F equen soma ic ans e o mi ochond ial DNA
in o he nuclea genome o human cance cells
Young Seok Ju,
1
Jose M.C. Tubio,
1,45
William Mi sud,
1,45
Beiyuan Fu,
2
Helen R. Da ies,
1
Manasa Ramak ishna,
1
Yilong Li,
1
Lucy Ya es,
1
Gunes Gundem,
1
Pa ick S. Ta pey,
1
Sam Behja i,
1
Elli Papaemmanuil,
1
Sancha Ma in,
1
An hony Fullam,
1
Mo i z Ge s ung,
1
ICGC P os a e Cance Wo king G oup,
46
ICGC Bone Cance
Wo king G oup,
46
ICGC B eas Cance Wo king G oup,
46
Jyo i Nangalia,
1,3,4
An hony R. G een,
3,4
Ca los Caldas,
3,5
Åke Bo g,
6,7,8
And ew Tu ,
9
Ming Ta
Michael Lee,
10,11
Lau a J. an’ Vee ,
12,13
Beni a K.T. Tan,
14
Samuel Apa icio,
15
Paul N. Span,
16
John W.M. Ma ens,
17
S ian Knappskog,
18,19
Anne Vincen -Salomon,
20
Anne-Lise Bø esen-Dale,
21,22
Jó unn E la Ey jö d,
23
Ola Myklebos ,
24
Ad ienne
M. Flanagan,
25,26
Ch is ophe Fos e ,
27
Da id E. Neal,
28,29
Colin Coope ,
30,31
Rosalind Eeles,
32,33
G. S e en Bo a,
34
Sunil R. Lakhani,
35,36,37
Ch is ine Desmed ,
38
Gilles Thomas,
39,44
And ea L. Richa dson,
40,41
Colin A. Pu die,
42
Alas ai
M. Thompson,
43
Ul an McDe mo ,
1
Feng ang Yang,
2
Se ena Nik-Zainal,
1
Pe e
J. Campbell,
1
and Michael R. S a on
1
1–43
[Au ho a ilia ions appea a end o pape .]
Mi ochond ial genomes a e sepa a ed om he nuclea genome o mos o he cell cycle by he nuclea double memb ane,
in e ening cy oplasm, and he mi ochond ial double memb ane. Despi e hese physical ba ie s, we show ha soma ically
acqui ed mi ochond ial-nuclea genome usion sequences a e p esen in cance cells. Mos occu in conjunc ion wi h in a-
nuclea genomic ea angemen s, and he ea u es o he usion agmen s indica e ha nonhomologous end joining and/o
eplica ion-dependen DNA double-s and b eak epai a e he dominan mechanisms in ol ed. Rema kably, mi ochond i-
al-nuclea genome usions occu a a simila a e pe base pai o DNA as in e ch omosomal nuclea ea angemen s, in-
dica ing he p esence o a high equency o con ac be ween mi ochond ial and nuclea DNA in some soma ic cells.
T ansmission o mi ochond ial DNA o he nuclea genome occu s in neoplas ically ans o med cells, bu we do no ex-
clude he possibili y ha some mi ochond ial-nuclea DNA usions obse ed in cance occu ed yea s ea lie in no mal
soma ic cells.
[Supplemen al ma e ial is a ailable o his a icle.]
Soma ically acqui ed s uc u al ea angemen s a e common ea-
u es o he nuclea genomes o cance cells. These may ange
om simple ch omosomal ea angemen s (Campbell e al.
2008) o mo e complex, compound pa e ns, such as ch omo-
h ipsis (S ephens e al. 2011) and ch omoplexy (Baca e al.
2013), o mobiliza ion o ansposable elemen s (Lee e al. 2012;
Tubio e al. 2014). In ach omosomal ea angemen s a e gene al-
ly mo e common han in e ch omosomal ea angemen s, indi-
ca ing a highe likelihood o joining a double-s and b eak in a
ch omosome o ano he b eak in he same ch omosome despi e
he a ailabili yo a much la ge quan i yo nuclea DNA om o h-
e ch omosomes (S ephens e al. 2009).
In addi ion o he nuclea genome, human cells ha e a ew
hund ed o a ew housand mi ochond ia, each ca ying one o a
ew copies o he 16,569-bp-long ci cula m DNA (Smei ink e al.
2001; F iedman and Nunna i 2014; Ju e al. 2014). Du ing endo-
symbio ic co-e olu ion, mos o he gene ic in o ma ion p esen
in he ances al mi ochond ion has ans e ed o he nuclea ge-
nome (G ay e al. 1999; Adams and Palme 2003; Timmis e al.
2004). An appa en bu s o m DNA ans e occu ed du ing p i-
ma e e olu ion ∼54 million yea s ago (Ghe man e al. 2007) and
occasional, p obably mo e ecen , ans e in humans has been
obse ed in he ge mline (Tu ne e al. 2003; Goldin e al. 2004;
Chen e al. 2005; Milla e al. 2010; Dayama e al. 2014).
44
Deceased.
45
These au ho s con ibu ed equally o his wo k.
46
A ull lis o membe s is p o ided in he Supplemen al Ma e ial.
Co esponding au ho : [email p o ec ed]
A icle published online be o e p in . A icle, supplemen al ma e ial, and publi-
ca ion da e a e a h p://www.genome.o g/cgi/doi/10.1101/g .190470.115.
F eely a ailable online h ough he Genome Resea ch Open Access op ion.
© 2015 Ju e al. This a icle, published in Genome Resea ch, is a ailable unde a
C ea i e Commons License (A ibu ion 4.0 In e na ional), as desc ibed a
h p://c ea i ecommons.o g/licenses/by/4.0/.
Resea ch
814 Genome Resea ch 25:814–824 Published by Cold Sp ing Ha bo Labo a o y P ess; ISSN 1088-9051/15; www.genome.o g
www.genome.o g
Cold Sp ing Ha bo Labo a o y P ess on Oc obe 16, 2016 - Published by genome.cshlp.o gDownloaded om Cold Sp ing Ha bo Labo a o y P ess on Oc obe 16, 2016 - Published by genome.cshlp.o gDownloaded om Cold Sp ing Ha bo Labo a o y P ess on Oc obe 16, 2016 - Published by genome.cshlp.o gDownloaded om
Al hough m DNA nuclea ans e in a HeLa cell line de i a i e,
and hus occu ing in i o, has been epo ed (Shay e al. 1991),
de no o nuclea ans e o m DNA in animal soma ic issues has
no p e iously been comp ehensi ely s udied o ou knowledge.
To in es iga e he possibili y o soma ic mi ochond ial-nuclea
DNA usion, we analyzed nex -gene a ion pai ed-end DNA
whole-genome sequencing da a om 559 p ima y cance s, 28 can-
ce cell lines ( e e ed as 587 cance whole genome below) and no -
mal DNAs om he same indi iduals (Supplemen al Table 1).
Resul s
Disco e y o soma ic m DNA ans e s o cance nuclea genomes
F om he 587 pai s o cance and no mal whole-genome se-
quencing da a, we sea ched o cance -speci ic clus e s o dis-
co dan pai ed-end sequence eads in which one membe o
he ead-pai mapped o he nuclea genome and he o he o
he mi ochond ial genome, and hen cha ac e ized he nuclea -
mi ochond ial genome junc ions o nucleo ide esolu ion using
indi idual sequence eads ha b idged he junc ion (Fig. 1A). In
12 samples (o e all posi i e a e 2.0%, 12 ou o 587 samples), we
obse ed 25 cance -speci ic mi ochond ial-nuclea DNA junc ions
(Table 1; Supplemen al Figs. 1–6). Gi en ha he e a e wo junc-
ions o a single in eg a ion e en , we conclude ha he e a e
mos likely 16 independen m DNA inse ions (Table 1). In addi-
ion o soma ic ans e s, we obse ed se e al no el a e ge mline
(inhe i ed) e en s ha we esha ed be ween cance and pai ed no -
mal samples (Supplemen al Table 2; Supplemen al Ma e ial).
B eas cance PD11372a showed a soma ically acqui ed in-
eg a ion o almos he en i e human m DNA sequence
(16,556 bp) in o a highly ampli ied 2.75-Mb-long egion o
Ch omosome 10q22.3. The in eg a ion e en was s ongly sup-
po ed by bo h disco dan and spli ead clus e s (Fig. 1B–D)
and was con i med by sho - and long- ange PCR ac oss he nu-
clea -mi ochond ial genome junc ions (Supplemen al Figs. 7, 8;
Supplemen al Table 3). I was no ound in no mal issue (blood)
om he same indi idual o om all he o he cases and did no
ma ch any known inhe i ed nuclea m DNA-like sequences
(known as num s) (Ghe man e al. 2007; Hazkani-Co o e al.
2010). Consis en wi h i s soma ic o igin, he m DNA used o
he nuclea genome ha bo ed sequence polymo phisms iden ical
o hose p esen in he mi ochond ia o his indi idual (14,905
G > A; 15,028 C > A; 15,043 G > A; 15,326 A > G; 15,452 C > A,
and 15,607 A > G). Fluo escence in si u hyb idiza ion (FISH) expe -
imen spe o medon o malin- ixedpa a inembedded issuecon-
i med ha he used DNA segmen exis s in he nuclei o cance
cells (Fig. 1E).
In o al, we ound 10 p ima y cance s (1.8%, 10/559) and wo
cance cell lines (7.1%, 2/28) wi h soma ic m DNA in eg a ions
in o hei nuclea genomes (Table 1; Supplemen al Figs. 1–6). O
he 12 cance s, wo (p ima y cance PD13296a and cance cell
line NCI-H2087) had mo e han one mi ochond ial-nuclea DNA
ansloca ion e en . All in eg a ions we e suppo ed by bo h dis-
co dan andspli eadsand u he con i medbyPCRac oss henu-
clea -mi ochond ial genome junc ions (Supplemen al Fig. 7;
Supplemen alTable3).Allinhe i edm DNAsubs i u ionpolymo -
phisms nea hese b eakpoin s we e de ec ed (Table 1). To u he
isualize he ans e e en s, we pe o med high- esolu ion FISH
on s e ched DNA ibe s ( ibe FISH) om he melanoma cell line,
CP66-MEL (Fig. 2A).
Soma ic nuclea in eg a ion o m DNA is equen ly combined
wi h o he ea angemen s o he nuclea genome
The a e o soma ic nuclea ans e o m DNA may a y acco ding
o umo ype. T iple-nega i e b eas cance showed a i e old
highe equency compa ed o es ogen- ecep o (ER) posi i e
b eas cance s (6.2% and 1.2%, espec i ely; Fishe ’s exac es P
= 0.002). T iple-nega i e b eas cance genomes ca y a highe
numbe o ch omosomal ea angemen s han ER-posi i e b eas
(a e age 254 and 94, espec i ely, in ou da a se ). As a esul , he e
was a sugges i e posi i e co ela ion be ween he numbe o ch o-
mosomal ea angemen s and m DNA ans e s (Mann-Whi ney
U es , one-sided P= 0.05) (Fig. 2B).
The leng h o m DNA agmen s ans e ed anged om 148
bp o en i e mi ochond ial genomes (16.5 kb) (Table 1). In e es -
ingly, b eakpoin s in m DNA we e en iched nea he mi ochond i-
al genome hea y s and o igin o eplica ion (χ
2
es , P= 0.0005)
(Fig. 2C). This sugges s ha he gene a ion o m DNA segmen s
o be in eg a ed in o he nuclea genome is no andom and may
occu in a m DNA eplica ion-dependen manne (Lenglez e al.
2010).
O he 25 mi ochond ial-nuclea DNA junc ions, a leas 17
(68.0%) we e clea ly associa ed wi h o he nuclea ch omosomal
ea angemen s (e.g., in e sions, ansloca ions, and la ge dele-
ions) in he icini y (Table 1; Supplemen al Figs. 1–6). Fo
example, wi h espec o PD11372a desc ibed ea lie , genomic
agmen s om Ch omosomes 10, 11, and m DNA gene a ed com-
plex de i a i e ch omosomes (Fig. 3A). In PD6047a, an m DNA
agmen was in ol ed in chains o complex genomic ansloca-
ions in ol ing Ch omosomes 6, 7, 11, 22, and X (Fig. 3B). In
PD10014a, a local in e sion was combined wi h he m DNA in e-
g a ion e en (Fig. 3C), and in PD4252a, a 16.5-kb m DNA in eg a-
ion was ound in a posi ion on he X Ch omosome om which
∼20 kb o nuclea DNA had been soma ically dele ed (Fig. 3D).
Thus, m DNA is o en in eg a ed in o nuclea genomes in he
icini y o , o as pa o , complex ea angemen s. Al hough ge m-
line num s end o occu nea ansposable elemen s such as
LINEs and SINEs (Mishma e al. 2004), we do no obse e
his associa ion o soma ic e en s (χ
2
es , wo-sided P= 0.33)
(Supplemen al Table 4).
The mechanism and iming o soma ic nuclea ans e o m DNA
The e was o e lapping sequence mic ohomology ( om 1 o 4 bp)
in 20/25 b eakpoin s (80%) (Fig. 4A,B; Table 1; Supplemen al Figs.
1–6), subs an ially mo e han expec ed by chance (χ
2
es , P=5×
10
−26
). Thus, DNA sequence mic ohomology plays an impo an
ole in mi ochond ial-nuclea DNA in eg a ion e en s, al hough
blun -end DNA epai was also obse ed. In wo b eakpoin s, we
also ound non empla ed sho -nucleo ide inse ions (1 and 4 bp
long) (Fig. 4A; Table 1). O e all, hese ea u es a e cha ac e is ic
o DNA double-s and b eak epai by nonhomologous end join-
ing (NHEJ) (Has ings e al. 2009). Howe e , hey do no ule ou
eplica ion-based mechanisms swi ching empla e be ween nucle-
a and m DNA, such as mic ohomology-media ed b eak-induced
eplica ion (MMBIR) (Liu e al. 2011).
We in es iga ed he iming o soma ic m DNA in eg a ion
in o he nuclea genome by assessing cases in which a me as a ic
sample had been sequenced in addi ion o he p ima y umo .
One such case (PD4252a) showed he mi ochond ial-nuclea in e-
g a ion e en in he p ima y bu no in he me as asis (Fig. 4C), in-
dica ing ha m DNA ans e o he nucleus can occu a e
Nuclea in eg a ion o mi ochond ial DNA in cance
Genome Resea ch 815
www.genome.o g
Cold Sp ing Ha bo Labo a o y P ess on Oc obe 16, 2016 - Published by genome.cshlp.o gDownloaded om
neoplas ic ans o ma ion and du ing he cou se o subclonal e o-
lu ion o he cance . The o he (PD6728b) showed i in bo h he
p ima y and me as asis (Fig. 4C), sugges ing ha his e en oc-
cu ed in he common ances al cance clone o in no mal soma ic
cells p io o neoplas ic change.
Nuclea ans e o m DNA is unexpec edly equen in human
soma ic cells
To ob ain a pe spec i e on he equency o mi ochond ial-nuclea
DNA ansloca ion, we compa ed i s a e o ha o in anuclea
1
0
20
40
60
80
100
120
140
160
180
200
220
240
2
0
20
40
60
80
100
120
140
160
180
200
220
240
3
0
20
40
60
80
100
120
140
160
180
4
0
20
40
60
80
100
120
140
160
180
5
0
20
40
60
80
100
120
140
160
180
6
0
20
40
60
80
100
120
140
160
7
0
20
40
60
80
100
120
140
8
0
20
40
60
80
100
120
140
9
0
20
40
60
80
100
120
140
10
0
20
40
60
80
100
120
11
0
20
40
60
80
100
120
12
0
20
40
60
80
100
120
13
0
20
40
60
80
100
14
0
20
40
60
80
100
15
0
20
40
60
80
100
16
0
20
40
60
80
17
0
20
40
60
80
18
0
20
40
60
19
0
20
40
20
0
20
40
60
21
0
20
40
22
0
20
40
x
0
20
40
60
80
100
120
140
umou
blood
1 Mb
10 kb
100 bp
1 bp
1 bp
100 bp
10 kb
1 Mb
PD11372a
CTCCTGGGTG AGAAA
CTCCTGGGTG TTGGCCTCAC
GATAT TTGGCCTCAC
Fusion
Ch 10
m DNA
TACTGTGGC CC AGACCTCTT
ACACT GC AGACCTCTT
TACTGTGGC CC CTCAG
x 60
x 27 x 25
x 76
ch 10 (+)
81,670,932
MT (-)
15,157
MT (-)
15,171
ch 10 (+)
78,920,385
x 24
x 28
16,556 bp
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
**
ch 10 posi ion (Mb)
0 20406080100120
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
**
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
**
*
*
*
*
*
*
*
*
**
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
**
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
**
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
**
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
0
2
4
6
8
10
Copy numbe
*
**
*
**
**
*
*
*
*
*
*
*
*
*
*
*
*
*
*
**
*
**
*
*
**
**
*
*
*
*
*******
****
*
*
*
*
**
*
*
**
**
*
*
*
*
**
*
*******
****
***
*
*
*
*
**
*
*
*
*
*
*
*
*
*
*
*
**
*
*
****
*
*
****
***
*
*
**
**
****
****
***
**
**
*
*
*
*
**
*
***
****
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
**
*
**
****
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
**
*
*
**
****
*
*
*
*
*
**
*
*
**
*
**
**
****
*
*
**
*
*
*
**
**
*
**
*
**
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
**
*
*
*
*
*
**
*
*
***
*
**
*
**
*
*
*
*
*
*
*
*
**
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
**
*
*
*
*
*
**
*
*
*
*
*
*
*
**
**
*
*
****
****
*
*
*
*
*
*
*
*
*
*
*
**
**
**
**
**
*
*
*
*
**
**
*
**
*
***
*
**
********
*
**
**
****
**
***
***
*
*
*
*
*
*
*
*
*
*
*
****
*
*
*
*****
*
***
*
**********
*
**
*
*
*
*
**
*
*
*
*
*
*
*
*
**
*
*
*
*
m DNA
PD11372a
dele ion ype
andem duplica ion ype
in e sion ype (head-head)
in e sion ype ( ail- ail)
m DNA ch 10: 80Mb ch 10: 78Mb me ged
E
DC
BA
mi ochond ial DNA
(inse ed segmen )
(2) DRs
(1) SRs (Nu)
nuclea DNA nuclea DNA
Mapping o he
e e ence genome
(using BWA)
cance genome (3) SRs (MT)
nuclea (Nu)
genome
mi ochond ial (MT)
genome
(2) DRs
(1) SRs (Nu)
mapped in Nu. genome
as ma e-unmmaped
(3) SRs (MT)
mapped in MT genome
as ma e-unmmaped
Figu e 1. Disco e y o soma ic nuclea m DNA ans e om PD11372a. (A) The s a egy o de ec ion o nuclea m DNA ans e e en s. See Me hods
o a de ailed desc ip ion. (SRs) Spli - eads, (DRs) disco dan eads, (Nu) nucleus, (MT) mi ochond ia. (B) G aphical ep esen a ion o disco dan ead clus-
e s in PD11372a and i s pai ed-no mal issue (PD11372b). The ed a ow indica es umo -speci ic disco dan - ead clus e s in Ch 10. Ch omosome ideo-
g ams a e shown in he ou e laye . The dis ance be ween each disco dan ead and one p io o i ( he in e - ead dis ance) is plo ed on he e ical axis on a
log-scale in he middle ( umo ) and inne laye (blood). Blue do s shown in he middle laye ep esen known num s. (C) m DNA in eg a ion in PD11372a.
B eakpoin sequences a e shown.Red ec anglehighligh s mic ohomology. Numbe s o disco dan spli eads a ep esen ed. Inhe i ed m DNA subs i u ion
polymo phisms a e shown by ed as e isks. (D) Rea angemen a chi ec u es o Ch omosome 10 o PD11372a. DNA copy numbe s a e shown by black
do s. The copy numbe o 2.75-Mb-long egion used wi h m DNA is colo ed in ed. Reads suppo ing ea angemen s (la ge dele ions, andem dupli-
ca ions, ail- ail and head-head in e sions) a e shown by a cs and e ical lines. Ch 10-m DNA usions a e shown wi h ed a ows. (E) Nuclea FISH con i ms
he mi ochond ial-nuclea DNA usion in he nucleus. (Red) Ch 10 (80 Mb), (blue) Ch 10 (78 Mb), and (g een) m DNA.
Ju e al.
816 Genome Resea ch
www.genome.o g
Cold Sp ing Ha bo Labo a o y P ess on Oc obe 16, 2016 - Published by genome.cshlp.o gDownloaded om
Table 1. Summa y o soma ic mi ochond ial-nuclea DNA usions iden i ied om 12 cance samples
Tissue Sample
Le junc ion Righ junc ion F ag. size
(bp)
Mic o-
homology
(bp,bp)
Va ian s
(#D/#P)
a
Con ex o ea angemen
Nuclea MT MT Nuclea
P ima y PD11372a 10+:81,670,932] [M−:15,157 M−:15,171] [10+:78,920,385 16,556 (0,1) 6/6 m DNA inse ion wi h complex
ea angemen s
PD4252a X+:45,631,665] [M+:14,450 M+:14,496] [X+:45,652,120 16,616 (2,1) 2/2 m DNA inse ion wi h la ge ch . dele ion
PD6047a X+:14,944,764] [M−:12,735 M−:16,128] [7+:96,923,229 13,177 (1,1) 6/6 Mul iple in e ch omosomal ansloca ions
PD10014a 17−:75,618,348] [M−:13,365 M−:9055] [17+:75,688,733 4311 (2,3) 0/0 m DNA inse ion wi h Ch 17 in e sion
PD13296a 4+:102,463,870] [M−:14,705 M−:13,235] [4+:102,464,084 1471 (4,0) 0/0 m DNA inse ion wi h la ge ch . dele ion
6+:103,639,248] [M+:14,692 M+:14,972]TA AT [6+:103,690,941 281 (2,0) 2/2 m DNA inse ion wi h la ge ch . dele ion
PD6728b 2+:138,664,890] [M−:13,199 M−:13,052] [2−:139,012,040 148 (4,2) 1/1 m DNA inse ion wi h complex
ea angemen s
PD11397a 19−:12,650,382] [M+:16,233 M+:96] [17+:40,005,738 433 (0,2) 1/1 Mul iple in e ch omosomal ansloca ions
PD7404a 1+:44,914,376] [M+:3732 –– >200 (1,–) 0/0 –
PD6733b 6−:45,823,498] [M+:16,107 –– >200 (0,–) 1/1 –
PD11768a 1−:144,944,326] [M+:16,104 –– >200 (4,–) 1/1 –
Cell line CP66-MEL 3+:47,419,506] [M−:7048 M−:16,193] [3+:47,419,447 7425 (1,1) 1/1 m DNA inse ion
NCI-H2087 10+:26,775,605] [M+:1690 –– >200 (1,–) 1/1 –
20−:33,836,717] [M−:5666 –– >200 (1,–) 1/1 –
17−:7,481,787] T [M−:3452 –– >200 (1,–) 1/1
b
Mul iple in e ch omosomal ansloca ions
17−:31,744,235] [M+:4346 –– >200 (3,–) 1/1 –
a
Inhe i ed m DNA polymo phisms in he icini y o b eakpoin s. (#D) Numbe o de ec ed, (#P) numbe o p esen .
b
A soma ically acqui ed he e oplasmic mu a ion in mi ochond ia.
Nuclea in eg a ion o mi ochond ial DNA in cance
Genome Resea ch 817
www.genome.o g
Cold Sp ing Ha bo Labo a o y P ess on Oc obe 16, 2016 - Published by genome.cshlp.o gDownloaded om
in e ch omosomal ansloca ion, aking in o accoun he sizes
and copy numbe s o he mi ochond ial and nuclea genomes.
Ou sequencing da a sugges ha each cance cell ca ies ∼500
copies o ci cula m DNA (median alue 495) (Fig. 5A), amoun -
ing in agg ega e o ∼8 million base pai s (bp) o m DNA (500 cop-
ies × 16.5 kb) enclosed by he mi ochond ial double memb ane
in he cy oplasm o each cance cell. The a e age equency in
he cance s analyzed o mi ochond ial-nuclea DNA usion was
5.1 × 10
−3
junc ions pe million bp o m DNA, only hal he a -
e age a e o in anuclea in e ch omosomal ansloca ion
(1.2 × 10
−2
junc ions pe million bp) and simila o ha o
Ch omosomes 2, 4, and 13 (Fig. 5B). Gi en he mul iple physical
ba ie s o con ac be ween he wo genomes, he esul s indica e
ema kably high a es o m DNA escape, con ac , and/o in eg a-
ion wi h nuclea DNA in human cance cells. These appea o be
conside ably highe han in he ge mline ac oss human e olu-
ion bu compa able o hose obse ed in Saccha omyces ce e isiae
(Tho sness and Fox 1990) and o chlo oplas DNA mig a ion
in o he nucleus in obacco plan s (Me hods; Supplemen al
Ma e ial; Huang e al. 2003).
Discussion
Despi e mul iple physical ba ie s, he e
a e plausible mechanisms by which
m DNA and nuclea DNA could come
in o con ac (Fig. 5C). F ee m DNA can
be eleased in o he cy oplasm om de-
g adingmi ochond iao a e mi ophagy
(Zhang e al. 2008; Eiyama e al. 2013;
Higgins and Coughlan 2014). Deg ada-
ion o mi ochond ia may be accele a ed
in cance cells due o hypoxia and in-
c eased ene gy demands (Zhang e al.
2008; Eiyama e al. 2013; Higgins and
Coughlan 2014). E en wi hou a bespoke
molecula p ocess o anspo a ion,
m DNA could hen, in p inciple, mig a e
o he nucleus du ing mi o ic me aphase
o anaphasewhen henuclea memb ane
has b oken down. When hese e en s a e
coupled wi h concu en double-s and
b eaks (DSBs) and/o eplica ion o k
s alling o nuclea ch omosomal DNA,
m DNA could be picked up and in eg a -
ed in o he nuclea genome as pa o
he p ocess o ejoining DSBs (NHEJ)
(Has ings e al. 2009) o used as an al-
e na i e DNA empla e in eplica ion
(MMBIR) (Liu e al. 2011). Mic onuclei
in cance cells, which can be gene a ed
bye o sinseg ega iono mi o icnuclea
ch omosomes, may con ibu e o he
e en s.Ch omosomesinmic onuclei e-
quen ly unde go de ec i e and delayed
DNA eplica ion, esul ing in ex ensi e
agmen a ion wi h subsequen jumbled
ejoiningcompa ed o hei o iginalo de
and o ien a ion (C as a e al. 2012; Fo -
men e al. 2012). Thus, m DNA ag-
men s inco po a ed in o mic onuclei
could end up used o sha e ed nuclea
ch omosomes. I is wo hy o no e ha
m DNA escaping o he nucleus canbe ac i elyused o DNA epai
in Saccha omyces ce e isiae (Ricche i e al. 1999; Yu and Gab iel
1999), pa icula ly when e o - ee DSB DNA epai is no possible.
Whe he his applies in mammalian cells is unknown.
Some o he soma ic nuclea m DNA in eg a ions we iden i-
ied a e di ec ly adjacen o nuclea genes. Fo example, nuclea -
m DNA usion in PD11372a occu ed in he i h in on o he
KCNMA1 gene, a po assium channel equen ly ampli ied in p os-
a e and b eas cance s (Oegge li e al. 2012). Howe e , we do no
ind ob ious en ichmen o he nuclea -m DNA usion b eak-
poin s nea human nuclea genes. RNA-seq om he NCI-H2087
cell-line indica es ha m DNA agmen s in he nucleus o he
cell line a e no exp essed as pa s o mi ochond ial-nuclea usion
ansc ip s.Thus, hemajo i yo henuclea m DNA ansloca ion
e en s a e likely o be passenge e en s, simila o mu a ions o all
o he ypes in mos cance genomes. Howe e , we do no ex-
clude hepossibili y ha someo hese e en smayha e unc ional
consequences in human cance by gene a ing usion mRNA an-
sc ip s (Shay e al. 1991) and/o unca ing cance genes by
m DNA inse ion wi hin exons.
ch 3 (downs eam)ch 3 (ups eam)
m DNA
(7.4kb)
A
CP66-MEL (melanoma cell-line)
250
500
1000
B
78
133
125
Cance issue ypes
B eas (ER + e)
O he b eas
B eas ( iple − e)
O he ypes
m DNA
ans e s -+
cance samples
0
1
2
3
4
0
2
4
6
8
10
12
14
0.5-3kb
3-5.5kb
5.5-8kb
8-11.5kb
11.5-14kb
14-0.5kb
Ra io (obs/exp)
F equency o e en s
Loca ion o b eakpoin s in mi ochond ial genome
Expec ed
Obse ed
Ra io (obs/exp)
P = 0.00052 *
D-Loop
RNAs
16s12s CO1 ND5ND4
ND6
Replica ion o igin
( H s and)
Replica ion o igin
( L s and)
C
CYB
Figu e 2. Fea u es o soma ic m DNA nuclea ans e in 12 cance samples. (A) Fibe FISH isualizes
he mi ochond ial-nuclea DNA usion om he CP66-MEL cell line. (B) Posi i e co ela ion be ween
m DNA ans e and numbe s o nuclea ch omosomal ea angemen s (la ge dele ion, andem duplica-
ion, in e sion, and ansloca ion) in cance genomes. Median alues a e shown. (C) m DNA b eak-
poin s a e en iched in he 14 kb- o 500-bp egion o he MT genome. (Top) Blue and ed ba s
ep esen he expec ed and obse ed numbe s o b eakpoin s in each in e al o MT genome, espec-
i ely. G een line shows a io be ween obse ed and expec ed numbe s. A χ
2
es was applied o es
en ichmen . (Bo om) Schema ic s uc u al ea u es o he MT genome co esponding o he in e als
a e shown.
Ju e al.
818 Genome Resea ch
www.genome.o g
Cold Sp ing Ha bo Labo a o y P ess on Oc obe 16, 2016 - Published by genome.cshlp.o gDownloaded om

This s udy has shown ha usion o m DNA o nuclea DNA
occu s in human soma ic cells a a a e simila o ha o ansloca-
ion be ween nuclea ch omosomes. Physical mig a ion o m DNA
in o he nucleus may be much mo e equen in s em cells han
ones in a e minally di e en ia ed s age (Schneide e al. 2014).
Fu he s udies will need o add ess he mechanisms by which
he appa en physical ba ie s o con ac be ween mi ochond ial
and nuclea DNA a e so e ec i ely o e come.
Me hods
Samples and sequencing da a
We analyzed 559 p ima y umo s and 28 cance cell-lines in his
s udy. Pai ed-no mal samples o all he cance s we e also included
in his s udy in pa allel. Whole-genome
sequencesusedin hiss udywe egene a -
ed by Illumina pla o ms (ei he Genome
Analyze o HiSeq 2000). Cance ge-
nomeswe esequenced oa leas 25×co -
e age. Wi h espec o TCGA da a, we
downloaded aligned BAM iles h ough
UCSC CGHub (h p://cghub.ucsc.edu).
Sequencing eads we e aligned on he
human e e ence genome build 37
(GRCh37) and human e e ence m DNA
sequence ( e ised Camb idge e e ence
sequence, CRS) (And ews e al. 1999),
mainly by he BWA alignmen ool (Li
and Du bin 2009). SAM ools (Li e al.
2009) was used o manipula ing se-
quence eads.
Calling mi ochond ial-nuclea DNA
usion e en s
We employed a pipeline o iden i ica-
ion o pu a i e m DNA ansloca ion
o ch omosomal DNA (Fig. 1A). F om
pai ed-end whole-genome sequencing
da a o umo s, we ex ac ed disco dan
eads (DRs), whe e one end aligned
uniquely o m DNA and he o he end
o nuclea DNA. In all cases, bo h ends
mus ha e a mapping quali y g ea e
han ze o. Those disco dan eads a e
clus e ed oge he using he ollowing
c i e ia: eads sha ing (1) close alignmen
posi ions (<500 nucleo ides) o bo h
ends on nuclea and m DNA, and (2)
he same o ien a ions.In o de o emo e
alse posi i es, we emo ed clus e s sup-
po ed by less han i e disco dan eads.
In o de o emo e po en ial ge m-
line calls, se e al il e s a e applied o
he umo candida e clus e . The clus e s
om umo cells we e emo ed i hey
o e lap wi h clus e s iden i ied om
ma ched and/o unma ched no mal is-
sues by mo e ole able c i e ia (suppo ed
by mo e han one disco dan ead)
om (1) i s pai ed-no mal issue, and (2)
om he o he 586 unma ched no mals.
Fil e ed clus e s we e u he e ined wi h known ge mline human
num s, a combined se om he human e e ence genome (hg19)
de ec ed by BLAT (Ken 2002) (n= 123) and om Simone e al.
(2011) (n= 766). Finally, 25 clus e s we e selec ed as soma ic
candida es.
Nucleo ide- esolu ion b eakpoin s o he ansloca ion junc ions
To ob ain nucleo ide- esolu ion b eakpoin s, we sea ched o spli -
eads(SRs) wi honeo he endsspanning hejunc iono he ans-
loca ion. We ex ac ed “o phan”o “ma e-unmapped” eads (one
end o a ead is unmapped by he BWA aligne ) in he icini y
(<1000bp)o disco dan - eadclus e sonnuclea andmi ochond i-
al genome sequences. Sequences om he unmapped end a e hen
e-aligned by BLAT (Ken 2002), which enablesspli - ead mapping.
A
10
0
5
5
10
15
20
25
30
35
3
40
45
4
50
55
60
65
6
70
75
80
85
8
90
95
100
105
110
115
1
120
125
130
135
11
0
5
5
10
15
5
20
25
5
30
35
40
45
5
50
55
5
60
65
5
70
75
5
80
85
5
90
95
5
100
105
110
115
5
120
125
130
135
5
MT
0
PD11372a
D
B
ch X(+):45,631,665 ch X(+):45,652,120
m DNA inse ion (16.5kb)
nuclea DNA dele ion (20 kb)

ch omosome X
0 20 40 60 80 100 120 140
p11.3
Copy numbe
45.61 45.62 45.63 45.64 45.65 45.66 45.67
024
o
oooo
o
oooooooo
o
ooooooooooooooo
o
o
o
o
oo
oo
o
ooo
o
oooooooooo
oooooooooooooo
o
ooooooooooooooooooooo
oo
oooooooooooooo
ooooooo
ooo
ooooooooooooooo
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
0
8
abe an ead
clus e s
PD4252a
PD6047a
2
0
10
20
30
40
50
60
70
80
90
100
110
120
130
140
150
160
170
180
190
200
210
220
230
240
3
0
10
20
30
40
50
60
70
80
90
100
110
120
130
140
150
160
170
180
190
4
0
10
20
30
40
50
60
70
80
90
100
110
120
130
140
150
160
170
180
190
5
0
10
20
30
40
50
60
70
80
90
100
110
120
130
140
150
160
170
180
6
0
10
20
30
40
50
60
70
80
90
100
110
120
130
140
150
160
170
7
0
10
20
30
40
50
60
70
80
90
100
110
120
130
140
150
8
0
10
20
30
40
50
60
70
80
90
100
110
120
130
140
10
0
10
20
30
40
50
60
70
80
90
100
110
120
130
11
0
10
20
30
40
50
60
70
80
90
100
110
120
130
12
0
10
20
30
40
50
60
70
80
90
100
110
120
130
13
0
10
20
30
40
50
60
70
80
90
100
110
16
0
10
20
30
40
50
60
70
80
90
22
0
10
20
30
40
50
x
0
10
20
30
40
50
60
70
80
90
100
110
120
130
140
150
MT
0
m DNA (-)
ch 7(+):96,923,229
ch X(+):14,944,764
13.1 kb
m DNA(-): 13,365 - 9,055
(4.3kb)
ch 17(+):75,688,733
local in e sion
C
ch 17(+):75,564,373
ch 17(-):75,655,898 - 75,618,348
(37.6kb)
m DNA inse ion
PD10014a
Figu e 3. Concu ence o soma ic m DNA nuclea ans e s wi h o he s uc u al a ia ions. The com-
plex web o ea angemen s in he icini y o mi ochond ial-nuclea DNA usions om ou examples. (A)
In PD11372a, m DNA in eg a ion wi h complex ea angemen s be ween Ch 10 and 11. (B)In
PD6047a, m DNA in eg a ion wi h complex ea angemen s among Ch 6, 7, 11, 22, and X. (A,B)
DNA copy numbe s a e shown by black do s wi h a log scale. Red lines ep esen ansloca ions in ol ing
m DNA. (C) In PD10014a, m DNA in eg a ion combined wi h a local in e sion (yellow). (D) In PD4252a,
m DNA in eg a ion wi h a local dele ion. DNA copy numbe s a e shown wi h blue do s and lines.
Abe an ead clus e s (disco dan and spli eads) a e shown by g een and ed a ows, espec i ely.
Nuclea in eg a ion o mi ochond ial DNA in cance
Genome Resea ch 819
www.genome.o g
Cold Sp ing Ha bo Labo a o y P ess on Oc obe 16, 2016 - Published by genome.cshlp.o gDownloaded om
Valida ion by PCR
A PCR alida ion assay o he soma ic m DNA ans e was pe -
o med using genomic DNA om bo h cance and pai ed-no mal
issues. P ime s we e designed o ampli y all he b eakpoin s
(Supplemen al Table 3). The sho - agmen PCR eac ions we e
pe o med as p e iously desc ibed (Tubio e al. 2014). Wi h espec
o long- ange PCR, elonga ion ime was inc eased 1 min pe 1 kb.
Gene a ion o FISH p obes
Human bac e ial a i icial ch omosomes (BAC) and osmid clones
used in his s udy we e ob ained om he clone a chi e eam o he
Wellcome T us Sange Ins i u e. Plasmid
DNA was p epa ed using he PhaseP ep
BAC DNA ki (Sigma-Ald ich). Human
m DNA was isola ed om lymphoblas-
oid cells using a Mi ochond ial DNA
Isola ion ki (Abcam).
P obes o use in FISH we e made as
desc ibed be o e (G ibble e al. 2013).
Pu i ied m DNA and plasmid DNA we e
i s ampli ied using a GenomePlex
Whole Genome Ampli ica ion (WGA)
ki (Sigma-Ald ich) ollowing he manu-
ac u e ’s p o ocols, hen labeled using
a WGA eampli ica ion ki (Sigma-
Ald ich) wi h a cus om-made dNTP mix.
P obes o in e phase FISH we e labeled
di ec ly wi h Aminoallyl-dUTPs - ATTO-
488, -Cy3, -Texas Red, and -Cy5 (Jena
Bioscience); p obes o ibe -FISH we e la-
beledwi hBio in-16-dUTP,Digoxigenin-
11-dUTP (Roche), and DNP-11-dUTP
(Pe kinElme ).
Valida ion by ibe -FISH wi h single-
molecule DNA ibe s gene a ed by
molecula combing
Single-molecule DNA ibe s om he
cance cell line, CP66-MEL, we e p e-
pa ed by molecula combing (Michale
e al. 1997) ollowing he manu ac u e ’s
ins uc ions (Genomic Vision). B ie ly,
he cells we e embedded in a low-mel -
poin aga ose plug (1 million cells pe
plug), ollowed by p o einase K diges-
ion, washing in 1 × TE (10 mM T is, 1
mM EDTA, pH 8.0) and be a-aga ose di-
ges ion s eps. The DNA ibe s we e me-
chanically s e ched on o saline-coa ed
co e slips using a Molecula Combing
Sys em (Genomic Vision).
Fo ibe -FISH, ∼500 ng o labeled
DNA om each p obe and 4 μg o human
Co -1 DNA (In i ogen) we e p ecipi-
a ed using e hanol, hen esuspended
in a mix (1:1) o hyb idiza ion bu e
(con aining 2 × SSC, 10% sa kosyl, 2 M
NaCl, 10% SDS, and blocking aid [In i-
ogen]) and deionized o mamide ( inal
concen a ion 50%). Co e slips coa ed
wi h combed DNA ibe s we e dehyd a ed h ough a 70%, 90%,
and 100% e hanol se ies and aged a 65°C o 30 sec, ollowed by
dena u a ion in an alkaline dena u e solu ion (0.5 M NaOH, 1.5
M NaCl) o 1–3 min, h ee washes wi h 1×PBS (In i ogen), and
dehyd a ion h ough a 70%, 90%, and 100% e hanol se ies. The
p obe mix was dena u ed a 65°C o 10 min be o e being applied
on o he co e slips, and he hyb idiza ion was ca ied ou in a 37°
C incuba o o e nigh . The pos -hyb idiza ion washes consis ed o
wo ounds o washes in 50% o mamide/2 × SSC ( / ), ollowed
by wo addi ional washes in 2 × SSC. All pos -hyb idiza ion washes
we e done a 25°C, 5 min each ime. Digoxigenin-11-dUTP (Roche)
labeled p obes we e de ec ed using a 1:100 dilu ion o monoclonal
ch 4: GGCGA AACCC CATT TCTACT
Fusion: GGCGA AACCC CATT GGTCGT
m DNA: TTTTT CATAT CATT GGTCGT
TAAT
A
ch 4(+)102,463,870 m DNA(-):14,705
PD13296a (2 m DNA nuclea in eg a ion e en s)
m DNA(-):13,235 ch 4(+):102,464,084
4bp mic ohomology blun -end DNA joining
GCTACGATTT CTTTTGATGT : m DNA
GCTACGATTT AAATAACCAC : Fusion
GTTATCTTCA AAATAACCAC : ch 4
In eg a ion #1
In eg a ion #2
ch 6(+):103,639,248 m DNA(+)
14,692
m DNA(+)
14,972
ch 6(+):103,690,941
ch 6: TTGTAAGA AC TAATAGAATG
Fusion: TTGTAAGA AC AACCACGACC
m DNA:CACGGACT AC AACCACGACC GTAAATTATG GCTGAATCAT : m DNA
GTAAATTATG TAAT AAAATATTTG : Fusion
CTGGGTCCTA AAAATATTTG : ch 6
2bp mic ohomology non- empla e 4bp inse ion
PD6728b
B
ch 2(+):138,664,890 m DNA(-)
13,199
m DNA(-)
13,052
ch 2(-):139,012,040
ch 2: TCATCT TGCT TGCGTTTTGC
Fusion: TCATCT TGCT GCGAACAGAG
m DNA:GCAGAC TGCT GCGAACAGAG
4bp mic ohomology
GGGGTGGGGC CT TCTATGGC : m DNA
GGGGTGGGGC CT GACTGCAG : Fusion
CTTGGTCTTG CT GACTGCAG : ch 2
2bp mic ohomology
C
PD4252
Fe ilized egg
Blood
(PD4252b)
MRCA
LN me as asis
(PD4252c)
P ima y locus
(PD4252a; subclonal)
m DNA ans e
PD6728
Fe ilized egg
Blood
(PD6728a)
MRCA
LN me as asis
(PD6728c; clonal)
P ima y locus
(PD6728b; clonal)
m DNA ans e
T ans o ma ion
T ans o ma ion
Figu e 4. Nucleo ide- esolu ion b eakpoin sequences and he iming o soma ic m DNA nuclea in e-
g a ion. (A) B eakpoin sequences o nuclea -m DNA usions in PD13296a. Red ec angles highligh se-
quence mic ohomology and non empla e nucleo ides inse ion. (B) B eakpoin sequences o nuclea -
m DNA usions in PD6728b. Red ec angles highligh sequence mic ohomology. (C) Phylogene ic ees
showing he iming o soma ic m DNA nuclea ans e s in PD4252 and PD6728 samples. (MRCA) Mos
ecen common ances o cell.
Ju e al.
820 Genome Resea ch
www.genome.o g
Cold Sp ing Ha bo Labo a o y P ess on Oc obe 16, 2016 - Published by genome.cshlp.o gDownloaded om
mouse an i-dig an ibody (Sigma-Ald ich) and a 1:100 o Texas Red-
X-conjuga ed goa an i-mouse IgG (Molecula P obes/In i ogen);
DNP-11-dUTP (Pe kinElme ) labeled p obes we e de ec ed using a
1:100 dilu ion o Alexa 488-conjuga ed abbi an i-DNP IgG and
1:100 Alexa 488-conjuga ed donkey an i- abbi IgG (Molecula
P obes/In i ogen); bio in-16-dUTP (Roche) labeled p obes we e
de ec ed wi h one laye 1:100 o Cy3-a idin (Sigma-Ald ich). A e
de ec ion, slides we e moun ed wi h SlowFade Gold moun ing sol-
u ion con aining 4′,6-diamidino-2-phenylindole (Molecula
P obes/In i ogen). Images we e isualized on a Zeiss AxioImage
D1 mic oscope. Digi al image cap u e and p ocessing we e ca ied
ou using he Sma Cap u e so wa e (Digi al Scien i ic UK).
Nuclea in e phase FISH
Nuclei ex ac ion om pa a in-embed-
ded issue o pa ien PD11372a and in e -
phase-FISH ollowed Pa e nos e e al.
(2002), wi h he excep ion ha 60-μm-
hick sec ions we e used in ou s udy.
The pos -hyb idiza ion washes consis ed
o wo ounds o washes in 50% o mam-
ide/2 × SSC ( / ), ollowed by wo ad-
di ional washes in 2 × SSC. Slides we e
moun ed wi h SlowFade Gold moun -
ing solu ion con aining 4′,6-diamid-
ino-2-phenylindole (Molecula P obes/
In i ogen). Images we e cap u ed and
p ocessed as desc ibed abo e.
Co ela ion be ween soma ic m DNA
in eg a ion si e and ansposable
elemen s
We pe o med a s udy simila o he p e-
ious epo (Mishma e al. 2004). We
calcula ed he dis ance be ween each
m DNA-inse ion si e (b eakpoin ) and
i s nea es ansposable elemen s (ei-
he o SINE, LINE, LTR, simple epea ,
o DNA ansposon by Repea Maske ,
downloaded om he UCSC Genome
B owse , June 6, 2013). Then, each
m DNA-inse ion si e was ca ego ized
in o one o ou g oups: (A) b eakpoin
wi hin a ansposable elemen ; (B) b eak-
poin wi hin 15 bp om a ansposable
elemen ; (C) wi hin 15–150 bp; and (D),
>150 bp. In o de o unde s and he posi-
ional en ichmen o b eakpoin s om
ansposable elemen s, we andomly
gene a ed in silico b eakpoin posi ions
40 imes as many ( o al n= 1000) as we
obse ed om each ch omosome in he
eal da a se . In silico b eakpoin s loca ed
wi hin gaps o he human e e ence ge-
nome we e emo ed and eplaced by
newly gene a ed inse ions. Fo hese in
silico-gene a ed b eakpoin s, he dis-
ances om he nea es ansposable
elemen s we e calcula ed and hen ca e-
go ized in o one o he ou g oups (A,
B,C, and D). Finally, he di e ence in
he equency o b eakpoin s in each
g oup be ween he obse ed and in silico-gene a ed da a se was
compa ed using a χ
2
es .
Assessmen o m DNA copy numbe s
To unde s and m DNA copy numbe s in a cance cell, we com-
pa ed a e age ead dep h o co e age be ween 22 au osomes and
m DNA. Wi h espec o he umo sequences by whole-genome
sequencing, a e age haploid au osomal co e age (RD
au osome
)
was ob ained om he ead dep h o 2.685-Gb-long au osomal e-
gions (excluding ch omosomal gaps). Likewise, a e age m DNA
co e age (RD
m DNA
) was ob ained om he ead dep h o he
500
1000
2000
Cance issue ypes
B eas
Os eosa coma
O he ypes
P os a e
A
Es ima ed ci cula m DNA copy numbe s (in cy oplasm)
pe cance cell
0.00
0.01
0.02
0.03
ch omosomes
T ansloca ion a e
(# o e en s pe Mb)
ch 2
ch 17
ch 19
m DNA
ch 4
ch 13
B
mi ophagy
mi ochond ial
deg ada ion
escape o
m DNA
nuclea memb ane b eaks down
beginning o mi osis
DNA double-s and b eaks
(ch omosome sha e ing)
and/o eplica ion o k s alling
m DNA mig a ion
o he nucleus (mic onucleus)
( a e> 2x10 pcpg)
cell memb ane
-4
nucleus
(mic onucleus)
m DNA in eg a ion
DSB epai
(NHEJ, MMBIR)
C
Figu e 5. F equency and po en ial mechanisms o soma ic m DNA nuclea ans e in human can-
ce . (A) Es ima ed ci cula m DNA copy numbe s (in he cy oplasm) pe cance cell om 587 cance is-
sues sequenced. The a io o ead dep hs be ween au osomes and m DNA was used (see Me hods). (B)
Simila equency o soma ic nuclea m DNA in eg a ions compa ed o he equency be ween au o-
somes (ch omosomal ansloca ion). (C) A model o soma ic m DNA ans e o he nuclea genomes.
Nuclea in eg a ion o mi ochond ial DNA in cance
Genome Resea ch 821
www.genome.o g
Cold Sp ing Ha bo Labo a o y P ess on Oc obe 16, 2016 - Published by genome.cshlp.o gDownloaded om
16.5-kb mi ochond ial genome. Finally, m DNA copy numbe in a
diploid cell (C
m
) is calcula ed as shown below:
Cm =2×RDm DNA
RDau osome .
Assessmen o ansloca ion a e o au osomes
and mi ochond ia
We iden i ied s uc u al a ia ions among nuclea ch omosomes
(la ge dele ions, andem duplica ions, in e sions, and in e ch o-
mosomal ansloca ions) using he BRASS II algo i hm (Nik-
Zainal e al. 2012), which iden i ies ea angemen s by clus e ing
disco dan ead pai s ha poin o he same junc ion and con i ms
b eakpoin s by local assembly o unmapped eads. The sensi i i y
and speci ici y o he BRASS II algo i hm is equi alen o hose al-
ues o he algo i hm used o mi ochond ial-nuclea DNA usions
(da a no shown). We ex ac ed in e ch omosomal ansloca ions
o calcula e he a e o such e en s. The a e o each haploid au o-
some (R ,ch) is calcula ed as shown below:
R ,ch(e en s pe megabase)=N ,ch/(2×Lch)/Nsam,
whe e N ,ch is he o al numbe o soma ic in e ch omosom-
al ansloca ion junc ions in ol ing a speci ic ch omosome, Lch is
heleng ho henon edundan egiono hech omosomeinmeg-
abases, and Nsam is he o al numbe o samples analyzed. To ob-
ain he unique egion leng h (Lch), we excluded edundan (o
highly epe i i e) sequence leng hs om he ungapped leng h o
each ch omosome. Genomic egions classi ied in one o mo e o
he h ee c i e ia shown below we e de ined as edundan , whe e
ansloca ion e en s could no be easilyde ec ed due o ambiguous
ead alignmen : (1) simple epea s, loca ed by Tandem Repea s
Finde (Benson 1999); (2) segmen al duplica ions wi h mode a e
o high sequence simila i y (≥95%) (Bailey e al. 2002), o (3) epe -
i i e sequences including up o 10 di e en classes o epea s (such
as SINE, LINE, LTR, DNA ansposons, andmic osa elli es), loca ed
by he Repea Maske p og am (h p://www. epea maske .o g),
wi h a low di e gence le el (di e gence < 5%). These non edun-
dan sequence egions we e downloaded om he UCSC Genome
B owse (h p://genome.ucsc.edu).
Simila ly, he a e o mi ochond ial-nuclea DNA ansloca-
ions (R ,m ) was calcula ed as below:
R ,m (e en s pe megabase)=N ,m /(Cm ×Lm DNA)/Nsam,
whe e N ,m is he o al numbe o junc ions o soma ic mi o-
chond ial-nuclea DNA usions iden i ied, C
m
is he median alue
o mi ochond ial genome copy numbe s in a diploid cance cell
calcula ed abo e (495 copies), and L
m DNA
is he leng h o he mi-
ochond ial genome in megabases (0.016569 Mb).
Assessmen o he a es o nuclea m DNA usion
and m DNA escape o he nucleus
Fusion o m DNA o he nuclea genome equi es a leas wo
e en s, each o which could in luence he a e o mi ochond ial-
nuclea DNA usion.Theseincludeescapeo m DNA o henucleus
and in eg a ion o nuclea DNA. Acco ding o his model, he o e -
all numbe o such usion e en s can be calcula ed using he a es
o hese p ocesses (ρ
escape
and ρ
in eg a ion
, espec i ely):
N =Nsam ×Ngen ×
escape ×
in eg a ion,
whe e N is he numbe o o al soma ic mi ochond ial-nu-
clea DNA usion e en s (n= 12), Nsam is he o al numbe o can-
ce issues (n= 587), and Ngen is he numbe o a e age cell
gene a ion om he e ilized egg. Using a easonable assump ion
ha Ngen= 1000, we ob ain he a e o soma ic m DNA usion o
he nuclea genome (ρ
escape
×ρ
in eg a ion
) obe2×10
−5
pe cell pe
cell gene a ion (pcpg). Wi h one mo e e y conse a i e assump-
ion ha ρ
in eg a ion
is 0.1, we ob ain ρ
escape
o be 2 × 10
−4
pcpg,
o a leas one escape e en pe 5000 cell gene a ions. We hypo h-
esize ha he eal ρ
in eg a ion
alue is hough o be much lowe
han 0.1, which esul s in a highe ρ
escape
. Fo example, du ing
he gene a ion o knockou mice, homologous ecombina ion al-
lows one ixa ion e en pe 1000–10,000 mic oinjec ed DNA cop-
ies (B ins e e al. 1985). The in eg a ion a e may, howe e , be
highe han he a e in cance cells wi h de ec i e homologous e-
combina ion-based epai and inc eased a ailabili y o nuclea
double-s and b eaks, which can be joined o by NHEJ o MMBIR.
The m DNA usion o he nuclea genome in he ge mline
( he a e o num s inse ion) is a ound 5 × 10
−6
pe ge m cell pe
indi idual gene a ion in p e ious phylogene ic s udies (Hazkani-
Co o e al. 2010). The a e is equi alen o ∼5×10
−8
pcpg, gi en
ha he numbe o ge m cell di isions pe human gene a ion is
∼100 (401 in males and 31 in emales [D os and Lee 1995]).
Da a access
Sequence da a o sample pai s wi h posi i e m DNA nuclea ans-
e ha e been submi ed o he Eu opean Genome-phenome
A chi e (EGA; h ps://www.ebi.ac.uk/ega/home). The s udy acces-
sion numbe is EGAS00001001234. Sample accession numbe s a e
a ailable in Supplemen al Table 1.
Lis o a ilia ions
1
Cance Genome P ojec , Wellcome T us Sange Ins i u e,
Hinx on, Camb idge CB10 1SA, Uni ed Kingdom;
2
Cy ogene ics
Facili y, Wellcome T us Sange Ins i u e, Hinx on, Camb idge
CB10 1SA, Uni ed Kingdom;
3
Camb idge Uni e si y Hospi als
NHS Founda ion T us , Camb idge CB2 0QQ, Uni ed Kingdom;
4
Depa men o Haema ology, Uni e si y o Camb idge,
Camb idge CB2 0XY, Uni ed Kingdom;
5
Cance Resea ch UK
(CRUK) Camb idge Ins i u e, Uni e si y o Camb idge,
Camb idge CB2 0RE, Uni ed Kingdom;
6
BioCa e, S a egic
Cance Resea ch P og am, SE-223 81 Lund, Sweden;
7
CREATE
Heal h, S a egic Cen e o T ansla ional Cance Resea ch,
SE-221 00 Lund, Sweden;
8
Depa men o Oncology and
Pa hology, Lund Uni e si y Cance Cen e , SE-221 85 Lund,
Sweden;
9
B eak h ough B eas Cance Resea ch Uni , Resea ch
Oncology, King’s College London, Guy’s Hospi al, London SE1
9RT, Uni ed Kingdom;
10
Labo a o y o In e na ional Alliance on
Genomic Resea ch, RIKEN Cen e o In eg a i e Medical
Sciences, 230-0045 Yokohama, Japan;
11
Na ional Cen e o
Genome Medicine, Ins i u e o Biomedical Sciences, Academia
Sinica, Taipei 115, Taiwan;
12
Depa men o Labo a o y
Medicine, Helen Dille Family Comp ehensi e Cance Cen e ,
Uni e si y o Cali o nia, San F ancisco, Cali o nia 94158, USA;
13
Ne he lands Cance Ins i u e, 1066 CX Ams e dam,
Ne he lands;
14
Depa men o Gene al Su ge y, Singapo e
Gene al Hospi al, Singapo e 169608;
15
Depa men o Molecula
Oncology, B i ish Columbia Cance Agency, Vancou e V5Z
1L3, Canada;
16
Depa men o Radia ion Oncology and
Depa men o Labo a o y Medicine, Radboud Uni e si y
Medical Cen e , 6525 HP Nijmegen, Ne he lands;
17
Depa men
o Medical Oncology, E asmus MC Cance Ins i u e, E asmus
Ju e al.
822 Genome Resea ch
www.genome.o g
Cold Sp ing Ha bo Labo a o y P ess on Oc obe 16, 2016 - Published by genome.cshlp.o gDownloaded om