Long term culture of mesenchymal stem cells in hypoxia promotes a genetic program maintaining their undifferentiated and multipotent status

Abstract

Background: In the bone marrow, hematopietic and mesenchymal stem cells form a unique niche in which the oxygen tension is low. Hypoxia may have a role in maintaining stem cell fate, self renewal and multipotency. However, whereas most studies addressed the effect of transient in vitro exposure of MSC to hypoxia, permanent culture under hypoxia should reflect the better physiological conditions.

Results: Morphologic studies, differentiation and transcriptional profiling experiments were performed on MSC cultured in normoxia (21% O2) versus hypoxia (5% O2) for up to passage 2. Cells at passage 0 and at passage 2 were compared, and those at passage 0 in hypoxia generated fewer and smaller colonies than in normoxia. In parallel, MSC displayed (>4 fold) inhibition of genes involved in DNA metabolism, cell cycle progression and chromosome cohesion whereas transcripts involved in adhesion and metabolism (CD93, ESAM, VWF, PLVAP, ANGPT2, LEP, TCF1) were stimulated. Compared to normoxic cells, hypoxic cells were morphologically undifferentiated and contained less mitochondrias. After this lag phase, cells at passage 2 in hypoxia outgrew the cells cultured in normoxia and displayed an enhanced expression of genes (4-60 fold) involved in extracellular matrix assembly (SMOC2), neural and muscle development (NOG, GPR56, SNTG2, LAMA) and epithelial development (DMKN). This group described herein for the first time was assigned by the Gene Ontology program to "plasticity".

Conclusion: The duration of hypoxemia is a critical parameter in the differentiation capacity of MSC. Even in growth promoting conditions, hypoxia enhanced a genetic program that maintained the cells undifferentiated and multipotent. This condition may better reflect the in vivo gene signature of MSC, with potential implications in regenerative medicine.

Figure 1

The effect of oxygen tension on CFU-F size and numbers. Cells were plated at 1000 (plain histograms) and 10000 MNC/60 cm² (dotted histograms) from total BM for P0. For the next passages, 100 (plain histograms) and 1000 cells/cm² (dotted histograms) were seeded. Cells were incubated under hypoxic (5% O₂, black histograms) or normoxic conditions (21% O₂, white histograms) respectively and colonies were counted and their size evaluated (A & B). Mean +/- of 3 to 5 independent experiments. Numbers above the histograms (A & B) represent the significance calculated using bilateral paired Student's t test. A representative aspect of colonies at various passages is shown (C).

Figure 2

Stability of MSC phenotype in culture. Flow cytometry analysis of surface markers at P0 and P2 in hypoxia and normoxia. Histograms show the intensity of several markers within the CD34-CD45- (dot plot), MSC-enriched population from the bone marrow. A representative experiment out of 3 from distinct donors is shown.

Figure 3

Morphological aspect of MSC. The amount of mitochondria was evaluated by flow cytometry (A) and optical microscopy (B) on MSC stained with Mitotracker Orange. TEM was performed on MSC at P2 culture in normoxic or hypoxic conditions (C). (N = 3)

Figure 4

Culture in hypoxia enhanced the potential of MSC to differentiate in adipocytes (A) and osteocytes (O). MSC were cultured in hypoxia or normoxia until P2, and shifted to adipocyte- or osteocyte-specific differentiation conditions respectively for 2-4 more weeks. The aspect of differentiated colonies under hypoxia is shown (A). In 2 independent experiments (B), MSC were harvested at P2, RNA from normoxic (N) or hypoxic (H) cells was extracted, reverse-transcribed and amplified by PCR with primers specific for control GAPDH and the indicated genes, representative of adipocytic (LPL, PPARγ) and osteocytic (ALPL, Runx2) lineages. The same cells were grown in osteocytic and adipocytic-specific medias and colonies were counted and compared to the numbers of CFU-F generated in MSC-specific medium (C).

Figure 5

Genetic reprogramming of MSC under hypoxia. MSC from 3 donors were cultured until P0 and P2. RNA was extracted from each culture, processed and hybridized on Agilent DNA microarrays. Diagram A shows the GO analysis of differentially expressed transcripts at P0. Six transcripts were chosen and amplified by qPCR in order to validate gene array results obtained after long term hypoxia using the same extracts. Mean +/- SD from 3 experiments.