Early embryonic chromosome instability results in stable mosaic pattern in human tissues

Abstract

The discovery of copy number variations (CNV) in the human genome opened new perspectives on the study of the genetic causes of inherited disorders and the aetiology of common diseases. Here, a single-cell-level investigation of CNV in different human tissues led us to uncover the phenomenon of mitotically derived genomic mosaicism, which is stable in different cell types of one individual. The CNV mosaic ratios were different between the 10 individuals studied. However, they were stable in the T lymphocytes, immortalized B lymphoblastoid cells, and skin fibroblasts analyzed in each individual. Because these cell types have a common origin in the connective tissues, we suggest that mitotic changes in CNV regions may happen early during embryonic development and occur only once, after which the stable mosaic ratio is maintained throughout the differentiated tissues. This concept is further supported by a unique study of immortalized B lymphoblastoid cell lines obtained with 20 year difference from two subjects. We provide the first evidence of somatic mosaicism for CNV, with stable variation ratios in different cell types of one individual leading to the hypothesis of early embryonic chromosome instability resulting in stable mosaic pattern in human tissues. This concept has the potential to open new perspectives in personalized genetic diagnostics and can explain genetic phenomena like diminished penetrance in autosomal dominant diseases. We propose that further genomic studies should focus on the single-cell level, to better understand the aetiology of aging and diseases mediated by somatic mutations.

Figure 1. Single-cell estimation of CNV by using pod-FISH.

(A) The parental origin of der(7) in an individual with Turner syndrome and the mosaic karyotype 45,X,der(7)t(Y;7)(p11.1∼11.2;p22.3)[122]/45,X[48] was detected with RP11-533E18 (red, 7q11.21–7q11.1) and RP11-45N9 (blue, 7q34). (B) The homologous chromosomes 7 of the father and der(7) of the child are marked with asterisks. The suggested development of this rare mosaic karyotype is shown schematically. (C) Different signal intensities for the CNV at 19p13.2 (RP11-367L15, blue) are apparent on the two homologous chromosomes 19 in metaphase, and in the interphase nuclei of all cells of an AML patient with t(8;21)(q22;q22.3) before bone-marrow transplantation (left). In the bone-marrow donor cells, 55% of the cells showed no signal intensity difference in the CNV region detected with RP11-367L15 (blue) at 19p13.2 (right, upper nucleus). However, the remaining 45% of the recipient cells showed signal intensity differences identical to those of the recipient (right, lower nucleus). As well as RP11-367L15 (blue), probes for AML1 (21q22.3, green) and ETO (8q22, red) were also applied, showing two fusion signals (f) in the patient cells. (D) pod-FISH using six BACs specific for different CNV regions on chromosome 5 in a patient with normal chromosome 5 and der(5) revealed different signal intensity patterns for BAC RP11-551B22 (green) and RP11-88L18 (blue).

Figure 2. Varying signal intensity patterns of 5 selected CNV loci.

Analysis of the signal intensity patterns of CNV at 2p11.2 (RP11-685N3: 88,981,161–89,122,3701bp; RP11-15J7: 88,979,594–89,141,7171bp), 4q12 (RP11-365H22: 52,354,875–52,530,859 bp), and 14q11.2 (RP11-645B7: 18,654,379–18,833,779 bp; RP11-831B15: 19,273,689–19,767,232 bp) in three cell types, each, from 10 different individuals. Notes: (1) X axis: BAC probes; Y axis: [%]  =  percentage of cells with different signal intensities. (2) In diagrams with no columns, no cells with different signal intensities for the BAC probes were detected. Abbreviations: T cells ( =  metaphases from T lymphocytes collected after PHA stimulation of the peripheral blood), B cells ( =  Epstein–Barr-virus-immortalized B-lymphoblastoid cell lines from the peripheral blood), and fibroblasts ( =  metaphases from skin biopsies). * Statistically significant difference (B cells vs T cells and B cells vs fibroblasts, P<0.05). ** Statistically significant difference (T cells vs B cells and T cells vs fibroblasts, P<0.017).

Figure 3. Scheme for the development of somatic CNV mosaicism.

Parental germ cells with a certain haploid pattern of CNV fuse to a zygote with a defined diploid pattern of CNV. It is hypothesized that in early embryogenesis, the mitotic recombination of single CNV occurs, leading to a somatic mosaicism in the CNV pattern, which is stable until at least the end of gastrulation and the formation of the germ layers. The ratio of this CNV mosaicism is stable over a lifetime and across different cell types, except when other factors directly influence DNA stability.