Cancer cell motility has been associated with metastatic potential of sublines of the Dunning R-3227 rat prostatic adenocarcinoma model. However, three sublines of high motility lacked the capacity for metastasis. In all previous works, motility has been studied upon plastic and only upon attached cells at least 18 h after gravity plating. We studied two highly motile sublines; MAT-LyLu metastasizes to lungs and lymph nodes whereas PIF-1 metastasizes rarely. We compared the motility and attachment capability of cells from the Dunning model in serum-free media upon plastic, glass, laminin, type IV collagen, fibronectin, and elastin to determine whether study upon more physiological surfaces could better explain the discrepancy in metastatic capability observed in vivo.

In the high-motility low-metastatic PIF-1 and high-motility high-metastatic MAT-Lylu sublines, membrane ruffling, pseudopodal extension, and cellular translation upon plastic in serum-free media were reduced (P < 0.0001) by 33 and 31, 70 and 60, and 77 and 65%, respectively. When returned to serum-containing media, biological characteristics (histology, chromosomal number, growth rate, host survival, and metastatic potential) were unchanged. Motility was affected by substrate (analysis of variance, P < 0.05); however, no consistent pattern of enhancement or detriment occurred in any substrate across both sublines. When motility was compared between sublines, membrane ruffling and cellular translation were relatively unaffected by substrate, whereas pseudopodal extension was altered significantly by different substrates. However, upon individual substrates, no significant differences in motility existed between the two sublines to resolve the inconsistency of high-motility but low-metastatic potential in the PIF-1 subline. Therefore, we examined cell attachment, since in order for a metastatic cell to be motile upon substrate, it must first attach to it.

Percentage of 105 cells attached was determined in serum-free media with the use of a cell counter after 1, 2, 4, 8, and 16 h on plastic and plastic coated with laminin, fibronectin, elastin, or type IV collagen. Cell attachment increased with time (P < 0.05) on all substrates for MAT-LyLu (r = 0.95) and PIF-1 (r = 0.98). Attachment of MAT-LyLu cells was impaired by fibronectin and enhanced by elastin, laminin, and type IV collagen compared to plastic at all time points. PIF-1 cell attachment was enhanced by substrate covering of plastic except with elastin at 1 and 2 h. The greatest discrepancy in attachment between sublines occurred upon elastin. Attachment of PIF-1 was less than MAT-LyLu at 1 h (7.7 ± 2.9% (SD) versus 61.6 ± 4.9%) and 16 h (60.1 ± 4.3% versus 97.9 ± 3.8%) (P < 0.05). The prevalence of elastin in lung may explain why MAT-LyLu attaches and moves through lung to form metastases whereas the equally motile but less able to attach to elastin subline PIF-1 cannot.

This content is only available via PDF.