Motility is a common property of animal cells. Cell motility is required for embryogenesis , tissue morphogenesis  and the immune response  but can also lead to disease processes such as metastasis of cancer cells . Analysis of cell migration in native tissue in vivo has yet to be fully explored, but motility can be relatively easily studied in vitro in isolated cells. Recent evidence suggests that cells plated in vitro on thin lines of adhesive proteins printed onto culture dishes can recapitulate many features of in vivo migration on collagen fibers [5, 6]. However, even with controlled in vitro measurements, the characteristics of motility are diverse and are dependent on the cell-type, origin and external cues. One objective of the World Cell Race was to perform a large-scale comparison of motility across many different adherent cell types under standardized conditions. To achieve a diverse selection, we enlisted the help of many international laboratories who submitted cells for analysis. The large-scale analysis, made feasible by this competition-oriented collaboration, demonstrated that higher cell speed correlates with the persistence of movement in the same direction irrespective of cell origin.
The race track consisted of 4 and 12 micron-wide fibronectin lines printed in multi-well glass-bottomed cell-culture wells (see Supplementary Methods and Fig. S1A). Fifty-four different cell-types from various animals and tissues were provided by forty-seven laboratories. Genotypically, cells were wild type, transformed or genetically engineered (see Supplementary Table 1). The cells were distributed to six organizing laboratories (two in the USA, and one each in the UK, France, Germany and Singapore) who prepared cell culture stocks using the frozen samples received from participating laboratories and plated these onto the race tracks under identical culture conditions. Cells were allowed to adhere overnight and cell motility was recorded for 24 hours using an inverted video microscope (Fig. 1B, movie S1-S3). Cell morphology (length, shape, symmetry, nucleus position) varied greatly from one cell-type to the other (Fig. 1A). Nuclear staining enabled individual cells to be automatically tracked (see Supplementary Methods, Fig. S1B and movie S4). The motility of over 7,000 cells was compared, with an average of 130 cells analyzed per cell-type. Detailed statistical analyses were used to characterize cell motility parameters for each cell-type (see http://www.worldcellrace.com/ResultFiles).
The mean instantaneous speed of individual cells is computed by averaging the cell displacements between consecutive frames over time. The distribution of cell mean instantaneous speeds for each cell-type was asymmetric and non-Gaussian (Fig. 1C and Fig. S1C). Interestingly, we observed that a higher mean speed for a given cell-type did not reflect a global shift of the speed distribution, but rather the spreading of the distribution due to the presence of faster moving cells (Fig. 1C and Fig. S1C). In order to identify the 2011 World Cell race winner, only cells with an effective overall displacement of at least 350 μm were considered. This cut-off was only reached by 26 of the 54 cell-types. The highest migration speed was recorded at 5.2 micron/minute by a human embryonic mesenchymal stem cell (movie S5).
Cell displacements on lines can be described by a 1D correlated random walk , simply derived from the 2D model, in which cells are more likely to move in the direction of the immediately preceding movement conserving their polarity. This can be quantified by a persistence probability (p) for a cell to maintain its direction of motion and keep the same front and rear. To calculate p for each cell-type, histograms built from the number of cell steps performed in the same direction were fitted to the 1D correlated random walk theory (Fig. S1D and S1E). A persistence path, defined as the ratio of the effective maximum displacement to the actual trajectory length, was further calculated to obtain a macroscopic measure of p (Fig. S1F). This ratio was strongly correlated with the persistence probability (Fig. S1G). Persistence path distributions for the 54 cell-types were typically non-Gaussian (Fig. S1H). Strikingly, the overall mean speeds for all cell-types correlated well with their mean persistence path (Fig. 1D), implying that fast-moving cell-types (mean cell speed >0.7 micron/minute) displayed high mean persistence path (>0.5). Cells moving rapidly, but only backward and forward were not observed.
Given the large and diverse sample of cell-types, this result may reveal a conserved mechanism that allows the coupling of the machinery controlling cell polarity (responsible for persistent oriented motion) to the one regulating instantaneous cell speed. Future experiments aimed at unraveling the associated molecular mechanisms shall now be performed.
Together, the results generated by the first world cell race highlight how scientific games involving large-scale experiments can lead to the identification of novel and relevant biological processes, which may otherwise escape observation.