Background:
The immunosuppressive myeloid tumor microenvironment represents a significant barrier to effective cancer immunotherapies. We and others have previously shown that viral vector-based engineering of mature myeloid cells to express chimeric antigen receptors (CAR-M) can overcome immune suppression and endow these cells with antitumor function in preclinical models. CAR-M mediate antitumor responses through both direct phagocytosis of tumor cells and ancillary engagement of the adaptive immune system. However, the limited in vivo half-life of current myeloid immunotherapies restricts their long-term potential activity. We therefore sought to improve the persistence of myeloid-based cellular immunotherapies by genetically engineering primary human hematopoietic stem/progenitor cells (HSPCs).
Methods and Results:
To improve the durability and effectiveness of CAR-M therapy, we exploited the dispensability of the myeloid lineage-restricted receptor CD33 for normal myelopoiesis and hematopoietic cell function. Using non-viral CRISPR/Cas9 knock-in to the CD33 locus of primary human hematopoietic stem cells (HSPCs), we replaced the normal CD33 gene with that of a tumor-specific CAR or reporter protein, successfully generating CD33CAR and CD33mCherry engineered HSPCs. We then proceeded to test this novel engineered cell therapy product. Immunodeficient mice engrafted with CD33 knock-in human HSPCs reconstituted hematopoiesis and demonstrated long-term, stable gene expression for >8 months during both primary and secondary engraftment. Furthermore, directed integration of the cassette into the endogenous CD33 locus restricted gene expression to the myeloid lineage in vivo.
To assess the antitumor function of CD33CAR engineered HSPCs, we tested their activity in a humanized xenograft mouse model of metastatic Her2-positive ovarian cancer. Mice engrafted with human CD33HER2CAR HSPCs, bearing CAR-expressing myeloid cells targeting the tumor associated antigen Her2, showed improved tumor control compared to control mice engrafted with CD33mCherry HSPCs that did not contain a CAR. Unexpectedly, despite equivalent human immune subsets and engraftment kinetics in the peripheral blood, flow cytometry and single-cell RNA sequencing of tumor digests revealed massive remodeling of the human immune tumor microenvironment within CD33HER2CAR engrafted mice compared to controls. CD33HER2CAR animals showed significantly increased total T cell infiltration, including T cell populations associated with responses to immune checkpoint blockade. We additionally found enhanced expression of interferon gamma-inducible chemokines within the peripheral blood of CD33HER2CAR engrafted mice, indicating a potential mechanism for enhanced T cell recruitment. These findings were particularly surprising given the absence of viral vector induced interferon in this system and suggest that durably engineered, CAR-expressing myelopoiesis is sufficient to remodel the tumor microenvironment toward more favorable immune archetypes.
Conclusions:
Nonviral gene knock-in at CD33 within primary human HPSCs reconstitutes long-term myelopoiesis and leads to myeloid lineage-restricted gene expression. Expression of a tumor-specific CAR by myeloid cells derived from CD33 engineered HPSCs remodels the tumor microenvironment by enhancing T cell infiltration. Genetic engineering of CD33 in HSPCs thus offers a tractable platform to improve the durability and effectiveness of myeloid-based cellular immunotherapies.
Gill:Mission Bio: Membership on an entity's Board of Directors or advisory committees; Asher Biotherapeutics: Research Funding; Novartis: Patents & Royalties, Research Funding; Carisma: Current holder of stock options in a privately-held company; Interius: Current holder of stock options in a privately-held company, Research Funding.
This feature is available to Subscribers Only
Sign In or Create an Account Close Modal