Evidence for independent evolution of functional progesterone withdrawal in primates and guinea ...

16 Pages·2013·0.88 MB·English

273
original
Evolution,Medicine,andPublicHealth[2013]pp.273–288 research
article
doi:10.1093/emph/eot022
Evidence for independent
evolution of functional
progesterone withdrawal in
primates and guinea pigs
Mauris C.Nnamani1, Silvia Plaza1,Roberto Romero2,3,4andGu¨nter P.Wagner*1,5
1YaleSystemsBiologyInstituteandDepartmentofEcologyandEvolutionaryBiology,YaleUniversity,NewHaven,
CT,USA;2PerinatologyResearchBranch,ProgramforPerinatalResearchandObstetrics,DivisionofIntramuralResearch,
EuniceKennedyShriverNationalInstituteofChildHealthandHumanDevelopment,NIH,Bethesda,MD,USA;
3DepartmentofObstetricsandGynecology,UniversityofMichigan,AnnArbor,MI,USA;4DepartmentofEpidemiology
andBiostatistics,MichiganStateUniversity,EastLansing,MI,USA;5DepartmentofObstetricsandGynecology,
WayneStateUniversity,Detroit,MI,USA
*Correspondenceaddress.YaleSystemsBiologyInstitute,YaleUniversityWestCampus,AdvanceBiosciencesCenter,
Room272B,POBOX27388,WestHaven,CT06516-7388,USA.Tel:þ12037373091;Fax:þ12037373109;
E-mail:[email protected]
Received8May2013;revisedversionaccepted21November2013
ABSTRACT
Backgroundandobjectives:Cervixremodeling(CRM)isacriticalprocessinpreparationforparturition.
Early cervix shortening is a powerfulclinical predictor ofpreterm birth, andthus understanding how
CRMisregulatedisimportantforthepreventionofprematurity.Humansandotherprimatesdifferfrom
most other mammals by the maintenance of high levels of systemic progesterone concentrations.
Humanshavebeenhypothesizedtoperformfunctionalprogesteronewithdrawal(FPW).Guineapigs
are similar to humans in maintaining high-progesterone concentrations through parturition, thus
making them a prime model for studying CRM. Here, we analyze the phylogenetic history of FPW
anddocumentgeneexpressionintheguineapiguterinecervix.
Methodology: Data on progesterone withdrawal were collected from the literature, and character
evolution was analyzed. Uterine cervixsamples were collected from non-pregnant, mid-pregnant and
latepregnantguineapigs.RNAwasextractedandsequenced.Relativetranscriptlevelswereestimated
andcomparedamongsamplegroups.
Results: The phylogenetic analysis shows that FPW evolved independently in primates and guinea
pigs. The transcriptome data confirms that guinea pigs down-regulate progesterone receptor toward
parturition,incontrasttohumans.Someofthesimilaritiesbetweenhumanandguineapigare:down-
regulation of estrogen receptor, up-regulation of VCAN and IGFBP4 as well as likely involvement
ofprostaglandins.
(cid:2)TheAuthor(s)2013.PublishedbyOxfordUniversityPressonbehalfoftheFoundationforEvolution,Medicine,andPublicHealth.ThisisanOpen
AccessarticledistributedunderthetermsoftheCreativeCommonsAttributionLicense(http://creativecommons.org/licenses/by/3.0/),whichpermits
unrestrictedreuse,distribution,andreproductioninanymedium,providedtheoriginalworkisproperlycited.
274 | Nnamanietal. Evolution,Medicine,andPublicHealth
Conclusions and implications: (i) FPW in guinea pigs evolved independently from that in primates.
(ii)Asmallsetofconservedgeneregulatorychangeshasbeendetected.
KEYWORDS: guinea pig cervix; gene expression; functional progesterone withdrawal; evolution of
parturition
INTRODUCTION
Cervical remodeling (CRM) is a necessary step In this communication we investigate gene
in preparation for successful parturition [1]. expression in the cervix of guinea pigs to assess
Premature softening and shorting of the cervix the similarity between human and guinea pig
is also a powerful predictor of preterm birth, CRM.Ouraimistoanswerthequestionofwhether
andthusunderstandingthemolecularmechanisms FPW in primates and guinea pigs is homologous,
regulatingCRMarecriticalforsuccessfulinterven- meaning that it was already present in the most
tion to prevent prematurity [2]. A major obstacle recent common ancestor of humans and guinea
in unraveling the mechanisms of human cervical pig. This question is important because molecular
ripening is that mammals are highly variable with mechanismsaremorelikelysharedamonghomolo-
respect to the mechanisms underlying parturition. gouscharactersthanamongcharactersthatevolved
Most notable is the fact that in humans and other independently.Guineapigbelongstoabasalrodent
primates the placenta produces large amounts of lineage[14],anditisthuspossiblethatFPWinpri-
progesteronethroughpregnancyandlabor.Incon- matesandguineapigscouldbehomologous.
trast, most other placental mammals, including Here, we report that there are extensive differ-
model organisms such as sheep, mouse, rabbit ences between the human and guinea pig cervical
and rat, have systemic progesterone withdrawal, geneexpressiondynamics,whichmakesitunlikely
meaningthattheserumconcentrationofprogester- that the mechanisms of ‘FPW’ are homologous
one declines toward term before the onset of between the two species. This inference is also
labor. Declining progesterone concentrations are supported by a phylogenetic analysis of serum
thought to remove the ‘progesterone block’ from progesterone concentrations at parturition among
theuterusandthecervixandthusallowprogression Archontolgires, the clade uniting primates and
toward parturition [3–5]. To explain parturition rodents.Nevertheless, itisstillpossible thatthere
in humans, it has been suggested that there have are conserved molecular mechanisms shared
to be changes downstream of the progesterone between primates and other placental mammals,
signal that undercut the progesterone block in its as for instance the down-regulation of estrogen
target tissues. Several mechanisms have been receptoralpha(ESR1),asreportedinourprevious
proposedforthisso-calledfunctionalprogesterone paperforhumans[17]andhereforguineapigs.
withdrawal(FPW)[6–9],butnoneofthemhaveyet
receiveddecisivesupport.
METHODOLOGY
A notable exception among model organisms
is the guinea pig that, like the human, maintains
Tissue harvesting and RNA extraction
high concentrations of serum progesterone
through parturition [10]. It has thus been sug- Hartley guinea pigs (pregnant and non-pregnant)
gested that guinea pig is a potential model organ- were obtained from a commercial supplier
ism to study the mechanisms of FPW as well as (CharlesRiver,Wilmington,MA,USA).Theanimals
prematurity, which is very rare in model species were housed in individual plastic cages with hay,
withfastgestationlikemice[11].Thisisanattract- branchesandotherenvironmentenhancingobjects.
ive possibility, as guinea pigs belong to the same Non-pregnant females were examined daily for
clade as other major model organisms, mice, rat vulva occlusion membrane. These animals were
andrabbit,theso-calledGlires[12–15].Itisalsoa sacrificed the day after the vulva occlusion dis-
model organism with a long tradition of experi- appeared indicating entry into estrus. Pregnant fe-
mental research [16] and a sequenced genome maleswereobtainedwithuncertainpregnancydates
(http://useast.ensembl.org/Cavia_porcellus/Info/ and were euthanized based on assessing overall
Index). body weight of the female with gestation range
Independentevolutionoffunctionalprogesteronewithdrawal Nnamanietal. | 275
providedbythesuppliers.For‘post-mortem’assess- In comparing transcript abundances we distin-
ment of the duration of pregnancy, we took meas- guishtwodifferentkindsofevents.Attheonehand
urementsofboththefetalsupra-occipital-premaxilla are induction and repression of gene expression
length and fetal weight at harvest (Supplementary (turninggeneexpressionONorOFF).Ontheother
Table S1). To estimate gestational age we used handaremodulationsofgeneexpression.Werefer
previouslypublisheddata[18–20].Estimatedgesta- to induction when the transcript abundance of a
tionalstageformid-pregnantandlatepregnant(but gene is below the operational threshold of three
not in labor) animals was 43.6±0.5 (mean±SD) TPM[26,27]intheearlierstageofthereproductive
days and 65.2±1.2days, respectively (average cyclebutabovethethresholdinthelaterstage.The
gestation period of 65.5±6.5days range [16]). thresholdofthreeTPMcorrespondsto(cid:3)1RPKMin
Tissueharvestingfocusedonthevaginalportionof terms of the traditional RNA abundance measure
the cervix to minimize contamination with uterine [26,27].Thethreshold isbasedon association be-
myometrial tissue. Samples were collected from tweenexpressionlevelandchromatinmodification
the three groups in triplicates for non-pregnant status as well as a statistical model of transcript
and mid-pregnant females and four replicates abundance.Incontrast,wecallchangesoftranscript
forlatepregnantfemales.Cervicaltissuewasimme- abundance‘expressionmodulation’iftheestimated
diately stored in RNAlater Solution (Ambion, cat# transcript abundance is above the operational
AM7020) until further processing. Total RNA was thresholdinbothstagescompared.Whenwerefer
extracted using the RNeasy extraction kit (Qiagen, toupordownregulation,wemeangeneexpression
cat# 75142) following manufactures instructions modulation.
that included an in-column DNAse digestion. Themotivationformakingthe(operational)dis-
Quality of the RNA was then confirmed using tinction between gene expression modulation and
Agilent2100Bioanalyzer(SantaClara,CA,USA). induction/repressionisthattheyrepresenttwodif-
ferent biological events. Induction/repression is
associated with qualitative changes in the nature
Sequencing and data processing
ofthechromatinmodificationandthekindoftran-
RNA library preparation and high-throughput scriptionfactorsandco-factorsassociatedwiththe
sequencing were performed on an Illumina HiSeq cis-regulatoryelements[27].Ontheotherhand,gene
2000sequencingsystemfollowingtheprotocolrec- expressionmodulationcanhaveavarietyofcauses,
ommended by Illumina for sequencing total RNA fromphosphorylationstatusoftranscriptionfactors
samples. Sequencing was done for each biological totheexpressionlevelofupstreamregulators.
replicates at 1(cid:2)75bp strand specific by the Yale
Center for Genome Analysis. Sequence reads were
Statistics
aligned to the Cavia porcellus reference genome
(cavPor3.69) using the splice junction mapper for When correlations were calculated or differential
RNA-seqreadsTopHat2[21–23].Sequencingdepth expression was tested, we transformed TPM data
forRNA-seqsamplesaveraged45millionreadsper bysquareroottransformationratherthanthemore
biologicalsamplewith>80%overallalignmentrate. widely used log-transformation for two reasons.
Afteralignment,readcountsweredeterminedwith First, TPM as well as RPKM data usually contains
HTSeq-countv0.5.4p1asdescribedbytheauthors. zero values. In a log-transformation these data
AllRNA-seqdataaredepositedinGEOunderacces- pointswillbetransformedintominusinfinity,with
sionnumberGSE47986. consequencesfordatahandlingandinterpretation.
p
ffiffiffi
Incontrast, 0¼0andnofurtherproblemsarise.
Thesecondreasontoprefersquareroottransform-
Data analysis
ationisthatitisinfactvariancestabilizing[26],while
Relative RNA abundance was measured as tran- thelog-transformationleadstoaninflationofvari-
scripts per million (TPM) transcripts as recom- anceatlowabundancevalues.Testsfordifferential
mended in [24, 25] rather than RPKM or FPKM. expressionwerenotusedasadiscoverytoolbutto
ThereasonisthatunitsofRPKMarenotconsistent testspecifichypotheses.Forinstant,wetesteddif-
between samples and are thus problematic when ferences between stages forthosegenes thathave
comparing RNA abundance between samples of beenfounddifferentiallyexpressedinhumans.For
differenttranscriptcomposition[25]. this reason, we did not use multiple comparison
276 | Nnamanietal. Evolution,Medicine,andPublicHealth
corrections.Weusedone-wayANOVAandt-testson downstream of the progesterone signal is respon-
squareroottransformedTPMdata. sibleforCRM,calledFPW.Itisthusinterestingto
askwhetherFPWintheguineapigishomologousto
thatinhumans.
BrdU labeling
In Fig. 1a and b, a survey of progesterone with-
Three guinea pigs 35–38weeks pregnant for drawal within Euarchontoglires is presented based
mid-pregnant and also three females 58–63weeks onthephylogenetichypothesisin[28].Lackofsys-
pregnant for late pregnancy were obtained from temicprogesteronewithdrawalisfoundinhumans,
Charles River laboratories. Two females from each apes(chimpanzee[29],gorilla[30,31])andoldworld
group were injected intraperitoneal with 10ml of monkeys(rhesusmonkey[32,33],baboon[34]).The
Bromodeoxyuridine (BrdU) labeling reagent Life situation in new world moneys is complicated.
Technology (Cat # 0103), and one control sample ThedetailedstudybyChambersandHearnonthe
with10mlPBS.After24htheanimalwassacrificed, common marmoset (Callithrix jacchus) shows a
andcervixwasrecoveredandfixedin4%paraform- steady decline in serum progesterone levels in the
aldehyde (24–48h). The tissue was processed for 4daysleadinguptoparturition[35].Otherauthors
paraffinembeddingandserialsectionof6mmwas didnotreportsuchadropbutalsodidnotexplicitly
produced.Theslideswereprocessedfollowingthe document the peripartum period [36]. In contrast,
LifeTechnologyprotocolforBrdU(Cat#93-3943). Corousos and collaborators report the peripheral
progesterone levels of two pregnant squirrel mon-
keys (Saimiri sciureus) but do not show a decline
RESULTS
of progesterone levels in the days leading up to
parturition [37], even though the levels at the day
Evolution of functional progesterone
of parturition are reported to be close to zero.
withdrawal
Nevertheless,itisclearthatthewithdrawal,ifany,
The guinea pig has attracted interest because wouldhavetobeveryprecipitousorabsent.Based
it shares with humans the feature of maintaining on comparison with the rate of decline in the
high concentrations of serum progesterone (P4) marmosetweconcludedthatthesquirrelmonkeys
(>200ng/ml)throughparturition[10].Forbothspe- donothavesystemicprogesteronewithdrawal.We
cies it has been suggested that some mechanism couldnotfinddataabouttarsiersandsomelimited
Figure 1. EvolutionofFPW.PhylogeneticreconstructionofserumprogesteroneconcentrationatparturitioninEuarchontoglires,thecladeunitingprimatesand
rodents.Speciesandlineageswithsustainedhighsystemicprogesteroneconcentrationsareindicatedinwhite,speciesandlineageswithadropofsystemic
progesteroneconcentrationsareindicatedinblack.Therootstatewasfixedasprogesteronewithdrawalbasedonoutgroupcomparisonwithhoofedanimalsand
carnivores.TreetopologyinGliresfollowsfig.6in[28].(a)Scenariothatassumesthatthemouselemurhasprogesteronewithdrawal.Underthisassumptionthe
lackofsystemicprogesteronewithdrawalisanancestralstateofprimatesandeventheEuarchonta.(b)Scenariothatassumesthatthemouselemurdoesnothave
progesteronewithdrawal.UnderthisscenariotheancestralstateforEuarchontaandprimatesisambiguous.Notethatinbothscenariosguineapigistheonly
lineagewithoutprogesteronewithdrawalwithintheGliresandthatguineapigisdeeplynestedwithinthisclade.ThisphylogeneticdistributionsuggeststhatFPW
evolvedindependentlyinprimates(Euarchonta)andguineapigwithintheGlires.
Independentevolutionoffunctionalprogesteronewithdrawal Nnamanietal. | 277
informationaboutlemurs.BlancoandMeyerreport were mapped to the guinea pig genome version
progesteronelevelsinvariousstagesofestrusand cavPor3.69 and read counts transformed to TPM
pregnancy of the brown mouse lemur (Microcebus transcripts[25].Belowwefocusontwocomparisons
rufus)butthegestationalstagesinthisstudyarevery oftranscriptabundance.Oneisthedifferencesbe-
unreliable [38]. In spite of that, the pre-partum co- tween non-pregnant and mid-pregnant stages and
hortshowshigherprogesteronethanthepre-estrus theotherthecomparisonofmid-pregnancyandlate
cohort,makingitpossiblethatthemouselemurlack pregnancy.
systemicprogesteronewithdrawal.Thus,weprovide
tworeconstructionsonewithandonewithoutsys-
Gene expression changes in the cervix from
temicprogesteronewithdrawalforthemouselemur
estrus to mid-pregnancy
(Fig.1aandb).
An outgroup clade of primates are the Tupaias Inthecomparisonfromnon-pregnantandmid-preg-
(Scandentia). In Tupaia belaneri, there is a drop in nantcervices,195genescrosstheoperationalcriter-
serum P4 only after parturition [39] but this paper ion from being repressed to being expressed.
reports unpublished data from Elger and Hasan Figure2ashowstheexpressionlevelsofgenesop-
aboutasingleanimal.However,thepeakprogester- erationallynon-expressedinestrusbutexpressedin
onelevelisreportedtooccur2daysbeforepartur- mid-pregnancy. Two genes stand out, CLCA1 and
ition and wethus classify it ashaving no systemic PLA2G10 (see Supplementary Table S2). CLCA1 is
progesterone withdrawal. Within the Glires, the a component of Ca2þ sensitive chlorine channels
clade including rodents and rabbits and squirrels, andisinvolvedinmucusproductioninotherorgans
all lineages have progesterone withdrawal except [42].PLA2G10isaphospholipasecatalyzingtherate-
guineapigs[40,41]. limiting step in prostaglandin synthesis [43]. The
InFig.1aandb,weprovidetwoparsimony-based other highly expressed and induced genes were:
ancestral character state reconstructions reflecting twomembersofthesodium-dependentdecarboxyl-
the ambiguity about the situation in the lemur. asetransporters(SLC13A2andSLC36A2),IGFBP1,
Under either scenario the lack of systemic proges- aswellasatumornecrosisfactor,TNFSF11.
teronewithdrawalintheguineapigisreconstructed There are 428 genes expressed during estrus
asindependentlyderived(Fig.1aandb).Thediffer- butoperationallyOFFatmid-pregnancy.Figure2b
ence between the two scenarios, lemurs with or shows their expression level during estrus. There
without progesterone withdrawal, only affects the are 11 genes highly expressed during estrus but
reconstruction of the ancestral character state in operationally turned OFF in mid-pregnancy with a
the primate lineage (Euarchonta). Assuming that discontinuity of lower expression level (big bold
thelemurhasnosystemicprogesteronewithdrawal arrow in Fig. 2b). The most highly expressed gene
(Fig. 1a) leads to a reconstruction where this con- turnedoffinpregnancyisinterleukin1a(IL1A).The
ditionisancestralfortheEuarchonta,andthesitu- other genes are TMPRSS11D, a trans-membrane
ation in the marmoset is inferred to be a reversal. serineproteaseinvolvedinglandsecretoryactivity,
Assuming that the lemur has systemic progester- interferon kappa, IFNK and an antagonist of IL1A,
one withdrawal (Fig. 1b) leads to a reconstruction IL1RNandothersmore(seelistinSupplementary
where the ancestral state of the stem lineage of Table S3). The gene ontology categories most
primates is ambiguous. From these data we con- over-represented are related to inhibition of cell
clude that FPW likely has evolved independently proliferation.
inprimatesandguineapigs.
Gene expression changes in the cervix from
Gene expression in guinea pig cervix mid- to late-pregnancy
Cervical tissue was harvested from guinea pigs in Inthetransitiontolateterm195genesareturned
threereproductivestages:non-pregnantandines- on (Fig. 2c). The strongest expressed gene, which
trus (NP), in mid-pregnancy (MT) and late preg- was not expressed in mid-pregnancy, is called
nancy/term (LT) (see ‘Methodology’ for details VISG8,V-setandimmunoglobulindomaincontain-
on the timing of sampling). The samples were ing 8. Very little is known about its function
transcribedandsequenceswith1(cid:2)75bptoanaver- (Supplementary Table S4). The second most
age sequencing depth of 40–50 Mio reads. Reads expressedgeneturnedontowardtermisSPC24,a
278 | Nnamanietal. Evolution,Medicine,andPublicHealth
Figure 2. ExpressionlevelsinTPMofgenesthatareturnedonoroffduringpregnancyintheguineapigcervix.(a)Expression
levelofgenesinmid-pregnancyofgenesthatarenotexpressedinnon-pregnantcervixandexpressedinmid-pregnancy.Note
thattherearetwogenesmuchmorehighlyexpressedthantheothers,CLCA1andPLA2G10.(b)Expressionlevelofgenesin
non-pregnantcervixthatarenotexpressedinmid-pregnantcervix.Thereisadistinctdifferenceinexpressionlevelbetweenthe
11highestexpressedgenesandtherest(largearrow).(c)Expressionlevelofgenesinlatepregnancyofgenesnotexpressed
inmid-pregnancy.(d)Expressionlevelofgenesduringmid-pregnancythatarenotexpressedinlatepregnancy.Fourgenes
standout:MGAMisthemaltase-glucoamylaseoralpha-glucosidase,anenzymeinvolvedinstarchdigestion;SCXBthebasic
helix-loop-helix transcription factor; COL9A2 is a type IX collagen mostly known from hyaline cartilage; and CSPG5 is the
chondroitinsulfateproteoglycan5(neuroglycanC).(e)Cervicalstromainmid-pregnancylabeledwithBrdU.Notesmalldense
nucleiwithveryfewBrdUlabeledcells.(f)CervicalstromainlatepregnancylabeledwithBrdU.Thecellnucleiarelessdense
andthereareabundantcellslabeledwithBrdU,consistentwiththefindingthatproliferationrelatedgenesarepreferentially
expressedinlatepregnancy
(continued)
Independentevolutionoffunctionalprogesteronewithdrawal Nnamanietal. | 279
kinetochore complex component, foreshadowing of ARG2 [44–46], FGFBP1 [47, 48] and FOXM1
the result of the gene ontology analysis. The over- [49–53],areassociatedwithseveralcancers.
whelmingmajorityofgenesarerelatedtoenhance- To determine whether the late term cervical
ment of cell proliferation, suggesting a very active stroma has proliferating cells we injected guinea
proliferative state of the guinea pig cervix at term. pigsinmid-andlate-termwithBrdUandharvested
Theseconclusionsaresupportedbytheinductionor thetissue24hlater(Fig.2eandf).Whileatmid-term
strongup-regulationofgenesknowntopromoteor the nuclei of stromal cells are small and dense, in
are associated with tumor growth. For instance, late term the cell nuclei are larger and less dense.
among the induced genes is FOXM1 and among BrdU stained cells are abundant in latter (Fig. 2f)
the strongest up-regulated genes are ARG2, consistentwiththetranscriptomicdatasuggesting
arginase type II and FGFBP1. The high expression higherproliferativeactivitytowardlateterm.
Figure 2. (Continued)
(continued)
280 | Nnamanietal. Evolution,Medicine,andPublicHealth
Figure 2. (Continued)
In the transition to term 449 genes meeting the increasefromestrustomid-pregnantstageandthen
operationalcriterionofexpressioninmid-pregnancy a0.55-folddecreasetowardterm,returningtoabout
are turned OFF toward term. Four of them have the same level as in estrus. A similar dynamics is
higher expression than the rest of the distribution observed for another small proteoglycan, testican
(Fig. 2d). These are: MGAM, SCXB, COL9A2 and 2,SPOCK2,althoughatlowerlevelsofexpression.
CSPG5 (Supplementary Table S5). MGAM is the Itincreases2.1(cid:2)fromestrustomid-pregnancyand
maltase-glucoamylaseoralpha-glucosidase,anen- towardtermisalmostcompletelysuppressed,going
zyme involved in starch digestion; SCXB the basic from 122 TPM at MT to 9 TPM at LT (0.07-fold
helix-loop-helix transcription factor; COL9A2 is a change). Repression of both FMOD and SPOCK2
typeIXcollagenmostlyknownfromhyalinecartilage transcriptionmayplayaroleinextracellularmatrix
andCSPG5isthechondroitinsulfateproteoglycan5 remodeling in preparation for parturition. In add-
(neuroglycan C). Repression of these genes sug- ition, the expression dynamics of FMOD suggests
gestsachangeintheextra-cellularmatrixcompos- lowerefficiency of TGF-beta signaling during preg-
itionduringtransitiontoterm. nancythanbeforeandtowardterm.
Wefoundsubstantialexpressionfortwolargepro-
Expression dynamics of candidate genes teoglycans,versican(VCAN)andperlecan(HSPG2,
heparansulfateproteoglycan2)(Fig.3c).VCANde-
Steroid receptors
creases from estrus to mid-pregnant by 0.38-fold
During estrus the expression of estrogen receptor,
and then increases 1.6-fold again toward term. In
ESR1,ishigh(>240TPM)andisdecreasingto<150
contrast,HSPG2ishighinestrusbutdeclinesslowly
TPM in mid-pregnancy and further decreases to
towardtermto59%fromitsvaluebeforepregnancy.
<100TPMtowardterm.Incontrast,theprogester-
The dominant matrix metallopeptidase mRNA
onereceptor(PR)mRNA,PGR,islowduringestrus
is MMP2 (Fig. 4a), which is known to break down
(27 TPM) and rises to a value of 56 TPM in mid-
collagen IV and gelatin. There is an apparent
pregnancybutfallsclosetopre-pregnancylevelsof
increaseinexpressionofMMP2betweenmid-and
30TPMtowardtheendofpregnancy(Fig.3a).
late-term, but the statistical support is weak
Extracellular matrix (P¼0.056).Thenexthighestexpressionwasfound
ThedominantcollagenmRNAintheguineapigcer- forMMP14,whichisactivatingMMP2andislikely
vix is type 1 collagen, COL1A1 and COL1A2. membrane bound. MMP14 expression in estrus
Expression is high for both type I collagen genes, is (cid:3)340 TPM and decreases to mid-pregnancy by
(cid:3)5000TPMand2000TPM,respectively,anddoes 0.55-fold(P¼0.016)andshowsaslightincreaseof
notchangeamongstagesofthereproductivecycle. 1.18-fold toward late term (P¼0.05). In contrast,
COL3A1,codingforcollagenIIIidentifiedinhuman MMP11 has a moderate expression in estrus (109
cervix,wasnotmappedinourtranscriptomedata. TPM), decreases toward mid-pregnancy by 90%
Thedominantsmallproteoglycanisfibromodulin (0.1-fold)andisbarelyabovetheoperationalthresh-
(FMOD)(Fig.3b),whichisknowntoassociatewith old for expressed genes at term (3.75 TPM). This
typeIcollagenandisalsoabletosequesterTGF-beta MMP11isnoteffectiveinextracellularmatrixbreak-
intheextracellularmatrix.FMODshowsa2.2-fold down, but is reported to cleave protease inhibitor
Independentevolutionoffunctionalprogesteronewithdrawal Nnamanietal. | 281
Figure 3. Expressiondynamicsofsteroidreceptorsandproteoglycans.(a)ExpressionofESR1andPR.ESR1ishighduring
estrusandissteadilydecliningtowardparturition.PGRisup-regulatedinmid-pregnancybutreturnstopre-pregnancylevels
towardterm.(b)mRNAexpressiondynamicsofsmallproteoglycangenes:FMOD,testican2(SPOCK2)andBGN.(c)mRNA
dynamicsoflargeproteoglycangenes:VCANandperlecan(HSPG2).ValuesattherightofeachgraphgivetheP-valueofan
ANOVA test including all three samples; the valued above the graphs give t-test based P-values for two-tailed pairwise
comparisons
A1AT, which is not included in our mapped consistentwithotherresultssuggestingproliferation
transcripts. inlatertermcervix(seeaboveandFig.2eandf).In
Among the ADAM metallopeptidases with contrast, TIMP3 shows a 50% reduction in expres-
thrombospondin motif, ADAMTS metallopepti- sionbetweenmid-andlate-term(0.51(cid:2),P¼0.026).
dases, ADAMTS1 and -2 are most expressed.
Signaling
ADAMTS1 shows a slow but insignificant increase
Prostaglandin signaling, in particular PGE
from non-pregnant to term, while ADAMTS2 is 2
and PGF , is associated with many aspects of
increasingfromnon-tomid-pregnantanddecreases 2a
female reproductive physiology including partur-
atlateterm(Fig.4b).
ition [54]. A rate-limiting step in prostaglandin
The most expressed inhibitor of matrix metallo-
synthesisistheliberationofarachidonicacidfrom
peptidaseisTIMP2,whichishighlyexpressedduring
estrus and shows a slow but insignificant phospholipids by phospholipase A and C. In the
decline toward term (Fig. 4c). The only highly guinea pig cervix we find a substantial increase
expressed protease inhibitor that shows a signifi- in PLA2G10 just before parturition (Fig. 5a).
cant increase from mid- to late-term is TIMP1 Interestingly, the expression of cyclo-oxygenase II,
(1.9(cid:2), P¼0.006). TIMP1 is associated with cell catalyzinganotherkeystepinprostaglandinsynthe-
proliferation and may have anti-apoptotic effects sis,isnotincreasedtowardterm.
282 | Nnamanietal. Evolution,Medicine,andPublicHealth
Figure 4. Expressiondynamicsofproteases.(a)mRNAexpressionofmatrixmetalloproteasegenes.(b)mRNAexpression
ofgenescodingforADAMmetallopeptidaseswiththrombospondinmotif,ADAMTSmetallopeptidases.(c)mRNAexpres-
sionofgenescodingforinhibitorsofmatrixmetallopeptidases,TIMPs.ValuesattherightofeachgraphgivetheP-valueof
anANOVAtestincludingallthreesamples;thevaluedabovethegraphsgivet-testbasedP-valuesfortwo-tailedpairwise
comparisons
TheexpressionofIGFsignalingmoleculescame 0.0022).Incontrast, IGFBP3isexpressedatabout
toourattentionthoughtheconservedup-regulation thesamelevelasIGFBP5atestrusandthenshows
of IGFBP4 in both human and guinea pig cervices a gradual decline during pregnancy with very low
(see below). mRNA expression of IGF1 and IGF2 expression at term (from 1500 TPM during estrus
towardtermfollowdifferenttrends,withIGF1being to 140 TPM at term). These results could point to
up-regulated (1.57(cid:2), P¼0.0074) and IGF2 down- anattenuationofIGF2signalinginthetermcervix
regulated (0.66(cid:2), P¼0.0038). In contrast, there oftheguineapigbecausetwobindingproteinsare
is no substantial change in IGF receptor mRNA increasing,andIGF2expressionisdecreasing(but
expression. IGF1mRNAisincreasing).
There is high expression of most IGF binding
proteins, with three having a dramatic regulatory
DISCUSSION
change (Fig. 5b). IGFBP5 is the most expressed
IGFBP overall but does not show a significant dif- Inthisstudyweaddressedtwoquestions.First,we
ference between pregnancy stages. IGFBP4 shows investigated whether FPW in primates and guinea
the most dramatic up-regulation (2.75(cid:2), P¼ pigs has evolved independently or whether these