Abundance of message levels for pyruvate kinase (PKM2) and thymidylate synthase (TYMS) in CNS cancer cell lines appears better correlated with their doubling times than do Ki67 or PCNA message levels

Objectives Cancer cell-lines, established from either primary or metastatic sites of cancer patients, have contributed to our understanding cancer biology including radiotracer selection and uptake mechanisms. Cell lines with shorter doubling time (DT) typically have more aggressive behavior. In the current study, we assessed correlations between the DTs of CNS cancer cell lines and the abundance of mRNA for Ki67, PCNA, TK1, thymidylate synthase (TYMS), and glucose transporters (SLC29A1-2)

Methods A NCI 60 exon dataset (interrogating over 1 million exon clusters within the known and predicted transcribed regions of the entire genome) was selected from the NIH Gene Expression Omnibus which contains 178 chips including 5 CNS cancer cell lines with triplicates for each: SF-268, SF-295, SNB-19, SNB-75, and U251. Using Partek, commercial software, we obtained the average abundance of transcripts of interest above. Correlations between the abundance of genes and NCI documented doubling times were plotted with Excel (α=0.05 for a one-tailed test)

Results The correlation coefficient r of PKM2 vs. DTs was -0.87 (p=0.03 < 0.05=α, one-tail). The r of TYMS (de novo synthesis of thymidine) vs. DTs was -0.86 (p= 0.03). The r of TK1 (phosphorylating FLT) vs. DTs was -0.79 (p=0.06). The r of Ki67 vs DTs was -0.44 (p=0.22). The r of PCNA vs. DTs was 0.31 (p=0.30). The r of SLC29A1 (ENT1, FLT transporter) vs. DTs was -0.49 (p=0.20). The r of SLC29A2 (ENT2) vs. DTs was -0.36 (p=0.28). Correlations between DTs and HK1 or glucose transporters were not observed.

Conclusions The abundance of PKM2 and TYMS appear to have better correlations to DTs of CNS cancer cell lines than TK1, and conventional proliferation markers Ki67 and PCNA. These correlations are negative, indicating higher PKM2 and TYMS abundance and shorter DT. Our conclusions must be viewed as preliminary until confirmed by proteomic data in larger number of cell lines.