Background: Cancer stem cells are believed to be a major reason for long-term therapy failure because they are multi-drug resistant and able to rest mitotically inactive in the hypoxic center of tumors. Due to their variable number and their often low proliferation rate, cancer stem cells are difficult to purify in decent quantities and to grow in cell culture systems, where they are easily outcompeted by faster growing more 'differentiated', i.e., less stem cell-like tumor cells.
Methods: Here we present a proof of principle study based on the idea to select cancer stem cells by means of the expression of a stem cell-specific gene. A selectable egfp-neo coding sequence was inserted in the last exon of the non-coding murine miR-302 host gene. As a stem cell specific regulatory element, 2.1 kb of the genomic region immediately upstream of the miR-302 host gene transcription start site was used. Stable transgenic CJ7 embryonic stem cells were used to induce teratomas.
Results: After three weeks, tumors were removed for analysis and primary cultures were established. Stem cell-like cells were selected from these culture based on G418 selection. When the selection was removed, stem cell morphology and miR-302 expression were rapidly lost, indicating that it was not the original ES cells that had been isolated.
Conclusions: We show the possibility to use drug resistance expressed from a regulatory sequence of a stem cell-specific marker, to isolate and propagate cancer stem cells that otherwise might be hidden in the majority of tumor cells.
Keywords: Cancer stem cells; G418 selection; MiR-302/367; Primary cell culture; Stem cell marker; Teratoma.