Sensitivity, recovery and cell cycle effects of cisplatin on VU-SCC-1131 and VU-SCC-OE
To assess in vitro drug response of our cell lines, we tested the sensitivity of both cell lines for cisplatin under continuous exposure, and the respective dose-response curves are shown in Fig. 1. As expected, the VU-SCC-1131 was more sensitive for cisplatin compared to VU-SCC-OE, with IC50 values of 0.21 µM versus 0.95 µM respectively. These results were in line with previous reported IC50 values (18). Effects after recovery were assessed by incubating cells for 4 hours with cisplatin and cell viability measured at 72 hours. The half maximal inhibitory concentration shifted for both cell lines ~ 3 times to 0.70 µM (VU-SCC-1131) and 3.3 µM (VU-SCC-OE) (Fig. 1A,B and C). The difference of 5x in response did not differ.
Next, we studied the dose-, cell line- and time-dependent effects of cisplatin on the cell cycle using flow-cytometry analysis after EdU labeling, which combines high resolution DNA content analysis with specific cell cycle markers. We treated cells for 4 hours with cisplatin: a low dose (0.04 µM) cisplatin which correlates on our recovery experiment with the IC1.7 value of both VU-SCC-1131 and VU-SCC-OE, and a high dose (0.2 µM) which correlates with the IC14.7 of VU-SCC-1131 and the IC7.0 of VU-SCC-OE. Independent of dose and cell line, cell cycle analysis directly after cisplatin exposure did not show any effects on the cell cycle compared to untreated samples.(Fig. 2A). At both the low and high dose VU-SCC-1131 accumulates in G1-phase after 72 hours recovery time, which prolongs after 96 hours.(Supplementary Fig. 1A) The number of cells in mitosis did not change due to the low dose of cisplatin in VU-SCC-1131, when the high dose is applied the fraction of cells in M-phase increased with longer recovery time (Supplementary Fig. 1B). VU-SCC-OE did not show major alterations in the distribution in either G1, S or G2/M-phase after exposure to both concentrations. Nonetheless, 2–3 times higher numbers of cells were in mitosis during recovery phase, dependent on the dose. (Supplementary Fig. 1B) Although the cell number did not change dramatically, the Edu incorporation during S-phase was severely hampered, particularly in VU-SCC-1131 cells. (Fig. 2B) The median fluorescence intensity (MFI) decreased substantially more in the cisplatin-sensitive VU-SCC-1131 cell line compared to the resistant VU-SCC-OE cell line after 72 hours recovery time (Fig. 2C).Likely as result of the DNA repair defect, the number of cross-links remains high, the replication forks collapse and DNA synthesis is inhibited.
Next we tested whether the differences in vitro in the cell lines, are also reflected in vivo. In total 10 mice were injected on both flanks with 2*10e6 VU-SCC-OE cells per site and randomized to either cisplatin or control group. In the control group all tumors (n = 10) developed, in the treatment arm, only 6 of 10 tumors. For VU-SCC-1131 6 mice were injected with cells on both flanks, and all tumors grew out sufficiently for further analysis. Mice were treated intra-peritoneal with cisplatin 0.5mg/kg or the control-vehicle at day 0 and 7. Growth curves are depicted in Fig. 1D and 1E. Analysis is hampered by the different growth rate of the two cell lines. We observed that VU-SCC-1131 xenografts were more sensitive to cisplatin compared to VU-SCC-OE xenografts, resulting in larger growth delay factor in VU-SCC-1131 (1.61 and 2.28 at day 7 and day 10 respectively) versus VU-SCC-OE (1.27 and 1.63). To analyze whether this was a significant difference, we fitted a linear mixed effects model for both cell lines with relative tumor volume as the outcome variable and the cell line (cisplatin vs vehicle) as the fixed effect, we included time in days, each individual mouse and growth speed as random effect. The overall model predicting relative tumor volume explained 74.1% of the variance. The variance explained by the fixed effects was 16.3%. The model’s intercept is at 2.13 (SE = 0.58, 95% CI [0.98–3.29]). The reduction in tumor growth by cisplatin treatment was significantly different between xenografts of both cell lines, VU-SCC-1131 showed significant increased growth delay (beta = 1.73, SE = 0.64, 95% CI [0.46–3.00], p < 0.01).
Uptake and retention of Pt-195m cisplatin
The experiments described, indicate that there is a significant difference in cisplatin sensitivity between the two cell lines, and the data are well explained by the known genetic defect. Next we analyzed using these models whether this is reflected by Pt-195 cisplatin biodistribution as well. In vitro distribution and retention was evaluated by direct measurement of intra-cellular amounts of Pt-195m cisplatin. Since total cellular uptake of cisplatin might be an important factor for the effect of cisplatin on cell survival between different cell lines, we measured the uptake of Pt-195m cisplatin at 4 hours incubation in both cell lines.
The amount of Pt-195m cisplatin adducts bound to the DNA was measured at the same time points (4 and 24 hours incubation), the amount of DNA-bound cisplatin increments with both the concentration and the incubation time as expected (Supplementary Fig. 2). However, against our expectations, both at 4 hrs and 24 hrs incubation the insensitive cell line VU-SCC-OE accumulated a higher amount of cisplatin in the DNA than VU-SCC-1131. This initial increase in accumulation needs to be taken into consideration when DNA repair is allowed by cisplatin removal, and to compensate for this effect we calculated the ratio with/without repair (Fig. 3). To obtain enough counts after 24 hrs repair we had to increase the dose to 75 uM. As the IC50 shifts 4 times with and without removal of cisplatin, we concluded that this was acceptable. Remarkably, when removing cisplatin from the medium, the accumulation in the DNA increased. The intracellular and protein bound pools apparently accumulated in the DNA. In VU-SCC-OE the additional accumulation was counterbalanced by crosslink repair, and the ratio was a mere 1.45. However the accumulation in VU-SCC-1131 increased tremendously with a ratio of 3.4, suggesting that the insensitive cell line VU-SCC-OE cleared DNA adducts from the DNA while the sensitive cell lines VU-SCC1131 did not, or considerably less. The higher sensitivity for cisplatin, the increased effect on S-phase arrest of cell line VU-SCC-1131, and the decreased removal of cisplatin adducts from the DNA, are all in line with and explainable by the FA phenotype.
Biodistribution of Pt-195m labeled cisplatin in mice
All data together made it difficult to predict whether Pt-195 displays a different biodistribution in vivo in the two cell lines. As the two different cell lines can be grown as xenograft in immune deficient mice, we tested radiolabeled cisplatin accumulation. Figure 4A shows the biodistribution of the radioactive cisplatin in the VU-SCC-1131-bearing mice after 1, 6 and 24 hrs. We also tested peritoneal and intravenous administration. The administered Pt-195m cisplatin is rapidly, within the first hour, distributed to each organ. After the first hour, we did not observe large changes in biodistribution. Most of the circulating Pt-195m cisplatin is cleared through glomular filtration, because of the high accumulation of Pt-195m cisplatin in the urine fraction. Accumulation of Pt-195m cisplatin in the main clearing organs remains relatively high over time, with median renal/blood and liver/blood AUC ratios of 4.31 (2.82–4.41) and 2.46 (2.33–3.04) respectively. There were no significant differences in the distribution to the large organs between intravenous (IV) and intra-peritoneal (IP) administration of the Pt-195m cisplatin solution.(Supplementary table 1 and supplementary Fig. 3) Likewise, the difference in amount of uptake and retention in the xenografts was not significant between IV and IP drug delivery.(Supplementary Fig. 4)
Uptake and retention of Pt-195m labeled cisplatin in xenografts
Figure 4B shows the uptake and retention of Pt-195m labeled cisplatin in xenografts of sensitive and non-sensitive cell lines. In the xenografts of both cell lines the uptake takes place within the first hour, with a maximum uptake at the 1 hour time point of 0.012 (0.010–0.014) for VU-SCC-1131 and 0.014 (0.012–0.018) for VU-SCC-OE median absorbed counts (as percentage of injected counts). There was a decline of retention of Pt-195m cisplatin, with lowest amounts after 48 hours in both cell lines (VU-SCC-1131 vs VU-SCC-OE; \(\stackrel{\sim}{\mathcal{x}}\) = .007 (0.007–0.009) vs \(\stackrel{\sim}{\mathcal{x}}\) = .007 (0.004–0.009)). A Wilcoxon signed-rank Test indicated that the maximum uptake in the sensitive cell line (VU-SCC-1131) was not statistically significant higher than the resistant cell line (VU-SCC-OE) Z = 7, p = .09 (Fig. 1B), and the amount of Pt-195m cisplatin remained equal at subsequent time points (6h: Z = 9, p = .33; 24h: Z = 15, p = .70; 48h: Z = 13, p = .91). The course of the uptake and retention measured by the AUC of the median percentage of absorbed counts-time curve was comparable between both cell lines (VU-SCC-1131: \(\stackrel{\sim}{\mathcal{x}}\) = 0.41 and VU-SCC-OE: \(\stackrel{\sim}{\mathcal{x}}\) = 0.45), the difference was not statistically different (Z = 19, p = .94).