Mosaic chromosomal alterations in peripheral blood leukocytes of children in sub-Saharan Africa

In high-income countries, mosaic chromosomal alterations in peripheral blood leukocytes are associated with an elevated risk of adverse health outcomes, including hematologic malignancies. We investigate mosaic chromosomal alterations in sub-Saharan Africa among 931 children with Burkitt lymphoma, an aggressive lymphoma commonly characterized by immunoglobulin-MYC chromosomal rearrangements, 3822 Burkitt lymphoma-free children, and 674 cancer-free men from Ghana. We find autosomal and X chromosome mosaic chromosomal alterations in 3.4% and 1.7% of Burkitt lymphoma-free children, and 8.4% and 3.7% of children with Burkitt lymphoma (P-values = 5.7×10−11 and 3.74×10−2, respectively). Autosomal mosaic chromosomal alterations are detected in 14.0% of Ghanaian men and increase with age. Mosaic chromosomal alterations in Burkitt lymphoma cases include gains on chromosomes 1q and 8, the latter spanning MYC, while mosaic chromosomal alterations in Burkitt lymphoma-free children include copy-neutral loss of heterozygosity on chromosomes 10, 14, and 16. Our results highlight mosaic chromosomal alterations in sub-Saharan African populations as a promising area of research.

: Genes on chromosome X that are either mutated in BL or indirectly linked with BL via Epstein-Barr Virus (EBV) overlapped by mCAs and their association with BL case status .................... Table S4: List of oncogenes and significantly mutated genes on chromosome 1q in BL cases...............           *Data from the Burkitt Lymphoma Genome Sequencing Project (BLGSP) that generated whole-genome sequencing (WGS) data for paired tumor-normal samples; genotype array data from normal samples were used in the current study of mCAs.

Table S2: Cell line fingerprinting in paired tumor-lymhoblastoid cell lines (LCLs) from corresponding normal (WW1-BL/LCLs and BL-2/IARC-304 LCL) to confirm their authenticity and genetic relationship.
The identity and relationship of the paired tumor-normal cultured cells was confirmed by short tandem repeat (STR) fingerprinting of 10 microsatellite markers (GenePrint 10 System, Promega).Samples labeled BL-2 and IARC-304 are from the same patient.*The STR results showed uniparental disomy (UPD) for the marker D1S539 in the WW1 BL/LCLs pairs, which was corroborated by WGS data that revealed UPD on 16q in both the tumor and WW1-LCLs, i.e., the tumor sample and LCL lost the other parental allele other than the normal sample, indicating the origin of the tumor is independent of LCL clones.* Genes implicated in BL tumor studies (see text)

Cell line
The list of oncogenes is from the oncogene database (https://ongene.bioinfo-minzhao.org/index.html) and the significantly mutated BL genes are from Lopez et al [1][2][3][4] .The flow chart shows a 10-step standard quality control pipeline applied to curate genotype data prior to analysis.Standard steps include filtering samples with low completion/call rates (<0.95), with high contamination rates (>=0.10),de-duplication, sex discordant samples, and loci with low (<0.95) completion rates prior to analysis.The box plots show the distributions of the sample completion/call rates all subjects (N=4,753) and for those with mCAs detected (n=209), stratified by sex (all females, N= 2,185 and females with mCAs n=45 and all males, N= 2,568, males with mCAs n=117).The two whiskers for each box plot mark the minima and maxima completion/call rates, and the dots beyond two whiskers are outlier completion/call rates.Analysis was limited to samples with a completion/call rate >=0.95.'Source data are provided as a Source Data file.'

Figure S5 .
Figure S5.Frequency of mCAs by age groups in cancer-free African children from Uganda, Tanzania, and Kenya in the EMBLEM study, cancer-free adult men from Ghana, and cancer-free European ancestry individuals from the US Prostate, Lung, Colon, and Ovarian Cancer Study (PLCO)..............

Figures Figure S1 .AFigure S2 .Figure S3 .Figure S4 .Figure S5 .
Figures Figure S1.Type of autosomal mCAs (n=438) plotted by the proportion of abnormal cells (p) on the x-axis versus relative copy number estimated from log R ratio on the y-axis.Results in (A) are for 931 BL cases and in (B) 3822 controls in EMBLEM and Malawi; 'Source data are provided as a Source Data file.'.A

Figure S6 .
Figure S6.Copy number changes in the BL cell line WW1 and three matching clones from several passages of the matching paired normal LCLs.Copy number profiles obtained from DNA methylation arrays.(A) Results for BL cell line WW1.(B) Results for different LCL clones created from the normal peripheral blood sample matching WW1 BL cell line: WW1-LCL-B06, WW1-LCL-U14, WW1-LCL-U20.For both A and B, the green colored dots in the upper panel represent gains, while red colored dots on the same panel represent losses of genetic material.The middle panel shows the results of the cytoScans.The colors (red, green, yellow, and blue) of the cytoscans represent different chromosomes and the WGS results are shown in the lower panel as Log2 in a whole-genome view, excluding sex chromosomes.(C) Results of a CytoScan Array showing chromosome 3q copy number changes in the WW1 BL and the three WW1 matching LCLs (WW1-LCL-B06, WW1-LCL-U14, WW1-LCL-U20).Colors represent different cell lines: violet-WW1-BL cell line; green-WW1-LCL-U20, orange-WW1-LCLU14, and blue-WW1-LCL-B06).Vertical dashed lines indicate the positions of the 3q gains, which vary in the genomic position in the WW1-BL tumor cell line versus the two WW1 LCLs (U20 and U14) in which they were detected; 'Source data are provided as a Source Data file.'.

Figure S7 .
Figure S7.Quality control steps in mCAs analysis for quality control assessment.

Figure S8 .
Figure S8.Box plots displaying sample completion/call rates for all subjects in the analysis versus subjects with mCAs

Table S3 : Genes on chromosome X that are either mutated in BL or indirectly linked with BL via Epstein-Barr Virus (EBV) overlapped by mCAs and their association with BL case status.
*All statistical tests used were two-sided.