Enhanced Bone Marrow Homing of Natural Killer Cells Following mRNA Transfection With Gain-of-Function Variant CXCR4R334X

Abstract

Adoptive transfer of natural killer (NK) cells can induce remission in patients with relapsed/refractory leukemia and myeloma. However, to date, clinical efficacy of NK cell immunotherapy has been limited to a sub-fraction of patients. Here we show that steps incorporated in the ex vivo manipulation/production of NK cell products used for adoptive infusion, such as over-night IL-2 activation or cryopreservation followed by ex vivo expansion, drastically decreases NK cell surface expression of the bone marrow (BM) homing chemokine receptor CXCR4. Reduced CXCR4 expression was associated with dampened in vitro NK cell migration toward its cognate ligand stromal-derived factor-1α (SDF-1α). NK cells isolated from patients with WHIM syndrome carry gain-of-function (GOF) mutations in CXCR4 (CXCR4^R334X). Compared to healthy donors, we observed that NK cells expanded from WHIM patients have similar surface levels of CXCR4 but have a much stronger propensity to home to BM compartments when adoptively infused into NOD-scid IL2Rgamma^null (NSG) mice. Therefore, in order to augment the capacity of adoptively infused NK cells to home to the BM, we genetically engineered ex vivo expanded NK cells to express the naturally occurring GOF CXCR4^R334X receptor variant. Transfection of CXCR4^R334X-coding mRNA into ex vivo expanded NK cells using a clinically applicable method consistently led to an increase in cell surface CXCR4 without altering NK cell phenotype, cytotoxic function, or compromising NK cell viability. Compared to non-transfected and wild type CXCR4-coding mRNA transfected counterparts, CXCR4^R334X-engineered NK cells had significantly greater chemotaxis toward SDF-1α in vitro. Importantly, expression of CXCR4^R334X on expanded NK cells resulted in significantly greater BM homing following adoptive transfer into NSG mice compared to non-transfected NK cell controls. Collectively, these data suggest up-regulation of cell surface CXCR4^R334X on ex vivo expanded NK cells via mRNA transfection represents a novel approach to improve homing and target NK cell-based immunotherapies to BM where hematological malignancies reside.

Keywords: CXCR4; NK cells; WHIM; homing; immunotherapy; mRNA transfection.

Figure 1

Commonly used methods to manipulate and prepare NK cells for adoptive infusion decrease surface expression of CXCR4 and reduce their ability to migrate toward SDF-1α in vitro. Surface expression of CXCR4 was evaluated by flow cytometry. (A) CXCR4 surface expression on fresh T cells, B cells and NK cells among PBMC. Representative histograms are shown (one representative donor). (B) CXCR4 intensity, as displayed as GMFI relative to FMO, on different NK cell preparations (n = 10 donors). (C) In vitro transwell migration of NK cells from populations in (B) toward SDF-1α (n = 10 donors). Statistical significance was evaluated with the paired t-test, two-tailed. Bar graphs present the mean and error bars report the SEM. **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 2

NK cells from patients with WHIM syndrome have similar intensity of CXCR4 cell surface expression but have significantly superior in vivo bone marrow homing compared to NK cells expanded from healthy donors. (A) Surface expression intensity of CXCR4 on expanded NK cells from patients with WHIM syndrome (white, n = 3 WHIM pts) vs. healthy donors (black, n = 9 donors). (B) Experiment schema illustrating injection of expanded human NK cells i.v. into NSG mice followed by a harvest 24 h later. (C) Proportion of CD56⁺ human NK cells in BM samples of mice injected with ex vivo expanded NK cells from healthy donors or from WHIM syndrome patients 24 h following infusion (one representative mouse from each group). (D) The fraction of human CD56⁺ cells recovered from the BM of mice injected with human NK cells (WHIM n = 15, healthy donors n = 9, data reported from 3 independent experiments; each experiment is represented by either white, gray or black circles). Statistical significance was evaluated by the Welch's T-test, two-tailed. Graphs present mean and error bars report SEM. *p < 0.05, ****p < 0.0001.

Figure 3

Ex vivo expanded NK cells transfected with CXCR4^R334X mRNA have significantly upregulated surface expression of the CXCR4 receptor without impacting cellular viability. (A) Titration of increasing amounts of electroporated CXCR4^R334X mRNA results in increased cell surface CXCR4 expression. (B) NK cell viability over time following mRNA transfection (n = 10 healthy donors). (C) CXCR4 surface expression kinetics over time reported as fold increase in CXCR4 on expanded NK cells transfected with either CXCR4^R334X or CXCR4^WT mRNA compared to their non-transfected control counterpart (n = 10 donors). Statistical significance was assessed by the Wilcoxon sign-rank test. Graphs present mean and error bars report SEM. **p < 0.01.

Figure 4

Transfection of ex vivo expanded NK cells with CXCR4^R334X-coding mRNA but not with CXCR4^WT-coding mRNA improves in vitro chemotaxis of cells toward SDF-1α. (A) In vitro migration of NK cells toward increasing concentration of SDF-1α (n = 10 donors, 4 independent experiments). (B) Migration toward SDF-1α when NK cells were pretreated with 100 μM of plerixafor to block the CXCR4 receptor (n = 9 donors, 1 experiment). Statistical significance was determined with the Student's paired t-test, two-tailed. Graphs present mean and error bars in panel A report SEM whereas error bars in (B) report SD. **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 5

NK cells transfected with CXCR4 coding mRNAs have no alterations in NK cell inhibitory and activating receptors and have preserved anti-tumor cytotoxic function. (A) Inhibitory and activation NK cell receptor expression 24 h post-transfection with CXCR4-coding mRNAs compared to non-transfected control NK cells (n = 6 donors, 1 experiment). (B–D) NK cell intracellular TNF-α and IFN-γ, and surface CD107a expression following a 4-h co-culture with K562 cells (n = 6 donors, 1 experiment). (E) NK cell degranulation measured by NK cell surface upregulation of CD107a in response to co-culture with various tumor cell lines (n = 4 donors, 4 independent experiments). Statistical significance was determined with the Student's paired t-test, two-tailed. Graphs present mean and error bars report SEM.

Figure 6

Ex vivo expanded NK cells transfected with CXCR4^R334X-coding mRNA home significantly more efficiently to bone marrow compartments following adoptive infusion into NSG mice when compared to their non-transfected counterpart. (A) Experimental schematic illustrating the infusion of ex vivo expanded NK cells that have been transfected or not with CXCR4^R334X-coding mRNA. (B–E) The frequency of recovered CD56⁺ human NK cells from the BM, blood, lung, and liver 24 h after infusion of the cells. Each dot represents one animal (n = 10 mice/group). Statistics were evaluated by the Welch's t-test, two-tailed. Graphs present mean and error bars report SEM. *p < 0.05, **p < 0.01.

Figure 7

Ex vivo expanded NK cells transfected with CXCR4^R334X-coding mRNA have superior trafficking to bone marrow compartments 24 h following adoptive transfer compared to their mock transfected counterpart. (A) Experimental schematic illustrating in vivo trafficking of infused NK cells transfected with mRNA coding for luciferase and CXCR4^R334X or luciferase alone. (B) X-ray image of a mouse demonstrating ventral (top) and dorsal (bottom) anatomical views of NK cell distributions observed in (C). (C) Ventral (top) and dorsal (bottom) bioluminescence images of luciferase expressing NK cells 2 and 24 h after infusion into mice (n = 3 mice/group).