Single-cell pharmacokinetic imaging reveals a therapeutic strategy to overcome drug resistance to the microtubule inhibitor eribulin - PubMed (original) (raw)

Single-cell pharmacokinetic imaging reveals a therapeutic strategy to overcome drug resistance to the microtubule inhibitor eribulin

Ashley M Laughney et al. Sci Transl Med. 2014.

Abstract

Eribulin mesylate was developed as a potent microtubule-targeting cytotoxic agent to treat taxane-resistant cancers, but recent clinical trials have shown that it eventually fails in many patient subpopulations for unclear reasons. To investigate its resistance mechanisms, we developed a fluorescent analog of eribulin with pharmacokinetic (PK) properties and cytotoxic activity across a human cell line panel that are sufficiently similar to the parent drug to study its cellular PK and tissue distribution. Using intravital imaging and automated tracking of cellular dynamics, we found that resistance to eribulin and the fluorescent analog depended directly on the multidrug resistance protein 1 (MDR1). Intravital imaging allowed for real-time analysis of in vivo PK in tumors that were engineered to be spatially heterogeneous for taxane resistance, whereby an MDR1-mApple fusion protein distinguished resistant cells fluorescently. In vivo, MDR1-mediated drug efflux and the three-dimensional tumor vascular architecture were discovered to be critical determinants of drug accumulation in tumor cells. We furthermore show that standard intravenous administration of a third-generation MDR1 inhibitor, HM30181, failed to rescue drug accumulation; however, the same MDR1 inhibitor encapsulated within a nanoparticle delivery system reversed the multidrug-resistant phenotype and potentiated the eribulin effect in vitro and in vivo in mice. Our work demonstrates that in vivo assessment of cellular PK of an anticancer drug is a powerful strategy for elucidating mechanisms of drug resistance in heterogeneous tumors and evaluating strategies to overcome this resistance.

Copyright © 2014, American Association for the Advancement of Science.

PubMed Disclaimer

Figures

Fig. 1

Fig. 1. Eribulin-BFL accumulation phenotype observed in native and taxol-resistant cells

(A) Structure of fluorescent eribulin-BFL (green shows BODIPY fluorescent tag). Synthesis is described in fig. S1. (B) The fluorescent eribulin analog, eribulin-BFL, retained an anticancer efficacy in HT1080 cells in vitro, comparable to the unlabeled compound. Data were measured in triplicate at escalating drug concentrations. (C) The fluorescent BODIPY tag did not impede drug uptake in wild type (taxane-sensitive) HT1080 cells in vitro. In contrast, sporadic accumulation of eribulin-BFL was observed in taxol-resistant HT1080 cells. Images were acquired at identical exposure settings.

Fig. 2

Fig. 2. Heterogeneous eribulin uptake and toxicity across human cancer cell lines

(A and B) Variable eribulin (unlabeled parent compound) cytotoxicity (A) and eribulin-BFL uptake (B) profiles across a human cell line panel, including breast (MDA-MB-231, MDA-MB-436, SK-BR3, 1834, 4175), ovarian (OVCAR-429 and A2780), pancreatic (PANC1), Ewing's scaroma (TC-71) and fibrosarcoma (HT1080: wild type, taxol resistant, and MDR1-expressing). Two human breast cancer lines were derived from metastases to the lung, 1834 and 4175. (C) Linear regression of eribulin-BFL accumulation (min-max normalized) and unlabeled eribulin efficacy, here defined as the inverse of its cellular IC50. Solid lines represent regression best-fit lines and dotted lines represent associated 95% confidence bands. Each data point represents the average of triplicate experiments; where mean drug accumulation was analyzed in 10,000 cells per experiment. (D and E) Heterogeneity in half maximal inhibitory concentration, IC50 (D), and in the surviving fraction of cells (E) in response to eribulin treatment. Data are means ± SEM (n=3). (F) Nonlinear regression of MDR1 protein expression and surviving cell fraction. MDR1 protein expression was normalized to GAPDH. Each data point represents the average of triplicate experiments. Solid lines represent regression best-fit lines and dotted lines represent associated 95% confidence bands. P values for the regression analyses in (C and F) were determined by the Student's t distribution.

Fig. 3

Fig. 3. System to quantify the in vivo pharmacokinetics of eribulin resistance using a fluorescent analog

(A) The partially (mosaic) taxane-resistant HT1080 tumor model engineered to heterogeneously express the MDR1-mApple fusion protein and uniformly express the fluorescent histone fusion protein (H2B-iRFP), was implanted subcutaneously in a mouse window chamber. The kinetics of eribulin-BFL tumor cell uptake were imaged intravitally, typically over the course of 2 hours. A segmented vessel and fiducial markers are outlined in yellow for visual co-registration between frames. Images are representative of 3 mice. (B) Unique cell objects identified by the automated segmentation algorithm are outlined and labeled with a color indicating their relative expression of the MDR1 protein. (C) Drug accumulation and efflux curves for 32 cells produced by the automated tracking algorithm, reporting mean cellular drug accumulation (single cells indicated by color) as a function of time. Curves indicated best-fit lines when fitting data to analytical model of drug diffusion and efflux.

Fig. 4

Fig. 4. Single cell and population pharmacokinetics of wild type and resistant tumor subpopulations in vivo

(A) Experimental data and analytical modeling of eribulin-BFL PK in wild-type and resistant (MDR1++) single cells, and in mixed populations of both wild-type and resistant cells. Resistant cells are defined as those fluorescently expressing MDR1-mApple. The vascular concentration shows drug concentration increase and loss within a segmented tumor vessel. Data are means ± SEM (n = 37 cells, trace from 1 mouse) with solid lines representing best-fit curves. Images were analyzed using automated cell segmentation and tracking algorithms. All cells were identified by the histone H2B-iRFP fusion protein, and the resistant subpopulation was identified by MDR1-mApple fluorescence. (B) The relative drug accumulation in resistant cells as compared to their wild type counterparts (ratio of means ± SEM, n = 3 mice) in a highly vascularized tumor, with a vessel in the imaging field and distant from detectable tumor vasculature, as determined by 3-dimensional imaging stack. Population pharmacokinetics near (< 200 μm) and far (> 200 μm) from vessels are displayed in the image insets from representative mice. (C) Relative eribulin-BFL accumulation in MDR1-expressing cells at steady state was quantified by the fit plateau of a one phase association (near to vessels) or decay (far from vessels). Data are mean ratios ± SEM, as well as individual animals (n = 6 independent time lapse z-stacks acquired in three mice). P value determined by a one-tailed t-test assuming equal SD.

Fig. 5

Fig. 5. Reversal of MDR1-mediated drug resistance in vitro using the experimental MDR1 inhibitor, HM30181

(A) A third-generation MDR1 inhibitor, HM30181 (16), restored sensitivity to eribulin in a dose-dependent manner in vitro in HT1080 cells over-expressing MDR1. (B) HM30181 does not alter the cytotoxicity of eribulin in the drug-sensitive parent cell line, HT1080 (human fibrosarcoma). (C) Sensitivity of resistant and wild-type HT1080 cells to eribulin in the presence of HM30181. Data in (A to C) are reported as averages of triplicate experiments. (D) Cellular eribulin-BFL uptake in resistant (MDR1++) and wild-type cells in response to increasing concentrations of HM30181. Single-cell levels of fluorescence drug were determined by confocal microscopy, where all images were analyzed at same exposure settings. Per-cell count was determined using the DRAQ5 nuclear label. Data are means ± SEM (an average of n = 38 cells evaluated per condition). (E) Eribulin-BFL is not taken up by MDR1-expressing cells (white arrows) in a mixed population of HT1080 cells. However, incubating resistant cells with HM30181 allowed eribulin influx (white arrows). (For the effects of other MDR1 inhibitors, see fig. S10.) Images were acquired at identical exposure settings. The nuclear stain is DRAQ5. Images are representative of n = 5 FOV.

Fig. 6

Fig. 6. Reversal of MDR1-mediated drug efflux in vivo with topical application of the MDR1 inhibitor, HM30181, to heterogeneous tumors

(A) Eribulin-BFL accumulation in heterogeneous tumors containing MDR1-expressing cells (arrows), with and without topical administration of HM30181 directly to the tumor site. Images are representative of 3 mice. (B) Single-cell quantification of eribulin-BFL accumulation in wild type and resistant single cells, with or without HM30181 (min-max normalized per frame). Horizontal lines indicate median uptake per unit area and all individual data points are shown; an average of 103 cells were analyzed per group. (C) Relative drug increase in wild type and MDR1-expressing (resistant) cells with topical application of HM30181. Data are the ratio of median drug accumulation per cell type before and after topical application of HM30181 in n = 3 mice. All P values determined according to a two-sided unpaired t-test, assuming equal SD.

Fig. 7

Fig. 7. Nanotherapeutic strategy for reversing MDR1-mediated drug resistance in vivo

(A) Schematic and characterization of the PLGA-PEG-BODIPY-630 nanoparticles used to encapsulate the MDR1 inhibitor, HM30181. The particles were 76 ± 7 nm (average weighted by volume ± SD across 6 synthesis batches). (B) Release of inhibitor from nanoparticles in saline at 37°C. Data are means ± 95% confidence intervals for duplicate measurements. (C) Vascular half-life (t1/2) following rapid clearance from the tumor vasculature into surrounding tissue. Data are means +/− SEM (n = 5 vessels). (D) Eribulin-BFL accumulation in MDR1-expressing HT1080 cells in vivo without HM30181 or with daily pre-injections of HM30181, in standard solution phase vehicles or nano-encapsulated. Images are representative of 3 mice. (E) Inhibitor efficacy was a function of delivery method. The fraction of wild-type (blue) or resistant (red) cells accumulating drug was quantified for each delivery method according to the cellular frequency of eribulin-BFL concentration in MDR1-expressing cells within 1.96 SD of the mean native drug accumulation. Data are mean cellular frequencies ± SEM per imaging frame (an average of n = 300 MDR1-expressing cells evaluated per delivery method). P values were determined by one-way ANOVA using Tukey's multiple comparison test.

Similar articles

Cited by

References

    1. Arao S, et al. Expression of multidrug resistance gene and localization of P-glycoprotein in human primary ovarian cancer. Cancer Res. 1994;54:1355–1359. - PubMed
    1. Arceci RJ. Clinical significance of P-glycoprotein in multidrug resistance malignancies. Blood. 1993;81:2215–2222. - PubMed
    1. Bell DR, Gerlach JH, Kartner N, Buick RN, Ling V. Detection of P-glycoprotein in ovarian cancer: a molecular marker associated with multidrug resistance. J Clin Oncol. 1985;3:311–315. - PubMed
    1. Dalton WS, et al. Immunohistochemical detection and quantitation of P-glycoprotein in multiple drug-resistant human myeloma cells: association with level of drug resistance and drug accumulation. Blood. 1989;73:747–752. - PubMed
    1. Pirker R, et al. MDR1 gene expression and treatment outcome in acute myeloid leukemia. J Natl Cancer Inst. 1991;83:708–712. - PubMed

Publication types

MeSH terms

Substances

Grants and funding

LinkOut - more resources