Resolving medulloblastoma cellular architecture by single-cell genomics

Article metrics

Abstract

Medulloblastoma is a malignant childhood cerebellar tumour type that comprises distinct molecular subgroups. Whereas genomic characteristics of these subgroups are well defined, the extent to which cellular diversity underlies their divergent biology and clinical behaviour remains largely unexplored. Here we used single-cell transcriptomics to investigate intra- and intertumoral heterogeneity in 25 medulloblastomas spanning all molecular subgroups. WNT, SHH and Group 3 tumours comprised subgroup-specific undifferentiated and differentiated neuronal-like malignant populations, whereas Group 4 tumours consisted exclusively of differentiated neuronal-like neoplastic cells. SHH tumours closely resembled granule neurons of varying differentiation states that correlated with patient age. Group 3 and Group 4 tumours exhibited a developmental trajectory from primitive progenitor-like to more mature neuronal-like cells, the relative proportions of which distinguished these subgroups. Cross-species transcriptomics defined distinct glutamatergic populations as putative cells-of-origin for SHH and Group 4 subtypes. Collectively, these data provide insights into the cellular and developmental states underlying subtype-specific medulloblastoma biology.

Access options

Rent or Buy article

Get time limited or full article access on ReadCube.

from$8.99

All prices are NET prices.

Fig. 1: Integrated analysis of MB and cerebellar single-cell transcriptomes.
Fig. 2: Intratumoral heterogeneity in WNT MB.
Fig. 3: Age-associated developmental hierarchies in SHH MB.
Fig. 4: Malignant transcriptional programs within Group 3/4.
Fig. 5: Cellular composition of Group 3/4 MBs.
Fig. 6: Subgroup-specific transcriptional programs correlate with distinct neuronal lineages.

Data availability

The scRNA-seq and array-based DNA methylation data of 36 patient and PDX samples described in this study have been deposited in the Gene Expression Omnibus (GEO) with the accession code GSE119926. The scRNA-seq data of the developing mouse cerebellum have been deposited to the European Nucleotide Archive (ENA) with the accession code PRJEB23051.

References

  1. 1.

    Gajjar, A. J. & Robinson, G. W. Medulloblastoma-translating discoveries from the bench to the bedside. Nat. Rev. Clin. Oncol. 11, 714–722 (2014).

  2. 2.

    Taylor, M. D. et al. Molecular subgroups of medulloblastoma: the current consensus. Acta Neuropathol. 123, 465–472 (2012).

  3. 3.

    Northcott, P. A. et al. The whole-genome landscape of medulloblastoma subtypes. Nature 547, 311–317 (2017).

  4. 4.

    Lin, C. Y. et al. Active medulloblastoma enhancers reveal subgroup-specific cellular origins. Nature 530, 57–62 (2016).

  5. 5.

    Hovestadt, V. et al. Decoding the regulatory landscape of medulloblastoma using DNA methylation sequencing. Nature 510, 537–541 (2014).

  6. 6.

    Northcott, P. A. et al. Subgroup-specific structural variation across 1,000 medulloblastoma genomes. Nature 488, 49–56 (2012).

  7. 7.

    Northcott, P. A. et al. Medulloblastomics: the end of the beginning. Nat. Rev. Cancer 12, 818–834 (2012).

  8. 8.

    Gibson, P. et al. Subtypes of medulloblastoma have distinct developmental origins. Nature 468, 1095–1099 (2010).

  9. 9.

    Yang, Z. J. et al. Medulloblastoma can be initiated by deletion of Patched in lineage-restricted progenitors or stem cells. Cancer Cell 14, 135–145 (2008).

  10. 10.

    Oliver, T. G. et al. Loss of patched and disruption of granule cell development in a pre-neoplastic stage of medulloblastoma. Development 132, 2425–2439 (2005).

  11. 11.

    Cho, Y. J. et al. Integrative genomic analysis of medulloblastoma identifies a molecular subgroup that drives poor clinical outcome. J. Clin. Oncol. 29, 1424–1430 (2011).

  12. 12.

    Tanay, A. & Regev, A. Scaling single-cell genomics from phenomenology to mechanism. Nature 541, 331–338 (2017).

  13. 13.

    Tirosh, I. et al. Single-cell RNA-seq supports a developmental hierarchy in human oligodendroglioma. Nature 539, 309–313 (2016).

  14. 14.

    Filbin, M. G. et al. Developmental and oncogenic programs in H3K27M gliomas dissected by single-cell RNA-seq. Science 360, 331–335 (2018).

  15. 15.

    Patel, A. P. et al. Single-cell RNA-seq highlights intratumoral heterogeneity in primary glioblastoma. Science 344, 1396–1401 (2014).

  16. 16.

    Venteicher, A. S. et al. Decoupling genetics, lineages, and microenvironment in IDH-mutant gliomas by single-cell RNA-seq. Science 355, eaai8478 (2017).

  17. 17.

    Hovestadt, V. et al. Robust molecular subgrouping and copy-number profiling of medulloblastoma from small amounts of archival tumour material using high-density DNA methylation arrays. Acta Neuropathol. 125, 913–916 (2013).

  18. 18.

    Picelli, S. et al. Full-length RNA-seq from single cells using Smart-seq2. Nat. Protocols 9, 171–181 (2014).

  19. 19.

    Northcott, P. A. et al. Rapid, reliable, and reproducible molecular sub-grouping of clinical medulloblastoma samples. Acta Neuropathol. 123, 615–626 (2012).

  20. 20.

    Carter, S. L. et al. Absolute quantification of somatic DNA alterations in human cancer. Nat. Biotechnol. 30, 413–421 (2012).

  21. 21.

    Carter, R. A. et al. A single-cell transcriptional atlas of the developing murine cerebellum. Curr. Biol. 28, 2910–2920 (2018).

  22. 22.

    Kool, M. et al. Molecular subgroups of medulloblastoma: an international meta-analysis of transcriptome, genetic aberrations, and clinical data of WNT, SHH, Group 3, and Group 4 medulloblastomas. Acta Neuropathol. 123, 473–484 (2012).

  23. 23.

    Waszak, S. M. et al. Spectrum and prevalence of genetic predisposition in medulloblastoma: a retrospective genetic study and prospective validation in a clinical trial cohort. Lancet Oncol. 19, 785–798 (2018).

  24. 24.

    Shih, D. J. et al. Cytogenetic prognostication within medulloblastoma subgroups. J. Clin. Oncol. 32, 886–896 (2014).

  25. 25.

    Northcott, P. A. et al. Pediatric and adult sonic hedgehog medulloblastomas are clinically and molecularly distinct. Acta Neuropathol. 122, 231–240 (2011).

  26. 26.

    Kool, M. et al. Genome sequencing of SHH medulloblastoma predicts genotype-related response to smoothened inhibition. Cancer Cell 25, 393–405 (2014).

  27. 27.

    Merk, D. J. et al. Opposing effects of CREBBP mutations govern the phenotype of Rubinstein–Taybi syndrome and adult SHH medulloblastoma. Dev. Cell 44, 709–724 (2018).

  28. 28.

    Northcott, P. A. et al. Medulloblastoma comprises four distinct molecular variants. J. Clin. Oncol. 29, 1408–1414 (2011).

  29. 29.

    Jones, D. T. et al. Dissecting the genomic complexity underlying medulloblastoma. Nature 488, 100–105 (2012).

  30. 30.

    Cavalli, F. M. G. et al. Intertumoral heterogeneity within medulloblastoma subgroups Cancer Cell 31, 737–754 (2017).

  31. 31.

    Schwalbe, E. C. et al. Novel molecular subgroups for clinical classification and outcome prediction in childhood medulloblastoma: a cohort study. Lancet Oncol. 18, 958–971 (2017).

  32. 32.

    Sharma, T. et al. Second-generation molecular subgrouping of medulloblastoma: an international meta-analysis of Group 3 and Group 4 subtypes. Acta Neuropathol. https://doi.org/10.1007/s00401-019-02020-0 (2019).

  33. 33.

    Chizhikov, V. V. et al. Lmx1a regulates fates and location of cells originating from the cerebellar rhombic lip and telencephalic cortical hem. Proc. Natl Acad. Sci. USA 107, 10725–10730 (2010).

  34. 34.

    Englund, C. et al. Unipolar brush cells of the cerebellum are produced in the rhombic lip and migrate through developing white matter. J. Neurosci. 26, 9184–9195 (2006).

  35. 35.

    Capper, D. et al. DNA methylation-based classification of central nervous system tumours. Nature 555, 469–474 (2018).

  36. 36.

    Rusch, M. et al. Clinical cancer genomic profiling by three-platform sequencing of whole genome, whole exome and transcriptome. Nat. Commun. 9, 3962 (2018).

  37. 37.

    Li, B. & Dewey, C. N. RSEM: accurate transcript quantification from RNA-seq data with or without a reference genome BMC Bioinformatics 12, 323 (2011).

  38. 38.

    Gaujoux, R. & Seoighe, C. A flexible R package for nonnegative matrix factorization. BMC Bioinformatics 11, 367 (2010).

  39. 39.

    Wolf, F. A., Angerer, P. & Theis, F. J. SCANPY: large-scale single-cell gene expression data analysis. Genome Biol. 19, 15 (2018).

  40. 40.

    Butler, A., Hoffman, P., Smibert, P., Papalexi, E. & Satija, R. Integrating single-cell transcriptomic data across different conditions, technologies, and species. Nat. Biotechnol. 36, 411–420 (2018).

  41. 41.

    Pöschl, J. et al. Genomic and transcriptomic analyses match medulloblastoma mouse models to their human counterparts. Acta Neuropathol. 128, 123–136 (2014).

  42. 42.

    Tamayo, P. et al. Metagene projection for cross-platform, cross-species characterization of global transcriptional states. Proc. Natl Acad. Sci. USA 104, 5959–5964 (2007).

Download references

Acknowledgements

P.A.N. is a Pew–Stewart Scholar for Cancer Research (Margaret and Alexander Stewart Trust) and recipient of The Sontag Foundation Distinguished Scientist Award. P.A.N. was also supported by the National Cancer Institute (R01CA232143-01), American Association for Cancer Research (NextGen Grant for Transformative Cancer Research), The Brain Tumour Charity (Quest for Cures), the American Lebanese Syrian Associated Charities (ALSAC), and St Jude. M. L. Suvà was supported by grants from the Howard Goodman Fellowship at MGH, the Merkin Institute Fellowship at the Broad Institute of MIT and Harvard, the Wang Family Fund, the V Foundation for Cancer Research, the Swiss National Science Foundation Sinergia program, and the Alex’s Lemonade Stand Foundation. M. L. Suvà is also recipient of The Sontag Foundation Distinguished Scientist Award. B.E.B. is the Bernard and Mildred Kayden Endowed MGH Research Institute Chair and an American Cancer Society Research Professor. This research was supported by a Pioneer Award from the NIH Common Fund and National Cancer Institute (DP1CA216873). V.H. is supported by a Human Frontier Science Program long-term fellowship (LT000596/2016-L). L.B. is supported by a Future Leaders Award from The Brain Tumour Charity (GN-000518). M.G.F. was supported by a Career Award for Medical Scientist from Burroughs Wellcome Fund, a K12 Paul Calabresi Career Award for Clinical Oncology (K12CA090354), a Harvard Brain Cancer SPORE—Career Enhancement Program Award, the National Institutes of Health (3P30 CA006516-53S6), The Cure Starts Now Foundation, Solving Kids’ Cancer/The Bibi Fund, The Andruzzi Foundation and Alex’s Lemonade Stand Foundation. I.S., D.K. and D.L. were supported by the Austrian National Bank (OeNB Jubiläumsfonds Project 15173). M.N.R. is supported by the ALSF, PBTF, AKBTC and CBJOLF. We are indebted to the Flow Cytometry Core Laboratory (Department of Developmental Neurobiology, St Jude) and the Core Flow Cytometry and Cell Sorting Shared Resource Facility (St Jude). From St Jude, we explicitly acknowledge the Hartwell Center, the Biorepository, members of the Clinical Genomics team, the Diagnostic Biomarkers Shared Resource in the Department of Pathology, and the Center for In Vivo Imaging and Therapeutics. We thank S. Pounds (Department of Biostatistics, St Jude) for valuable discussions and B. Stelter for assistance with artwork.

Author information

Study design: V.H., K.S.S., L.B., M.G.F., B.E.B., M. L. Suvà and P.A.N. Generation of human transcriptome data: L.B., M.G.F., M. L. Shaw, A.B., J.C.D., A. Groves, L.M., H.R.W., A.R.R, M.E.S., J. H., R.A.A., J.G., D.K., D.L., R.G. and A.H. Generation of mouse transcriptome data: L.B., C.R., T.N.P., J.L.H., Y.T. and J.E. Analysis of human transcriptome data: V.H. and K.S.S. Analysis of mouse transcriptome data: V.H., K.S.S., R.A.C. and C.G. Generation and analysis of genome data: V.H., K.S.S., L.B., T.S., D.F., A.S., S.M.P, A. Gajjar and G.W.R. Immunohistochemistry experiments: L.B. and B.A.O. RNA in situ hybridization: H.R.W. and M.E.S. Procurement of patient and PDX samples: L.B., M.G.F., L.G., J.L.H., M.D., K.L.L., J.M.R., R.J.W.-R., X.-N.L., A.P., T.C., C.D., C.H., A. Gajjar, B.A.O., I.S. and G.W.R. Project support: S.L.P., M.N.R., O.R.-R. and A. R. Manuscript preparation (with feedback from all authors): V.H., K.S.S., L.B., M.G.F., B.E.B., M. L. Suvà and P.A.N. Study supervision and funding: M.G.F., B.E.B., M. L. Suvà and P.A.N.

Correspondence to Bradley E. Bernstein or Mario L. Suvà or Paul A. Northcott.

Ethics declarations

Competing interests

B.E.B. discloses financial interests in Fulcrum Therapeutics, 1CellBio, HiFiBio, Arsenal Biosciences, Cell Signaling Technologies and Nohla Therapeutics. A.R. is a founder and equity holder of Celsius Therapeutics and an SAB member of ThermoFisher Scientific and Syros Pharmaceuticals.

Additional information

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Peer review information Nature thanks Xing Fan and the other anonymous reviewer(s) for their contribution to the peer review of this work.

Extended data figures and tables

Extended Data Fig. 1 Characteristics of the MB single-cell cohort.

a, Haematoxylin and eosin-stained sections from all St Jude single-cell samples (n = 12). Tumours demonstrated large cell/anaplastic morphology (LCA, top), classic morphology (middle), or desmoplastic/nodular morphology (D/N, bottom). Scale bars, 50 µm. b, Detailed characterization of the PDX single-cell dataset. Subgroup prediction scores35 derived by DNA methylation profiling are indicated in the top panel (light shade, low probability; dark shade, high probability). The heat map shows expression levels of previously described subgroup-specific marker genes19 in 946 PDX-derived single-cells. c, Heat map shows expression levels of previously described subgroup-specific marker genes19 in 7,788 tumour-derived single-cells. d, Heat maps show pairwise correlation of aggregated scRNA-seq data (top) and bulk DNA methylation data (bottom) of all patient (n = 25) and PDX (n = 11) samples. For each PDX sample, the patient sample with the highest correlation coefficient is indicated by a black circle. e, Scatter plots show expression scores for published subgroup-specific gene sets for all single cells in the patient cohort (n = 7,788). Cells from WNT and SHH subgroups score only for their respective gene set. Some overlap is observed between cells from Group 3 and 4 subgroups and their respective gene sets, warranting the combined analysis of these subgroups in this study.

Extended Data Fig. 2 Copy-number analysis distinguishes malignant from non-malignant single cells.

ae, Heat maps show scRNA-seq-derived copy-number profiles of every cell in each sample (y axis) along the genome (x axis) for WNT (a), SHH (b), Group 3 (c) and Group 4 (d) patient MBs as well as PDX samples (e). Copy-number profiles derived from array-based DNA methylation profiling from the same sample are shown above. CNVs are observed in 21/25 patient tumour samples (all except MUV34, MUV41, SJ577 and SJ625). Generally, we observe a high concordance between single-cell and DNA methylation array-derived copy-number profiles. Genetic subclones at the level of broad copy-number changes are detected in samples SJ99 and BCH825. Cells without detected CNVs from samples that showed CNVs in the majority of cells are indicated for samples in which at least four non-malignant cells were detected (BCH807 and SJ454). Amplifications of the MYC and MYCN oncogenes detected by DNA methylation array are indicated.

Extended Data Fig. 3 Unsupervised clustering and detection of expressed SNVs in MB single-cells.

a, t-SNE visualization of the entire single-cell dataset (n = 8,924 cells). WNT (blue), SHH (red), Group 3 (yellow) and Group 4 (green) patient samples are indicated. PDX models are shown in pink. Non-neoplastic oligodendrocytes and immune cells are included for comparison. Generally malignant cells are expected to cluster by patient sample, whereas non-malignant cells are expected to cluster by cell type. Only few cells from different samples cluster with oligodendrocytes (n = 22) or immune cells (n = 6) and were classified as non-malignant. No additional clusters of cells from different samples were identified, indicating the absence of additional non-malignant cell populations in our dataset. b, Identical t-SNE visualization as in a, coloured by copy-number state. CNVs were detected in most single cells, facilitating their classification as malignant. A small number of cells did not show CNVs, even though CNVs were detected in the majority of cells from the respective sample (n = 38). These cells were classified as non-malignant. Most cells with without CNVs clustered with normal oligodendrocytes (n = 21), supporting their initial classification as non-malignant. Remaining cells without CNVs did not form clusters and likely represent poor-quality cells. c, Identical t-SNE visualization as in a, coloured by detected mutant and wild-type transcripts. Cells classified as non-malignant are depleted for mutant transcripts (P < 0.01, binomial test), supporting their initial classification. d, Heat map shows detected mutant and wild-type transcripts for 39 variants (columns) in each cell (n = 1,780, rows) of the WNT MB dataset. If both mutant and wild-type transcripts are detected in a single cell, only the mutant transcript is shown. Variants were initially detected by genome sequencing and subsequently quantified in the scRNA-seq data. Sample BCH807 was not subjected to genome sequencing, and the CTNNB1 variant was manually detected by examining scRNA-seq alignments. Mutations are detected almost exclusively in single cells from samples in which they were detected by genome sequencing, illustrating the high specificity of single-cell variant detection. e, Heat map shows mutant and wild-type transcripts for 15 variants in each cell (n = 1,135, rows) of the SHH MB dataset. Sample SJ454 was not subjected to genome sequencing, and the TP53 mutation was manually identified by examining scRNA-seq alignments. f, Heat map shows mutant and wild-type transcripts for 28 variants in each cell (n = 3,172, rows) of the Group 3/4 MB samples that were subjected to genome sequencing.

Extended Data Fig. 4 Single-cell mapping of mouse cerebellar development.

ac, Two-dimensional representation of the cerebellar (CB) scRNA-seq dataset by t-SNE. Each dot represents one cell. In a, colours represent 13 different embryonic and early postnatal time points. In b, colours indicate the differentiation score across the entire dataset. In c, colours indicate cell types identified by Louvain clustering using the top 3,000 overdispersed genes. The main CB lineages were assigned on the basis of published lineage markers. d, Annotation of 18 CB cell types based on the expression of lineage specific marker genes shown as violin plot. Violin plots represent kernel density estimation showing the distribution shape of the data. e, Lineage tree reconstruction using partition-based graph abstraction. The abstracted graph shows all cell types (nodes) as identified in c and d. The size of the nodes is related to the number of cells in the defined cell type. The width of edges connecting cell types reflects the probability of the path. f, Radar plot showing CCA coefficients between each mouse CB cell type and human MB subgroup scRNA-seq.

Extended Data Fig. 5 Characterization of WNT MB single-cell programs.

a, Expression scores for individual programs identified by unsupervised NMF analysis in each sample. Cells are ordered as in Fig. 2a (n = 1,780). Metaprograms WNT-A, WNT-B, WNT-C, and WNT-D were identified by hierarchical clustering of individual programs. b, Heat maps show pairwise correlation (left), principal component analysis (PCA, centre), and expression scores for NMF-derived metaprograms (right) for 301 cells from WNT MB sample MUV44. The ordering of cells (rows) is maintained between the heat maps. A two-dimensional representation of the same cells using t-SNE is shown on the far right (coloured by expression scores for each metaprogram). This analysis shows that the same programs and cell populations that are identified by the NMF analysis are also supported by PCA and t-SNE clustering. Furthermore, no additional programs and cell populations are identified (starting from PC5, components are less informative). c, Scatter plot shows isometric projection of average gene expression levels for cells with highest expression score for WNT-B (undifferentiated, proliferating), WNT-C (neuron-like), or WNT-D (undifferentiated, post-mitotic). WNT-B metaprogram genes are indicated in red, WNT-C metaprogram genes are indicated in green, and WNT-D metaprogram genes are indicated in blue. Genes that are higher in both undifferentiated cell populations compared to neuron-like cells are indicated in black. d, Images show RNA in situ hybridization experiments of five marker genes representative for the four WNT MB metaprograms in two samples of the single-cell cohort. Results confirm expression of these genes independently of the scRNA-seq experiments.

Extended Data Fig. 6 Characterization of SHH MB single-cell programs.

a, Expression scores for individual programs identified by unsupervised NMF analysis in each sample. Cells are ordered as in Fig. 3a (n = 1,135). Metaprograms SHH-A, SHH-B, and SHH-C were identified by hierarchical clustering of individual programs. b, Heat maps show pairwise correlation (left), PCA (centre), and expression scores for NMF-derived metaprograms (right) for 493 cells from SHH MB sample SJ577. The ordering of cells (rows) is maintained between the heat maps. A two-dimensional representation of the same cells using t-SNE is shown on the far right (coloured by expression scores for each metaprogram). This analysis shows that the same programs and cell populations that are identified by the NMF analysis are also supported by PCA and t-SNE clustering. Furthermore, no additional programs and cell populations are identified (starting from PC3, components are less informative). c, Pairwise correlations between the expression profiles of 303 single-cells (rows, columns) from two SHH PDX samples (RCMB18 and RCMB24) (left). Expression scores for each of the NMF-derived metaprograms SHH-A, SHH-B, and SHH-C (columns) (right). Cells are ordered as in the left panel (rows). d, Heat maps show the relative expression of the 60 genes representing the metaprograms SHH-B and SHH-C (rows), across 303 cells for RCMB18 and RCMB24. Cells are sorted by the difference between the two scores. Cells positive for the cell cycle program (SHH-A) are indicated by red bars. Similar cell populations as in the primary samples (undifferentiated GNP-like and differentiated neuron-like cells) are identified in RCMB18. No differentiated cells are identified in RCMB24.

Extended Data Fig. 7 Cross-species mapping of SHH MB origins.

a, Heat map shows average expression levels of 29 GNP-associated genes (rows) in cell types identified in the mouse CB dataset (columns). Genes are ordered by their relative expression in GNPs. b, Left, the relative expression of orthologous genes in a in all cells from the single-cell cohort (n = 7,745; columns). Cells are ordered by increasing GNP CCA cosine correlation coefficients. Cells expressing high levels of GNP-associated genes are predominantly from SHH tumours. Right, the relative expression of the same genes in the bulk microarray cohort (n = 392). c, d, Heat maps as in a and b, but showing 30 genes associated with the UBC/GN intermediate cell type. e, Two-dimensional representation of GNPs/granule neurons from the cerebellar atlas by t-SNE. Each dot represents one cell (n = 35,013). Colours represent the assigned cerebellar cell types (left), as well as the expression of Atoh1 and Neurod1 (middle and right). f, Box plots of select granule lineage marker genes in the mouse CB cohort (left), MB single-cell cohort (middle) and MB bulk microarray cohort (right). g, Box plot of patient age associated with infant and adult/child subtypes of SHH MB. h, Box plot of the number of coding mutations associated with SHH MB subtypes. The median is shown as a thick line; box limits are 25th and 75th percentiles; whiskers denote 1.5 times the interquartile range. i, Expression of Barhl1 (left) and Pde1c (right) at P4 during CB development. In situ hybridization data were obtained from the Allen Developing Mouse Brain Atlas (© 2008 Allen Institute for Brain Science. Allen Developing Mouse Brain Atlas http://developingmouse.brain-map.org). j, Radar plot showing the CCA cosine correlation coefficients between each mouse CB cell type and the MB single-cell cohort from cells scoring highest for metaprograms SHH-B (GNP-like cells) and SHH-C (granule neuron-like cells).

Extended Data Fig. 8 Characterization of Group 3/4 MB single-cell programs.

a, Top, Group 3/4 subtype prediction scores derived by DNA methylation profiling32 (light shade, low probability; dark shade, high probability). Expression scores for individual programs identified by unsupervised NMF analysis in each sample are indicated in the bottom. Cells are ordered as in Fig. 4a (n = 4,873). Metaprograms Group 3/4-A, Group 3/4-B, and Group 3/4-D were identified by hierarchical clustering of individual programs. b, Expression scores across 4,873 single cells (columns) for each of the NMF-derived metaprograms Group 3/4-A, Group 3/4-B, and Group 3/4-C (rows). Cells are ordered as in a. c, Heat maps show pairwise correlation (left), PCA (centre) and expression scores for NMF-derived metaprograms (right) for 400 cells from Group 3 MB sample SJ617. The ordering of cells (rows) is maintained between the heat maps. A two-dimensional representation of the same cells using t-SNE is shown on the far right (coloured by expression scores for each metaprogram). This analysis shows that the same programs and cell populations that are identified by the NMF analysis are also supported by PCA and t-SNE clustering. Furthermore, no additional programs and cell populations are identified (starting from PC4 components are less informative). d, Pairwise correlations between the expression profiles of 643 single cells (rows, columns) from nine patient-derived xenograft models (Med114FH, Med2112FH, Med211FH, Med411FH, RCMB20, Icb1299, Icb1572, Med2312FH and DMB006). Left, Group 3/4 subtype prediction scores derived by DNA methylation profiling. Right, expression score for the NMF-derived metaprograms Group 3/4-A, Group 3/4-B and Group 3/4-C (columns). e, Heat maps show the relative expression of the 60 genes representing the metaprograms Group 3/4-B and Group 3/4-C (rows) across 140 cells for RCMB20 and DMB006. Cells are sorted by the difference between the two scores. Cells positive for the cell cycle program (Group 3/4-A) are indicated by red bars. Group 3 PDX samples are predominantly undifferentiated, with the exception of Med2312FH, which is predominantly differentiated (classified by DNA methylation array as an intermediate Group 3/4 sample). This parallels the high frequency of MYC amplifications in our Group 3 PDX cohort (5/8). Group 4 PDX sample DMB006 is also predominantly differentiated. These results are supportive of the cellular compositions detected in primary Group 3/4 samples.

Extended Data Fig. 9 Analysis of Group 3/4 intermediate samples and pan-subgroup comparison.

a, Scatter plot of the metaprogram Group 3/4-C (x axis) and Group 3/4-B (y axis) expression scores for Group 3 and Group 4 bulk MBs3 (yellow and green dots, respectively; n = 248). Samples that score similarly for both programs are classified as intermediate samples (n = 49) b, Representative MYC and TUJ1 immunohistochemistry images of seven Group 3/4 samples. Four of these samples are shown at higher magnification in Fig. 5b (SJ17, SJ617, SJ625, SJ723). c, Two-dimensional representation of 740 Group 3/4 MB samples analysed by DNA methylation profiling using t-SNE3. Eight subtypes are delineated by curved lines. Samples are coloured by their predicted subgroup35. d, Heat map showing expression of transcripts coding for ribosomal proteins (n = 75, rows). Cells positive for the cell cycle programs, and cells classified as neuron-like cells are indicated on top. Cells are ordered as in Fig. 6b (n = 7,745). e, Heat map showing relative expression levels of genes that are specific to neuron-like cells and are shared between multiple subgroups (n = 134, rows). Cells are ordered as in d. f, Heat map shows the relative expression of UBC-specific genes in Fig. 6d (n = 30; rows) in the bulk expression array cohort (n = 392; columns). Samples are ordered by increasing CCA cosine correlation coefficient.

Extended Data Fig. 10 Cross-species mapping of Group 4 MB origins.

a, Top, expression of TBR1 and EOMES in bulk Group 4 MB expression array data (n = 149). Middle, Group 3/4 DNA methylation-based subtype annotations for each sample. Bottom, CCA scores from comparison of bulk MB expression data and UBCs and GluCN late populations from the cerebellar single-cell dataset. b, t-SNE visualization shows clustering of glutamatergic populations correlated with Group 4 MBs. c, Box plot of CCA cosine correlation coefficients from comparison of bulk MB expression data and UBCs, according to Group 3/4 subtypes. The median is shown as a thick line; box limits are 25th and 75th percentiles; whiskers denote 1.5 times the interquartile range. d, e, Left, in situ hybridization data for Tbr1 (d) and Eomes (e) in the developing mouse cerebellum at the indicated time point. Data were obtained from the Allen Developing Mouse Brain Atlas (© 2008 Allen Institute for Brain Science. Allen Developing Mouse Brain Atlas http://developingmouse.brain-map.org). Right, expression of Tbr1 (d) and Eomes (e) in the mouse single-cell dataset according to the t-SNE structure shown in b. f, Radar plot showing CCA cosine correlation coefficients between each mouse CB cell type and Group 3 MB (top) or Group 4 MB (bottom) cells scoring highest for metaprograms Group 3/4-B or Group 3/4-C. g, Graphical summary of subgroup-specific cellular hierarchies identified in MB.

Supplementary information

Reporting Summary

Supplementary Table 1

Cohort details.

Supplementary Table 2

Transcriptional programs.

Supplementary Table 3

Comparison of neuronal-like cells between MB subgroups.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Hovestadt, V., Smith, K.S., Bihannic, L. et al. Resolving medulloblastoma cellular architecture by single-cell genomics. Nature 572, 74–79 (2019) doi:10.1038/s41586-019-1434-6

Download citation

Further reading

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.