Advertisement

Dual-action CXCR4-targeting liposomes in leukemia: function blocking and drug delivery

Catriona McCallion, Anna D. Peters, Andrew Booth, Karen Rees-Unwin, Julie Adams, Raisa Rahi, Alain Pluen, Claire V. Hutchinson, Simon J. Webb and John Burthem

Data supplements

Article Figures & Data

Figures

  • Figure 1.

    Molecular structures of the compounds used in this article, with the symbols used for schematic representations. (A) Structure of BAT1, and its derivatives, BAT1-Cy5 (fluorescently labeled BAT1 conjugate) and BAT1-cholesterol (used to decorate liposomes with BAT1). (B) Structure of the fluorescently labeled lipid, TopFluor, which was used to label liposomal membranes for ease of tracking. (C) Structure of doxorubicin, the drug cargo loaded into liposomes and delivered using BAT1 targeting.

  • Figure 2.

    CLL cells consistently express CXCR4 in the peripheral blood and lymphoid organs, whereas CXCR7 expression is significantly lower, (A) Healthy tonsil and CLL spleen sections were stained with antibodies against CXCR4 or CXCR7. In normal tonsil, CXCR4 is widely expressed but with noticeably stronger expression in follicles. In contrast, expression of CXCR7 is weak, and it is expressed primarily by vascular endothelial cells. In organs infiltrated by CLL (shown in spleen), expression of CXCR4 and CXCR7 is observed across the entire sample. CXCR4 staining in proliferation zones is weaker than in other regions of the sample but is still observed. CXCR7 expression is detected, but it is far weaker than CXCR4, with no strong differences observed between proliferation centers and the surrounding tissue. The insets in the upper panels are shown at increased magnification in the lower panels. HRP/DAB technique was used with hematoxylin counterstain. Scale bars, 500 µm. (B) Immunoblot showing relative protein expression levels of CXCR4 and CXCR7 in primary CLL cells, with total ERK1/2 presented as a protein expression control. Densitometric quantification of the blots is shown with adjustment relative to the protein expression control on each blot. Percentage errors were calculated as the average variation between equivalent blots across several experiments. (C) Immunocytofluorescence of CXCR4 assessed by flow cytometry using a phycoerythrin-conjugated CXCR4 antibody for a representative case (shaded graph), with the isotype-control peak shown as a dotted line. (D) Median CXCR4 staining of CLL PBMCs from 10 cases was assessed with flow cytometry using a phycoerythrin-conjugated CXCR4 antibody (●) compared with isotype control (⃝). The median and range of each distribution are represented by horizontal lines. Strong CXCR4 expression is consistently observed, although a significant variation in median staining level was also seen.

  • Figure 3.

    BAT1 binds tightly to CXCR4, leading to dose-dependent inhibition of antibody binding and delivery of fluorescent cargo. (A) Phycoerythrin-CXCR4 antibody competition assay against BAT1 (red line) and plerixafor (blue line), assessed using flow cytometry. Primary CLL cells from a representative case were incubated with BAT1 or plerixafor at concentrations between 0 and 20 µM for 3 hours and then stained with phycoerythrin-conjugated CXCR4 antibodies. Experiment was performed in triplicate, and the mean was taken of the medians of each fluorescent distribution; errors were calculated as standard deviation of the mean. Dose-response curve fitted using a modified Hill equation, as detailed in supplemental Methods and data. A dose-dependent reduction in fluorescence was observed for BAT1 and plerixafor, with IC50(BAT1) = 138 nM and IC50(plerixafor) = 11.7 nM. (B) Analogous PerCP-CXCR7 antibody competition assay against BAT1 (red line) and plerixafor (black line). A dose-dependent reduction in antibody-staining was not observed for BAT1 or plerixafor. Fluorescence due to bound anti-CXCR7 antibody decreased at the highest concentrations of plerixafor or BAT1, but this decrease is within the measurement error, determined using the standard deviation of the mean. (C) Flow cytometric analysis of dose-dependent and selective BAT1-Cy5 targeting to CLL cells. Primary CLL cells were incubated with Cy5-conjugated BAT1 at 5 µM (solid red line) and 10 µM (solid blue line) for 3 hours. To demonstrate selectivity, the cells were also preincubated with 20 µM plerixafor as a comparison (dashed lines). BAT1-Cy5 was found to bind to CLL cells in a dose-dependent manner, and cells preincubated with 20 µM plerixafor present significantly reduced fluorescence, as indicated by the arrow. These data indicate that BAT1 can be used to specifically target functional molecules, such as dyes, to CXCR4-expressing cells.

  • Figure 4.

    BAT1 acts as a pure antagonist against CXCL12, reducing migration along the chemokine gradient and cell viability in vitro. (A) Fluorescence microscopy analysis of receptor redistribution following exposure to CXCL12, BAT1, or plerixafor. Primary CLL cells were incubated with vehicle, 20 µM BAT1, or 20 µM plerixafor for 3 hours before exposure to CXCL12 for 10 minutes. Control cells were incubated with vehicle alone (no CXCL12). Cells were then stained for membrane and cytoplasmic CXCR4 to assess its distribution. Unstimulated CLL lymphocytes (top row) express CXCR4 at the cell surface and in the cytoplasm. Binding of CXCL12 causes receptor internalization with intracellular redistribution, with possible trafficking of cytoplasmic CXCR4 to the membrane (second row). Plerixafor (third row) and BAT1 (bottom row) bind to surface CXCR4, blocking immunoreactivity, but they do not induce receptor internalization or redistribution. Original magnification ×600 (60× 1.4 N.A. objective lens). Scale bars, 20 µm. (B) Immunoblotting demonstrates BAT1’s antagonism of CXCL12-induced signaling. Primary CLL cells were incubated with BAT1, plerixafor, or vehicle for 3 hours before exposure to CXCL12 for 10 minutes. Lysates were prepared, and signaling levels were assessed by immunoblotting for ERK phosphorylation (p-ERK). Incubation with BAT1 led to a dose-dependent reduction in p-ERK, with saturating levels of BAT1 reducing p-ERK to similar levels as 20 µM plerixafor or no stimulation. Densitometric quantification of the blots is shown with adjustment relative to total ERK for each lane. Percentage errors were calculated based on the average variation between equivalent blots across several experiments. (C) Effects of BAT1 and plerixafor on cell viability were assessed over 24 hours using flow cytometry. Primary CLL cells were incubated with plerixafor (i) or BAT1 (ii) at a range of concentrations over 24 hours. A progressive decrease in cellular viability was demonstrated at concentrations exceeding 20 nM BAT1 or plerixafor. Experiments were performed in triplicate, and the mean was taken. Errors are the standard deviation of the mean, added in quadrature to the standard error on the gating. Curves were fitted using nonlinear regression with software’s in-built log(agonist) vs response curve with variable slope, based on the Hill equation. (D) Inhibition of chemotaxis in response to CXCL12 was assessed using a filter migration assay, where cell migration was assessed using flow cytometry. BAT1 significantly reduces chemotactic migration of CLL lymphocytes. No significant difference in migration levels was seen between the BAT1-treated cells and those treated with plerixafor or vehicle alone. ***P < .001, 1-way analysis of variance with the Holm-Sidak multiple-comparisons test. ns, not significant.

  • Figure 5.

    BAT1-cholesterol can be used to decorate fluorescently labeled liposomes, resulting in significantly higher uptake into cells. (A) Illustration of the prepared liposomes, indicating BAT1-cholesterol decoration of the fluorescently labeled liposomal membrane using the postinsertion method. (B) Liposome sizes were assessed using DLS. A representative line graph from the analysis is shown: intensity is related to particle size using the Mie scattering function. (C) CXCR4 antibody staining was used to assess the specificity of liposome binding. Cells were stained with phycoerythrin-CXCR4 antibodies following incubation with bare liposomes (black line) or BAT1-decorated liposomes (filled graph). In the presence of bare liposomes, ∼94% of the live cells are stained to a high level. In the presence of BAT1-labeled liposomes, ∼37% of cells are stained, and the median fluorescence is significantly lower. Cell staining reduction is indicated by the arrow. (D) Cell migration in response to CXCL12 was measured using a filter migration assay. Data shown are from 1 representative case. Experiments were performed in triplicate, means were taken across the triplicates, and errors were calculated as the standard deviation of the mean. BAT1-decorated liposomes inhibit migration along a CXCL12 gradient: no significant difference in migration inhibition is observed between BAT1-decorated liposomes and 20 μM plerixafor or the negative control. Further, no significant difference in migration was observed between cells incubated with bare liposomes and the positive control. **P < .01, 1-way analysis of variance with the Holm-Sidak multiple-comparisons test. (E) Flow cytometry was used to quantify levels of liposomal attachment and uptake. At only 30 µM liposomes, the cells incubated with decorated liposomes presented significantly higher levels of fluorescence (**P < .01, Student t test). (F) Laser deconvolution microscopy was used to show in which region of the cell the liposomes are located after a 3-hour incubation. Liposomes fluoresce in green, whereas nuclei are labeled with DAPI (blue). Images were prepared by building a z-projection (median intensity) from the image stack comprising the central region of the cell. The liposomes are observed throughout the cytoplasm. Scale bar, 50 µm.

  • Figure 6.

    BAT1 decoration of dox-liposomes leads to increased uptake in primary CLL cells, causing significantly increased rates of death compared with bare liposomes. (A) Schematic illustrating the postinsertion process for decorating dox-liposomes. (B) Primary CLL cells from 10 subjects were incubated with dox-liposomes, with and without BAT1 decoration. (Bi) Liposomal uptake and doxorubicin delivery were determined by assessing the doxorubicin-associated fluorescence in dead cells using flow cytometry. Each point represents the median fluorescence from an individual case. Median doxorubicin-associated fluorescence in dead CLL cells is significantly higher when CLL cells are incubated with BAT1-decorated dox-liposomes than with bare dox-liposomes. P = .0020, Wilcoxon matched-pairs signed-rank test. (Bii) Dox-liposome–associated death was measured in primary CLL cells across 10 cases using flow cytometry. The number of dead cells was calculated using forward/side scatter. BAT1-decorated liposomes consistently led to increased levels of cell death across all cases tested. P = .0039, Wilcoxon matched-pairs signed-rank test (**P < .01). (C) Cellular localization of the doxorubicin was assessed using laser deconvolution microscopy. Images were prepared by building a z-projection (intensity sum) from the image stack comprising the central region of the cell. Colocalization of the doxorubicin-associated red fluorescence and the DAPI staining was used as a measure of delivery to the cell nucleus. Doxorubicin is delivered to the cell nucleus in large quantities when cells are incubated for 3 hours with BAT1-decorated liposomes. Scale bar, 20 µm.