The first high-density genetic map of common cockle (Cerastoderma edule) reveals a major QTL controlling shell color variation

1
Vol.:(0123456789)
Scien i ic Repo s | (2022) 12:16971 | h ps://doi.o g/10.1038/s41598-022-21214-3
www.na u e.com/scien i ic epo s
The i s high‑densi y gene ic map
o common cockle (Ce as ode ma
edule) e eals a majo QTL
con olling shell colo a ia ion
Miguel He mida1, Diego Robledo2, Seila Díaz3,4, Damián Cos as5, Alicia L. B uzos3,6,
And és Blanco1, Belén G. Pa do1 & Paulino Ma ínez1*
Shell colo shows b oad a ia ion wi hin mollusc species and despi e in o ma ion on he gene ic
pa hways in ol ed in shell cons uc ion and colo has ecen ly inc eased, mo e s udies a e needed o
unde s and i s gene ic a chi ec u e. The common cockle (Ce as ode ma edule) is a aluable species
om ecological and comme cial pe spec i es which shows impo an a ia ion in shell colo ac oss
No heas A lan ic. In his s udy, we cons uc ed a high‑densi y gene ic map, as a ool o sc eening
common cockle genome, which was applied o asce ain he gene ic basis o colo a ia ion in he
species. The consensus gene ic map comp ised 19 linkage g oups (LGs) in acco dance wi h he cockle
ka yo ype (2n = 38) and spanned 1073 cM, including 730 ma ke s pe LG and an in e ‑ma ke dis ance
o 0.13 cM. Fi e ull‑sib amilies showing seg ega ion o se e al colo ‑associa ed ai s we e used o
a genome‑wide associa ion s udy and a majo QTL on ch omosome 13 associa ed o di e en colo ‑
ai s was de ec ed. Mining on his genomic egion e ealed se e al candida e genes ela ed o shell
cons uc ion and colo . A genomic egion p e iously epo ed associa ed wi h di e gen selec ion in
cockle dis ibu ion o e lapped wi h his QTL sugges ing i s pu a i e ole on adap a ion.
The common cockle, Ce as ode ma edule, is a bi al e mollusc na u ally dis ibu ed along he No heas A lan ic
coas , om Senegal in he Sou h o No way and Iceland in he No h, inhabi ing on in e idal so sedimen
egions1. This species has an impo an ecological ole on ma ine sedimen enewal and ep esen s a ood sou ce
o bi ds, c us aceans and ish, hus playing an impo an ole o coas al ecosys ems and ma ine communi ies2.
The species is conside ed a delicacy and i is comme cially ished mainly in I eland, Uni ed Kingdom, Ne he -
lands, F ance, Spain and Po ugal, whe e i ep esen s a aluable species o coas al ishe ies3.
As bi al e molluscs, he shell is a undamen al pa o he cockle, se ing as p o ec ion agains p eda o s,
desicca ion a in e idal zones o mechanical damage, enabling beha iou s such as being swep by cu en s o
bu owing. This mul i-laye ed exoskele on is cons i u ed mainly o calcium ca bona e deposi ed in o an o ganic
ma ix o p o eins and pigmen s sec e ed by specialized epi helial cells on he do sal man le4. Al hough a con-
se ed se o egula o y genes appea s o unde lie man le p ogeni o cell speci ica ion, he genes ha con ibu e
o he o ma ion o he ma u e shell a e di e se5. Technical inno a ions ha e allowed o disco e su p ising pa -
e ns o shell pigmen a ion and apid di e gences in he mix o pigmen s used o achie e simila colo pa e ns6,7.
Indeed, he shells o di e en bi al es a e ema kably di e se, and shell pigmen a ion a ies d ama ically e en
wi hin species. Va ia ion in shell colo and i s pa e n may be associa ed wi h di e en bio ic o abio ic ac o s
such as p eda ion, subs a e, die o en i onmen al condi ions8,9. Since colo a ia ion has been epo ed o be
con olled o some ex en by gene ic ac o s, shell colo migh be impo an o adap a ion o bi al e popula-
ion o selec i e p essu es10,11. Mo eo e , as a comme cialized ood esou ce, shell colo can be impo an o
OPEN
1Depa men o Zoology, Gene ics and Physical An h opology, Acuigen G oup, Facul y o Ve e ina y, Uni e sidade
de San iago de Compos ela, Campus o Lugo, 27002 Lugo, Spain. 2The Roslin Ins i u e and Royal (Dick) School
o Ve e ina y S udies, Uni e si y o Edinbu gh, Midlo hian, UK. 3Genomes and Disease G oup, Depa men o
Zoology, Gene ics and Physical An h opology, Cen e o Resea ch in Molecula Medicine and Ch onic Diseases
(CiMUS), Uni e sidade de San iago de Compos ela, 15782 San iago de Compos ela, Spain. 4ECOMARE,
CESAM-Cen e o En i onmen al and Ma ine S udies, Depa men o Biology, Uni e si y o A ei o, San iago
Uni e si y Campus, 3810-193 A ei o, Po ugal. 5Cen o de In es igación Ma iña, Uni e sidade de Vigo, ECIMAT,
36331 Vigo, Spain. 6Mosaicism and P ecision Medicine G oup, Depa men o Gene ics and Genomic Medicine, The
F ancis C ick Ins i u e, Uni e si y College o London, London, UK. *email: [email p o ec ed]
2
Vol:.(1234567890)
Scien i ic Repo s | (2022) 12:16971 | h ps://doi.o g/10.1038/s41598-022-21214-3
www.na u e.com/scien i ic epo s/
consume pleasan ness and accep abili y, a ec ing o a ce ain ex en he sale alue. Whe he common cockle
popula ions show a ia ion in hei colo a ion pa e ns, i i has an unde lying gene ic basis and o wha ex en
i could be ela ed o adap i e a ia ion is unknown.
The apid expansion o nex -gene a ion sequencing (NGS) in he las decade has allowed he de elopmen o
geno yping-by-sequencing (GBS) me hods, which ha e been employed o disco e and geno ype housands o
single nucleo ide polymo phisms (SNPs) in a cos -e ec i e manne , enabling popula ion-scale gene ic s udies
in non-model species12. These me hods, including Res ic ion-si e Associa ed DNA (RAD) sequencing13 and
i s de i a ions, ddRAD14, 2b-RAD15 o SLAF16, ha e been success ully used o high- h oughpu geno yping in
many aquacul u e species17, including se e al impo an comme cial molluscs18. GBS me hods ha e been ecen ly
applied o unde s and adap i e a ia ion o common cockle om No heas A lan ic, and consis en signals o
adap i e a ia ion we e de ec ed bo h a mic ogeog aphic (dd-RAD;19) and mac ogeog aphic (2b-RAD;20) scales.
GBS ha e acili a ed he cons uc ion o high- esolu ion linkage maps21–23, which a e impo an ools o genome
sca olding and assembly24 and ha e aided o disen angle he gene ic basis o ele an e olu iona y o p oduc i e
ai s h ough quan i a i e ai locus (QTL) sc eening25,26. Gene ic maps ha e been used o s udy he gene ic
a chi ec u e o ai s o in e es in a ious bi al e species, such as g ow h in Zhikong scallop (Azumapec en a -
e i;27), bay scallop (A gopec en i adians;28) o Paci ic oys e (C assos ea gigas29;), a ious pea l-quali y ai s in
iangle sail mussel (Hy iopsis cumingii;30) and esis ance o pa hologies in Eu opean la oys e (Os ea edulis;31).
P e ious s udies ha e also iden i ied QTL o shell colo a ion in se e al bi al es including Manila clam (Rudi apes
philippina um;32–34), ha d clam (Me cena ia me cena ia;35), Paci ic oys e (C assos ea gigas;36–39), black-lip pea l
oys e (Pinc ada ma ga i i e a;40), Akoya pea l oys e (Pinc ada uca a;41) and Yesso scallop (Mizuhopec en yes-
soensis;42,43), and he e o e, simila s a egies migh be employed o asce ain he gene ic componen unde lying
di e ences in shell colo a ion in common cockle.
In his s udy, we in es iga ed he a ia ion o shell colo in No heas A lan ic popula ions o Eu opean com-
mon cockle and s udied i s gene ic a chi ec u e h ough a genome-wide associa ion s udy (GWAS) on se e al
ull-sibs amilies using he i s common cockle high-densi y linkage map he e cons uc ed using 2b-RAD SNP
geno yping. A majo QTL unde lying shell colo and i s pa e n in his species was de ec ed on ch omosome 13
and se e al candida e genes iden i ied. This in o ma ion should be conside ed as a po en ial sou ce o adap i e
a ia ion in common cockle and could be exploi ed in b eeding p og ams o adap p oduc ion o consume
demands.
Resul s
Colo and pa e n a ia ion in common cockle Eu opean popula ions. Shell colo showed g ea
di e si y among he 270 cockle indi iduals analysed in No heas A lan ic popula ions (Fig.1). Al hough a p e-
dominan colo was displayed in each popula ion, impo an a ia ion was also obse ed wi hin he nine cockle
beds analysed (Fig.2).
Yellow (1) was he only colo de ec ed in he popula ion om F ance and was he mos abundan in he
popula ions om Denma k, Po ugal (Alga e) and UK (Wales); whi e (2) was p edominan in popula ions om
Spain, Ne he lands and Ge many; b own (4) was he leas equen colo in his na u al su ey, only de ec ed
in one o he popula ions om Po ugal (Alga e); and o ange (6) was a he common in he o he popula ion
om Po ugal (A ei o), Ne he lands and UK (Plymou h) (Fig.2). Finally, g ay (3) and black (5) colo s we e
only de ec ed in he amilies s udied a ha che y using b eede s om NW Spain (see below), so appa en ly a e
in Eu opean popula ions.
Shell pa e ns in wild adul s (Fig.1) we e no as ma ked as in ju enile samples ob ained om c osses in
he ha che y (see below) and consequen ly we e no sys ema ically eco ded. None heless, di e en colo pa -
e ns simila o hose obse ed in he ha che y we e also iden i ied: (i) a ligh e colo a ion in he ci cle-shaped
umbo in he popula ion om Po ugal (Alga e); (ii) a da ke band (black o o ange) wi h blu ed bounda ies
in he la e al a ea o he shell on he opposi e side o he ligamen in popula ions om UK (Wales, black) and
Po ugal (A ei o, o ange); (iii) changes in he a angemen o he pe ios acum (p o ein laye ) de ec ed in he
en al ma gin associa ed wi h he las g ow h ings in all popula ions, excluding Denma k, whe e emnan s o
he pe ios acum we e de ec ed on he en i e su ace o he shell. Addi ionally, mal o ma ion o he shell ha
a ec ed he las g ow h ings, likely ela ed o en i onmen al ac o s, we e de ec ed in indi iduals om F ance
and Po ugal (Alga e).
2b‑RAD sequencing. Cockle b eede s o p oduce amilies we e collec ed in he na u al bed o Noia (NW
Spain) and ans e ed o he ECIMAT-CIM-UVigo (Vigo, NW Spain), whe e i e amilies we e used o GWAS
on shell colo and i s pa e n; u he , he wo mo e nume ous amilies (F6 and F8) we e used o gene ic map
cons uc ion. A o al o 275 samples we e sequenced in h ee 2b-RAD lib a ies: (i) he wo pa en s and 97
o sp ing o F6; (ii) he wo pa en s and 99 o sp ing o F8; and (iii) 25 o sp ing om each o F2, F3 and F7.
A ound ~ 575 million aw eads we e ob ained in he i s wo lib a ies o he wo la ge amilies: on a e age ~ 6.9
million o pa en s ( ange: 5,635,542—8,343,112) and ~ 2.8 million o o sp ing ( ange: 3,680—5,991,704).
A e il e ing, ~ 75% o he eads we e e ained and aligned o he cockle genome (Tubío e al., unpublished).
An impo an numbe o eads we e disca ded due o mapping o wo o mo e genomic posi ions (40.74%),
esul ing in ~ 200 million single-si e aligned eads: ~ 3 million om each pa en ( ange: 2,101,828 o3,843,549)
and ~ 900,000 om each o sp ing ( ange: 1,317 o 2,158,061).
The hi d un, wi h he emaining 75 samples, yielded ~ 275 million aw sequences (~ 3.6 million pe o -
sp ing), and a e he il e ing and alignmen s eps, ~ 92 million high-quali y aligned eads we e e ained (~ 1.2
million pe o sp ing).
3
Vol.:(0123456789)
Scien i ic Repo s | (2022) 12:16971 | h ps://doi.o g/10.1038/s41598-022-21214-3
www.na u e.com/scien i ic epo s/
Geno yping and linkage map cons uc ion. The gs acks module using he ma ukilow model applied o
all amilies yielded 318,755 loci, which esul ed in 85,078 polymo phic SNPs using he popula ions module. One
indi idual om each o he wo amilies used o mapping showed a low numbe o alid geno ypes (< 30% o
SNPs geno yped) and hey we e emo ed. A e quali y con ol, 7,094 and 8,439 SNPs we e e ained o F6 and
F8, espec i ely. The la ges numbe o il e ed SNPs in ou s udy was due o missing geno ypes and de ia ions
om Mendelian seg ega ion, which can be explained by he p esence o null alleles ela ed o polymo phism in
he es ic ion enzyme a ge s, as p e iously epo ed in molluscs44,45. The e we e 1,329 common in o ma i e
SNPs be ween he wo amilies, which means ha in o al 14,204 SNPs we e used o he cons uc ion o he
gene ic map.
Sepa a e male and emale maps we e buil o each amily (Supplemen al TablesS1–S4; Supplemen al Fig.1).
To achie e an app op ia e numbe o LGs close o he numbe o cockle ch omosomes (n = 19), di e en LOD
sco es we e explo ed o each gene ic map ( anging be ween 7.0 and 9.0), esul ing in 21 LGs in all maps. Fo
he ma e nal map o F6, 3,514 ma ke s we e mapped o a o al leng h o 12,572cM, whe eas he pa e nal map
included 3952 ma ke s spanning 16,692cM (Supplemen al TableS5). In F8, 4,698 and 4,368 ma ke s we e
mapped in he ma e nal and pa e nal maps, spanning 25,017 and 23,266cM, espec i ely. Sha ed ma ke s
among pa en al maps we e used o build a single consensus map. As a esul , 13,874 SNPs we e assigned o 19
LGs in he inal common cockle gene ic map wi h a o al leng h o 51,778cM, in acco dance wi h he 19 cockle
ch omosomes (Supplemen al TablesS5–S6). The leng h o he maps exceeded ha expec ed based on he genome
size conside ing a s anda d ela ionship be ween physical (Mb) and gene ic dis ance (cM) o 1.2 and a C- alue
o 1.37 pg46. The obse ed elonga ion o he gene ic maps is he consequence o he high numbe o ma ke s,
se e al mapping amilies and he limi a ions o he so wa e used (JoinMap o mapping and Me geMaps o
consensing), consis en wi h p e ious obse a ions21,47,48.
To build a eliable amewo k gene ic map we used he Reg ession Mapping app oach, a simila app oxi-
ma ion o ha ollowed in C. gigas by Hedgecock e al.49. Acco dingly, a o al o 831 and 340 ma ke s wi hou
Figu e1. Rep esen a i e indi iduals o each colo and pa e n iden i ied in shells o C. edule om No heas
A lan ic. F om up o down and le o igh : colo pheno ypes yellow (1), whi e (2), g ay (3), b own (4), black (5),
o ange (6), and shell pa e ns ci cle (7), line (8) and s ipe (9).
4
Vol:.(1234567890)
Scien i ic Repo s | (2022) 12:16971 | h ps://doi.o g/10.1038/s41598-022-21214-3
www.na u e.com/scien i ic epo s/
missing da a we e selec ed in o sp ing o F6 and F8, espec i ely. Sepa a e maps we e buil o each pa en in
each amily wi h a LOD sco e ≥ 5.0. G aphical ep esen a ion was no implemen ed, bu he indi idual g oups
we e me ged di ec ly wi h he “Combine G oups” op ion o he JoinMap o build a consensus map. The eg es-
sion o common ma ke s dis ance was used o co ec he dis ances in he o iginal consensus map o build a new
co ec ed-leng h consensus map, which included he 19 LGs bu wi h a new o al leng h o 1073cM (Fig.3). In
he educed consensus map, he es ima ed in e -ma ke dis ance dec eased d ama ically ( om 4.34 o 0.13cM),
compa able o o he gene ic linkage maps cons uc ed using 2b-RAD50,51, and he a e age a io be ween physical
and gene ic dis ance (0.74Mb/cM) was also much close o ha expec ed.
GWAS on shell colo ype and pa e n a ha che y. Shell colo and i s pa e n showed a ema kable
a ia ion wi hin and among he i e amilies ea ed in he same en i onmen al condi ions (Fig.4). Shell colo o
he o sp ing o he i e amilies was classi ied as ou lined be o e ( om 1 o 6) and conside ed as a con inuous
ai o analyses (Table1). Fi e o he six colo s iden i ied in common cockle (Fig.1) we e iden i ied in amilies.
Black was he mos equen colo (31.6%), bu i was de ec ed only in wo amilies (F2 and F8), whe eas g ay
was he leas equen and de ec ed only in F8. Whi e was he only p esen in all amilies and he second mos
abundan in he whole sample (29.8%), and b own was missed only in one amily. O ange was no de ec ed in
any o he amilies, only in he wild, as ou lined be o e. Colo pa e ns we e only de ec ed in h ee amilies and
we e qui e he e ogeneous; o ins ance, s ipe was only obse ed in F6 a a a io close o 1:1 (Fig.4; Table1).
The es ima ed he i abili ies we e high o colo and s ipe, 0.755 and 0.657, espec i ely, and mode a e-high
o ci cle and line, 0.537 and 0.506, espec i ely. A gene ic co ela ion o almos 1 was obse ed be ween colo
and ci cle (0.998), while hese pa e ns showed a mode a e nega i e gene ic co ela ion wi h s ipe (− 0.327 and
-0.379, espec i ely) (Table2). Line did no show signi ican gene ic co ela ion wi h any o he o he ai s. The
ci cle in he umbo was associa ed wi h da ke shell colo s, which sugges s a pa icula pa e n no isible in he
whi ish pheno ypes.
Figu e2. Geog aphical ep esen a ion o he Eu opean samples o C. edule, including pie cha colo
dis ibu ions (yellow (1), whi e (2), b own (4), o ange (6)) and a pho og aph o a ep esen a i e indi idual.
F om no h o sou h: Nykobing Mo s (Denma k), Syl (Ge many), Slikken an Viane (Ne he lands), Wales and
Plymou h (Uni ed Kingdom), A cachon (F ance), Baiona (Spain), A ei o and Alga e (Po ugal).
5
Vol.:(0123456789)
Scien i ic Repo s | (2022) 12:16971 | h ps://doi.o g/10.1038/s41598-022-21214-3
www.na u e.com/scien i ic epo s/
The comple e da ase con ained colo pheno ypes o 275 indi iduals om i e amilies, geno yped o 13,874
SNPs mapped in he Consensus Map (be ween 4,643 and 8,439 in o ma i e ma ke s pe amily). GWAS e ealed
a highly signi ican genome-wide QTL o colo and s ipe, and o a less ex en o ci cle, a ch omosome C13
(Fig.5). O he SNP associa ions a ch omosome-le el we e de ec ed o ci cle a C8, o line a C2, C9, C12 and
C14, and o s ipe a C1, C5, C9, and C17 (Supplemen al TableS7).
Gene mining. GWAS iden i ied a con incing QTL associa ed wi h colo , s ipe and ci cle loca ed a C13
be ween 14,367,847 and 33,654,270bp (Fig.5). The highes signi ican SNP associa ed wi h he h ee ai s
was loca ed a 30,286,849bp, bu ac oss his wide egion, he e we e se e al s e ches de ined by highly sig-
ni ican SNPs associa ed wi h colo and s ipe, he mos signi ican -associa ed ai s (Fig.6). The wo end
sub egions we e mainly associa ed wi h colo , and mining a ound he mos signi ican SNPs (14,778,145 and
33,225,111bp ± 500kb) e ealed 16 and 17 anno a ed genes, espec i ely (Supplemen al TableS7. Hal o he
genes in he i s window, ela ed o shell a chi ec u e and colo , clus e ed on a ~ 250kb egion (Fig.6; Supple-
men al TableS7): i e we e ela ed o chi in binding ( h ee mic o ib il-associa ed glycop o ein 4, one includ-
ing a ib inogen domain, and one DNA damage- egula ed au ophagy modula o ), one o calcium binding
(ependymin- ela ed), one o mucin sec e ion, and one o ammonium anspo . The window a he o he end
(33,225,111kb) included wo genes ela ed o i on binding ( wo s e oid 17-alpha-hyd oxylase/17,20 lyase-like)
and ano he one o calcium anspo (phospha idylinosi ol 4,5-bisphospha e phosphodies e ase).
On he o he hand, wo consecu i e sub egions associa ed wi h s ipe colo pa e n we e loca ed a ound he
mos signi ican SNPs a 22,683,835 and 30,286,849bp (± 500kb) on he same ch omosome and comp ised a
o al o 36 anno a ed genes. These g oup o genes included h ee cell memb ane o ganic anspo e s (solu e
ca ie o ganic anion anspo e amily membe 4A1-like, u ea-p o on sympo e DUR3-like and sodium/
glucose co anspo e 4-like); one chi in-binding (sushi, on Willeb and ac o ype A, EGF and pen axin
domain-con aining p o ein 1-like); se e al in ol ed in ansi ac oss he memb ane by endo-exocy osis mecha-
nisms (p enylcys eine oxidase 1-like, glucose-6-phospha e exchange SLC37A4-like, and AP-2 complex subuni
alpha-2-like); and one ela ed o i on chela ion (pu a i e e ic-chela e educ ase 1) (Supplemen al TableS7).
Discussion
Molluscs ep esen a highly di e se Phylum o in e eb a es comp ising an es ima ed numbe o 200,000 spe-
cies, dis ibu ed ac oss almos e e y ype o habi a wo ldwide18. They ha e key oles as ecosys em enginee s,
wa e il e ing and pollu ion moni o ing, jewel y and, o cou se, as an impo an ood sou ce. Mo e han 17
million me ics ons o molluscs we e a med wo ldwide in 2018 and mos o his p oduc ion concen a ed in
Figu e3. G aphical ep esen a ion o he consensus map o C. edule including 13,874 ma ke s encompassing
1073cM.. The ule in he le indica es leng h in cen imo gans (cM).

6
Vol:.(1234567890)
Scien i ic Repo s | (2022) 12:16971 | h ps://doi.o g/10.1038/s41598-022-21214-3
www.na u e.com/scien i ic epo s/
a hand ul o species o he class Bi al ia52. Bi al e aquacul u e mainly elies on ex ensi e a ming based on he
collec ion o wild seed and ha es ing in na u al beds, which means ha wild popula ions a e unde impo an
human al e a ions45.
Shell cons i u es a main s uc u e o mollusc ana omy ha p o ec s hem agains p eda o s and desicca ion,
bu ha also plays o he impo an unc ions depending on species. Shells a e sec e ed by he man le, so colo
Figu e4. Pho og aph composi ion showing ep esen a i e indi iduals o he i e amilies (F2, F3, F6, F7 and
F8) o C. edule used o GWAS on shell colo and pa e n. Las panel: summa y o he a ia ion wi hin and
be ween amilies: yellow (1), whi e (2), g ay (3), b own (4), black (5), ci cle (7), line (8) and s ipe (9).
Table 1. Dis ibu ion o shell colo and pa e n pheno ypes in he o sp ing o he i e amilies o C. edule.
Family
Shell colo Shell colo pa e n
Yellow Whi e G ay B own Black Ci cle Line S ipe
Family 2 3 1 0 8 13 7 11 0
Family 3 7 12 0 6 0 0 0 0
Family 6 24 45 0 31 0 0 0 54
Family 7 0 13 0 12 0 0 0 0
Family 8 4 11 11 0 74 36 8 0
To al (%) 38 (13.9) 82 (29.8) 11 (4.0) 57 (20.7) 87 (31.6) 43 (37.1) 19 (16.4) 54 (46.6)
7
Vol.:(0123456789)
Scien i ic Repo s | (2022) 12:16971 | h ps://doi.o g/10.1038/s41598-022-21214-3
www.na u e.com/scien i ic epo s/
and i s pa e n a e mainly de e mined by pigmen s p oduced by his issue, al hough he mic os uc u e o he
shell may also con ibu e o colo a ion4,53. Despi e ecen e o s, he e is s ill a knowledge gap ega ding he
gene ics unde pinning shell colo in Mollusca and i s po en ial ole on adap a ion. While i is well-known ha
shell colo can be unde gene ic con ol, bio ic and abio ic ac o s, such as die , empe a u e, salini y o pH can
also play a ole, and he in e ac ion be ween hem is no well unde s ood ye 8. To add ess he gene ic a chi ec u e
o complex ai s, such as shell colo , comp ehensi e genomics app oaches a e essen ial.
Genomic esou ces ha e inc eased exponen ially in he las decade as consequence o he lowe ing cos o
sequencing echnologies and he new bioin o ma ic ools ha enabled he assembly o genomes a ch omo-
some le el, he cons uc ion o high-densi y gene ic maps and he geno yping o millions o SNPs o genomic
sc eening54. Gene ic maps a e essen ial o he iden i ica ion o genomic egions unde lying pheno ypic a ia ion
Table 2. Pheno ypic and gene ic co ela ions (uppe iangle and lowe iangle, espec i ely), be ween colo
ai s o C. edule included in his s udy. (***p < 0.001, *p < 0.05).
T ai Colo Ci cle Line S ipe
Colo 0.465*** 0.147* − 0.501***
Ci cle 0.998 0.049n.s − 0.208***
Line 0.173 0.137 − 0.130*
S ipe − 0.379 − 0.327 − 0.217
Figu e5. Ci cula Manha an plo o SNP signi ican ly associa ed wi h shell colo and pa e n pheno ypes
in C. edule based on he Mixed Linea Model (MLM). SNP-GWAS ci cle ep esen a ions: inne -mos : colo ;
i s -middle: ci cle: second-middle: line; and ou e -mos : s ipe. Fo Manha an plo , Y-axis ep esen s − log 10
(p- alue) o he associa ion wi h each SNP and X-axis is physical posi ion in bp. The dashed ed lines ep esen
he Bon e oni h eshold a genome-wide le el (p ≤ 0.05). SNP ma ke s a e ep esen ed by g ay do s and hose
abo e his h eshold in ed do s.
8
Vol:.(1234567890)
Scien i ic Repo s | (2022) 12:16971 | h ps://doi.o g/10.1038/s41598-022-21214-3
www.na u e.com/scien i ic epo s/
o ele an ai s unde domes ic o na u al selec ion44,45,55. The i s mollusc gene ic maps we e published in
C assos ea i ginica56 and C assos ea gigas57,58 using AFLPs and mic osa elli es, espec i ely, bu he lowe ing
cos o sequencing echnologies enabled geno yping housands o SNPs o imp o ing map densi y29,32. The
common cockle gene ic map he e cons uc ed comp ehends 13,874 ma ke s wi h an in e -ma ke dis ance o
0.13cM and comp ises he 19 expec ed LGs ma ching wi h he haploid ka yo ype o he species59, being, o ou
knowledge, he denses gene ic map published o da e in molluscs. Besides i s impo ance o genomic sc eening,
he common cockle gene ic map is an in aluable ool o genome sca olding, as has been p e iously epo ed in
o he species23,60, and in ac , i e o he se en mollusc genomes assembled a ch omosome-le el ook ad an age
o high-densi y linkage maps45.
We used he common cockle gene ic map o asce ain he gene ic componen unde lying he b oad di e si y o
shell colo and pa e n on he wild popula ions o his species in he No heas A lan ic. Rica do e al.61,62 epo ed
an impo an a ia ion in shell ion composi ion in common cockle, appa en ly ela ed o en i onmen al ac-
o s, ha allowed acing back he geog aphic o igin o specimens, bu hey did no associa e his a ia ion wi h
colo . Mos pheno ypes obse ed in wild popula ions in ou s udy could be iden i ied in he amilies p oduced
a ha che y, al hough hei in ensi y was somewha aded, likely due o en i onmen al ac o s ope a ing ac oss
he li espan o he adul specimens collec ed. Indeed, di e en ia ion o indi iduals by colo in amilies was e y
clea , some imes esembling single gene Mendelian seg ega ion, and he i abili ies o all colo ai s e alua ed
we e high (h2 > 0.5), suppo ing a signi ican gene ic componen unde lying colo a ia ion in common cockle.
Bo h colo , ci cle and s ipe showed highly signi ican gene ic and pheno ypic co ela ions, sugges ing ha he
same genes (o genomic egions) could unde lie colo a ia ion o hese h ee ai s. Mo eo e , he ac ha
mos pheno ypes obse ed in he A lan ic dis ibu ion appea o be seg ega ing in a single popula ion om NW
Spain suppo s an impo an in apopula ion a ia ion in common cockle and sugges ha mo e de ailed s udies
ac oss he ull li espan could disclose mo e a ia ion han obse ed in ou p elimina y sc eening.
In acco dance wi h hese obse a ions, he GWAS pe o med on i e ull-sib amilies p oduced a ha ch-
e y iden i ied a majo QTL a C13 o wo o he ou ai s e alua ed (colo and s ipe). This egion encom-
passed ~ 13Mb, al hough wi h di e en s e ches o he same o he di e en ai s s udied, which sugges s he
exis ence o a b oad gene clus e ela ed o colo pigmen a ion in common cockle wi h di e en genes playing
di e se unc ions on simila ai s. A o al o 69 anno a ed genes we e ound on his egion a e mining he
cockle genome (Tubío e al., unpublished). We iden i ied a no able p opo ion o genes ela ed o ion binding
and anspo /sec e ion ac oss he cell memb ane, such as ammonium o o ganic anspo e s, calcium binding
o i on chela ion, mucin p oduc ion and se e al ela ed o endo-exocy osis mechanisms, which play an impo -
an ole in he de elopmen o he shell and ha has been ela ed wi h shell colo in o he molluscs35,38,41,42,63,64.
Mo eo e , shells a e sec e ed by he man le in a p ocess called biomine aliza ion, whe e chi in ep esen s an
impo an componen 40,65,66. We could iden i y six chi in-binding ela ed genes, ou o hem in a small genomic
egion encompassing ~ 110kb including h ee mic o ib il-associa ed glycop o ein 4 genes. Genes ela ed o chi in
and calcium me abolism in ol ed in di e en shell colo lines ha e been epo ed in Mizuhopec en yessoensis42.
Ne e heless, i is impo an o no e ha p e ious s udies ha e shown ha some o he genes ela ed o shell
a chi ec u e and colo a e species-speci ic8,53, so u he s udies should be conduc ed in he u u e o asce ain
he oles o he genes loca ed a he C13 clus e o a deep unde s anding o colo ype and pa e n di e si y in
common cockle.
A e lec ion on he pu a i e adap i e ole o colo di e si y in he common cockle is wo h a inal hough .
While shell colo can be impo an o he adap a ion o bi al e popula ions o selec i e p essu es, his migh no
be a majo ac o in molluscs ha li e bu ied in he sedimen , such as C. edule. Ne e heless, cockles li e in he
in e idal zone and cons i u e he eed o di e en p eda o s, such as bi ds, mammals and c us aceans, among
Figu e6. Manha an plo s o SNP signi ican ly associa ed wi h colo (A, D) and s ipe pa e n (B, C)
pheno ypes in ou ch omosome 13 sub egions o C. edule based on he Mixed Linea Model (MLM). Y-axis
ep esen s − log 10 (p- alue) o he associa ion wi h each SNP and X-axis i s physical posi ion in bp. Codes o
genes mined in hose egions a e shown wi h unde lined ba s ep esen ing hei leng h in Mb; yellow and ed
do s highligh associa ed SNPs wi h s ipe and colo , espec i ely; ed ba s a each plo show he mos in e es ing
candida es in he sub egion.
9
Vol.:(0123456789)
Scien i ic Repo s | (2022) 12:16971 | h ps://doi.o g/10.1038/s41598-022-21214-3
www.na u e.com/scien i ic epo s/
o he s. Mo eo e , a b oad colo di e si y exis s in his species ac oss i s ull dis ibu ion ange unde pinned by a
subs an ial gene ic a ia ion, as he high he i abili ies es ima ed in ou s udy demons a e, so i s pu a i e adap i e
ole should dese e u he s udies. In e es ingly, he C13 colo associa ed QTL o e laps wi h a genomic egion
in he same ch omosome which showed e y consis en signals o di e gen selec ion in he whole No heas
A lan ic and in he No he n Region (abo e he Ushan F on ), including se e al ou lie s abo e he neu al
backg ound and highly signi ican linkage disequilib ium sugges i e o selec i e sweeping20.
Conclusion
He e we p esen ed a high-densi y gene ic map in common cockle, he i s epo ed in he species and, o ou
knowledge, he highes dense map epo ed in molluscs o da e. The consis ency o he map was shown by i ing
he numbe o linkage g oups o he haploid ch omosome numbe o he species and by he consis en esul o
he GWAS on shell colo and pa e n. This map was used o asce ain he gene ic componen and a chi ec u e
unde lying he b oad colo di e si y obse ed in common cockle in he No heas A lan ic. High he i abili ies
we e es ima ed o all ai s e alua ed, suppo ing an impo an gene ic componen unde lying colo a ion pa -
e n. This gene ic a ia ion was mainly associa ed wi h a single genomic egion a C13, whe e a clus e o genes
ela ed o speci ic en iched unc ions on shell a chi ec u e and colo we e de ec ed. Ou p elimina y esul s in
he wild sugges a po en ial adap i e ole o he colo a ia ion obse ed and highligh s he impo ance o deepe
s udies a popula ion le el ac oss he cockle li espan o unde s and he signi icance o he a ia ion obse ed.
Ma e ial and me hods
Cockle amilies. In May 2018, 300 ma u e adul C. edule cockles we e collec ed in Noia, Galicia (NW Spain)
and ans e ed o he ECIMAT-CIM-UVigo ma ine acili ies (Vigo, NW Spain). Cockles we e kep indi idually
in glasses wi h 0.3L o 1μm il e ed seawa e a 20°C. Spawning was induced by he mal shocks be ween 10 and
22°C o 10h and he quali y o he oocy es and spe m was e alua ed unde a ligh mic oscope. Con olled e -
iliza ion was ca ied ou by adding spe m o oocy es, one male x one emale, a a a io o 1:10. D-shaped la ae
we e ob ained 24h a e e iliza ion wi h a ans o ma ion a e om ochopho e la ae o 42% ± 19. Following
his p o ocol, a o al o ele en ull-sib amilies we e ob ained by c ossing one male x one emale, in ol ing a o al
o 11 emales and 7 males.
La ae om each amily we e cul u ed in indi idual 150L cylind ical-conical anks a a densi y o 8 ± 3 la ae
mL−1 wi h sea wa e il e ed a 1µm and ea ed wi h UV, sligh ae a ion, and empe a u e 19.0 ± 1.4°C in an
open ci cui wi h a enewal o 5% olume / hou . The die consis ed o Tisoch ysis lu ea (ECC038), Chae oce os
neog acile (ECC007), Phaeodac ylum ico nu um (ECC028) and Rhodomonas lens (ECC030) in a a io o 1:1:1:1
(acco ding o he cell coun ), and Te aselmis suecica (ECC036) was included om he se en h day o cul u e. The
daily die was adminis e ed au oma ically e e y 4h in 6 daily in akes, main aining a cons an concen a ion in he
ank o 20–40 cells µl−1. A 14days pos - e iliza ion (dp ), pedi elige la ae om each amily we e ans e ed
o sepa a e 50L anks in suspended baske s wi h cons an ae a ion, empe a u e 18.4°C ± 0.5 and a enewal a e
o 50L day-1. The animals we e ed wi h he same die as desc ibed abo e bu main aining a cons an densi y o
168 ± 48 seeds cm2. Me amo phosis o la ae ook place in hose anks.
A 112 dp one hund ed indi iduals om each o wo amilies, F6 (12.74 ± 0.73mm) and F8 (12.23 ± 1.48mm),
we e selec ed o gene ic mapping and o GWAS on colo pa e ns, while wen y- i e indi iduals om each o
h ee addi ional amilies, F2 (9.82 ± 0.85mm), F3 (12.58 ± 0.90mm) and F7 (13.12 ± 1.67mm), we e sampled
o inc easing s a is ical powe o GWAS. The mea o each indi idual was ixed in pu e e hanol and sen o he
Genomics Pla o m o Uni e si y o San iago de Compos ela (Campus de Lugo) o DNA ex ac ion.
Colo a ia ion o common cockles. A e he isual inspec ion o all he shells in he s udy, shells we e
classi ied acco ding wi h hei colo in o six pheno ypic classes: yellow (1), whi e (2), g ey (3), b own (4), black
(5), o ange (6). Fu he , speci ic colo pa e ns o he shell we e consis en ly iden i ied and eco ded as h ee
o he ai s: (i) a ci cle in he umbo, di e en ia ed om he es o he shell and gene ally yellow (ci cle); (ii) a
b oken whi e line (line); and (iii) a s ipe wi h whi e line edge (s ipe) (Fig.1). A ypical indi idual om each
class was selec ed o de ine a s anda d pa e n o ca ego ize he pheno ype o each shell in he s udy. This e alu-
a ion was pe o med by a single obse e , who e ised and sco ed all he shells in wo independen ounds.
Acco dingly, colo and pa e n o he shell we e assigned o all ha che y specimens (F2, F3, F6, F7, F8 amilies),
while mainly only he colo was assessed in he Eu opean wild popula ion samples p o ided by he Scuba Can-
ce s P ojec (ERC-2016-STG). Despi e simila colo pa e ns we e de ec ed in he ha che y and wild indi iduals,
i s classi ica ion was no s aigh o wa d in wild specimens, likely due o en i onmen al ac o s ac oss hei li e
span, and he e o e hey we e no sys ema ically eco ded.
2b‑RAD lib a y cons uc ion and sequencing. DNA was ex ac ed om he whole mea using he
E.Z.N.A. E-96 mollusc DNA ki (Omega Bio- ek) ollowing manu ac u e ecommenda ions. Lib a y p epa-
a ion ollowed he 2b-RAD p o ocol15 wi h sligh modi ica ions21. B ie ly, DNA samples we e adjus ed o
80ngµL−1 and diges ed using he IIb ype es ic ion enzyme Al I (The mo Fishe ). As a esul , he genome was
cu in agmen s o 36bp o leng h, wi h he es ic ion enzyme ecogni ion si e in he middle. Speci ic adap-
o s, also including indi idual sample ba codes, we e liga ed and he esul ing agmen ampli ied. A e PCR
pu i ica ion, samples we e quan i ied using Qubi 2.0 luo ome e (Li e Technologies, Ca lsbad, CA, USA) and
equimola ly pooled. The pools we e sequenced on a Nex Seq 500 Illumina sequence using he 50bp single-end
chemis y in he acili ies o FISABIO Sequencing and Bioin o ma ics Se ice (Valencia, Spain). The wo bigge
amilies including ~ 100 o sp ing pe amily, wi h pa en s a double concen a ion, we e mul iplexed each in one
un, whe eas he o he se en y- i e samples, wi hou pa en s, we e mul iplexed in a hi d un.