mRNA engineering in prostate cancer restored an immune signal in mice

mRNA engineering in prostate cancer strengthened antitumour immunity in two mouse models. Rather than altering DNA, the researchers made tumour cells display more MHC-I — the molecular identification system that allows cytotoxic T cells to inspect what is happening inside a cell.

MHC-I molecules present small protein fragments on the cell surface. When those fragments include tumour antigens, CD8 T cells can recognize and kill the abnormal cell. Many cancers reduce MHC-I expression and become harder for the immune system to see. This is one reason prostate cancer is usually an immunologically “cold” tumour and often responds poorly to immune-checkpoint blockade.

The researchers first analysed RNA-sequencing data from 501 androgen-dependent and 76 castration-resistant prostate cancers. They found 370 messenger RNAs with shortened 3′ untranslated regions, or 3′UTRs, in the advanced tumours. For 46 genes, this shortening was associated with the time to biochemical recurrence. SPSB1 emerged as a leading candidate. These patient data establish an association; they do not by themselves prove that the RNA change causes disease progression.

A 3′UTR is a non-coding tail that helps control an mRNA’s stability and how efficiently it is translated into protein. Cells can finish the same transcript at different polyadenylation sites. Selecting an earlier site produces a shorter tail and can remove regulatory sequences that restrain protein production. In prostate tumour cells, the shortened SPSB1 transcript produced more SPSB1 protein.

Excess SPSB1 did more than support tumour growth. It recruited an enzyme complex that attached ubiquitin to MHC-I, marking the molecule for degradation. With less MHC-I on the surface, antigen presentation weakened and CD8 T-cell attack declined. In the tested models, this intervention did not increase the immune-suppressive ligand PD-L1.

The team built a system called 3′UTRCES to reverse the RNA-processing change. A catalytically inactive, RNA-binding dCas13b protein was guided to the proximal polyadenylation site of SPSB1. By physically obstructing the cleavage and polyadenylation machinery, it forced the cell to use a more distant site and restore the longer 3′UTR. SPSB1 protein production fell, while surface MHC-I increased.

In cultured human and mouse prostate cancer cells, the system improved antigen presentation and made tumour cells more vulnerable to CD8 T-cell killing. In one specificity experiment, targeting SPSB1 did not measurably alter the use of roughly 10,000 other polyadenylation sites, although this limited assay cannot rule out every possible off-target effect.

For the animal experiments, dCas13b mRNA and guide RNA were packaged in lipid nanoparticles and injected directly into tumours. In two immunocompetent prostate cancer models, with seven mice in each treatment group, combining 3′UTRCES with antibodies against PD-1 and CTLA-4 produced the strongest tumour control. In the Pten- and P53-deficient model, checkpoint blockade alone had little effect. Removing the relevant MHC-I gene or disrupting CD8 T-cell activity abolished the combination’s advantage, supporting the proposed mechanism.

This is still preclinical work. The platform was tested in cells and mice, delivery was local rather than systemic, and the durability of the effect, repeat dosing, whole-body delivery and possible immune responses to the bacterial dCas13b protein remain unresolved. The study therefore does not establish a treatment for patients. It demonstrates a broader principle: tumour immune evasion may sometimes be weakened by reprogramming the processing of a specific mRNA without permanently editing the genome.