Molecular karyotyping and genetic stability of human embryonic stem cells

Introduction
Human embryonic stem cells (hESCs) are a renewable source of different cell types that could be used for research and cell therapy, and are therefore often kept in culture for long periods of time. It is important to guarantee the general stability of these cells, to ensure the reproducibility and reliability of the results obtained by using them. Recently, several research groups have reported the appearance of culture-induced genetic abnormalities, such as aberrant karyotypes. Although performing a karyotype is part of the routine tests carried out on a putative hESC line during its initial characterization, these findings indicate that the genetic constitution of hESC lines must be regularly followed up. G-banding is the standard method for karyotyping the hESCs, but this was proven difficult and inefficient. We therefore decided to evaluate the use of array-based comparative genomic hybridization to regularly karyotype our hESC lines.

Materials and methods
Thirteen hESC lines were used in this study, all of them derived in our laboratory using human preimplantation embryos donated for research. Genomic DNA was phenol/chloroform extracted from the cell cultures and used to perform CGH on 1 MB BAC/PAC arrays.

Results
The following results were obtained (p stands for passage number, b for number of passages after a freeze-thaw procedure). VUB01 carried a monosomy 18 (45, XY, -18) at p55b14 and was 46, XY, +20q11.21 at p135 and p276. This region was amplified as shown by FISH, and contains the DNMT3B gene. VUB02 had a normal karyotype, 46, XY, at p21b15, p109b77 and p219b187. VUB04 was 46, XX, +5q14.2-ter, -18q21.2-ter at p117. A G-banding showed a derivative chromosome 18. VUB05_HD had a normal 46, XY karyotype at p51b21, p58b28, p65b35 and p75b43. VUB07 was 46, XX, +20q11.21 at p77b7, and 46, XX, +3q26.33-q27.2, +20q11.21 at p122b52. These regions contain the genes DNMT3B and SOX2. VUB08_MFS carried a trisomy 22 at p49b11 (47, XX, +22). VUB11_FXS had a normal female karyotype at p45. VUB13_FXS had a normal female karyotype at p28 and p30, but mutated to 46, XX, +5q21.3-ter, -18q12.1-ter at p43, and finally to 45, X, +5q21.3-ter, -18q12.1-ter at p45. G-banding showed a derivative chromosome 18. VUB15, VUB17 and VUB19_DM1 showed normal karyotypes, once 46, XY, twice 46, XX, and at p18, p7 and p5, respectively. VUB20_CMT1A showed a duplication in 5q13.2 at p7 and p18 (46, XX, +5q13.2), that proved to be a polymorphism inherited from the female donor. VUB26 showed a derivative chromosome 18 at p10 (46, XX, +7q33-ter, -18q23-ter), which was visualized by G-banding.

Conclusions
Molecular karyotyping has proven cost-effective and very accurate. Not only were aneuploidies detected, but small aberrations were also detected, such as the amplifications of the stemness genes DNMT3B and SOX2, which would have passed undetected by G-banding. These amplifications are likely to provide the cells with a better adaptation to the in vitro culture system, and have already proven to influence the results of differentiation experiments in our laboratory. The recurrent appearance of the derivative chromosome 18 is striking, and further experiments are planned to study its biological significance.