We first tested whether trypsin digestion affected the replicative capacity of fibroblasts at the first, second and 9th transfers, by comparing two and four days of digestion to no trypsin treatment. The overall cumulative lifespan for undigested explants was shorter (Supplementary Figure 1), indicating that there was no detrimental effect of trypsin digestion on the final cumulative lifespans.
We determined the effect of serial transfer on replicative lifespan by calculating the total cumulative PDs of cells obtained from outgrowths at approximately every 10 transfers. In order to attain the maximum replicative lifespan for each donor culture, a variable split ratio was used because this strategy avoids the loss of proliferative capacity that results from seeding cells at constant density (for example, see Figure 3 in reference 8, where it was shown that seeding adult skin fibroblasts at a density of 1000 cells/cm2 shortened the lifespan by 10 PDs). In Figure 1A the final cumulative lifespan of Fre96s over 70 transfers is shown. For the first 20 transfers, the final lifespans were similar but with subsequent transfers there was an overall trend for the replicative capacity of Fre96s cultures to decrease (Figure 1B). This was also the case for other donors examined (Supplementary Figure S2). The observed reduction in replicative capacity of the HSFs most likely indicates that the fibroblasts continue dividing within the explants during the course of the experiment. However, the lifespan of the culture obtained from the final Fre96s-70 explant outgrowth was greater than those for the 40th to the 60th transfers, which may reflect outgrowth of different sub-populations from within the explant. The final cumulative PDs for each transfer can be used to obtain a more accurate estimate of the lifespan by regression analysis of final lifespans of all explanted cultures (Figure 1B). The lifespan calculated by regression analysis is used in the remainder of this study.
Although transferring explant cultures for three years is unlikely to become routine, our results demonstrate the capacity of this protocol to increase the number of early passage cells obtained from explants at least 100-fold. At later transfers the HSF cultures had shorter lifespans, but for the first 20 transfers the decrease in lifespan was modest (see Figure 1B, where the first 20 transfers show approximately 10% reduction in overall lifespan). Therefore HSFs isolated from the first 20 transfers would provide a large supply of HSFs with similar growth potentials for experimentation. The lifespan of cultures obtained after 50 to 60 transfers dropped by about 50% compared with the first transfer (Figure 1B; data not shown). Because the flask of cells obtained from tissue outgrowth is usually regarded as PD0, the size of the explanted tissue has an effect on the apparent culture lifespan because more cell generations are required to obtain the first flask of cells growing out from smaller pieces of tissue. In agreement with this, a study using a different explant technique found that a 1mm2 skin biopsy per 25cm2 flask could be transferred four times; the lifespans from the first three transfers were similar, but the replicative capacity of the cultures from the last transfer dropped by 72%, presumably because cells had been depleted in the tissue fragment by the fourth transfer (8).Characterization of HSFs isolated using the serial transfer protocol
HSFs obtained from donors of different chronological ages were used to determine whether age correlated with proliferative potential or cellular stress response. HSFs were obtained from early transfers (first, second, or third) from 16 donors aged 18 to 66 years. The final cumulative PDs ranged from 36.4 to 67.2 and did not correlate with donor age (Table 1 and Figure 2A). The relationship between HSF replicative capacity and donor age has been contentious. Early studies showed a negative correlation between donor age and lifespan (9), but this has not been confirmed by later studies (10, 11). Our study, which analyzed a relatively small cohort, supports the conclusion that there is no correlation between the in vitro growth potential of fibroblasts and donor age (Figure 2B).