Methodological aspects of analysing human breast cancer cell lines by NMR spectroscopy.

In an attempt to identify the factors which might affect the measurement of water proton relaxation times in cultured cells, we have begun a long-term study of two human breast cancer cell lines, MDA-MB-231 and MDA-MB-435s. We tested growth rates and cell cycle distribution as intrinsic properties of the cells as well as methodological steps which might affect the measurement of T1 and T2. A detailed examination of the growth rates of the two cell lines, easily recognized as slow (231) and fast (435s) in culture, revealed that this attribute is difficult to correlate precisely with T1s or T2s. The reason is that the relaxation times are necessarily measured at one point in time while the growth rates are a summation of ongoing processes occurring over hours. Cell cycle distribution, on the other hand, can be measured simultaneously with the relaxation times by using cells quick-frozen from the same suspension. By this method, cell cycle distribution appears to be reflected through an effect on T1s. For example, cell pellets distributed 72:15:14 in G0G1:S:G2M has longer T1s (p less than 0.01) than those distributed 43:34:23 in G0G1:S:G2M. Regarding methodological factors, trypsin appeared to lower water content and T2s in the 231 cell line. Drift in the cell cycle distribution after sample preparation did not become significant until after 2 hours in the NMR tube. It was important to standardize the force and duration of centrifugation of the cell pellets to minimize the contribution of the suspending medium without affecting cell viability. We conclude that, given careful control of methodological factors, differences in T1 may reflect metabolic differences as demonstrated by T1 differences in cell pellets showing divergent cell cycle distribution.