Patient-Specific Nanoparticle Targeting in Human Leukemia Blood

Abstract

Antibody-directed targeting of chemotherapeutic nanoparticles to primary human cancers holds promise for improving efficacy and reducing off-target toxicity. However, clinical responses to targeted nanomedicines are highly variable. Herein, we prepared and examined a matrix of 9 particles (organic and inorganic particles of three surface chemistries with and without antibody functionalization) and developed an ex vivo model to study the person-specific targeting of nanoparticles in whole blood of 15 patients with chronic lymphocytic leukemia (CLL). Generally, anti-CD20-functionalized poly(ethylene glycol) (PEG) nanoparticles efficiently targeted CLL cells, leading to low off-target phagocytosis by granulocytes and monocytes in the blood. However, there was up to 164-fold patient-to-patient variability in the CLL targeting. This was further exemplified through using clinically relevant PEGylated doxorubicin-encapsulated liposomes, which showed high interpersonal differences in CLL targeting (up to 234-fold differences) and off-target phagocytosis (up to 65- and 112-fold differences in granulocytes and monocytes, respectively). Off-target phagocytosis led to almost all monocytes being killed within 24 h of treatment. Variance of the off-target association of PEGylated liposomes with granulocytes and monocytes significantly correlated to anti-PEG immunoglobulin G levels in the blood of CLL patients. A negative correlation between CLL targeting of PEG particles and anti-PEG immunoglobulin M levels was found in the blood. Taken together, our study identifies anti-PEG antibodies as key proteins in modulating patient-specific targeting of PEGylated nanoparticles in human leukemia blood. Other factors, such as the antigen expression of targeted cells and fouling properties of nanoparticles, also play an important role in patient-specific targeting. The human leukemia blood assay we developed provides an ex vivo model to evaluate interpersonal variances in response to targeted nanomedicines.

Keywords: biomolecular coronas; human blood model; immunoglobulins; off-targeting cytotoxicity; particle−immune cell interactions; precision medicine.

Figure 1

Characterization of PEG, PEG-MS, Doxil nanoparticle systems. (a–c) TEM images of unfunctionalized PEG and PEG-MS nanoparticles and a cryo-TEM image of unfunctionalized Doxil nanoparticles. Scale bars are 500 nm (PEG and PEG-MS) and 100 nm (Doxil). (d–l) SIM images of the three particles types with and without functionalization of anti-PEG/anti-CD20 or anti-PEG/anti-CD28 BsAbs. Scale bars are 2 μm (PEG and PEG-MS nanoparticle systems) and 1 μm (Doxil nanoparticle system). (m) DLS and (n) diameter of PEG-MS and Doxil nanoparticles before and after functionalization with CD20 or CD28 BsAbs. The sizes of the PEG-MS and Doxil (with and without BsAb functionalization) nanoparticles were determined by DLS and presented as the mean ± standard deviation (SD) of three independent measurements. As PEG particles have negligible light scattering, their size was determined by SIM images and shown as the mean ± SD (n = 30). No significant difference in particle size before and after functionalization of BsAbs (two-way analysis of variance with Tukey’s multiple comparisons test) was observed.

Figure 2

Evaluation of cancer cell targeting in whole human blood model. (a) Schematic illustration of an ex vivo human blood model, which was established by spiking cancer cells in healthy human blood to assess nanoparticle cancer targeting in the presence of healthy blood immune cells. Created with BioRender.com. (b–g) Cancer cell targeting of nonfunctionalized and BsAb-functionalized PEG (b, e), PEG-MS (c, f), and Doxil (d, g) nanoparticles in cell culture media versus in whole human blood. Raji (b–d) or Jurkat (e–g) cells were spiked into cell culture media or fresh human blood from a healthy donor, followed by incubation with nanoparticles for 1 h at 37 °C and subsequent analysis by flow cytometry. Cell association (%) refers to the proportion of each cell type with fluorescence above background, stemming from fluorescence-labeled particles (see gating strategy in Figure S3 and S4). Cell association (%) data are shown as the mean ± SD of three independent experiments (using fresh blood from the same donor), with at least 400,000 leukocytes analyzed for each experimental condition studied. Cell only control groups represent the respective cell populations without particle incubation. NK, natural killer cell; DC, dendritic cell.

Figure 3

Cancer cell targeting of BsAb-functionalized PEG particles in the whole blood of CLL patients. (a) Schematic illustration of a whole human CLL blood model, where fresh whole blood from CLL patients was collected and subsequently incubated with nanoparticles to assess leukemia targeting in the presence of blood immune cells. Created with BioRender.com. (b) Flow cytometry histograms representing cell association of PEG, CD20-functionalized PEG (PEG-CD20), and CD-28-functionalized PEG (PEG-CD28) particles with targeted CLL and other blood immune cells in the blood from CLL patients. (c) CLL targeting of BsAb-functionalized PEG particles in the whole blood of 15 CLL patients (see Table 1 for patient details). Cell association (%) refers to the proportion of each cell type with fluorescence above background, stemming from fluorescence-labeled particles (see gating strategy in Figure S7). Cell association (%) data are shown as the mean ± SD of three independent experiments (using fresh blood from each donor), with at least 400,000 leukocytes analyzed for each experimental condition studied. Cell only control groups represent the respective cell populations without particle incubation. Patients are identified based on monocyte–PEG particle association (from low to high).

Figure 4

Confocal microscopy images showing the association of BsAb-functionalized PEG or PEG-MS nanoparticles with blood cells from a CLL patient. (a–c) PEG and (d–f) PEG-MS nanoparticles functionalized with or without anti-PEG/anti-CD20 or anti-PEG/anti-CD28 BsAbs were incubated with PBMCs from a CLL patient for 1 h at 37 °C, followed by phenotyping cells with antibody cocktails and imaging by confocal microscopy. Green, CD20+ CLL cells; blue, CD3+ T cells; yellow, CD14+ monocytes; red, fluorescence-labeled PEG or PEG-MS nanoparticles. Scale bars are 20 μm.

Figure 5

Cell association summary of PEG, PEG-MS, and Doxil nanoparticle systems (with and without BsAb functionalization) in the whole blood of 15 CLL patients. (a–c) Violin plots summarizing the association of nonfunctionalized and BsAb-functionalized PEG, PEG-MS, and Doxil nanoparticles with CLL cells, T cells, monocytes, granulocytes, and NK cells in CLL patient blood after 1 h incubation at 37 °C. (d) Cross-comparison of the three particle systems of their association with CLL cells, T cells, monocytes, and granulocytes in CLL patient blood. Each data point represents the mean of three independent experiments (using the same batch of fresh blood from each donor). The median cell association across 15 CLL patients is shown as a solid line in the violin plots.

Figure 6

Anti-PEG antibodies influence particle phagocytosis and CLL targeting in the blood of CLL patients. (a,b) Anti-PEG IgG levels in the plasma positively correlate with the phagocytosis of Doxil nanoparticles with and without BsAb (CD20 or CD28) functionalization by monocytes and granulocytes. (c) Anti-PEG IgM levels in the plasma negatively correlate with CLL targeting of CD20-functionalized PEG particles. Cell association (MFI) refers to the median fluorescence index of each cell type, stemming from fluorescence-labeled particles. The anti-PEG IgG and IgM levels in the plasma were determined by ELISA (Figure S11). Statistics were assessed by Spearman correlation analysis (n = 14).

Figure 7

Targeted and off-targeted cytotoxicity of nonfunctionalized and BsAb-functionalized Doxil nanoparticles to CLL and healthy immune cells in blood of CLL patients. (a) Flow cytometry histograms represent the cell death of CLL cells, T cells, monocytes, granulocytes, and NK cells after treating with Doxil nanoparticles in the blood of a CLL patient (Patient 2). (b) Cytotoxicity and (c) cell association of nonfunctionalized and BsAb-functionalized Doxil nanoparticles in the whole blood of four CLL patients. The nonfunctionalized and BsAb-functionalized Doxil nanoparticles were incubated with the whole blood of CLL patients for 24 h at 37 °C, after which the cells were phenotyped with antibody cocktails and dead cells were labeled with SYTOX AADvanced dead cell stain. Data in (b,c) are shown as the mean ± SD of three independent experiments (using fresh blood from each donor), with at least 400,000 leukocytes analyzed for each experimental condition studied.