Preprint: the second CAR receptor sometimes appears on T-cells before the first receptor recognizes its target
Preprint: the second CAR receptor sometimes appears on T-cells before the first receptor recognizes its target
On August 31, the authors published a preprint on CAR-T cells, which are T-cells equipped with an engineered receptor that recognizes a target and triggers its destruction. In their two-stage system, the first receptor, SynNotch, recognizes a marker on the surface of another cell and induces expression of the second receptor, CAR. In some T-cells, CAR appeared on the surface before this step occurred. An mScarlet tag attached to the end of CAR reduced this background expression in cell experiments and mouse xenografts, helping the cells distinguish between high and low HER2 density.
HER2 can be particularly abundant on tumor cells, but it is also present at lower levels in some healthy tissues. A CAR-T cell must distinguish a cell with high HER2 density from one with low HER2 density so that it attacks the former while sparing the latter.
In a 2021 study, Rogelio Hernandez-Lopez and colleagues built a two-receptor circuit for this purpose. SynNotch, a synthetic switch receptor, was designed to respond to high HER2 density and induce expression of a CAR that binds this protein more strongly. CAR therefore appears after contact with a cell that carries sufficient HER2.
The authors measured what happens before such contact. In three constructs, CAR was already present on the surface of approximately 6%, 2%, and 0.2% of resting T-cells. Within this series, the construct with the highest background expression also showed the weakest target discrimination. In a two-tumor mouse model, the system with the highest background reduced the volume of both the tumor with high HER2 density and the tumor with low HER2 density.
When CAR binds its target, it activates the T-cell and initiates its proliferation. A small initial fraction of CAR-positive T-cells can therefore alter the behavior of the entire population. The authors' mathematical model incorporated the fraction of these cells, HER2 density, and the ratio of T-cells to target cells. Experiments showed that each system achieved its best discrimination at a different T-cell dose.
The authors then sought to reduce CAR on the surface of resting T-cells. mScarlet reduced its baseline surface expression by approximately 60%, while largely preserving CAR expression after signaling. In the two-tumor model, the mScarlet version suppressed the high-antigen tumor while the low-antigen tumor continued to grow.
Degrons, which are tags that accelerate protein degradation, reduced background expression more strongly in culture. After CAR induction, T-cells carrying degron tags lost their ability to kill the target more quickly. In mouse models, mScarlet provided better discrimination between high-antigen and low-antigen targets. The performance of a two-stage CAR-T system depends both on when CAR expression begins and on how much CAR remains on the surface after the correct signal.