Ex vivo hematopoietic stem cell (HSC) gene therapy can cure various blood and immune disorders. However, barriers persist regarding the accessibility of these treatments and current cytotoxic conditioning regimens used to deplete host HSCs for the engraftment of gene-modified cells. In vivo delivery strategies can broaden the accessibility of gene therapy but require new target-specific delivery platforms to ensure safety and efficiency. Here, we report the development of HSC/CD90-targeted Multiplexed Virus-Like Particles (MVPs), enabling concomitant in vivo transduction and multiplex gene-editing in HSCs. In addition, we developed a CD90 ablation/epitope editing strategy to either selectively enrich in vivo gene-modified HSCs or protect HSCs from cancer immunotherapy using a CD90 antibody-drug conjugate (ADC) or anti-CD90 CAR T cells.

HSC targeting was enabled by studding a cocal-pseudotyped MVPs with a recently redesigned, 2nd-generation anti-CD90 single-chain variable fragment (scFv). Administration of transgene-loaded CD90-MVPs in humanized NBSGW mice displayed ~14.8% transduction of long-term engrafting human HSC with serial reconstitution potential in secondary mice. Chemotherapy-mediated enrichment of MGMT(P140K) transgene expressing cells led to an increase up to ~65% transduction efficiency in HSCs without perturbing multilineage differentiation in the blood and bone marrow. Most encouraging, in vivo administration of CD90-MVPs loaded with Cas9 protein for HBG editing and transgene led to ~18% transduction efficiency and up to 30.3% allelic editing at the HBG locus of human CD90+ HSCs engrafted in primary NBSGW mice reaching the presumed therapeutic threshold for sickle cell treatment even without further chemo selection.

To avoid genotoxic MGMT(P140K)-mediated selection regimens for the in vivo enrichment of gene-modified cells, we confirmed that the knock-out of CD90 does not perturb HSC engraftment and multilineage differentiation in the mouse xenograft as well as autologous nonhuman primate (NHP) model. No impact of short-term recovery or long-term multilineage engraftment in the blood and bone marrow were seen in the mouse or NHP indicating that CD90 is not essential for adult hematopoiesis. In parallel, we developed anti-CD90 CAR/ADC allowing selective killing of remaining CD90-WT HSCs and CD90-expressing cancer/tumor cells for selection/enrichment and consolidation therapy, respectively. CD90 KO in combination with CD90-CAR/ADC was comprehensively tested ex vivo and in vivo using a humanized NBSGW tumor model. We observed enrichment of edited HSCs from ~15% to over 70% and reached entire clearance of CD90-expressing tumor cells without impacting the human engraftment in mice. Finally, utilizing CryoEM in combination with an adenine base editor (ABE) screen on exon 2 of the CD90 gene delivered with virus-like particles (VLPs), we identified the binding epitopes of our CD90-targeted MVPs, CAR, and ADC. Precise mutation of the recognized epitope using a single ABE, we were able to shield human and NHP HSCs from CD90-CAR/ADC without impacting their mouse xenograft or autologous engraftment potential, respectively.

In summary, CD90-MVPs are a novel gene delivery platform, allowing efficient, portable, and highly versatile multiplex gene engineering of HSC in vivo. Furthermore, knock out as well as epitope editing of CD90 provides complete protection from CD90-CAR/ADCs without disrupting functionality or CD90 protein structure. Both technologies combined offer a robust strategy to gene modify and enrich gene-corrected HSCs by removing uncorrected HSCs and provide on-target specificity as gene therapy moves toward in vivo administration and enables a wider application of HSC gene therapy. Combination of CD90-KO/epitope editing with CD90-CAR/ADCs further provides a novel strategy to selectively remove CD90-expressing cancer stem cells while protecting healthy HSCs.

This content is only available as a PDF.
Sign in via your Institution