CD90-targeted multiplex virus-like particles (MVPs) for concomitant In Vivo gene editing, transduction, and transcription factor delivery into HSCs Free

In vivo delivery strategies will substantially broaden the accessibility of Hematopoietic Stem Cell (HSC) gene therapy. However, current platforms lack the capacity for iteration, requiring binary choices in gene modification strategies. Here, we report Multiplexed Virus-Like Particles (MVPs) enabling concomitant in vivo transduction, gene-editing, and recombinant protein delivery to HSCs. Depending on the disease pathology, some disorders require a minimal threshold of gene modification for pathophysiological relief, necessitating post-gene modification enrichment. To enrich for gene-modified cells, we evaluated the co-delivery of recombinant constitutively active STAT5 in our viral particles (MVP-S), a downstream target triggered by cytokines activating HSCsand a suppressor of p53-mediated DNA damage response. We hypothesized that transient delivery of STAT5 would induce self-renewal of HSPCs, thereby facilitating positive enrichment synergizing with negative selection (i.e., chemo selection, CAR-T, ADCs).

To confirm functionality of MVP-S, HSPCs from healthy donors in the absence of cytokines were treated with MVP-Cas9 (B2M KO + mNeonGreen transgene) or MVP-C+S (B2M KO + mNeonGreen transgene + STAT5). Editing analysis, immunoblotting, and bulk RNA-seq of samples were performed. An average allelic B2M editing rate of 17.6% and 22.5% was observed for MVPs without and with STAT5A, respectively. While both treatment groups showed similar levels of phosphorylated H2AX, HSPCs treated with MVP-C+S had a lower abundance of phosphorylated p53. Finally, gene set enrichment analysis of MVP-C+S HSPCs confirmed the activation of gene sets associated with proliferation (mitotic spindle), while differentiation-associated pathways (oxidative phosphorylation, oxygen species pathways) were downregulated to maintain stemness. These results suggest that STAT5 signaling mitigates the genotoxic effects of CRISPR editing in HSPCs.

We next sought to evaluate the impact of STAT5 programming on in vivo gene modification. Humanized mice were treated with either MVP-Cas9 or MVP-C+S and euthanized 2 weeks post-transduction. Transgene expression of 30-45% was observed in human CD45⁺ WBCs and HSPC subsets in the bone marrow of mice. Additionally, an increase in B2M editing was observed by flow cytometry in human CD45⁺ cells (11.2% vs. 20.2%, p = 0.0232) and CD90⁺ HSCs (1.2% vs. 6.5%, p = 0.0318) in mice that received MVP-C+S, suggesting a STAT5-enhanced contribution of gene-edited cells to steady-state hematopoiesis. Biallelic editing was confirmed by single-colony DNA sequencing of HSPCs.

We next evaluated whether gene enrichment strategies could be multiplexed in vivo using MVPs. MVP-S, delivering mCherry and the chemoprotective MGMT(P140K) transgene was injected into gene barcoded humanized mice, underwent two rounds of chemo selection, and gene marking was followed up to 10 weeks post-injection. At necropsy, the bone marrow of mice was analyzed for gene marking in HSPC subsets, and single-cell RNAseq was performed to analyze clonality. MVP-S-treated mice without chemo selection showed an average of 16% gene marking in CD34⁺CD90⁺ HSCs, which was 2.5 times higher than in mice that underwent no enrichment. Furthermore, in mice that received MVP-S and underwent two rounds of chemoselection, we observed an 85.5% HSC gene marking rate, compared to mice that underwent chemoselection without STAT5 enrichment, which had an average gene marking rate of 30.7%. STAT5-induced gene enrichment did not perturb or bias hematopoietic output as gene marking in all hematopoietic lineages (B cells, T cells, etc.) remained similar between mice. Finally, single-cell RNA-seq analysis showed a high polyclonal distribution, as indicated by the presence of hundreds of unique barcodes.

In summary, MVPs represent a novel in vivo HSC gene delivery platform, enabling the multiplexing of HSC engineering strategies. STAT5 programming of HSCs can be used in place of, or concurrently with, negative selection strategies for replacing endogenous HSCs. Furthermore, the in vivo delivery of transcription factors to a specific population opens new possibilities for in vivo transdifferentiation and programming of pathologic subpopulations. MVPs, therefore, have broad utility not only for improving current gene therapy strategies but can be further utilized in a variety of regenerative medicine platforms.