Nanoengineered Therapeutic Stealth Cells

Abstract

Disclosed herein is a method to "reprogram" highly motile cells found in tumors, such as these highly motile GSC and / or MDSC clones, into "self-destructive" cellular "missiles" (referred to herein as therapeutic stealth cells) that can seek out and destroy new recurrent foci in the body, such as the brain. Cells with enhanced motility can be selected from a heterogeneous population and then rendered "self-destructive" by the decisive delivery of anti-cancer drugs, such as oncolytic viral plasmid cocktails.

Description

[Background technology]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims priority to U.S. Provisional Patent Application No. 62 / 747,980, filed October 19, 2018, which is incorporated by reference in its entirety.

[0002] Tumor cell metastasis is the major cause of cancer-related deaths (>90%) (Gallego-Perez, D. et al. Lab Chip 12:4424-4432 (2012); Fidler, IJ Nat Rev Cancer 3:453-458 (2003); Gupta, GP & Massague, J. Cell 127:679-695 (2006)). In particular, glioblastoma multiforme (GBM) is a highly aggressive and lethal type of brain tumor (Bellail, AC, et al. Int J Biochem Cell Biol 36:1046-1069 (2004)). This aggressive tumor exhibits a distinct intracranial spreading pattern and effectively metastasizes as single cells along pre-aligned white matter tracts (Gallego-Perez, D. et al. Lab Chip 12:4424-4432 (2012); Bellail, A. C., et al. Int J Biochem Cell Biol 36:1046-1069 (2004)). Increasing evidence indicates that the invasive phenotype of GBM is regulated by cell motility (Giese, A., et al. J Clin Oncol 21:1624-1636 (2003)). Moreover, recurrence appears to be driven primarily by a subset of highly motile tumor-initiating cells, known as glioma stem cells (GSCs), that are resistant to conventional therapies (Calabrese, C. et al. Cancer Cell 11:69-82 (2007); Ghotra, VP, et al. Int J Radiat Biol 85, 955-962 (2009)). As GSCs continue to attract great interest from the scientific and medical communities, new analytical and engineering tools are needed to better understand and counteract the mechanisms by which GSCs spread and generate new tumor foci in the brain. However, research on GSC motility and therapeutic resistance has been limited compared to ongoing efforts on oncogenic transformation. This is in part due to the lack of effective tools to identify, study, and manipulate specific subsets of GSCs, or other cells of interest from the GBM niche, for research, diagnostic, and / or therapeutic purposes.Identifying tumors at the single clone level by in vivo imaging is extremely challenging (Irimia, D. & Toner, M. Integr Biol (Camb) 1:506-512 (2009); Condeelis, J. & Segall, JENat Rev Cancer 3:921-930 (2003)). Furthermore, current techniques for ex vivo analysis of tissue explants tend to be laborious and limited (Johnson, J. et al. Tissue Eng Part C Methods 15:531-540 (2009)). On the other hand, traditional in vitro assays (Boyden, SJ Exp Med 115, 453-466 (1962); Albini, A. & Benelli, R. Nat Protoc 2, 504-511 (2007); Rao, JS Nat Rev Cancer 3, 489-501 (2003); Yamada, KM & Cukierman, E. Cell 130, 601-610 (2007); Liang, CC, Park, AY & Guan, JL Nat Protoc 2, 329-333 (2007)) are endpoint tests that are not physiologically relevant and / or focus only on the bulk behavior of highly heterogeneous cell populations. Summary of the Invention [Problem to be solved by the invention]

[0003] There are subsets of GSCs and MDSCs that display enhanced metastatic and therapeutic resistance. These findings suggest that GSCs and MDSCs are not monolithic populations, but that specific clonal subsets display a significantly more "aggressive" phenotype, possibly contributing to driving disease recurrence. [Means for solving the problem]

[0004] Disclosed herein is a method to "reprogram" highly motile cells found in tumors, such as these highly motile GSC and / or MDSC clones, into "self-destructive" cellular "missiles" (referred to herein as therapeutic stealth cells) capable of searching for and destroying new recurrent foci in the body, such as the brain. Cells with enhanced motility can be selected from the heterogeneous population and then rendered "self-destructive" by the decisive delivery of anti-cancer drugs, such as oncolytic viral plasmid cocktails.

[0005] The disclosed methods can include sorting cells from a subject to generate therapeutic stealth cells. In some embodiments, the cells are autologous, such as blood cells or a tumor biopsy from the subject to be treated. In some cases, however, the cells are allogeneic.

[0006] The disclosed methods may include sorting cells from a subject for a highly motile subpopulation and then reprogramming the subpopulation to deliver an anti-cancer drug. In some embodiments, the subpopulation may be sorted in a migration assay using a chemoattractant gradient. In particular, the chemoattractant gradient may include chemokines produced by the tumor being treated. For example, in some embodiments, the chemoattractant comprises Matrigel®. In some embodiments, the chemoattractant comprises tumor cell conditioned medium.

[0007] In some embodiments, the subpopulations are sorted in migration assays using nanotextured and / or biomimetic surfaces. For example, MDSCs can respond to and be guided along pre-aligned structural cues in the absence of biochemical stimuli. Thus, in some embodiments, the surface comprises micro- or nanoscale protrusions / grooves. For example, the depth and width of the protrusions / grooves can have dimensions between 100 nm and 10 μm, including 100 nm to 1 μm, 1 μm to 10 μm, 500 nm to 5 μm. The protrusions / grooves can have a variety of shapes and patterns, including straight grooves.

[0008] In some embodiments, the subpopulations are sorted in a transwell migration assay or a cell invasion assay. The transwell migration assay measures the number of cells passing through a porous membrane, while the cell invasion assay focuses on invasive cell migration through the extracellular matrix.

[0009] Once the subpopulation has been selected and optionally expanded, the cells can then be reprogrammed to heterologously express a transgene encoding an anti-tumor protein, an oligonucleotide, or a combination thereof.

[0010] The introduction of an efficient "safety switch" can be used to reduce the risk of severe graft-versus-host disease in some cases. Thus, in some embodiments, the subpopulation is also reprogrammed with a kill switch system. The most widely studied safety switch to date is the HSV I-derived thymidine kinase (HSV-TK) gene product. Non-immunogenic safety switch systems have also been developed, which include fusion proteins composed of human pro-apoptotic molecules (e.g., caspase 9) linked to modified human FK506 binding proteins (i.e., iCasp9). These safety switches can be activated by injection of a chemical dimerization inducer (CID) consisting of a dimer of two synthetic mutants of FK506. Other inducible self-destructive kill switches are in development and can be used in the disclosed therapeutic stealth cells.

[0011] Also disclosed is a composition comprising a plurality of therapeutic stealth cells produced by the disclosed methods. In certain embodiments, the composition further comprises a pharma- ceutically acceptable excipient.

[0012] Also disclosed is a method for treating a tumor in a subject, comprising administering to the subject an effective amount of the disclosed pharmaceutical composition. The disclosed method can be used to treat any solid tumor. In certain embodiments, the tumor matches the cell source used to generate therapeutic stealth cells. For example, highly motile MDSCs obtained from breast cancer biopsy can be reprogrammed to treat breast cancer. Similarly, highly motile GSCs / MDSCs obtained from glioblastoma multiforme (GBM) biopsy can be reprogrammed to treat GBM.

[0013] The details of one or more embodiments of the invention are set forth in the accompanying drawings and the description below. Other features, objects, and advantages of the invention will become apparent from the description and drawings, and from the claims. [Brief description of the drawings]

[0014] [Figure 1A] Figures 1A-1E show the results of migration-based sorting / "chromatography". Figure 1A shows that nanotextured surfaces induce guided motility. Clones with high motility are attracted into the collection chamber by chemoattraction. Figure 1B shows that studies with GSCs reveal that highly motile clones were resistant to anti-miR363 therapy. Figure 1C shows that migration-based sorting of MDSCs revealed clonal subsets with superior motility compared to bulk MDSCs. Figure 1D shows that sorting revealed diverse phenotypes with granulocytic (P4) and monocytic (P5) subtypes. There was also an unclassified subtype showing low (P6) Ly6-C / G or high (P3) Ly6-C / G. Figure 1E shows that MDSC clones with high motility were either P4 or P3. *p<0.05. [Figure 1B] Same as above. [Figure 1C] Same as above. [Figure 1D] Same as above. [Figure 1E] Same as above. [Diagram 2]FIG. 2 illustrates the intracranial injection of reprogrammed / controlled GSCs and / or MDSCs into GBM-bearing mice and monitoring tumor progression. [Diagram 3] Figure 3 shows that MDSCs exhibit significant (i.e., guided) motility on textured / biomimetic surfaces, whereas MDSCs cultured on TCPs exhibit limited motility. These results suggest that, like tumor cells, MDSCs may respond to the same structural cues that enhance tumor cell metastasis in vivo. *p<0.05. [Figure 4A-B] Figures 4A and 4B show a migration chromatography setup where MDSCs are selectively seeded on one side of the platform and allowed to migrate unidirectionally by chemotaxis. Surface nanotexture triggers clonal separation based on guided motility. Figure 4C shows the speed of fast vs. slow migrating cells. Figures 4D and 4E include flow cytometry analysis showing that fast migrating clones have a distinct phenotype compared to slow migrating clones (e.g., bulk MDSCs). *p<0.05. [Figure 4C] Same as above. [Figure 4D] Same as above. [Figure 4E] Same as above. [Figure 5A]Figures 5A and 5B show single clone motility assays of circulating MDSCs from melanoma patients. Differences in speed (Figure 5A) and effective displacement (Figure 5B) for each patient. Results show that MDSCs from certain patients show enhanced speed. However, when considering effective displacement (i.e., geometric distance from start to end position), certain MDSC batches with low speed show significant displacement, which may reflect higher directional / sustained motility (no chemotaxis). P1: stage IIIC, tx nivolumab + surgery; P2: stage IV, tx nivolumab; P3: stage IV V600E / BRAF, tx radiation + pembrolizumab; P4: stage IV, tx nivolumab + ipilumamab; P5: stage IV, tx pembrolizumab / ipilumamab / nivolumab; P6: stage IV V600E / BRAF, treated with INF-α). [Figure 5B] Same as above. [Figure 6A] Figures 6A and 6B show that nanotextured surfaces can be used to reveal drug sensitivity not observed with standard TCP. Figure 6A shows that single clone motility assays of patient-derived MDSCs show inhibition of specific clonal subsets (mean velocity >40 μm / h) in response to ibrutinib. Figure 6B shows that single clone motility on TCPs did not reveal an effect of ibrutinib on MDSC metastasis. [Figure 6B] Same as above. [Figure 7A] Figures 7A and 7B show a migration chromatography device with integrated microfluidics to enable automated detachment of clones of interest. Figures 7C and 7D show that once migration-based separation has occurred, the intrinsic microfluidic system can be used to continuously flush cold water into place, and thermal actuation of the PINIPAM layer promotes selective detachment of MDSC clones of interest. [Figure 7B] Same as above. [Figure 7C] Same as above. [Figure 7D] Same as above. [Figure 8A] Figure 8A shows that co-culturing MDSCs with non-cancerous MCF10A enhanced motility in the MCF10A clone population. Figure 8B shows that immunofluorescence analysis shows that the co-culturing conditions caused a decrease in the expression of epithelial markers such as E-cadherin and an increase in mesenchymal markers (e.g., vimentin) in certain clones. Figure 8C shows that co-culturing MDSCs with already highly malignant MDAMB-231 cells did not significantly alter motility in the MDA-MB-231 population. ●: culture alone, ■: 50:50 co-culture, ▲: 90:10 co-culture (MDSC: breast cancer / tissue cells). [Figure 8B] Same as above. [Figure 8C] Same as above. [Figure 9] Figure 9 shows that co-culture of MDSCs with breast tissue / cancer cells resulted in a significant increase in the velocity of certain MDSC clones, especially when co-cultured with MDA-MB-231. These results potentially suggest that MDSC motility is up-regulated in the presence of metastatic cells, facilitating co-metastasis outside the tumor and continued immune protection during the tumor cell metastatic process. [Figure 10A-B]Figures 10A-10E show that MDSCs respond to aligned structural cues and display guided metastatic patterns. Figure 10A is a schematic of the tumor microenvironment showing invasive cancer cells and invasive MDSCs using pre-aligned structural cues (e.g., remodeled ECM, vascular wall) to evade and invade the tumor stroma, respectively. Figure 10B is an SEM micrograph (overlaid with MDSC mockup) of a PDMS-based biomimetic textured surface used to evaluate structurally guided MDSC migration at the single clone level. Figure 10C shows actin-nuclear staining of MDSCs cultured on textured vs. control / TCP surfaces. MDSCs adopt an aligned / more migratory morphology on textured surfaces compared to TCP. Figures 10D and 10E show single clone metastatic tracks (Figure 10D) and quantification of MDSCs (Figure 10E) on textured vs. control / TCP surfaces, confirming the enhanced metastatic potential of MDSCs (i.e., mean single clone velocity and net track distance) when exposed to pre-aligned structural cues. Net track distance reflects the geometric distance traveled by cells during the tracking period. *p<0.01 and ‡p<0.02 (t-test, n=4). [Figure 10C] Same as above. [Figure 10D] Same as above. [Figure 11A-B]Figures 11A-11I show that MDSC subpopulations display distinct migration and gene expression patterns. Figures 11A and 11B are schematics of the experimental design. Here, MSC-2 cultures were sorted by flow cytometry into three different subpopulations, including granulocytic MDSCs (CD11b+Ly6CloLy6G+) and monocytic MDSCs (CD11b+Ly6ChiLy6G-), as well as CD11b+Ly6C+Ly6G+ cells. Each population was then subjected to single-clone motility assays on textured PDMS and qRT-PCR analysis of pro- and anti-inflammatory markers. Figure 11C shows actin-nuclear staining of different MSC-2 subtypes cultured on textured surfaces. Granulocytic MDSCs tended to display a more aligned and migratory-prone morphology compared to their counterparts. Figure 11D shows the quantification of single-clone migration (i.e., mean speed and net track distance) of each subtype on textured surfaces. *p=0.006, **p<0.001, ψp=0.001, ‡p=0.09 (2-way ANOVA, n=4). Figure 11E shows single clonal tracks for each population. Figure 11F shows fluorescently labeled flow-sorted MDSCs vs "new" / unsorted MDSCs injected (i.e., via tail vein) into tumor-bearing mice (i.e., orthotopic breast cancer developed from human cells in nude mice). Pictures on the right show tumor progression / growth from 1 to 4 weeks. Figure 11G shows tumors and other target organs imaged to detect the extent of MDSC infiltration 24 hours after injection. Figures 11H and 11I show qRT-PCR analysis of pro-inflammatory genes (Figure 11H) and anti-inflammatory genes (Figure 11I) for each subtype. *p<0.001, **p<0.0001, ‡p=0.03 (two-way ANOVA, n=3-4). [Figure 11C] Same as above. [Figure 11D] Same as above. [Figure 11E] Same as above. [Figure 11F] Same as above. [Figure 11G] Same as above. [Figure 11H] Same as above. [Figure 11I] Same as above. [Figure 12A]Figures 12A-12G show that a single MDSC subpopulation appears to exhibit phenotypic plasticity capable of driving recruitment across the entire phenotypic range. Figure 12A is a schematic of the experimental design. Figure 12B shows that single clonal migration (i.e., mean velocity and net track distance) studies showed no significant differences in all three populations by 7 days. Figures 12C-12E show that flow cytometry analysis suggests that all subpopulations remained relatively pure up to 1 day after sorting, but by 7 days, the full range of phenotypes was recruited regardless of the starting cell population. *p<0.0001, ‡p=0.01, #p=0.03, ψp=0.0001 (2-way ANOVA / Tukey's multiple comparisons, n=3-4). Figures 12F and 12G show qRT-PCR analysis of pro-inflammatory (Figure 12F) and anti-inflammatory (Figure 12G) genes at 7 days after sorting. *p=0.006, **p=0.01 (two-way ANOVA / Tukey's multiple comparisons, n=3~6). [Figure 12B] Same as above. [Figure 12C] Same as above. [Figure 12D] Same as above. [Figure 12E] Same as above. [Figure 12F] Same as above. [Figure 12G] Same as above. [Figure 13A] Figures 13A and 13B show that circulating MDSCs from melanoma patients show distinct metastatic profiles at the single clone level. Figures 13A and 13B show that the mean single clone velocity (Figure 13A) and net track distance (Figure 13B) tended to be significantly higher for certain patients compared to the rest of the patient population, which may reflect the patient background. [Figure 13B] Same as above. [Figure 14A]Figures 14A-14F show that different subpopulations of patient-derived MDSCs display distinct metastatic potential. Figures 14A-14C show that melanoma patient MDSCs were sorted by flow cytometry into granulocytic (CD11b+CD15+CD14-) and monocytic (CD11b+CD15-CD14+) subpopulations. Figures 14D-14F show that similar to the observations in mouse MDSCs, the granulocytic subpopulation of patient-derived MDSCs also displays increased metastatic (i.e., mean single clone velocity and net track distance) potential compared to the monocytic subtype. *p=0.0005, **p=0.002 (Mann-Whitney, n=3). [Figure 14B-C] Same as above. [Figure 14D] Same as above. [Figure 14E] Same as above. [Figure 14F] Same as above. [Figure 15A] Figures 15A and 15B show differential gene expression of pro-inflammatory (Figure 15A) and anti-inflammatory (Figure 15B) markers as a function of time for flow cytometry sorted subpopulations. *p<0.0001, ‡p=0.06 (2-way ANOVA / Sidak's multiple comparisons, n=3-6). [Figure 15B] Same as above. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0015] Before describing the present disclosure in further detail, it is to be understood that this disclosure is not limited to particular embodiments described, as such may, of course, vary. It is also to be understood that the terminology used herein is for the purpose of describing particular embodiments only, and is not intended to be limiting, since the scope of the present disclosure will be limited only by the appended claims.

[0016] Where a range of values ​​is stated, it is understood that each intervening value between the upper and lower limit of that range, to the tenth of the unit of the lower limit unless the context clearly dictates otherwise, and any other stated or intervening value in the stated range, is included within the disclosure. The upper and lower limits of these smaller ranges may independently be included in the smaller ranges, which are also included within the disclosure, unless a specifically excluded limit in the stated range is present. Where the stated range includes one or both of the limits, ranges excluding either or both of those included limits are also included in the disclosure.

[0017] Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure belongs. Although any methods and materials similar or equivalent to those described herein can be used in the practice or testing of this disclosure, the preferred methods and materials are described herein.

[0018] All publications and patents mentioned in this specification are incorporated by reference as if each individual publication or patent was specifically and individually indicated to be incorporated by reference herein, and are incorporated by reference herein to disclose and describe the methods and / or materials in connection with the publications mentioned. The reference to any publication is for its disclosure prior to the filing date and should not be construed as an admission that the present disclosure is not entitled to antedate such publication by prior disclosure. Further, the publication dates provided may be different from the actual publication dates which may need to be independently confirmed.

[0019] As will be apparent to one of ordinary skill in the art upon reading this disclosure, each of the individual embodiments described and illustrated herein has separate components and features which may be readily separable from or readily combinable with any of the features of the other several embodiments without departing from the scope or spirit of the disclosure. Any recited method may be carried out in the order of events recited or in any other order which is logically possible.

[0020] The embodiments of the present disclosure employ, unless otherwise indicated, chemical, biological and other techniques within the skill of the art.

[0021] The following examples are provided to provide one of ordinary skill in the art with a complete disclosure and description of how to practice the methods and use the probes disclosed and claimed herein. Efforts have been made to ensure accuracy with respect to numbers (e.g., amounts, temperatures, etc.), but some errors and deviations may occur. Unless otherwise indicated, parts are parts by weight, temperature is in °C, and pressure is at or near atmospheric. Standard temperature and pressure are defined as 20 °C and 1 atmosphere.

[0022] Before describing embodiments of the present disclosure, it is to be understood that, unless otherwise indicated, this disclosure is not limited to particular materials, reagents, reaction materials, manufacturing processes, etc., as these may vary. It is also to be understood that the terminology used herein is for the purpose of describing particular embodiments only, and is not intended to be limiting. The present disclosure also allows for steps to be executed in different order, where this is logically possible.

[0023] Please note that as used in the specification and the appended claims, the singular forms "a," "an," and "the" include plural referents unless the context clearly dictates otherwise.

[0024] "Suicide genes", as used herein, refer to genes that cause cells to commit suicide by apoptosis. Activation of these genes can result from many processes, but the main cellular "switch" that induces apoptosis is the p53 protein. Stimulation or introduction (by gene therapy) of suicide genes can be used to treat cancer or other proliferative diseases by making cancer cells more vulnerable and more sensitive to chemotherapy. Some of the genes expressed in cancer cells are linked to other genes for enzymes not found in mammals that can convert harmless substances into substances that are toxic to the tumor. The suicide genes that mediate this sensitivity can code for viral or bacterial enzymes that convert inactive drugs into toxic antimetabolites that inhibit the synthesis of nucleic acids.

[0025] Highly motile GSC and / or MDSC clones can be selected from heterogeneous populations by migration-based sorting, such as nanotip-supported single clone motility chromatography. Migration-based cell sorting methods involve identifying clonal subsets and / or cell subpopulations that exhibit enhanced metastatic potential compared to the rest of the population. Such cells are inherently more prone to homing / invasion and / or metastatic growth into primary tumors and therefore may serve as more efficient drug / gene delivery vehicles. Identifying highly metastatic clonal subsets can be accomplished in a variety of ways.

[0026] One option is to seed cell mixtures in lines onto micro- or nano-textured surfaces, which induces contact-guided directional migration of cells. For example, MDSCs can respond to and be guided along pre-aligned structural cues in the absence of biochemical stimuli.

[0027] Cells can be exposed to a chemoattractant gradient that defines a particular direction in which the cells migrate, and "fast-moving" clones can be gradually collected in a container as they migrate toward the chemoattractant. Performing this sorting in the absence of chemoattractant can also be used as a way to identify clonal subsets that may be more likely to exhibit unidirectional motility (i.e., toward the collection container) even in the absence of chemoattractant. Even if these cells are not necessarily the fastest movers, their ability to exhibit sustained motility in a single direction can translate into enhanced ability to metastasize in vivo (e.g., cells with high migration speed but lacking directionality may not necessarily be the most effective "infiltrators"), which also makes these clones desirable for enhanced drug / gene delivery to primary tumors and / or metastatic growth.

[0028] Another method for selecting cells with enhanced metastatic potential can be by translocation assay in a transwell system (e.g., 8 micron pores). For example, cells can be seeded onto the top chamber of a transwell, and cells with enhanced metastatic potential will gradually translocate across the pores to the bottom chamber, where they can be collected for further modification (for gene / drug delivery applications).

[0029] As disclosed herein, the disclosed subpopulation of cells is CD11b + Ly6C lo Ly6G + Myeloid-derived suppressor cells. Thus, in some embodiments, the highly motile cells are obtained by cell sorting of tumor-derived GSCs and / or MDSCs using a combination of antibodies that selectively bind to CD11b, Ly6C, and Ly6G.

[0030] In some embodiments, the cells are derived from primary tumor cells (e.g., isolated from routine biopsies). In some embodiments, the cells are derived from myeloid-derived suppressor cells (e.g., isolated from the circulation). However, the disclosed methods can be applied to any other cell type (e.g., other monocytes, T cells, etc.) that is prone to infiltrating cancerous tissues.

[0031] Once pre-selection of highly disseminated clones is complete, these cells can be first expanded and then genetically engineered via various routes, including viral or non-viral (e.g., bulk electroporation, tissue nanotransfection) delivery of transgenes, and / or CRISPR / CAS9-driven transgene insertion. The goal of this step is to induce the production of anti-tumor proteins, oligos, and / or other entities (e.g., glut1, mir146, oncolytic viruses, etc.) by these cells. Once genetic engineering of these highly motile subpopulations is complete, these cells can then be returned to the patient either systemically (e.g., in the blood, lymphatic system) or locally (to the primary or metastatic tumors) with the goal of eradicating the cancerous growth. In some embodiments, the transgene encodes tissue inhibitor of metalloproteinase-3 (TIMP-3).

[0032] For cancer applications, these cells can be engineered (via transfection) to express pro-inflammatory molecules (e.g., ccl4, mir146, glut1) to promote T cell infiltration into tumors, or anti-metastatic components (e.g., timp3) to prevent cancer metastasis.

[0033] In some embodiments, MDSCs are used to provide therapy for other conditions such as Alzheimer's disease or diabetes that deliver anti-inflammatory molecules, or other forms of brain injury (e.g., ischemic stroke), where MDSCs can naturally home in and deliver therapeutic cargo, such as pro-angiogenic and / or proneural, or anti-inflammatory agents.

[0034] These autologous cells can be further engineered (before being injected back into the patient) with a drug-inducible (e.g., doxycycline) "kill switch" system to eradicate the therapeutic cells once their action is no longer required. Kill switch systems are known in the art, and therefore it is within the purview of one of ordinary skill in the art to select and employ a suitable kill switch system.

[0035] Although a number of embodiments of the invention have been described, it will be understood that various modifications can be made without departing from the spirit and scope of the invention. Accordingly, other embodiments are within the scope of the following claims. EXAMPLES

[0036] Example 1: Nanoengineered Therapeutic Stealth Cells The metastatic potential of GSCs and their ability to evade the immune system or standard treatments remain major lethal factors. Using nanoscale tools, we isolated and studied a specific subset of GSCs that exhibits high metastatic and treatment-resistant potential (Figure 1).

[0037] Pilot studies with MDSCs also revealed clonal subsets with significant metastatic potential, similar to GSCs (Figure 1). These findings suggest that GSCs and MDSCs are not monolithic populations, and that certain clonal subsets display significantly more "aggressive" phenotypes that may possibly be responsible for causing disease recurrence. Disclosed in this example is the development of a novel high-risk / high-reward approach to minimize GBM recurrence by reprogramming highly motile GSC and / or MDSC clones into "self-destructive" cellular "missiles" capable of searching for and destroying new recurrent lesions in the brain. Motility-enhanced GSCs and MDSCs are selected from heterogeneous populations by nanotip-supported single-clone motility "chromatography" as shown in Figure 1. These cells are then subjected to limited clonal expansion (passages 2-5) and subsequently rendered "self-destructive" by definitive nanochannel-based delivery of oncolytic viral plasmid cocktails (Gallego-Perez, D. et al. Nanomedicine 12:399-409 (2016)). These viruses are able to kill cancer cells by mechanisms distinct from conventional therapies and have therefore been proposed as a promising therapeutic option to eradicate GSCs (Cripe, TP, et al. Mol Ther 17:1677-1682 (2009)).

[0038] A series of in vitro studies are performed to determine the optimal plasmid dose and ratio at which selected subgroups of GSCs or MDSCs become "self-destructive" while retaining good motility over time. These "controlled" but highly motile GSC / MDSC populations are then injected intracranially (together and separately) into GBM-bearing mice with the goal of effectively transferring and strategically releasing therapeutic virions throughout the diseased brain (Figure 2).

[0039] Also, comparative experiments of systemic delivery of reprogrammed / drugged MDSCs will be performed in GBM-bearing mice to verify that these cells can home to the diseased brain and disrupt tumor progression. State-of-the-art imaging techniques (e.g., IVIS, PET, MRI) will be used to monitor the outcome of therapeutic GSCs / MDSCs. Although cell-based oncolytic virus therapy has previously shown promising results compared to direct treatment with oncolytic virus particles (Power, AT& Bell, JCMol Ther 15:660-665(2007)), the main limitation is that most of the cells studied so far have reduced metastatic potential, especially when compared to the intracranial metastatic pace of GSCs. On the other hand, controlled / reprogrammed GSCs or MDSCs may have inherently high intracranial motility in addition to their stealth ability against the immune system, thus allowing them to colonize and monitor and treat diseased brains more effectively. Example 2: Structurally guided translocation of mouse MDSCs

[0040] Recent studies have shown that MDSCs can respond to and be guided along pre-aligned structural cues in the absence of biochemical stimuli (Figures 3-4). In contrast, single clone motility assays on tissue culture polystyrene (TCP) revealed low metastatic potential (Figure 3), suggesting that, similar to tumor cells, MDSC metastasis / invasion is probably favored by structurally guided migration. Chip-supported migration chromatography studies revealed a clonal subset with enhanced metastatic potential compared to the rest of the population, corresponding to high-grade cancerous cells (Figures 4A-4C). Flow cytometric analysis of fast-migrating clones revealed that such populations were predominantly Ly6-G high / Ly6-C low (granulocytes) and Ly6-G high / Ly6-C high The slow-migrating clones were phenotyped as monocytic (Ly6-G low / Ly6-C high) and granulocytes, as well as the unidentified variant Ly6-G low / Ly6-C low and Ly6-G high / Ly6-C high MDSCs were more evenly distributed among the MDSCs. Thus, MDSCs apparently have specialized clonal subsets with improved metastatic potential, which are presumably more prone to colonize tumors / ganglia and exert immunosuppression. Such clonal subsets may therefore be new therapeutic targets in the fight against cancer. Example 3: Structurally guided transfer of patient MDSCs

[0041] We next tested whether patient-derived MDSCs also exhibited structurally guided migration in the absence of biochemical stimuli. MDSCs isolated from peripheral blood of melanoma patients at different stages under different treatments were followed for approximately 24 hours. Each patient's MDSCs displayed a unique metastatic pattern / signature, with some patients exhibiting clonal subsets with enhanced mobility compared to the bulk population (Figure 5), which remained clustered at less than 25 μm / h. Notably, some patients had MDSCs that were clustered overall with velocities less than 25 μm / h, indicative of a seemingly "quiescent" population, possibly reflecting the type of malignancy and / or the modality / stage of therapy. Although further studies are required to establish clear correlations between these factors and single clonal metastatic potential / signatures of MDSCs, which may serve as surrogates for their levels of in vivo activity and / or disease outcome, these data further support the idea that MDSCs are not monolithic, and that investigating mechanisms related to motility could potentially pave the way for the development of new diagnostic / prognostic tools as well as improved therapies. Example 4: Structurally guided migration reveals drug sensitivity

[0042] Impairing MDSC migration / infiltration into tumors / ganglia may be a viable strategy to reduce the immunosuppressive burden. Inhibitors of Bruton's tyrosine kinase (BTK) have typically been used in the treatment of hematological cancers. BTK plays a role in many biological processes, including cell migration. Although MDSCs express BTK, single-clone motility assays on TCPs in the presence of ibrutinib (a BTK-inhibitor) showed no significant effect on migration (Figure 6). In contrast, motility assays on biomimetic surfaces seem to indicate selective targeting of highly motile subsets of MDSCs (Figure 5). Such results indicate that migration-driven changes (e.g., cytoskeleton alignment) may in part regulate drug sensitivity in MDSCs. Example 5: A biomimetic platform for migration chromatography

[0043] Utilizing in-house nanofabrication expertise (i.e., contact / projection-based lithography, and soft lithography), we fabricate pre-aligned structural cues (~300 nm wide) from polydimethylsiloxane (PDMS). The textured surface is then functionalized with thermoresponsive poly(N-isopropylacrylamide) (PNIPAM) under argon plasma (30 watts, ~1000 microTorr). The PINIPAM-coated substrate (~100 μm thick) is then coupled to a microfluidic system (50 μm wide, 500 μm pitch, independently operated, Figure 7) with aligned microchannels. These channels are used to selectively flow cold water under the textured PDMS to facilitate selective thermally activated cell detachment from the PINIPAM (i.e., switching from weakly hydrophobic to highly hydrophilic). The nanotextured surface is characterized by scanning electron microscopy (SEM) and atomic force microscopy (AFM). The PNIPAM coatings are characterized by contact angle measurements at different temperatures, X-ray photoelectron spectroscopy (XPS) and Fourier transform infrared spectroscopy (FTIR). Example 6: MDSC Motility

[0044] MDSCs are isolated from freshly obtained tissues (i.e., peripheral blood, tumor and lymphoid tissues) of breast cancer tumor patients under protocol OSU-09142 using standard procedures. Tumor cells / tissues are also collected using standard procedures36. Migration chromatography is performed on biomimetic surfaces (Figures 4, 7) using GM-CSF (200 ng / mL) as a chemoattractant. Single clonal migration of different source MDSCs (i.e., circulating vs tumor vs lymphoid tissue-resident) is recorded by time-lapse microscopy on a confocal microscope fitted with a culture chamber. Images are collected every 10 min for 24-72 h and post-processed / analyzed using a manual tracker plugged into Fiji. MDSC clones exhibiting different degrees of motility are isolated by selectively "manipulating" the microchannels of the platform. Cells are segmented as highly motile vs intermediately motile vs lowly motile depending on the distance traveled from the starting position (Figure 7). The biomechanics (i.e., stiffness and contractility) of these clonal subsets will then be analyzed by vibration AFM, a technique developed by co-I Ghadiali to analyze the viscoelastic properties of single cells, and traction force microscopy (TFM), as described below. Additionally, flow cytometry will be used to analyze the viscoelastic properties of monocytes (CD15 + / CD14 + ) and granulocytes (CD15 + / CD14 - ) markers, and their combinations. To evaluate the immunosuppressive activity in each clonal subset, they are cultured with CFSE-labeled T cells (at different concentrations) and T cell proliferation is evaluated by flow cytometry. RMPI medium alone and SIINFEKL peptide are used as negative and positive controls, respectively. Finally, the clonal subset with the strongest suppressive activity is further analyzed by single-cell sequencing, as described below. Example 7: BTK Inhibition

[0045] Once high migration and / or immunosuppressive clones have been identified from different source MDSCs, we assess the extent to which a BTK inhibitor (i.e., ibrutinib) interferes with guided metastasis. First, we use immunoblotting to assess the levels of BTK and phosphorylated BTK (p-BTK) in each clonal subset exposed to 0-10 μM ibrutinib. Each clonal subset is then seeded (approximately 10 3 ~10 4 cells / cm 2 ), and guided migration is monitored by time-lapse microscopy while exposed to 0-10 μM ibrutinib. Images are processed / analyzed by Fiji. Experiments with ACP-196 (selective and irreversible BTK inhibitor) and GDC-0853 (selective and reversible BTK inhibitor) are performed for comparison purposes. After exposure to ibrutinib, AFM and TFM are again used to evaluate single cell stiffness and contractility, respectively. The effect of BTK inhibition on MDSC motility and biomechanics is further evaluated in breast cancer patients who received ibrutinib under the auspices of an open, OSU CCC-supported clinical study (OSU-18015) conducted at OSU. After obtaining informed consent, 30 cc of peripheral blood is collected pre-treatment and at 2 and 4 weeks after initiation of therapy. MDSCs are isolated and single clone motility and biomechanics (i.e., AFM and TFM) are evaluated as previously described. Example 8: Reciprocal regulation of MDSC / tumor cell metastasis

[0046] Preliminary studies have shown that MDSCs have specialized clonal subsets with improved motility, likely more likely to colonize tumor / lymphoid tissues or co-metastasize with highly invasive tumor cells, providing a "protective" immune suppression early in metastasis. Pilot studies on biomimetic surfaces (Figures 8-9) showed that the presence of MDSCs triggered enhanced motile epithelial to mesenchymal transition in a clonal subset of non-cancerous breast tissue cells (Figures 8A-8B), suggesting that MDSCs may have the ability to induce potentially cancerous changes in healthy tissue. In contrast, MDSCs did not appear to induce significant changes in the overall motility pattern of highly aggressive breast cancer cells (Figure 8C). Interestingly, the most aggressive tumor cells induced the strongest changes in single clonal motility in MDSCs (Figure 9). MDSCs migrate from single clonal velocities clustering near / below 40 μm / h to velocities that can occasionally reach approximately 100 μm / h, likely due to enhanced cytokine / chemokine secretion from aggressive tumor cells. Example 9: Guided migration studies at the single clone level reveal interesting potential therapeutic targets in tumor-associated myeloid-derived suppressor cell populations

[0047] method

[0048] Textured PDMS surfaces: Microtextured PDMS surfaces were fabricated from photolithographically patterned silicon masters by a replica molding process. Parallel arrays of protrusions and grooves (2 μm wide, 2 μm high, 2 μm apart) were first patterned on the silicon master by standard UV photolithography using S1813 photoresist. A 10:1 mixture of PDMS and curing agent was then cast onto the master and allowed to degas and cure for several hours. The PDMS was then demolded from the master and placed on a multiwell plate for single cell migration experiments. Scanning electron microscopy (SEM) was used to characterize the surface morphology.

[0049] MDSC culture: A murine MDSC cell line (MSC-2) was kindly donated by Gregoire Mignot. MSC-2 cells were cultured in RPMI 1640 medium supplemented with 25 mM HEPES, 10% heat-inactivated fetal bovine serum (FBS), 1% antibiotic-antimycotic, and 1 mM sodium pyruvate. Patient-derived MDSCs were enriched from peripheral blood using a RosetteSep HLA-myeloid cell enrichment kit (Stemcell Technologies), followed by Ficoll-Paque centrifugation (GE healthcare). MDSCs were then isolated by negative selection of HLA-DRneg cells using anti-HLA-DR MicroBeads (Miltenyi Biotec) for 15 min at 4°C and isolated using an MS-MACS column. Samples were collected under an IRB-approved protocol for human subjects research with informed consent.

[0050] Single cell migration assay: approximately 1.5 x 10 5 MSC-2 cells were seeded and allowed to adhere on the textured PDMS surfaces or TCP controls in regular medium for several hours. Cells were imaged by time-lapse microscopy every 10 min for 16 h using a cell culture chamber (Okolab) attached to an inverted microscope. Images were analyzed using a manual tracker plugged into Fiji. Single cell displacement data were then analyzed by MATLAB to determine velocity and net track distance traveled.

[0051] Flow cytometry-based analysis and sorting: The following antibodies were used for MSC-2 cells: anti-CD11b-FITC, anti-Ly6-C-APC and anti-Ly6-G-PE, all purchased from Biolegend. For patient-derived MDSCs, we used anti-CD33-APC, anti-CD11b-AP and anti-HLA-DR-PECy7 purchased from Beckman Coulter. Data were acquired using an LSRII flow cytometer (BD Biosciences). All colors were evaluated against their respective isotype controls and unstained samples.

[0052] Gene expression analysis: Total RNA was extracted using TRizol reagent (ThermoFisher). Reverse transcription reactions were performed using 500–1000 ng of RNA in a 20 μl reaction with the Superscript VILO cDNA synthesis kit (ThermoFisher). cDNA was used as a template to measure expression levels of pro- and anti-inflammatory genes by quantitative real-time PCR using pre-specified primers. Real-time PCR reactions were performed using a QuantStudio 3 real-time PCR System with TaqMan fast advance chemistry (Thermo Scientific) with the following conditions: 95°C for 10 min, 40 cycles of 95°C for 1 min, 60°C for 1 min, and 72°C for 1 min. Gene expression was normalized to the housekeeping genes GAPDH and ATP-6.

[0053] Orthotopic tumor xenografts: Immunodeficient nude mice (Jackson Laboratory), 6-8 weeks of age, were first injected with 1 million human breast cancer cells (MDA-MB-231) in the mammary fat pad to generate tumors. Four weeks after tumor development, the selected MDSC subpopulations were stained using the PKH67 green fluorescent cell linker kit for general cell membrane labeling (Millipore Sigma) according to the instructions suggested by the manufacturer. Tumor-bearing mice were then injected with approximately 2.5 × 10 5 MDSCs were injected via the tail vein. Mice were then harvested 1 day after injection, and tumors, lungs, and spleens were characterized with an IVIS Imaging System (Xenogen Imaging Technologies). All animal experiments were performed in accordance with protocols approved by the Ohio State University Laboratory Animal Care and Use Committee.

[0054] Statistical Analysis: All statistical analyses were performed in Sigma Plot 12 or GraphPad Prism 7. We used n=3-6 replicates per experiment. Specific information regarding the number of replicates, statistical tests, and levels of significance can be found in the figure legends.

[0055] Results and Discussion

[0056] MDSCs respond to topographical cues and display structurally guided metastatic patterns. Structurally guided cell migration is known to play a role in the escape of cancerous cells from primary tumors and the establishment of metastatic growth in peripheral organs and tissues. Highly aggressive cancer cells tend to display distinct spreading patterns, preferentially diffusing along pre-aligned anatomical microstructures within tissues, including extracellular matrix (ECM), white matter tracts, basement membranes of blood vessels, and radially oriented fibrils from the subpial / subperitoneal space, among others (Figure 10A) (Gallego-Perez D, et al. Lab Chip 2012,12:4424-4432; Bellail AC, et al. Int J Biochem Cell Biol 2004,36:1046-1069; Johnson J, et al. Tissue Eng Part C Methods 2009,15:531-540). Using micro- and nanoscale tools, we have developed systems that can be used to logically probe cancer cell motility under these physiologically relevant conditions (Gallego-Perez D, et al. Lab Chip 2012,12:4424-4432; Bellail AC, et al. Int J Biochem Cell Biol 2004,36:1046-1069; Johnson J, et al. Tissue Eng Part C Methods 2009,15:531-540; Irimia D, et al. Integr Biol(Camb) 2009,1:506-512; Doyle AD, et al. J Cell Biol 2009,184:481-490; Petrie RJ, et al. Nat Rev Mol Cell Biol 2009,10:538-549).Although topographical or cell-localized cues have been used to mimic rapid and highly directional motility in various cancerous cells (Gallego-Perez D, et al. Lab Chip 2012,12:4424-4432; Johnson J, et al. Tissue Eng Part C Methods 2009,15:531-540; Irimia D, et al. Integr Biol (Camb) 2009,1:506-512; Sidani M, et al. J Mammary Gland Biol Neoplasia 2006,11:151-163; Provenzano PP, et al. BMC Med 2006,4:38; Wong IY, et al. Nat Mater 2014,13:1063-1071), no studies have examined the impact of such cues on the metastatic / invasive potential of tumor-associated MDSCs. We next tested whether MDSCs respond to topographical cues by exhibiting structurally guided metastatic patterns similar to invasive cancer cells.The murine MDSC cell line, MSC-2, was used as a model (Stiff A, et al. Cancer Res 2016, 76:2125-2136; Trikha P, et al. Oncoimmunology 2016,5:e1214787). These cells were placed on a microtextured polydimethylsiloxane (PDMS) surface (Figure 10B), which was fabricated via replica molding from a photolithographically generated silicon master and designed as an array of parallel projections and grooves with dimensions (approximately 2 μm × 2 μm, 2 μm spacing) previously tested in cancer cell metastasis studies (Gallego-Perez D, et al. Nano Lett 2016, 16:5326-5332; Gallego-Perez D, et al. Lab Chip 2012, 12:4424-4432; Kim SH, et al. Cancer Cell 2016, 29:201-213; Gu SQ, et al. Nucleic Acids Res 2016, 44:5811-5819). MDSC motility was monitored in real time at the single clone level by time-lapse microscopy.Cells seeded on standard cell culture surfaces (i.e., tissue culture polystyrene or TCP) were used for comparison purposes. Results show that MDSCs exhibit limited motility at the single clone level on TCP (Figures 10C-10E), with most cells exhibiting a rounded morphology (Figure 10C). On the other hand, the textured surface clearly induced cytoskeletal and morphological rearrangements (i.e., alignment) in a portion of MDSCs (Figure 10C), which led to increased motility (Figures 10D, 10E). The average single clone velocity reached a maximum of approximately 40 μm / h on the textured surface compared to approximately 20 μm / h on TCP. The net track distance, a measure of the effective displacement of single clones, reached a maximum of approximately 400 μm over a 16-h period on the textured surface compared to <100 μm on TCP. Notably, MDSCs migrating on textured surfaces showed significant inter-clonal variability in metastatic potential, with cells spanning the full range of low to high motility. In contrast, MDSC migrating on TCP showed significantly lower inter-clonal variability. Studies on circulating MDSC derived from cancer patients (Figure 13) further confirmed the presence of a highly motile MDSC population that showed significant inter-clonal variability, with some clones ranging from approximately 200 μmh. -1 The MDSCs exhibited a mean guided migration velocity of up to 1 mm and an overall net displacement approaching 1 mm over a 16 hour period. However, certain populations of circulating MDSCs from patients exhibited limited overall motility, which could potentially be a direct reflection of the primary malignancy (e.g., type, stage, mutation) and / or current therapy (Tables 1-3). [Table 1] [Table 2] [Table 3]

[0057] MDSC subpopulations display distinct metastatic potential. Based on the apparent interclonal variability in motility, we define MDSC subpopulations as follows: granulocytes (CD11b + Ly6C lo Ly6G + ) and monocytes (CD11b + Ly6C hi Ly6G - The MDSC population was further stratified and probed by flow cytometry-based sorting into CD11b subpopulations (Figures 11A-11C) (Bronte V, et al. Nat Commun 2016,7:12150). + Ly6C + Ly6G + Subpopulations of cells were also identified from the flow cytometry data and included in our analysis. Flow-sorted subpopulations were then subjected to structurally guided motility studies on textured surfaces as previously described, in addition to qRT-PCR analysis of pro- and anti-inflammatory markers. Single-clonal migration studies demonstrated that granulocytic MDSC outnumbered monocytic MDSC and CD11b MDSC when probed in isolation. + Ly6C + Ly6G + The subpopulations showed superior metastatic potential (Figure 11D), with single clones occasionally reaching average velocities and net displacements of >100 μm / h and approximately 1.5 mm over a 16-h period. + Ly6C + Ly6G +Although some clones within the subpopulation displayed a relatively high average migration speed of about 50 μm / h, the net displacement was rather limited, thus suggesting that these cells tend to display very short-distance and / or disorganized motility patterns compared to granulocytic MDSCs (Figure 11E). These observations were further confirmed by in vivo studies (Figures 11F, 11G), in which tumor-bearing mice were systemically injected with fluorescently labeled suspensions of sorted vs. "fresh" / unsorted MDSCs and IVIS was used to document MDSC accumulation within the tumor niche vs. peripheral organs / tissues. Mice injected with granulocytic MDSCs showed a more prominent fluorescent signal accumulation within the tumor (Figure 11G). Parallel single-clone motility studies (Figure 14) with circulating MDSCs from cancer patients also demonstrated a greater concentration of granulocytic subpopulations (CD11b + CD15 + CD14 - ) is a monocyte subpopulation (CD11b + CD15 - CD14 + Gene expression analysis of MSC-2 cells for proinflammatory markers indicates that the “fresh” (i.e., unsorted) MDSC population exhibits enhanced motility compared to purified granulocytes, monocytes, and CD11b + Ly6C + Ly6G + The results show that there are no statistically significant differences in the expression of TNF-α, iNOS, and IL-27 between the subpopulations. However, IL-6 was significantly overexpressed in the fresh population vs. the flow-sorted subpopulation. Meanwhile, gene expression analysis of anti-inflammatory markers revealed that the flow-sorted granulocyte subpopulation was significantly overexpressed compared to the fresh flow-sorted monocytic MDSCs and CD11b + Ly6C + Ly6G + These results suggest that the granulocytic MDSC subpopulation is not only more likely to spread and colonize cancerous tissues, but also has a tendency to overexpress arginase and IL-10 compared to the monocytic MDSC and CD11b subpopulations. + Ly6C + Ly6G + These results suggest that the IL-16 subpopulations appear to be more prone to overexpressing anti-inflammatory / inhibitory markers compared to the IL-16 subpopulations.

[0058] MDSC subpopulations display phenotypic plasticity that drives population homeostasis under long-term culture conditions. Distinct subpopulations of granulocytic and monocytic MDSCs in MSC-2 cells, as well as CD11b + Ly6C + Ly6G + After flow-based purification into cells, cells were maintained in culture for 1-7 days. Phenotypic plasticity was assessed by flow cytometry on days 1 and 7. Single clone motility assays and gene expression analysis were performed on day 7 (Figure 12A). Surprisingly, and in contrast to what we found immediately after flow-based sorting; no significant differences in metastatic properties were detected across all three populations by day 7 (Figure 12B). The average single clone velocity for all populations was approximately 50 μmh. -1 The mean track distance remained within 100 μm, while the overall net track distance remained below 200 μm. Flow cytometry analysis showed that one day after sorting, purified populations still constituted the majority of the cultures (~80%), but by day 7, the entire hierarchy of populations had been re-established (Figures 12C-12E), possibly suggesting a role for cellular plasticity in maintaining population homeostasis / heterogeneity in MDSC populations. Cell cultures from purified granulocyte subpopulations (Figure 12C) showed that, for example, monocytic MDSCs and CD11b + Ly6C + Ly6G + The monocyte subpopulation showed the steepest increase from day 1 to day 7 (approximately 7-fold change), and CD11b + Ly6C + Ly6G + The purified monocytic MDSC-derived cultures showed an approximately three-fold increase in CD11b expression by day 7 compared to the granulocyte population. + Ly6C + Ly6G + The population was easily cleared (increased by about 2.5 fold). Finally, purified CD11b + Ly6C + Ly6G +Cultures from the populations were more likely to give rise to granulocytic MDSCs (approximately 3-fold increase) by day 7 compared to monocytic MDSCs, which did not show a significant increase between days 1 and 7. Gene expression profiles of proinflammatory (FIG. 12F) and anti-inflammatory (FIG. 12G) markers at day 7 showed more subtle differences across populations, with increased and decreased iNOS and IL-6 expression, respectively, in the "fresh" MDSC population compared to the sorted / purified subpopulations. However, when comparing expression profiles between day 0 (i.e., the day of sorting / purification) and day 7, more striking differences were seen, with an overall increase in proinflammatory iNOS expression for all three populations and a significant decrease in arginase 1 and Il-10 for the granulocytic subpopulation only (FIG. 15).

[0059] Micro- and nanoscale techniques are widely used to investigate and / or modulate various aspects of cell biology for medical applications (Gallego-Perez D, et al. Nano Lett 2016,16:5326-5332; Gallego-Perez D, et al. Lab Chip 2012,12:4424-4432; Kim SH, et al. Cancer Cell 2016,29:201-213; Gu SQ, et al. Nucleic Acids Res 2016,44:5811-5819; Minata M, et al. Cell reports 2019,26:1893-1905; Shukla VC, et al. Trends in biotechnology 2018,36:549-561; Benavente-Babace A, et al. Biosens Bioelectron 2014,61:298-305;Fei Z,et al.Analytical Chemistry 2013,85:1401-1407;Chang L,et al.Small 2016,12:5971-5980;Chang L,et al.Lab Chip 2015,15:3147-3153;Gallego-Perez D,et al.Biomed Microdevices 2012,14:779-789;Gallego-Perez D,et al.Nanomedicine 2016,12:399-409;Gallego-Perez D,et al.Nature nanotechnology 2017,12:974;Wu Y,et al.Small 2013,9:2358-2367;Zhao X,et al.Advanced Science 2015,2; Zhao X,et al.Anal Chem 2015,87:3208-3215). Using microscale engineering tools, we demonstrated that tumor-associated MDSCs exhibited structurally guided migration patterns similar to those of invasive cancerous cells.Single clonal motility analysis revealed clear heterogeneity within and across MDSC populations (i.e., for patient-derived MDSCs) and confirmed the presence of clonal subsets with enhanced metastatic potential in both murine- and patient-derived MDSCs. Motility tracking studies coupled with flow cytometry-based sorting, gene expression analysis, and orthotopic tumor xenograft experiments in nude mice suggest that granulocytic subpopulations are likely to display increased metastatic and tumor-infiltrating capacity, as well as enhanced anti-inflammatory activity, making this population an attractive therapeutic target in cancer. Subsequent studies, however, revealed the remarkably dynamic and plastic nature of such clonal subsets, with purified MDSC subpopulations rapidly reaching collective homeostasis by giving rise to the full range of MDSC phenotypes. Although there are conflicting reports regarding the dominant phenotype of tumor-resident MDSCs (i.e., granulocytes vs. monocytes) (Kumar V, et al. Trends in immunology 2016,37:208-220; Hossain F, et al. Cancer immunology research 2015,3:1236-1247; Haverkamp JM, et al. European journal of immunology 2011,41:749-759; Mairhofer DG, et al. Journal of Investigative Dermatology 2015,135:2785-2793; Bozkus CC, et al. The Journal of Immunology 2015,195:5237-5250), our monoclonal migration and phenotypic plasticity results point to a potential mechanism by which granulocytic MDSCs may possibly be better equipped to infiltrate the tumor niche, where they subsequently remain granulocytic and / or give rise to monocytic MDSCs, depending on multiple factors including tumor type.Interestingly, monoclonal metastasis using circulating MDSCs from cancer patients suggests that MDSC motility may potentially be influenced by patient background (e.g., cancer type / stage, treatment, etc.) and therefore further studies are required to determine whether the metastatic patterns of circulating MDSCs ex vivo can be used to monitor disease and / or treatment progression.

[0060] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the disclosed invention belongs. All publications mentioned herein and the materials cited therein are specifically incorporated by reference.

[0061] Those skilled in the art will know, or be able to ascertain using no more than routine experimentation, many equivalents to the specific embodiments of the invention described herein which equivalents are intended to be encompassed by the following claims.

Claims

1. 1. A method for producing tumor-specific therapeutic cells, comprising: (a) selecting a highly motile subpopulation of myeloid derived suppressor cells (MDCS) from a tumor sample of a subject; and (b) genetically engineering said subpopulation to heterologously express a transgene encoding an anti-tumor protein, an anti-tumor oligonucleotide, or a combination thereof, thereby producing said tumor-specific therapeutic cells; The subpopulation is granulocyte CD11b ＋ Ly6C ｌｏ Ly6G ＋ A method that is an MDCS.

2. The method of claim 1, wherein the subpopulation is sorted in a migration assay using a chemoattractant gradient.

3. The method of claim 1 or 2, wherein the subpopulation is selected in a migration assay using nanotextured and / or biomimetic surfaces.

4. The method of any one of claims 1 to 3, wherein the subpopulation is selected in a Transwell migration assay or a Boyden chamber assay.

5. The method of claim 1, wherein the transgene encodes tissue inhibitor of metalloproteinase-3 (TIMP-3).

6. The method of any one of claims 1 to 5, wherein the subpopulation is genetically engineered with a kill switch system.

7. The method of claim 1 , wherein the subpopulation is sorted by flow cytometry.

8. A composition comprising a plurality of tumor-specific therapeutic cells produced by the method of any one of claims 1 to 7.

9. The composition of claim 8 further comprising a pharma- ceutically acceptable excipient.

10. 10. The composition of claim 9 for treating a solid tumor in a subject, wherein the solid tumor to be treated is matched with the tumor-specific therapeutic cells.

11. The composition of claim 10, wherein the solid tumor is breast cancer.

Images