You are on page 1of 29

Article doi:10.

1038/nature25154

Alcohol and endogenous aldehydes damage
chromosomes and mutate stem cells
Juan I. Garaycoechea1, Gerry P. Crossan1, Frédéric Langevin1, Lee Mulderrig1, Sandra Louzada2, Fentang Yang2,
Guillaume Guilbaud1, Naomi Park2, Sophie Roerink2, Serena Nik-Zainal2, Michael R. Stratton2 & Ketan J. Patel1,3

Haematopoietic stem cells renew blood. Accumulation of DNA damage in these cells promotes their decline, while
misrepair of this damage initiates malignancies. Here we describe the features and mutational landscape of DNA
damage caused by acetaldehyde, an endogenous and alcohol-derived metabolite. This damage results in DNA double-
stranded breaks that, despite stimulating recombination repair, also cause chromosome rearrangements. We combined
transplantation of single haematopoietic stem cells with whole-genome sequencing to show that this damage occurs
in stem cells, leading to deletions and rearrangements that are indicative of microhomology-mediated end-joining
repair. Moreover, deletion of p53 completely rescues the survival of aldehyde-stressed and mutated haematopoietic
stem cells, but does not change the pattern or the intensity of genome instability within individual stem cells. These
findings characterize the mutation of the stem-cell genome by an alcohol-derived and endogenous source of DNA damage.
Furthermore, we identify how the choice of DNA-repair pathway and a stringent p53 response limit the transmission of
aldehyde-induced mutations in stem cells.

The consumption of alcohol contributes to global mortality and ­cancer and repair mechanisms that counteract this. We also establish a method
development1. Most of the toxic effects of alcohol are probably caused that allows us to determine the mutational landscape of individual
by its oxidation product acetaldehyde, which is highly reactive towards HSCs, and in doing so, provide new insight into the p53 response in
DNA2. The enzyme aldehyde dehydrogenase 2 (ALDH2) p ­ revents mutagenized stem cells.
acetal­dehyde accumulation by oxidizing it efficiently to acetate, but
around 540 million people carry a polymorphism in ALDH2 that Ethanol stimulates homologous recombination repair
encodes a dominant-negative variant of the enzyme3. Alcohol con- Aldh2−/−Fancd2−/− mice develop severe HSC attrition, causing spon-
sumption in these individuals induces an aversive reaction and predis- taneous bone marrow failure, which can also be induced by exposing
poses them to oesophageal cancer4. Nevertheless, ALDH2 deficiency these mice to ethanol5,6. This genetic interaction suggests that in the
is surprisingly well tolerated in humans. This could be because of the absence of aldehyde catabolism (such as in Aldh2−/− mice), DNA repair
additional tier of protection provided by FANCD2, a DNA-crosslink- is engaged to maintain blood homeostasis. To test this theory, we set out
repair protein. In fact, genetic inactivation of Aldh2 and Fancd2 in to monitor DNA repair activity in vivo. The Fanconi anaemia pathway
mice leads to cancer and a profound haematopoietic phenotype5,6. repairs DNA crosslinks by using a replication-coupled excision mecha-
In humans, deficiency in DNA-crosslink repair causes the inherited nism that is completed by homologous recombination14,15. We therefore
illness Fanconi anaemia, a devastating condition that leads to abnormal used a method to visualize sister-chromatid exchange (SCE) events in
develop­ment, bone-marrow failure and cancer7. Acetaldehyde geno- bone marrow cells of living mice; these represent recombination repair
toxicity is likely to contribute to this phenotype, as Japanese children transactions coupled to replication (Fig. 1a). The number of SCE events
who are afflicted with Fanconi anaemia and carry the ALDH2 poly- is elevated 2.3-fold in Aldh2−/− mice, indicating that recombination
morphism display earlier-onset bone marrow failure8. Together, these repair is stimulated in response to endogenous aldehydes (Fig. 1b, c).
data suggest that endogenous aldehydes are a ubiquitous source of DNA Moreover, a single exposure to alcohol causes a fourfold increase
damage that impairs blood production. in SCE events in Aldh2−/− mice (Fig. 1b, c, Extended Data Fig. 1a),
It is likely that some of this damage occurs in haematopoietic stem suggesting that physiological acetaldehyde accumulation in blood cells
cells (HSCs), which are responsible for lifelong blood production. HSC is not sufficient to inactivate the homologous recombination repair
attrition is a feature of ageing, and mutagenesis in the remaining HSCs ­factor BRCA216. Fancd2−/− mice do not show a similar induction
promotes dysfunctional haematopoiesis and leukaemia. Moreover, both ­following exposure to ethanol; therefore, detoxification is the primary
humans and mice that lack DNA repair factors are prone to HSC loss, mechanism that prevents DNA damage by aldehydes and alcohol.
and in some cases, bone marrow failure9,10. HSCs employ DNA repair Finally, the number of SCE events in Aldh2−/−Fancd2−/− mice is indis-
and respond to damage in a distinct manner compared to later pro- tinguishable from that in Aldh2−/− mice, showing that homologous
genitors11,12. While these observations point to a fundamental role for recombination repair occurs despite inactivation of FANCD2 (Fig. 1c,
DNA repair in HSCs, recent work has highlighted that the response to Extended Data Fig. 1b).
replication stress maintains HSC function and integrity13. However, The repair of aldehyde-induced DNA damage is therefore not
there is a key gap in our knowledge regarding the identity of the endog- ­limited to the Fanconi anaemia crosslink-repair pathway. As the
enous factors that damage DNA and lead to replication stress. Here recombination machinery is essential for mouse development, we used
we show that alcohol-derived and endogenous aldehydes damage the the isogenic chicken B-cell line DT40, which has been used to define
genomes of haematopoietic cells, and we characterize the surveillance the involvement of homologous recombination in crosslink repair14.
1
MRC Laboratory of Molecular Biology, Cambridge Biomedical Campus, Francis Crick Avenue, Cambridge CB2 0QH, UK. 2Wellcome Trust Sanger Institute, Hinxton, Cambridge CB10 1SA, UK.
3
Department of Medicine, University of Cambridge, Addenbrooke’s Hospital, Hills Rd, Cambridge CB2 0QQ, UK.

0 0 M o n t h 2 0 1 7 | VO L 0 0 0 | NAT U R E | 1
© 2018 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.

RESEARCH Article

a c d 100 e 100
BrdU pellet Colchicine IP NS (P = 0.9067) DT40 DT40
40
10 10 ∆XRCC2

Survival (%)
Survival (%)
∆XRCC3
12 h, 30 min ∆RAD51C
± Ethanol, 12 h Harvest 1 ∆PALB2 1
BM cells ∆RAD51D
b 30 P < 0.0001 ∆RAD51B

SCEs per metaphase
0.10 ∆FANCC 0.10 ∆FANCC
P < 0.0001
∆BRCA2

20 P < 0.0001 0.01 0.01
0 2 4 6 8 0 2 4 6 8
Acetaldehyde (mM) Acetaldehyde (mM)

(n = 2) (n = 6) (n = 4) f 100 g 100
10 DT40
Wild type Fancd2–/– Wild type + ethanol DT40
∆XRCC2
10

Survival (%)

Survival (%)
∆FANCC ∆XRCC2
10 ∆FANCC/ 1
0 ∆FANCC
– – ∆XRCC2
–/ –/
– –/ –/


pe

pe
2 2 2 2

2 –/

–/

2 –/

–/
dh d dh d
ty

ty
d2

d2
Al anc Al anc 0.10 ∆FANCC/

dh

dh
ild

ild
nc

nc
Al

Al
∆XRCC2
W

W
Fa

Fa
(n = 6) (n = 7) (n = 21) F F
Aldh2–/– Aldh2–/– Fancd2–/– Aldh2–/– + ethanol + Ethanol 1 0.01
0 0.5 1.0 1.5 2.0 0 2 4 6 8
Cisplatin (μM) Acetaldehyde (mM)

Figure 1 | Ethanol induces potent homologous recombination in vivo. the bone marrow of Aldh2−/−Fancd2−/− and control mice (triplicate
a, Treatment of mice with BrdU for differential labelling of sister experiments, 25 metaphases per mouse, n =​  75; P calculated by two-sided
chromatids of bone marrow cells in vivo. Some mice were also treated Mann–Whitney test; data shown as mean and s.e.m.). NS, not significant.
with ethanol, a precursor of acetaldehyde. IP, intraperitoneal injection; d–g, Clonogenic survival of DT40 DNA-repair mutants (triplicate
BM, bone marrow. b, Representative images of bone-marrow metaphase experiments; data shown as mean and s.e.m.).
spreads (n, number of SCEs per metaphase). c, Number of SCEs in

DT40 cells carrying disruptions of key homologous recombination FANCD2 prevents alcohol-induced genomic instability
genes show hypersensitivity to acetaldehyde (Fig. 1d, e), in a similar The active DNA recombination in bone marrow cells indicates that
way to cells lacking the Fanconi anaemia gene FANCC. To test the rela- even in the absence of FANCD2, there is an alternative repair response
tionship between the Fanconi anaemia and homologous recombination to both endogenous and ethanol-derived aldehydes. However, our
pathways, we analysed the sensitivity of cells deficient in both FANCC ­previous work has shown that Aldh2−/−Fancd2−/− mice lose the a­ bility
and XRCC2. These cells showed the same sensitivity to cisplatin as the to maintain blood production5,6. To determine whether this is due to
­single knockout cells (Fig. 1f), but were much more sensitive to acetal­ the accumulation of damaged DNA, we examined ­haematopoietic
dehyde (Fig. 1g), indicating that homologous recombination repair cells for evidence of broken chromosomes. One marker of genetic
confers additional acetaldehyde resistance beyond that provided by instability is the formation of micronuclei, which are formed from
Fanconi anaemia crosslink repair. In summary, detoxification provides lagging or broken chromosomes. Micronuclei are easily quantified
the dominant protection mechanism against endogenous aldehydes; in normochromic erythrocytes (NCEs) in vivo, because they persist
however, when aldehydes damage DNA, cells use both DNA-crosslink following enucleation (Fig. 2a, Extended Data Fig. 1c). There is a
and homologous recombination repair. significant increase in the proportion of NCEs with micronuclei in

a Wild type Aldh2–/– Fancd2–/– b 2.5
9.5-fold,
P < 0.0001
c 25 P = 0.0004 d e 25 Duplication
105 Aneuploidy
1.9-fold, NS

Aberrations in 90 metaphases
Micronucleated NCEs (%)

Abnormal metaphases (%)

2.0 20 Translocation
P < 0.0001 (P = 0.12) 20
der(1,5) 2 3 4 der(1,5)
T(1,5) Deletion
104 Chromatid break
CD71 (FITC)

1.5 2.9-fold, 15 NS 15
P < 0.0001 (P = 0.25)
6 7 8 9 10
103
1.0 10 10
11 12 der(2,13)
T(2,13) 14 15
102 Mn-NCE Mn-NCE 0.5
0.13 ± 0.005% 1.24 ± 0.05% 5 5
0
0.0 16 17 18 19 X Y
0 103 104 105 0 103 104 105 –/

/–
0 0
40, XY, der(1)T(1;5), der(5)T(1;5), der(13)T(2;13)
Fa h2 –/–

– –
Al pe

2 2– –/ –/
–/

– –
–/ –/
nc –

nc –

Al pe

Al pe

DNA (PI) DNA (PI) dh d
Fa h2 –/

–/

Fa h2 –/

–/

2 2 2 2
ty

d2

Al anc dh d dh d
ty

ty
d2

d2
ild

d
nc

Al anc Al anc
ild

ild
d

d
W

F
W

W

F F
f g 8
h 80 P = 0.0017
3.9-fold,
P < 0.0001
reticulocytes (%)

metaphases (%)
Micronucleated

6 1.9-fold, 3.0-fold, NS 60
Abnormal

P = 0.0001 P < 0.0001 (P = 0.1)
48 30
hours minutes 4 40
or
Ethanol IP Blood sample Colchicine IP Bone marrow 2 20
5.8 g kg–1 measure Mn-Ret M-FISH
0 0
Ethanol – + – + – + – + Ethanol – + – + – + – +
Wild type Aldh2–/– Fancd2–/– Aldh2–/– Wild type Aldh2–/– Fancd2–/– Aldh2–/–
Fancd2–/– Fancd2–/–

Figure 2 | Spontaneous and ethanol-induced genomic instability chromosomal aberrations (90 metaphases per genotype). f, Treatment
in Aldh2−/−Fancd2−/− mice. a, Quantification of micronucleated of mice with ethanol to assess genomic instability with the micronucleus
normochromic erythrocytes (Mn-NCE, CD71− PI+) by flow cytometry. assay (g) or M-FISH karyotyping (h). g, Percentage of micronucleated
b, Percentage of micronucleated normochromic erythrocytes (P calculated reticulocytes (Mn-Ret, CD71+ PI+) after ethanol treatment. P calculated
by two-sided Mann–Whitney test; data shown as mean and s.e.m.; n =​  28, by two-sided Mann–Whitney test; data shown as mean and s.e.m.; n =​  29,
28, 25 and 37 mice, left to right). c, Percentage of abnormal metaphases 15, 25, 15, 20, 10, 28 and 9 mice, left to right. h, Abnormal metaphases
in bone marrow cells (P calculated by one-sided Fisher’s exact test; data in bone marrow cells after ethanol treatment. P calculated by one-sided
shown as mean and s.e.m.; three mice per genotype, 30 metaphases per Fisher’s exact test; data shown as mean and s.e.m.; 3 mice per genotype,
mouse). d, A Aldh2−/−Fancd2−/− metaphase, showing two translocations, 30 metaphases per mouse.
see Extended Data Fig. 1f–i for the complete list of aberrations. e, Types of

2 | NAT U R E | VO L 0 0 0 | 0 0 m o n t h 2 0 1 7
© 2018 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.

data shown as mean and s. However.0001 f P = 0. 2b). NHEJ is required mosomes in bone marrow cells. 3b–e).0013 100 Micronucleated NCEs (%) response to aldehydes. Fancafl/-Ku70−/− Vav1-iCre ST-HSCs were much Our results so far indicate that endogenous aldehydes give rise to more sensitive to acetaldehyde than either of the single mutant ST-HSCs DSBs in the absence of Fanconi anaemia repair. which could be 5 0.9-fold) mice compared to a Fancafl/+ Fancafl/– Fancafl/+ Ku70–/– Fancafl/– Ku70–/– (all Vav1-iCre+) wild-type controls. blood counts show that Fancafl/-Ku70−/− Vav1-iCre mice are ­anaemic (Fig. A single dose of ethanol 0 caused a marked increase in the proportion of reticulocytes containing 0 103 104 105 103 104 105 micronuclei in Aldh2−/− mice.0101 P = 0. engagement of DSBs by NHEJ leads nc nc nc nc nc nc 7 7 7 n n n Fa Fa Fa to further genomic instability18. 3a) and have fewer HSCs compared to congenic controls (Fig. 2h.to 12-week old mice (P calculated by maintaining resistance to acetaldehyde. such as ionizing irradiation or vincris.8 Ku70 contributes to repair of aldehyde-induced DSBs 10 0.. 6.008% HSCs 0. 2c–e). also known as Ku70).5) 100 type or Fancd2−/− mice to mitomycin C (Extended Data Fig. b Fancafl/+ Vav1-iCre+ Fancafl/– Ku70–/– Vav1-iCre+ c 150 P = 0. but the increase is much larger in Aldh2−/− 15 P = 0.15) Mean corpuscular volume (μm3) White blood cells (×103 mm–3) 54 Fancd2−/− mice (9. Frequency of Mn-NCE (P calculated by compartment (Fancafl/-Ku70−/− Vav1-iCre). Therefore. part of Springer Nature. and control mice.e. the Fanconi anaemia crosslink-repair pathway is essential for e P < 0. b. Article RESEARCH both Aldh2−/− (2.0063 10 2. with increased frequency of micronuclei-containing NCEs active NHEJ on the viability of Fanconi anaemia-deficient chicken (Fig. 3f). in the absence (all Vav1-iCre+) Fa a fl/+ Ku a fl/– 70 /– Fa a fl/+ Ku a fl/– 70 /– Fa a fl/+ Ku a fl/– 70 /– Fa 0 –/– Fa 0 –/– Fa 0 –/– – – – Ku ca fl Ku ca fl Ku ca fl –/ –/ –/ of the Fanconi anaemia pathway.m. 1e). there was a stronger induc. encompassing all classes of cytogenetic change 46 (Fig. Each point represents the number of CFU-S12 in a single out mice (Extended Data Fig. More 49 4 5 than 10% of Aldh2−/−Fancd2−/− bone-marrow cells carried chromo.0 (% relative to untreated) 15 CFU-S12 survival 0.19. we crossed mice deficient in the known Fanconi and 5 mice. n =​  10 mice).000 NS (P = 0.0001 P = 0. These micronuclei could represent Red blood cells (×106 mm–3) P = 0. which was accompanied by a striking increase in 103 the ­number of abnormal metaphases.0043 g P = 0.5-fold. Lin−Kit+Sca-1+. Notably. genomic stability and cellular resistance to genous DNA damage in HSCs. despite activation of homologous recombi. LKS. Previous 0 0 1 studies in cell lines and nematodes have indicated that.m. As a control. c. due to failed blood production (Supplementary Information Table 1).0485 HSCs (% of wild type) tion of micronucleus formation in Aldh2−/−Fancd2−/− mice than in 104 100 CD48 (BV421) controls (Fig. 2f) shows how we deter- mined the prevalence of micronuclei in reticulocytes and aberrant 50 metaphases following exposure to ethanol.m.0101 55 P = 0. 105 150 tine (Extended Data Fig. and whether there is a role for NHEJ in aldehydes. Finally. 2). Fancafl/-Ku70−/− Vav1-iCre mice also display genomic damage.. These results indicate that in the mouse ­haematopoietic homologous recombination and NHEJ. 1). These mice rapidly lost the ability to pro. Fig. .19. Representative flow cytometry plot of anaemia repair gene Fanca with mice lacking the key NHEJ ­factor haematopoietic stem and progenitor cells (HSPCs) of Fancafl/-Ku70−/− Vav1-iCre mice and control.e. 7 To do this. we tested whether Ku70 was required to maintain DT40 cells and worms18. d. 2g).. 1d). we generated blood-specific Fanca knock. but ultimately rearrange chro- ­system. data shown as to produce mice that had the double mutation in HSCs and the blood mean and s. These mice were viable.2 processed by non-homologous end-joining (NHEJ) repair17. this induction was comparable 0 0 Fa a fl/+ Ku fl/– 70 /– Fa 0 –/– – Ku ca fl –/ a Sca-1 (PE-Cy7) Sca-1 (PE-Cy7) nc nc 7 n to that observed in wild-type mice following exposure to agents known Fa d to induce genome instability. Extended Data Fig.04% c-Kit (PerCP-Cy5.0002 1.e. We failed to obtain of HSPCs (Lin−Kit+Sca-1+) and HSCs (Lin−Kit+Sca-1+CD48−CD150+) Fanca −/−Ku70 −/− mice. Quantification Ku70 (encoded by Xrcc6. 0 103 104 105 0 103 104 105 0 Fa a fl/+ Ku fl/– 70 /– Fa 0 –/– – Ku ca fl –/ These results show that. This result contrasts with the reported negative impact of instability.001% duce blood and died from bone-marrow failure (Extended Data Fig. we investigated whether Figure 3 | NHEJ cooperates with the Fanconi anaemia pathway to NHEJ and Fanconi anaemia repair are redundant in resolving endo­ maintain HSC integrity.2 CFU-S12 per 105 nucleated BM cells 20 preventing chromosome breakage and loss of blood homeostasis in P < 0. f. LKS 0. n =​  20 mice). 1g–i). 3) and crossed them with Ku70+/− mice recipient (P calculated by two-sided Mann–Whitney test. data shown lethal interaction between Ku70-dependent NHEJ and Fanconi as mean and s. We therefore examined 10 52 NS cells in metaphase obtained directly from the bone marrow of these 51 50 6 (P > 0. To (CFU-S) colonies from the bone marrow of Fancafl/-Ku70−/− Vav1-iCre bypass embryonic lethality. data shown as mean and s.e.0088 53 8 Platelets (×103 mm–3) 1. P = 0. ­indicating that the embryonic lethality of Fanca−/−Ku70−/− is not g. n as in a).500 genomic instability during blood production. relative to untreated samples (P calculated by two-sided Mann–Whitney test.e. Survival of CFU-S12 after treatment with 4 mM acetaldehyde for 4 h.m. 0 HSCs 0. two-sided Mann–Whitney test. which are engaged by (Fig. e. we investigated whether this chromosome damage was HSPCs (% of wild type) 104 exacerbated by exposure to ethanol. a. 103 The experimental scheme (outlined in Fig. with almost 60% of metaphases 50 having ­damaged chromosomes following ethanol exposure (Fig. However. These aberrations are not clonal events. 3g). data shown as mean and s. CD150 (PE) CD150 (PE) a nc nc 7 n Fa nation. we exposed wild.m. 48 47 2 500 somal aberrations. A key question is whether e­ ndogenous to provide resistance to endogenous and a­ cetaldehyde-induced DNA DNA damage and subsequent mutations accumulate in the HSC 0 0 M o n t h 2 0 1 7 | VO L 0 0 0 | NAT U R E | 3 © 2018 Macmillan Publishers Limited. resistance of short-term (ST)-HSCs to aldehydes by exposing bone marrow cells to acetaldehyde in vitro before injecting them into lethally Aldehyde-damaged HSCs are functionally compromised irradiated recipients. Counts of colony-forming unit-spleen ­anaemia-crosslink repair (Supplementary Information Table 1).. because each karyo­ 0 45 0 0 type was unique (Extended Data Fig. Blood parameters of 8. All rights reserved.16% LKS 0. two-sided Mann–Whitney test.99) 1.6 10 The presence of chromosome breaks and translocations suggests 0.0556 Next. left to right).4 that aldehydes cause double-stranded breaks (DSBs).. n =​ 8. indicating that there was a synthetic by flow cytometry (P calculated by two-tailed Student’s t-test. in the absence of Fanconi anaemia repair. 1. n as in a).9-fold) and Fancd2−/− (1.000 mice with multiplex fluorescence in situ hybridization (M-FISH).

to ­establish whether endogenous aldehydes mutate the genomes of Although the number of single-base substitutions was significantly these vital cells. we found that microhomology-­mediated whole-genome sequencing to obtain the mutational landscape of deletions are the main contributors to the mutations observed in stem cells. 5k–l). in functional stem cells. four months after transplantation. b. We also wanted to ascertain whether mutations arise cII Big Blue in vivo reporter assay (Extended Data Fig. (Fig.0007 75 75 10 Contribution (%) 12 / 36 1 12 / 40 50 50 11 / 12 10 / 12 12 / 21 16 / 79 5 / 38 0. 5c). Extended Data Fig. This is a critical question because there is evidence mutations in the HSC. RESEARCH Article a Sort single HSCs Sort HSC progeny c-Kit (PerCP-Cy5. contributed less to haematopoiesis and were breaks. low and no changes were detected in the type of substitutions (Fig. production following transplantation into a lethally irradiated The mean variant allele frequency (VAF) for all filtered indels was mouse20. we decided to define HSCs functionally and nificant increases in the frequency of deletions. irradiated recipients that were positive for reconstitution by one or five data shown as mean and s. j).). 16 and 6 HSC clones. 0.0009 P = 0.0001 P < 0. we isolated the CD45. e. A limitation of our approach is that cells with the capacity to engraft fication or short-term in vitro expansion of cells isolated by flow may represent the least mutated HSCs.0001 100 Fancd2–/– (n = 21) 100 100 Aldh2–/–Fancd2–/– (n = 38) Reconstitution efficiency (%) Reconstitution efficiency (%) P = 0. 5g.22. 4a).m. First. which were more preva­ exploit the ability of a single HSC to reconstitute long-term blood lent (Fig. columns represent individual were recovered after four months and analysed by whole-genome recipients or clones.. We carried out transplants with one or five (Fig. These stem cells rarely engrafted suggests a role for alternative end-joining in the repair of some of these (with a frequency of 4. indicative of end-joining repair of DSBs23 mutant HSCs (Fig. ments per genome. Transplantation of single HSCs (data shown as mean and s. 4a). Contribution to blood production over time by five transplanted HSCs HSCs are functionally compromised. as recent work has suggested a role and share features with aged HSCs21. which provided us with regions or transcribed genes (Fig. right). c­ ompartment. Instead.10 6 / 123 25 25 0 0 0. All rights reserved. Nevertheless.e.1 Wild type (n = 12) b 1 HSC 5 HSCs c Aldh2–/– (n = 12) P < 0.0138 e W W d Gr-1+ Mac-1+ B220+ CD4+ CD8+ NS (P = 0. for the Fanconi anaemia pathway in preventing genomic instability at t­ ranscription–replication collisions25.33) 100 100 Myeloid chimaerism (%) Lineage distribution (%) 75 75 50 50 25 25 0 0 nc 2 –/– – nc 2 –/– Fa ldh /– Fa ldh e –/ 1 2 3 4 5 6 7 8 910 12 1 2 3 4 5 6 7 8 910 12 1 2 3 4 5 6 7 8 910 12 14 16 1 2 3 4 5 6 A d2 – A typ d2 Wild type Aldh2–/– Fancd2–/– Aldh2–/– ild Fancd2–/– W Figure 4 | Single HSC transplantation reveals that Aldh2−/−Fancd2−/− c. rearrangements and trans- to later progenitors11. 5i. tail DNA from the donor mouse served as the germline Aldh2−/−Fancd2−/− stem cells contained on average two rearrange- ­reference.01 0 4 8 12 16 nc 2 –/– – nc 2 –/– – Fa 2 –/– – Fa 2 –/– – Al pe Al pe –/ –/ –/ –/ Time (weeks) d2 d2 d2 d2 ty ty Fa ldh Fa dh dh dh nc nc ild ild Al A P = 0.5) CD45. In summary. left to transplanted HSCs (P calculated by two-sided Fisher’s exact test). 4b–e). the total numbers were unlikely that the same mutation will occur in multiple cells. especially in the case of Aldh2−/−Fancd2−/− We also found no difference in the frequency or pattern of point muta- mice. and therefore avoided whole-genome ampli.2+) HSC clones four months after transplant. 16 weeks CD150+ CD48– WGS: c-Kit+ Sca-1+ WGS: Germline Sca-1 (PE-Cy7) HSC clone CD45. suggesting that DSB formation a physiological method to amplify stem-cell genomes. precludes the use of most conventional techniques to assess tions in bone marrow cells of Aldh2−/−Fancd2−/− mice using the Select- DNA damage. h).2+ The most striking change in Aldh2−/−Fancd2−/− HSCs was the HSC progeny and performed whole-genome sequencing at 20×​ presence of rearrangements that were not detected in most controls.2 Visual confirmation Lineage. we observed sig- cytometry. these data 4 | NAT U R E | VO L 0 0 0 | 0 0 m o n t h 2 0 1 7 © 2018 Macmillan Publishers Limited. This allowed us to detect heterozygous somatic changes. Lineage composition of single- for the generation of HSC clones in vivo. n =​ 12. . 12. However. a. 5f) Aldh2−/−Fancd2−/− HSCs (Fig. as alternative end-joining is characterized by increased resec- myeloid-biased compared to controls (Fig. part of Springer Nature. alongside a germline reference. 5d. Percentage and number of derived white blood cells (P calculated by two-sided Mann–Whitney test. Proportion of myeloid (Gr-1+Mac-1+) HSC- sequencing. 5). 4b).26. we observed only two large deletions which are absent in the matched germline reference and represent among all ten control HSC genomes (Fig.47. Two obstacles had to be overcome in order locations (Fig.8%). The HSC progeny (CD45. 5a. in contrast. 4). e) and larger (Fig. the scarcity of HSCs. Genomes of Aldh2−/−Fancd2−/− HSCs were that HSCs differ in their DNA repair capacity and response compared mutated with increased prevalence of indels. d. establishing that these changes are of clonal origin. These results indicate tion in comparison to classical NHEJ24. Additionally. By exami­ Our approach combines transplantation of single HSCs with ning the flanking regions. we analysed the loca- that Aldh2−/−Fancd2−/− HSCs are severely functionally compromised tion of indels across the genome.m.e. 5b). the increase in the size of the deletions (Fig. we found no evi- Mutational landscape of aldehyde-damaged stem cells dence of microhomology-mediated deletions being enriched at coding Our ultimate goal was to obtain clonal blood. ­coverage. while also allowing us to assess the functional capacity of Aldh2−/−Fancd2−/− HSCs. 5f) in Aldh2−/−Fancd2−/− genomes. As outlined in Aldh2−/−Fancd2−/− HSCs is stochastic. Next. the stochastic nature of DNA damage makes it higher in Aldh2−/−Fancd2−/− genomes (Fig. Second.

4 and 5 HSC genomes. most Aldh2−/−Fancd2−/− HSCs failed to engraft (Fig.m. recently reported27. left to right). 8a).8 2–10 bp 25 25 MH-mediated 11–50 bp NS 20 20 20 (P = 0. the haematopoietic stem and progenitor cells (HSPCs) accumulated p53 genomes of Aldh2−/−Fancd2−/−Trp53−/− stem cells did not carry a and cleaved caspase-3. we HSCs (Fig. Moreover. n = 323 substitutions. but this analysis showed no significant difference transplantation (Extended Data Fig. Point mutation classes in HSC genotypes. One plausible explanation for the macrophage colony forming units (Extended Data Fig. b). substitutions. All rights reserved.34) 10 NS (P = 0. Extended Data Fig. It in Aldh2 −/−Fancd2 −/−Trp53 −/− mice. 0 0 M o n t h 2 0 1 7 | VO L 0 0 0 | NAT U R E | 5 © 2018 Macmillan Publishers Limited. when activated. Number Aldh2−/−Fancd2−/− and Aldh2−/− HSCs at the indicated locations. n is number of indels). Notably.73 Gb 25 25 n = 20 n = 24 n = 42 n = 10 n=7 1 n=7 4 Aldh2–/– 0 0 0 2 c13:1. f. Circos of repeat-mediated deletions per genome. as described earlier. as the mice showed this. Article RESEARCH a Wild-type HSC b c P = 0. while Trp53 deletion completely res- gesting that a different checkpoint might mediate developmental failure cues HSC depletion. Furthermore.055) P = 0. All HSCs are shown in Extended (MH)-mediated deletions per genome.86 Gb n = 10 n = 11 n = 19 Non-coding Transcribed P = 0.0159 P = 0.5 Substitutions per genome Relative contribution P = 0.1 T>G 0 0 n = 161 substitutions.0343 Deletion size g h P = 0. which is inconsistent Data Fig. We found that Aldh2−/−Fancd2−/− defect of Aldh2 −/−Fancd2 −/− HSCs (Fig. n =​ 3. cycles. 4b). induces observed that p53 deficiency partially rescued the engraftment restorative processes or apoptosis. Indels in Aldh2−/−Fancd2−/− Data Fig. Fancd2−/−Trp53−/− mice must be occurring at the cost of genome nous aldehydes induce a tapestry of inter-chromosomal changes that integrity. i. . 6d. Surprisingly. The severe HSC depletion of Aldh2−/−Fancd2−/− mice was Aldh2−/−Fancd2−/− mice might have undergone more replicative completely rescued in the triple knockouts (Fig. b. ysis of total bone marrow cells (Fig. The p53 protein regu.4 10 (P = 0.. It was therefore important to determine the mechanism of HSC is similarly compromised in Aldh2−/−Fancd2−/−Trp53−/− HSCs. lack of increased mutagene­s is in Aldh2 −/−Fancd2 −/−Trp53 −/− We therefore generated Aldh2−/−Fancd2−/−Trp53−/− triple-knock. n shows number of deletions).453 x 108 11744 n=3 n=3 n=9 0. part of Springer Nature. sug. HSCs is the ­possibility that the very small number of HSCs in out mice. g. respectively (Extended hypersensitivity of Fancd2-deficient splenic B cells and granulocyte/ Data Figs 9. and if attenuated.0179 Aldh2–/– Fancd2–/– 100 100 5 2 Coding Silenced c5:1.0 1.0079 0 Proportion (%) 4 Proportion (%) Copy number 75 75 4 3 Aldh2–/– Fancd2–/– c15 50 50 2 c4:2.0357 Wild type Aldh2–/– Fancd2–/– Aldh2–/– Fancd2–/– 200 0. bined with whole-genome sequencing.0357 30 30 1. Number of deletions per genome. n = 664 substitutions. Distribution of rearrangements per genome.e. a. In addition. indicating that endogenous-aldehyde stress greater mutation burden compared to those of Aldh2−/−Fancd2−/− ­activates the p53 response (Extended Data Fig. 6c.2 T>A ements 50 T>C 0. Indel and rearrangement calls were validated found that genetic ablation of p53 partially suppressed the acetaldehyde by targeted deep sequencing and PCR. neither of these analyses tell us whether genome stability cells. this does not occur at the expense of genome (Supplementary Information Table 2). 6d).054) n = 35 n = 99 0. Although the level of micronucleated NCEs in the blood are mediated by mutagenic end-joining of DNA DSBs.915 x 107 d2 – – (M ns (M ns se e le s le ns pe – Al –/– De pe s De pea s se e In nomC le at) le t) nc 2 –/ –/ 0 ) ) 2 –/ De tion H H (re tion om (re tion ge HS tio tio De rtio ty d2 Position (50 kbp) Fa dh dh en r ild nc c10 c9 Al G W In Fa Figure 5 | Endogenous aldehydes mutate the HSC genome.4 C>G Inde titut 150 P = 0. To address the triple-knockout stem cells were functional. nated.975 x 107 n = 20993 2 0 2. These results indicate that aldehyde-induced DNA and made these mice more resistant to alcohol exposure (Extended damage induces p53 leading to HSC attrition. d. stability in blood stem cells. 6a. d2 – – pe Fa 2 –/– Al –/– nc 2 –/ –/ 3 genomes 3 genomes 4 genomes 5 genomes ty d2 Fa dh dh ild nc Al W Aldh2–/–Fancd2–/– HSCs d NS (P = 0. and selection pressure favours the survival of less damaged stem However. Mutations of different classes per genome HSCs are randomly distributed: within or outside genes (i) (P calculated (P calculated by two-sided Mann–Whitney test.0357 f P = 0. provide the first whole-genome sequences obtained from single stem We reasoned that the rescue of haematopoiesis in Aldh2 −/− cells propagated in vivo. Number of microhomology plots showing mutations in three HSCs.34 0. 7). 10 and Methods). 4. of Aldh2−/−Fancd2−/−Trp53−/− mice appeared similar to that of Aldh2−/−Fancd2−/− mice (Extended Data Fig. and lates the cellular response to DNA damage and. Numbers above columns. 6a–c). 3.34 P = 0. Trp53 deletion did not rescue the embryonic with p53 being a negative regulator of Fanconi anaemia repair. 8).3 on 100 C > T at CpG s ng 0. Number of expressed or silenced genes (j) (P calculated by binomial distribution. Number of n is number of indels).0286 Rearls i C>T ra 0.09) 15 NS 15 c16 0. it would be impor. k. h. we quantified the frequency of ‘clock’ mutations (C to T at a complete rescue in the frequency of ST-HSCs upon bone-marrow CpG sites). l.0 30 30 Insertions per genome Relative contributution deletions per genome Deletions per genome 1 bp deletions per genome NS (P = 0.13) e P = 0. we noticed a A p53 response removes aldehyde-damaged HSCs significant (P =​ 0. e). k. p53 deficiency fully between Aldh2 −/−Fancd2 −/− and Aldh2 −/−Fancd2 −/−Trp53 −/− restored the blood cytopenias of untreated Aldh2−/−Fancd2−/− mice HSCs (Fig.48 1. as lethality of Aldh2−/−Fancd2−/− embryos (in Aldh2−/− mothers). c.0159 C>A Subs 0. n = 156 substitutions.0 0. d. However.6 > 50 bp 15 NS 15 (P = 0.2 n=9 n=7 5 5 5 5 0 0 0 0 0 pe Fa 2 –/– Al –/– d2 – – pe – Al –/– d2 – – pe Fa 2 –/– – d2 – – pe – – d2 – – pe – Al –/– d2 – – nc 2 –/ –/ –/ nc 2 –/ –/ –/ nc 2 –/ –/ –/ –/ nc 2 –/ –/ –/ nc 2 –/ –/ ty ty ty ty ty d2 h2 d2 d2 h2 d2 h2 d2 dh Fa dh Fa dh dh Fa ldh Fa ldh Fa dh ild ild ild ild ild d d d nc nc nc nc nc Al Al Al Al Al A A W W W W W Fa Fa Fa i NS NS NS j NS NS NS k l 4 Rearrangements per genome 0. loss in Aldh2−/−Fancd2−/− mice. j. b. f ). Large copy-number losses in of the size of deletions (χ2 test. Number insertions per genome. or between and s. These stem cell genomes show that endoge. We therefore performed transplantation of single HSCs com- tant to determine the mutagenic consequences. g. thereby accruing a larger number of mutations.0034) increase in chromosome rearrangements Strikingly.43 0. h. 6e. as seen by M-FISH anal- is possible that HSCs that carry heavy DNA-damage burdens are elimi. 6f). 6f). data shown as mean by hypergeometric distribution. P values. e.99) 10 10 (P = 0.054) 20 NS 0.0079 Repeat-mediated c2 25 25 0. e.

DNA repair is limiting for haematopoietic stem cells during mediated deletions. irradiated recipients that were positive for reconstitution by transplantation 8. et al. Alcohol as a risk factor for cancer: existing 4 30 Substitutions 3 Rearrangements C > T at CpG evidence in a global perspective.. 174–185 (2010). are available in the online version of the paper. f. Ku70 corrupts DNA repair in the absence of the Fanconi anemia in HSCs (Extended Data Fig. Long-term diverse range of DNA lesions—from base adducts to DNA–DNA or lymphohematopoietic reconstitution by a single CD34-low/negative DNA–­protein crosslinks. 20 2 2. substitutions and clock substitutions ageing.e. et al. M. 38. & Weiderpass. L. A distinctive DNA damage response in human hematopoietic stem cells reveals an apoptosis-independent role for p53 in Discussion self-renewal.0061 c P = 0. Cell Stem Cell 7. It is important to emphasize. 6. Yokoyama. Nature 489. 4. T. Joukov. Hamada. this research provides a simple plausible explanation for the established epidemiological link 7 / 42 7 / 41 25 6 / 123 between alcohol consumption and enhanced cancer risk4. paradoxically. 3. et al. then the factors involved in the c-Kit (PerCP-Cy5. Res. C. S. E. Rosado. M. DNA repair and mutagenesis. Garaycoechea.e15 can also deal with these lesions. Cell 39. Jpn. that the metaphases (%) 30 Abnormal 10 20 pool of HSCs is larger.14.53% if it is indeed an interstrand crosslink. & Nakauchi. LKS: 2. e. C. V..0075 50 14 / 42 age-related blood disorders. and therefore there is still an overall increase in 10 mutation. Aldehydes are capable of forming a pathway. 198–202 (2014). Nature 447. Hematopoietic stem cell quiescence promotes error-prone These results outline the mechanisms by which the mouse haemato. that coordinated pathway choice is critical for maintaining genome 17. defects of Fanconi anemia. 3. 6. S. F. I. h ­ owever. 3. 607–620 (2004).. Identification of DNA adducts of acetaldehyde. Science 333. a. data shown as 10. haematopoietic stem cells with age. 20. Lai. It will be important to resolve the nature of the lesion and the precise mechanics of its repair. Chem. 30 metaphases per mouse). T. Long. rearrangements. b.. The p53 response is critical in driving the loss in b P = 0. detoxification and. Bunting. Rossi. J. The Fanconi anaemia pathway is the principal mechanism to dependent DNA interstrand cross-link repair. et al.06% LKS: 0. Crossan. 44–50 (2014). Langevin. 3. P. blood stem cells respond to an 13. left to right). MH-mediated deletions Substitutions 15 100 Published online 3 January 2018. Adamo. J. 18. & Patel. Cell 169. this. Frequency of abnormal metaphases in bone marrow counteracts the toxic effects of naturally produced aldehydes in mice. Pace. M. Science 273. et al. 1105–1118.. A.. Nevertheless. Milyavsky. data shown as mean and 475. 0 2 –/ Trp 2 –/– nc – – – –– – Fa 53 –/ 3 –/ –/ 3 –/ Online Content Methods. 53–58 (2011). mate repair of which leads to a characteristic pattern of mutagenesis 19. et al. Dominance of the inactive Asian variant over activity and protein Trp53 +/+ –/– +/+ –/– +/+ –/– +/+ –/– Trp53 +/+ –/– +/+ –/– +/+ –/– +/+ –/– contents of mitochondrial aldehyde dehydrogenase 2 in human liver. 5 and 4 HSC genomes. Blood 122. n =​ 3. dehydrogenases and risk for esophageal and head and neck cancers.m. I. part of Springer Nature.. 3206–3209 (2013). 186–197 (2010). A. s. left to right). A. 1149–1157 (2000). 3. Replication stress is a potent driver of functional decline in endogenous and alcohol-derived source of DNA damage. Representative flow cytometry plot of HSPCs (LKS). 725–729 (2007).0034 number and function of HSCs.29. Nature 512. Nijnik. dh nc nc dh Al Fa Fa NS (P = 0. Flach. references unique to Tr Tr – Aldh2–/– Fancd2–/– Trp53–/– – –/ – 2 –/ d2 these sections appear only in the online paper. I. Tan. poietic system and.5) 104 translesion synthesis or homologous recombination processes are dis- 103 tinct from the previously described mechanism of interstrand-crosslink repair6. 111–121 (2003). Roswall. 26–34 (2014). Nature 447.e. G. 16. Exp. per genome (P calculated by two-sided Mann–Whitney test.. Number per genome Number per genome 10 5 50 0 0 NS (P = 0. J. Wild type Aldh2–/– Fancd2–/– Aldh2–/– Wild type Aldh2–/– Fancd2–/– Aldh2–/– Fancd2–/– Fancd2–/– Clin. M. Primary ageing haematopoietic stem cells. does not result in further genomic instabi­lity HSCs (% of wild type) 100 at the single HSC level. N. Cell Stem Cell 7. 242–245 (1996). 1–9 (2015). 219–223 (2010). Alcohol. the illegiti. 13. Niedzwiedz. M. Mol..28. Deficiencies in DNA damage repair limit the function of mean and s. 3 mice per genotype. 243–254 (2010). Räschle. 6. Mohrin. Preventing nonhomologous end joining suppresses DNA repair ­prevents aldehyde lesions from degenerating into DSBs. L. RESEARCH Article a Wild type Aldh2–/– Fancd2–/– Aldh2–/– Fancd2–/– Trp53–/– anaemia pathway suggests that the most physiologically toxic lesion 105 caused by aldehydes may be a DNA interstrand crosslink. Number of microhomology. but NHEJ and homologous recombination BRCA2 haploinsufficiency and genome instability. All rights reserved. & Omori. A class of environmental and endogenous toxins induces counteract this damage. Proportion and number of anemia? Blood 123. 0 HSCs mutated by aldehydes are functionally compromised and 0 103 104 Sca-1 (PE-Cy7) 105 0 103 104 Sca-1 (PE-Cy7) 105 0 103 104 Sca-1 (PE-Cy7) 105 ­display myeloid bias. Cell 15. acetaldehyde ­damages 15. See Extended Data 7. HSCs as determined by flow cytometry (P calculated by two-sided Mann– 5. et al. However. Med. D..056) f 20 150 received 31 January. These results therefore illustrate (2017). our work implies that the relationship between 1 p53. 12. when this is lost or saturated. J. Garaycoechea. W. d. Public Health 48. M. 571–575 (2012). P. 25–35 (2010).28) NS (P = 0. The central role for ALDH2 in removing Wild type Aldh2–/– Fancd2–/– Aldh2–/– Fancd2–/– Wild type Aldh2–/– Fancd2–/– Aldh2–/– Fancd2–/– genotoxic aldehydes has implications for the more than 540 million d NS (P = 0.53) NS (P = 0. J. Variant ALDH2 is associated with accelerated progression of bone of single HSCs (P calculated by two-sided Fisher’s exact test). 5 and on mouse haematopoietic stem cell function. F. 8. Mol. Reconstitutionefficiency (%) 75 NS (P = 0. J. More generally. Hanada.78) This large population may also be susceptible to alcohol-induced 16 / 42 P = 0.49) 1. Cell 141. accepted 21 November 2017. J. et al. 11). et al. The Fanconi anaemia pathway deficient cells by blocking resection of DNA breaks. 6. Mutations marrow failure in Japanese Fanconi anemia patients. The known molecular function of the Fanconi hematopoietic stem cell. K. D. Why does the bone marrow fail in Fanconi Fig. J. A. 14. et al. Although Trp53 deletion rescues HSC 40 defects. J. 0 0 3. J.08% LKS: 2. et al. Genetic polymorphisms of alcohol and aldehyde Figure 6 | A p53 response depletes aldehyde-damaged HSCs. Res. W. H. Nature cells (P calculated by two-sided Fisher’s exact test. 84–87 (2011). Prev. 6 | NAT U R E | VO L 0 0 0 | 0 0 m o n t h 2 0 1 7 © 2018 Macmillan Publishers Limited. et al. Arends. in two Aldh2−/−Fancd2−/−Trp53−/− HSCs. Science 329. Mechanism of RAD51- DNA. n =​ 9. 1 10 Toxicol. Quantification of Clin. & Walter.m. 53BP1 inhibits homologous recombination in Brca1- ­stability upon aldehyde exposure. 686–690 (2007). Fancd2 7 mice. The Fanconi anaemia gene FANCC promotes homologous ­protection against acetaldehyde is provided by ALDH2-mediated recombination and error-prone DNA repair. 4. more specifically. V. DNA repair and genome stability is more complex in stem cells Trp53 +/+ –/– +/+ –/– +/+ –/– +/+ +/– –/– 0 Trp53 +/+ –/– +/+ –/– +/+ –/– +/+ –/– than previously appreciated. Alcohol exposure in such individuals may cause DNA DSBs and chromosome rearrangements3. c. 11.15.m. 33.0898) e people who are deficient in ALDH2 activity. data shown as mean and s. & Patel. along with any additional Extended Data display items and 3 –/ d2 – dh p5 p5 p5 Al d2 –/ Al Tr Source Data. K. K. 8b–d for a complete list of rearrangements. H. Oncol. 9. . et al. Wang. Hira. Osawa. et al.e. 4. Genotoxic consequences of endogenous aldehydes Whitney test.

Ritsema.I. R. et al.P.. J.G.-Z. and F. and K. Br.Y. part of Springer Nature. 351–361 (2015). the Fanconi anaemia DNA repair pathway. Roe.P. Y. et al. Lejour. A. G. Med. Haematol. Bardot.L. the manuscript.. to K.. performed the M-FISH karyotyping of mouse metaphases. All rights reserved. F. Rada. were supported by CRUK. Willems. Mol. 460–463 (2015). (kjp@mrc-lmb.cam. C.G. e1005674 (2015).nature.. Correspondence and requests for materials should be addressed the ARES staff and Biomed for their help with mouse work. analysed the 26.C. 208.com/reprints. Mol. N. C. Mol. D. Sci. N. L. conceived the study and wrote repair of chromosome breaks. conducted the majority of the experiments and analysed 25. Maternal aldehyde elimination during pregnancy preserves N. Cambridge. is supported by the MRC and the Jeffrey Cheah Foundation. Cell 63.C. Schwab. & Toledo. E. et al. V. Commun.J. PLoS Genet. 20012–20017 (2011). is supported by the Wellcome Trust and King’s 23.. Proc. J. Taylor. L.. F. reveals multiple functional defects of aged murine hematopoietic stem cells.P.M. W. 807–817 (2014). J. B. Sale.J. F. The Human Research Tissue Bank (supported by 22. W. in frequency and myeloid-biased with age. J. Wyatt. the fetal genome. Daly. Nat. and M. The Fanconi anemia pathway maintains genome stability survival of chicken DT40 cells. H. Balmont for help with flow cytometry. Berks. Kawashima. Author Information Reprints and permissions information is available at www. the analysis of micronucleus samples.. et al. assisted with single cell 27. and J. García-Rubio. assisted in the characterization of genomic instability in integrity from R-loops. Essential roles for Polymerase θ​-mediated end joining in the Author Contributions J. transplants. p53 downregulates transplantation and performed the BigBlue in vivo point-mutation analysis. B. S. Gong. and L. the progression of bone marrow failure in children with idiopathic aplastic anaemia. 829–842 (2014).I.P. Aldh2−/−Fancd2−/− mice and single HSC transplantation. Cell 55. Dykstra. J. Knox.P. 24. G.S.R. interests. Wiles. Mol. The authors declare no competing financial Supplementary Information is available in the online version of the paper. . L. USA 108. G. 2691–2703 (2011). & de Haan. Clonal analysis V. 11. R. M.P. A.N. M. et al. S. Ghezraoui.uk). M.P. S. The Fanconi anemia pathway protects genome the data. by canonical nonhomologous end-joining. Pang. Wu and members of the Patel laboratory for critical J. reading of the manuscript. Olthof. Schreuder. K. Middleton. Human bone marrow hematopoietic stem cells are increased the NIHR Cambridge Biomedical Research Centre) processed histology. Publisher’s note: Springer Nature remains neutral with regard Acknowledgements We thank D. Cell 55.C..M. et al. M. Aldehyde dehydrogenase-2 polymorphism contributes to sequencing data. Natl Acad. 0 0 M o n t h 2 0 1 7 | VO L 0 0 0 | NAT U R E | 7 © 2018 Macmillan Publishers Limited. 168.I. 7. Toufektchan. X. performed validations of indels by targeted deep sequencing. 28.R. Readers are welcome to comment on the online version of the paper. W. J.G. Oberbeck. S. Romashova and M. Cell 60. 11091 (2016).. provided assistance with the analysis and interpretation of 29. performed western blotting and assisted with by coordinating replication and transcription. Jaber.G. J. G. Zhang. et al. G. S.ac. 662–673 (2016).L. Kent for technical advice with single-HSC to jurisdictional claims in published maps and institutional affiliations. Article RESEARCH 21.J. Exp. Chromosomal translocations in human cells are generated College.

05% Tween-20 (VWR) and a © 2018 Macmillan Publishers Limited. to measure full followed by a 5 min rinse in 2×​ SSC containing 0.8 g kg−1 was split between FL040. Aldh2−/−Fancd2−/−Trp53−/− mice blocked in PBS. 10:15:75 blackcurrant Ribena:ethanol:water. leaving 50 μ​l and (129S4S6/Sv) and Aldh2+/−Fancd2+/− males (C57BL/6). The Vav1-iCre allele Molecular Probes) in 2×​SSC. For single HSC transplantation experiments. 3 ml of fixative were added dropwise and the volume was made up to Fanca+/− mice (Fancatm1a(EUCOMM)Wtsi. Sigma). Thirty metaphases were captured per sample using a Zeiss LSM C57BL/6 ×​ 129S4S6/Sv hybrid background. mounted with ProLong Gold For the in vivo point-mutation assay. Likewise. Sigma) cells suspended in fixative as described above (3:1 methanol:acetic acid) were was diluted to 28% v/v in saline.05. on each occasion. 3. The slides were then dehydrated by passing them through an ethanol Fanca∆/∆ mice phenocopy the Fanca−/− mice reported previously. 1% BSA. ATTO 488-. Trp53+/− mice were then intercrossed with Aldh2−/−Fancd2+/− mice The slides were then washed three times with PBS. 0. Treated or untreated ID: 2673422. 1% BSA. Pups from these crosses were genotyped at between two and three quantification of SCEs was adapted from published protocols34. The number of sister-chromatid exchanges per to obtain Aldh2−/−Fancd2−/− BigBlue λ​LIZ and control mice. through a 70-μ​m cell strainer and incubated for 15 min in a water bath at 37 °C. we used CD45. sample size was determined M-FISH karyotyping. mice were injected with colchicine for 90% and 100%).2 antibodies (104.5% Tween-20 for 6 h at room temperature. ­ethanol-precipitated together with mouse Cot-1 DNA (Thermo Fisher Scientific). The slides were then washed three mice on a C57BL/6J ×​  129S6/Sv F1 background as recipients. PE-Cy7. dislocation. Metaphase spreads on slides were denatured by immersion in an the preparation of metaphase spreads for M-FISH. 15 min at room ­temperature with 1 μ​g ml−1 Hoechst 33258 pentahydrate (H3569. The cells were incubated at were crossed with Ku70+/−  mice (Xrcc6tm1Fwa. The the last five days. Cy5. The cells were resuspended by flicking the base of the tube To generate Fanca −/−Ku70 −/− mice on a pure C57BL/6 background. The slides were then dried for 1 h at 62 °C in a hybridization oven. The slides were then washed three times in PBS for 10 min BioLegend) and anti-CD45. 1% BSA. Cells were stained for Fancafl/flKu70+/− mice. Molecular Probes) diluted 1:2000 in PBS for 15 min at with selection at each generation by serotyping with anti-CD45. Mice were dropped onto precleaned microscope slides. Bone marrow cells the previously reported Aldh2 allele (Aldh2tm1a(EUCOMM)Wtsi. Bone marrow of 5.0 M NaCl) for 2 min. these two doses as described previously. The total dose sulfate.5% Tween-20 at room temperature. washed once in water for 5 min. using power of 0.4). denatured at 65 °C for 10 min before being applied onto the denatured slides. The slides were then transferred to a Petri dish with directs the expression of the iCre recombinase to HSCs and haematopoietic t­ issues32. Cy3-. Ethanol treatment. No randomization was employed. BD allele reported previously33 was backcrossed onto 129S6/Sv or C57BL/6J for six Biosciences) diluted 1:1 in PBS. GCCTTTGCTGCTCTAATTCCATGT. The staining of metaphase spreads for the genotypes. 1% BSA. sterility and sensitivity to mitomycin C (Fig. 10% dextran appropriate controls were injected intraperitoneally with ethanol. CD45. TCAGCTCACTGAGACGCAACCTTTT ACACT. the cell pellet in the tube. Cells were then dropped from a height of 30 cm onto chilled.8 g kg−1 was split into two injections separated by 4 h. Aldh2−/−Fancd2−/− mice were generated on a C57BL/6 ×​  129S4S6/Sv F1 bone marrow cellularity. 1×​ PBS for 3 min and dehy- For chronic ethanol treatment of Aldh2−/−Fancd2−/−Trp53−/− and control mice. part of Springer Nature. mice carrying BigBlue λ​LIZ shuttle Antifade Mountant (P36930. Aldh2−/−Fancd2−/− mice and and resuspended in a hybridization buffer (50% formamide. Unchallenged mice were injected with first crossed with FLP deletor mice31 to produce the Fanca floxed allele  (Fancafl colchicine 24 h later and metaphases were prepared after 30 min. Life Technologies) diluted 1:500 in PBS. 3% BSA.5% w/v in water.5×​ SSC at 75 °C.htm. and reconstitution of FANCA expression above. A mouse 21-colour painting the other animal experiments. Fanca+/−Ku70+/− progeny were then intercrossed to generate all possible SCE assay for mouse bone marrow. Sufficient mice generated from flow-sorted chromosomes provided by the Flow Cytometry Core to detect a 50% reduction in expected frequency were used. Metaphases were prepared as outlined GCTTCACTGAGTCTCTGGCATCTC). the drinking water supply was replaced by a solution of hybridization area was sealed with a 22 mm ×​ 22 mm coverslip and rubber cement. The M-FISH probes were ethanol was administered in drinking water for ten days as reported previously6. In brief. 1. and administered twice at 13 ml kg−1. trihydrochloride (H3570. 3). A total ethanol dose of 5. To this end. Finally. All animal The labelled products were pooled and sonicated to achieve a size range of experiments undertaken in this study were done so with the approval of the UK 200–1. respectively. 2×​SSC and exposed to UV irradiation for 30 min in a Stratalinker Crosslinker and in this case yields the Fanca-null allele (Fanca∆ or Fancatm1d(EUCOMM)Wtsi). followed by fixation in acetone exsanguinated 48 h after the second injection and peripheral blood was analysed (Sigma–Aldrich) for 10 min and dehydration through an ethanol series (70%. using the GenomePlex WGA reamp­ All animals were maintained in specific pathogen-free conditions. were flushed with 10 ml of pre-warmed hypotonic solution (75 mM KCl.35.1 (or Ptprca) times in PBS.000 bp. the previously reported Fancd2 allele (Fancd2tm1Hou.4) solution for 3 min. All rights reserved. No statistical methods were used to predetermine sample size in whole-genome amplification kit (Sigma–Aldrich). Alternatively. 0. dration through a 70%. Fig. and En2A. chromosome-specific DNA libraries were by power analysis using http://biomath. and the room temperature for 30 min and stored at −​20 °C until further use.5% Tween-20 for 1 h at room temperature and stained were also generated in a C57BL/6 ×​  129S4S6/Sv F1 background. rinsing in 1 M Tris-HCl (pH 7. these mice were crossed with Fanca+/−Ku70+/− Vav1. Five mouse-chromosome were blinded to the genotypes of mice throughout the study and data were acquired pools were each labelled with ATTO 425-. FITC. a gift from M.5 M NaOH. 96% and 100%) and placed in PBS for 5 min at room temperature. and by cardiac puncture at the end of the experiment. the Trp53 overnight with a FITC-conjugated mouse anti-BrdU antibody (Clone B44. 0.5% Tween-20 for 15 min and stained with Hoechst 33342 had been serially backcrossed from B6. followed by a 10:20:80 solution for Hybridization was carried out in a 37 °C incubator for approximately 44–48 h. Finally. For acute ethanol exposure. generations. with the micronucleus assay. For M-FISH. metaphase was then counted blind. In individual lification kit (Sigma–Aldrich) and a dNTP mixture as described previously37. For SCE analysis. Femurs were dissected for histological analysis and to determine Mice. iCre to generate Fancafl/−Ku70−/−Vav1-iCre and control mice. experiments all mice were matched for gender and age (8–12 weeks). The FANCA expression. Molecular Probes) and coverslips were lowered vector repeats (Stratagene) were crossed with Aldh2−/−Fancd2+/− mice on a onto the slides. which were finally 5 min at room temperature and stained with goat anti-mouse Alexa Fluor-488 crossed to obtain Aldh2−/−Fancd2−/−Trp53−/− and control F1 mice. very gently. femurs were harvested and placed in ice-cold PBS. tions and crossed with Fancd2+/− mice to generate Aldh2+/−Fancd2+/− mice After the incubation. . 3). The resulting mice were intercrossed 780 confocal microscope (Zeiss). 1 ml of fixative (3:1 methanol:acetic acid) was added dropwise on a 129S4S6/Sv background. 2×​SSC. and mixed by gentle inversion of the tube. C57BL/6N.07 N NaOH for 2 min at room temperature. Recombination of the Frt sites was v­ erified by with ethanol were injected intraperitoneally with ethanol 8 h and 12 h after implan- PCR (using the primers FL033. 90% and 100% ethanol series. exposure. A 50 mg BrdU weeks old. BioLegend). The slides were then Similarly to Aldh2−/−Fancd2−/− F1 mice. The tubes were were ­generated as F1 hybrids from crosses between Aldh2+/−Fancd2+/− females spun down for 10 min at 250 g and the supernatant was aspirated. EUCOMM5) 10 ml by pipetting fixative down the side of the tube. using the GenomePlex Complete and alpha 0. secondary antibody (A-11001.SJL onto 129S6/Sv for six generations. RESEARCH Article Methods blood counts. followed by injected with colchicine 12 h after the second ethanol injection. Extended Data DNA was denatured by exposure to 0. The investigators probe was prepared following the pooling strategy36. humidified was verified by western blotting spleen extracts of Fancafl/fl mice (Extended Data slides. MGI ID: 217995430). Ethanol (96%.5 M phosphate buffer. Fanca+/fl mice were then crossed with Ku70+/− mice to eventually produce were washed in 2×​SSC for 5 min at room temperature. For the analysis of Mendelian segregation of alleles.1 (A20.8 Facility of the Wellcome Trust Sanger Institute. pH 7. Fanca+/− mice were ­subcutaneously into 8-to-12-week-old mice. For the generation of Fancafl/-Ku70−/−Vav1-iCre+ tissue-­specific slow-release pellet (Innovative Research of America) was surgically implanted double mutants (also on a pure C57BL/6 background). Aldh2−/−Fancd2−/− and control mice to the hypotonic buffer. Mice ­challenged or Fancatm1c(EUCOMM)Wtsi). The sonicated DNA was Home Office. 37 °C) EUCOMM6) was backcrossed from C57BL/6N onto 129S6/Sv for five genera. room temperature. 0. The For the first five days. Grompe) was back-crossed onto the C57BL/6Jo1a young mice (8–12 weeks of age) were injected intraperitoneally with 100 μ​l of background for 10 generations and crossed with Aldh2+/− (C57BL/6N) mice to ­colchicine (0.info/power/chsq1gp. mice were alkaline denaturation solution (0. The slides were then washed three times in PBS for 5 min. Cells Fig. 0. background. (Stratagene). as judged by series (70%. MGI Preparation of mouse bone marrow for metaphase spreads.1 homozygous 0. A 50 μ​l blood sample was taken from tail veins before alcohol post-hybridization washes included a 5-min stringent wash in 0.and Texas by relying purely on identification numbers. MGI ID: 4434431. optimal for use in chromosome painting. MGI ID: 4431566. 1×​Denhardt’s solution.5% Tween-20 for to establish parental (F0) strains on both genetic backgrounds. Red-dUTPs (Jena Bioscience). After 30 min the mice were culled by cervical generate Aldh2+/−Fancd2+/− mice on a pure C57BL/6 background. tation of the BrdU pellet.

7. clone MWReg30. eBioscience). clone M1/70. Finally. Each data point represents the a­ verage of experimental caspase-3 (AlexaFluor647. BioLegend). and the bone marrow through a 70-μ​m cell strainer (Falcon). After centrifugation. slides were mounted with D7. and processed using the SmartType Karyotyper software (Digital Scientific LKS cells were defined as lineage− CD41−Sca-1+ Kit+ and HSCs were defined as UK). overnight.1+).200 r. cells were stained with the lineage-depletion kit (130-090-858. Cell Signalling) antibodies. Fcε​ R1α​(clone MAR-1. spread on ten 100-mm TB1-agar plates and incubated at FITC-conjugated lineage cocktail with antibodies against CD4 (clone H129. clone Cells were resuspended to make up 1. Gating strategies are described in Supplementary Fig. 360 μ​l of blood suspension was then added to 3. CD11c (clone N418.5.9% NaCl. 105 cells were temperature. DEAC. 50 μ​M β​-mercaptoethanol and penicillin/streptomycin). coli were resuspended in 200 μ​l of staining buffer containing the following antibody solution: diluted in TB1 top agar. 61870). RB6-8C5. 10-and-50-fold dilutions of the stock of infected E. CD150 (PE. In parallel. 4 and 8 mM acetaldehyde. eBioscience) serial dilutions were made. Thirty metaphases for each sample were karyotyped by M-FISH. These assays were performed as chimaerism was calculated using FlowJo v10. Cells (4 ×​  105) were then plated with acetaldehyde in as the proportion of cells derived from the single HSC (CD45. D1L2Z) was used at 1:1000 (Calbiochem). Each data point represents the mean of staining buffer before being resuspended in 250 μ​l of the same buffer. the RecoverEase DNA-isolation kit (Stratagene) was used to extract through 70-μ​m meshes. Sca-1 (PE-Cy7. After antibody staining.0. 1×​TBS. using standard methods. BD Pharmingen) and Ly-6G/Gr-1 (FITC. Survival is plotted (APC. Lineage-depleted cells were spun down for 5 min at 1. The mutation © 2018 Macmillan Publishers Limited. which was then used to infect E. Ly-6G/Gr-1 (clone they were lytic. M-FISH cells were washed. CD11b/Mac-1 (clone M1/70. Dilutions were plated in 6-well BD Pharmingen). The cells were incubated at 37 °C for 4 h in sealed tubes. clone A20. used for subsequent staining. The FANCA antibody (Cell Signalling. ­eBioscience).7. both at room temperature. B220+ B cells. followed by addition of 1 ml bicarbonate buffer and centrifugation. clone TER-119. Article RESEARCH 2 min rinse in 1×​ PBS. (PerCP-Cy5. blood was mixed with 338 μ​l PBS supplemented with 1. 400 μ​l of cells were then added to 4 ml of MethoCult for monocyte/granulocyte progenitors. clone H129. c-Kit (PerCP-Cy5. 16110).000 U ml−1 of heparin Western blot.1 mM Na2EDTA. TER-119 (PE-Cy7. Lymphocytes purified from the spleen using Lympholyte RBC debris and chimaerism was calculated for each of the WBC lineages (CD45+ M (Cederlane) were stimulated with lipopolysaccharide (L4391. picked and replated to confirm that BD Pharmingen).5. BioLegend) and CD48 (biotin. CD45R/B220 plates containing semi-solid medium (4000 cP methyl cellulose (M0512 Sigma).4) and 7 μ​l RNase A (Sigma) were for at least 24 h. BD Pharmingen) for erythroid maturation. BioLegend) and CD45. 7% FBS.7.2 7–10 days. 0.7. The femur samples were then decalcified and embedded in ­premixed and added to 20 μ​l of each cell suspension. Plates were incubated for clone TER-119. clone D3E9. the colonies were counted and the with either anti-p53 (AlexaFluor647. M3534 (StemCell Technologies). CD45. SlowFade Gold mounting solution containing DAPI (Thermo Fisher Scientific). Bone marrow cells were ­isolated Biotec) following the manufacturer’s instructions and passed through LS magnetic using IMDM medium. clone RA3-6B2.3 mM NaHCO3). were run on the Fortessa analyser (BD) with the HTS module and the multilineage Sensitivity assays of primary mouse B cells. 24 °C for 48 h. cells were washed. 0.2+) over the total one well of a 24-well plate.p. Flow cytometry. BD Pharmingen). BioLegend). CD8a (FITC. CD4+CD8+ T cells and Gr-1+ Mac-1+ myeloid cells) concentration of 40 μ​g ml−1. In brief.000 in the same conditions. femurs with staining buffer (PBS supplemented with 2. Samples three independent experiments each carried out in quadruplicate. clone the total number of replication-competent phage to be assessed. BD Pharmingen) for and 250 μ​l of each suspension was mixed with 250 μ​l of IMDM containing B cell progenitors.5% FCS) and strained In brief. clone 1C12. data were acquired on a digital images were captured using the SmartCapture software (Digital Scientific Fortessa FACS analyser (Becton Dickinson) and analysed with FlowJo v10. DT40 cells were grown in RPMI Medium 1640 To assess engraftment of single HSCs into irradiated recipients. Red cells were lysed by resuspending the cells in 10 ml genomic DNA from the bone marrow of 8-to-12-week old Aldh2−/−Fancd2−/− and red cell lysis buffer (MACS Miltenyi Biotec) for 10 min at room temperature. 1 ml of fixed blood cells was then The β​-actin antibody (Abcam. Tissue samples were fixed in 10% neutral-buffered ­formalin CD71 antibody (GenTex. BD Pharmingen) and CD41 (FITC. Texas Red containing streptavidin-BV421 and incubated for another 15 min at 4 °C. The plaques were enumerated. after which time colonies were counted manually.19. 10270). 1 μ​l of FITC-conjugated Histological analysis. UK). part of Springer Nature. 2. After incubation. The cells Swine anti-rabbit immunoglobulins HRP (Dako) was used as secondary antibody were resuspended in 150 μ​l of bicarbonate buffer and 20 μ​l of this suspension was at 1:2. Sigma) at a final total WBCs. The samples were incubated for 15 min at 4 °C and M-FISH images were visualized on a Zeiss AxioImager D1 fluorescent microscope washed with 2 ml buffer. The λ​select-cII (BigBlue) mutagenesis assay was with 5 μ​g ml−1 propidium iodide (Sigma). After seven days. Ly-6G/Gr-1 (PE. 8227) was used at 1:3. DT40 clonogenic survival. which includes cells derived from the recipient or carrier blue exclusion from 100 images on a Vi-Cell XR cell viability counter (Beckman cells (CD45. 5. and 4 μ​m sections were cut before staining with haematoxylin and eosin for 45 min. . Finally. 105 and 104 cells. clone R17217.0. Each data point represents the mean of three independent experiments For intracellular staining of p53 and cleaved caspase-3. Finally.1 (BV421. total bone marrow each ­carried out in quadruplicate. 3% chicken serum. 10 mM KHCO3. at 37 °C in a 5% CO2 incubator. FITC. and Cy5 fluorescence and an ORCA-EA CCD camera (Hamamatsu). BD Pharmingen). The shuttle vector that was recovered ­centrifugation. MACS Miltenyi Survival assays of colony-forming units (CFU). cells were washed with 3 ml of as a percentage relative to untreated cells. and the total volume of each dilution was plated then fixed and permeabilized with BD Cytofix/Cytoperm solution (554722. 72 μ​l of bicarbonate buffer. RBCs were Techonologies. cell viability counter (Beckman Coulter). cell pellets were resuspended in 500 μ​l bicarbonate buffer supplemented In vivo point mutation assay. TER-119 top agar were spread on two 100-mm TB1-agar plates and incubated at 37 °C for (clone Ter119. eBioscience). BD Pharmingen). Cell Signalling) or anti-cleaved- relative survival was plotted.19.2) and incubated for 10 min at room Sensitivity assays were performed as previously described6. 2. eBioscience). Pharmingen) following the manufacturer’s instructions. The cell pellets were resuspended in 200 μ​l staining buffer equipped with narrow band-pass filters for DAPI. resuspended in 500 μ​l staining buffer. Treated or untreated mice (8–12 weeks of age) were bled and 62 μ​l bio-one) and analysed on a VetABC analyser (Horiba). pH 7. Coulter). (PE. clone 2B8. BD DMEM/F-12 powder (Life Technologies. The cells were stained at 4 °C ­paraffin. eBioscience). at −​80 °C and stored at −​80°C for at least 12 h.1% Tween-20 at 4 °C with gentle shaking. duplicates carried out on three mice of each genotype. Pharmingen). stained with antibodies against committed lineages: CD45R/B220 (PE. CD45R/ To determine the total number of phage undergoing the lysogenic cycle and the B220 (clone RA3-6B2.0. Incubation at 37 °C switches all phage to the lytic cycle. clone II/41. supplemented with 7% fetal bovine serum (FBS. BioLegend). clone 53-6.1. This assay allows the detec- on an LSRII FACS analyser (BD) and the data analysed with FlowJo v10.5 ml of IMDM containing 30 ×​  106 cells RA3-6B2. 3% chicken serum (Life Technologies. 3 ×​  106 total bone marrow cells were Cell Technologies) using a Vi-Cell XR cell viability counter (Beckman Coulter). bone marrow cells were isolated from tibiae and either through the lytic or lysogeneic cycles in Escherichia coli. TER-119 (FITC. coli G1250 (Stratagene). Nucleated cells were counted pellets were resuspended in 200 μ​l MACS buffer with the antibodies described by diluting cells tenfold in a 3% solution of acetic acid with methylene blue (Stem above for HSC quantification. and therefore allows Pharmigen). and CD11b/Mac-1 (PE. Life was obtained from the tail vein of recipient mice every two weeks.m. Total blood was collected in K3EDTA MiniCollect tubes (Greiner previously38. β​-mercaptoethanol and penicillin/streptomycin. BD Pharmingen).6 ml of methanol in 5% w/v BSA. The micronucleus assay was performed essentially as described Blood counts. the cell pellets were resuspended in 100 μ​l incubated with drug-containing medium in a sealed FACS tube at 37 °C for 2 h of staining buffer containing antibodies against: CD4 (FITC. clone 1A8. clone HM48-1. coli in TB1 CD8a (clone 53-6. BD Pharmingen) and IgM (FITC. Cy3.7. after which two tenfold clone M1/70. cells were stained After seven days of culture at 37 °C with 5% CO2. infected E. CD11b/Mac-1 (PE. 32500-043). This provided the number of mutated phage within the sample. 1. washed with 6 ml of bicarbonate buffer (0. Bone marrow cells (10 ×​  106 cells) were To assess the frequency of mutations within the cII gene. clone 1A8. LKS CD48− CD150+. and single cell suspensions were obtained by passing the ­columns. The samples were analysed immediately performed following the manufacturer’s instructions. Finally. clone C2. respectively. All rights reserved.2. 50 μ​M lysed by the addition of 1 ml of ammonium chloride lysis buffer (155 mM NH4Cl. the cell pellet was resuspended in staining buffer and n ­ ucleated from mouse genomic DNA was packaged into phage with the Transpack packaging cells were counted with 3% acetic acid (StemCell Technologies) on a Vi-Cell XR extract (Stratagene).000 for 1 h at room temperature. clone 104. TER-119 was used to exclude described previously5. 50 μ​l of blood (Life Technologies. viable cells were counted using trypan number of CD45+ cells. After control mice that carry Big Blue λ​LIZ repeats. BD 24 h. (acetaldehyde) or 1 h (mitomycin C or cisplatin). BD in two wells of a six-well plate each containing 106. BD Pharmingen) and CD71 2×​acetaldehyde to give final concentrations of 0. BD Pharmingen). clone TC15-12F12. eBioscience). BioLegend). CD3e (clone 145-2C11. tion of point mutations in vivo and is based on the ability of coliphage λ​to multiply For HSC quantification. mutation frequency.

Primers were designed using MPprimer44 to 1. we asked that all variants were unique fixed in Bouin’s solution (Sigma). together with their matched germ-line references.200 r. The reactions were diluted to 1 in 50 and 1 μ​l of the avoiding the creation of bubbles. followed by 23 cycles of kit (130-090-858. would not have arisen after the transplant). Lineage-depleted cells were spun pooled. 0.8%) calls had VAF values <​0. CD45. 5l) and the number of reads involved in the rearrangement at (Qiagen) following the manufacturer’s instructions.5% then 72 °C 60 s. The full content of selected wells or tails) or 400 ng (donor bone marrow) of genomic DNA and GoTaq G2 Hot was loaded into insulin syringes (29G. indels and structural changes were called with the CaVEMan. We did not apply a VAF filter for clonality because when we To assess the survival of CFU-S12 after exposure to acetaldehyde. In brief. Validation of indel and rearrangement calls. Therefore. All tissues were prepared for flow cytometry as described above and be clonal (that is. CFU-S12 assays were performed as described previously5. BSA. 60 °C for 60 s. Sample number (n) indicates the number of independent short-insert 500-bp genomic libraries were constructed according to Illumina biological samples in each experiment. With this approach. another 15-ml tube and topped up with ice-cold MACS buffer (PBS pH 7. as well as the sensitivity of each PCR. and resuspended in 6 ml IMDM medium at room Group 1. Short-insert tested using the D’Agostino–Pearson omnibus normality test and variance was paired-end reads were aligned to the reference mouse genome (NCBIM38) using estimated before deciding on a statistical test.net/). tailed gene ­primers and was individually barcoded by a second round of PCR with ground (CD45. noted a big overlap between the variants found at 20×​and 40×​coverage (data not peritoneal injection on one occasion. these were centered around total bone marrow cells with 4 mM acetaldehyde for 4 h in vitro before injecting 0. . 95 °C for 2 min. Survival was expressed relative to the untreated control for each bowtie2/index. and were deemed false positives. using a high fidelity polymerase (Q5 Hot extracted with IMDM medium (GIBCO). We exposed to 150 mg kg−1 N-ethyl-N-nitrosourea (ENU. to assess the frequency of that 20×​whole-genome sequencing provided sufficient coverage to allow us to CFU-S in mutant mice. post-transplant) events.8) and quantified before being stored at −​20  °C down for 5 min at 1. Recipient mice that were positive for reconstitution were euthanized detected in the bone marrow of donor mice at the time the HSCs were transplanted four months after transplantation.2. for 30 s.400g at room temperature with 60 s. 10 ng DNA input in 5 μ​l PCR reaction. spleen and thymus (Extended Data Fig. clone 104. defined as lineage−c-Kit+Sca-1+CD41−CD48−CD150+.net/ were counted. the reaction was then held at 4 °C until addition of barcoded second-round the brake off. The interface containing the mononucleated cells was transferred into primers. All rights reserved. 3 h apart) using a 137Cs GSR C1m source (GSR Gmbh). 0. bands of the expected sizes were Whole-genome sequencing of mouse HSC clones. The the coverage was higher than 100×​(159/172). Any additional calls found when the coverage was increased to 40×​ CFU-S12 assays. previously reported them into lethally irradiated recipient mice. the breakpoint for copy number-neutral changes. and the blood.m. 100 ng DNA input in 25 μ​l PCR reaction. BioLegend). New England Biolabs).1 (Sony Biotechnology Inc. showing that doubling the coverage did not actually uncover many more ­euthanized 3 weeks later. then 72 °C for 5 min. size selected by SPRI (×​0.40. and the overlap between indels and ­genotype. to increase specificity and sensitivity. 58 °C for 17 min and 70 °C for 60 s.sourceforge. 50 °C for 60 s and 70 °C for a 15-ml Falcon tube and spun down for 20 min at 1. a solution of 3% acetic acid and methylene blue and injected intravenously into Pindel and BRASS algorithms. Nucleated cells were enumerated using Substitutions. with a mean VAF of 0. The cell if chimaerism for CD45.2+ cells were sorted from each tissue on a Synergy (that is. we found that 16/27 (59%) of rearrangements could be stem cell. 10). The PCR reaction conditions were as follows: down for 5 min at 1.1 (FITC. part of Springer Nature.5 is Single HSCs. Whole-genome sequencing was performed as described previously41. see Fig. 35 using flow cytometry. and the number of colonies were counted and to each HSC clone and not found in unrelated HSC or tail samples. or in ­unrelated expressed relative to the number of total bone marrow cells injected. The failure to amplify stained with CD45. After 5 min of centrifugation at 1. then 72 °C 60 s. described previously39. The lethal radiation was delivered as a split dose of of the previously identified indels. 55 °C for 60 s. 65 °C for 60 s. except were predominantly subclonal. Mice were allowed to recover and were shown). CD45.5 (Extended Data Fig. reactions were analysed on 2% agarose gels. clone A20.p.2.000 rad (500 rad each. 98 °C for 20 s.22.m. we treated ­examined the VAF distribution of the final datasets.m. We sequenced two of the Aldh2−/−Fancd2−/− As a positive control for the in vivo λ​ select-cII mutagenesis assay. 60 °C for 20 s and 72 °C for 30 s. mouse strains.200 r. Sigma) in a single intra. 2 mM EDTA). Each sample was tions and passed through LS magnetic columns. to expand in vivo for four months to guarantee that the transplanted cell was a Using this approach. spun Start HF. Therefore. 35 cycles of 95 °C for 30 s. stained with the l­ ineage-depletion 4 °C until addition of barcoded second-round primers. we designed nested PCRs surrounding the well of a round bottom. 62 °C for 15 s and 72 °C for 30 s. MACS Miltenyi Biotec) following the manufacturer’s instruc. filtered through a 70-μ​m strainer. Recipients were considered reconstituted by the single stem cycles of 95 °C for 30 s. covered by genes. Indel calls of less than 50 bp 8-to-12-weeks old) were fed with water supplemented with the antibiotic enro. performed at 95 °C for 2 min. then held at were resuspended in 320 μ​l of MACS buffer. Normality of data distribution was 2000 or HiSeq X genome analysers to an average of 20×​coverage. 95 °C for 2 min. mice were HSC clones to 40×​coverage. bone marrow. The normal distribution of VAFs around 0. RESEARCH Article frequency was then calculated as the number of cII-mutated plaques (24 °C)/total BWA-MEM (http://bio-bwa. Genomic DNA was extracted with the Puregene Cell and Tissue kit deletions. Bayer Corporation) for seven days before irradiation and for the ­multiplex-primer combinations to capture and simultaneously amplify 172 (50%) duration of the experiment. To calculate the fraction of the genome mice. except that ­confirmation. Unless otherwise stated in the © 2018 Macmillan Publishers Limited. Promega). total bone marrow was flushed from the femora and tibiae uncover mutations present in the transplanted HSC. Statistical analysis. Genomic DNA transplanted HSC was contributing to blood production in the donor animal at the was then extracted from the CD45. were validated using multiplex PCR and targeted re-sequencing. PCR amplification will depend on how much the cell sorter (Sony Biotechnology).sourceforge.2+ bone-marrow cells and from a tail biopsy time of the transplant.p. Each data point represents the mean CFU-S survival in ten recipient transcribed genes was calculated in R. After 12 days the spleens were ­addition to previously reported filtering41. The plates were spun down for 5 min at 180g and the presence upon request). 62 °C for 15 s and 72 °C for 30 s. BioLegend) and CD45. the cell pellets seven cycles of 95 °C for 20 s.5-inch needle) containing 2 ×​  105 carrier Start Polymerase (M7401. spun down and frozen at −​80 °C. six cycles of temperature. In brief..2+ WBCs was ≥​  0. tail sample was used as the reference. mutations. 8-to-12 weeks old). The first of round multiplex-PCR amplifications was performed with Aldh2−/−Fancd2−/− and control mice on a C57BL/6Jo1a ×​ 129S4S6/Sv F1 back. Overlapping Single HSC transplants.2% of the original calls: 14/159 (8. For the transcriptional analysis. which are described in detail elsewhere42. We used 13 floxacin (Baytril. The contents ­performed at 95 °C for 2 min. The single-stem-cell transplants were performed as regions between two genes were taken into account and counted only once. the number of CFU-S HSC RNA-seq data was aligned with Bowtie2 (http://bowtie-bio. positions of genes were retrieved from Ensembl. Group 2. of mutant mice and appropriate controls. followed by 19 cycles of 98 °C for 20 s.shtml) against NCBIM3843. After 12 days. 55 °C for 20 s and 72 °C of the syringe were used to dislodge the single HSC from the bottom of the well. 96-well plate (Costar) using a Synergy cell sorter (Sony breakpoints determined by the BRASS algorithm (primer sequences available Biotechnology). we concluded that CFU-S colonies were counted after 12 days. PCR reactions were carried out in 20 μ​l. and the pellets were resuspended in 200 μ​l MACS until ­sequencing. sorted into 100 μ​l StemSpan SFEM medium (StemCell Technologies) in each For validation of rearrangements. The entire volume (400 μ​l) was injected into the diluted reaction was used as template for a second round of PCR with nested tail veins of irradiated recipients and chimaerism was measured every 2 weeks ­primers. Single HSCs were allowed excised and the identities of all products were confirmed by Sanger sequencing. We analysed positions where anti-CD48 was directly conjugated to BV421 (clone HM48-1. the remaining 11 rearrangements by PCR does not mean that these are sub-clonal BioLegend) antibodies. Any rearrangements present before transplantation must were collected. we inferred that had been collected at 2 weeks of age from the same mouse that provided the clonality for the remaining calls by looking at loss of copy number (in the case of single HSC. expressed relative to untreated samples of the corresponding genotype. The first round of PCR amplifications was cells in 300 μ​l Hank’s balanced salt solution (StemCell technlogies). using 10 ng (HSC clones of a single cell per well was confirmed visually.1%. 9).p. were consistent with most indels being clonal in origin. Sample numbers and experimental repeats library protocols and 100-base paired-end sequencing was performed on HiSeq are indicated in figure legends or Methods. the matched number of plaques (37 °C).). Bone marrow cells from these mice were pre-­validated MiSeq-ready primers45.200 r. and reads were mapped with BWA. capture regions of between 190 and 250 bp in size (sequences available upon The lethally irradiated recipients were injected with single HSCs sorted from request). In 20 recipient mice that had been lethally irradiated.1 recipient mice (C57BL/6J ×​  129S6/Sv F1.2 (APC. For each HSC clone. then 72 °C for 5 min. we were able cells were resuspended in 500 μ​l of MACS buffer and run on a Synergy sorter to validate 91. These cells were then overlaid onto 6 ml Lympholyte M (Cederlane) in 98 °C for 20 s. Two MiSeq runs (300-bp paired-end) were used for variant buffer with the antibodies described previously (see Flow cytometry).

© 2018 Macmillan Publishers Limited. T. Krishna. B. Genome sequencing of normal cells reveals developmental (2000). Protoc.. Z. Nat. S. Multiplex-fluorescence in situ aneugenicity and toxicity in the mouse micronucleus assay using hybridization for chromosome karyotyping. 33. Widespread progenitors. et al. Protoc. 1172–1184 (2006). Geigl. et al. 35. Methods Mol. Mutagenesis 30. All rights reserved. Transgenic mice with hematopoietic and lymphoid specific 42. 1005–1010 (2008). Mol. Stem Cell Biol. M. M. and the two-sided nonpara. Soriano. Oguro. S. Identification of regulatory networks in HSCs and susceptible to spontaneous tumours. P. Nature 356. et al. Mutat. H. N.m. 87–107 (2004). 262. model of Down syndrome. Isolation and assessment of long-term reconstituting hematopoietic stem cells from adult mouse bone marrow. R. D. Massively parallel sequencing reveals the complex structure metric Mann–Whitney test was used to assess statistical significance. A large genome center’s improvements to the Illumina development and genome maintenance in the absence of poly(ADP)ribosyl sequencing system. Schimenti. J. 33. Differential requirement for H2AX and 53BP1 in organismal 45. Modulation of mitomycin C-induced sister chromatid methylome analysis. . F. Whole-genome sequencing data have been deposited in the meiotic and mitotic recombination-defective mutants in mice. et al. S. and DNA 34. J. Cell Stem Cell 13. Behjati. Forward genetic screens for Data availability. Article RESEARCH f­ igure legend.. Nature 513. 159–167 (1992). A. 36. Methods 5. (2007). part of Springer Nature. J. Cabezas-Wallscheid. C. M. Cell. expression of Cre. 413. D.. Fiedler. B. e60482 (2013). 2341–2352 (2010). Biol. Curr. Y. B. The genome as a record of environmental exposure. L. et al. 38. Steffen. & Chatterjee. Res. A. 215–221 (1992). Uhrig. & Theiss. Simultaneous evaluation of clastogenicity. Donehower.. 507–522 (2014). C. 30. in mouse bone marrow cells in vivo. S. & Speicher. Kent. lineages and mutational processes. W. data are shown as the mean ±​ s. polymerase 1. 32. et al. 30. immunofluorescence. recombinase expression using FLPeR (flipper) mice. Cell Stem Cell 15. et al. Growth retardation and leaky SCID phenotype of Ku70-deficient 40. Quail. J.. & Dymecki. SLAM family markers resolve functionally mice. Mice deficient for p53 are developmentally normal but 43. 422–425 (2014). All other data are available upon reasonable request from the authors. Ashley. 763–770 (2015). S. Orsburn. Immunity 7. de Boer. J. L. 282. A. Dykstra. exchanges and cell cycle delay by buthionine sulfoximine and reduced glutathione 44. transcriptome. L. R. Reinholdt. S. J. their immediate progeny via integrated proteome. MPprimer: a program for reliable multiplex PCR primer design. Res. Ding. Genesis 28. distinct subpopulations of hematopoietic stem cells and multipotent 31. S. Giri.e. Immunol. et al. Farley. et al. Gu. Nik-Zainal. Analysis was of an irradiated human chromosome on a mouse background in the Tc1 performed using GraphPad Prism. & Shima. BMC Bioinformatics 11. Biol. 39. N. M. et al. 314–325 (2003).. 227–234 (1998). 106–110 41. PLoS ONE 8. Shen. 37. S. 143 (2010). G... 653–665 (1997). 46. 102–116 (2013). EMBL European Nucleotide Archive (ENA) under the accession code ERP009447. Eur. & Morrison. & Eaves. 1. Nat. Mutat. L. Gribble.

Dp(4).15) 40.XY. Chtb(7).fold 3. Chtb(16) 2/30 40. Del(5).XX. Del(2).XY. Chtb(14) 40. Chtb(7) 40.XY. Chtb(X) 40.XX. Del(1).XX. XY.XX. . Der(16)T(1. Der(14)T(10.X) 40.XY. 4/30 40. + . Del(10). Dp(2) 40. Chtb(9) 40.10).XY.13) 40.15) 40.XY.XX.XX.Chtb(14) 40. Chtb(1) 39. Chtb(19) 40. + . Chtb(1) 40.14). +Del(2). Chtb(15) 40. Der(12)T(6.XY. (n = 8) .6) 40.XY. Chtb(12) 40.7)x2 39. Del(2) 40. +T(3. 0/30 Aldh2 -/. Der(8)T(8. Chtb(X). Del(X). Del(X) 40.Del(4).0001 1.Der(3)T(3.XX. Chtb(2) aberrations in 90 metaphases 100 Translocation 40.XY. dic(18. Chtb(10) 25 40. Chtb(15) 17/30 40.XY.XY.XX. + .18). Del(5).XY. -9. Chtb(12).XY. Chtb(13) 40. Chtb(9) 40.7).T(7. Chte(9.16). Del(2). -6.5).XY. Der(6)T(1.22/30 40.0001 P < 0. Chtb(4) 1/30 40. Chtb(8).XX.XX. +11 40. -2. Del(6). Del(5).XX.XY.XX.12) 40. Del(2) Deletion 40. Dp(16) 40.XX. Del(14). Der(10.XX.17) 75 Chromatid exchange Number of 40.2-fold e 5 10. +15. Der(2)T(2. Del(13).XX. Chte(2. Aldh2-/- Fancd2-/- i Balanced Unbalanced Chromatid Chromatid Complex Polyploidy Deletions exchange translocations translocations breaks rearrangements Aneuploidy Extended Data Figure 1 | See next page for caption.12) 40. -6.2-fold.-Y 0/30 0/30 0/30 Aldh2 -/. Chtb(3) 40.XX. Der(2)T(X. Fancd2-/. All rights reserved.XY. Der(13)T(2.0016 4 15 1. T(X. -11.XY.XY. Der(5)T(1.XY.Fancd2 -/.0001 P < 0. Del(2).18) Wild type Aldh2-/.12) 40.0001 P < 0.XX.XX.3 . Del(2). -12.13) 40.XY.Del(17). Der(12)T(9.XX. Chtb(X) 40. Chtb(2). P < 0.XX.XX.2).XX. +12 2/30 40. 0/30 0/30 0/30 2/30 40.-Y 0/30 39.11) 40.4)x1 2/30 41. Chtb(5) 41.XY.7/30 40. Der(7)T(4.17). RESEARCH Article a 40 4-fold P < 0.XY.+Del(15) 39.X. -17. Chtb(10) 15/30 39.0001 b 20 c Red cell progenitor d 2.X.0 2 10 5 Mn-Ret 0. Chtb(6) 40. Der(8)T(8. Del(1) 40.XX. 0/30 Fancd2 -/.19).XX.32) SCEs / metaphase 3 Wild type (n = 5) 20 Wild type (n = 5) 10 Nucleus extrusion 1./- Fa pe pe /- 2- W ty 2- d2 dh ty ty dh ild nc Al ild ild Al W Fa W W f Untreated g 48 hours after Ethanol IP (5.+9. Del(4) 2/30 40.5).14) 40. Der(17)T(16.0002 10. Chtb(12) 40. -1.0001 Frequency of Mn-Ret (%) Frequency of Mn-Ret (%) 30 P = 0.+14.XY.XY.XY.19) 40. Del(6). + . Chtb(9) 39.XY. Del(7).XX.0 0 Wild type + MMC Fanca-/. Chtb(9). Chtb(6) …Chtb(1). +Del(17). Chtb(6) 0 40. Chtb(14) Chromatid break 40. + 40. Chtb(4) 40. Chtb(4) 40.18) h 125 Duplication 40.+Del(8). Chtb(4) 40. -10. Del(2) 1/30 41.XX.XX.18)x2. Chtb(15). Del(10).XY. Del(7). Chte(11.XY. Chtb(13) 40.XX. Dp(2). -1.XY. Del(14).XY.fold P < 0.14) 40.XY.10).XX.15). Del(13) Fancd2 -/. Del(12). Chtb(3). Chtb(1).Y.XY.7). Del(6) Aneuploidy 41. Der(1)T(1. EtOH EtOH Vcn IR EtOH (n = 18) MMC .XY.2) Aldh2 -/-Fancd2 -/.Chtb(1) 40. Rb(12./- Mn-NCE pe NCE a ty DNA+ nc ild pe /- . -19 Ethanol . Der(13)T(2.XY. Chtb(1).XX.XX.XY.XY. part of Springer Nature.… 50 40. Der(3. Chte(13. Del(7). Chtb(17) 1/30 40. Der(X)T(X. © 2018 Macmillan Publishers Limited.XY. Chtb(5) 40.XX. Chte(4.5 SCEs / metaphase ns (P = 0. T(12.XX.XX. EtOH MMC . Chtb(2) 41.16) 40.11) 40. Chtb(16) 40. Chtb(2) 40.17). Del(5). -8. + .XY. Der(X)T(6. Der(15)t(7.XX. Del(13) 40. +7 Aldh2 -/.8 g/kg) Wild type 0/30 Wild type 2/30 39.0 3.17).19) 40. Del(X). Der(7)t(7. Del(X). -19. Del(16) 39.8 . Der(18)T(6.Der(8)T(8.… 40. Dp(16). Chte(5.XY.XX.T(8. Chtb(2). Del(12) …Der(10)T(3.5 CD71+ 1 Ret DNA+ CD71+ 0 0 0. Del(6) 40. - .XX.6-fold P = 0.

400 rad. comparable to to potent micronucleus formation. were analysed per condition. and s. d. part of Springer Nature. 8. Article RESEARCH Extended Data Figure 1 | Ethanol-induced genomic instability. f.8 g kg−1) or MMC (1 mg kg−1). red-cell progenitors lose CD71 a. three mice and 30 metaphases per mouse c. P calculated by two-sided Mann–Whitney test. Examples after extrusion of the main nucleus.m.. IP). i.). data shown as mean quantification of SCEs (duplicate experiments. All rights reserved. n =​ 29. 48 h)46. Left. These fragments can be detected by of different types of chromosomal aberrations. g. h. b.m. 11.m.8 g kg−1) leads homologous recombination response in Aldh2−/− mice. Therefore. This induction is comparable to that that observed in wild-type mice exposed to MMC. e. and the numbers represent the fraction Micronuclei (Mn) generated by fragmentation or mis-segregation of of abnormal metaphases per mouse. List of chromosomal aberrations Fancd2−/− in Fig. indicating that a homologous recombination and control mice 48 h after ethanol treatment (5. intraperitoneally. Left. P calculated by two-sided Mann–Whitney test.2 mg kg−1. Whitney test. Right. n shows the number of SCE events per metaphase. © 2018 Macmillan Publishers Limited.e. Ethanol causes a strong left to right. 0. data shown as mean and s. P calculated by two-sided Mann– Whitney test. Right. Triplicate experiments. Scheme depicting the formation of micronucleated erythrocytes. Proof-of-principle of SCEs in the bone marrow of wild-type and Aldh2−/− mice treated with experiment showing the induction of micronucleated reticulocytes 48 h ethanol (5. 48 h) or clastogenic γ​-irradiation (IR.e. 10. injected repair response occurs in the absence of the Fanconi anaemia pathway. . or Aldh2−/−Fancd2−/− and control mice.e. n =​  75. List of chromosomal n =​  50. Treatment of Aldh2−/− mice with ethanol (5. 20 and 9 mice. a DNA stain (PI+). data shown aberrations observed in the bone marrow of 8-to-12-week-old untreated as mean and s. 25 and 15 mice. wild-type mice treated with mitomycin C (MMC). after MMC treatment (1 mg kg−1).e. In f and g. 15. representative images of bone marrow metaphase spreads from expression.. n =​ 29. During maturation. n shows the number short-lived reticulocytes (Ret) and CD71− cells represent mature. P calculated by two-sided Mann– 25 metaphases per mouse.m. Bar chart classifying the type of chromosomes during erythrocyte maturation remain in the erythrocyte aberrations for each genotype (90 metaphases per condition). comparison between number long-lived normochromic erythrocytes (NCEs). 1a) show a small but significant increase in the number observed in the bone marrow of 8-to-12-week-old Aldh2−/−Fancd2−/− of spontaneous SCE events. peripheral CD71+ red cells represent immature. 25 metaphases per mouse. representative observed in wild-type mice that were treated with the aneugen vincristine images of bone marrow metaphase spreads from wild-type and (Vcn.8 g kg−1. data shown as mean and s. of SCE events per metaphase. Fanca−/− mice. Mice deficient in cross-link repair (Fanca−/−.

+ . b. Fancd2-/- NS (P = 0. + .Aldh2-/. + .(n = 5) Survival (%) 60 Fancd2-/. by Mantel-Cox test. as in a).. before injection and terminal bleeds after ethanol Aldh2−/−Fancd2−/− mice one to two months after treatment (P calculated treatment (P calculated by paired t-test.e. + . + . + Ethanol Wild type Aldh2-/.Fancd2-/. data shown as mean and s. + . Wild type Aldh2-/. Haematoxylin and eosin n =​ number of mice. Fancd2-/- Extended Data Figure 2 | A single dose of ethanol precipitates staining of bone marrow sections 30 days after ethanol treatment (original bone-marrow failure in Aldh2−/−Fancd2−/− mice.(n = 7) Aldh2-/. + .0001 Mantel-Cox test 0 0 30 60 90 120 150 180 Time after ethanol IP (days) c 80 65 15 Mean Corpuscular Volume (μm3) NS P < 0.38) 1500 10 1000 5 500 0 0 . A single dose magnification. + . + . + .Aldh2-/. c. Wild type Aldh2-/. Fancd2-/. + Ethanol . + Ethanol .0143 (P = 0.62) Red blood cells (106/mm3) 60 60 Haematocrit (%) 10 55 40 50 5 20 45 0 0 40 . + .Fancd2-/.Aldh2-/- Fancd2-/. ×​100). injected intraperitoneally) leads to anaemia in and control mice. a. + . n =​ number of mice).Fancd2-/.Fancd2-/. part of Springer Nature.8 g kg−1. + Ethanol Wild type Aldh2-/. + . Wild type Aldh2-/. Aldh2-/. © 2018 Macmillan Publishers Limited.Fancd2-/. + . + Ethanol . .Aldh2-/- Fancd2-/.8 g/kg) b Wild type Aldh2-/- 100 Wild type (n = 7) 80 Aldh2-/. Full blood-count analysis for Aldh2−/−Fancd2−/− of ethanol (5.Aldh2-/. RESEARCH Article a Ethanol IP (5.m. All rights reserved. + .Fancd2-/- 40 20 P < 0.Fancd2-/.(n = 7) Fancd2-/.0001 P = 0.79) 15 2000 White blood cells (103/mm3) Platelets (103/mm3) NS (P = 0.

s. showing hypersensitivity of both Fanca−/− and FancaΔ/Δ cells and Fancafl/− Vav1-iCre mice. ×​50) from yields the FancaΔ allele (Fancatm1d(EUCOMM)Wtsi). PCR. fl/fl fl/- FL033 En2A KO (-) band: 692 bp FL040 or tm1a(EUCOMM)Wtsi iCre - loxP loxP loxP 1 2 En2A IRES lacZ NeoR 3 45 6 iCre + FRT FRT FANCA FLP recombinase Fanca fl allele β-ACTIN FL033 FL040 Flox (F) band: 965 bp or tm1c(EUCOMM)Wtsi loxP loxP 1 2 3 45 6 7 8 9 d Fanca exon 3 Taqman 2 Exon 3 Copy number exon3 Cre recombinase Fanca ∆ allele loxP ∆/+ loxP or tm1d(EUCOMM)Wtsi Cre 1 1 2 45 6 7 8 9 loxP ∆/∆ 0 nc /+ s BM s BM r r nc /∆ ∆ Ea Ea BC BC ∆/ Fa a + Fa a + W W a e Wild type Fanca -/- nc Fanca fl/+ Fanca fl/- Fa Vav1-iCre Vav1-iCre f 100 Fanca fl/fl (n = 6) 80 Survival (%) Fanca fl/fl Fanca ∆/∆ 60 40 20 Fanca -/. 3). Fancafl/fl and FancaΔ/Δ showing complete absence of FANCA protein in the spleens of Fanca−/− embryos. Fanca+/Δ and FancaΔ/Δ mice carry 2.m.). part of Springer Nature. Western blot (single experiment) fibroblasts (MEFs) derived from Fanca−/−. s. which lacks exon 3 and wild-type. Fancafl Vav1-iCre mice show tissue-specific deletion of were crossed with mice carrying the FLP recombinase. Sensitivity assay of transformed mouse-embryonic showing bands of the expected sizes.(n = 6) Fanca ∆/∆ (n = 2) 0 0 10 20 30 40 Mitomycin C (ng/ml) Extended Data Figure 3 | Generation of a conditional Fanca allele. fl/+ fl/- a FL033 FL040 WT (+) band: 800 bp 1000 bp 1 2 3 45 6 7 8 9 800 bp 700 bp Fanca . see Supplementary to the cross-linking agent mitomycin C (n =​ number of experiments. c. Determination of the number of exon 3 copies by quantitative carried out in quadruplicate. e. 1 and 0 copies. For gel source data. yielding the Fancafl exon 3 in white blood cells (WBCs) and bone marrow (n =​  4 technical allele (Fancatm1c(EUCOMM)Wtsi). All rights reserved. (one experiment). b. Mice carrying the previously reported Fanca− allele (Fancatm1a(EUCOMM)Wtsi) respectively. impaired spermatogenesis in testes of Fanca−/− and FancaΔ/Δ mice type. Genotyping PCRs for the wild. Microscopic analysis of haematoxylin shown by western blot (Fig. Fanca− and Fancafl alleles with primers FL033. Cre-mediated recombination of Fancafl and eosin-stained sections of testes (original magnification.allele c Fanca +/+ -/.e. © 2018 Macmillan Publishers Limited. This allele restores FANCA expression as replicates. -/. bars: mean. Article RESEARCH Fanca + allele b Fanca +/+ +/. 3). FL040 and En2A. bars: mean. showing leads to loss of FANCA protein (Fig. Fanca−/−. 1. f. d. each Fig. Fancafl/fl and FancaΔ/Δ males at 12 weeks.).d. . a. Wild-type.

Circos plots showing the mutations observed in all sequenced HSC clones (wild type. RESEARCH Article Extended Data Figure 4 | Endogenous aldehydes mutate the HSC genome. . © 2018 Macmillan Publishers Limited. All rights reserved. n =​  3. indels and rearrangements are plotted. Substitutions. n =​ 5 HSC genomes). and Aldh2−/−Fancd2−/−. n =​  3. part of Springer Nature. Fancd2−/−. n =​  4. Aldh2−/−.

mutant plaques Incubate at 37°C All λ plaques c Wild type Aldh2-/.m.mutant plaques/106 plaques) 300 Mutation frequency Extract DNA NS 200 (P = 0. a. of cII− -mutant phage recovered from the bone marrow of young n is the number of sequenced cII− mutant phage. coli G1250 U ild /- + pe Al -/- /- d - pe W 2- nc 2-/ EN 2- 2 ty ty cd dh Fa dh ild n Al Fa W Incubate at 24°C λ cII. Aldh2-/-Fancd2-/. 6. n =​ 7.2 T>C T>G 0. c.8 C>A Relative contribution n = 66 n = 52 n = 33 n = 45 n = 113 0. packaged into phage and of the indicated mutation classes to the point-mutation spectra of used to infect bacteria. Quantification of the frequency spectrum is characterized by T to A transversions and T to  C transitions.4 T>A 0. The ENU-mutation these phage to form plaques at 24 °C. b. Wild type + ENU 0. part of Springer Nature. All rights reserved.6 C>G C>T 0. 7. Relative contribution phage DNA can be recovered from mouse tissues.0 Extended Data Figure 5 | Detection of point mutations in mice with the ENU-treated mice serve as positive controls for the assay BigBlue reporter system. Phage cII mutants can be detected by the ability of cII−-mutant phage isolated from the bone marrow. 7 and 6 mice. .e.0012 bone marrow (cII. left to right). Article RESEARCH a Harvest BigBlue b 400 P = 0. Aldh2−/−Fancd2−/− and control mice carrying the BigBlue transgene.45) (P = 0. data shown as mean harbours a λ​-phage transgene that contains the mutational target.. The and s.53) NS NS Package (P = 0.32) phage DNA 100 Infect 0 E. Chromosome 4 of the BigBlue reporter mouse (P calculated by two-sided Mann–Whitney test. © 2018 Macmillan Publishers Limited. Fancd2-/.

b.m. All rights reserved.Fancd2 -/.5 Fancd2 -/.60% 0.Fancd2-/.01 % 12. RESEARCH Article a Wild type Wild type. f. 2 hours..5 20 1. . Quantification of the frequency of of Aldh2−/−Fancd2−/−Trp53−/− and control mice. 2. data shown as mean and s.(n = 3) 30 15 p53 + cells (%) 40 2.0 10 5 10 0. the number of p53+ or cleaved-caspase-3+ HSCs © 2018 Macmillan Publishers Limited.0 Wild type (n = 3) 10 Gy. Trp53 -/- Fancd2 -/.+/+ +/.(n = 5) 10 Gy. a. e. Cells were collected from wild-type and Trp53−/− mice 2 h after Fancd2−/−Trp53−/− and control mice. Quantification of the frequency of p53+ cells in different shown as mean and s.m. c. Trp53 -/. respectively.+/+ -/.3 ± 0. 2 hours Aldh2-/.e. each carried out in quadruplicate. n =​  number response.(n = 3) Aldh2 -/. In b and c.(n = 3) Trp53 -/. Cells were obtained from mice. Owing to the low numbers of LKS CD48− CD150+ cells in mice). data the assay.5 0 0 0 0. 2 hours Trp53-/-. Each point represents the mean of 10 Gy irradiation as positive and negative controls. Fancd2 -/. 10Gy. n =​  10–15 as controls.Fancd2 -/.e.Fancd2-/- 250K 10 5 Lineage. 10Gy. Fancd2 -/- Fancd2 -/. part of Springer Nature. 2 hours (n = 3) Aldh2 -/.0 20 10 30 1.+/+ -/.(n = 3) Cleaved caspase-3 + cells (%) 10 Gy.55% 4 LKS 5. Each point represents cleaved-caspase-3+ cells in different bone marrow populations by flow the number of CFU-S12 in the spleen of a single recipient (P calculated by cytometry.-/- Acetaldehyde (mM) Acetaldehyde (mM) Wild type Aldh2-/. Representative flow cytometry plots for the quantification of of mice).Trp53 -/.(n = 3) Fancd2 -/.. Survival of B cells and myeloid progenitors (CFU-GM) p53+ LKS cells from 8-to-12-week-old Aldh2−/−Fancd2−/− and control following exposure to acetaldehyde in vitro.Trp53 -/- 1 1 0. irradiated wild-type and Trp53−/− mice were used two-sided Mann–Whitney test. 2 hours (n = 5) Aldh2 -/.cells p53+ LKS p53+ LKS p53+ LKS p53+ LKS c-Kit (PerCP-Cy5.e.5) 200K 0. Aldh2−/−Fancd2−/− mice. for three independent experiments.m.(n = 5) 50 10 Gy.33% 10 150K FCS 3 10 100K 50K 0 0 4000 LKS cells 4000 LKS cells 4000 LKS cells 4000 LKS cells 0 10 3 10 4 10 5 0 10 3 10 4 10 5 0 10 3 10 4 10 5 0 10 3 10 4 10 5 0 10 3 10 4 10 5 Sca-1 (PE-Cy7) p53 (AlexaFluor 647) b 40 Wild type (n = 4) 20 Wild type (n = 8) c 60 Wild type (n = 3) 3.1 0 1 2 3 4 0 2 4 6 8 Trp53 +/+ -/. Frequency of CFU-S12 in the bone marrow bone-marrow populations. 2 hours.22 ± 0.0 - G -1 + -1 + e- - G -1 + G -1 + -1 + -1 + + + + e- S G -1 + BM -1 + + + + + + + S S BM BM + + + S BM e R 71 e LK R 71 R 71 r-1 20 LK LK R 71 19 r-1 r-1 20 20 LK ag 19 19 r-1 20 ag ag 19 ag ac ac ac TE CD ac al TE CD TE CD B2 al al TE CD B2 B2 al ne B2 ne ne ne M t M M t t M To t To To To Li Li Li Li 17-fold d 100 e 100 f P < 0.(n = 3) Aldh2 -/.06) 10 10 1 Wild type Wild type Trp53 -/.07% 20.1 ± 0. Aldh2-/- Fancd2-/- Extended Data Figure 6 | Aldehyde-induced stress elicits a p53 could not be determined (data shown as mean and s.0001 CFU-S12/105 nucleated BM cells CFU-GM clonogenic survival (%) 10 B cell growth (%) NS (P = 0. d.

Fancd2-/- Mean Corpuscular Volume (µm3) 80 30 NS 80 NS P = 0.0052 White blood cells (103/mm3) Red blood cells (106/mm3) Platelets (103/mm3) 1500 10 10 1000 5 5 500 0 0 0 +/+ -/. Fancd2-/.+/+ -/.0043 BM cellularity (106 cells/femur) 40 30 20 Trp53 -/.+/+ -/. Wild type Aldh2-/. 16. d.+/+ -/. Aldh2-/. 18 and 12 mice.0276 2000 P = 0.+/+ -/. 21. n =​ 17.0043 P = 0. Aldh2-/- Fancd2-/. left was observed in Aldh2−/−Fancd2−/−Trp53−/− mice compared to to right. Aldh2-/- Fancd2-/. Trp53 Wild type Aldh2-/. Trp53 +/+ -/.+/+ -/. Full blood count analysis of as described previously6. Trp53 +/+ -/.+/+ -/.+/+ -/. 18. Wild type Aldh2-/.+/+ -/. Wild type Aldh2-/. Aldh2−/−Fancd2−/−.+/+ -/.+/+ -/.+/+ -/. Fancd2-/. Fancd2-/. Wild type Aldh2-/. Fancd2-/. Fancd2-/. Aldh2-/- Fancd2-/. Trp53 +/+ -/. Trp53 +/+ -/. Aldh2-/- Fancd2-/. Fancd2-/. 6 and 5 mice.+/+ -/. Fancd2-/. Fancd2-/..0043 P = 0. 4. Fancd2-/. data shown as mean and s. Wild type Aldh2-/.+/+ -/.Fancd2 -/. A significant increase in the of ethanol treatment. data shown as mean and s. Article RESEARCH a 15 P = 0.4935) Haemoglobin (g/dl) Haematocrit (%) 60 60 20 40 40 10 20 20 0 0 +/+ -/. Fancd2 -/.+/+ -/.+/+ -/. 14. 6. Trp53 +/+ -/. Trp53 +/+ -/.m.+/+ -/. Trp53 Wild type Aldh2-/. Trp53 Wild type Aldh2-/. Bone marrow cellularity after 10 days a C57BL/6 ×​ 129S4S6/Sv F1 background). Fancd2 -/. Wild type Aldh2-/.Trp53-/.+/+ -/. red blood cells. All rights reserved.+/+ -/.+/+ -/. a.609) Haemoglobin (g/dl) 60 Haematocrit (%) 60 20 40 40 10 20 20 0 0 +/+ -/. c. n =​ 5.+/+ -/.+/+ -/. Aldh2−/−Fancd2−/−Trp53−/− and blood cytopenias and ethanol-induced bone-marrow failure control mice were treated with ethanol in their drinking water for 10 days in Aldh2−/−Fancd2−/− mice. Fancd2-/.+/+ -/.+/+ -/.0310 (P = 0. P calculated by two-sided Mann–Whitney number of white blood cells.Trp53 -/- 10 0 +/+ -/.+/+ -/..+/+ -/.+/+ -/.+/+ -/. Aldh2-/. 12.+/+ -/. Aldh2-/.0884) (P = 0.Fancd2 -/- P = 0. Fancd2-/- Mean Corpuscular Volume (µm3) 80 30 80 NS P = 0. Fancd2-/. Aldh2-/. 8.+/+ -/. Wild type Aldh2-/. Wild type Aldh2-/. on 10 days of ethanol treatment.m. Trp53 +/+ -/. 6. Aldh2-/. Haematoxylin and eosin staining of bone-marrow sections Aldh2−/−Fancd2−/− mice (P calculated by two-sided Mann–Whitney test. Trp53 +/+ -/.+/+ -/. Fancd2-/. Fancd2-/. Aldh2 -/.e.+/+ -/.+/+ -/. Trp53 Wild type Aldh2-/. Aldh2 -/. part of Springer Nature. left to right).Trp53-/. c. © 2018 Macmillan Publishers Limited.0043 (P = 0.+/+ -/. 10 days after ethanol treatment (original magnification. Full blood-count analyses were carried out after Aldh2−/−Fancd2−/−Trp53−/− and control mice (8-to-12 weeks old. .+/+ -/. In b. Aldh2-/.0002 15 P = 0.0087 Red blood cells (106/mm3) Platelets (103/mm3) 1500 10 10 1000 5 5 500 0 0 0 +/+ -/. Fancd2-/. Aldh2-/- Fancd2-/- Extended Data Figure 7 | p53 deficiency suppresses peripheral. 6.+/+ -/. Fancd2-/. Trp53 Wild type Aldh2-/.+/+ -/.4286) P = 0. Fancd2-/- b 15 NS 15 2000 White blood cells (103/mm3) (P = 0. Aldh2-/. Fancd2-/- c 50 d Wild type Aldh2 -/.+/+ -/. Fancd2-/. platelets and haematocrit test. b. Fancd2-/.+/+ -/. ×​100). Aldh2 -/.+/+ -/.e. Aldh2-/.

XY.XY. Trp53 -/.13) Fancd2 -/.XY. der(3)T(3. n =​  8 (90 metaphases per condition). 1/30 40. +/+ -/.XY.XY.8691) c 50 Duplication Aneuploidy aberrations in 90 metaphases 1.17).18) 41.XX. Del(17) 1/30 40. . T(3. T(2.8 30 Number of 0. der(2)T(2. +6 40.XY.XY. Del(2).X.XY. +/+ -/. part of Springer Nature.XY.XY.. +/+ -/- Wild type Aldh2-/.XY. Del(4) 40. der(9)T(4.4 10 0. b. T(5. RESEARCH Article NS a 1. T(13. List of chromosomal aberrations observed in the bone marrow Aldh2−/−Fancd2−/−Trp53−/− mouse. Quantification of micronucleated NCEs in the genotype.?).XY.XY. Trp53 +/+ -/.11/30 40.XX. Del(1). -19 [x2] 41.Trp53 -/. Del(Y) 41.XY. Del(19) 39. Wild type Aldh2-/.XX.+6 [x2] 40.-X [x2] 11/30 39. Three mice and 30 metaphases per mouse were analysed per Trp53−/− mice. Fancd2-/.XX.2 0 0 Trp53 +/+ -/.m. Del(1) 40. Del(14) 6/30 40.+/+ -/.XY. Del(15) 40. data shown as mean and s. T(10.der(13)T(13. Dp(1) 41. der(11)T(X.Dp(Y) -/- Aldh2 Fancd2 -/.XY.XY.XY.11) 1/30 40.16) 40. -10. Fancd2-/- b Trp53 -/. -Y [x2] 41.4). Del(10) 40.3) d 1/30 40. +4.Trp53 -/.XY. Fancd2-/.11). 0/30 40.XY. d. der(2)T(2. All rights reserved. Aldh2-/. Bar chart classifying the types of aberrations for each genotype by two-sided Mann–Whitney test. Aldh2-/- Fancd2-/. Examples of two metaphases from an mice). +/+ -/. Del(10) 0/30 40.18) 4/30 39. Del(2).e.XY.XY.11) 40.XY. der(2)T(2. c. 3/30 42.+8 40. der(12)T(12.13) Extended Data Figure 8 | Genomic instability in Aldh2−/−Fancd2−/− mice. -1 40.X.9) [x2] 40. T(2.2 (P = 0. der(3)T(2. Del(2) 2/30 39. Del(6) 40. of 8-to-12 week-old untreated Aldh2−/−Fancd2−/−Trp53−/− and control © 2018 Macmillan Publishers Limited.XY.6 20 0. Del(9) 40. the numbers represent the fraction of abnormal metaphases per blood of Aldh2−/−Fancd2−/−Trp53−/− and control mice (P calculated mouse.0 40 Translocation Ddeletion Chromatid break Mn-NCE (%) 0. Chrb(5) 40. +/+ -/.XY .XY.7) Aldh2 -/. Dp(9) 39.+8 40. a. Chrb(3) 39. +4 [x2] 39.XY.16) 40.XY.XY.XY.XY.

000X b 40 0.3 0. indicative of ‘clonal haematopoiesis’ in these mice (accounting for represent the first and third quartiles.1 20 0. . confirming the clonal nature of coverage (100–100.1 0.6 0. bone-marrow cells from the mice that provided the transplanted HSCs.1 and were of the indel location from 20×​whole-genome sequencing. Coverage depth and VAF of the filtered set of indel calls from whole.000×) to confirm that the calls were not sequencing most indels. box edges © 2018 Macmillan Publishers Limited. On the basis 91.05 103 0.5. part of Springer Nature.10 104 0. All rights reserved.001 Depth Depth 0.0 0.0 0.20 c 106 0. c.0 0. 14/159 (8.2 0.7 and 21.0 nc 2. box plot shows the mean.4% of blood production.1 #2 Fa dh # d /- Al 2-/- d2 /- -/- VAF VAF nc 2- Fa ldh A Extended Data Figure 9 | Validation of indels by targeted deep represent the first and third quartiles.4 0.7 0. We could validate the presence of and validation of indel calls by amplicon deep sequencing.2% of the initial calls.2 0.. validation (n =​ 159 locations.1 0 0. a.15 105 0.00 102 0. we attempted to detect indels in DNA samples of bone-marrow samples from the mice that provided the transplanted HSCs. respectively).2 0. subsequent whole-genome sequencing 100×​and were used for the analysis. we could detect indels from two Aldh2−/−Fancd2−/− genome sequencing (n =​ 342 indels.8 0. the calls were below the detection limit of the assay (VAF b.m. Coverage depth and VAF of the indel calls from deep sequencing s.cells a Single HSC 16 weeks Germline BM transplants reference Lineage+ cells HSC Tail 20X WGS ? Multiplex PCR Indels seen Amplicon sequencing in bone marrow (BM) before transplant? > 100-100. box plot shows the mean.4 d 1 Relative frequency (fractions) Relative frequency (fractions) 30 0.5 0. n =​ 13 and 7 indels. Data shown as mean and data).3 0. Article RESEARCH Lineage. In addition.9 1. whiskers extend over 10–90% of sequencing.7 0.0 0.1 0. However.8 0.3 VAF in donor BM Coverage (X) Coverage (X) 0. <​0. d. we designed deemed false positives (indicated by grey shading). Note that the VAF multiplex PCRs and deep sequenced the PCR products to higher distribution is centred tightly around 0. In most cases. box edges HSCs. We used targeted deep sequencing to look for indel calls in artefacts.5 0.9 1.e.6 0. Scheme depicting the generation of HSC clones by data).0001). whiskers extend over 10–90% of 0.8%) calls had VAF <​0.4 0.01 10 0. One hundred and fifty-nine locations had coverage greater than transplantation of single stem cells.

part of Springer Nature. deletion c5:145424065 24239 336 Fancd2-/.dupl HSC HSC HSC HSC HSC HSC HSC HSC HSC Tail BM Tail BM Tail BM Tail BM Tail BM Tail BM Tail BM Tail BM Tail BM 1000 bp 500 bp 100 bp c HSC HSC Tail Tail BM Sample Type Start location Size (bp) PCR Sample Type Start location Size (bp) PCR BM MD5045a deletion c13:112244133 2999 161 MD5445a translocation c4:128611975-c8:106880788 . 1. deletion c4:124986607 24307 164 deletion c6:86545624 47656 177 MD5054a deletion c5:59377956 1148 268 Aldh2-/. duplication c14:64748311 1651 153 MD5057a deletion c16:56319262 10056 131 Aldh2-/.Fancd2-/.p53-/. absence in matched germline samples from single stem cells.p53-/- MD5052a deletion c4:29837258 9957 173 MD5448a deletion c3:81290371 10604 167 Aldh2-/. in some cases. c. and the sensitivity of each PCR. PCR amplification in these samples flanking the breakpoints calculated by the BRASS algorithm. Aldh2-/-Fancd2-/-p53-/- c11 del c4 del c6 del c17 del c8 del c8 inv1 c8 inv2 c13 del c14 t.Fancd2-/.p53-/. In production. 16/27 (59%) and were present in the stem cell at the time of transplantation. Aldh2-/. Aldh2-/-Fancd2-/. Aldh2-/-Fancd2-/.c8 c3 del HSC HSC HSC HSC HSC HSC HSC HSC HSC HSC Tail BM Tail BM Tail BM Tail BM Tail BM Tail BM Tail BM Tail BM Tail BM Tail BM 1000 bp 500 bp 100 bp MD5450 MD5454 MD5455 MD5456 Aldh2-/-Fancd2-/-p53-/. gels (one experiment) showing presence of specific PCR amplification a. 161 MD5455a deletion c8:44991181 7530 191 translocation c15:80199437-c9:77318662 . 152 Fancd2-/. .Fancd2-/.Fancd2-/- c13 del c6 inv c4 del c4 del c5 del a c5 del b c5 del c9 del HSC HSC HSC HSC HSC HSC HSC HSC Tail BM Tail BM Tail BM Tail BM Tail BM Tail BM Tail BM Tail BM 1000 bp 500 bp 100 bp MD5056 MD5057 MD5445 MD5448 Aldh2-/-Fancd2-/. 199 Aldh2-/. we attempted to detect the rearrangements in DNA samples of in Supplementary Fig. RESEARCH Article Lineage.Fancd2-/. bal c16 del c2 del c4 . 226 translocation (bal) c15:80200668-c9:77317689 .cells a Single HSC 16 weeks Germline BM transplants reference Lineage+ cells HSC Tail 20X WGS ? Rearrangement seen PCR validation in bone marrow (BM) before transplant? Sanger sequencing b MD5045 MD5052 MD5053 MD5054 Aldh2-/-Fancd2-/. Fancd2-/-p53-/. Scheme depicting the generation of HSC clones by transplantation of from DNA of HSC clones. deletion/duplication c9:117677119 342 198 MD5454a deletion c17:46760471 88447 321 MD5056a translocation c10:82901864-c9:49792783 . 163 inversion c8:45070602 69445 244 translocation c15:80200657-c9:77317682 .p53-/- Aldh2-/.p53-/- MD5053a deletion c13:19185501 24841 103 MD5450a deletion c11:37239409 6950 477 Aldh2-/. Fancd2-/-p53-/- c10-c9 c10-c9 bal c15-c9 c15-c9 bal c15-c9 b c9-c15 b.Fancd2-/. deletion c13:19311304 25273 132 Aldh2-/. Aldh2-/. © 2018 Macmillan Publishers Limited.Fancd2-/. b. inversion c6:69207252 36822 135 Aldh2-/.p53-/.Fancd2-/. Gel source data is shown addition. deletion c2:177355146 3657 229 Extended Data Figure 10 | Validation of rearrangements by PCR. 169 Aldh2-/. Aldh2-/-Fancd2-/-p53-/. and is dependent on the contribution of the transplanted HSC to blood the identity of the products was confirmed by Sanger sequencing. Agarose rearrangements could be detected before transplantation. detection in bone-marrow of rearrangement calls by PCR. subsequent whole-genome sequencing and validation the tail of the same mouse and. Aldh2-/-p53-/. List summarizing the rearrangements found bone-marrow cells from the mice that provided the transplanted HSCs. 176 MD5456a deletion c13:84517083 54157 212 Aldh2-/.Fancd2-/. We designed primers for nested PCRs tissue that predates the transplants. inversion c8:45073529 72375 180 translocation (bal) c15:80199443-c9:77318664 . translocation (bal) c10:82901948-c9:49793011 . All rights reserved. in 28 HSC clones and the results from b. All 27 rearrangements could demonstrating that these changes did not arise during clonal expansion be detected by PCR and confirmed by Sanger sequencing.

However. Passage of mutated genetic information is prevented by end-joining. aldehyde lesions degenerate into DNA DSBs a. leading to the engagement of both classical and alternative and mutagenesis. Aldehyde catabolism and Fanconi anaemia (FA)-pathway-mediated that can be repaired through error-free recombination. and subsequent mutagenic repair. Fancd2-/. . this DNA repair constitute two distinct tiers of protection against aldehyde mechanism is not sufficient to fully compensate for Fanconi anaemia damage. © 2018 Macmillan Publishers Limited. In the absence of a functional and suppress mutagenesis by endogenous aldehydes in HSCs. Fanconi anaemia pathway. All rights reserved. Loss of this protection leads to the accumulation of DNA damage inactivation. part of Springer Nature. p53 HSC loss b Aldehyde-induced DSBs FA pathway HR cNHEJ Alt-EJ (FA independent) (Ku70) (MH) Repair Repair Mutagenesis Extended Data Figure 11 | Mechanisms to maintain genetic integrity the activation of p53. b. leading to HSC loss. Article RESEARCH a Aldh2-/.

Note: all studies involving animals and/or human research participants must disclose whether blinding and randomization were used. This form is intended for publication with all accepted life science papers and provides structure for consistency and transparency in reporting. reproducible on all occasions reproduced. data was acquired relying purely on identification numbers. some list items might not apply to an individual manuscript. mean) and variation (e.8 and alpha 0.or two-sided (note: only common tests should be described solely by name. such as an adjustment for multiple comparisons The test results (e. median. no statistical methods were used to predetermine sample size. interquartile range) Clearly defined error bars See the web collection on statistics for biologists for further resources and guidance.htm.05.g. For all other animal experiments. etc. ` Experimental design 1. 3. but all fields must be completed for clarity. samples size was determined by power analysis using the following site http://biomath. noting whether measurements were taken from distinct samples or whether the same sample was measured repeatedly A statement indicating how many times each experiment was replicated The statistical test(s) used and whether they are one.g. Sample size Describe how sample size was determined. 1 . given as a discrete number and unit of measurement (animals. cultures. For further information on Nature Research policies. using power of 0. Data exclusions Describe any data exclusions. 4. For further information on the points included in this form. nature research | life sciences reporting summary Corresponding author(s): Patel KJ Initial submission Revised version Final submission Life Sciences Reporting Summary Nature Research wishes to improve the reproducibility of the work that we publish. Replication Describe whether the experimental findings were All experiments were replicated as stated in the text. litters.info/power/chsq1gp. confirm that the following items are present in relevant figure legends (or in the Methods section if additional space is needed). 5. Every life science submission will use this form. see Reporting Life Sciences Research. more complex techniques should be described in the Methods section) A description of any assumptions or corrections. n/a Confirmed The exact sample size (n) for each experimental group/condition. 6. see Authors & Referees and the Editorial Policy Checklist. Blinding Describe whether the investigators were blinded to The investigators were blinded to the genotypes of mice throughout the study and group allocation during data collection and/or analysis. P values) given as exact values whenever possible and with confidence intervals noted June 2017 A clear description of statistics including central tendency (e.g. Statistical parameters For all figures and tables that use statistical methods. No data were excluded from analyses. All experiments were reliably reproduced. 2.) A description of how samples were collected. including our data availability policy. For the analysis of Mendelian segregation of alleles. Randomization Describe how samples/organisms/participants were No randomisation was employed allocated into experimental groups. Sufficient animals were used in order to detect a 50% reduction in expected frequency. standard deviation.

Flow cytometry data was analyzed with FlowJo.7. anti-CD45R/B220 (PerCP-Cy5. anti-cleaved caspase-3 (AlexaFluor647.4). clone A20. clone C2. BioLegend).2 (APC. clone R17217.1. anti-TER-119 (FITC. NGS data which was analysed with study. eBioscience).e. anti-Ly-6G/Gr-1 (PE. TER-119 (clone Ter119. INTRACELLULAR ANTIGENS: anti-p53 (AlexaFluor647. clone 104. WESTERN BLOT: The FANCA antibody (Cell Signalling.19. clone RA3-6B2. CD11c (clone N418. BioLegend) and anti-CD48 (biotin. FITC-CD71 antibody (GenTex.5. anti-CD8a (FITC. D1L2Z) was used at 1:1000 in 5% w/v BSA. BD Pharmingen). anti-CD150 (PE. anti-CD11b/Mac-1 (PE. HSCs FITC-conjugated lineage cocktail with antibodies anti-CD4 (clone H129. at 1:2000 for one hour at room temperature. Fcε R1α (clone MAR-1. BD Pharmingen). A-11001). clone H129. clone 1A8. clone RA3-6B2. clone M1/70. BD Pharmingen). eBioscience).1% Tween20 at 4°C with gentle shaking. FLOW CYTOMETRY: Micronucleus assay. CD3e (clone 145-2C11. BD Biosciences) for use in the system under study (i. assay and species). BioLegend). eBioscience). clone TER-119. anti-c-Kit (PerCP-Cy5. BioLegend). eBioscience). BD Pharmingen) and anti-CD41 (FITC. CD45R/B220 (clone RA3-6B2. anti-CD11b/Mac-1 (PE. BD Pharmingen) for erythroid maturation. 129002) was used at 1:4000 in the same conditions.1 (BV421. eBioscience). clone D3E9. anti-TER-119 (PE- Cy7. clone D7. Software Describe the software used to analyze the data in this Data was analysed using Graphpad prism. MATURE LINEAGES: anti-CD45R/B220 (PE. BD Pharmigen). eBioscience) for monocyte/granulocyte progenitors. Cell Signalling). clone TER-119. Nature Methods guidance for providing algorithms and software for publication provides further information on this topic. BioLegend) and anti-CD45. Materials availability Indicate whether there are restrictions on availability of All materials are available unique materials or if these materials are only available for distribution by a for-profit company. The anti- VINCULIN (abcam. 9. BioLegend). Antibodies Describe the antibodies used and how they were validated SCE ASSAY: mouse FITC-conjugated anti-BrdU antibody (Clone B44. clone 1C12. 0. 1X TBS. We strongly encourage code deposition in a community repository (e. BD Pharmingen). Ly-6G/Gr-1 (clone RB6-8C5. clone 1A8. CD8a (clone 53-6.7. GitHub). clone 2B8. Dako swine anti-rabbit immunoglobulins HRP was used as secondary antibody. Cell Signalling). CaVEman. clone MWReg30. clone M1/70.5. software must be made available to editors and reviewers upon request. pindel and Brass. BD Pharmingen) and anti-CD71 (PE. clone HM48-1. and goat anti-mouse Alexa fluor-488 secondary antibody (Life Technologies. anti-CD45. BD Pharmingen) and anti-IgM (FITC.` Software nature research | life sciences reporting summary Policy information about availability of computer code 7. BD Pharmingen). BD Pharmingen). CD11b/Mac-1 (clone M1/70.2. overnight.g. ` Materials and reagents Policy information about availability of materials 8. June 2017 2 . For manuscripts utilizing custom algorithms or software that are central to the paper but not yet described in the published literature. eBioscience). clone TC15-12F12. BD Pharmingen) for B cell progenitors. clone 53-6. clone II/41. ENGRAFMENT anti-CD4 (FITC. BD Pharmingen). anti-Sca-1 (PE-Cy7. The antibody was validated with FANCA KO samples. BD Pharmingen).19. BD Pharmingen) and anti-Ly-6G/ Gr-1 (FITC.

If any of the cell lines used are listed in the database N/A of commonly misidentified cell lines maintained by ICLAC. see above reports for detailed information. Mol Cell. Al Abo. Caner Research.. provide a scientific rationale for their use. c. Eukaryotic cell lines nature research | life sciences reporting summary a. 2004) or the lab of Shunichi Takeda (HR mutants: Takata. Report whether the cell lines were tested for All cells lines tested mycoplasma negative mycoplasma contamination. DT40 cell lines used in this study were described previously. Mol Cell Biol. Nidezwiedz et al. b.10. They were generated in our lab (FANCC. FANCC/XRCC2. Describe the method of cell line authentication used.. 2014). DT40 genetic mutants were validated by southern blot and sensitivity to DNA damaging agents. d. 2001. June 2017 3 . State the source of each eukaryotic cell line used.

UK ) and Aldh2+/- Fancd2+/. Nautre 2012) were crossed with Ku70+/- mice (Xrcc6tm1Fwa. MGI ID: 2673422. sterility and sensitivity to mitomycin C (Figure 3. For the in vivo point mutation assay. Fanca∆/∆ mice phenocopy the Fanca-/.BigBlue λLIZ and control mice. For single HSC transplantation experiments. the p53 allele reported previously 42 was backcrossed onto 129S6/Sv or C57BL/6J for six generations.mice to establish parental (F0) strains on both genetic backgrounds. with selection at each generation by serotyping with anti-CD45.1 (A20. which were finally crossed to obtain Aldh2-/-Fancd2-/-p53-/. Similarly to Aldh2-/- Fancd2-/.mice in a pure C57BL/6 background. EUCOMM8) was backrossed from C57BL/6N onto 129S6/Sv for five generations and crossed with Fancd2+/. these mice were crossed with Fanca+/- Ku70+/-Vav1-iCre to generate FancaF/- Ku70-/. Harwell. In individual experiments all mice were matched for gender and age (8-12 weeks). CD45. as judged by FANCA expression.` Animals and human research participants nature research | life sciences reporting summary Policy information about studies involving animals. C57BL/6N. FITC. Description of human research participants Describe the covariate-relevant population N/A characteristics of the human research participants. mice carrying BigBlue λLIZ shuttle vector repeats (Stratagene) were crossed with Aldh2-/-Fancd2+/.mice to eventually produce FancaF/F Ku70+/- mice.mice were also generated in a C57BL/6 x 129S4S6/Sv F1 background. and the Fanca+/- Ku70+/. Fanca+/- mice (Fancatm1a(EUCOMM)Wtsi.males (C57BL/6. To this end. The resulting mice were inter-crossed to obtain Aldh2-/-Fancd2-/. Aldh2-/- Fancd2-/-p53-/.mice reported previously. Finally.Vav1- iCre+ tissue-specific double mutants (also pure C57BL/6). Fanca+/F  mice were then crossed with Ku70+/. FW Genesis 2000) to produce the Fanca floxed allele  (FancaF or Fancatm1c(EUCOMM)Wtsi). Extended Data Figure 5). the previously reported Fancd2 allele (Fancd2tm1Hou. FL040: TCAGCTCACTGAGACGCAACCTTTT ACACT.F1 mice.and control F1 mice.progeny were then intercrossed to generate all possible genotypes. and in this case yields the Fanca null allele  (Fanca∆ or Fancatm1d(EUCOMM)Wtsi). Recombination of the Frt sites was verified by PCR (FL033: GCCTTTGCTGCTCTAATTCCATGT. MGI ID: 4434431. The Vav1-iCre allele directs the expression of the iCre recombinase to HSCs and hematopoietic tissues 41.mice on a C57BL/6 x 129S4S6/Sv hybrid background. MGI ID: 4431566. we employed CD45. Description of research animals Provide details on animals and/or animal-derived Aldh2-/- Fancd2-/. the previously reported Aldh2 allele (Aldh2tm1a(EUCOMM)Wtsi.SJL onto 129S6/Sv for six generations. for this study.1 homozygous mice on a C57BL/6J x 129S6/Sv F1 background as recipients. MGI ID: 2179954 the generation of which was described in Gu. Fanca+/- mice were first crossed with FLP deletor mice (Farley. when reporting animal research.females (129S4S6/Sv. BioLegend) and anti-CD45. and En2A: GCTTCACTGAGTCTCTGGCATCTC) and reconstitution of FANCA expression was verified by western blotting FancaF/F spleen extracts (Figure 3). Harwell UK). a gift from M. EUCOMM the generation of which was described in Garaycoechea JI. p53+/. Y Immunitry 1997). 4 . Aldh2-/- Fancd2-/. All animals were maintained in specific pathogen-free conditions. Grompe) was backrossed onto the C57BL/6Jo1a background for ten generations and crossed with Aldh2+/. BioLegend).mice on a pure C57BL/6 background. For the generation of FancaF/- Ku70-/.(C57BL/6N) mice to generate Aldh2+/- Fancd2+/. To generate Fanca-/- Ku70-/.mice on a 129S4S6/Sv background.mice to generate Aldh2+/- Fancd2+/.mice on a C57BL/6 x 129S4S6/Sv F1 background were generated materials used in the study. Likewise. June 2017 Policy information about studies involving human research participants 12.1 (or Ptprca) had been serially backcrossed from B6.mice were then intercrossed with Aldh2-/- Fancd2+/.Vav1-iCre and control mice.2 antibodies (104. PE-Cy7. follow the ARRIVE guidelines 11. Briefly. Pups from these crosses were genotyped at between 2-3 weeks old. Finally.and control mice were generated as F1 hybrids from crosses between Aldh2+/- Fancd2+/.

1 μl of FITC-CD71 antibody (GenTex. 1 ml of fixed blood cells were then washed with 6 ml of bicarbonate buffer (0.5 % FCS) and strained through 70 μm meshes. Finally. 360 μl of blood suspension were then added to 3. the cell pellet was resuspended in staining buffer and nucleated cells were counted with 3% acetic acid (StemCell Technologies) on a Vi-Cell XR cell viability counter (Beckman Coulter). bone marrow cells were isolated from tibiae and femurs with staining buffer (PBS supplemented with 2. 4. The axis labels state the marker and fluorochrome used (e. CD4-FITC). nature research | flow cytometry reporting summary Corresponding author(s): KJ Patel Initial submission Revised version Final submission Flow Cytometry Reporting Summary Form fields will expand as needed. Treated or untreated mice (8-12 weeks of age) were bled and 62 μl of blood were mixed with 338 μl of PBS supplemented with 1000 U/mL of heparin (Calbiochem). the cell pellets were resuspended in 100 μl of staining 1 . After centrifugation. The cell pellets were resuspended in 200 μl of staining buffer containing streptavidin-BV421 and incubated for another 15 minutes at 4°C. All plots are contour plots with outliers or pseudocolor plots. Red cells were lysed by resuspending the cells in 10 ml of red cell lysis buffer (MACS Miltenyi Biotec) for 10 minutes at room temperature. ` Methodological details 5. confirm that: 1. pH 7.2) and incubated for 10 minutes at room temperature. Red blood cells (RBC) were lysed by the addition of 1 ml of ammonium chloride lysis buffer (155 mM NH4Cl.1. 72 μl of bicarbonbate buffer.7 For HSC quantification. 10x106 bone marrow cells were resuspended in 200 μl of staining buffer and stained as described in the methods.6 ml of -80°C methanol and placed at -80°C for at least 12 hours.9% NaCl and 5. ` Data presentation For all flow cytometry data. Finally. 2. 10 mM KHCO3.3 mM NaHCO3). 3. 0.1 mM Na2EDTA. 50 μl of blood were obtained from the tail vein of recipient mice every two weeks.4) and 7 μl of RNaseA (Sigma) were premixed and added to 20 μl of each cell suspension. resuspended in 500 μl of staining buffer. After centrifugation. The axis scales are clearly visible. clone R17217. cells pellets were resuspended in 500 μl of bicarbonate buffer supplemented with 5 μg/ml propidium iodide (Sigma). The cells were stained at 4°C for 45 minutes.g.0. data acquired on a Fortessa analyser (BD) June 2017 To assess the engraftment of single HSCs into irradiated recipients. Describe the sample preparation. The cells were resuspended in 150 μl of bicarbonate buffer and 20 μl of this suspension was used for subsequent staining. followed by addition of 1 ml of bicarbonate buffer and centrifugation. Please do not leave fields blank. Include numbers along axes only for bottom left plot of group (a 'group' is an analysis of identical markers). The samples were analysed immediately on an LSRII analyser (BD) and data analysed with FlowJo 10. A numerical value for number of cells or percentage (with statistics) is provided. The samples were incubated for 15 minutes at 4°C and washed with 2 ml of buffer. cells were washed.

0. Describe the abundance of the relevant cell Post sort analysis was not conducted on single cells for HSC transplant. The gating strategies are outline in the Suppl Figure. Samples were run on the Fortessa analyser (BD) with the HTS module. LSRII analyser (BD) Fortessa Analyser (BD) iCyt Synergy Sorter (Sony) 7. 9. Identify the instrument used for data collection.7 the flow cytometry data. buffer and stained with antibodies described in the methods. After the incubation. Describe the software used to collect and analyze Data was collected using FACSDiva (BD) and processed using FlowJo 10. cells were washed with 3 ml of staining buffer before being nature research | flow cytometry reporting summary resuspended in 250 μl of the same buffer. 6. These cells were assessed for their functional ability to engraft and provide long term multi-lineage reconsitution. Describe the gating strategy used. June 2017 2 . 8. populations within post-sort fractions. Tick this box to confirm that a figure exemplifying the gating strategy is provided in the Supplementary Information.