Preprint Club
A cross-institutional Journal Club Initiative

Keywords:

• CD8+ T lymphocytes

• Major Histocompatibility Complex I

• Retrovirus

Main findings:

Pseudotyping a fusogen-displaying murine leukemia gammaretrovirus (MuLV) with class I MHC H2-Db single chain trimer (SCT) enables targeting of antigen-specific T cells for gene cargo delivery, activation and expansion. The authors used TRP1high TCR transnuclear mice as a source of CD8 T cells that recognize tyrosine-related protein 1 (TRP1), a well-described melanoma antigen. Administering a virus pseudotyped with a SCT carrying a TRP1 mimitope (termed “A1”) together with an ecotropic envelope protein (Eco), enabled cognate T cell activation and efficient delivery of a ZsGreen cargo to target cells within a polyclonal CD8 cell population in vitro Specific cargo delivery, activation and expansion were achieved even when antigen-specific T cells were present at low starting frequencies (1-5%) within the bulk population, showing potential for activating rare, cognate T cells. To evaluate the generality of the pMHC-pseudotyping approach, viruses bearing two different SCTs were shown to exhibit similarly high specificity for their cognate T cells among off-target cells in vitro.

A1-virus carrying a gene encoding tethered murine IL-12 with downstream ZsGreen (A1-mIL12 virus) was used to demonstrate the delivery of a therapeutic cargo. The virus was confirmed to induce IL-12 expression in antigen-specific T cells associated with STAT4 phosphorylation and IFNg production in vitro. When TRP1high CD8 T cells were transduced ex vivo and injected into B12F10 melanoma-tumour inoculated mice, survival was significantly prolonged relative to animals receiving untransduced TRP1high T cells. Increased frequencies of antigen-specific T cells were observed within tumours, and the proportion of these cells retaining transgene expression (ZsGreen+) mirrored transduction frequencies pre-transfer, demonstrating maintenance of gene expression in vivo.

When A1-mIL12 virus was directly injected into tumour-bearing B6 mice reconstituted with 1-5% TRP1high CD8s, survival was improved compared to that of A1-ZsGreen or Eco-mIL-12 virus-treated mice, demonstrating the need for gene targeting to tumour-specific cells for therapeutic efficacy.   Survival  was not significantly increased by co-administration of A1-mIL12 virus with anti-PD1 antibody. Antigen-specific T cell transduction, activation and expansion was achieved in vivo, accompanied by intra-tumoural IFNg expression and MHC II upregulation in macrophages. In a further cohort of mice receiving an increased viral dose and monitored until day 40, CD44+ CD62L- effector-memory phenotype transduced cells were detected in some animals, suggesting generation of long-lived anti-tumour responses by engineered T cells in vivo.

Limitations:

• Compared to virus displaying A1 single-chain trimers, the transduction efficiencies achieved with the two other single-chain trimers-pseudotyped viruses appeared notably lower (Figure 2C, 2D). Hence, although pMHC-pseudotyped viruses can be readily adapted to target different TCRs, transduction efficiencies may vary greatly depending on peptide-MHC and pMHC - TCR affinity, or other unelaborated factors. This may impede development of pMHC-pseudotyped viruses against certain cognate TCRs.

• Analysis of the expression of tethered mIL-12 versus the downstream ZsGreen reporter in the A1-mIL12 virus suggests that ZsGreen is expressed in a lower proportion of cells than mIL-12 (Figure 3A). This may be due to the internal ribosomal entry site (IRES) enabling only relatively inefficient ZsGreen expression. Use of a 2A ribosomal skipping sequence in place of the IRES may improve ZsGreen expression and have the added benefit of being shorter, which could improve viral packaging. Inefficient ZsGreen expression may have led to underestimation of the true level of transduction of TRP1high CD8 T cells in mice.

• In the cohort of mice that received a high dose of virus, although treatment with the A1-mIL12 virus significantly improved survival relative to control groups, there was a trend for an initial drop in survival in the A1-mIL-12-treated group before day 15 post tumour inoculation, when survival began to decline in the PBS-only group. This may indicate some form of toxicity, possibly caused by potent induction of IFNg. Careful dosing will need to be considered to balance possible acute toxicity versus long-lasting effector T cell functions.

Novelty and significance:

Tumour-infiltrating lymphocyte therapy (TIL), in which autologous T cells are isolated from tumour biopsies, expanded, then reintroduced into patients, is hampered by T cell exhaustion caused by intra-tumoural antigen exposure and long in vitro T cell expansion regimes. To overcome the drawbacks of ex vivo T cell expansion and engineering, gene delivery to T cells in vivo offers an attractive alternative. mRNA-carrying lipid nanoparticles displaying pMHCs have been used to target cognate T cells, but result in only transient mRNA expression that may not enable sustained anti-tumour activity. pMHC-pseudotyped lentiviral gene delivery has previously been shown to stably transduce cognate T cells in vitro; but the novelty in this preprint lies in demonstrating that this approach is able to target long-lasting expression of therapeutic cargo in specific cognate T cells in vivo. This method of cargo delivery streamlines otherwise labour-intensive and time-consuming manufacturing processes in TIL production, downsides often seen as major bottlenecks in adoptive cell transfer therapies. While the authors have selected tethered mIL-12 as an anti-tumour therapeutic cargo, pMHC-pseudotyped viruses can be applied to a range of diseases requiring antigen-specific gene delivery and could significantly improve the precision and quality of cell therapy products.

Credit

Reviewed by Hugo Kwong as part of a cross-institutional journal club between the Icahn School of Medicine at Mount Sinai, the University of Oxford, the Karolinska Institute, the University of Toronto and the University of Texas MD Aderson Cancer Center.

The author declares no conflict of interests in relation to their involvement in the review.