SIMS ion microscopy imaging of BPA and 13C15N-labeled phenylalanine in human glioblastoma cells: Relevance of subcellular scale observations to BPA-mediated boron neutron capture therapy of cancer

Subhash Chandra Daniel R. Lorey II
Int. J. Mass Spectrom. 2007, 260, pp. 90-101

p-Boronophenylalanine (BPA) is a clinically approved boron neutron capture therapy (BNCT) agent currently being used in clinical trials of glioblastoma multiforme, melanoma and liver metastases. Secondary ion mass spectrometry (SIMS) observations from the Cornell SIMS Laboratory provided support for using a 6 h infusion of BPA, instead of a 2 h infusion, for achieving higher levels of boron in brain tumor cells. These observations were clinically implemented in Phase II experimental trials of glioblastoma multiforme in Sweden. However, the mechanisms for higher BPA accumulation with longer infusions have remained unknown. In this work, by using ¹³C¹⁵N-labeled phenylalanine and T98G human glioblastoma cells, comparisons between the ¹⁰B-delivery of BPA and the accumulation of labeled phenylalanine after 2 and 6 h treatments were made with a Cameca IMS-3f SIMS ion microscope at 500 nm spatial resolution in fast frozen, freeze-fractured, freeze-dried cells. Due to the presence of the Na–K-ATPase in the plasma membrane of most mammalian cells, the cells maintain an approximately 10/1 ratio of K/Na in the intracellular milieu. Therefore, the quantitative imaging of these highly diffusible species in the identical cell in which the boron or labeled amino acid was imaged provides a rule-of-thumb criterion for validation of SIMS observations and the reliability of the cryogenic sampling. The labeled phenylalanine was detected at mass 28, as the ²⁸(¹³C¹⁵N)⁻ molecular ion. Correlative analysis with optical and confocal laser scanning microscopy revealed that fractured freeze-dried glioblastoma cells contained well-preserved ultrastructural details with three discernible subcellular regions: a nucleus or multiple nuclei, a mitochondria-rich perinuclear cytoplasmic region and the remaining cytoplasm. SIMS analysis revealed that the overall cellular signals of both ¹⁰B from BPA and ²⁸CN⁻ from labeled phenylalanine increased approximately 1.6-fold between the 2 and 6 h exposures. However, the subcellular distribution of ¹⁰B was different than the ²⁸CN in the mitochondria-rich perinuclear cytoplasmic region: ¹⁰B was reduced in this region, but ²⁸CN was not. These observations indicate that: (i) a comparable higher accumulation of BPA and phenylalanine at 6 h versus 2 h plausibly represents a similar time-dependent entry mechanism through the plasma membrane in response to cellular requirements for the amino acid in glioblastoma cells and (ii) intracellular processes, especially those implicated with mitochondria, can plausibly recognize BPA as a different molecule than phenylalanine and may significantly differ in its sequestration and metabolism. For further understanding cell cycle influence on BPA accumulation, DNA-synthesizing S-phase cells were compared with non-S-phase cells. SIMS observations revealed that after 1 h exposure to BPA, S-phase cells contained elevated levels of ¹⁰B in their nucleus in comparison to the nucleus of non-S-phase cells. Consequently, one reason that longer BPA exposures would increase its accumulation in most tumor cells will be the movement of the cell cycle through the S-phase. These observations suggest further cell cycle research in BPA-mediated BNCT and may have special significance for brain tumors since tumor cells are primarily the only cells in the brain with active proliferation characteristics. This study also shows the need for cryogenic sampling for subcellular measurements in BNCT, as even a brief thaw of frozen samples can result in gross redistribution of boron in subcellular compartments.

http://dx.doi.org/10.1016/j.ijms.2006.09.006