Isolation of tumor-initiating cells currently relies on markers that do not reflect essential biologic functions of these cells. We proposed to overcome this limitation by isolating tumor-initiating cells based on enhanced migration, a function tightly linked to tumor-initiating potential through epithelial-to-mesenchymal transition (EMT). We developed a high-throughput microfluidic migration platform with automated cell tracking software and facile recovery of cells for downstream functional and genetic analyses. Using this device, we isolated a small subpopulation of migratory cells with significantly greater tumor formation and metastasis in mouse models. Whole transcriptome sequencing of migratory versus non-migratory cells from two metastatic breast cancer cell lines revealed a unique set of genes as key regulators of tumor-initiating cells. We focused on phosphatidylserine decarboxylase (PISD), a gene downregulated by 8-fold in migratory cells. Breast cancer cells overexpressing PISD exhibited reduced tumor-initiating potential in a high-throughput microfluidic mammosphere device and mouse xenograft model. PISD regulated multiple aspects of mitochondria, highlighting mitochondrial functions as therapeutic targets against cancer stem cells. This research establishes not only a novel microfluidic technology for functional isolation of tumor-initiating cells regardless of cancer type, but also a new approach to identify essential regulators of these cells as targets for drug development.

Summary Infection of BALB/c mice with murine cytomegalovirus (MCMV) decreased the numbers of cells recovered from the thymus by 80–90% after 4–7 days, although less than 10 thymocytes per million were productively infected with the virus. A loss of cortical thymocytes was evident in histologic sections and correlated with depletion of CD4+CD8+ cells. Thymic involution was minimal in C57BL/6 mice. This resistance was not H‐2b‐associated, as BALB.B (H‐2b) mice were severely affected. In CXB recombinant inbred mice, thymic involution and MCMV replication co‐segregated with atrophy and infection of the spleen and bone marrow. This suggests common regulation by natural killer (NK)1.1+ cells, consistent with the enhanced thymic involution demonstrated in NK‐deficient bg/bg mice. However, CD4− CD8− cells were not depleted, so bone marrow hypoplasia may not be the proximal cause of thymic involution. MCMV infection activated CD4+, CD8+ and CD4+ CD8+ thymocytes, as expression of MEL14, major histocompatibility complex class I (H‐2) and Sca‐1 antigens increased on these cells. In vitro lymphoproliferation and interleukin (IL)‐3 release were enhanced in unseparated and CD4+ ‐enriched thymus preparations. Maturation of the thymus population was also evident from the high frequencies of single positive CD4+ and CD8+ cells and the decline in Sca‐2 expression. However, unlike peripheral T cells, thymocytes from infected mice did not release IL‐2. The results suggest that thymic involution accelerates the transit of cells through the thymus. The possibility that this impairs the elimination of autoreactive T cells within the thymus and promotes the autoimmune manifestations of MCMV disease is discussed.

Peritoneal (PM) and bone marrow-derived (BMM) macrophages and lung fibroblasts (LF) from inbred, intra-H-2 recombinant, H-2 mutant, and hybrid mice were infected with murine cytomegalovirus (MCMV) under centrifugal enhancement. At the concentration of virus employed, peritoneal macrophages from strains carrying Kd, Kb, Dd, Ks and/or Ds, Kq and/or Dq alleles could be infected to a level of 80%-100%, as assessed by viral antigen expression or loss of Fc receptors. Cells lacking these haplotypes and carrying Kk, Kj, Dk, Dj, or Db were resistant, yielding levels of infection below 20%. The background (non-H-2) and class II genotype and the S allele did not influence the proportions of cells infected. Furthermore, sensitivity was dominant in the F1 progeny of H-2b X H-2k and H-2d X H-2k crosses, and was not compromised by the bm1, bm3, bm10, or bm14 mutations in the alpha 1 or alpha 2 regions of Kb or Db. The proportions of cells able to release infectious virus were low, but paralleled the frequencies of viral antigen expression. The class I genotype also determined susceptibility to MCMV infection in BMM and LF, although up to 35% of H-2k BMM and 46% of H-2k LF could be infected. The findings are consistent with an association between K and D antigens and a cellular receptor for MCMV on all three cell types.

Genetically-encoded fluorescence resonance energy transfer (FRET) reporters are powerful tools to analyze cell signaling and function at single cell resolution in standard two-dimensional cell cultures, but these reporters rarely have been applied to three-dimensional environments. FRET interactions between donor and acceptor molecules typically are determined by changes in relative fluorescence intensities, but wavelength-dependent differences in absorption of light complicate this analysis method in three-dimensional settings. Here we report fluorescence lifetime imaging microscopy (FLIM) with phasor analysis, a method that displays fluorescence lifetimes on a pixel-wise basis in real time, to quantify apoptosis in breast cancer cells stably expressing a genetically encoded FRET reporter. This microscopic imaging technology allowed us to identify treatment-induced apoptosis in single breast cancer cells in environments ranging from two-dimensional cell culture, spheroids with cancer and bone marrow stromal cells, and living mice with orthotopic human breast cancer xenografts. Using this imaging strategy, we showed that combined metabolic therapy targeting glycolysis and glutamine pathways significantly reduced overall breast cancer metabolism and induced apoptosis. We also determined that distinct subpopulations of bone marrow stromal cells control resistance of breast cancer cells to chemotherapy, suggesting heterogeneity of treatment responses of malignant cells in different bone marrow niches. Overall, this study establishes FLIM with phasor analysis as an imaging tool for apoptosis in cell-based assays and living mice, enabling real-time, cellular-level assessment of treatment efficacy and heterogeneity.

We have studied the effects of murine cytomegalovirus (MCMV) infection on bone marrow stem and progenitor cell populations to find an explanation for the defects in hematopoiesis that accompany CMV infections in patients. Sublethal MCMV infection of BALB/c mice resulted in a 5- to 10-fold decrease in the numbers of myeloid (colony- forming unit-granulocyte-macrophage [CFU-GM]) and erythroid (burst- forming unit-erythroid [BFU-E]) progenitor cells in the marrow, but not in primitive myeloerythroid progenitor cell (colony-forming unit-spleen [CFU-S]) numbers. In contrast, we observed a 10- to 20-fold reduction in CFU-S as well as CFU-GM and BFU-E in lethally infected mice. Depletion of marrow CFU-GM was less severe in C57BL/10 and C3H/HeJ mice, which are more resistant to the effects of MCMV infection. Treatment of bone marrow cells with MCMV preparations in vitro did not reduce the numbers of CFU-GM, although up to 10% of the cells were productively infected. This finding suggests that CFU-GM were not susceptible to lytic MCMV infection in vitro and are probably not eliminated by lytic infection in vivo. Increases in the frequencies of Sca-1+Lin- marrow cells, a population that includes cells with the characteristics of pluripotential stem cells, were observed in MCMV- infected BALB/c, C57BL/10, and DBA/2J mice. Increases in the frequencies of c-kit+Lin- marrow cells were only seen in DBA/2J mice. MCMV infection did not impair the function of pluripotential stem cells because transplantation of marrow from MCMV-infected donors into irradiated recipient mice resulted in successful reconstitution of the T, B, and myeloid cell lineages.