanteffectsonmusclecapillarityandoxidativecapacity.This muscleofhighlanderscouldhavecontributedtotheheightened
is consistent with many (but not all) previous studies of capillarity in this population. In contrast, transcripts of many
high-altitudeacclimationinlowlandmammals,inwhichmus- other regulators of energy metabolism (i.e., most other PPAR
cle capillarity and oxidative capacity do not change or even and PGC transcripts measured) declined in hypoxia and were
decrease in response to hypoxia (34, 43). expressedatsimilarlevelsbetweenpopulations(Fig.8),coun-
Musclephenotypeandgeneexpression.Thehighlycapillar- ter to the general expectation that these regulators will be
ized and oxidative phenotype of the skeletal muscle of high- expressed at higher levels in muscles enriched with slow
landers was associated with elevated mRNA expression and oxidative fibers (38).
AJP-RegulIntegrCompPhysiol•doi:10.1152/ajpregu.00362.2014•www.ajpregu.org
R788 HIGH-ALTITUDEADAPTATIONINDEERMICE
A
2.0 Lowlander
s Highlander
el
v
e 1.5
pt l
cri
ns 1.0
a
e tr
v
ati 0.5
el *
R
0.0
Vegfa Angpt1 Angpt2 Flt1 Kdr Tek
B
3.0
*
s *
el 2.5
v
e
pt l 2.0
cri
ns 1.5
a
e tr 1.0
v
ati
el 0.5
R
0.0
Ppargc1a Ppargc1b Pprc1 Ppara Ppard Pparg Nrd1d1
Fig.8.Expressionofcandidategenesinvolvedinregulatingangiogenesis(A)andenergymetabolism(B).Reactionnormsforgeneexpressionineachacclimation
environmentareshownwithnormoxiaontheleftandhypoxiaontheright.Therelativeabundanceofvascularendothelialgrowthfactor(VEGF)A(Vegfa)and
peroxisomeproliferator-activatedreceptor(PPAR)(cid:3)(Pparg)transcriptswassignificantlyaffectedbyaltitudeofancestry(VegfaF (cid:8)13.56;P(cid:8)0.002;Pparg
1,20
F (cid:8)14.48;P(cid:8)0.001).Expressionofseveralgenesdecreasedinresponsetohypoxiaacclimation,includingVegfa,PPAR(cid:11)(Ppara),PPAR(cid:12)(Ppard),PPAR(cid:3)
1,20
coactivator(PGC)1(cid:7)(Ppargc1b),PGC-relatedprotein1(Pprc1),andnuclearreceptorsubfamily1groupDmember1(Rev-ErbA(cid:11);Nrd1d1).Theexpression
oftheremainingcandidategeneswereunaffectedbybothancestryandacclimationenvironment,includingangiopoietin1(Angpt1),angiopoietin2(Angpt2),
VEGF receptor 1 (Flt1), VEGF receptor 2 (Kdr), angiopoietin receptor (Tek), and PGC-1(cid:11) (Ppargc1a). Two-factor ANOVA results are shown in Table 2.
*Significantpairwisedifferencebetweenhighlandersandlowlanderswithinanacclimationenvironment(P(cid:4)0.05).†Significantmaineffectofacclimation
environment(seeTable3).Normoxiclowlanders,n(cid:8)7;hypoxiclowlanders,n(cid:8)4;normoxichighlanders,n(cid:8)7;hypoxichighlanders,n(cid:8)6.
The highly capillarized muscle of highlanders was not as- pika (Ochotona curzoniae) (35). It is possible that the higher
sociated with a persistent upregulation of genes involved in muscle capillarity in highland mice is initiated by enhanced
angiogenesis in the muscle (22). In fact, the combined tran- VEGFexpressioninearlydevelopment,butwehadanticipated
script abundance of all vegfa splice variants declined in hyp- that persistent differences in VEGF expression might contrib-
oxia and was lower in highlanders (Fig. 8). This is consistent
withpreviousobservationsthatchronichypoxiareducesVEGF
expressionintheskeletalmuscleofrats(49),butitcontrastsa 4 Lowlander
recent suggestion that high expression of some VEGF splice Highlander
variants is associated with highland residency in the plateau
e
anc 3 *
Table 3. Two-factor ANOVA results for gene expression d
n
data bu
a
AltitudeofAncestry AcclimationEnvironment ein 2
ot
Transcript F P F P pr
γ
R
Angpt1 2.734 0.116 0.599 0.449 A
Angpt2 0.229 0.638 0.018 0.895 P 1
P
Flt1 0.083 0.776 0.736 0.401
Kdr 0.059 0.810 0.835 0.371
Nrd1d1 0.177 0.679 10.80 0.004
Ppara 0.014 0.907 5.650 0.028 0
Pparg 14.48 0.001 0.106 0.748
Normoxia Hypoxia
Ppargc1a 0.175 0.680 0.232 0.636
Ppargc1b 0.768 0.391 8.454 0.009 Acclimation Environment
Ppard 1.398 0.250 6.446 0.019
Pprc1 8.047 0.545 8.047 0.011 Fig. 9. Highland deer mice have higher PPAR(cid:3) protein abundance in the
Tek 0.128 0.724 3.385 0.081 gastrocnemiusmuscle.TherewasasignificantmaineffectonPPAR(cid:3)protein
Vegfa 13.56 0.002 7.502 0.014 abundance of both altitude of ancestry (*F1,24 (cid:8) 5.36; P (cid:8) 0.030) and
acclimation environment (†F (cid:8) 6.25; P (cid:8) 0.020), with no significant
1,24
Onedegreeoffreedomforeachmaineffectvariable(altitudeofancestry, interaction between each factor (F (cid:4) 0.001; P (cid:8) 0.982) (n (cid:8) 7 for all
1,24
acclimationenvironment)andtheirinteraction,and20degreesoffreedomfor groups).Insetsarerepresentativeimmunoreactivebandsforahighlander(left)
theresidual. andalowlander(right),eachacclimatedtohypoxia.
AJP-RegulIntegrCompPhysiol•doi:10.1152/ajpregu.00362.2014•www.ajpregu.org