The earliest stages of neoplastic transformation in Familial Adenomatous Polyposis

The earliest stages of neoplastic transformation in Familial Adenomatous Polyposis | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article The earliest stages of neoplastic transformation in Familial Adenomatous Polyposis Michael Stratton, Suet Yi Leung, Yichen Wang, Philip Robinson, and 22 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6581155/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract The succession of somatic genetic events associated with the conversion of a normal colorectal epithelial cell into a colorectal carcinoma constitutes a paradigmatic model of cancer development. Familial Adenomatous Polyposis (FAP) is caused by constitutional inactivating mutations in APC , the central gatekeeper gene of colorectal cancer, and is associated with a substantially increased lifetime-risk of colorectal cancer. To investigate the earliest stages of neoplastic change due to APC inactivation, we microdissected and individually whole genome sequenced 279 histologically normal and abnormal colorectal crypts from 15 individuals with FAP. Histologically normal crypts generally exhibited similar mutation burdens and mutational signatures to normal crypts from wild-type individuals of the same age, with 1/110 carrying a somatic inactivating APC mutation. By contrast, 9/18 aberrant crypt foci carried somatic APC mutations and exhibited modestly increased burdens of some mutational signatures found in normal crypts. 12/13 diminutive adenomatous polyps (< 5mm diameter) showed somatic APC mutations and carried substantially increased mutation loads of most mutational signatures present in normal crypts. Phylogenetic trees of crypts from aberrant crypt foci and adenomatous polyps revealed that some had acquired their initiating somatic APC mutations decades previously during the first few years of life. The results catalogue the changes in somatic mutation rates, mutational processes and “driver” mutations in cancer genes during the earliest stages of colorectal neoplastic transformation initiated by APC inactivation and highlight the long periods of clonal evolution required for a cancer to develop. Biological sciences/Genetics/Genomics Biological sciences/Cancer/Gastrointestinal cancer/Colorectal cancer Figures Figure 1 Figure 2 Figure 3 Figure 4 INTRODUCTION The progression of somatic genetic changes associated with, and underlying, the conversion of a normal colorectal epithelial cell into a colorectal carcinoma has served as a general model of cancer development 1 . Through genetic characterisation of the succession of morphologically distinguishable intermediate neoplasms, the consensus sequence of somatic genetic changes leading to colorectal carcinoma has been described 2 , 3 . This progression incorporates an ordered accumulation of “driver” mutations in cancer genes commonly mutated in colorectal carcinoma, including APC , KRAS , TP53 , genome-wide changes in mutation burdens and signatures of all classes of somatic mutation, and their associations with macroscopically visible adenomatous polyps of increasing size and histological abnormality. Iterative revision of this progression has been enabled by advances in DNA sequencing technology resulting in increasingly comprehensive characterisations of adenoma 4 , 5 , 6 , 7 , 8 and carcinoma genomes 9 , 10 , 11 , 12 , 13 . The earliest stages in colorectal cancer progression have, however, been less well characterised. These may be of particular importance as potentially reversible endogenous and exogenous risk factors for colorectal cancer may exert their influence during this period. Familial Adenomatous Polyposis (FAP) is characterised by the presence of a single constitutional mutation in the APC gene, generally resulting in APC protein truncation and functional inactivation. This constitutional mutation is usually present in every cell of the body and somatic acquisition of inactivating mutations in the remaining wild type allele of individual cells results in biallelic APC inactivation, initiating colorectal neoplastic transformation. Individuals with FAP develop hundreds-to-thousands of intestinal adenomas over the first three decades of life. Without prophylactic surgical intervention, such polyposis substantially increases the risk of colorectal carcinoma, with the penetrance approaching 100% by 60 years of age 14 . Biallelic somatic APC mutations are also thought to be the earliest step of the chromosomal instability pathway of sporadic colorectal cancer development, which accounts for ∼80% of cases 10 , 11 , 15 , 16 . In consequence, study of FAP has been widely used to advance understanding of sporadic colorectal cancer development due to APC inactivation. In addition to profuse colorectal adenomatous polyps, individuals with FAP show elevated frequencies of aberrant crypt foci (ACF), microscopic areas of flat colorectal epithelium incorporating crypts exhibiting a range of abnormal architectural and/or nuclear cytological histological characteristics 17 . Microscopic examination of ACF has revealed heterogeneous histology, ranging from the presence of very early regenerative crypts (type A) to hyperplastic crypts (type B) to dysplastic crypts (type C) 18 . It has been proposed that ACF represent early stages of the normal cell-to-carcinoma progression before the development of macroscopically visible colorectal adenomatous polyps 19 , 20 , 21 . Early studies reported “driver” mutations in APC and KRAS 22 , 23 , 24 , 25 , 26 , 27 , 28 , 29 in ACF in different contexts, but few are from individuals with FAP 24 , 25 , 27 and these do not report the genome-wide features of neoplastic evolution, including mutation burdens and mutational signatures, that might inform on changes in mutational processes following biallelic APC inactivation. Characterisation of the somatic mutational landscapes of these very early stages of colorectal cancer development presents the technical challenge of generating genome sequences from the small amounts of DNA present in individual normal cells or microscopic clones. Recently, we combined laser capture microdissection with a bespoke library construction protocol that enables high quality whole genome sequences to be obtained from individual colorectal crypts 30 . Each crypt is a population of ~ 2,000 colorectal epithelial cells derived from a single, common, recent ancestor stem cell, with high variant allele fraction (VAF) somatic mutations found in DNA sequences from an individual crypt representing the mutations present in that recent ancestor stem cell. The method has been applied to normal colorectal crypts from individuals who are healthy or have colorectal cancer 31 , from individuals with colorectal cancer predisposition syndromes due to inherited DNA polymerase proof reading deficiencies 32 , DNA base excision repair deficiencies 33 , DNA mismatch repair deficiencies 34 , and from individuals with inflammatory bowel disease 35 . By cataloguing somatic mutations in whole genome sequences of normal crypts, crypts from ACF and crypts from very small polyps from individuals with FAP, before and after biallelic APC inactivation, we characterise the progression of “driver” mutation acquisition and changes in somatic mutation rates and mutational signatures during the earliest stages of colorectal neoplastic change due to APC inactivation at a much higher level of resolution than previously achieved. Furthermore, through the phylogenetic crypt analysis enabled by this approach we provide insights into the long time period over which evolution of neoplastic colorectal clones can occur. RESULTS Patients and samples 15 individuals (five of whom were from the same family) aged 18 to 44 years with intestinal adenomatous polyposis due to inherited heterozygous APC mutations were studied. Ten exhibited features of classical FAP with > 100 colorectal polyps and five had attenuated FAP (AFAP) with < 100 colorectal polyps (Supplementary Table 1). None had developed invasive colorectal carcinoma at the time of tissue sampling. All constitutional APC mutations were predicted to inactivate the APC protein and included nonsense substitutions (seven individuals), small frame-shifting insertions or deletions (five), essential splice site substitutions (two) and whole gene deletion (one) (Supplementary Table 1). 279 colorectal crypts were isolated by laser capture microdissection and individually whole genome sequenced: 110 were from histologically normal epithelium; 83 were from 18 ACF, of which three were composed of just a single bifurcating crypt with no other nearby abnormality, thus representing the simplest stereotypic morphological microscopic abnormality we were able to identify; 86 were from 16 adenomatous polyps, including 70 from 13 diminutive polyps ( 10mm) (Fig. 1 a, Supplementary Table 2). Crypts with normal or uncertain histology adjacent to, or entrapped within each type of lesion were also included to explore their origins. Crypts were individually whole genome sequenced to a median of 31-fold read coverage and somatic mutations of all classes were identified using previously described approaches 31 , 32 , 33 , 34 , 35 . Crypts with complex or branching morphologies were, where feasible, sampled and genome sequenced in separate segments. For clarity of narrative and analysis, when groups of crypts apparently from the same ACF/adenomatous polyp had different “driver” and genome-wide mutations, and thus different clonal origins, they were given different identifiers and treated as separate groups. Somatic mutations in APC and other cancer genes Like normal crypts from wild-type individuals 31 , crypts from individuals with FAP/AFAP, whether normal, from ACF or from adenomatous polyps, showed somatic mutations with VAF peaks of 0.4–0.5 indicating that they originate from recent single common ancestors (Extended Data Fig. 1 ). One out of 110 morphologically normal crypts (PD44897b_lo0001) from individuals with FAP/AFAP carried a high VAF somatic heterozygous inactivating APC mutation (R213*). Using short-read sequencing it was not possible, either in this or any other crypt genome sequence, to determine whether somatic APC mutations were in the constitutional mutant or wild type APC gene copy. Nevertheless, it is probable that they were in the wild type allele, resulting in biallelic APC mutation, cellular impairment of APC function and initiation of neoplastic transformation. The presence of this somatic APC mutation in most cells of this crypt likely occurred through the mutation arising in one of the ~ 5 stem cells at the crypt base with subsequent complete crypt colonisation by its progeny. This raised the possibility that other normal crypts had been incompletely colonised by APC somatic mutant stem cells, with the VAF of the somatic APC mutation consequently falling below standard thresholds for detection. However, a systematic search through the sequence reads of all normal crypts did not reveal any further somatic inactivating APC mutations, even at very low VAFs, indicating that incomplete colonisation is rare, and suggesting that colonisation occurs relatively quickly. Inactivating mutations, each restricted to one normal crypt, were found in other tumour suppressor genes ( RNF43 (R454Afs*48), FBXW7 (R224*), KMT2C (R2257*) and KMT2D (X1565_splice)). Their biological significance is unclear. Nine out of 18 ACF contained crypts with a somatic inactivating APC mutation (nine different APC mutations in 36/83 crypts). Microscopy images of the ACF and their constituent crypts were reviewed by two clinically practising colorectal histopathologists, without prior knowledge of the mutational results, and classified as 10 type A-I ACF (with crypt elongation, branching or dilatation, no to minimal regenerative changes, and absence of serration or dysplastic changes), seven type A-II ACF (with crypt elongation, branching or dilatation, moderate to marked regenerative changes, but no serration or dysplastic changes), and one type C ACF (with dysplastic change) based on established standards from previous literature 18 (Supplementary Table 3, Extended Data Fig. 2 ). There were no type B ACFs (with serration characteristic of early hyperplastic polyp) consistent with these being associated with an APC -mutation independent serrated neoplasm pathway to colorectal cancer 36 . Comparing sequenced crypts from ACF with a somatic APC mutation to those without showed that presence of the somatic inactivating APC mutation is associated with regenerative changes ( P = 0.001, Chi-squared test). None of the three crypt clusters composed of just a single bifurcating crypt had a somatic inactivating APC mutation. Among the nine ACF without a somatic APC mutation, a known hotspot mutation in ALK (R395H) was found in one crypt together with a splice site mutation in CDH1 (X17_splice). Two crypts with somatic inactivating APC mutations also had additional inactivating somatic mutations in tumour suppressor genes, one in SMARCA4 (X648_splice) and NCOR1 (R694*), and the other in KMT2C (L2662Cfs*16) (Fig. 1 b-c, Supplementary Table 4). These mutations were not shared with other crypts in the same ACF. 12/13 diminutive adenomatous polyps and 3/3 large adenomatous polyps contained crypts with inactivating somatic APC mutations. (Fig. 1 b-c). Two diminutive polyps and one large polyp showed individual crypts carrying two distinct inactivating somatic APC mutations (and therefore, in total, three inactivating APC mutations when the germline mutation is also considered), a phenomenon that has been previously reported 37 . Two diminutive polyps (PD40734C_PLP_007, PD42778D_PLP_003) each carried more than one independent APC mutant clone, and another one (PD42778D_PLP_002) had a type A-II ACF with an independent APC somatic mutation at the edge of the polyp, consistent with previous studies of early adenomas in mice and human 13 , 38 , 39 (Extended Data Fig. 3 ). Mutations in several other cancer genes, including canonical hotspot mutations, were found in crypts from polyps ( KRAS (G12S, G12D), TP53 (R202C, R64*), PIK3CA (E542K, E545K), SMAD4 (R361H) and FBXW7 (R479Q))(Supplementary Table 4). Some crypts adjacent to, or entrapped within, APC mutation carrying ACF and adenomatous polyps lacked second APC mutations despite exhibiting elongation, branching, dilatation, or regenerative changes. We termed these ‘bystander’ crypts to distinguish them from normal crypts and independent ACF. Elevated somatic mutation burdens in aberrant crypt foci and polyps Somatic mutation burdens were examined in 276 crypts from FAP patients with a coverage of > 10-fold, including 110 normal crypts, 102 crypts in ACF and adenomatous polyps with confirmed somatic APC mutations (39 crypts from 10 ACF and 63 crypts from 15 adenomatous polyps), 30 crypts from 9 ACF without APC somatic mutations, 4 adenomatous crypts from the diminutive polyp without somatic APC mutation, and 30 ‘bystander crypts’. In normal crypts from FAP/AFAP individuals, the total burdens of somatic single base substitutions (SBS, also sometimes termed single nucleotide variants or SNV) showed a linear accumulation with age of 41.7/year (95% C.I., 23–61, P-adjust = 1.5 × 10 − 3 ), which did not differ from that of normal crypts from wild-type individuals (likelihood ratio test, P = 0.59) (Fig. 2 a). Similarly, the burdens of small insertions and deletions (indels) in normal crypts from FAP/AFAP individuals did not differ from those of wild type individuals (likelihood ratio test, P = 0.22) (Fig. 2 a). Copy number variants (CNVs) and structural variants (SVs) were rare, but 132 retrotransposition events (RTs) were found in 110 normal crypts, all at a similar frequency to wild type individuals (Supplementary Table 5–7). Thus, across the spectrum of somatic mutation types, no differences were identified between the mutation rates in normal crypts from individuals with FAP and those from wild-type controls. The single morphologically normal AFAP crypt with an inactivating somatic APC mutation ( R213* ) showed a similar number of single base substitutions and indels to normal crypts without somatic APC mutations, with no copy number changes or rearrangements. Crypts from ACF which carried somatic APC mutations showed modest increases in total small indel mutation burdens (35 more indels, 95% C.I., 12–58, P-adjust = 7.0 × 10 − 3 ) (Fig. 2 b-c). Copy number variants, structural variants and retrotransposition events were at a similar level to normal crypts. In contrast, crypts with somatic APC mutations from diminutive adenomas (< 5mm) showed markedly increased burdens of both single base substitutions (1580 more substitutions, 95% C.I., 1236–1924, P-adjust = 5.3 × 10 − 17 ) and indels (186 more indels, 95% C.I., 158–215, P-adjust = 2.0 × 10 − 30 ), and these burdens were further increased in the three large adenomas (6728 more substitutions, 95% C.I., 5820–7636, P-adjust = 7.1 × 10 − 38 ; 794 more indels, 95% C.I., 730–857, P-adjust = 8.5 × 10 − 81 ). Copy number variants, structural variants and retrotransposition events were more commonly found in diminutive polyps compared with normal crypts (CNV: increased by 0.6/crypt, 95% C.I.: 0.2–0.9, P-adjust = 4.2 × 10 − 3 ; SV: increased by 0.7/crypt, 95% C.I.: 0.3–1.2, P-adjust = 4.2 × 10 − 3 ; RT: increased by 6.6/crypt, 95% C.I.: 3.8–9.4, P-adjust = 3.8 × 10 − 5 ), with more significant increases in large polyps (CNV: increased by 5.0/crypt, 95% C.I.: 4.3–5.7, P-adjust = 6.0 × 10 − 25 ; SV: increased by 4.3/crypt, 95% C.I.: 3.4–5.2, P-adjust = 1.3 × 10 − 14 ; RT: increased by 36.4/crypt, 95% C.I.: 30.5–42.1, P-adjust = 8.0 × 10 − 21 ). Overall, ACF from FAP patients, including the one dysplastic ACF, appear to represent an early stage of neoplastic progression following APC inactivation in which some mutational processes are mildly accelerated, but large-scale chromosome instability has not yet been triggered. ‘Bystander crypts’ within or immediately adjacent to ACFs and polyps did not show increased mutation burdens, confirming their likely status as entrapped normal crypt epithelium. Mutational signatures Mutational signatures were extracted from the full set of FAP/AFAP crypt mutation catalogues (ie including morphologically normal and abnormal crypts) together with a previously published set of normal crypts from wild-type individuals 31 . Two different methods were used (HDP 40 and SigProfiler 41 ) which yielded similar repertoires of signatures. Using HDP, 11 SBS signature components were identified (Extended Data Fig. 4 ) which could be explained by 10 previously described mutational signatures: SBS1, SBS2, SBS5, SBS13, SBS17b, SBS18, SBS88, SBS89, SBS93 and SBSC (Fig. 3 a, Extended Data Fig. 5–6). SBS1 is caused by deamination of 5-methylcytosine, a process which occurs in most human cell types throughout life at a more-or-less constant rate 42 , 43 . SBS2 and SBS13 are caused by activity of the AID/APOBEC family of cytidine deaminases, and were only found in three crypts from the previously published control group 31 . SBS5 is also ubiquitous in normal human tissues, and accrues in a clock-like manner with age, but its aetiology is not well understood 42 , 43 . SBS18 is attributed to the effects of reactive oxygen species on DNA, principally the formation of 8-oxoguanine 44 , and is present in some normal cell types including human placenta 45 , and large intestine epithelium 46 , 31 . SBS88 is due to colibactin, a mutagenic agent produced by strains of E.coli often present in the large intestine microbiome 47 . SBS89 is found in normal colorectal crypts, but not thus far in other cell types, is of unknown origin but may plausibly also be caused by a microbiome-derived mutagen 31 . SBS17b is observed in chemotherapy naïve oesophageal adenocarcinomas 48 and in tissues exposed to the chemotherapeutic agent 5-fluorouracil (5-FU) 43 , 49 . SBS93 was previously discovered in stomach and oesophageal cancers and is of unknown aetiology 41 . SBSC is a predominantly C > T signature previously reported in wild-type colorectal crypts 31 . Seven small insertion and deletion mutational signatures, ID1, ID2, ID5, ID14, ID18, IDA and IDB, were also extracted (Fig. 3 a, Extended Data Fig. 7–9). ID1 is characterised by insertions of T at T homopolymer tracts, is thought to arise from polymerase-related slippage of the replicated DNA strand and is the predominant indel signature in normal cells 43 . ID2 is characterised by T deletions at T homopolymer tracts, is thought to arise from polymerase-related slippage of the template DNA strand and is predominant in cancer genomes 50 . ID5 is an age-associated signature present in cancer and normal cells. ID14 is a signature common in colorectal cancer and was extracted from polyps from one individual (PD42778). ID18 is associated with colibactin exposure and SBS88 47 . IDA primarily consists of 1bp insertions at C homopolymer tracts and was only found in one individual (PD40734). IDB is characterised by 1bp deletions of T at non-homopolymer regions, and has previously been reported in wild-type colorectal crypts 31 . The repertoire and contributions of mutational signatures in normal crypts from individuals with FAP/AFAP were similar to those in normal crypts from wild-type individuals and included SBS1, SBS5, SBS18, SBS88, ID1, ID2, ID5 and ID18. The single exception was individual PD40734 who showed elevated burdens and proportions of SBS5 (or a signature resembling it) in all normal crypts, and the presence of an indel signature, IDA, not found in any other individual. PD40734 had not been exposed to cancer chemotherapy or to any other known systemic mutagen. The cause of the unusual signature repertoire in this individual is unknown, but IDA has been previously reported in an individual with defective base excision repair 33 . Differences in signature contributions were observed between normal crypts and crypts with somatic APC mutations in ACF (Fig. 3 b-c). These included modest additional burdens of SBS18 (128, 95% C.I., 38–218, P-adjust = 1.0 × 10 − 2 ) and ID1 (29, 95% C.I., 21–37, P-adjust = 5.0 × 10 − 11 ). In crypts with somatic APC mutations from diminutive and large adenomas there were greater increases in SBS18 (diminutive polyps: 552, 95% C.I., 460–644, P-adjust = 8.8 × 10 − 27 ; large polyps: 1868, 95% C.I., 1680–2055, P-adjust = 2.0 × 10 − 59 ) and ID1 (diminutive polyps: 97, 95% C.I., 85–108, P-adjust = 2.5 × 10 − 45 ; large polyps: 420, 95% C.I., 393–448, P-adjust = 1.2 × 10 − 95 ) mutation burdens than in ACF, and the burdens of SBS1 (diminutive polyps: 563, 95% C.I., 426–699, P-adjust = 3.4 × 10 − 14 ; large polyps: 1286, 95% C.I., 1018–1554, P-adjust = 1.4 × 10 − 18 ), SBS5 (diminutive polyps: 399, 95% C.I., 280–518, P-adjust = 5.7 × 10 − 10 ; large polyps: 1711, 95% C.I., 1439–1982, P-adjust = 3.8 × 10 − 29 ), ID2 (diminutive polyps: 52, 95% C.I., 44–60, P-adjust = 1.3 × 10 − 30 ; large polyps: 264, 95% C.I., 245–282, P-adjust = 1.2 × 10 − 95 ) and ID5 (diminutive polyps: 24, 95% C.I., 17–32, P-adjust = 8.8 × 10 − 9 ; large polyps: 81, 95% C.I., 65–96, P-adjust = 3.0 × 10 − 22 ) were also increased (Fig. 3 c). In addition to the increased burdens of signatures present in normal crypts, SBS17b was found in a single large adenomatous polyp from PD42778 (who had not been treated with 5-fluorouracil) and SBS93 was found in the same large polyp and a diminutive polyp with two somatic APC mutations from the same individual. Crypt phylogenies Using somatic mutations as unique markers, we constructed phylogenetic trees of crypts for each individual in this study 51 (Fig. 4 , Extended Data Fig. 10). As in wild-type individuals, morphologically normal crypts from individuals with FAP/AFAP, whether physically adjacent or distant from one another in the epithelial lining, typically shared only a small number of somatic mutations, indicating a common ancestor that existed in the distant past, possibly during embryonic development. By contrast, crypts from each ACF with the same APC somatic mutation usually shared substantial numbers of somatic mutations indicating comparatively recent common ancestors, as did crypts from adenomatous polyps. Adenomatous polyps with recent common ancestors carrying more than one driver mutation had longer trunks (reflecting a larger number of mutations in the most recent common ancestor) than ACF or adenomatous polyps with just a single somatic APC mutation ( P = 0.0028). However, in polyps with just a single somatic mutation in APC , the higher total crypt mutation burdens than those found in ACF could not be explained by either truncal or branch mutations alone (branch mutations being defined by mutations not in the most recent common ancestor). The genome-wide somatic mutation burdens in these phylogenies can serve as molecular clocks allowing estimates of the timing of occurrence of the somatic APC mutations during the life of the individual 42 , and thus the length of time that individual neoplastic lesions have taken to develop. Some APC somatic mutations appear to have arisen early in life, with cells carrying them progressing to a limited extent. For example, the somatic APC R1114* mutation in an ACF consisting of two bifurcating crypts (PD42778D_ACF_007_2) from a 44-year-old with FAP is likely to have occurred within the first five years of life. Therefore, over a period of at least 39 years, the neoplastic clone initiated by this mutation has only developed into an ACF (Fig. 4 a). Similarly, in this 44-year-old individual, a somatic APC loss of heterozygosity event (PD42778D_PLP_003_1) is likely to have occurred by age five years and over 39 years has only progressed into a diminutive adenomatous polyp. DISCUSSION We have investigated the earliest stages of colorectal neoplastic progression due to biallelic APC mutation by whole genome sequencing of 279 crypts microdissected from normal colorectal epithelium, ACF and adenomas from individuals with Familial Adenomatous Polyposis. The results inform on the earliest stages of the most common pathway of sporadic colorectal carcinogenesis which is also thought to be initiated by APC inactivation. Almost all histologically normal crypts remained heterozygous for the inherited APC gene mutation and exhibited mutation rates and signatures indistinguishable from those found in normal crypts from wild type individuals. The results show that an increased genome-wide mutation rate in normal colorectal cells is not a factor contributing to the high risk of neoplastic change in cells and individuals with heterozygous APC inactivating mutations. They are, therefore, consistent with the orthodoxy that APC inactivating mutations act in a recessive manner and engender biological consequences for an individual cell only if biallelic. This overall pattern is similar to that observed in the normal colorectal epithelium of individuals with colorectal cancer predisposition due to inherited mutations in the DNA Mismatch Repair genes MSH2 and MLH1 , in whom the large majority of crypts show normal mutation rates and signatures, with only rare crypts showing loss of the wild type allele and initiation of neoplastic change 34 . However, it differs from the pattern in individuals with inherited monoallelic mutations in genes encoding the replicative polymerases POLE and POLD1, which cause DNA proofreading defects and in whom all normal colorectal crypts, and all other normal cell types thus far sampled, show elevated mutation rates with distinctive mutational signatures 32 . It also differs from the pattern in individuals with inherited biallelic mutations in MUTYH in whom all normal colorectal crypts also show elevated mutation rates and abnormal signatures 33 . Collectively, these studies depict the different ways in which inherited mutations can influence the rate of neoplastic change in colorectal cells through mutagenesis of the somatic genome. One FAP patient with a germline ~ 5Mb deletion on chromosome 5 including all of APC (PD40734, chr5: 110,610,575 − 115,126,084) exhibited a subtly different somatic mutation landscape in all analysed crypts due to the elevated burden of an SBS5-like mutational signature and an indel mutational signature, IDA, not found in wild type individuals. The cause of this unusual mutation pattern is unclear. It could be due to haploinsufficiency of another gene within the inherited 5Mb deletion (although there are no striking candidates), coincidental occurrence of a constitutional mutation in a gene elsewhere in the genome that alters mutagenesis (although, if so, we have been unable to identify it), or an exogenous mutagenic exposure (although there is no record of this and the pattern of mutations does not correspond with the mutational signatures of any of the known lifestyle, environmental or iatrogenic mutagenic exposures thus far catalogued). One out of 110 (1%) normal crypts from individuals with FAP/AFAP carried a somatic APC mutation. This compares to 0/1,974 whole or targeted genome sequenced crypts from wild-type individuals 31 . Whether this represents a biological difference or chance is uncertain. Nevertheless, it is plausible that a colorectal stem cell from an individual with FAP/AFAP carrying biallelic APC inactivating mutations has selective advantage in colonising a crypt compared to other crypt stem cells with just the monoallelic constitutional APC mutation 52 , whereas a colorectal stem cell from a wild-type individual carrying a monoallelic somatic APC inactivating mutation, may have little selective advantage over biallelic wild-type cells. Although detection of a single normal crypt with biallelic APC mutations does not permit robust estimates of their prevalence, the result suggests that a colorectum of ~ 15,000,000 crypts in an adult with FAP could harbour tens of thousands of such crypts. In contrast to normal cells (of which < 1% harbour somatic APC mutations), 9/18 (50%) ACF included crypts with a somatic APC mutation. Thus, by virtue of their biallelic mutations in APC , some ACF represent an early stage of colorectal neoplasia. Early studies found few APC mutations in ACF and predominantly in type C ACF with dysplasia 22 , 23 , 25 . Here, we show APC mutations in type A-I and A-II ACF characterised just by crypt fission and early regenerative changes. Whether ACF are actually the progenitor lesions from which adenomatous polyps develop remains uncertain. However, their size, microscopic morphology, and non-polypoid structure have previously been suggested as indications that they are an intermediate stage of neoplastic transformation between normal crypts and diminutive polyps 19 , 20 , 21 . This conclusion is supported by results from this study in which the pattern of increased mutation burdens of specific ubiquitous mutational signatures (SBS18, ID1) in ACF are further increased in diminutive polyps alongside increased loads of other ubiquitous signatures (SBS1, SBS5, ID2). It is also compatible with the appearance of additional driver mutations in APC , KRAS and TP53 in diminutive polyps which are not seen in ACF. The changes in the overall somatic mutation landscapes of ACF may, at least in part, be due to their stem cells having undergone more mitotic divisions than the stem cells of normal crypts (potentially resulting in higher ID1 burdens) but may also reflect changes in cell function, state or microenvironment as suggested by increased reactive oxygen species damage (resulting in increased SBS18 burdens). The extent to which these accelerated mutational processes contribute to acquisition of further driver mutations and continuation of cancer evolution is uncertain. The phylogenetic analyses using somatic mutations of individual crypts from ACF in this study provide insight into their natural history. Several ACF and diminutive adenomatous polyps acquired a somatic APC mutation within the first decade of life. Despite the early advent of APC loss of function, progression over decades to cancer was sometimes absent. Thus, stem cells with biallelic inactivating APC mutations can remain in a relatively early stage of neoplastic change for an extended duration and, notably, without obvious genome instability. The findings support the possibility that some cancers slowly evolve for several decades following their initiating somatic cancer gene driver mutation, and highlight the relatively low chance of any individual neoplastic clone ultimately developing into a cancer. METHODS Ethics and overview Usage of patient samples in this study were approved by IRB of the University of Hong Kong/Hospital Authority Hong Kong West Cluster (UW14-257) and West Midlands - Coventry and Warwickshire Research Ethics Service with United Kingdom Research Ethics Committee (REC) reference 17/WM/0295 (PD40734, PD42778, PD44720, PD44721), as well as East of Scotland Research Ethics Service with REC reference 18/ES/0133 (PD44892, PD44893, PD44894, PD44895, PD44896, PD44897, PD44898, PD44899, PD44900, PD44901, PD44902). The first cohort (PD40734, PD42778, PD44720, PD44721) consists of patients with a genetic diagnosis of FAP and operated on in Queen Mary Hospital of Hong Kong for total colectomy or total proctocolectomy. Specimens were collected fresh from the operating theatre and examined by a pathologist. Part of large polyps and random blocks of colon containing variable sizes of small polyps as well as normal colon mucosa were sampled, snap frozen and kept in -80 o C until use. The second cohort (PD44892, PD44893, PD44894, PD44895, PD44896, PD44897, PD44898, PD44899, PD44900, PD44901, PD44902) consists of FAP patients from UK and biopsies were taken from prophylactic colectomy. All samples were collected with informed consent from the patients. Laser-microdissection and low-input whole-genome sequencing We followed the standard protocol established at the Wellcome Sanger Institute for tissue processing, laser-microdissection and low-input whole-genome sequencing 30 . Fresh frozen blocks were cryostat sectioned to 20 mm, fixed to 4 mm PEN membrane slides (11600288, Leica) and stained with hematoxylin and eosin (H&E). Crypts were isolated using laser capture microscopy (LMD7000, Leica) and collected in separate wells of a 96-well plate. Collected samples were lysed using ARCTURUS PicoPure DNA extraction kit (Applied Biosystems) according to the manufacturer’s instructions. DNA library concentration was measured following library preparation and used to guide the choice of samples subject to DNA sequencing. The minimum library concentration was 5 ng/µl, and libraries with > 15 ng/µl were preferably picked for an aim of ~ 30x coverage. Paired-end sequencing reads (150 bp) were generated on Illumina NovaSeq platform and aligned to the human reference genome (NCBI build38) using BWA-MEM (versions 0.7.17). Crypt classification Normal crypts, ACF and adenomatous polyps were classified according to their microscopic morphologies after H&E staining. Normal crypts had normal architecture lined by orderly differentiated cells. ACF were non-polypoid lesions including either a singleton or cluster of branching/budding crypts and/or cluster of enlarged crypts with wider calibre, tortuosity and/or changes in proportion or shapes of cell types. Nanozoomer-scanned images for all crypts from ACF were reviewed and scored without knowledge of the mutation results by two clinically practising gastrointestinal histopathologist (S.Y.L. and S.T.Y.) and with discrepancy discussed and reviewed to arrive at a consensus. Presence or absence of dysplasia, serration, crypt elongation (> 500 um length), crypt dilatation (> 100 um diameter) and branching were recorded. Their levels of regenerative features were graded as follow: 0, quiescent, small basally located nuclei even at crypt base with no mitosis; 1, mild elongation and crowding of nuclei with rare mitosis restricted to basal 1/3 of crypt; 2, elongation and crowding of nuclei with mitosis extending to middle third of crypt, with evidence of maturation on top 1/3 of crypt. Each group of ACF was then classified based on the most advanced lesion in the group using established standard from previous literature 18 . Whilst there were no type B lesions (serrated type), type A lesions were further distinguished into two groups, type A lesion with no to minimal regeneration (type A-I) and Type A lesion with moderate to marked regeneration (type A-II). Representative pathology images of different types of ACF and the one normal crypt with a somatic APC nonsense mutation, along with their pathology grading can be found in Extended Data Fig. 2 . Individual crypts microdissected from polyps were also graded based on whether they correspond to an adenoma, ACF or normal. The ACF were graded using similar criteria as above. Single-base substitution calling and phylogenetic tree building Cancer Variants through Expectation Maximization (CaVEMan) 53 (cgpCaVEMan version 1.15.2) and Strelka2 54 (version 2.9.10) were used to call single-base somatic substitutions against a matched blood sample from the same individual. For individuals without blood samples, we first called variants against an in silico human reference genome, reconstructed the phylogeny, and called variants against a phylogenetically unrelated crypt. The consensus variants called by both CaVEMan and Strelka2 further went through a series of postprocessing filters and the final substitutions were used for constructing phylogenies: (1) The first filter removed mapping artifacts associated with BWA-MEM as follows: the median alignment score of reads supporting a mutation should be greater than or equal to 140, and fewer than half of these reads should be clipped. (2) The second filter was applied to remove artifacts that are associated with the LCM library preparation: https://github.com/cancerit/hairpin2 . (3) A binomial distribution test was applied to remove possible remaining germline variants and a beta-binomial test was applied to filter out the common low frequency artifacts. Phylogenetic trees then were generated from the filtered substitutions using a maximum parsimony algorithm (MPBoot version 1.1.0) 55 . Substitutions were mapped onto tree branches using a maximum likelihood approach (TreeMut version 1.1) and visualized using ggtree (version 3.3.1) 56 and ape (version 5.6.1) 57 . The code for this part of the pipeline was integrated into Sequoia 51 : https://github.com/TimCoorens/Sequoia . Indel calling and mapping to the phylogenetic trees Pindel 58 , 59 (cgpPindel version 3.10.0) and Strelka2 54 (version 2.9.10) were used for indel calling. Consensus indels from the two algorithms that possessed a minimum quality score of ≥ 300 at positions covered by at least 15 reads were kept and subject to the same binomial and beta-binomial filters as single-base substitutions. Subsequently, indels were mapped to the corresponding phylogenetic trees reconstructed from single-based substitutions using a maximum likelihood approach (TreeMut version 1.1) and visualized using ggtree (version 3.3.1) 56 and ape (version 5.6.1) 57 . Annotation of putative driver mutations Mutations were overlapped with a known list of genes under positive selection in human cancers 60 and were further annotated using the cBioPortal MutationMapper cancer hotspot mutation database v5.1.7 ( http://www.cbioportal.org/mutation_mapper ) 61 . Known cancer mutation hotspots and protein-truncating variants in tumour suppressor genes were reported as putative driver mutations. Screening for subclonal APC mutations To inspect possible subclonal APC mutations, all variants called by cgpCaVEMan and cgpPindel were intersected with the APC gene and annotated by ANNOVAR 62 (version 2020-06-08, database refGene,dbnsfp41c,avsnp150,cosmic89), including those that did not pass the two algorithms’ quality control filters for reasons such as insufficient number of reads supporting the mutation. We then checked the annotation of all APC mutations to see if there were any nonsense, splice site or hotspot missense mutations (hotspot annotation from cBioPortal MutationMapper cancer hotspot mutation database). Copy number variant calling Somatic copy-number variants (CNVs) were called using the Allele-Specific Copy number Analysis of Tumours (ASCAT) algorithm 63 as part of the ascatNGS package (version 4.3.2) 64 . ASCAT was run with default parameters with the exception of a segmentation penalty of 100. A bespoke filtering algorithm - ascatPCA - was used to reduce the number of false-positive calls that can arise when analysing genome sequences from low-input WGS ( https://github.com/hj6-sanger/ascatPCA ). ascatPCA extracts a noise profile by aggregating the LogR ratio from across a panel of unrelated normal samples and subtracts this signature from that observed in the sample being analysed using principal component analysis. Structural variant and retrotransposition calling Somatic structural variants (SVs) and retrotranspositions were called using the Genomic Rearrangement Identification Software Suite (GRIDSS) 65 with default settings (version 2.9.4). All variants were confirmed by visual inspection and by checking if they fit the distribution expected based on the SNV-derived phylogenetic tree. Specifically, SVs (> 1kb) and retrotranspositions with QUAL > = 250 were included. For SVs smaller than 30kb, SVs with QUAL > = 300 were only included. Furthermore, SVs required breakpoint assembly on both sides with at least four discordant and two reads supporting them. Where breakpoints were imprecise, defined as the start and end positions being > 10bp apart, the SV was filtered out. We further filtered out SVs for which the standard deviation of the alignment positions at either end of the discordant read pairs was smaller than five. We also applied this standard deviation filter to single-break end calls in calling retrotranspositions and used Repeatmasker 66 (version 4.1.7) to annotate repetitive elements. Modelling disease effects on mutational burdens Raw mutation burden was normalised by sensitivity of mutation detection in each crypt: we calculated the possibility of detecting a variant with a given coverage, median variant allele frequency of the sample and algorithm settings (at least four mutant reads required CaVEMan and five for Pindel) by running 100,000 simulations. Each simulation was a set of Bernoulli tests with a success probability equal to the median VAF of the sample, and the number of tests was drawn from a Poisson-distributed depth given the median coverage. Crypts with coverages less than 10 were excluded in downstream analysis for mutation rate and disease effect. With corrected mutation counts, we then fitted linear mixed-effects models using the nlme R package (version 3.1.166) to estimate the contribution of age, constitutional APC mutation and disease condition to mutation burden. Crypts from ACF and adenomatous polyps without a somatic APC inactivating mutation were excluded in estimation of disease effect. Because crypts from the same individual/ACF/adenomatous polyps are not independent, we controlled these factors by estimating a random effect for each patient/biopsy/group of crypts (crypts from the same ACF or adenomatous polyps belong to the same group). ANOVA tests between models were used to test whether modifications on fixed and random effects should be included into the model or not. For the final models, multiple test correction with the Benjamini–Hochberg method was applied before reporting the significant coefficients. The workflow and code for this part of analysis will be put in: https://github.com/YichenWang1/FAP_colon . Mutational signature extraction and attribution Mutational signatures were first extracted using HDP 40 (version 0.1.5) without hierarchy and prior information. Mutations mapped to branches on the phylogenetic tree were used as input to avoid counting the same mutation within one patient multiple times. To avoid overfitting, we only kept branches with > 50 single-base substitutions (or > 30 for indels) as input. The HDP was run in 10 independent chains for 120,000 iterations and with a burn-in of 20,000. Identified HDP signatures were compared against COSMIC reference signatures 67 as well as signatures identified in normal colorectal crypts 31 . For confirmation of potential novel signatures, HDP extraction results were compared with mutational signature extraction results from another independent method SigProfilerExtractor 41 (version 1.1.24) with parameters min_sigs = 2, max_sigs = 20 for substitutions (8 for indels). Non-COSMIC signature SBSC, IDA and IDB were extracted by both algorithms and had been reported in past studies 31 , 33 , therefore were also kept in the final list of signatures. For HDP signatures with ≥ 0.9 cosine similarity with a reference signature, branches with such HDP signatures were thought to have the corresponding reference signatures. Remaining HDP signatures were deconvoluted into a shortlist of candidate reference signatures previously found in colorectal cancer and normal colorectal crypts 31 , 33 , 32 , 35 , 34 , 50 (SBS1, SBS2, SBS5, SBS13, SBS17a, SBS17b, SBS18, SBS28, SBS31, SBS35, SBS88, SBS89, SBS93 and ID1, ID2, ID5, ID14) using expectation maximization. A second round of expectation maximization was further run on reference signatures with > 10% contributions for each HDP signature to reduce overfitting. Branches with a HDP signature that contributed to more than 5% of mutations were thought to possess the corresponding deconvoluted reference signatures for that HDP signatures, unless the length of the branch was ≤ 200 for single-base substitutions (or ≤ 50 for indels), in which case the HDP signature need to contribute to at least 5% of mutations in at least one longer branch from the same individual to be allowed in. In this way, the final mutational signatures present in each branch were defined, and the final proportion of reference signatures was estimated using sigfit (version 2.0) 68 . If any reference signatures contributed no more than 5% of total mutations of the branch, the signatures were removed followed by a second round of fitting using the remaining reference signatures. For the very short branches without an HDP signature extraction result, their signature attributions were approximated by their immediate descendent/ascendent branch when reconstructing the signature burdens for the whole crypt. The workflow and code for this part of analysis will be put in: https://github.com/YichenWang1/FAP_colon Timing the onset for somatic APC inactivation Burdens of lock-like signature SBS5 was used to estimate the age of the somatic APC inactivation. A liner mixed-effect models with age as fixed effect, patient as random effect was used to estimate SBS5 mutation rates in all normal crypts (nlme R package version 3.1.166). SBS5 mutation rate for each individual was adjusted by the patient random effect term, and the age of onset for the earliest common ancestor with APC inactivation was estimated using the SBS5 burden on the trunk of the phylogenetic tree (where the APC somatic mutation was placed on) divided by the SBS5 mutation rate of the corresponding individual. This estimation is an upper limit because the crypt fission might have happened later than the APC inactivation. Declarations Competing interests P.J.C. is a co-founder and CSO of Quotient Therapeutics. M.R.S, and I.M. are co-founders of Quotient Therapeutics. The other authors declare no competing interests. Author contributions P.S.R., S.Y.L. and M.R.S. conceived the study design, Y.W., P.S.R., S.Y.L. and M.R.S designed the analysis strategy. M.R.S. and S.Y.L. obtained funding. S.Y.L., S.T.Y., H.H.N.Y., C.C.F., A.S.Y.C., A.K.W.C., W.Y.T., J.R.S., H.D.W., and L.T. recruited participants, collected samples and curated sample and clinical data. S.Y.L, S.T.Y. and O.T.G. performed histological analysis of samples. P.S.R., B.C.H.L., Y.H. and K.R. undertook laboratory work. Y.W., P.S.R., H.J. performed data analysis with help and input from S. F., T.H.H.C., A.R.J.L., and S.O.. L.A. contributed to sample management. H.L.-S. and S.O. contributed and analysed the control data. M.R.S., P.J.C. and I.M. oversaw statistical analysis. M.R.S. and S.Y.L oversaw the study. All authors were involved in the preparation and review of the manuscript. Acknowledgement We thank the staff of the Wellcome Sanger Institute Sample Logistics, Genotyping, Pulldown, Sequencing and Informatics facilities for their many contributions, especially L. O’Neill, C. Latimer, K. Roberts for their support with sample management and laboratory work. We thank S. Moody, A. P. Le for discussion of the results. We thank Dorothy H.T. Cheng for support with patient coordination and the clinicians in Hong Kong Hospital Authority for clinical care. We thank all the patients and their families, without their support this work would not have been possible. This research is supported by core funding from the Wellcome Trust (206194), the Kadoorie Charitable Foundation, the Hong Kong Cancer Fund, the Centre for Oncology and Immunology under the Health@InnoHK Initiative funded by the Innovation and Technology Commission, the Government of the Hong Kong SAR, China. Y.W., T.H.H.C., A.R.J.L., and S.O. was supported by Wellcome Ph.D. Studentships and P.S.R. by a Wellcome Clinical Ph.D. fellowship. L.E.T was supported by a Health Fellowship from the Welsh Government (HF-14-10, Health and Care Research Wales). J.R.S. was supported by a grant from Health and Care Research Wales. Data availability DNA sequencing data generated for this study are deposited in the European Genome-Phenome Archive (EGA) with accession code EGAD00001015471, and aligned BAM files with accession code EGAD00001015477. The guidelines for patient consent prevent the derived data files from being dispersed by open access. To ensure the data is used for academic and research purposes, controlled access of the data will be available indefinitely upon request made to the WTSI CGP Data access committee. Code availability Code required to reproduce the analyses in this paper is available online. Mutation-calling algorithms are available through GitHub (https://github.com/cancerit). Variant calling filters can be found at https://github.com/cancerit/hairpin2 and https://github.com/hj6-sanger/ascatPCA). The phylogeny reconstruction pipeline is available at https://github.com/TimCoorens/Sequoia. Code for mutational signature analysis and other customised analyses in this study is available online at https://github.com/YichenWang1/FAP_colon. References Luebeck, E. G. & Moolgavkar, S. H. Multistage carcinogenesis and the incidence of colorectal cancer. Proc. Natl. Acad. Sci. U.S.A. 99 , 15095–15100 (2002). Fearon, E. R. & Vogelstein, B. A genetic model for colorectal tumorigenesis. Cell 61 , 759–767 (1990). Vogelstein, B. et al. Genetic Alterations during Colorectal-Tumor Development. N Engl J Med 319 , 525–532 (1988). Lamlum, H. et al. APC mutations are sufficient for the growth of early colorectal adenomas. Proc. Natl. Acad. Sci. U.S.A. 97 , 2225–2228 (2000). Li, J. et al. Genomic and transcriptomic profiling of carcinogenesis in patients with familial adenomatous polyposis. Gut 69 , 1283–1293 (2020). Rashid, M. et al. Adenoma development in familial adenomatous polyposis and MUTYH ‐associated polyposis: somatic landscape and driver genes. The Journal of Pathology 238 , 98–108 (2016). Gausachs, M. et al. Mutational Heterogeneity in APC and KRAS Arises at the Crypt Level and Leads to Polyclonality in Early Colorectal Tumorigenesis. Clinical Cancer Research 23 , 5936–5947 (2017). Klos, C. L. & Dharmarajan, S. Polyp Genetics. Clin Colon Rectal Surg 29 , 289–295 (2016). Fearon, E. R. Molecular Genetics of Colorectal Cancer. Annu. Rev. Pathol. Mech. Dis. 6 , 479–507 (2011). The Cancer Genome Atlas Network. Comprehensive molecular characterization of human colon and rectal cancer. Nature 487 , 330–337 (2012). Cornish, A. J. et al. The genomic landscape of 2,023 colorectal cancers. Nature 633 , 127–136 (2024). Brenner, H., Kloor, M. & Pox, C. P. Colorectal cancer. The Lancet 383 , 1490–1502 (2014). Lu, Z. et al. Polyclonal-to-monoclonal transition in colorectal precancerous evolution. Nature 636 , 233–240 (2024). Zaffaroni, G. et al. Updated European guidelines for clinical management of familial adenomatous polyposis (FAP), MUTYH-associated polyposis (MAP), gastric adenocarcinoma, proximal polyposis of the stomach (GAPPS) and other rare adenomatous polyposis syndromes: a joint EHTG-ESCP revision. British Journal of Surgery 111 , znae070 (2024). Dunne, P. D. & Arends, M. J. Molecular pathological classification of colorectal cancer—an update. Virchows Arch 484 , 273–285 (2024). Yan, H. H. N. et al. RNF43 germline and somatic mutation in serrated neoplasia pathway and its association with BRAF mutation. Gut 66 , 1645–1656 (2017). Roncucci, L., Stamp, D., Medline, A., Cullen, J. B. & Robert Bruce, W. Identification and quantification of aberrant crypt foci and microadenomas in the human colon. Human Pathology 22 , 287–294 (1991). Di Gregorio, C. et al. Histology of aberrant crypt foci in the human colon. Histopathology 30 , 328–334 (1997). Takayama, T. et al. Aberrant Crypt Foci of the Colon as Precursors of Adenoma and Cancer. N Engl J Med 339 , 1277–1284 (1998). Cheng, L. Aberrant crypt foci as microscopic precursors of colorectal cancer. WJG 9 , 2642 (2003). Orlando, F. A. et al. Aberrant crypt foci as precursors in colorectal cancer progression. Journal of Surgical Oncology 98 , 207–213 (2008). Otori, K. et al. Infrequent somatic mutation of the adenomatous polyposis coli gene in aberrant crypt foci of human colon tissue. Cancer 83 , 896–900 (1998). Jen, J., Powell, S. M., Smith, K. J., Hamilton, S. & Vogeistein, B. Molecular Determinants of Dysplasia in Colorectal Lesions. Smith, A., Stern, H. & Ga, S. Somatic APC and K-ras Codon 12 Mutations in Aberrant Crypt Foci from Human Colons’. Takayama, T. et al. Analysis of K-ras, APC, and β-catenin in aberrant crypt foci in sporadic adenoma, cancer, and familial adenomatous polyposis. Gastroenterology 121 , 599–611 (2001). Losi, L., Roncucci, L., Di Gregorio, C., De Leon, M. P. & Benhattar, J. K-ras AND p53 mutations in human colorectal aberrant crypt foci. J. Pathol. 178 , 259–263 (1996). Nucci, M. R., Robinson, C. R., Longo, P., Campbell, P. & Hamilton, S. R. Phenotypic and genotypic characteristics of aberrant crypt foci in human colorectal mucosa. Human Pathology 28 , 1396–1407 (1997). Pretlow, T. P., Brasitus, T. A., Fulton, N. C., Cheyer, C. & Kaplan, E. L. K-ras Mutations in Putative Preneoplastic Lesions in Human Colon. JNCI Journal of the National Cancer Institute 85 , 2004–2007 (1993). Luo, L. et al. Loss of heterozygosity in human aberrant crypt foci (ACF), a putative precursor of colon cancer. Carcinogenesis 27 , 1153–1159 (2006). Ellis, P. et al. Reliable detection of somatic mutations in solid tissues by laser-capture microdissection and low-input DNA sequencing. Nat Protoc 16 , 841–871 (2021). Lee-Six, H. et al. The landscape of somatic mutation in normal colorectal epithelial cells. Nature 574 , 532–537 (2019). Robinson, P. S. et al. Increased somatic mutation burdens in normal human cells due to defective DNA polymerases. Nat Genet 53 , 1434–1442 (2021). Robinson, P. S. et al. Inherited MUTYH mutations cause elevated somatic mutation rates and distinctive mutational signatures in normal human cells. Nat Commun 13 , 3949 (2022). Lee, B. C. H. et al. Mutational landscape of normal epithelial cells in Lynch Syndrome patients. Nat Commun 13 , 2710 (2022). Olafsson, S. et al. Somatic Evolution in Non-neoplastic IBD-Affected Colon. Cell 182 , 672-684.e11 (2020). Rosenberg, D. W. et al. Mutations in BRAF and KRAS Differentially Distinguish Serrated versus Non-Serrated Hyperplastic Aberrant Crypt Foci in Humans. Cancer Research 67 , 3551–3554 (2007). Segditsas, S. et al. APC and the three-hit hypothesis. Oncogene 28 , 146–155 (2009). Novelli, M. R. et al. Polyclonal Origin of Colonic Adenomas in an XO/XY Patient with FAP. Science 272 , 1187–1190 (1996). Sadien, I. D. et al. Polyclonality overcomes fitness barriers in Apc-driven tumorigenesis. Nature 634 , 1196–1203 (2024). Li, Y. et al. Patterns of somatic structural variation in human cancer genomes. Nature 578 , 112–121 (2020). Islam, S. M. A. et al. Uncovering novel mutational signatures by de novo extraction with SigProfilerExtractor. Cell Genomics 2 , 100179 (2022). Alexandrov, L. B. et al. Clock-like mutational processes in human somatic cells. Nat Genet 47 , 1402–1407 (2015). Moore, L. et al. The mutational landscape of human somatic and germline cells. Nature 597 , 381–386 (2021). Kucab, J. E. et al. A Compendium of Mutational Signatures of Environmental Agents. Cell 177 , 821-836.e16 (2019). Coorens, T. H. H. et al. Inherent mosaicism and extensive mutation of human placentas. Nature 592 , 80–85 (2021). Wang, Y. et al. APOBEC mutagenesis is a common process in normal human small intestine. Nat Genet (2023) doi:10.1038/s41588-022-01296-5. Pleguezuelos-Manzano, C. et al. Mutational signature in colorectal cancer caused by genotoxic pks+ E. coli. Nature 580 , 269–273 (2020). Secrier, M. et al. Mutational signatures in esophageal adenocarcinoma define etiologically distinct subgroups with therapeutic relevance. Nat Genet 48 , 1131–1141 (2016). Christensen, S. et al. 5-Fluorouracil treatment induces characteristic T>G mutations in human cancer. Nat Commun 10 , 4571 (2019). Alexandrov, L. B. et al. The repertoire of mutational signatures in human cancer. Nature 578 , 94–101 (2020). Coorens, T. H. H. et al. Reconstructing phylogenetic trees from genome-wide somatic mutations in clonal samples. Nat Protoc (2024) doi:10.1038/s41596-024-00962-8. Vermeulen, L. et al. Defining Stem Cell Dynamics in Models of Intestinal Tumor Initiation. Science 342 , 995–998 (2013). Jones, D. et al. cgpCaVEManWrapper: Simple Execution of CaVEMan in Order to Detect Somatic Single Nucleotide Variants in NGS Data. Current Protocols in Bioinformatics 56 , 15.10.1-15.10.18 (2016). Kim, S. et al. Strelka2: fast and accurate calling of germline and somatic variants. Nat Methods 15 , 591–594 (2018). Hoang, D. T. et al. MPBoot: fast phylogenetic maximum parsimony tree inference and bootstrap approximation. BMC Evol Biol 18 , 11 (2018). Yu, G. Using ggtree to Visualize Data on Tree‐Like Structures. Current Protocols in Bioinformatics 69 , (2020). Paradis, E. & Schliep, K. ape 5.0: an environment for modern phylogenetics and evolutionary analyses in R. Bioinformatics 35 , 526–528 (2019). Ye, K., Schulz, M. H., Long, Q., Apweiler, R. & Ning, Z. Pindel: a pattern growth approach to detect break points of large deletions and medium sized insertions from paired-end short reads. Bioinformatics 25 , 2865–2871 (2009). Raine, K. M. et al. cgpPindel: Identifying Somatically Acquired Insertion and Deletion Events from Paired End Sequencing. CP in Bioinformatics 52 , (2015). Martincorena, I. et al. Universal Patterns of Selection in Cancer and Somatic Tissues. Cell 171 , 1029-1041.e21 (2017). Gao, J. et al. Integrative Analysis of Complex Cancer Genomics and Clinical Profiles Using the cBioPortal. Science Signaling 6 , pl1–pl1 (2013). Wang, K., Li, M. & Hakonarson, H. ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data. Nucleic Acids Research 38 , e164–e164 (2010). Van Loo, P. et al. Allele-specific copy number analysis of tumors. Proceedings of the National Academy of Sciences 107 , 16910–16915 (2010). Raine, K. M. et al. ascatNgs: Identifying Somatically Acquired Copy‐Number Alterations from Whole‐Genome Sequencing Data. CP in Bioinformatics 56 , (2016). Cameron, D. L. et al. GRIDSS2: comprehensive characterisation of somatic structural variation using single breakend variants and structural variant phasing. Genome Biol 22 , 202 (2021). Tarailo‐Graovac, M. & Chen, N. Using RepeatMasker to Identify Repetitive Elements in Genomic Sequences. CP in Bioinformatics 25 , (2009). Sondka, Z. et al. COSMIC: a curated database of somatic variants and clinical data for cancer. Nucleic Acids Research 52 , D1210–D1217 (2024). Gori, K. & Baez-Ortega, A. Sigfit: Flexible Bayesian Inference of Mutational Signatures . http://biorxiv.org/lookup/doi/10.1101/372896 (2018) doi:10.1101/372896. Additional Declarations Yes there is potential Competing Interest. P.J.C. is a co-founder and CSO of Quotient Therapeutics. M.R.S, and I.M. are co-founders of Quotient Therapeutics. The other authors declare no competing interests. Supplementary Files SupplementaryTables.xlsx Supplementary Tables ExtendedDataFigs.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6581155","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":459402574,"identity":"ef321899-4595-4ede-bbb2-b792197f3bc2","order_by":0,"name":"Michael Stratton","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAu0lEQVRIiWNgGAWjYHACxgfIPANitDCjqCJKC5sEaVp0Zx9/Vvmj5o5dv/QBxg8/GA4bE9Ridi7H7DbPsWfJM/sSmCV7GA6bEdZyhoftNgPb4WSDMwwM0gwMh22I0ML+rPDHv8PJ9mcYmH8TqYXBjIG37bCdAQ8DG8gWohxmLM3bdzhB4gxjm2WPQToR3j/D/vDjj2+H7fl7mA/f+FFhbdhAUA8UJDYwMDYQGZFQYE+C2lEwCkbBKBhpAAAFXTdoG4coXQAAAABJRU5ErkJggg==","orcid":"https://orcid.org/0000-0001-6035-153X","institution":"Sanger Institute","correspondingAuthor":true,"prefix":"","firstName":"Michael","middleName":"","lastName":"Stratton","suffix":""},{"id":459402575,"identity":"ede0c259-1e68-4557-b285-fccc35d0f28c","order_by":1,"name":"Suet Yi Leung","email":"","orcid":"https://orcid.org/0000-0001-8614-4619","institution":"The University of Hong Kong","correspondingAuthor":false,"prefix":"","firstName":"Suet","middleName":"Yi","lastName":"Leung","suffix":""},{"id":459402576,"identity":"76580306-f184-4ff2-8040-2d0270fedfc0","order_by":2,"name":"Yichen Wang","email":"","orcid":"https://orcid.org/0000-0001-5122-7545","institution":"Wellcome Sanger Institute","correspondingAuthor":false,"prefix":"","firstName":"Yichen","middleName":"","lastName":"Wang","suffix":""},{"id":459402577,"identity":"c3b18218-3df3-4275-89a1-e7a5d8751aba","order_by":3,"name":"Philip Robinson","email":"","orcid":"https://orcid.org/0000-0002-6237-7159","institution":"Wellcome Sanger Institute","correspondingAuthor":false,"prefix":"","firstName":"Philip","middleName":"","lastName":"Robinson","suffix":""},{"id":459402578,"identity":"2c465df7-9b7e-450d-9213-1cc4533ce733","order_by":4,"name":"Laura Thomas","email":"","orcid":"https://orcid.org/0000-0002-8621-5285","institution":"Swansea University","correspondingAuthor":false,"prefix":"","firstName":"Laura","middleName":"","lastName":"Thomas","suffix":""},{"id":459402579,"identity":"1574939f-ea84-4e1c-8ad4-aabdaca630b7","order_by":5,"name":"Hannah West","email":"","orcid":"","institution":"Institute of Medical Genetics, Cardiff University School of Medicine","correspondingAuthor":false,"prefix":"","firstName":"Hannah","middleName":"","lastName":"West","suffix":""},{"id":459402580,"identity":"3d7712e8-c511-4dfc-8556-e9078c8c2a12","order_by":6,"name":"Julian Sampson","email":"","orcid":"","institution":"Institute of Medical Genetics,Cardiff University School of Mediicine","correspondingAuthor":false,"prefix":"","firstName":"Julian","middleName":"","lastName":"Sampson","suffix":""},{"id":459402581,"identity":"d406c822-8b90-404e-8215-95c1eefbe196","order_by":7,"name":"Hyunchul Jung","email":"","orcid":"","institution":"Wellcome Sanger Institute","correspondingAuthor":false,"prefix":"","firstName":"Hyunchul","middleName":"","lastName":"Jung","suffix":""},{"id":459402582,"identity":"23e48a7d-8f18-481c-baa1-4d8a273f31c4","order_by":8,"name":"Stephen Fitzgerald","email":"","orcid":"","institution":"Wellcome Trust Sanger Institute, Wellcome Genome Campus, Hinxton, Cambridge, CB10 1SA, UK","correspondingAuthor":false,"prefix":"","firstName":"Stephen","middleName":"","lastName":"Fitzgerald","suffix":""},{"id":459402583,"identity":"72c7a2cd-cd50-462e-99f9-ce2fab009810","order_by":9,"name":"Henry Lee-Six","email":"","orcid":"https://orcid.org/0000-0003-4831-8088","institution":"Wellcome Sanger Institute","correspondingAuthor":false,"prefix":"","firstName":"Henry","middleName":"","lastName":"Lee-Six","suffix":""},{"id":459402584,"identity":"ad7ce632-b31b-40f0-b249-5cfef658963c","order_by":10,"name":"Tim Coorens","email":"","orcid":"https://orcid.org/0000-0002-5826-3554","institution":"EMBL-EBI","correspondingAuthor":false,"prefix":"","firstName":"Tim","middleName":"","lastName":"Coorens","suffix":""},{"id":459402585,"identity":"b6e98cd2-9476-4f6c-8fc0-2cbf222e8c8f","order_by":11,"name":"Sigurgeir Olafsson","email":"","orcid":"","institution":"Wellcome Sanger Institute","correspondingAuthor":false,"prefix":"","firstName":"Sigurgeir","middleName":"","lastName":"Olafsson","suffix":""},{"id":459402586,"identity":"2bacf94c-2dad-4f42-9a09-749092537be1","order_by":12,"name":"Andrew Lawson","email":"","orcid":"https://orcid.org/0000-0003-3592-1005","institution":"Wellcome Sanger Institute","correspondingAuthor":false,"prefix":"","firstName":"Andrew","middleName":"","lastName":"Lawson","suffix":""},{"id":459402587,"identity":"6663a5f8-f527-4c47-9455-d1ebbd140bbe","order_by":13,"name":"Yvette Hooks","email":"","orcid":"","institution":"Wellcome Trust Sanger Institute","correspondingAuthor":false,"prefix":"","firstName":"Yvette","middleName":"","lastName":"Hooks","suffix":""},{"id":459402588,"identity":"bf4b210b-bed5-48a7-9d04-74206662dfd1","order_by":14,"name":"Laura Allen","email":"","orcid":"","institution":"Wellcome Sanger Institute","correspondingAuthor":false,"prefix":"","firstName":"Laura","middleName":"","lastName":"Allen","suffix":""},{"id":459402589,"identity":"a82537be-57d9-4760-ba1b-2a936c694cff","order_by":15,"name":"Kirsty Roberts","email":"","orcid":"","institution":"Wellcome Sanger Institute","correspondingAuthor":false,"prefix":"","firstName":"Kirsty","middleName":"","lastName":"Roberts","suffix":""},{"id":459402590,"identity":"6b1f7fa6-26b7-44fc-b06a-0c0d2c5b4ecf","order_by":16,"name":"Siu Tsan Yuen","email":"","orcid":"","institution":"The University of Hong Kong","correspondingAuthor":false,"prefix":"","firstName":"Siu","middleName":"Tsan","lastName":"Yuen","suffix":""},{"id":459402591,"identity":"01bcc732-78e5-4030-957f-d00c61e9cf67","order_by":17,"name":"Helen Yan","email":"","orcid":"https://orcid.org/0000-0001-5693-8231","institution":"Department of Pathology, School of Clinical Medicine, The University of Hong Kong, Queen Mary Hospital, Pokfulam, Hong Kong","correspondingAuthor":false,"prefix":"","firstName":"Helen","middleName":"","lastName":"Yan","suffix":""},{"id":459402592,"identity":"c89a71b9-8e60-4957-983e-c31f848ee6d7","order_by":18,"name":"Chi Chung Foo","email":"","orcid":"https://orcid.org/0000-0001-8849-6597","institution":"The University of Hong Kong","correspondingAuthor":false,"prefix":"","firstName":"Chi","middleName":"Chung","lastName":"Foo","suffix":""},{"id":459402593,"identity":"fa8292e0-c58a-4e1b-a15e-e01147c19979","order_by":19,"name":"Bernard Lee","email":"","orcid":"","institution":"The University of Hong Kong","correspondingAuthor":false,"prefix":"","firstName":"Bernard","middleName":"","lastName":"Lee","suffix":""},{"id":459402594,"identity":"1211bafe-dc92-4622-b180-678eaf782519","order_by":20,"name":"Annie Chan","email":"","orcid":"","institution":"The University of Hong Kong","correspondingAuthor":false,"prefix":"","firstName":"Annie","middleName":"","lastName":"Chan","suffix":""},{"id":459402595,"identity":"b1a3ffa2-f2d6-4bea-8dcf-7311ebe41264","order_by":21,"name":"Anthony Chan","email":"","orcid":"","institution":"The University of Hong Kong","correspondingAuthor":false,"prefix":"","firstName":"Anthony","middleName":"","lastName":"Chan","suffix":""},{"id":459402596,"identity":"c2063314-8ff8-43cb-ad39-04c1bc196fdc","order_by":22,"name":"Wai Tsui","email":"","orcid":"https://orcid.org/0009-0001-4900-3929","institution":"The University of Hong Kong","correspondingAuthor":false,"prefix":"","firstName":"Wai","middleName":"","lastName":"Tsui","suffix":""},{"id":459402597,"identity":"d70a59c8-0892-4001-b9d4-9bc7aa2a6e38","order_by":23,"name":"Olivier Giger","email":"","orcid":"","institution":"The University of cambridge","correspondingAuthor":false,"prefix":"","firstName":"Olivier","middleName":"","lastName":"Giger","suffix":""},{"id":459402598,"identity":"37473201-ad97-4830-847b-e123521ae022","order_by":24,"name":"Inigo Martincorena","email":"","orcid":"https://orcid.org/0000-0003-1122-4416","institution":"Wellcome Sanger Institute","correspondingAuthor":false,"prefix":"","firstName":"Inigo","middleName":"","lastName":"Martincorena","suffix":""},{"id":459402599,"identity":"93b656ec-5463-4f10-bd94-aba7e6859571","order_by":25,"name":"Peter Campbell","email":"","orcid":"https://orcid.org/0000-0002-3921-0510","institution":"Cancer Genome Project, Wellcome Sanger Institute","correspondingAuthor":false,"prefix":"","firstName":"Peter","middleName":"","lastName":"Campbell","suffix":""}],"badges":[],"createdAt":"2025-05-02 23:55:11","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6581155/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6581155/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":83296190,"identity":"84c70216-f92d-430c-84cc-f1c867a1b058","added_by":"auto","created_at":"2025-05-22 13:53:41","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":110885,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSomatic mutations in \u003c/strong\u003e\u003cem\u003e\u003cstrong\u003eAPC\u003c/strong\u003e\u003c/em\u003e\u003cstrong\u003e and other cancer genes in FAP patients. a\u003c/strong\u003e, Representative images of H\u0026amp;E staining of crypts with different morphologies. \u003cstrong\u003eb, \u003c/strong\u003eDriver mutations in aberrant crypt foci (ACF) and adenomatous polyps from FAP patients (each column represents one ACF or adenomatous polyp: ACFs without any cancer gene mutation are not shown)).\u003cstrong\u003e \u0026nbsp;c, \u003c/strong\u003eDistribution through the \u003cem\u003eAPC\u003c/em\u003e gene coding sequence of somatic \u003cem\u003eAPC\u003c/em\u003e mutations in FAP patients.\u003c/p\u003e","description":"","filename":"1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6581155/v1/0c6c7cfd2fb985a105ae2eb0.jpg"},{"id":83296506,"identity":"abc3b9c4-6150-4d59-882e-98e4c7b3a821","added_by":"auto","created_at":"2025-05-22 14:01:42","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":100685,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMutation burdens in normal crypts, crypts from aberrant crypt foci and crypts from adenomatous polyps from FAP patients. a\u003c/strong\u003e, Single base substitution and indel burdens for \u003cem\u003eAPC\u003c/em\u003e\u003csup\u003emut/wt \u003c/sup\u003enormal crypts, \u003cem\u003eAPC\u003c/em\u003e\u003csup\u003emut/wt \u003c/sup\u003ebystander crypts and ACF, a single \u003cem\u003eAPC\u003c/em\u003e\u003csup\u003emut/mut \u003c/sup\u003enormal crypt, \u003cem\u003eAPC\u003c/em\u003e\u003csup\u003emut/mut \u003c/sup\u003ecrypts from ACF from FAP patients, and normal crypts from a control group of germline wild type individuals. \u003cstrong\u003eb, \u003c/strong\u003eSingle base substitution burdens in \u003cem\u003eAPC\u003c/em\u003e\u003csup\u003emut/wt \u003c/sup\u003enormal crypts, \u003cem\u003eAPC\u003c/em\u003e\u003csup\u003emut/wt \u003c/sup\u003ebystander crypts and ACF, \u003cem\u003eAPC\u003c/em\u003e\u003csup\u003emut/mut \u003c/sup\u003eACF, \u003cem\u003eAPC\u003c/em\u003e\u003csup\u003emut/mut \u003c/sup\u003ediminutive and large adenoma polyps for the four extensively sampled FAP patients.\u003cstrong\u003e \u0026nbsp;c, \u003c/strong\u003eIndel burdens in \u003cem\u003eAPC\u003c/em\u003e\u003csup\u003emut/wt \u003c/sup\u003enormal crypts, \u003cem\u003eAPC\u003c/em\u003e\u003csup\u003emut/wt \u003c/sup\u003ebystander crypts and ACF, \u003cem\u003eAPC\u003c/em\u003e\u003csup\u003emut/mut \u003c/sup\u003eACF, \u003cem\u003eAPC\u003c/em\u003e\u003csup\u003emut/mut \u003c/sup\u003ediminutive and large adenoma polyps for the four extensively sampled FAP patients.\u003cstrong\u003e \u0026nbsp;\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6581155/v1/873c9e7e27c55e75087a77a7.jpg"},{"id":83295547,"identity":"9f4a7c08-dec8-4d1e-b9fb-f68a8fc203e1","added_by":"auto","created_at":"2025-05-22 13:45:41","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":96201,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMutational signatures in FAP patients. a\u003c/strong\u003e, Substitution and Indel mutational signatures for normal crypts, \u003cem\u003eAPC\u003c/em\u003e\u003csup\u003emut/wt \u003c/sup\u003ebystander crypts and ACF, a single \u003cem\u003eAPC\u003c/em\u003e\u003csup\u003emut/mut \u003c/sup\u003enormal crypt, and \u003cem\u003eAPC\u003c/em\u003e\u003csup\u003emut/mut \u003c/sup\u003ecrypts from aberrant crypt foci from FAP patients, with normal crypts from control group. Grey bars: crypts with no somatic \u003cem\u003eAPC\u003c/em\u003e mutation. Brown bar:\u0026nbsp; \u003cem\u003eAPC\u003c/em\u003e\u003csup\u003emut/mut \u003c/sup\u003enormal crypt. \u003cstrong\u003e\u0026nbsp;\u003c/strong\u003eBlack bars: \u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cem\u003eAPC\u003c/em\u003e\u003csup\u003emut/mut \u003c/sup\u003ecrypts from\u003cstrong\u003e \u003c/strong\u003eaberrant crypt foci.\u003cstrong\u003e b, \u003c/strong\u003eSBS1, SBS5 and SBS18 burdens for \u003cem\u003eAPC\u003c/em\u003e\u003csup\u003emut/wt \u003c/sup\u003enormal crypts, \u003cem\u003eAPC\u003c/em\u003e\u003csup\u003emut/wt \u003c/sup\u003ebystander crypts and ACF, a single \u003cem\u003eAPC\u003c/em\u003e\u003csup\u003emut/mut \u003c/sup\u003enormal crypt, \u003cem\u003eAPC\u003c/em\u003e\u003csup\u003emut/mut \u003c/sup\u003ecrypts from ACF from FAP patients, and normal crypts from a control group of germline wild type individuals. \u003cstrong\u003ec, \u003c/strong\u003eDisease effects (additional mutations per crypt) on substitution and indel signature burdens with 95% CI. Dash line represents control level (0 mutations).\u003c/p\u003e","description":"","filename":"3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6581155/v1/e3da0f4030074b4ad9541eff.jpg"},{"id":83295549,"identity":"5d9062ce-43b9-49d2-81d3-238621d1f2fb","added_by":"auto","created_at":"2025-05-22 13:45:41","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":109775,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePhylogenetic trees for extensively sampled FAP patients. \u003c/strong\u003ePLP: adenomatous polyps. ACF: aberrant crypt foci. For an ACF, its bifurcating crypts are labelled ACF-BC and the rest are labelled ACF. For a polyp, its bystander crypts are labelled PLP-ACF or PLP-normal and the rest are labelled PLP. The prefix indicates the biopsy block ID. Crypts without a label are normal crypts.\u003cstrong\u003e a\u003c/strong\u003e, Phylogenetic tree for a 44-year old male. An ACF D_ACF_007_2 and a diminutive adenomatous polyp D_PLP_003_1 lost the 2\u003csup\u003end\u003c/sup\u003e copy of wild type \u003cem\u003eAPC\u003c/em\u003e by the age of five but did not progress into cancer.\u003cstrong\u003e b\u003c/strong\u003e, Phylogenetic tree for a 32-year old male. A diminutive adenomatous polyp N_PLP_006 lost the 2\u003csup\u003end\u003c/sup\u003e copy of wildtype \u003cem\u003eAPC\u003c/em\u003e by the age of seven.\u003cstrong\u003e c\u003c/strong\u003e, Phylogenetic tree for an 18-year old male. A diminutive adenomatous polyp K_PLP_004 had acquired somatic mutations in both \u003cem\u003eAPC\u003c/em\u003e and \u003cem\u003eKRAS \u003c/em\u003ebefore the age of seven.\u003cstrong\u003e d\u003c/strong\u003e, Phylogenetic tree for a 23-year old male. Multiple diminutive adenomatous polyps had acquired somatic mutations in \u003cem\u003eAPC\u003c/em\u003e and before the age of ten.\u003c/p\u003e","description":"","filename":"4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6581155/v1/eec2407ceabaad14a6f23186.jpg"},{"id":86498289,"identity":"5c1349f0-465f-4329-9791-f18789c2550d","added_by":"auto","created_at":"2025-07-11 10:33:51","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1809831,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6581155/v1/22874f90-3359-4f7e-9f90-0a8817ef5b85.pdf"},{"id":83296195,"identity":"71a6a706-6055-464f-9445-8e7ac2a51a8a","added_by":"auto","created_at":"2025-05-22 13:53:42","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":240240,"visible":true,"origin":"","legend":"Supplementary Tables","description":"","filename":"SupplementaryTables.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-6581155/v1/aed2ef5b1e411ad269bc3192.xlsx"},{"id":83296201,"identity":"b80f2104-37da-4e03-9150-40dd2b8239e2","added_by":"auto","created_at":"2025-05-22 13:53:42","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":9606994,"visible":true,"origin":"","legend":"","description":"","filename":"ExtendedDataFigs.docx","url":"https://assets-eu.researchsquare.com/files/rs-6581155/v1/d759810984c4e5bf6dbc4f0e.docx"}],"financialInterests":"\u003cb\u003eYes\u003c/b\u003e there is potential Competing Interest.\nP.J.C. is a co-founder and CSO of Quotient Therapeutics. M.R.S, and I.M. are co-founders of Quotient Therapeutics. The other authors declare no competing interests.","formattedTitle":"The earliest stages of neoplastic transformation in Familial Adenomatous Polyposis","fulltext":[{"header":"INTRODUCTION","content":"\u003cp\u003eThe progression of somatic genetic changes associated with, and underlying, the conversion of a normal colorectal epithelial cell into a colorectal carcinoma has served as a general model of cancer development\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. Through genetic characterisation of the succession of morphologically distinguishable intermediate neoplasms, the consensus sequence of somatic genetic changes leading to colorectal carcinoma has been described\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e,\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. This progression incorporates an ordered accumulation of \u0026ldquo;driver\u0026rdquo; mutations in cancer genes commonly mutated in colorectal carcinoma, including \u003cem\u003eAPC\u003c/em\u003e, \u003cem\u003eKRAS\u003c/em\u003e, \u003cem\u003eTP53\u003c/em\u003e, genome-wide changes in mutation burdens and signatures of all classes of somatic mutation, and their associations with macroscopically visible adenomatous polyps of increasing size and histological abnormality. Iterative revision of this progression has been enabled by advances in DNA sequencing technology resulting in increasingly comprehensive characterisations of adenoma\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e,\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e,\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e,\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e and carcinoma genomes \u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e,\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e,\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e,\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e,\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. The earliest stages in colorectal cancer progression have, however, been less well characterised. These may be of particular importance as potentially reversible endogenous and exogenous risk factors for colorectal cancer may exert their influence during this period.\u003c/p\u003e \u003cp\u003eFamilial Adenomatous Polyposis (FAP) is characterised by the presence of a single constitutional mutation in the \u003cem\u003eAPC\u003c/em\u003e gene, generally resulting in APC protein truncation and functional inactivation. This constitutional mutation is usually present in every cell of the body and somatic acquisition of inactivating mutations in the remaining wild type allele of individual cells results in biallelic \u003cem\u003eAPC\u003c/em\u003e inactivation, initiating colorectal neoplastic transformation. Individuals with FAP develop hundreds-to-thousands of intestinal adenomas over the first three decades of life. Without prophylactic surgical intervention, such polyposis substantially increases the risk of colorectal carcinoma, with the penetrance approaching 100% by 60 years of age\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. Biallelic somatic \u003cem\u003eAPC\u003c/em\u003e mutations are also thought to be the earliest step of the chromosomal instability pathway of sporadic colorectal cancer development, which accounts for \u0026sim;80% of cases\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e,\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e,\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e,\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. In consequence, study of FAP has been widely used to advance understanding of sporadic colorectal cancer development due to \u003cem\u003eAPC\u003c/em\u003e inactivation.\u003c/p\u003e \u003cp\u003eIn addition to profuse colorectal adenomatous polyps, individuals with FAP show elevated frequencies of aberrant crypt foci (ACF), microscopic areas of flat colorectal epithelium incorporating crypts exhibiting a range of abnormal architectural and/or nuclear cytological histological characteristics\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. Microscopic examination of ACF has revealed heterogeneous histology, ranging from the presence of very early regenerative crypts (type A) to hyperplastic crypts (type B) to dysplastic crypts (type C)\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. It has been proposed that ACF represent early stages of the normal cell-to-carcinoma progression before the development of macroscopically visible colorectal adenomatous polyps\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e,\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e,\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. Early studies reported \u0026ldquo;driver\u0026rdquo; mutations in \u003cem\u003eAPC\u003c/em\u003e and \u003cem\u003eKRAS\u003c/em\u003e\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e,\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e,\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e,\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e,\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e,\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e,\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e,\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e in ACF in different contexts, but few are from individuals with FAP\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e,\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e,\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e and these do not report the genome-wide features of neoplastic evolution, including mutation burdens and mutational signatures, that might inform on changes in mutational processes following biallelic \u003cem\u003eAPC\u003c/em\u003e inactivation.\u003c/p\u003e \u003cp\u003eCharacterisation of the somatic mutational landscapes of these very early stages of colorectal cancer development presents the technical challenge of generating genome sequences from the small amounts of DNA present in individual normal cells or microscopic clones. Recently, we combined laser capture microdissection with a bespoke library construction protocol that enables high quality whole genome sequences to be obtained from individual colorectal crypts\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e. Each crypt is a population of ~\u0026thinsp;2,000 colorectal epithelial cells derived from a single, common, recent ancestor stem cell, with high variant allele fraction (VAF) somatic mutations found in DNA sequences from an individual crypt representing the mutations present in that recent ancestor stem cell. The method has been applied to normal colorectal crypts from individuals who are healthy or have colorectal cancer\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e, from individuals with colorectal cancer predisposition syndromes due to inherited DNA polymerase proof reading deficiencies\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e, DNA base excision repair deficiencies\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e, DNA mismatch repair deficiencies\u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e, and from individuals with inflammatory bowel disease\u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eBy cataloguing somatic mutations in whole genome sequences of normal crypts, crypts from ACF and crypts from very small polyps from individuals with FAP, before and after biallelic \u003cem\u003eAPC\u003c/em\u003e inactivation, we characterise the progression of \u0026ldquo;driver\u0026rdquo; mutation acquisition and changes in somatic mutation rates and mutational signatures during the earliest stages of colorectal neoplastic change due to \u003cem\u003eAPC\u003c/em\u003e inactivation at a much higher level of resolution than previously achieved. Furthermore, through the phylogenetic crypt analysis enabled by this approach we provide insights into the long time period over which evolution of neoplastic colorectal clones can occur.\u003c/p\u003e"},{"header":"RESULTS","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003ePatients and samples\u003c/h2\u003e \u003cp\u003e15 individuals (five of whom were from the same family) aged 18 to 44 years with intestinal adenomatous polyposis due to inherited heterozygous \u003cem\u003eAPC\u003c/em\u003e mutations were studied. Ten exhibited features of classical FAP with \u0026gt;\u0026thinsp;100 colorectal polyps and five had attenuated FAP (AFAP) with \u0026lt;\u0026thinsp;100 colorectal polyps (Supplementary Table\u0026nbsp;1). None had developed invasive colorectal carcinoma at the time of tissue sampling. All constitutional \u003cem\u003eAPC\u003c/em\u003e mutations were predicted to inactivate the APC protein and included nonsense substitutions (seven individuals), small frame-shifting insertions or deletions (five), essential splice site substitutions (two) and whole gene deletion (one) (Supplementary Table\u0026nbsp;1).\u003c/p\u003e \u003cp\u003e279 colorectal crypts were isolated by laser capture microdissection and individually whole genome sequenced: 110 were from histologically normal epithelium; 83 were from 18 ACF, of which three were composed of just a single bifurcating crypt with no other nearby abnormality, thus representing the simplest stereotypic morphological microscopic abnormality we were able to identify; 86 were from 16 adenomatous polyps, including 70 from 13 diminutive polyps (\u0026lt;\u0026thinsp;5mm diameter) and 16 from three large polyps (\u0026gt;\u0026thinsp;10mm) (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea, Supplementary Table\u0026nbsp;2). Crypts with normal or uncertain histology adjacent to, or entrapped within each type of lesion were also included to explore their origins. Crypts were individually whole genome sequenced to a median of 31-fold read coverage and somatic mutations of all classes were identified using previously described approaches\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e,\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e,\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e,\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e,\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e. Crypts with complex or branching morphologies were, where feasible, sampled and genome sequenced in separate segments. For clarity of narrative and analysis, when groups of crypts apparently from the same ACF/adenomatous polyp had different \u0026ldquo;driver\u0026rdquo; and genome-wide mutations, and thus different clonal origins, they were given different identifiers and treated as separate groups.\u003c/p\u003e \u003cp\u003e \u003cb\u003eSomatic mutations in\u003c/b\u003e \u003cb\u003eAPC\u003c/b\u003e \u003cb\u003eand other cancer genes\u003c/b\u003e\u003c/p\u003e \u003cp\u003eLike normal crypts from wild-type individuals\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e, crypts from individuals with FAP/AFAP, whether normal, from ACF or from adenomatous polyps, showed somatic mutations with VAF peaks of 0.4\u0026ndash;0.5 indicating that they originate from recent single common ancestors (Extended Data Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). One out of 110 morphologically normal crypts (PD44897b_lo0001) from individuals with FAP/AFAP carried a high VAF somatic heterozygous inactivating \u003cem\u003eAPC\u003c/em\u003e mutation (R213*). Using short-read sequencing it was not possible, either in this or any other crypt genome sequence, to determine whether somatic \u003cem\u003eAPC\u003c/em\u003e mutations were in the constitutional mutant or wild type \u003cem\u003eAPC\u003c/em\u003e gene copy. Nevertheless, it is probable that they were in the wild type allele, resulting in biallelic \u003cem\u003eAPC\u003c/em\u003e mutation, cellular impairment of APC function and initiation of neoplastic transformation. The presence of this somatic \u003cem\u003eAPC\u003c/em\u003e mutation in most cells of this crypt likely occurred through the mutation arising in one of the ~\u0026thinsp;5 stem cells at the crypt base with subsequent complete crypt colonisation by its progeny. This raised the possibility that other normal crypts had been incompletely colonised by \u003cem\u003eAPC\u003c/em\u003e somatic mutant stem cells, with the VAF of the somatic \u003cem\u003eAPC\u003c/em\u003e mutation consequently falling below standard thresholds for detection. However, a systematic search through the sequence reads of all normal crypts did not reveal any further somatic inactivating \u003cem\u003eAPC\u003c/em\u003e mutations, even at very low VAFs, indicating that incomplete colonisation is rare, and suggesting that colonisation occurs relatively quickly. Inactivating mutations, each restricted to one normal crypt, were found in other tumour suppressor genes (\u003cem\u003eRNF43\u003c/em\u003e (R454Afs*48), \u003cem\u003eFBXW7\u003c/em\u003e (R224*), \u003cem\u003eKMT2C\u003c/em\u003e (R2257*) and \u003cem\u003eKMT2D\u003c/em\u003e (X1565_splice)). Their biological significance is unclear.\u003c/p\u003e \u003cp\u003eNine out of 18 ACF contained crypts with a somatic inactivating \u003cem\u003eAPC\u003c/em\u003e mutation (nine different \u003cem\u003eAPC\u003c/em\u003e mutations in 36/83 crypts). Microscopy images of the ACF and their constituent crypts were reviewed by two clinically practising colorectal histopathologists, without prior knowledge of the mutational results, and classified as 10 type A-I ACF (with crypt elongation, branching or dilatation, no to minimal regenerative changes, and absence of serration or dysplastic changes), seven type A-II ACF (with crypt elongation, branching or dilatation, moderate to marked regenerative changes, but no serration or dysplastic changes), and one type C ACF (with dysplastic change) based on established standards from previous literature\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e (Supplementary Table\u0026nbsp;3, Extended Data Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). There were no type B ACFs (with serration characteristic of early hyperplastic polyp) consistent with these being associated with an \u003cem\u003eAPC\u003c/em\u003e-mutation independent serrated neoplasm pathway to colorectal cancer\u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e. Comparing sequenced crypts from ACF with a somatic \u003cem\u003eAPC\u003c/em\u003e mutation to those without showed that presence of the somatic inactivating \u003cem\u003eAPC\u003c/em\u003e mutation is associated with regenerative changes (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.001, Chi-squared test). None of the three crypt clusters composed of just a single bifurcating crypt had a somatic inactivating \u003cem\u003eAPC\u003c/em\u003e mutation. Among the nine ACF without a somatic \u003cem\u003eAPC\u003c/em\u003e mutation, a known hotspot mutation in \u003cem\u003eALK\u003c/em\u003e (R395H) was found in one crypt together with a splice site mutation in \u003cem\u003eCDH1\u003c/em\u003e (X17_splice). Two crypts with somatic inactivating \u003cem\u003eAPC\u003c/em\u003e mutations also had additional inactivating somatic mutations in tumour suppressor genes, one in \u003cem\u003eSMARCA4\u003c/em\u003e (X648_splice) and \u003cem\u003eNCOR1\u003c/em\u003e (R694*), and the other in \u003cem\u003eKMT2C\u003c/em\u003e (L2662Cfs*16) (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb-c, Supplementary Table\u0026nbsp;4). These mutations were not shared with other crypts in the same ACF.\u003c/p\u003e \u003cp\u003e12/13 diminutive adenomatous polyps and 3/3 large adenomatous polyps contained crypts with inactivating somatic \u003cem\u003eAPC\u003c/em\u003e mutations. (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb-c). Two diminutive polyps and one large polyp showed individual crypts carrying two distinct inactivating somatic \u003cem\u003eAPC\u003c/em\u003e mutations (and therefore, in total, three inactivating \u003cem\u003eAPC\u003c/em\u003e mutations when the germline mutation is also considered), a phenomenon that has been previously reported\u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e. Two diminutive polyps (PD40734C_PLP_007, PD42778D_PLP_003) each carried more than one independent \u003cem\u003eAPC\u003c/em\u003e mutant clone, and another one (PD42778D_PLP_002) had a type A-II ACF with an independent \u003cem\u003eAPC\u003c/em\u003e somatic mutation at the edge of the polyp, consistent with previous studies of early adenomas in mice and human\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e,\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e,\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e (Extended Data Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Mutations in several other cancer genes, including canonical hotspot mutations, were found in crypts from polyps (\u003cem\u003eKRAS\u003c/em\u003e (G12S, G12D), \u003cem\u003eTP53\u003c/em\u003e (R202C, R64*), \u003cem\u003ePIK3CA\u003c/em\u003e (E542K, E545K), \u003cem\u003eSMAD4\u003c/em\u003e (R361H) and \u003cem\u003eFBXW7\u003c/em\u003e (R479Q))(Supplementary Table\u0026nbsp;4). Some crypts adjacent to, or entrapped within, \u003cem\u003eAPC\u003c/em\u003e mutation carrying ACF and adenomatous polyps lacked second \u003cem\u003eAPC\u003c/em\u003e mutations despite exhibiting elongation, branching, dilatation, or regenerative changes. We termed these \u0026lsquo;bystander\u0026rsquo; crypts to distinguish them from normal crypts and independent ACF.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eElevated somatic mutation burdens in aberrant crypt foci and polyps\u003c/h3\u003e\n\u003cp\u003eSomatic mutation burdens were examined in 276 crypts from FAP patients with a coverage of \u0026gt;\u0026thinsp;10-fold, including 110 normal crypts, 102 crypts in ACF and adenomatous polyps with confirmed somatic \u003cem\u003eAPC\u003c/em\u003e mutations (39 crypts from 10 ACF and 63 crypts from 15 adenomatous polyps), 30 crypts from 9 ACF without \u003cem\u003eAPC\u003c/em\u003e somatic mutations, 4 adenomatous crypts from the diminutive polyp without somatic \u003cem\u003eAPC\u003c/em\u003e mutation, and 30 \u0026lsquo;bystander crypts\u0026rsquo;. In normal crypts from FAP/AFAP individuals, the total burdens of somatic single base substitutions (SBS, also sometimes termed single nucleotide variants or SNV) showed a linear accumulation with age of 41.7/year (95% C.I., 23\u0026ndash;61, \u003cem\u003eP-adjust\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1.5 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;3\u003c/sup\u003e), which did not differ from that of normal crypts from wild-type individuals (likelihood ratio test, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.59) (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea). Similarly, the burdens of small insertions and deletions (indels) in normal crypts from FAP/AFAP individuals did not differ from those of wild type individuals (likelihood ratio test, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.22) (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea). Copy number variants (CNVs) and structural variants (SVs) were rare, but 132 retrotransposition events (RTs) were found in 110 normal crypts, all at a similar frequency to wild type individuals (Supplementary Table\u0026nbsp;5\u0026ndash;7). Thus, across the spectrum of somatic mutation types, no differences were identified between the mutation rates in normal crypts from individuals with FAP and those from wild-type controls. The single morphologically normal AFAP crypt with an inactivating somatic \u003cem\u003eAPC\u003c/em\u003e mutation (\u003cem\u003eR213*\u003c/em\u003e) showed a similar number of single base substitutions and indels to normal crypts without somatic \u003cem\u003eAPC\u003c/em\u003e mutations, with no copy number changes or rearrangements.\u003c/p\u003e \u003cp\u003eCrypts from ACF which carried somatic \u003cem\u003eAPC\u003c/em\u003e mutations showed modest increases in total small indel mutation burdens (35 more indels, 95% C.I., 12\u0026ndash;58, \u003cem\u003eP-adjust\u003c/em\u003e\u0026thinsp;=\u0026thinsp;7.0 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;3\u003c/sup\u003e) (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb-c). Copy number variants, structural variants and retrotransposition events were at a similar level to normal crypts. In contrast, crypts with somatic \u003cem\u003eAPC\u003c/em\u003e mutations from diminutive adenomas (\u0026lt;\u0026thinsp;5mm) showed markedly increased burdens of both single base substitutions (1580 more substitutions, 95% C.I., 1236\u0026ndash;1924, \u003cem\u003eP-adjust\u003c/em\u003e\u0026thinsp;=\u0026thinsp;5.3 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;17\u003c/sup\u003e) and indels (186 more indels, 95% C.I., 158\u0026ndash;215, \u003cem\u003eP-adjust\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2.0 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;30\u003c/sup\u003e), and these burdens were further increased in the three large adenomas (6728 more substitutions, 95% C.I., 5820\u0026ndash;7636, \u003cem\u003eP-adjust\u003c/em\u003e\u0026thinsp;=\u0026thinsp;7.1 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;38\u003c/sup\u003e; 794 more indels, 95% C.I., 730\u0026ndash;857, \u003cem\u003eP-adjust\u003c/em\u003e\u0026thinsp;=\u0026thinsp;8.5 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;81\u003c/sup\u003e). Copy number variants, structural variants and retrotransposition events were more commonly found in diminutive polyps compared with normal crypts (CNV: increased by 0.6/crypt, 95% C.I.: 0.2\u0026ndash;0.9, \u003cem\u003eP-adjust\u003c/em\u003e\u0026thinsp;=\u0026thinsp;4.2 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;3\u003c/sup\u003e; SV: increased by 0.7/crypt, 95% C.I.: 0.3\u0026ndash;1.2, \u003cem\u003eP-adjust\u003c/em\u003e\u0026thinsp;=\u0026thinsp;4.2 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;3\u003c/sup\u003e; RT: increased by 6.6/crypt, 95% C.I.: 3.8\u0026ndash;9.4, \u003cem\u003eP-adjust\u003c/em\u003e\u0026thinsp;=\u0026thinsp;3.8 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;5\u003c/sup\u003e), with more significant increases in large polyps (CNV: increased by 5.0/crypt, 95% C.I.: 4.3\u0026ndash;5.7, \u003cem\u003eP-adjust\u003c/em\u003e\u0026thinsp;=\u0026thinsp;6.0 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;25\u003c/sup\u003e; SV: increased by 4.3/crypt, 95% C.I.: 3.4\u0026ndash;5.2, \u003cem\u003eP-adjust\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1.3 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;14\u003c/sup\u003e; RT: increased by 36.4/crypt, 95% C.I.: 30.5\u0026ndash;42.1, \u003cem\u003eP-adjust\u003c/em\u003e\u0026thinsp;=\u0026thinsp;8.0 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;21\u003c/sup\u003e). Overall, ACF from FAP patients, including the one dysplastic ACF, appear to represent an early stage of neoplastic progression following \u003cem\u003eAPC\u003c/em\u003e inactivation in which some mutational processes are mildly accelerated, but large-scale chromosome instability has not yet been triggered. \u0026lsquo;Bystander crypts\u0026rsquo; within or immediately adjacent to ACFs and polyps did not show increased mutation burdens, confirming their likely status as entrapped normal crypt epithelium.\u003c/p\u003e\n\u003ch3\u003eMutational signatures\u003c/h3\u003e\n\u003cp\u003eMutational signatures were extracted from the full set of FAP/AFAP crypt mutation catalogues (ie including morphologically normal and abnormal crypts) together with a previously published set of normal crypts from wild-type individuals\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e. Two different methods were used (HDP\u003csup\u003e\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e and SigProfiler\u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e) which yielded similar repertoires of signatures. Using HDP, 11 SBS signature components were identified (Extended Data Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e) which could be explained by 10 previously described mutational signatures: SBS1, SBS2, SBS5, SBS13, SBS17b, SBS18, SBS88, SBS89, SBS93 and SBSC (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea, Extended Data Fig.\u0026nbsp;5\u0026ndash;6). SBS1 is caused by deamination of 5-methylcytosine, a process which occurs in most human cell types throughout life at a more-or-less constant rate\u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e,\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e. SBS2 and SBS13 are caused by activity of the AID/APOBEC family of cytidine deaminases, and were only found in three crypts from the previously published control group\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e. SBS5 is also ubiquitous in normal human tissues, and accrues in a clock-like manner with age, but its aetiology is not well understood\u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e,\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e. SBS18 is attributed to the effects of reactive oxygen species on DNA, principally the formation of 8-oxoguanine\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e, and is present in some normal cell types including human placenta\u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e, and large intestine epithelium\u003csup\u003e\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e,\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e. SBS88 is due to colibactin, a mutagenic agent produced by strains of \u003cem\u003eE.coli\u003c/em\u003e often present in the large intestine microbiome\u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e. SBS89 is found in normal colorectal crypts, but not thus far in other cell types, is of unknown origin but may plausibly also be caused by a microbiome-derived mutagen\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e. SBS17b is observed in chemotherapy na\u0026iuml;ve oesophageal adenocarcinomas\u003csup\u003e\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e\u003c/sup\u003e and in tissues exposed to the chemotherapeutic agent 5-fluorouracil (5-FU)\u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e,\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e. SBS93 was previously discovered in stomach and oesophageal cancers and is of unknown aetiology\u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e. SBSC is a predominantly C\u0026thinsp;\u0026gt;\u0026thinsp;T signature previously reported in wild-type colorectal crypts\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e. Seven small insertion and deletion mutational signatures, ID1, ID2, ID5, ID14, ID18, IDA and IDB, were also extracted (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea, Extended Data Fig.\u0026nbsp;7\u0026ndash;9). ID1 is characterised by insertions of T at T homopolymer tracts, is thought to arise from polymerase-related slippage of the replicated DNA strand and is the predominant indel signature in normal cells\u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e. ID2 is characterised by T deletions at T homopolymer tracts, is thought to arise from polymerase-related slippage of the template DNA strand and is predominant in cancer genomes\u003csup\u003e\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e\u003c/sup\u003e. ID5 is an age-associated signature present in cancer and normal cells. ID14 is a signature common in colorectal cancer and was extracted from polyps from one individual (PD42778). ID18 is associated with colibactin exposure and SBS88\u003csup\u003e47\u003c/sup\u003e. IDA primarily consists of 1bp insertions at C homopolymer tracts and was only found in one individual (PD40734). IDB is characterised by 1bp deletions of T at non-homopolymer regions, and has previously been reported in wild-type colorectal crypts\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe repertoire and contributions of mutational signatures in normal crypts from individuals with FAP/AFAP were similar to those in normal crypts from wild-type individuals and included SBS1, SBS5, SBS18, SBS88, ID1, ID2, ID5 and ID18. The single exception was individual PD40734 who showed elevated burdens and proportions of SBS5 (or a signature resembling it) in all normal crypts, and the presence of an indel signature, IDA, not found in any other individual. PD40734 had not been exposed to cancer chemotherapy or to any other known systemic mutagen. The cause of the unusual signature repertoire in this individual is unknown, but IDA has been previously reported in an individual with defective base excision repair\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eDifferences in signature contributions were observed between normal crypts and crypts with somatic \u003cem\u003eAPC\u003c/em\u003e mutations in ACF (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb-c). These included modest additional burdens of SBS18 (128, 95% C.I., 38\u0026ndash;218, \u003cem\u003eP-adjust\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1.0 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e) and ID1 (29, 95% C.I., 21\u0026ndash;37, \u003cem\u003eP-adjust\u003c/em\u003e\u0026thinsp;=\u0026thinsp;5.0 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;11\u003c/sup\u003e). In crypts with somatic \u003cem\u003eAPC\u003c/em\u003e mutations from diminutive and large adenomas there were greater increases in SBS18 (diminutive polyps: 552, 95% C.I., 460\u0026ndash;644, \u003cem\u003eP-adjust\u003c/em\u003e\u0026thinsp;=\u0026thinsp;8.8 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;27\u003c/sup\u003e; large polyps: 1868, 95% C.I., 1680\u0026ndash;2055, \u003cem\u003eP-adjust\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2.0 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;59\u003c/sup\u003e) and ID1 (diminutive polyps: 97, 95% C.I., 85\u0026ndash;108, \u003cem\u003eP-adjust\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2.5 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;45\u003c/sup\u003e; large polyps: 420, 95% C.I., 393\u0026ndash;448, \u003cem\u003eP-adjust\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1.2 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;95\u003c/sup\u003e) mutation burdens than in ACF, and the burdens of SBS1 (diminutive polyps: 563, 95% C.I., 426\u0026ndash;699, \u003cem\u003eP-adjust\u003c/em\u003e\u0026thinsp;=\u0026thinsp;3.4 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;14\u003c/sup\u003e; large polyps: 1286, 95% C.I., 1018\u0026ndash;1554, \u003cem\u003eP-adjust\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1.4 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;18\u003c/sup\u003e), SBS5 (diminutive polyps: 399, 95% C.I., 280\u0026ndash;518, \u003cem\u003eP-adjust\u003c/em\u003e\u0026thinsp;=\u0026thinsp;5.7 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;10\u003c/sup\u003e; large polyps: 1711, 95% C.I., 1439\u0026ndash;1982, \u003cem\u003eP-adjust\u003c/em\u003e\u0026thinsp;=\u0026thinsp;3.8 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;29\u003c/sup\u003e), ID2 (diminutive polyps: 52, 95% C.I., 44\u0026ndash;60, \u003cem\u003eP-adjust\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1.3 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;30\u003c/sup\u003e; large polyps: 264, 95% C.I., 245\u0026ndash;282, \u003cem\u003eP-adjust\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1.2 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;95\u003c/sup\u003e) and ID5 (diminutive polyps: 24, 95% C.I., 17\u0026ndash;32, \u003cem\u003eP-adjust\u003c/em\u003e\u0026thinsp;=\u0026thinsp;8.8 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;9\u003c/sup\u003e; large polyps: 81, 95% C.I., 65\u0026ndash;96, \u003cem\u003eP-adjust\u003c/em\u003e\u0026thinsp;=\u0026thinsp;3.0 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;22\u003c/sup\u003e) were also increased (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec). In addition to the increased burdens of signatures present in normal crypts, SBS17b was found in a single large adenomatous polyp from PD42778 (who had not been treated with 5-fluorouracil) and SBS93 was found in the same large polyp and a diminutive polyp with two somatic \u003cem\u003eAPC\u003c/em\u003e mutations from the same individual.\u003c/p\u003e\n\u003ch3\u003eCrypt phylogenies\u003c/h3\u003e\n\u003cp\u003eUsing somatic mutations as unique markers, we constructed phylogenetic trees of crypts for each individual in this study\u003csup\u003e\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e, Extended Data Fig.\u0026nbsp;10). As in wild-type individuals, morphologically normal crypts from individuals with FAP/AFAP, whether physically adjacent or distant from one another in the epithelial lining, typically shared only a small number of somatic mutations, indicating a common ancestor that existed in the distant past, possibly during embryonic development. By contrast, crypts from each ACF with the same \u003cem\u003eAPC\u003c/em\u003e somatic mutation usually shared substantial numbers of somatic mutations indicating comparatively recent common ancestors, as did crypts from adenomatous polyps. Adenomatous polyps with recent common ancestors carrying more than one driver mutation had longer trunks (reflecting a larger number of mutations in the most recent common ancestor) than ACF or adenomatous polyps with just a single somatic \u003cem\u003eAPC\u003c/em\u003e mutation (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.0028). However, in polyps with just a single somatic mutation in \u003cem\u003eAPC\u003c/em\u003e, the higher total crypt mutation burdens than those found in ACF could not be explained by either truncal or branch mutations alone (branch mutations being defined by mutations not in the most recent common ancestor).\u003c/p\u003e \u003cp\u003eThe genome-wide somatic mutation burdens in these phylogenies can serve as molecular clocks allowing estimates of the timing of occurrence of the somatic \u003cem\u003eAPC\u003c/em\u003e mutations during the life of the individual\u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e, and thus the length of time that individual neoplastic lesions have taken to develop. Some \u003cem\u003eAPC\u003c/em\u003e somatic mutations appear to have arisen early in life, with cells carrying them progressing to a limited extent. For example, the somatic \u003cem\u003eAPC\u003c/em\u003e R1114* mutation in an ACF consisting of two bifurcating crypts (PD42778D_ACF_007_2) from a 44-year-old with FAP is likely to have occurred within the first five years of life. Therefore, over a period of at least 39 years, the neoplastic clone initiated by this mutation has only developed into an ACF (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea). Similarly, in this 44-year-old individual, a somatic \u003cem\u003eAPC\u003c/em\u003e loss of heterozygosity event (PD42778D_PLP_003_1) is likely to have occurred by age five years and over 39 years has only progressed into a diminutive adenomatous polyp.\u003c/p\u003e"},{"header":"DISCUSSION","content":"\u003cp\u003eWe have investigated the earliest stages of colorectal neoplastic progression due to biallelic \u003cem\u003eAPC\u003c/em\u003e mutation by whole genome sequencing of 279 crypts microdissected from normal colorectal epithelium, ACF and adenomas from individuals with Familial Adenomatous Polyposis. The results inform on the earliest stages of the most common pathway of sporadic colorectal carcinogenesis which is also thought to be initiated by \u003cem\u003eAPC\u003c/em\u003e inactivation.\u003c/p\u003e \u003cp\u003eAlmost all histologically normal crypts remained heterozygous for the inherited \u003cem\u003eAPC\u003c/em\u003e gene mutation and exhibited mutation rates and signatures indistinguishable from those found in normal crypts from wild type individuals. The results show that an increased genome-wide mutation rate in normal colorectal cells is not a factor contributing to the high risk of neoplastic change in cells and individuals with heterozygous \u003cem\u003eAPC\u003c/em\u003e inactivating mutations. They are, therefore, consistent with the orthodoxy that \u003cem\u003eAPC\u003c/em\u003e inactivating mutations act in a recessive manner and engender biological consequences for an individual cell only if biallelic. This overall pattern is similar to that observed in the normal colorectal epithelium of individuals with colorectal cancer predisposition due to inherited mutations in the DNA Mismatch Repair genes \u003cem\u003eMSH2\u003c/em\u003e and \u003cem\u003eMLH1\u003c/em\u003e, in whom the large majority of crypts show normal mutation rates and signatures, with only rare crypts showing loss of the wild type allele and initiation of neoplastic change\u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e. However, it differs from the pattern in individuals with inherited monoallelic mutations in genes encoding the replicative polymerases POLE and POLD1, which cause DNA proofreading defects and in whom all normal colorectal crypts, and all other normal cell types thus far sampled, show elevated mutation rates with distinctive mutational signatures\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e. It also differs from the pattern in individuals with inherited biallelic mutations in \u003cem\u003eMUTYH\u003c/em\u003e in whom all normal colorectal crypts also show elevated mutation rates and abnormal signatures\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e. Collectively, these studies depict the different ways in which inherited mutations can influence the rate of neoplastic change in colorectal cells through mutagenesis of the somatic genome.\u003c/p\u003e \u003cp\u003eOne FAP patient with a germline\u0026thinsp;~\u0026thinsp;5Mb deletion on chromosome 5 including all of \u003cem\u003eAPC\u003c/em\u003e (PD40734, chr5: 110,610,575\u0026thinsp;\u0026minus;\u0026thinsp;115,126,084) exhibited a subtly different somatic mutation landscape in all analysed crypts due to the elevated burden of an SBS5-like mutational signature and an indel mutational signature, IDA, not found in wild type individuals. The cause of this unusual mutation pattern is unclear. It could be due to haploinsufficiency of another gene within the inherited 5Mb deletion (although there are no striking candidates), coincidental occurrence of a constitutional mutation in a gene elsewhere in the genome that alters mutagenesis (although, if so, we have been unable to identify it), or an exogenous mutagenic exposure (although there is no record of this and the pattern of mutations does not correspond with the mutational signatures of any of the known lifestyle, environmental or iatrogenic mutagenic exposures thus far catalogued).\u003c/p\u003e \u003cp\u003eOne out of 110 (1%) normal crypts from individuals with FAP/AFAP carried a somatic \u003cem\u003eAPC\u003c/em\u003e mutation. This compares to 0/1,974 whole or targeted genome sequenced crypts from wild-type individuals\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e. Whether this represents a biological difference or chance is uncertain. Nevertheless, it is plausible that a colorectal stem cell from an individual with FAP/AFAP carrying biallelic \u003cem\u003eAPC\u003c/em\u003e inactivating mutations has selective advantage in colonising a crypt compared to other crypt stem cells with just the monoallelic constitutional \u003cem\u003eAPC\u003c/em\u003e mutation\u003csup\u003e\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e\u003c/sup\u003e, whereas a colorectal stem cell from a wild-type individual carrying a monoallelic somatic \u003cem\u003eAPC\u003c/em\u003e inactivating mutation, may have little selective advantage over biallelic wild-type cells. Although detection of a single normal crypt with biallelic \u003cem\u003eAPC\u003c/em\u003e mutations does not permit robust estimates of their prevalence, the result suggests that a colorectum of ~\u0026thinsp;15,000,000 crypts in an adult with FAP could harbour tens of thousands of such crypts.\u003c/p\u003e \u003cp\u003eIn contrast to normal cells (of which\u0026thinsp;\u0026lt;\u0026thinsp;1% harbour somatic \u003cem\u003eAPC\u003c/em\u003e mutations), 9/18 (50%) ACF included crypts with a somatic \u003cem\u003eAPC\u003c/em\u003e mutation. Thus, by virtue of their biallelic mutations in \u003cem\u003eAPC\u003c/em\u003e, some ACF represent an early stage of colorectal neoplasia. Early studies found few \u003cem\u003eAPC\u003c/em\u003e mutations in ACF and predominantly in type C ACF with dysplasia\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e,\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e,\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. Here, we show \u003cem\u003eAPC\u003c/em\u003e mutations in type A-I and A-II ACF characterised just by crypt fission and early regenerative changes. Whether ACF are actually the progenitor lesions from which adenomatous polyps develop remains uncertain. However, their size, microscopic morphology, and non-polypoid structure have previously been suggested as indications that they are an intermediate stage of neoplastic transformation between normal crypts and diminutive polyps\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e,\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e,\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. This conclusion is supported by results from this study in which the pattern of increased mutation burdens of specific ubiquitous mutational signatures (SBS18, ID1) in ACF are further increased in diminutive polyps alongside increased loads of other ubiquitous signatures (SBS1, SBS5, ID2). It is also compatible with the appearance of additional driver mutations in \u003cem\u003eAPC\u003c/em\u003e, \u003cem\u003eKRAS\u003c/em\u003e and \u003cem\u003eTP53\u003c/em\u003e in diminutive polyps which are not seen in ACF. The changes in the overall somatic mutation landscapes of ACF may, at least in part, be due to their stem cells having undergone more mitotic divisions than the stem cells of normal crypts (potentially resulting in higher ID1 burdens) but may also reflect changes in cell function, state or microenvironment as suggested by increased reactive oxygen species damage (resulting in increased SBS18 burdens). The extent to which these accelerated mutational processes contribute to acquisition of further driver mutations and continuation of cancer evolution is uncertain.\u003c/p\u003e \u003cp\u003eThe phylogenetic analyses using somatic mutations of individual crypts from ACF in this study provide insight into their natural history. Several ACF and diminutive adenomatous polyps acquired a somatic \u003cem\u003eAPC\u003c/em\u003e mutation within the first decade of life. Despite the early advent of APC loss of function, progression over decades to cancer was sometimes absent. Thus, stem cells with biallelic inactivating \u003cem\u003eAPC\u003c/em\u003e mutations can remain in a relatively early stage of neoplastic change for an extended duration and, notably, without obvious genome instability. The findings support the possibility that some cancers slowly evolve for several decades following their initiating somatic cancer gene driver mutation, and highlight the relatively low chance of any individual neoplastic clone ultimately developing into a cancer.\u003c/p\u003e "},{"header":"METHODS","content":"\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\n \u003ch2\u003eEthics and overview\u003c/h2\u003e\n \u003cp\u003eUsage of patient samples in this study were approved by IRB of the University of Hong Kong/Hospital Authority Hong Kong West Cluster (UW14-257) and West Midlands - Coventry and Warwickshire Research Ethics Service with United Kingdom Research Ethics Committee (REC) reference 17/WM/0295 (PD40734, PD42778, PD44720, PD44721), as well as East of Scotland Research Ethics Service with REC reference 18/ES/0133 (PD44892, PD44893, PD44894, PD44895, PD44896, PD44897, PD44898, PD44899, PD44900, PD44901, PD44902). The first cohort (PD40734, PD42778, PD44720, PD44721) consists of patients with a genetic diagnosis of FAP and operated on in Queen Mary Hospital of Hong Kong for total colectomy or total proctocolectomy. Specimens were collected fresh from the operating theatre and examined by a pathologist. Part of large polyps and random blocks of colon containing variable sizes of small polyps as well as normal colon mucosa were sampled, snap frozen and kept in -80 \u003csup\u003eo\u003c/sup\u003eC until use. The second cohort (PD44892, PD44893, PD44894, PD44895, PD44896, PD44897, PD44898, PD44899, PD44900, PD44901, PD44902) consists of FAP patients from UK and biopsies were taken from prophylactic colectomy. All samples were collected with informed consent from the patients.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\n \u003ch2\u003eLaser-microdissection and low-input whole-genome sequencing\u003c/h2\u003e\n \u003cp\u003eWe followed the standard protocol established at the Wellcome Sanger Institute for tissue processing, laser-microdissection and low-input whole-genome sequencing\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e. Fresh frozen blocks were cryostat sectioned to 20 mm, fixed to 4 mm PEN membrane slides (11600288, Leica) and stained with hematoxylin and eosin (H\u0026amp;E). Crypts were isolated using laser capture microscopy (LMD7000, Leica) and collected in separate wells of a 96-well plate. Collected samples were lysed using ARCTURUS PicoPure DNA extraction kit (Applied Biosystems) according to the manufacturer\u0026rsquo;s instructions. DNA library concentration was measured following library preparation and used to guide the choice of samples subject to DNA sequencing. The minimum library concentration was 5 ng/\u0026micro;l, and libraries with \u0026gt;\u0026thinsp;15 ng/\u0026micro;l were preferably picked for an aim of ~\u0026thinsp;30x coverage. Paired-end sequencing reads (150 bp) were generated on Illumina NovaSeq platform and aligned to the human reference genome (NCBI build38) using BWA-MEM (versions 0.7.17).\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\n \u003ch2\u003eCrypt classification\u003c/h2\u003e\n \u003cp\u003eNormal crypts, ACF and adenomatous polyps were classified according to their microscopic morphologies after H\u0026amp;E staining. Normal crypts had normal architecture lined by orderly differentiated cells. ACF were non-polypoid lesions including either a singleton or cluster of branching/budding crypts and/or cluster of enlarged crypts with wider calibre, tortuosity and/or changes in proportion or shapes of cell types. Nanozoomer-scanned images for all crypts from ACF were reviewed and scored without knowledge of the mutation results by two clinically practising gastrointestinal histopathologist (S.Y.L. and S.T.Y.) and with discrepancy discussed and reviewed to arrive at a consensus. Presence or absence of dysplasia, serration, crypt elongation (\u0026gt;\u0026thinsp;500 um length), crypt dilatation (\u0026gt;\u0026thinsp;100 um diameter) and branching were recorded. Their levels of regenerative features were graded as follow: 0, quiescent, small basally located nuclei even at crypt base with no mitosis; 1, mild elongation and crowding of nuclei with rare mitosis restricted to basal 1/3 of crypt; 2, elongation and crowding of nuclei with mitosis extending to middle third of crypt, with evidence of maturation on top 1/3 of crypt. Each group of ACF was then classified based on the most advanced lesion in the group using established standard from previous literature\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. Whilst there were no type B lesions (serrated type), type A lesions were further distinguished into two groups, type A lesion with no to minimal regeneration (type A-I) and Type A lesion with moderate to marked regeneration (type A-II). Representative pathology images of different types of ACF and the one normal crypt with a somatic \u003cem\u003eAPC\u003c/em\u003e nonsense mutation, along with their pathology grading can be found in Extended Data Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e. Individual crypts microdissected from polyps were also graded based on whether they correspond to an adenoma, ACF or normal. The ACF were graded using similar criteria as above.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\n \u003ch2\u003eSingle-base substitution calling and phylogenetic tree building\u003c/h2\u003e\n \u003cp\u003eCancer Variants through Expectation Maximization (CaVEMan)\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e (cgpCaVEMan version 1.15.2) and Strelka2\u003csup\u003e54\u003c/sup\u003e (version 2.9.10) were used to call single-base somatic substitutions against a matched blood sample from the same individual. For individuals without blood samples, we first called variants against an \u003cem\u003ein silico\u003c/em\u003e human reference genome, reconstructed the phylogeny, and called variants against a phylogenetically unrelated crypt. The consensus variants called by both CaVEMan and Strelka2 further went through a series of postprocessing filters and the final substitutions were used for constructing phylogenies: (1) The first filter removed mapping artifacts associated with BWA-MEM as follows: the median alignment score of reads supporting a mutation should be greater than or equal to 140, and fewer than half of these reads should be clipped. (2) The second filter was applied to remove artifacts that are associated with the LCM library preparation: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/cancerit/hairpin2\u003c/span\u003e\u003c/span\u003e. (3) A binomial distribution test was applied to remove possible remaining germline variants and a beta-binomial test was applied to filter out the common low frequency artifacts. Phylogenetic trees then were generated from the filtered substitutions using a maximum parsimony algorithm (MPBoot version 1.1.0)\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e55\u003c/span\u003e\u003c/sup\u003e. Substitutions were mapped onto tree branches using a maximum likelihood approach (TreeMut version 1.1) and visualized using ggtree (version 3.3.1)\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e56\u003c/span\u003e\u003c/sup\u003e and ape (version 5.6.1)\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e57\u003c/span\u003e\u003c/sup\u003e. The code for this part of the pipeline was integrated into Sequoia\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e : \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/TimCoorens/Sequoia\u003c/span\u003e\u003c/span\u003e.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\n \u003ch2\u003eIndel calling and mapping to the phylogenetic trees\u003c/h2\u003e\n \u003cp\u003ePindel\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e58\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e59\u003c/span\u003e\u003c/sup\u003e (cgpPindel version 3.10.0) and Strelka2\u003csup\u003e54\u003c/sup\u003e (version 2.9.10) were used for indel calling. Consensus indels from the two algorithms that possessed a minimum quality score of \u0026ge;\u0026thinsp;300 at positions covered by at least 15 reads were kept and subject to the same binomial and beta-binomial filters as single-base substitutions. Subsequently, indels were mapped to the corresponding phylogenetic trees reconstructed from single-based substitutions using a maximum likelihood approach (TreeMut version 1.1) and visualized using ggtree (version 3.3.1)\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e56\u003c/span\u003e\u003c/sup\u003e and ape (version 5.6.1)\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e57\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\n \u003ch2\u003eAnnotation of putative driver mutations\u003c/h2\u003e\n \u003cp\u003eMutations were overlapped with a known list of genes under positive selection in human cancers\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e60\u003c/span\u003e\u003c/sup\u003e and were further annotated using the cBioPortal MutationMapper cancer hotspot mutation database v5.1.7 (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.cbioportal.org/mutation_mapper\u003c/span\u003e\u003c/span\u003e)\u003csup\u003e61\u003c/sup\u003e. Known cancer mutation hotspots and protein-truncating variants in tumour suppressor genes were reported as putative driver mutations.\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003eScreening for subclonal\u003c/strong\u003e \u003cstrong\u003eAPC\u003c/strong\u003e \u003cstrong\u003emutations\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003eTo inspect possible subclonal \u003cem\u003eAPC\u003c/em\u003e mutations, all variants called by cgpCaVEMan and cgpPindel were intersected with the \u003cem\u003eAPC\u003c/em\u003e gene and annotated by ANNOVAR\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e62\u003c/span\u003e\u003c/sup\u003e (version 2020-06-08, database refGene,dbnsfp41c,avsnp150,cosmic89), including those that did not pass the two algorithms\u0026rsquo; quality control filters for reasons such as insufficient number of reads supporting the mutation. We then checked the annotation of all \u003cem\u003eAPC\u003c/em\u003e mutations to see if there were any nonsense, splice site or hotspot missense mutations (hotspot annotation from cBioPortal MutationMapper cancer hotspot mutation database).\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec16\" class=\"Section2\"\u003e\n \u003ch2\u003eCopy number variant calling\u003c/h2\u003e\n \u003cp\u003eSomatic copy-number variants (CNVs) were called using the Allele-Specific Copy number Analysis of Tumours (ASCAT) algorithm\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e63\u003c/span\u003e\u003c/sup\u003e as part of the ascatNGS package (version 4.3.2)\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e64\u003c/span\u003e\u003c/sup\u003e. ASCAT was run with default parameters with the exception of a segmentation penalty of 100. A bespoke filtering algorithm - ascatPCA - was used to reduce the number of false-positive calls that can arise when analysing genome sequences from low-input WGS (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/hj6-sanger/ascatPCA\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e).\u003c/span\u003e ascatPCA extracts a noise profile by aggregating the LogR ratio from across a panel of unrelated normal samples and subtracts this signature from that observed in the sample being analysed using principal component analysis.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec17\" class=\"Section2\"\u003e\n \u003ch2\u003eStructural variant and retrotransposition calling\u003c/h2\u003e\n \u003cp\u003eSomatic structural variants (SVs) and retrotranspositions were called using the Genomic Rearrangement Identification Software Suite (GRIDSS)\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e65\u003c/span\u003e\u003c/sup\u003e with default settings (version 2.9.4). All variants were confirmed by visual inspection and by checking if they fit the distribution expected based on the SNV-derived phylogenetic tree. Specifically, SVs (\u0026gt;\u0026thinsp;1kb) and retrotranspositions with QUAL\u0026thinsp;\u0026gt;\u0026thinsp;=\u0026thinsp;250 were included. For SVs smaller than 30kb, SVs with QUAL\u0026thinsp;\u0026gt;\u0026thinsp;=\u0026thinsp;300 were only included. Furthermore, SVs required breakpoint assembly on both sides with at least four discordant and two reads supporting them. Where breakpoints were imprecise, defined as the start and end positions being \u0026gt;\u0026thinsp;10bp apart, the SV was filtered out. We further filtered out SVs for which the standard deviation of the alignment positions at either end of the discordant read pairs was smaller than five. We also applied this standard deviation filter to single-break end calls in calling retrotranspositions and used Repeatmasker\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e66\u003c/span\u003e\u003c/sup\u003e(version 4.1.7) to annotate repetitive elements.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec18\" class=\"Section2\"\u003e\n \u003ch2\u003eModelling disease effects on mutational burdens\u003c/h2\u003e\n \u003cp\u003eRaw mutation burden was normalised by sensitivity of mutation detection in each crypt: we calculated the possibility of detecting a variant with a given coverage, median variant allele frequency of the sample and algorithm settings (at least four mutant reads required CaVEMan and five for Pindel) by running 100,000 simulations. Each simulation was a set of Bernoulli tests with a success probability equal to the median VAF of the sample, and the number of tests was drawn from a Poisson-distributed depth given the median coverage. Crypts with coverages less than 10 were excluded in downstream analysis for mutation rate and disease effect.\u003c/p\u003e\n \u003cp\u003eWith corrected mutation counts, we then fitted linear mixed-effects models using the nlme R package (version 3.1.166) to estimate the contribution of age, constitutional \u003cem\u003eAPC\u003c/em\u003e mutation and disease condition to mutation burden. Crypts from ACF and adenomatous polyps without a somatic \u003cem\u003eAPC\u003c/em\u003e inactivating mutation were excluded in estimation of disease effect. Because crypts from the same individual/ACF/adenomatous polyps are not independent, we controlled these factors by estimating a random effect for each patient/biopsy/group of crypts (crypts from the same ACF or adenomatous polyps belong to the same group). ANOVA tests between models were used to test whether modifications on fixed and random effects should be included into the model or not. For the final models, multiple test correction with the Benjamini\u0026ndash;Hochberg method was applied before reporting the significant coefficients. The workflow and code for this part of analysis will be put in: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/YichenWang1/FAP_colon\u003c/span\u003e\u003c/span\u003e.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec19\" class=\"Section2\"\u003e\n \u003ch2\u003eMutational signature extraction and attribution\u003c/h2\u003e\n \u003cp\u003eMutational signatures were first extracted using HDP\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e (version 0.1.5) without hierarchy and prior information. Mutations mapped to branches on the phylogenetic tree were used as input to avoid counting the same mutation within one patient multiple times. To avoid overfitting, we only kept branches with \u0026gt;\u0026thinsp;50 single-base substitutions (or \u0026gt;\u0026thinsp;30 for indels) as input. The HDP was run in 10 independent chains for 120,000 iterations and with a burn-in of 20,000.\u003c/p\u003e\n \u003cp\u003eIdentified HDP signatures were compared against COSMIC reference signatures\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e67\u003c/span\u003e\u003c/sup\u003e as well as signatures identified in normal colorectal crypts\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e. For confirmation of potential novel signatures, HDP extraction results were compared with mutational signature extraction results from another independent method SigProfilerExtractor\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e (version 1.1.24) with parameters min_sigs\u0026thinsp;=\u0026thinsp;2, max_sigs\u0026thinsp;=\u0026thinsp;20 for substitutions (8 for indels). Non-COSMIC signature SBSC, IDA and IDB were extracted by both algorithms and had been reported in past studies\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e31\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e, therefore were also kept in the final list of signatures. For HDP signatures with \u0026ge;\u0026thinsp;0.9 cosine similarity with a reference signature, branches with such HDP signatures were thought to have the corresponding reference signatures. Remaining HDP signatures were deconvoluted into a shortlist of candidate reference signatures previously found in colorectal cancer and normal colorectal crypts\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e31\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e33\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e32\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e35\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e34\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e50\u003c/span\u003e\u003c/sup\u003e (SBS1, SBS2, SBS5, SBS13, SBS17a, SBS17b, SBS18, SBS28, SBS31, SBS35, SBS88, SBS89, SBS93 and ID1, ID2, ID5, ID14) using expectation maximization. A second round of expectation maximization was further run on reference signatures with \u0026gt;\u0026thinsp;10% contributions for each HDP signature to reduce overfitting. Branches with a HDP signature that contributed to more than 5% of mutations were thought to possess the corresponding deconvoluted reference signatures for that HDP signatures, unless the length of the branch was \u0026le;\u0026thinsp;200 for single-base substitutions (or \u0026le;\u0026thinsp;50 for indels), in which case the HDP signature need to contribute to at least 5% of mutations in at least one longer branch from the same individual to be allowed in.\u003c/p\u003e\n \u003cp\u003eIn this way, the final mutational signatures present in each branch were defined, and the final proportion of reference signatures was estimated using sigfit (version 2.0)\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e68\u003c/span\u003e\u003c/sup\u003e. If any reference signatures contributed no more than 5% of total mutations of the branch, the signatures were removed followed by a second round of fitting using the remaining reference signatures. For the very short branches without an HDP signature extraction result, their signature attributions were approximated by their immediate descendent/ascendent branch when reconstructing the signature burdens for the whole crypt. The workflow and code for this part of analysis will be put in: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/YichenWang1/FAP_colon\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003eTiming the onset for somatic\u003c/strong\u003e \u003cstrong\u003eAPC\u003c/strong\u003e \u003cstrong\u003einactivation\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003eBurdens of lock-like signature SBS5 was used to estimate the age of the somatic \u003cem\u003eAPC\u003c/em\u003e inactivation. A liner mixed-effect models with age as fixed effect, patient as random effect was used to estimate SBS5 mutation rates in all normal crypts (nlme R package version 3.1.166). SBS5 mutation rate for each individual was adjusted by the patient random effect term, and the age of onset for the earliest common ancestor with \u003cem\u003eAPC\u003c/em\u003e inactivation was estimated using the SBS5 burden on the trunk of the phylogenetic tree (where the \u003cem\u003eAPC\u003c/em\u003e somatic mutation was placed on) divided by the SBS5 mutation rate of the corresponding individual. This estimation is an upper limit because the crypt fission might have happened later than the \u003cem\u003eAPC\u003c/em\u003e inactivation.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec21\" class=\"Section2\"\u003e\u003cbr\u003e\u003c/div\u003e"},{"header":"Declarations","content":"\u003ch2\u003eCompeting interests\u003c/h2\u003e\n\u003cp\u003eP.J.C. is a co-founder and CSO of Quotient Therapeutics. M.R.S, and I.M. are co-founders of Quotient Therapeutics. The other authors declare no competing interests.\u003c/p\u003e\n\u003ch2\u003eAuthor contributions\u003c/h2\u003e\n\u003cp\u003eP.S.R., S.Y.L. and M.R.S. conceived the study design, Y.W., P.S.R., S.Y.L. and M.R.S designed the analysis strategy. M.R.S. and S.Y.L. obtained funding. S.Y.L., S.T.Y., H.H.N.Y., C.C.F., A.S.Y.C., A.K.W.C., W.Y.T., J.R.S., H.D.W., and L.T. recruited participants, collected samples and curated sample and clinical data. S.Y.L, S.T.Y. and O.T.G. performed histological analysis of samples. P.S.R., B.C.H.L., Y.H. and K.R. undertook laboratory work. Y.W., P.S.R., H.J. performed data analysis with help and input from S. F., T.H.H.C., A.R.J.L., and S.O.. L.A. contributed to sample management. H.L.-S. and S.O. contributed and analysed the control data. M.R.S., P.J.C. and I.M. oversaw statistical analysis. M.R.S. and S.Y.L oversaw the study. All authors were involved in the preparation and review of the manuscript.\u003c/p\u003e\n\u003ch2\u003eAcknowledgement\u003c/h2\u003e\n\u003cp\u003eWe thank the staff of the Wellcome Sanger Institute Sample Logistics, Genotyping, Pulldown, Sequencing and Informatics facilities for their many contributions, especially L. O\u0026rsquo;Neill, C. Latimer, K. Roberts for their support with sample management and laboratory work. We thank S. Moody, A. P. Le for discussion of the results. We thank Dorothy H.T. Cheng for support with patient coordination and the clinicians in Hong Kong Hospital Authority for clinical care. We thank all the patients and their families, without their support this work would not have been possible. This research is supported by core funding from the Wellcome Trust (206194), the Kadoorie Charitable Foundation, the Hong Kong Cancer Fund, the Centre for Oncology and Immunology under the Health@InnoHK Initiative funded by the Innovation and Technology Commission, the Government of the Hong Kong SAR, China. Y.W., T.H.H.C., A.R.J.L., and S.O. was supported by Wellcome Ph.D. Studentships and P.S.R. by a Wellcome Clinical Ph.D. fellowship. L.E.T was supported by a Health Fellowship from the Welsh Government (HF-14-10, Health and Care Research Wales). J.R.S. was supported by a grant from Health and Care Research Wales.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDNA sequencing data generated for this study are deposited in the European Genome-Phenome Archive (EGA) with accession code EGAD00001015471, and aligned BAM files with accession code EGAD00001015477. The guidelines for patient consent prevent the derived data files from being dispersed by open access. To ensure the data is used for academic and research purposes, controlled access of the data will be available indefinitely upon request made to the WTSI CGP Data access committee.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCode availability\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCode required to reproduce the analyses in this paper is available online. Mutation-calling algorithms are available through GitHub (https://github.com/cancerit). Variant calling filters can be found at https://github.com/cancerit/hairpin2 and https://github.com/hj6-sanger/ascatPCA). The phylogeny reconstruction pipeline is available at https://github.com/TimCoorens/Sequoia. \u0026nbsp;Code for mutational signature analysis and other customised analyses in this study is available online at https://github.com/YichenWang1/FAP_colon.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eLuebeck, E. G. \u0026amp; Moolgavkar, S. H. Multistage carcinogenesis and the incidence of colorectal cancer. \u003cem\u003eProc. Natl. Acad. Sci. U.S.A.\u003c/em\u003e \u003cstrong\u003e99\u003c/strong\u003e, 15095\u0026ndash;15100 (2002).\u003c/li\u003e\n\u003cli\u003eFearon, E. R. \u0026amp; Vogelstein, B. A genetic model for colorectal tumorigenesis. \u003cem\u003eCell\u003c/em\u003e \u003cstrong\u003e61\u003c/strong\u003e, 759\u0026ndash;767 (1990).\u003c/li\u003e\n\u003cli\u003eVogelstein, B. \u003cem\u003eet al.\u003c/em\u003e Genetic Alterations during Colorectal-Tumor Development. \u003cem\u003eN Engl J Med\u003c/em\u003e \u003cstrong\u003e319\u003c/strong\u003e, 525\u0026ndash;532 (1988).\u003c/li\u003e\n\u003cli\u003eLamlum, H. \u003cem\u003eet al.\u003c/em\u003e \u003cem\u003eAPC\u003c/em\u003e mutations are sufficient for the growth of early colorectal adenomas. \u003cem\u003eProc. Natl. Acad. Sci. U.S.A.\u003c/em\u003e \u003cstrong\u003e97\u003c/strong\u003e, 2225\u0026ndash;2228 (2000).\u003c/li\u003e\n\u003cli\u003eLi, J. \u003cem\u003eet al.\u003c/em\u003e Genomic and transcriptomic profiling of carcinogenesis in patients with familial adenomatous polyposis. \u003cem\u003eGut\u003c/em\u003e \u003cstrong\u003e69\u003c/strong\u003e, 1283\u0026ndash;1293 (2020).\u003c/li\u003e\n\u003cli\u003eRashid, M. \u003cem\u003eet al.\u003c/em\u003e Adenoma development in familial adenomatous polyposis and \u003cem\u003e MUTYH \u003c/em\u003e ‐associated polyposis: somatic landscape and driver genes. \u003cem\u003eThe Journal of Pathology\u003c/em\u003e \u003cstrong\u003e238\u003c/strong\u003e, 98\u0026ndash;108 (2016).\u003c/li\u003e\n\u003cli\u003eGausachs, M. \u003cem\u003eet al.\u003c/em\u003e Mutational Heterogeneity in \u003cem\u003eAPC\u003c/em\u003e and \u003cem\u003eKRAS\u003c/em\u003e Arises at the Crypt Level and Leads to Polyclonality in Early Colorectal Tumorigenesis. \u003cem\u003eClinical Cancer Research\u003c/em\u003e \u003cstrong\u003e23\u003c/strong\u003e, 5936\u0026ndash;5947 (2017).\u003c/li\u003e\n\u003cli\u003eKlos, C. L. \u0026amp; Dharmarajan, S. Polyp Genetics. \u003cem\u003eClin Colon Rectal Surg\u003c/em\u003e \u003cstrong\u003e29\u003c/strong\u003e, 289\u0026ndash;295 (2016).\u003c/li\u003e\n\u003cli\u003eFearon, E. R. Molecular Genetics of Colorectal Cancer. \u003cem\u003eAnnu. Rev. Pathol. Mech. Dis.\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 479\u0026ndash;507 (2011).\u003c/li\u003e\n\u003cli\u003eThe Cancer Genome Atlas Network. Comprehensive molecular characterization of human colon and rectal cancer. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e487\u003c/strong\u003e, 330\u0026ndash;337 (2012).\u003c/li\u003e\n\u003cli\u003eCornish, A. J. \u003cem\u003eet al.\u003c/em\u003e The genomic landscape of 2,023 colorectal cancers. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e633\u003c/strong\u003e, 127\u0026ndash;136 (2024).\u003c/li\u003e\n\u003cli\u003eBrenner, H., Kloor, M. \u0026amp; Pox, C. P. Colorectal cancer. \u003cem\u003eThe Lancet\u003c/em\u003e \u003cstrong\u003e383\u003c/strong\u003e, 1490\u0026ndash;1502 (2014).\u003c/li\u003e\n\u003cli\u003eLu, Z. \u003cem\u003eet al.\u003c/em\u003e Polyclonal-to-monoclonal transition in colorectal precancerous evolution. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e636\u003c/strong\u003e, 233\u0026ndash;240 (2024).\u003c/li\u003e\n\u003cli\u003eZaffaroni, G. \u003cem\u003eet al.\u003c/em\u003e Updated European guidelines for clinical management of familial adenomatous polyposis (FAP), MUTYH-associated polyposis (MAP), gastric adenocarcinoma, proximal polyposis of the stomach (GAPPS) and other rare adenomatous polyposis syndromes: a joint EHTG-ESCP revision. \u003cem\u003eBritish Journal of Surgery\u003c/em\u003e \u003cstrong\u003e111\u003c/strong\u003e, znae070 (2024).\u003c/li\u003e\n\u003cli\u003eDunne, P. D. \u0026amp; Arends, M. J. Molecular pathological classification of colorectal cancer\u0026mdash;an update. \u003cem\u003eVirchows Arch\u003c/em\u003e \u003cstrong\u003e484\u003c/strong\u003e, 273\u0026ndash;285 (2024).\u003c/li\u003e\n\u003cli\u003eYan, H. H. N. \u003cem\u003eet al.\u003c/em\u003e RNF43 germline and somatic mutation in serrated neoplasia pathway and its association with BRAF mutation. \u003cem\u003eGut\u003c/em\u003e \u003cstrong\u003e66\u003c/strong\u003e, 1645\u0026ndash;1656 (2017).\u003c/li\u003e\n\u003cli\u003eRoncucci, L., Stamp, D., Medline, A., Cullen, J. B. \u0026amp; Robert Bruce, W. Identification and quantification of aberrant crypt foci and microadenomas in the human colon. \u003cem\u003eHuman Pathology\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e, 287\u0026ndash;294 (1991).\u003c/li\u003e\n\u003cli\u003eDi Gregorio, C. \u003cem\u003eet al.\u003c/em\u003e Histology of aberrant crypt foci in the human colon. \u003cem\u003eHistopathology\u003c/em\u003e \u003cstrong\u003e30\u003c/strong\u003e, 328\u0026ndash;334 (1997).\u003c/li\u003e\n\u003cli\u003eTakayama, T. \u003cem\u003eet al.\u003c/em\u003e Aberrant Crypt Foci of the Colon as Precursors of Adenoma and Cancer. \u003cem\u003eN Engl J Med\u003c/em\u003e \u003cstrong\u003e339\u003c/strong\u003e, 1277\u0026ndash;1284 (1998).\u003c/li\u003e\n\u003cli\u003eCheng, L. Aberrant crypt foci as microscopic precursors of colorectal cancer. \u003cem\u003eWJG\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 2642 (2003).\u003c/li\u003e\n\u003cli\u003eOrlando, F. A. \u003cem\u003eet al.\u003c/em\u003e Aberrant crypt foci as precursors in colorectal cancer progression. \u003cem\u003eJournal of Surgical Oncology\u003c/em\u003e \u003cstrong\u003e98\u003c/strong\u003e, 207\u0026ndash;213 (2008).\u003c/li\u003e\n\u003cli\u003eOtori, K. \u003cem\u003eet al.\u003c/em\u003e Infrequent somatic mutation of the adenomatous polyposis coli gene in aberrant crypt foci of human colon tissue. \u003cem\u003eCancer\u003c/em\u003e \u003cstrong\u003e83\u003c/strong\u003e, 896\u0026ndash;900 (1998).\u003c/li\u003e\n\u003cli\u003eJen, J., Powell, S. M., Smith, K. J., Hamilton, S. \u0026amp; Vogeistein, B. Molecular Determinants of Dysplasia in Colorectal Lesions.\u003c/li\u003e\n\u003cli\u003eSmith, A., Stern, H. \u0026amp; Ga, S. Somatic APC and K-ras Codon 12 Mutations in Aberrant Crypt Foci from Human Colons\u0026rsquo;.\u003c/li\u003e\n\u003cli\u003eTakayama, T. \u003cem\u003eet al.\u003c/em\u003e Analysis of K-ras, APC, and \u0026beta;-catenin in aberrant crypt foci in sporadic adenoma, cancer, and familial adenomatous polyposis. \u003cem\u003eGastroenterology\u003c/em\u003e \u003cstrong\u003e121\u003c/strong\u003e, 599\u0026ndash;611 (2001).\u003c/li\u003e\n\u003cli\u003eLosi, L., Roncucci, L., Di Gregorio, C., De Leon, M. P. \u0026amp; Benhattar, J. K-ras AND p53 mutations in human colorectal aberrant crypt foci. \u003cem\u003eJ. Pathol.\u003c/em\u003e \u003cstrong\u003e178\u003c/strong\u003e, 259\u0026ndash;263 (1996).\u003c/li\u003e\n\u003cli\u003eNucci, M. R., Robinson, C. R., Longo, P., Campbell, P. \u0026amp; Hamilton, S. R. Phenotypic and genotypic characteristics of aberrant crypt foci in human colorectal mucosa. \u003cem\u003eHuman Pathology\u003c/em\u003e \u003cstrong\u003e28\u003c/strong\u003e, 1396\u0026ndash;1407 (1997).\u003c/li\u003e\n\u003cli\u003ePretlow, T. P., Brasitus, T. A., Fulton, N. C., Cheyer, C. \u0026amp; Kaplan, E. L. K-ras Mutations in Putative Preneoplastic Lesions in Human Colon. \u003cem\u003eJNCI Journal of the National Cancer Institute\u003c/em\u003e \u003cstrong\u003e85\u003c/strong\u003e, 2004\u0026ndash;2007 (1993).\u003c/li\u003e\n\u003cli\u003eLuo, L. \u003cem\u003eet al.\u003c/em\u003e Loss of heterozygosity in human aberrant crypt foci (ACF), a putative precursor of colon cancer. \u003cem\u003eCarcinogenesis\u003c/em\u003e \u003cstrong\u003e27\u003c/strong\u003e, 1153\u0026ndash;1159 (2006).\u003c/li\u003e\n\u003cli\u003eEllis, P. \u003cem\u003eet al.\u003c/em\u003e Reliable detection of somatic mutations in solid tissues by laser-capture microdissection and low-input DNA sequencing. \u003cem\u003eNat Protoc\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 841\u0026ndash;871 (2021).\u003c/li\u003e\n\u003cli\u003eLee-Six, H. \u003cem\u003eet al.\u003c/em\u003e The landscape of somatic mutation in normal colorectal epithelial cells. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e574\u003c/strong\u003e, 532\u0026ndash;537 (2019).\u003c/li\u003e\n\u003cli\u003eRobinson, P. S. \u003cem\u003eet al.\u003c/em\u003e Increased somatic mutation burdens in normal human cells due to defective DNA polymerases. \u003cem\u003eNat Genet\u003c/em\u003e \u003cstrong\u003e53\u003c/strong\u003e, 1434\u0026ndash;1442 (2021).\u003c/li\u003e\n\u003cli\u003eRobinson, P. S. \u003cem\u003eet al.\u003c/em\u003e Inherited MUTYH mutations cause elevated somatic mutation rates and distinctive mutational signatures in normal human cells. \u003cem\u003eNat Commun\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 3949 (2022).\u003c/li\u003e\n\u003cli\u003eLee, B. C. H. \u003cem\u003eet al.\u003c/em\u003e Mutational landscape of normal epithelial cells in Lynch Syndrome patients. \u003cem\u003eNat Commun\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 2710 (2022).\u003c/li\u003e\n\u003cli\u003eOlafsson, S. \u003cem\u003eet al.\u003c/em\u003e Somatic Evolution in Non-neoplastic IBD-Affected Colon. \u003cem\u003eCell\u003c/em\u003e \u003cstrong\u003e182\u003c/strong\u003e, 672-684.e11 (2020).\u003c/li\u003e\n\u003cli\u003eRosenberg, D. W. \u003cem\u003eet al.\u003c/em\u003e Mutations in \u003cem\u003eBRAF\u003c/em\u003e and \u003cem\u003eKRAS\u003c/em\u003e Differentially Distinguish Serrated versus Non-Serrated Hyperplastic Aberrant Crypt Foci in Humans. \u003cem\u003eCancer Research\u003c/em\u003e \u003cstrong\u003e67\u003c/strong\u003e, 3551\u0026ndash;3554 (2007).\u003c/li\u003e\n\u003cli\u003eSegditsas, S. \u003cem\u003eet al.\u003c/em\u003e APC and the three-hit hypothesis. \u003cem\u003eOncogene\u003c/em\u003e \u003cstrong\u003e28\u003c/strong\u003e, 146\u0026ndash;155 (2009).\u003c/li\u003e\n\u003cli\u003eNovelli, M. R. \u003cem\u003eet al.\u003c/em\u003e Polyclonal Origin of Colonic Adenomas in an XO/XY Patient with FAP. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e272\u003c/strong\u003e, 1187\u0026ndash;1190 (1996).\u003c/li\u003e\n\u003cli\u003eSadien, I. D. \u003cem\u003eet al.\u003c/em\u003e Polyclonality overcomes fitness barriers in Apc-driven tumorigenesis. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e634\u003c/strong\u003e, 1196\u0026ndash;1203 (2024).\u003c/li\u003e\n\u003cli\u003eLi, Y. \u003cem\u003eet al.\u003c/em\u003e Patterns of somatic structural variation in human cancer genomes. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e578\u003c/strong\u003e, 112\u0026ndash;121 (2020).\u003c/li\u003e\n\u003cli\u003eIslam, S. M. A. \u003cem\u003eet al.\u003c/em\u003e Uncovering novel mutational signatures by de novo extraction with SigProfilerExtractor. \u003cem\u003eCell Genomics\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e, 100179 (2022).\u003c/li\u003e\n\u003cli\u003eAlexandrov, L. B. \u003cem\u003eet al.\u003c/em\u003e Clock-like mutational processes in human somatic cells. \u003cem\u003eNat Genet\u003c/em\u003e \u003cstrong\u003e47\u003c/strong\u003e, 1402\u0026ndash;1407 (2015).\u003c/li\u003e\n\u003cli\u003eMoore, L. \u003cem\u003eet al.\u003c/em\u003e The mutational landscape of human somatic and germline cells. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e597\u003c/strong\u003e, 381\u0026ndash;386 (2021).\u003c/li\u003e\n\u003cli\u003eKucab, J. E. \u003cem\u003eet al.\u003c/em\u003e A Compendium of Mutational Signatures of Environmental Agents. \u003cem\u003eCell\u003c/em\u003e \u003cstrong\u003e177\u003c/strong\u003e, 821-836.e16 (2019).\u003c/li\u003e\n\u003cli\u003eCoorens, T. H. H. \u003cem\u003eet al.\u003c/em\u003e Inherent mosaicism and extensive mutation of human placentas. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e592\u003c/strong\u003e, 80\u0026ndash;85 (2021).\u003c/li\u003e\n\u003cli\u003eWang, Y. \u003cem\u003eet al.\u003c/em\u003e APOBEC mutagenesis is a common process in normal human small intestine. \u003cem\u003eNat Genet\u003c/em\u003e (2023) doi:10.1038/s41588-022-01296-5.\u003c/li\u003e\n\u003cli\u003ePleguezuelos-Manzano, C. \u003cem\u003eet al.\u003c/em\u003e Mutational signature in colorectal cancer caused by genotoxic pks+ E. coli. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e580\u003c/strong\u003e, 269\u0026ndash;273 (2020).\u003c/li\u003e\n\u003cli\u003eSecrier, M. \u003cem\u003eet al.\u003c/em\u003e Mutational signatures in esophageal adenocarcinoma define etiologically distinct subgroups with therapeutic relevance. \u003cem\u003eNat Genet\u003c/em\u003e \u003cstrong\u003e48\u003c/strong\u003e, 1131\u0026ndash;1141 (2016).\u003c/li\u003e\n\u003cli\u003eChristensen, S. \u003cem\u003eet al.\u003c/em\u003e 5-Fluorouracil treatment induces characteristic T\u0026gt;G mutations in human cancer. \u003cem\u003eNat Commun\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, 4571 (2019).\u003c/li\u003e\n\u003cli\u003eAlexandrov, L. B. \u003cem\u003eet al.\u003c/em\u003e The repertoire of mutational signatures in human cancer. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e578\u003c/strong\u003e, 94\u0026ndash;101 (2020).\u003c/li\u003e\n\u003cli\u003eCoorens, T. H. H. \u003cem\u003eet al.\u003c/em\u003e Reconstructing phylogenetic trees from genome-wide somatic mutations in clonal samples. \u003cem\u003eNat Protoc\u003c/em\u003e (2024) doi:10.1038/s41596-024-00962-8.\u003c/li\u003e\n\u003cli\u003eVermeulen, L. \u003cem\u003eet al.\u003c/em\u003e Defining Stem Cell Dynamics in Models of Intestinal Tumor Initiation. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e342\u003c/strong\u003e, 995\u0026ndash;998 (2013).\u003c/li\u003e\n\u003cli\u003eJones, D. \u003cem\u003eet al.\u003c/em\u003e cgpCaVEManWrapper: Simple Execution of CaVEMan in Order to Detect Somatic Single Nucleotide Variants in NGS Data. \u003cem\u003eCurrent Protocols in Bioinformatics\u003c/em\u003e \u003cstrong\u003e56\u003c/strong\u003e, 15.10.1-15.10.18 (2016).\u003c/li\u003e\n\u003cli\u003eKim, S. \u003cem\u003eet al.\u003c/em\u003e Strelka2: fast and accurate calling of germline and somatic variants. \u003cem\u003eNat Methods\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 591\u0026ndash;594 (2018).\u003c/li\u003e\n\u003cli\u003eHoang, D. T. \u003cem\u003eet al.\u003c/em\u003e MPBoot: fast phylogenetic maximum parsimony tree inference and bootstrap approximation. \u003cem\u003eBMC Evol Biol\u003c/em\u003e \u003cstrong\u003e18\u003c/strong\u003e, 11 (2018).\u003c/li\u003e\n\u003cli\u003eYu, G. Using ggtree to Visualize Data on Tree‐Like Structures. \u003cem\u003eCurrent Protocols in Bioinformatics\u003c/em\u003e \u003cstrong\u003e69\u003c/strong\u003e, (2020).\u003c/li\u003e\n\u003cli\u003eParadis, E. \u0026amp; Schliep, K. ape 5.0: an environment for modern phylogenetics and evolutionary analyses in R. \u003cem\u003eBioinformatics\u003c/em\u003e \u003cstrong\u003e35\u003c/strong\u003e, 526\u0026ndash;528 (2019).\u003c/li\u003e\n\u003cli\u003eYe, K., Schulz, M. H., Long, Q., Apweiler, R. \u0026amp; Ning, Z. Pindel: a pattern growth approach to detect break points of large deletions and medium sized insertions from paired-end short reads. \u003cem\u003eBioinformatics\u003c/em\u003e \u003cstrong\u003e25\u003c/strong\u003e, 2865\u0026ndash;2871 (2009).\u003c/li\u003e\n\u003cli\u003eRaine, K. M. \u003cem\u003eet al.\u003c/em\u003e cgpPindel: Identifying Somatically Acquired Insertion and Deletion Events from Paired End Sequencing. \u003cem\u003eCP in Bioinformatics\u003c/em\u003e \u003cstrong\u003e52\u003c/strong\u003e, (2015).\u003c/li\u003e\n\u003cli\u003eMartincorena, I. \u003cem\u003eet al.\u003c/em\u003e Universal Patterns of Selection in Cancer and Somatic Tissues. \u003cem\u003eCell\u003c/em\u003e \u003cstrong\u003e171\u003c/strong\u003e, 1029-1041.e21 (2017).\u003c/li\u003e\n\u003cli\u003eGao, J. \u003cem\u003eet al.\u003c/em\u003e Integrative Analysis of Complex Cancer Genomics and Clinical Profiles Using the cBioPortal. \u003cem\u003eScience Signaling\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, pl1\u0026ndash;pl1 (2013).\u003c/li\u003e\n\u003cli\u003eWang, K., Li, M. \u0026amp; Hakonarson, H. ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data. \u003cem\u003eNucleic Acids Research\u003c/em\u003e \u003cstrong\u003e38\u003c/strong\u003e, e164\u0026ndash;e164 (2010).\u003c/li\u003e\n\u003cli\u003eVan Loo, P. \u003cem\u003eet al.\u003c/em\u003e Allele-specific copy number analysis of tumors. \u003cem\u003eProceedings of the National Academy of Sciences\u003c/em\u003e \u003cstrong\u003e107\u003c/strong\u003e, 16910\u0026ndash;16915 (2010).\u003c/li\u003e\n\u003cli\u003eRaine, K. M. \u003cem\u003eet al.\u003c/em\u003e ascatNgs: Identifying Somatically Acquired Copy‐Number Alterations from Whole‐Genome Sequencing Data. \u003cem\u003eCP in Bioinformatics\u003c/em\u003e \u003cstrong\u003e56\u003c/strong\u003e, (2016).\u003c/li\u003e\n\u003cli\u003eCameron, D. L. \u003cem\u003eet al.\u003c/em\u003e GRIDSS2: comprehensive characterisation of somatic structural variation using single breakend variants and structural variant phasing. \u003cem\u003eGenome Biol\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e, 202 (2021).\u003c/li\u003e\n\u003cli\u003eTarailo‐Graovac, M. \u0026amp; Chen, N. Using RepeatMasker to Identify Repetitive Elements in Genomic Sequences. \u003cem\u003eCP in Bioinformatics\u003c/em\u003e \u003cstrong\u003e25\u003c/strong\u003e, (2009).\u003c/li\u003e\n\u003cli\u003eSondka, Z. \u003cem\u003eet al.\u003c/em\u003e COSMIC: a curated database of somatic variants and clinical data for cancer. \u003cem\u003eNucleic Acids Research\u003c/em\u003e \u003cstrong\u003e52\u003c/strong\u003e, D1210\u0026ndash;D1217 (2024).\u003c/li\u003e\n\u003cli\u003eGori, K. \u0026amp; Baez-Ortega, A. \u003cem\u003eSigfit: Flexible Bayesian Inference of Mutational Signatures\u003c/em\u003e. http://biorxiv.org/lookup/doi/10.1101/372896 (2018) doi:10.1101/372896.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-6581155/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6581155/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe succession of somatic genetic events associated with the conversion of a normal colorectal epithelial cell into a colorectal carcinoma constitutes a paradigmatic model of cancer development. Familial Adenomatous Polyposis (FAP) is caused by constitutional inactivating mutations in \u003cem\u003eAPC\u003c/em\u003e, the central gatekeeper gene of colorectal cancer, and is associated with a substantially increased lifetime-risk of colorectal cancer. To investigate the earliest stages of neoplastic change due to \u003cem\u003eAPC\u003c/em\u003e inactivation, we microdissected and individually whole genome sequenced 279 histologically normal and abnormal colorectal crypts from 15 individuals with FAP. Histologically normal crypts generally exhibited similar mutation burdens and mutational signatures to normal crypts from wild-type individuals of the same age, with 1/110 carrying a somatic inactivating \u003cem\u003eAPC\u003c/em\u003e mutation. By contrast, 9/18 aberrant crypt foci carried somatic \u003cem\u003eAPC\u003c/em\u003e mutations and exhibited modestly increased burdens of some mutational signatures found in normal crypts. 12/13 diminutive adenomatous polyps (\u0026lt;\u0026thinsp;5mm diameter) showed somatic \u003cem\u003eAPC\u003c/em\u003e mutations and carried substantially increased mutation loads of most mutational signatures present in normal crypts. Phylogenetic trees of crypts from aberrant crypt foci and adenomatous polyps revealed that some had acquired their initiating somatic \u003cem\u003eAPC\u003c/em\u003e mutations decades previously during the first few years of life. The results catalogue the changes in somatic mutation rates, mutational processes and \u0026ldquo;driver\u0026rdquo; mutations in cancer genes during the earliest stages of colorectal neoplastic transformation initiated by \u003cem\u003eAPC\u003c/em\u003e inactivation and highlight the long periods of clonal evolution required for a cancer to develop.\u003c/p\u003e","manuscriptTitle":"The earliest stages of neoplastic transformation in Familial Adenomatous Polyposis","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-05-22 13:45:37","doi":"10.21203/rs.3.rs-6581155/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"035e3769-fc18-40d4-8da0-a0288206acfb","owner":[],"postedDate":"May 22nd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":48803488,"name":"Biological sciences/Genetics/Genomics"},{"id":48803489,"name":"Biological sciences/Cancer/Gastrointestinal cancer/Colorectal cancer"}],"tags":[],"updatedAt":"2025-07-11T10:25:44+00:00","versionOfRecord":[],"versionCreatedAt":"2025-05-22 13:45:37","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6581155","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6581155","identity":"rs-6581155","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}