Is there a common developmental mechanism underlying childhood cancers?

Xiao-Lin Guo; Hua Zhu; Deng-Li Hong

doi:10.15302/vita.2025.12.0003

Vita > Expert Views > DOI: 10.15302/vita.2025.12.0003

Vita Published:

Is there a common developmental mechanism underlying childhood cancers?

Xiao-Lin Guo ¹^,^# ,
Hua Zhu ²^,^# ,
Deng-Li Hong ¹^,³

Author information +

History +

Vita () Cite this article

PDF (1200KB)

登录浏览全文

4963

注册一个新账户忘记密码

Childhood cancers are closely linked to embryonic and early postnatal development¹, distinguishing them from adult cancers, which are primarily driven by aging. These malignancies mainly affect the hematopoietic and neural systems² — tissues inherently susceptible to oncogenesis during early development — and highlight the developmental origins of pediatric cancers. Based on insights from childhood leukemia and tissue development³, we propose a unified mechanism for pediatric cancer etiology. Tissue-forming cells (TFCs) migrate through developing organs and undergo proliferation, differentiation, DNA replication, and transcription processes, requiring precise microenvironmental regulation to maintain genomic integrity. When this control fails, replication–transcription conflicts cause genomic instability; initiating mutations can then transform TFCs into pre-cancer stem cells, which evolve into malignant cells through additional genetic alterations.

Early development involves rapid proliferation of the three germ layers (ectoderm, mesoderm, and endoderm), generating diverse functional cell types. During this phase, TFCs produce mature lineages through tightly regulated proliferation and differentiation. These processes involve repeated DNA replication and active transcription, placing high demands on genomic stability. Thus, preserving genomic integrity in TFCs is essential.

Oncogenic mutations in pediatric cancers largely arise in TFCs during development^4,5. However, how developing tissues protect these genomes and how defects in this protective mechanism lead to DNA damage under stress remain unclear. Dividing cells are prone to DNA damage during replication, especially when replication and transcription occur simultaneously, causing collisions between molecular machineries⁶. TFCs undergo both processes at once — as seen in fetal liver hematopoiesis, where hematopoietic stem and progenitor cells (HSPCs) show coordinated replication and transcription³ — greatly increasing their risk of genomic instability.

To reduce this risk, proliferating cells rely on intrinsic genome-protective mechanisms shaped by external signals^3,7,8. Developmental microenvironments play key roles in maintaining genomic stability³. In the fetal liver, hepatocytes act as niche cells that protect HSPCs by secreting fetuin-A, a glycoprotein with regulatory functions. Fetuin-A activates Toll-like receptor (TLR) signaling, increasing expression of the helicase BLM, which resolves R-loops, nucleic acid structures formed by RNA hybridizing with template DNA. R-loops are strongly associated with replication–transcription conflicts and genomic instability³.

Here, we summarize key findings from studies of fetal liver hematopoiesis and leukemogenesis³ and compare developmental processes in the hematopoietic and neural systems, the two main targets of pediatric cancers. From this analysis, we propose a shared mechanistic framework for childhood cancer initiation. Importantly, this framework may extend beyond blood tissues to nervous and other organ systems, offering a broader understanding of the developmental origins of pediatric malignancies.

DEVELOPMENTAL TFC MIGRATION AND CHILDHOOD CANCER VULNERABILITY

TFCs from all three germ layers expand significantly during development. Yet this expansion does not explain why childhood cancers occur predominantly in the hematopoietic and nervous systems². This pattern is likely due to unique developmental features of these tissues that inherently increase their susceptibility to genomic instability and oncogenic transformation.

HSPC migration and leukemogenesis vulnerability

Hematopoietic development has three key features⁹ (Fig. 1a). First, hematopoiesis shifts across sites: in mice, the fetal liver is the main site from E12.5 to E18.5, with earlier activity in the placenta and aorta–gonad–mesonephros (AGM) region, and the bone marrow takes over after birth (from E19.5). Second, HSPCs arise de novo primarily in the AGM, but not in the fetal liver or bone marrow. Third, HSPCs migrate via blood in two major waves: from the placenta to the fetal liver (E11.5–E12.5) and from the fetal liver to the bone marrow (E16.5–E19.5). These transitions expose HSPCs to genotoxic stress, especially during the early fetal liver phase (E12.5–E14.5 in mice; first trimester in humans), a critical window of susceptibility³.

The fetal liver supports both liver and blood development. Fetal hepatocytes help maintain HSPC genomic stability through paracrine signaling, but this protection is limited early in gestation due to the few hepatocytes relative to expanding HSPCs. During this period, HSPCs are highly vulnerable to DNA damage and accumulate mutations in genes critical for hematopoiesis, DNA replication, and epigenetic regulation, including Runx1, Pax5, Mll, Af4/9, Dnmt3b, Tet1/2/3, and Flt3, genes commonly altered in childhood leukemias³. These mutations increase postnatal leukemia risk. Maternal exposure to harmful substances worsens DNA damage in fetal HSPCs, further weakening genomic integrity³.

In pathological conditions, this vulnerable period may extend or protective mechanisms may fail, leading to persistent DNA damage and sustained genomic instability, promoting oncogenic mutations. Children with leukemia have lower bone marrow fetuin-A levels than those with benign conditions³, suggesting prolonged developmental vulnerability, though the cause remains unclear. Similarly, HSPCs from fetuin-A-deficient mice show persistent genomic instability and higher leukemia susceptibility³.

Studying hematopoietic development³, especially when microenvironments fail to protect TFC genomes, can improve understanding of childhood blood cancers. This knowledge may inform prevention strategies and clarify the developmental basis of clinical differences, such as why leukemia differs between children and adults, why childhood leukemias vary in subtype, pathology, and prognosis, and why infant leukemia has particularly poor outcomes. By comparing hematopoietic and neural development, we propose a shared mechanism for tumor initiation in the nervous system.

Neural TFC migration and cancer vulnerability

The nervous system shares key developmental features with hematopoiesis (Fig. 1b, c), a similarity that may explain the high incidence of tumors during neurodevelopment². Both systems are widely distributed: blood cells reach nearly all tissues, and nerve fibers innervate almost every organ. This shared biological demand suggests similar developmental trajectories and comparable susceptibility to malignant transformation.

First, neurogenesis involves dynamic spatial transitions^10,11. It begins with the neural plate in the dorsal ectoderm, which folds to form the neural tube and neural crest. The neural tube gives rise to the central nervous system (CNS)¹⁰, while the neural crest generates the peripheral nervous system¹¹, including sensory, autonomic, and enteric ganglia, and the adrenal medulla. The neural tube bends and expands to form major brain structures such as the cerebrum and cerebellum¹⁰.

Second, neural TFCs — neural stem and progenitor cells (NSPCs) — arise in discrete niches^10-12: the ventricular zone (VZ), cerebellar rhombic lip (CRL), and neural crest.

Third, NSPCs migrate extensively between neurogenic regions^10-12, much like HSPCs⁹. VZ- and CRL-derived NSPCs migrate radially to form the cerebral and cerebellar cortices; neural crest cells undergo long-distance migration to target tissues.

These developmental transitions and migrations expose NSPCs to genotoxic stress, particularly in new microenvironments that lack robust genome-protective mechanisms. This vulnerability likely contributes to the most common neural tumors²: supratentorial gliomas in the cerebrum¹³, medulloblastomas in the cerebellum¹⁴, and neuroblastomas outside the CNS¹⁵.

In cerebral development¹⁰ (Fig. 1b), NSPCs in the VZ generate radial glial cells (RGCs) that extend processes from the ventricle to the pial surface. RGCs divide asymmetrically to produce neurons or intermediate progenitor cells (IPCs), whose progeny migrate along RGC fibers to form the cortical plate in an inside-out manner: early-born neurons settle deep, while later-born ones migrate past them. Microenvironmental failure to protect NSPC genomes can lead to supratentorial gliomas. Two observations¹³ support this: 1) these tumors — astrocytomas, oligodendrogliomas, and glioblastomas — reflect the glial identity of cerebrum-derived NSPCs; 2) diverse glial differentiation pathways explain their histological and molecular heterogeneity.

In cerebellar development¹² (Fig. 1b), NSPCs arise in two regions: the cerebellar ventricular zone (CVZ) and CRL. CVZ-derived cells become GABAergic neurons, including Purkinje cells, and migrate into the cerebellar anlage. The CRL first produces glutamatergic neurons for deep nuclei, then generates granule neuron progenitors (GNPs) by E11.5 in mice. GNPs migrate over the anlage to form the external granule layer (EGL), where they proliferate intensely, peaking at P7 and ending by P15, driving cerebellar growth. Genomic instability in GNPs or NSPCs during migration, proliferation, or differentiation, especially due to insufficient microenvironmental protection, can lead to medulloblastoma. Medulloblastoma has four subgroups¹⁴ — WNT, SHH, Group 3, and Group 4 — each linked to recurrent mutations in key developmental pathways. These mechanisms resemble those in fetal liver leukemia³.

In neural crest development¹¹, neural crest progenitors (NCPs) migrate under local signals (Fig. 1c). In sympathoadrenal development, they move ventrally from the dorsal neural crest, guided by somites, neural tube, notochord, and dorsal aorta. After forming primary sympathetic ganglia (PSGs), precursor cells differentiate into sympathetic neurons or chromaffin cells of the adrenal medulla. MYCN is highly expressed in migrating NCPs, promoting migration and proliferation, and is regulated by the ARF-p53 pathway. Recurrent neuroblastoma alterations¹⁵ — MYCN amplification, TERT rearrangements, ATRX mutations, 17q gain, 1p/11q loss — reflect genomic instability in NCPs at critical stages, especially when microenvironmental support is insufficient. This vulnerability initiates neuroblastoma.

In summary, the neural and hematopoietic systems share core traits — widespread germinal zones and persistent TFC migration — that likely increase cancer risk. Yet the precise developmental vulnerabilities — similar to those in the early fetal liver — that cause genomic instability remain poorly understood.

FUTURE PERSPECTIVES

Our findings in hematopoietic development³, combined with nervous system features, reveal a shared mechanism in childhood cancers. TFCs migrate across organs and undergo proliferation, differentiation, DNA replication, and transcription, making them vulnerable to genomic instability and cancer development when microenvironments fail to protect their genomes (Fig. 2). This risk is highest in multi-site systems like the hematopoietic and nervous systems. These insights highlight the need to understand how genetic and epigenetic stabilities are established during development. Next steps include identifying disease-driving developmental constraints and strategies to reduce cancer risk.

Uncovering TFC genome protection in developing tissues

The mechanisms maintaining genetic and epigenetic stability in TFCs during organogenesis are not fully understood. As TFCs undergo repeated DNA replication and transcriptional reprogramming, local microenvironments must safeguard their genomes, as seen in the fetal liver³.

In the fetal liver, hepatocytes secrete fetuin-A, which protects HSPCs by activating TLR signaling³. Fetuin-A enters circulation after release¹⁶. Whether this protection extends to other organs remains unknown. HSPCs in fetuin-A knockout mice show increased genomic instability in neonatal bone marrow³, confirming its protective role. It is unclear whether similar mechanisms exist elsewhere.

TLRs are widely expressed in embryonic tissues^17,18 and contribute to development¹⁹, beyond their known role in inflammation²⁰. In the fetal liver, TLR activation maintains HSPC genome stability³, suggesting broader roles in developmental genome homeostasis.

Fetuin-A cannot cross the placental or blood–brain barriers^21,22, but TLRs are present in neural TFCs (unpublished data). If TLR4 has a similar protective function in neural TFCs, alternative endogenous ligands must be identified.

Identifying disease-vulnerable developmental features

Complex adaptations in higher mammals improve survival but involve trade-offs — developmental constraints — that can create vulnerabilities²³. Hematopoietic development relies on stage-specific microenvironments that support HSPC expansion and blood cell production⁹. The fetal liver becomes active at E12.5, with HSPC numbers rising by E16.5 before migration to bone marrow⁹. A key trade-off exists: HSPCs are exposed to genotoxic stress when protective mechanisms lag³. Delayed hepatocyte development leaves HSPCs without fetuin-A, increasing childhood leukemia risk³.

Given the similarities between hematopoietic and nervous systems, and potentially others, similar constraints likely exist in neural and other tissues. Germinal zones host proliferating and differentiating TFCs^24,25. It is critical to map niche composition and determine whether it forms before or after TFC arrival. Timing mismatches may leave TFCs unprotected. Identifying missing protective factors is essential for understanding developmental cancer risks.

Addressing developmental constraints to prevent pediatric cancers

Our studies show that fetuin-A deficiency in the early fetal liver — a developmental constraint — can be corrected by intraplacental delivery of recombinant fetuin-A in mice³ (Fig. 2). This suggests that targeting the fetuin-A-TLR-BLM axis with drug-like compounds could enable maternal therapies that stabilize TFC genomes, prevent mutations, and reduce pediatric cancer risk.

Once similar constraints, such as missing genome-protective factors, are identified in the brain or other organs prone to childhood cancers, similar interventions should be applied (Fig. 2). Restoring deficient factors or using engineered analogs that cross the placenta or blood–brain barrier could protect vulnerable neural TFCs during rapid fetal and neonatal development. This approach may prevent mutation accumulation and reduce the incidence of pediatric brain tumors, a leading cause of childhood cancer death.

CONCLUSIONS

In summary, shared developmental features — TFC migration, concurrent replication and transcription, and dependence on tissue-specific microenvironmental protection — across hematopoietic, nervous, and other organ systems prone to childhood cancers collectively shape their susceptibility. Understanding developmental constraints — the trade-offs inherent in essential developmental functions — provides key mechanistic insights into pediatric cancer origins. Identifying these context-specific vulnerabilities and developing strategies to restore missing protective factors or target their pathways could enable the prevention of childhood cancers and related developmental disorders (Fig. 2), translating this framework into clinical potential.

[1]	Marshall, G.M. et al. Nat. Rev. Cancer 14, 277–289 (2014).

[2]	Ni, X. et al. Lancet 400, 1020–1032 (2022).

[3]	Guo, X.L. et al. Nature 637, 402–411 (2025).

[4]	Hong, D.L. et al. Science 319, 336–339 (2008).

[5]	Jansky, S. et al. Nat. Genet. 53, 683–693 (2021).

[6]	Goehring, L., Huang, T.T. & Smith, D.J. Annu. Rev. Genet. 57, 157–179 (2023).

[7]	De Laval B. et al. Cell Stem Cell 12, 37–48 (2013).

[8]	Zhou, W.J., Geng, Z.H., Spence, J.R. & Geng, J.G. Nature 501, 107–111 (2013).

[9]	Mikkola, H.K.A. & Orkin, S.H. Development 133, 3733–3744 (2006).

[10]	Ahirwar, L.K., Blackburn, S.L., McBride, D.W. & Kumar, T.P. Mol. Psychiatry 30, 5953–5966 (2025).

[11]	Murtazina, A. & Adameyko, I. Development 150, dev201164 (2023).

[12]	Consalez, G.G., Goldowitz, D., Casoni, F. & Hawkes, R. Front. Neural Circuits 14, 611841 (2021).

[13]	Laug, D., Glasgow, S.M. & Deneen, B. Nat. Rev. Neurosci. 19, 393–403 (2018).

[14]	Kiang, K.M., Wong, Y.K.H., Sengupta, S., Roussel, M.F. & Lu, Q.R. Annu. Rev. Neurosci. 48, 85–102 (2025).

[15]	Zeineldin, M., Patel, A.G. & Dyer, M.A. Neuron 110, 2916–2928 (2022).

[16]	Icer, M.A. & Yıldıran, H. Clin. Biochem 88, 1–10 (2021).

[17]	Arnaboldi, F. et al. Ann. Anat. 231, 151526 (2020).

[18]	Hamonic, G., Pasternak, J.A. & Wilson, H.L. Cell Tissue Res. 372, 1–11 (2018).

[19]	He, Q.P. & Liu, F. Curr. Opin. Hematol. 23, 18–22 (2016).

[20]	Kawai, T., Ikegawa, M., Ori, D. & Akira, S. Immunity 57, 649–673 (2024).

[21]	Desforges, M. & Sibley, C.P. Int. J. Dev. Biol. 54, 377–390 (2010).

[22]	Bradbury, M.W. Fed. Proc. 43, 186–190 (1984).

[23]	Williams, G.C. Evolution 11, 398–411 (1957).

[24]	Lathia, J.D., Rao, M.S., Mattson, M.P. & Ffrench-Constant, C. Dev. Dyn. 236, 3267–3282 (2007).

[25]	Cain, T.L., Derecka, M. & McKinney-Freeman, S. Nat. Rev. Mol. Cell Biol. 26, 32–50 (2025).

RIGHTS & PERMISSIONS

The Author(s) 2026. Published by Higher Education Press. This is an Open Access article distributed under the terms of the CC BY license (https://creativecommons.org/licenses/by/4.0/).

Cite this article

Download citation ▾

[Author(id=1221046087034163477, tenantId=1045748351789510663, journalId=1169329684812255232, articleId=1217837792865592187, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1221046087097078040, tenantId=1045748351789510663, journalId=1169329684812255232, articleId=1217837792865592187, authorId=1221046087034163477, language=EN, stringName=Xiao-Lin Guo, firstName=Xiao-Lin, middleName=null, lastName=Guo, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=¹^,^#, address=¹. Institute for Translational Medicine on Cell Fate and Disease, Shanghai Ninth People's Hospital; Key Laboratory of Cell Differentiation and Apoptosis of Ministry of Education, Department of Pathophysiology; Shanghai Institute of Hematology, National Research Center for Translation Medicine, Ruijin Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1221046086858002700, tenantId=1045748351789510663, journalId=1169329684812255232, articleId=1217837792865592187, xref=1, ext=[AuthorCompanyExt(id=1221046086874779917, tenantId=1045748351789510663, journalId=1169329684812255232, articleId=1217837792865592187, companyId=1221046086858002700, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=¹. Institute for Translational Medicine on Cell Fate and Disease, Shanghai Ninth People's Hospital; Key Laboratory of Cell Differentiation and Apoptosis of Ministry of Education, Department of Pathophysiology; Shanghai Institute of Hematology, National Research Center for Translation Medicine, Ruijin Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China)])]), Author(id=1221046087151603994, tenantId=1045748351789510663, journalId=1169329684812255232, articleId=1217837792865592187, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1221046087214518556, tenantId=1045748351789510663, journalId=1169329684812255232, articleId=1217837792865592187, authorId=1221046087151603994, language=EN, stringName=Hua Zhu, firstName=Hua, middleName=null, lastName=Zhu, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=²^,^#, address=². Shanghai Children’s Medical Center, Shanghai Jiao Tong University School of Medicine, Shanghai, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1221046086916722958, tenantId=1045748351789510663, journalId=1169329684812255232, articleId=1217837792865592187, xref=2, ext=[AuthorCompanyExt(id=1221046086933500175, tenantId=1045748351789510663, journalId=1169329684812255232, articleId=1217837792865592187, companyId=1221046086916722958, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=². Shanghai Children’s Medical Center, Shanghai Jiao Tong University School of Medicine, Shanghai, China)])]), Author(id=1221046087260655902, tenantId=1045748351789510663, journalId=1169329684812255232, articleId=1217837792865592187, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=dlhong@sjtu.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1221046087336153377, tenantId=1045748351789510663, journalId=1169329684812255232, articleId=1217837792865592187, authorId=1221046087260655902, language=EN, stringName=Deng-Li Hong, firstName=Deng-Li, middleName=null, lastName=Hong, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=¹^,³, address=¹. Institute for Translational Medicine on Cell Fate and Disease, Shanghai Ninth People's Hospital; Key Laboratory of Cell Differentiation and Apoptosis of Ministry of Education, Department of Pathophysiology; Shanghai Institute of Hematology, National Research Center for Translation Medicine, Ruijin Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China
³. Shanghai Key Laboratory of Reproductive Medicine, Shanghai Jiao Tong University School of Medicine, Shanghai, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1221046086858002700, tenantId=1045748351789510663, journalId=1169329684812255232, articleId=1217837792865592187, xref=1, ext=[AuthorCompanyExt(id=1221046086874779917, tenantId=1045748351789510663, journalId=1169329684812255232, articleId=1217837792865592187, companyId=1221046086858002700, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=¹. Institute for Translational Medicine on Cell Fate and Disease, Shanghai Ninth People's Hospital; Key Laboratory of Cell Differentiation and Apoptosis of Ministry of Education, Department of Pathophysiology; Shanghai Institute of Hematology, National Research Center for Translation Medicine, Ruijin Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China)]), AuthorCompany(id=1221046086975443217, tenantId=1045748351789510663, journalId=1169329684812255232, articleId=1217837792865592187, xref=3, ext=[AuthorCompanyExt(id=1221046086992220434, tenantId=1045748351789510663, journalId=1169329684812255232, articleId=1217837792865592187, companyId=1221046086975443217, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=³. Shanghai Key Laboratory of Reproductive Medicine, Shanghai Jiao Tong University School of Medicine, Shanghai, China)])])]

Guo, X., Zhu, H., Hong, D. Is there a common developmental mechanism underlying childhood cancers? Vita https://doi.org/10.15302/vita.2025.12.0003 ()