Back to Annual Meeting Program
Human Keratinocytes have Two Distinct Cell Proliferation Patterns on Timelapse Imaging in Vitro with 8-Cell Microarrays Revealing Early Molecular Identifiers
Amit Roshan, MRCS(Eng)1, Benjamin D. Simons, PhD2, Kasumi Murai, PhD3, Philip H. Jones, PhD, FRCP3.
1Departments of Plastic Surgery and Oncology, University of Cambridge, Cambridge, United Kingdom, 2Cavendish Laboratory, TCM, University of Cambridge, Cambridge, United Kingdom, 3Hutchison/MRC Research Laboratories, Cambridge, United Kingdom.
Human keratinocytes are commonly used during reconstruction either as sheet grafts or passaged in vitro cultures, but direct evidence for the drivers of proliferation after transplantation are limited. Recent genetic lineage tracing studies in vivo have defined the cellular behavior in homeostatic mouse epidermis. Here, the inter-follicular epidermis is renewed by stochastically dividing progenitors and also contains a smaller second population of slow cycling stem cells which self renew and generate progenitor cells. Slow cycling cells do not participate in homeostasis but proliferate extensively following wounding. We examine the proliferation patterns in primary and cultured human keratinocytes to inform regenerative strategies.
Using in vitro timelapse imaging, we observed primary adult and cultured neonatal keratinocytes at single-cell resolution over 7 days. We recorded spatial and division outcome for related cells. Monte Carlo simulations and combinatorial statistics were used to predict population colony size distributions. Supplementation or withdrawal of growth factors(Epidermal Growth factor(EGF) or R-Spondin) were used to assess environmental responses to proliferation. Microarrays of 8-cell clones at 60 hours post-plating were performed on Affymetrix GeneChip® Human Gene 1.0ST Array.
Two distinct growth patterns emerged from 2335 divisions in 83 colonies. The first group of cells showed behaviour similar to murine progenitors, dividing to generate two differentiated daughters, two progenitor daughters or one progenitor and one differentiated daughter. Division outcome in cell pairs were independent in neighbours or closely related cells. The probabilities of generating two progenitor daughters or two differentiated daughters were similar(38±3% or 34±2%), as in vivo. In consequence, colonies derived from these progenitor-like cells increasingly accumulated differentiated cells and eventually stopped dividing.
The second group exhibited no divisions producing two differentiated daughters in the first four rounds of division, with 90% of divisions producing two dividing cells and 10% one differentiated and one proliferating cell. Here too, division outcomes were independent of neighbours or close relations. These stem-like cells generated colonies containing hundreds of cells, mostly proliferating.
Combinatorial statistics using the timelapse data predicted the distribution of clone sizes in a large sample of cells (n=1462) cultured for 7days, indicating the probabilities of each type of division were balanced as in vivo. Hypersupplementation (20ng/mL) or withdrawal (0ng/mL) of EGF, or addition of R-Spondin to the culture imbalanced the generation of proliferating or differentiating cells indicating that the fate of the progenitor-like cells was subject to external regulation.
Transcriptional analysis revealed a distinct pattern of gene expression in 8-cell colonies derived from stem and progenitor-like cells. Validation of these results by immunostaining revealed the proportion of cells expressing differentiation associated genes in progenitor-derived colonies in accordance with the timelapse data.
In summary, this is the first direct observation of two modes of cell proliferation in human keratinocytes, revealing a stem-like and progenitor-like population similar to mouse in vivo. We identify potential therapeutic targets for patients with large area skin loss or with systemic healing impairments. This also provides mechanistic insights why EGF supplementation in skin grafting has been of limited benefit after promising initial human trials.
Back to Annual Meeting Program