Na Re Mol Cell Biol T ansc ip ion- eplica ion
con lic s
1
T ansc ip ion-- eplica ion con lic s: how hey occu and
how hey a e esol ed
Ta iana Ga cía-Muse & And és Aguile a
Cen o Andaluz de Biología Molecula y Medicina Regene a i a-CABIMER,
Uni e sidad de Se illa, Se ille, Spain
co esponding au ho : [email protected]
Abs ac
The equen occu ence in cells o ansc ip ion and DNA eplica ion esul s in many
encoun e s and hus con lic s be ween he ansc ip ion and eplica ion machine ies.
These con lic s cons i u e a majo in insic sou ce o genome ins abili y, which is a
hallma k o cance cells. How he eplica ion machine y p og esses h ough a DNA
occupied by an RNA polyme ase is an old ques ion. He e we e iew ecen da a on he
biological ele ance o ansc ip ion- eplica ion con lic s and he ac o s and
mechanisms in ol ed in ei he p e en ing o esol ing hem in euka yo es. Wi h his we
aim o p o ide ou cu en iew o how ansc ip ion can gene a e obs acles o
eplica ion, including o sional s ess and non-B DNA s uc u es, and o he di e en
cellula p ocesses ha ha e e ol ed o sol e hem.
In oduc ion
Genomes a e empla es o mul iple biological p ocesses, including ansc ip ion,
epigene ic modi ica ions, DNA eplica ion, DNA epai and ch omosome seg ega ion. In
a numbe o cases, c oss alk be ween di e en p ocesses occu ing a he DNA may
ha e a posi i e e ec , as in he case o ansc ip ion-coupled epai [G] 1. Howe e , in
Na Re Mol Cell Biol T ansc ip ion- eplica ion
con lic s
2
o he cases, he co- empo al ac i i y o wo cellula machine ies a he same genomic
egion may cause a con lic wi h nega i e consequences. This is he case o DNA
eplica ion and ansc ip ion. Resea ch in he pas wo decades has p o ided e idence
ha ansc ip ion and eplica ion con lic s cons i u e a conside able na u al in insic
sou ce o genome ins abili y, which is a hallma k o cance cells 2. Gi en ha
ansc ip ion and eplica ion a e wo essen ial p ocesses o cell iabili y and
p oli e a ion unc ion and ha hey occu equen ly, a high incidence o encoun e s
be ween he ansc ip ion and eplica ion machine ies is o be expec ed. Al hough
ansc ip ion can ha e a posi i e ole e ec on eplica ion ini ia ion h ough
ansc ip ion-media ed ch oma in changes ha may acili a e i ing o o igins o DNA
eplica ion o igins 3, collisions a e a po en ial h ea o genome in eg i y and cell
iabili y.
How he eplica ion machine y p og esses h ough a double-s anded DNA
occupied by an RNA polyme ase is an old ques ion. Albe s and colleagues elegan ly
add essed his ques ion using he T4 bac e iophage sys em in i o 4; bu ou ac ual
knowledge o he RNA polyme ase s uc u e and mechanisms o ansc ip ion
elonga ion sugges s ha he ac o s and mechanisms used by cells o sol e such
con lic s a e mo e complex han p e iously o eseen. The ele ance o ansc ip ion as
a sou ce o genome ins abili y, as measu ed by he a e o poin mu a ions o o
ecombina ion and ch omosome ea angemen s, and he pu a i e mechanisms by
which such ins abili y is media ed ha e been e iewed ecen ly 5-8. He e we e iew
ecen da a on he ac o s and mechanisms in ol ed in ei he p e en ing o esol ing
ansc ip ion-- eplica ion collisions, and on hei po en ial consequences. In pa icula ,
we discuss how ansc ip ion may hinde he p og ession o he eplica ion o ks i sel
o how ansc ip ion ac i i y gene a es obs acles o eplica ion, including o sional
s ess [G] and non-B DNA [G] s uc u es, and he di e en solu ions he cells ha e
e ol ed o a oid, minimize o esol e hese collisions o hei consequences.
[H1] How do collisions occu ?
A basic di e ence be ween he ansc ip ion and eplica ion machine ies is ha he
elonga ing RNA polyme ase holoenzyme, which comp ises one polyme ase subuni ,
emb aces he double-s anded DNA. The nascen RNA chain is syn hesized in he
ac i e pocke o he RNA polyme ase, whe e i o ms a dynamic 9-11-n RNA-DNA
hyb id (Figu e 1A). The elonga ing DNA polyme ase holoenzyme, on he o he hand,
consis s o wo polyme ase subuni s (o DNA pol III in bac e ia; DNA pol epsilon and
Na Re Mol Cell Biol T ansc ip ion- eplica ion
con lic s
3
del a in euka yo es), each wo king on a single-s anded DNA (ssDNA) empla e (Figu e
1B). Fu he mo e, whe eas se e al ac i e RNA polyme ases can simul aneously
ansc ibe he same gene, eplisomes [G] mo e alone and a e no ollowed by a
second eplica ion o k. No ma e whe he collisions a e co-di ec ional o in head-on
o ien a ion (Figu e 2), he eplica ion o k canno go h ough an elonga ing RNA
polyme ase and so hei encoun e s will cause con lic s 9, 10. Al hough eplica ion o k
p og ession may be a ec ed by collisions in bo h o ien a ions, da a sugges ha he
consequences o collisions a e mo e d ama ic in he head-on o ien a ion 11, 12. When
encoun e s we e p omo ed in yeas a i icial sys ems in a head-on o ien a ion,
eplica ion pause si es we e de ec ed by wo-dimensional (2D)-gel elec opho esis and
hype - ecombina ion was obse ed 12. By con as , co-di ec ional o ien a ion did no
lead o eplica ion pauses o high le els o hype - ecombina ion. This di e ence can be
explained i co-di ec ional encoun e s may in pa be esol ed once he RNA
polyme ase e mina es ansc ip ion.
Cells ha e de eloped di e en s a egies o educe o p e en collisions. In
bac e ia, he e is a genome-wide bias owa ds co-o ien a ion o eplica ion and
ansc ip ion 13 and in e ing ansc ip ional uni s o p o oke head-on collisions causes
eplica ion impai men , p oli e a ion de ec s and genome ins abili y 11. In euka yo es a
bias owa ds co-di ec ional eplica ion and ansc ip ion is no ob ious, bu cells seem
o ha e e ol ed o he s a egies o educe head-on collisions. Fo example, in he
Saccha omyces ce e isiae highly ansc ibed ibosomal DNA ( DNA) genes, eplica ion
o k blocking (RFB) si es exis ha block o k p og ession and p e en ha m ul
encoun e s wi h RNA polyme ases 14. In he mammalian DNA loci eplica ion and he
ansc ip ion seem o be e icien ly spa ially sepa a ed in he nucleoli as a way o a oid
collisions 15. In o he egions o he genome, ansc ip ion and eplica ion seem o be
sepa a ed empo aly 16. Analysis o nascen mRNAs in genes encoding eplica ion
ac o s e ealed ha ac i e genes ansc ibed du ing ea ly eplica ion a e eplica ed
la e in S-phase and ice e sa 17.
Impo an ly, howe e , i is no clea whe he he RNA and DNA polyme ases
e e ac ually make con ac . I is plausible ha be o e he physical connec ion occu s,
ansc ip ion- and eplica ion-media ed changes in ch oma in and DNA s uc u es
a enua e he p og ession o he polyme ases. To ully unde s and con lic s we
he e o e need o iden i y he elemen s and condi ions ha a ec hei occu ence.
[H1] Cis-elemen s a ec ing collisions
Na Re Mol Cell Biol T ansc ip ion- eplica ion
con lic s
4
The ansc ip ion machine y may cons i u e a na u al obs acle o eplica ion o k
p og ession, bu his in e e ence can be di ec o indi ec , since he ansc ip ion
p ocess may also gene a e s uc u al ea u es ha ha e he capaci y o hinde
eplica ion o k p og ession, like changes in DNA supe coiling o seconda y DNA
s uc u es such as hai pins[G], iplex DNA (H-DNA) [G], G-quad uplexes[G] o RNA-
DNA hyb ids.
[H3] DNA supe coiling
T ansc ip ion and eplica ion equi e he unwinding o he DNA molecule. This
unwinding leads o posi i e and nega i e supe coiling ahead and behind he RNA
polyme ase, espec i ely (Figu e 3A). The esul ing o sional s ess [G] is elie ed by
DNA opoisome ases, which a e o Type I o Type II depending on whe he hey
ca alyse b eakage o one o bo h DNA s ands, espec i ely. In budding yeas
opoisome ase mu an s accumula e supe coiling and his o sional s ess p e en s bo h
ansc ip ion and eplica ion o he highly ansc ibed DNA 18, sugges ing ha
supe coiling can cause ansc ip ion and eplica ion block. La e s udies in yeas and
human cells ha e shown ha bo h opoisome ase 1 (human TOP1, yeas Top1) and
opoisome ase 2 (human TOP2, yeas Top2) a e c ucial o p e en ansc ip ion-
eplica ion collisions 19, 20, indica ing ha un esol ed o sional s ess can a enua e he
p og ession o bo h DNA and RNA polyme ases and p omo e ansc ip ion- eplica ion
con lic s. Genome-wide analysis o Top1 and Top2 dis ibu ion in eplica ing budding
yeas cells e ealed an associa ion o hese enzymes wi h mo ing eplica ion o ks 19,
21. Mo eo e , op1 op2 double mu an cells also accumula e DNA damage19.
Consis en ly, analyses o eplica ion by DNA combing [G] ha e e ealed ha
eplica ion o ks a e slowe in yeas and human Top1-de icien cells 20. In e es ingly, in
TOP1-deple ed human cells he e was an inc ease in o k s alling ha co ela ed wi h
he accumula ion o γ-H2AX oci [G] in S-phase and ha was supp essed by inhibi ion
o ansc ip ion elonga ion wi h co dycepin. The e o e, TOP1 ac i i y can p e en
ansc ip ion- eplica ion con lic s and hei ha m ul consequences 20.
Based on hese esul s we sugges ha DNA supe coiling is ansien ly
accumula ing be ween he ad ancing ansc ip ion and eplica ion machine ies and
may be impo an in he con ol o hei collisions and hei ad e se e ec s (Figu e 3A).
In heo y, his phenomenon should be exace ba ed in he case o he head-on
o ien a ion. Howe e , con e gen ansc ip ion, which would c ea e he same
opological cons ain as a head-on collisions, does no pose a majo h ea o genome
in eg i y and ansc ip ion in budding yeas 22. Al e na i ely, ine icien esolu ion o he
nega i e supe coiling accumula ing behind he elonga ing RNA polyme ase may also
Na Re Mol Cell Biol T ansc ip ion- eplica ion
con lic s
5
acili a e local mel ing o he DNA duplex and, consequen ly, o ma ion o non-B DNA
s uc u es ha can block eplica ion o k p og ession (see below). Indeed, in yeas ,
di e gen ansc ip ion was ound o enhance ch omosome ea angemen s 23.
[H3] non-B DNA s uc u es and RNA-DNA hyb ids
Some DNA sequences, especially epe i i e sequences, can assume non-B DNA
s uc u es such as hai pins, iplex DNA (H-DNA) o G-quad uplexes (also e med G-
qua e s). Such non-B DNA s uc u es ha e he capaci y o s all eplica ion o ks and
ha e been co ela ed wi h ho spo s o DNA double-s and b eaks (DSBs) and
ch omosomal dele ions, ansloca ions and o he ea angemen s 24. These seconda y
DNA s uc u es a e belie ed o o m p e e en ially a he ssDNA ha is exposed du ing
DNA eplica ion, bu hey can also be o med du ing ansc ip ion, a ou ed by he
nega i e supe coiling ha is ansien ly accumula ed behind he elonga ing RNA
polyme ase (Figu e 3). A good example o he pu a i e ele ance o non-B s uc u es is
p o ided by G-quad uplexes, which consis o ou epea s o a leas h ee guanines
ha can o m ou s and-in e ac ions (Figu e 3B). G-quad uplexes can o m du ing
lagging-s and eplica ion as shown a elome es 25, as well as du ing ansc ip ion:
human cells ea ed wi h he G-quad uplex ligand py idos a in show a igh co ela ion
be ween py idos a in binding and γH2AX oci [G] o ma ion, which is educed by
ea men wi h he ansc ip ion inhibi o 5,6-dichlo o-1-β-D- ibo u anosylbenzimidazole
(DRB) 26. The idea ha G-quad uplex o ma ion can be po en ia ed behind an
elonga ing RNA Polyme ase II (Pol II) has been indi ec ly in e ed in yeas exp essing
he mu ine G-quad uplex-p one Sμ Ig swi ch egion, which s imula es ecombina ion in
combina ion wi h condi ions o high ansc ip ion le els27. The ac i i y o Top1 in hese
condi ions supp esses G-quad uplex-associa ed ecombina ion, consis en wi h
nega i e supe coiling enhancing G-quad uplex accumula ion 28. I appea s ha he
genomic ins abili y is highe when he o ien a ion o he G- ich s and o he Sµ
sequence, wi h espec o ansc ip ion, lea es he G- ich s and in he non- ansc ibed
s and, sugges ing ha he ssDNA ha allows quad uplex o ma ion o igina es om
ansc ip ion. Addi ional suppo o he idea ha non-B DNA s uc u e may con ibu e o
ansc ip ion-media ed eplica ion o k s alling comes om s udies in mu an s o he
budding yeas DNA helicase Pi 1 (pe i e in eg a ion equency) and he ission yeas
P h1 (Pi 1-homolog), which unwind G-quad uplexes in i o. Absence o Pi 1 o P h1
a enua es o hal s eplica ion in egions o high G-quad uplex densi y and in RNA Pol
II- and Pol III-highly ansc ibed genes 29, 30.

Na Re Mol Cell Biol T ansc ip ion- eplica ion
con lic s
6
Ano he ype o ansc ip ion-media ed s uc u es able o cause o k s alling ha
can s ongly con ibu e o ansc ip ion- eplica ion con lic s a e co- ansc ip ional RNA-
DNA hyb ids (also known as R loops when o med ou side o he ansc ip ion bubble;
Figu e 3C). Al hough hey a e na u al in e media es in class swi ching ecombina ion
and in ini ia ion o mi ochond ial DNA eplica ion, RNA-DNA hyb ids a e o med in
condi ions ha p e en he p ope o ma ion o he ibonucleop o ein pa icle, as shown
in yeas and human cells 31, 32. E idence om yeas o mammalian cells sugges ha
RNA-DNA hyb ids can o m na u ally and may cons i u e an impo an ansc ip ion
in e media e ha can p o oke eplica ion o k s alling a elome es, he DNA egions,
CpG islands [G] and o he si es a speci ic Pol II- ansc ibed genes, including 3’-end
egions. RNA-DNA hyb ids ha e been ho oughly and ex ensi ely e iewed ecen ly 33-
36 and will no be discussed u he he e. I is impo an o no e, howe e , ha an
en ichmen o sequences wi h high p obabili y o o ming non-B DNA s uc u es o
RNA-DNA hyb ids and o unde go ansc ip ion- eplica ion con lic s a e obse ed a
some agile si es (Box 1).
[H1] Mechanisms o p e en ing con lic s
Owing o he impac o ansc ip ion- eplica ion collisions on genomic s abili y and
he eby po en ially on disease, cells ha e e ol ed mechanisms o p e en such
encoun e s. The ac o s ha minimize collisions include he ansc ip ion machine y
i sel , as well as ac o s ha help o acili a e eplica ion p og ession h ough
ansc ibed DNA.
[H3] The RNA polyme ase ansc ip ion appa a us
Some clues o unde s and how he RNA polyme ase di ec ly con ibu e o ansc ip ion-
eplica ion con lic s a e s a ing o eme ge om he analysis o se e al RNA
polyme ase mu an s in bac e ia and yeas , bu we a e s ill a om ha ing a comple e
iew. A c i ical s ep in ansc ip ion ha seems o be ele an o con lic s is RNA
polyme ase ‘back acking’, which e e s o he p ocess by which he RNA polyme ase
e e ses i s p og ession o enzyma ically emo e he las inco po a ed ibonucleo ide.
This allows es a ing ansc ip ion elonga ion ollowing a pause p o oked by
hind ances du ing ansc ip ion elonga ion o as pa o a egula o y p ocess o
coo dina e he di e en s eps o ansc ip ion and RNA p ocessing 37. A back acked
RNA polyme ase is able o block eplica ion p og ession in Esche ichia coli. Using
speci ic p omo e sequences ha allow modula ion o he polyme ase ac i i y i was
shown ha a pe manen ly-a es ed elonga ing polyme ase causes DSBs 38, 39. Such
Na Re Mol Cell Biol T ansc ip ion- eplica ion
con lic s
7
b eaks we e in e ed o be eplica ion-dependen since ea men wi h he eplica ion
inhibi o hyd oxyu ea p io o ansc ip ion ac i a ion a oided hei appea ance. The
clash be ween he eplisome and he back acked RNA polyme ase was in e p e ed o
be esponsible o he o ma ion o DSBs 39. Consis en wi h his iew, he E. coli
ansc ip ion elonga ion ac o s G eA and G eB , which p omo e he elease o
back acked and s alled RNA polyme ases, seem o educe he consequences o
con lic s (Figu e 4A). In he absence o G eA and G eB and unde subs an ial
ansc ip ion ac i i y induced by s a a ion, eplica ion p og ession was comple ely
blocked 40. A simila ole was p oposed o he yeas ansc ip ion elonga ion ac o
TFIIS 41, bu i emains o be seen i his ac i i y has any e ec on pu a i e collisions.
Di ec in ol emen o he ansc ip ion appa a us in modula ing ansc ip ion-
eplica ion con lic s was demons a ed in ecen s udies u ilizing RNA polyme ase
mu an s ha comp omise he s abili y o ansc ip ion complexes. These RNA
polyme ase mu an s we e shown o supp ess g ow h de ec s o E. coli cells lacking
ac o s ha help esol e collisions such as he DNA helicases Rep (Replicase), U D
(u esis an p o ein D) and DinG (damage-inducible p o ein G) 42. These esul s
sugges ha less s able ansc ip ion complexes may no comp omise eplica ion
p og ession since hey do no seem o o m s ong eplica ion obs acles 42. Also,
se e al yeas RNA Pol II mu an s wi h ansc ip ion elonga ion de ec s exhibi ed
eplica ion impai men , in e ed by 2D-gels, b omodeoxyu idine [G] inco po a ion by
DNA polyme ases o by al e ed dis ibu ion o R m3 ( DNA ecombina ion mu a ion
p o ein 3), which is a eplica i e helicase equi ed o eplica ion p og ession h ough
DNA obs acles 43. I is likely ha ollowing a collision he RNA Pol II is eleased om
he DNA o allow passage o he eplisome, as is he case in bac e ia 38. In e es ingly,
one o hese yeas RNA Pol II mu an s, he yeas pb1-1 mu an (o he la ges RNA Pol
II subuni ), has igh e a achmen o ch oma in han wild- ype RNA Pol II, as
de e mined by ch oma in immunop ecipi a ion, suppo ing he idea ha RNA Pol II
mu an s wi h inc eased a achmen o ch oma in could agg a a e he consequences o
a ansc ip ion- eplica ion encoun e 43. These esul s sugges ha he ansc ip ion
machine y, and RNA Pol II i sel , may pa icipa e in managing ansc ip ion- eplica ion
con lic s h ough he easibili y o hei e ic ion om DNA ollowing a collision. The
ecen obse a ion ha PAF1C (RNA polyme ase II-associa ed ac o 1 complex)
igge s RNA Pol II deg ada ion a si es o collisions 44 suppo s his iew.
[H3] Replica ion o k ba ie s
Na Re Mol Cell Biol T ansc ip ion- eplica ion
con lic s
8
Replica ion o ks ha e o deal along hei pa h wi h non-nucleosomal p o ein–DNA
complexes ha assemble a genes and egula o y elemen s. In bac e ia he ba ie
o med by he ansc ip ion complex is able o pause eplica ion o ks, and esump ion
o eplica ion equi es speci ic DNA helicases 45, 46. In yeas , di e en egions ha
impede eplica ion o k p og ession in i o ha e been iden i ied, he mos
ep esen a i e being he o k ba ie ound in he 35S RNA gene in he DNA 47. The
DNA egion p o ides he bes model o s udy he impac o eplica ion s ess
gene a ed by ansc ip ion owing o i s high ansc ip ion a e and high densi y o
eplica ion o igins. The eplica ion ba ie consis s o DNA eplica ion o k-blocking
p o ein Fob1 bound o he speci ic RFB sequence, which p e en s head-on collisions
be ween RNA and DNA polyme ases (Figu e 4B). In e es ingly, eplica ion o k
p og ession h ough he RFB–Fob1 complex equi es he helicase R m3. Dele ion o
he m3 gene ( m3∆) inc eases eplica ion pauses a DNA, esul ing in b eakage and
accumula ion o excised DNA ci cles 14, 48. Al hough R m3 could be seen he e o e as
a complemen a y ac o ac ing in ans o p omo e eplica ion o k passage h ough
p o ein ba ie s, o k pausing in m3∆ mu an s is also inc eased in o he pause si es
such as in RNA genes o elome es 49, 50. No ably, hough, o he pause si es a e ound
a highly ansc ibed genes, bu a e no exace ba ed in m3∆ mu an s, sugges ing ha
o he ac o s may ha e oles in he p e en ion o esolu ion o collisions 9.
[H3] The RecQL5 DNA helicase
Human RecQ-like ATP-dependen DNA helicase Q5 (RECQL5), a membe o he
RecQ amily [G] o DNA helicases, is o da e he p o ein wi h he bes -cha ac e ised
ac i e ole in p e en ing ansc ip ion- eplica ion collisions. RECQL5 o ms a s able
complex wi h RNA Pol II and se e al in i o and in i o s udies indica ed i has a
nega i e egula o y ole in ansc ip ion elonga ion 51. ChIP-seq analysis wi h an RNA
Pol II an ibody e ealed ha ansc ip ion up- egula ion in cells lacking RECQL5
inc eases ansc ip ion pausing, a es and back acking, sugges ing ha uncon olled
and high ansc ip ion a es lead o ansc ip ional s ess 52. RECQL5 associa es wi h
he eplica i e DNA sliding clamp PCNA (p oli e a ing cell nuclea an igen), which
sugges s ha RECQL5 is also in ol ed in eplica ion. Acco dingly, RECQL5-de icien
cells ail o inco po a e b omodeoxyu idine in condi ions o eplica ion s ess and
apidly accumula e DNA damage, e ec s which can be alle ia ed by ully a es ing
eplica ion wi h he eplica ion inhibi o aphidicolin 53. Impo an ly, spon aneous DSBs
accumula e in RECQL5-deple ed cells du ing eplica ion, bu only in associa ion wi h
RNA Pol II ansc ip ion, as he spon aneous b eaks a e loca ed in ansc ibed genes
Na Re Mol Cell Biol T ansc ip ion- eplica ion
con lic s
9
and ansc ip ion inhibi ion elimina ed hei appea ance 54. Fu he mo e, RECQL5 has
been shown o ha e a ole in supp essing genome ea angemen s associa ed
p e e en ially wi h common agile si es (Box 1) and ansc ibed genes 52. These da a
ha e led o he p oposal ha RECQL5 p e en s ansc ip ion- eplica ion collisions.
Recen ly, RECQL5 was also asc ibed a ole in p e en ing he o ma ion o RNA-DNA
hyb ids. Appa en ly, RECQ5 p omo es TOP1 SUMOyla ion by acili a ing he
in e ac ion wi h he PIAS1-SRSF1 E3 ligase complex. This modi ica ion is necessa y
o he binding o TOP1 o RNA Pol II and o he e icien ec ui men o mRNA
p ocessing ac o s o ansc ip ionally ac i e si es, he eby educing he o ma ion o
RNA-DNA hyb ids, as in e ed by he inc eased le els o such hyb ids in cells de ec i e
in RECQ5-dependen TOP1 SUMOyla ion 55. The e o e, RecQL5 may main ain
genome in eg i y by ac i ely limi ing he occu ence o ansc ip ion- eplica ion con lic s
and/o by educing he accumula ion o non-B DNA s uc u es gene a ed du ing
ansc ip ion ha could enhance eplica ion blockage (Figu e 3).
[H3] Ch oma in emodelling
In euka yo es, ansc ip ion and eplica ion occu in he con ex o highly s uc u ed
ch oma in. Following eplica ion he ch oma in s a e is main ained by coupling he
deposi ion o ecycled pa en al his ones wi h newly-syn hesized his ones on he
duplica ed DNA, which is ca ied ou by his one chape ones and ch oma in emodelling
complexes. E en hough no much wo k has been done on he ole o ch oma in
emodelling in diminishing ansc ip ion- eplica ion con lic s, e idence exis s abou he
impo ance o such ac o s on collisions. This has been clea ly shown o he his one
chape one FACT ( acili a es ch oma in ansc ip ion) complex, which was ini ially ound
o be equi ed o ch oma in emodelling du ing ansc ip ion bu is also in ol ed in
DNA eplica ion 56-58. Yeas and human cells lacking FACT complex ac i i y ha e high
le els o ansc ip ion– eplica ion collisions, exhibi ing o k p og ession impai men ha
co ela es wi h inc eased genomic ins abili y. Howe e , when ansc ip ion was
inhibi ed wi h co dycepin in FACT-deple ed human cells, he a e o o k p og ession
was es o ed o no mal 59. The e o e, ch oma in- eo ganizing ac o s, such as FACT,
can p e en collisions by p omo ing he eplica ion o k p og ession h ough ansc ibed
egions (Figu e 4C). I would ce ainly be in e es ing o see whe he his iew would
also apply o he INO80 emodelling complex since i was shown in E. colibudding
yeas ha Ino80 (inosi ol equi ing mu an 80) coope a es wi h he ansc ip ion ac o
PAF1C o igge RNA Pol II deg ada ion a si es o collisions 44.
Na Re Mol Cell Biol T ansc ip ion- eplica ion
con lic s
16
This wo ks shows ha co-di ec ional con lic s a highly ansc ibed RNA
genes can s all eplica ion in i o in B. sub ilis.
11. S i a san, A., Teh anchi, A., MacAlpine, D.M. & Wang, J.D. Co-o ien a ion o
eplica ion and ansc ip ion p ese es genome in eg i y. PLoS Gene . 6,
e1000810 (2010).
12. P ado, F. & Aguile a, A. Impai men o eplica ion o k p og ession media es
RNA pol II ansc ip ion-associa ed ecombina ion. EMBO J. 24, 1267-1276
(2005).
This s udy demons a es ha head-on ansc ip ion and eplica ion
encoun e s cause eplica ion o k impai men ha leads o genome
ins abili y.
13. Me ikh, H., Zhang, Y., G ossman, A.D. & Wang, J.D. Replica ion- ansc ip ion
con lic s in bac e ia. Na u e Re . Mic obiol. 10, 449-458 (2012).
14. I essa, A.S., Zhou, J.Q. & Zakian, V.A. The Saccha omyces Pi 1p DNA
helicase and he highly ela ed R m3p ha e opposi e e ec s on eplica ion o k
p og ession in ibosomal DNA. Cell 100, 479-489 (2000).
15. Smi no , E. e al. Sepa a ion o eplica ion and ansc ip ion domains in
nucleoli. J. S uc . Biol. 188, 259-266 (2014).
16. Gilbe , D.M. Replica ion iming and ansc ip ional con ol: beyond cause and
e ec . Cu . Opin. Cell Biol. 14, 377-383 (2002).
17. Me ye -Figuie e, M. e al. Tempo al sepa a ion o eplica ion and ansc ip ion
du ing S-phase p og ession. Cell Cycle 13, 3241-3248 (2014).
18. B ill, S.J., DiNa do, S., Voelkel-Meiman, K. & S e nglanz, R. Need o DNA
opoisome ase ac i i y as a swi el o DNA eplica ion o ansc ip ion o
ibosomal RNA. Na u e 326, 414-416 (1987).
19. Be mejo, R. e al. Top1- and Top2-media ed opological ansi ions a
eplica ion o ks ensu e o k p og ession and s abili y and p e en DNA damage
checkpoin ac i a ion. Genes De . 21, 1921-1936 (2007).
20. Tudu i, S. e al. Topoisome ase I supp esses genomic ins abili y by p e en ing
in e e ence be ween eplica ion and ansc ip ion. Na u e Cell Biol. 11, 1315-
1324 (2009).
This genome-wide analysis o TOP1 de icien cells shows co ela ion
be ween o k s alling and DNA b eaks a ansc ibed genes, sugges ing
ha TOP1 educes R-loops-dependen ansc ip ion– eplica ion con lic s.
21. Be mejo, R. e al. Genome-o ganizing ac o s Top2 and Hmo1 p e en
ch omosome agili y a si es o S phase ansc ip ion. Cell 138, 870-884 (2009).
22. Ga cía-Rubio, M.L. & Aguile a, A. Topological cons ain s impai RNA
polyme ase II ansc ip ion and causes ins abili y o plasmid-bo ne con e gen
genes. Nucleic Acids Res. 40, 1050-1064 (2012).
23. Pannunzio, N.R. & Liebe , M.R. Dissec ing he oles o di e gen and
con e gen ansc ip ion in ch omosome ins abili y. Cell Rep. 14, 1025-1031
(2016).
24. Zhao, J., Bacolla, A., Wang, G. & Vasquez, K.M. Non-B DNA s uc u e-induced
gene ic ins abili y and e olu ion. Cell Mol. Li e Sci. 67, 43-62 (2010).
25. Aud y, J. e al. RPA p e en s G- ich s uc u e o ma ion a lagging-s and
elome es o allow main enance o ch omosome ends. EMBO J. 34, 1942-1958
(2015).
26. Rod iguez, R. e al. Small-molecule-induced DNA damage iden i ies al e na i e
DNA s uc u es in human genes. Na u e Chem. Biol. 8, 301-310 (2012).

Na Re Mol Cell Biol T ansc ip ion- eplica ion
con lic s
17
27. Kim, N. & Jinks-Robe son, S. Guanine epea -con aining sequences con e
ansc ip ion-dependen ins abili y in an o ien a ion-speci ic manne in yeas .
DNA Repai (Ams ) 10, 953-960 (2011).
28. Yada , P. e al. Topoisome ase I plays a c i ical ole in supp essing genome
ins abili y a a highly ansc ibed G-quad uplex- o ming sequence. PLoS Gene .
10, e1004839 (2014).
29. Sabou i, N., McDonald, K.R., Webb, C.J., C is ea, I.M. & Zakian, V.A. DNA
eplica ion h ough ha d- o- eplica e si es, including bo h highly ansc ibed RNA
Pol II and Pol III genes, equi es he S. pombe P h1 helicase. Genes De . 26,
581-593 (2012).
30. Paeschke, K., Cap a, J.A. & Zakian, V.A. DNA eplica ion h ough G-quad uplex
mo i s is p omo ed by he Saccha omyces ce e isiae Pi 1 DNA helicase. Cell
145, 678-691 (2011).
31. Hue as, P. & Aguile a, A. Co ansc ip ionally o med DNA:RNA hyb ids
media e ansc ip ion elonga ion impai men and ansc ip ion-associa ed
ecombina ion. Mol. Cell 12, 711-721 (2003).
32. Li, X. & Manley, J.L. Inac i a ion o he SR p o ein splicing ac o ASF/SF2
esul s in genomic ins abili y. Cell 122, 365-378 (2005).
33. San os-Pe ei a, J.M. & Aguile a, A. R loops: new modula o s o genome
dynamics and unc ion. Na u e Re . Gene . 16, 583-597 (2015).
34. Skou i-S a haki, K. & P oud oo , N.J. A double-edged swo d: R loops as h ea s
o genome in eg i y and powe ul egula o s o gene exp ession. Genes De .
28, 1384-1396 (2014).
35. Sollie , J. & Cimp ich, K.A. B eaking bad: R-loops and genome in eg i y. T ends
Cell Biol. 25, 514-522 (2015).
36. Cos an ino, L. & Koshland, D. The Yin and Yang o R-loop biology. Cu . Opin.
Cell Biol. 34, 39-45 (2015).
37. Nudle , E. RNA polyme ase back acking in gene egula ion and genome
ins abili y. Cell 149, 1438-1445 (2012).
38. Pome an z, R.T. & O'Donnell, M. Di ec es a o a eplica ion o k s alled by a
head-on RNA polyme ase. Science 327, 590-592 (2010).
39. Du a, D., Sha alin, K., Epsh ein, V., Go esman, M.E. & Nudle , E. Linking RNA
polyme ase back acking o genome ins abili y in E. coli. Cell 146, 533-543
(2011).
40. Teh anchi, A.K. e al. The ansc ip ion ac o DksA p e en s con lic s be ween
DNA eplica ion and ansc ip ion machine y. Cell 141, 595-605 (2010).
This wo k illus a es he unc ion o RNA polyme ases, ansc ip ion
ac o s and DNA epai p o eins in p e en ing and/o esol ing
ansc ip ion– eplica ion con lic s.
41. Du a, A. e al. Cc 4-No and TFIIS unc ion coope a i ely o escue a es ed
RNA polyme ase II. Mol. Cell Biol. 35, 1915-1925 (2015).
42. Baha oglu, Z., Les ini, R., Duigou, S. & Michel, B. RNA polyme ase mu a ions
ha acili a e eplica ion p og ession in he ep u D ecF mu an lacking wo
accesso y eplica i e helicases. Mol. Mic obiol. 77, 324-336 (2010).
43. Felipe-Ab io, I., La uen e-Ba que o, J., Ga cía-Rubio, M.L. & Aguile a, A. RNA
polyme ase II con ibu es o p e en ing ansc ip ion-media ed eplica ion o k
s alls. EMBO J. 34, 236-250 (2015).
44. Poli, J. e al. Mec1, INO80, and he PAF1 complex coope a e o limi
ansc ip ion eplica ion con lic s h ough RNAPII emo al du ing eplica ion
s ess. Genes De . 30, 337-354 (2016).
45. Me ikh, C.N., B ewe , B.J. & Me ikh, H. The B. sub ilis accesso y helicase
Pc A acili a es DNA eplica ion h ough ansc ip ion uni s. PLoS Gene . 11,
e1005289 (2015).
Na Re Mol Cell Biol T ansc ip ion- eplica ion
con lic s
18
46. Gup a, M.K. e al. P o ein-DNA complexes a e he p ima y sou ces o
eplica ion o k pausing in Esche ichia coli. P oc. Na l. Acad. Sci. 110, 7252-
7257 (2013).
47. B ewe , B.J. & Fangman, W.L. A eplica ion o k ba ie a he 3' end o yeas
ibosomal RNA genes. Cell 55, 637-43 (1988).
48. To es, J.Z., Bessle , J.B. & Zakian, V.A. Local ch oma in s uc u e a he
ibosomal DNA causes eplica ion o k pausing and genome ins abili y in he
absence o he S. ce e isiae DNA helicase R m3p. Genes De . 18, 498-503
(2004).
49. I essa, A.S. e al. The Saccha omyces ce e isiae helicase R m3p acili a es
eplica ion pas nonhis one p o ein-DNA complexes. Mol. Cell 12, 1525-1536
(2003).
50. I essa, A.S., Zhou, J.-Q., Schulz, V.P., Monson, E.K. & Zakian, V.A.
Saccha omyces R m3p, a 5' o 3' DNA helicase ha p omo es eplica ion o k
p og ession h ough elome ic and sub elome ic DNA. Genes De . 16, 1383-
1396 (2002).
51. Popu i, V., Tadoko o, T., C o eau, D.L. & Boh , V.A. Human RECQL5: gua ding
he c oss oads o DNA eplica ion and ansc ip ion and p o iding backup
capabili y. C i . Re . Biochem. Mol. Biol. 48, 289-299 (2013).
52. Sapona o, M. e al. RECQL5 con ols ansc ip elonga ion and supp esses
genome ins abili y associa ed wi h ansc ip ion s ess. Cell 157, 1037-1049
(2014).
This s udy shows ha RECQL5 a enua es ansc ip ion elonga ion and
supp esses genome ea angemen s a common agile si es, indica ing
ha RECQL5 p e en s ansc ip ion- eplica ion con lic s.
53. Hu, Y., Lu, X., Zhou, G., Ba nes, E.L. & Luo, G. Recql5 plays an impo an ole
in DNA eplica ion and cell su i al a e camp o hecin ea men . Mol. Biol. Cell
20, 114-123 (2009).
54. Li, M., Xu, X. & Liu, Y. The SET2-RPB1 in e ac ion domain o human RECQ5 is
impo an o ansc ip ion-associa ed genome s abili y. Mol. Cell. Biol. 31,
2090-2099 (2011).
55. Li, M., Pokha el, S., Wang, J.-T., Xu, X. & Liu, Y. RECQ5-dependen
SUMOyla ion o DNA opoisome ase I p e en s ansc ip ion-associa ed
genome ins abili y. Na u e Commun. 6, 6720 (2015).
56. O phanides, G., Wu, W.H., Lane, W.S., Hampsey, M. & Reinbe g, D. The
ch oma in-speci ic ansc ip ion elonga ion ac o FACT comp ises human
SPT16 and SSRP1 p o eins. Na u e 400, 284-288 (1999).
57. Fol man, M. e al. Euka yo ic eplisome componen s coope a e o p ocess
his ones du ing ch omosome eplica ion. Cell Rep. 3, 892-904 (2013).
58. Abe, T. e al. The his one chape one acili a es ch oma in ansc ip ion (FACT)
p o ein main ains no mal eplica ion o k a es. J. Biol. Chem. 286, 30504-
30512 (2011).
59. He e a-Moyano, E., Me gui, X., Ga cia-Rubio, M.L., Ba oso, S. & Aguile a, A.
The yeas and human FACT ch oma in- eo ganizing complexes sol e R-loop-
media ed ansc ip ion- eplica ion con lic s. Genes De . 28, 735-748 (2014).
This wo k p o ides compelling e idence o he ole o ch oma in-
eo ganizing complex FACT in he esolu ion o ansc ip ion- eplica ion
con lic s.
60. Cas ellano-Pozo, M. e al. R loops a e linked o his one H3 S10 phospho yla ion
and ch oma in condensa ion. Mol. Cell 52, 583-590 (2013).
Na Re Mol Cell Biol T ansc ip ion- eplica ion
con lic s
19
61. G oh, M., Lu ino, M.M., Wade-Ma ins, R. & G omak, N. R-loops associa ed
wi h iple epea expansions p omo e gene silencing in F ied eich a axia and
agile X synd ome. PLoS Gene . 10, e1004318 (2014).
62. Za a iegui, M. e al. RNAi p omo es he e och oma ic silencing h ough
eplica ion-coupled elease o RNA Pol II. Na u e 479, 135-138 (2011).
The au ho s show ha a pe icen ome ic egions he co- ansc ip ional
RNA in e e ence machine y eleases he RNA polyme ase o a oid
con lic s wi h he eplica ion machine y.
63. Cas el, S.E. e al. Dice p omo es ansc ip ion e mina ion a si es o eplica ion
s ess o main ain genome s abili y. Cell 159, 572-583 (2014).
64. Lopes, M., Foiani, M. & Sogo, J.M. Mul iple mechanisms con ol ch omosome
in eg i y a e eplica ion o k uncoupling and es a a i epa able UV lesions.
Mol. Cell 21, 15-27 (2006).
65. Helle , R.C. & Ma ians, K.J. Replica ion o k eac i a ion downs eam o a
blocked nascen leading s and. Na u e 439, 557-562 (2006).
66. Deshpande, A.M. & Newlon, C.S. DNA eplica ion o k pause si es dependen
on ansc ip ion. Science 272, 1030-1033 (1996).
67. Nguyen, V.C. e al. Replica ion s ess checkpoin signaling con ols RNA gene
ansc ip ion. Na . S uc . Mol. Biol. 17, 976-981 (2010).
68. Molla-He man, A., Vallés, A.M., Ganem-Elbaz, C., An oniewski, C. & Huynh, J.-
R. RNA p ocessing de ec s induce eplica ion s ess and Chk2-dependen
dis up ion o piRNA ansc ip ion. EMBO J. (2015).
69. Im, J.S. e al. ATR checkpoin kinase and CRL1be aTRCP collabo a e o
deg ade ASF1a and hus ep ess genes o e lapping wi h clus e s o s alled
eplica ion o ks. Genes De . 28, 875-887 (2014).
70. Yeo, C.Q. e al. p53 Main ains Genomic s abili y by p e en ing in e e ence
be ween ansc ip ion and eplica ion. Cell Rep. 15, 132-146 (2016).
71. Cabal, G.G. e al. SAGA in e ac ing ac o s con ine sub-di usion o ansc ibed
genes o he nuclea en elope. Na u e 441, 770-773 (2006).
72. Blobel, G. Gene ga ing: a hypo hesis. P oc. Na l. Acad. Sci. USA 82, 8527-8529
(1985).
73. Be mejo, R. e al. The eplica ion checkpoin p o ec s o k s abili y by eleasing
ansc ibed genes om nuclea po es. Cell 146, 233-246 (2011).
74. Rossi, S.E., Ajazi, A., Ca o enu o, W., Foiani, M. & Gianna asio, M. Rad53-
media ed egula ion o R m3 and Pi 1 DNA helicases con ibu es o p e en ion
o abe an o k ansi ions unde eplica ion s ess. Cell Rep. 13, 80-92 (2015).
75. Bha ia, V. e al. BRCA2 p e en s R-loop accumula ion and associa es wi h
TREX-2 mRNA expo ac o PCID2. Na u e 511, 362-365 (2014).
76. Hill, S.J. e al. Sys ema ic sc eening e eals a ole o BRCA1 in he esponse
o ansc ip ion-associa ed DNA damage. Genes De . 28, 1957-1975 (2014).
77. Schlache , K. e al. Double-s and b eak epai -independen ole o BRCA2 in
blocking s alled eplica ion o k deg ada ion by MRE11. Cell 145, 529-542
(2011).
78. Schwab, R.A. e al. The Fanconi anemia pa hway main ains genome s abili y by
coo dina ing eplica ion and ansc ip ion. Mol. Cell 60, 351-361 (2015).
79. Ga cía-Rubio, M.L. e al. The Fanconi anemia pa hway p o ec s genome
in eg i y om R-loops. PLoS Gene . 11, e1005674 (2015).
Re e ences 78 and 79 p o ide e idence o a no el ole o Fanconi anemia
ac o s in p e en ing ansc ip ion- eplica ion con lic s media ed by R-
loops.
Na Re Mol Cell Biol T ansc ip ion- eplica ion
con lic s
20
80. Ha chi, E. e al. BRCA1 ec ui men o ansc ip ional pause si es is equi ed o
R-loop-d i en DNA damage epai . Mol. Cell 57, 636-647 (2015).
81. Yüce, Ö. & Wes , S.C. Sena axin, de ec i e in he neu odegene a i e diso de
a axia wi h oculomo o ap axia 2, lies a he in e ace o ansc ip ion and he
DNA damage esponse. Mol. Cell. Biol. 33, 406-417 (2013).
82. Alzu, A. e al. Sena axin associa es wi h eplica ion o ks o p o ec o k in eg i y
ac oss RNA-polyme ase-II- ansc ibed genes. Cell 151, 835-846 (2012).
83. Maldonado, E. e al. A human RNA polyme ase II complex associa ed wi h SRB
and DNA- epai p o eins. Na u e 381, 86-89 (1996).
84. Duch, A. e al. Coo dina ed con ol o eplica ion and ansc ip ion by a SAPK
p o ec s genomic in eg i y. Na u e 493, 116-119 (2013).
85. Alcasabas, A.A. e al. M c1 ansduces signals o DNA eplica ion s ess o
ac i a e Rad53. Na u e Cell Biol. 3, 958-965 (2001).
86. Glo e , T.W., Be ge , C., Coyle, J. & Echo, B. DNA polyme ase alpha inhibi ion
by aphidicolin induces gaps and b eaks a common agile si es in human
ch omosomes. Hum. Gene . 67, 136-142 (1984).
87. Zhang, H. & F euden eich, C.H. An AT- ich sequence in human common agile
si e FRA16D causes o k s alling and ch omosome b eakage in S. ce e isiae.
Mol. Cell 27, 367-379 (2007).
88. Ge ha d , J. e al. The DNA eplica ion p og am is al e ed a he FMR1 locus in
agile X emb yonic s em cells. Mol. Cell 53, 19-31 (2014).
89. Le essie , A. e al. Cell- ype-speci ic eplica ion ini ia ion p og ams se agili y o
he FRA3B agile si e. Na u e 470, 120-123 (2011).
90. Oze i-Galai, E. e al. Failu e o o igin ac i a ion in esponse o o k s alling leads
o ch omosomal ins abili y a agile si es. Mol. Cell 43, 122-131 (2011).
91. Le Tallec, B. e al. Common agile si e p o iling in epi helial and e y h oid cells
e eals ha mos ecu en cance dele ions lie in agile si es hos ing la ge
genes. Cell Rep. 4, 420-428 (2013).
92. Wilson, T.E. e al. La ge ansc ip ion uni s uni y copy numbe a ian s and
common agile si es a ising unde eplica ion s ess. Genome Res. 25, 189-200
(2015).
93. Helm ich, A., Balla ino, M. & To a, L. Collisions be ween eplica ion and
ansc ip ion complexes cause common agile si e ins abili y a he longes
human genes. Mol. Cell 44, 966-977 (2011).
This wo k documen s ansc ip ion- eplica ion con lic s in e y long
human genes and shows ha he ins abili y o common agile si es
loca ed wi hin hose genes is dependen on ansc ip ion and R-loop
o ma ion.
94. Ba low, J.H. e al. Iden i ica ion o ea ly eplica ing agile si es ha con ibu e o
genome ins abili y. Cell 152, 620-632 (2013).
95. Ho man, E.A., McCulley, A., Haa e , B., A nak, R. & Feng, W. B eak-seq
e eals hyd oxyu ea-induced ch omosome agili y as a esul o unscheduled
con lic be ween DNA eplica ion and ansc ip ion. Genome Res. 25, 402-412
(2015).
96. Halazone is, T.D., Go goulis, V.G. & Ba ek, J. An oncogene-induced DNA
damage model o cance de elopmen . Science 319, 1352-1355 (2008).
97. Mi on, K., Golan-Le , T., D i , R., Ben-Da id, E. & Ke em, B. Oncogenes c ea e
a unique landscape o agile si es. Na u e Commun. 6, 7094 (2015).
98. Ba ko a, J. e al. DNA damage esponse as a candida e an i-cance ba ie in
ea ly human umo igenesis. Na u e 434, 864-870 (2005).
99. Valo ka, T. e al. T ansc ip ional con ol o DNA eplica ion licensing by Myc.
Sci. Rep. 3, 3444 (2013).
Na Re Mol Cell Biol T ansc ip ion- eplica ion
con lic s
21
100. S ini asan, S.V., Dominguez-Sola, D., Wang, L.C., Hy ien, O. & Gau ie , J.
Cdc45 is a c i ical e ec o o myc-dependen DNA eplica ion s ess. Cell Rep.
3, 1629-1639 (2013).
101. Jones, R.M. e al. Inc eased eplica ion ini ia ion and con lic s wi h ansc ip ion
unde lie Cyclin E-induced eplica ion s ess. Oncogene 32, 3744-53 (2013).
Acknowledgemen s
The au ho s would like o hank B. Gómez-González o he commen s on he
manusc ip and D. Haun o s yle supe ision. Resea ch in A.A.'s labo a o y is unded
by g an s om he Spanish Minis y o Economy and Compe i i eness, he Jun a de
Andaluc a, he u opean nion , o ldwide ance esea ch, and he
Eu opean Resea ch Council. The au ho s apologize o hose whose wo k could no be
ci ed owing o space limi a ions.
Compe ing in e es s s a emen
The au ho s decla e no compe ing in e es s.
BOX 1. F agile si es as ho spo s o ansc ip ion- eplica ion collisions
F agile si es a e genomic egions exhibi ing cons ic ions o gaps in me aphase
ch omosomes ollowing eplica ion s ess. They a e ca ego ized in o wo classes: a e
agile si es a e ound in <5% o indi iduals and a ise om inucleo ide epea
expansion, and common agile si es (CFSs), which a e ound in all indi iduals and a e
no associa ed wi h epea expansion 86. F agile si es a e equen ly en iched in
sequences ha can s all DNA eplica ion, such as AT-dinucleo ide- ich sequences o
high DNA lexibili y in CFSs, as shown in yeas a FRA16D 87, o in o he eplica ion-
a enua ing sequences, as shown in human cells exp essing he a e agile si e
FRAXA a he FMR1 locus 87, 88. Sca ci y in o igins o DNA eplica ion o ine icien
eplica ion ac i a ion in la ge genomic egions may explain he agili y o some CFSs
89, 90. In e es ingly, mapping o CFSs in se e al human cell lines e ealed hey a e
loca ed mos ly wi hin la ge genes. Also, he e is a high co ela ion be ween CFSs and
ecu en ch omosomal ea angemen s obse ed in cance cells, and a simila
co ela ion eme ged om he analysis o copy numbe a ian s, whose ho spo s
ma ched CFSs when loca ed in la ge egions o ac i e ansc ip ion in bo h human and
mouse cells 91, 92. These da a sugges ha concomi an ansc ip ion and eplica ion
may lead o agili y. Rela ed o his is he obse a ion ha RNA-DNA hyb ids

Na Re Mol Cell Biol T ansc ip ion- eplica ion
con lic s
22
accumula e in he long FHIT, WWOX o IMMP2L genes, which ha bou he CFSs
FRA3B, FRA16D and FRA7K, espec i ely 93, as well as in he a e agile si es FXN
and FRAXA 61, among o he s 2. Recen ly, ch oma in immunop ecipi a ion analysis wi h
he single-s and DNA binding p o ein eplica ion p o ein A (RPA) ollowing eplica ion
s ess has allowed he iden i ica ion and mapping o a new class o agile si es, e med
ea ly eplica ion agile si es (ERFSs), which con a y o CFSs a e loca ed nea
eplica ion o igins and wi hin ac i ely ansc ibed genes, s eng hening he possibili y
ha hei agili y esul om ansc ip ion- eplica ion con lic s 94. Simila ly, using he
B eak-seq [G] echnique in cells ollowing exposu e o he eplica ion-s ess agen
hyd oxyu ea, eplica ion-induced double s and b eaks we e mapped p e e en ially a
genes whose exp ession is[Au: OK?] induced unde eplica ion s ess condi ions 95.
The e o e, genomes con ain ho spo s o ansc ip ion- eplica ion collisions, which can
mani es as di e en o ms o agili y.
BOX 2. Oncogenes and ansc ip ion- eplica ion collisions.
An oncogene e e s o a gene ha when mu a ed con ibu es o he de elopmen o
p og ession o cance , whe eas he e m p o o-oncogene is ese ed o i s wild- ype
allele. Oncogenes gene ally egula e cell di ision, cell di e en ia ion and/o cell dea h.
This is he case o he oncogene c-Myc, which egula es ansc ip ion o se e al genes
ha con ol cell g ow h and cell cycle p og ession2. As eplica ion s ess and genomic
ins abili y a e hallma ks o cance cells 96, i seems plausible ha oncogenes may
inc ease he a e o ansc ip ion- eplica ion con lic s, which will se e as a sou ce o
genomic ins abili y. Al e ed exp ession o cyclin E o oncogenic Ras induces
ch omosomal agili y a si es ha co-localize wi h la ge genes and only pa ially
o e lap wi h he canonical, eplica ion s ess-induced agile si es 97. Oncogene
exp ession can nega i ely a ec eplica ion by p omo ing eplica ion o igin ac i a ion,
as shown o c-Myc o cyclin E 98, 99. Using DNA combing and cell- ee ex ac s de i ed
om Xenopus lae is eggs i has been shown ha c-Myc inc eases ac i a ion o ea ly-
eplica ing o igins, esul ing in ele a ed o k collapse and subsequen DNA damage
accumula ion 100. Al hough such eplica ion o k collapses may occu independen ly o
ansc ip ion, he excess o ac i e eplica ion o ks may inc ease he p obabili y o
collisions. Consis en wi h his iew, DNA damage esul ing om eplica ion impai men
by cyclin E o e exp ession was pa ially supp essed by he ansc ip ion inhibi o
co dycepin, sugges ing ha collisions can indeed con ibu e o oncogene-induced
eplica ion s ess101. I would be impo an o de e mine he gene al ele ance o his
Na Re Mol Cell Biol T ansc ip ion- eplica ion
con lic s
23
phenomenon in cance cells and o explo e he possibili y o using ansc ip ion-
eplica ion collisions as a selec i e a ge in cance he apy.
Figu e legends
Figu e 1. T ansc ip ion and eplica ion. A. A small po ion o he double DNA helix is
unwound by he RNA polyme ase (RNAP) o enable ansc ip ion (known as
" ansc ip ion bubble"). DNA unwinding by he RNAP gene a es posi i e and nega i e
supe coiling[G], which is alle ia ed by opoisome ases. In euka yo es, ansc ip ion
also in ol es ch oma in modi ica ion and emodelling. The p og ession o RNA
polyme ase equi es he ac i i y o ansc ip ion elonga ion ac o s. The nascen RNA is
co- ansc ip ionally p ocessed by di e en ac o s. B. A he eplica ion o k he DNA
helicase minich omosome main enance complex (MCM) opens he double helix and
he NA polyme ases Polε and Polδ ex end he leading and lagging s and,
espec i ely. Syn hesis o each new NA molecule is ini ia ed by he Polα-P imase
complex Polα-P i). Lagging s and syn hesis leads o he o ma ion o ssDNA, which is
coa ed wi h eplica ion p o ein A (RPA). Fo k p og ession equi es he ac i i y o
se e al eplica ion co ac o s, including he clamp p oli e a ing cell nuclea an igen
(PCNA). DNA unwinding by he eplica ion o k gene a es posi i e supe coiling, which
is alle ia ed by opoisome ases. Replica ion also en ails eassembly o ecycled and de
no o-syn hesized nucleosomes a he newly syn hesized DNA. Dashed aA ows
indica e he di ec ion o o k p og ession andRNA and DNA polyme ases syn hesis.
Figu e 2. Head-on and co-di ec ional ansc ip ion- eplica ion collisions. A.
P og ession in opposi e di ec ions o an RNA Polyme ase (RNAP) and a eplica ion
o k leads o head-on collisions, which induce pausing and blockage o he eplica ion
o k and may lead o i s collapse and he o ma ion o DNA b eaks . B. P og ession o
an RNA Polyme ase and a eplica ion o k in he same di ec ion leads o co-di ec ional
collisions i he o k mo es as e han he RNA Polyme ase. Co-di ec ional collisions
can be esol ed by displacemen o he RNA Polyme ase om he DNA. MCM,
minich omosome main enance complex; Polε, DNA polyme ase ε; Polδ, DNA
polyme ase δ.
Figu e 3. Condi ions ha a ec he occu ence o ansc ip ion- eplica ion
collisions. A. Con e gence o an RNA Polyme ase (RNAP) and a eplica ion o k
when o ien ed head-on can lead o he accumula ion o posi i e DNA supe coiling
Na Re Mol Cell Biol T ansc ip ion- eplica ion
con lic s
24
be ween hem, which induces pausing o he o k. B. The pa ial unwinding o DNA by
he nega i e supe coiling gene a ed behind he RNA polyme ase can enable he
o ma ion o non-B DNA s uc u es, such as G-quad uplexes, which may cons i u e an
obs acle o eplica ion o k p og ession. C. O he non-B DNA s uc u es include RNA-
DNA hyb ids, which also may cons i u e an obs acle o o k p og ession. (B and C)
Once a s able non-B DNA s uc u e capable o blocking o k p og ession is co-
ansc ip ionally o med, he di ec ion o ansc ip ion o he p esence o he RNA
Polyme ase i sel would be in p inciple i ele an o he o ma ion o he ansc ip ion-
eplica ion collision. MCM, minich omosome main enance complex; Polε, DNA
polyme ase ε; Polδ, DNA polyme ase δ.
Figu e 4. Mechanisms p e en ing ansc ip ion- eplica ion collisions. A. Pausing
o RNA polyme ase (RNAP) is no mally esol ed by back acking, which disengages
he 3'-end o he RNA molecule om he ac i e si e and leads o back and o h sliding
o he RNA polyme ase. The G eA and G eB RNA clea age ac o s s imula e he
emo al o he ex uded RNA and he eac i a ion o ansc ip ion. Wi hou G eA and
G eB he RNA polyme ase migh s all and become an obs acle o he eplica ion o k,
leading o ansc ip ion- eplica ion collisions. B. The eplica ion o k ba ie (RFB) si e is
a DNA sequence loca ed nea he 3'-end o RNA genes and p e en s ansc ip ion–
eplica ion con lic s in he budding yeas . DNA eplica ion o k-blocking p o ein Fob1 is
equi ed o RFB ac i i y as wi hou i he e is no eplica ion o k a es a he RFB,
esul ing in ansc ip ion- eplica ion collisions. C. Ch oma in emodelling by he FACT
complex acili a es ansc ip ion as well as eplica ion o k p og ession. Wi hou FACT
al e ed ch oma in eo ganiza ion esul s in ansc ip ion- eplica ion collisions. D. A
pe icen ome ic egions co- ansc ip ional RNAi eleases he RNA polyme ase he eby
allowing comple ion o DNA eplica ion. Wi hou he RNAi machine y, ailu e o elease
he RNA polyme ase du ing S-phase esul s in ansc ip ion– eplica ion collisions.
DnaB, DNA eplicase B; Polα, DNA polyme ase III α; MCM, minich omosome
main enance complex; Polε, NA polyme ase ε; Polδ, DNA polyme ase δ.
Figu e 5. Resol ing ansc ip ion– eplica ion collisions o a oid genome
ins abili yby he DNA damage esponse[Au: OK?]. A s alled eplica ion o k can
ac i a e he ATR-dependen checkpoin o sol e collisions and a oid hei
consequences. Speci ic ac o s a e ec ui ed o esol e o p e en he obs acle,
including DNA helicases (R m3), mRNA p ocessing p o eins o ch oma in emodelling
complexes (FACT), DNA helicases (R m3) o mRNA p ocessing p o eins [Au: THO is
no men ioned in he main ex no explained he e, so we emo ed i , OK?] . In
Na Re Mol Cell Biol T ansc ip ion- eplica ion
con lic s
25
addi ion, he RNA Polpolyme ase (RNAP) could be eleased a ansc ip ion
e mina ion si es, wi h he help o BRCA1 and helicase SETX, o he INO80 his one
emodelling complex and ansc ip ion ac o PAF1C complex, as a way o a oid he
collisions. Simila ly, a RNAs ansc ip ion si es de RNA Pol III is di ec ly e ic ed du ing
S-phase. S abiliza ion and esump ion o s alled o ks a ansc ibed DNA egions can
occu ia DNA epai ac o s, such as hose o he Fanconi anemia epai pa hway
including umou supp esso s BRCA1 and BRCA2. Finally di e en DNA epai
pa hways can ac a collision si es, i hese degene a e in o DNA lesions. Al hough in
some examples di ec in ol emen o he ATR-dependen checkpoin has been
epo ed, in o he cases is ye unknown. MCM, minich omosome main enance
complex; Polε, NA polyme ase ε; Polδ, DNA polyme ase δ.
Glossa y:
B eak-seq. Technique o map ch omosome b eaks based on DSB labelling and nex
gene a ion sequencing.
B omodeoxyu idine. Syn he ic analogue o he hymidine nucleoside used o ollow
DNA syn hesis.
CpG islands. Ch omosomal egions wi h high densi y o non-me hyla ed CpG
sequences, o en loca ed a gene p omo e s.
DNA combing. A me hod o he analysis o single DNA molecules; used o s udying
DNA eplica ion.
DNA damage esponse (DDR). Ne wo k o DNA damage epai and checkpoin
ac o s ha oge he o deal wi h DNA lesions.
G-quad uplexes. Fou epea s o a leas h ee guanines ha can in e ac o o m ou -
s anded DNA s uc u es.
Hai pins. DNA s uc u es in which a s and olds on i sel and o ms in as and base
pai ing.
non-B DNA. Any DNA s uc u e ha is di e en om igh -handed double helix wi h 10
nucleo ides pe u n.
Phosphomime ic. P o eins wi h amino acid subs i u ions ha simula e hei
phospho yla ed s a e.