Brain Cell Clusters Mirror Changes in a Newborn’s Brain

Researchers have shown in mice that designer antibodies can curb the growth of tumors by targeting two of cancer’s most infamous offenders—the proteins RAS and p53, which are mutated in many tumors but have largely defied drug development efforts. If their promise holds up in clinical trials, such drugs could make it possible to unleash the body’s immune system on hard-to-treat cancers including pancreatic and ovarian.

The new drugs can lock onto a few snippets of p53 or RAS jutting out from a tumor cell’s surface—and then trigger immune cells to attack them. The therapies have the potential to work when there’s a really low number of the target on cells. P53 is a tumor suppressor, and the intact protein helps healthy cells repair DNA—or self-destruct if the damage can’t be fixed. When p53 is turned off in tumors, they can grow unchecked. But targeting p53 with drugs is difficult, because restoring its activity is much harder than inhibiting its activity or turning off its production—more typical drug strategies against misbehaving proteins.

RAS, which signals cells to grow uncontrollably when mutated, has been hard to target with inhibitors because of its smooth shape and lack of obvious binding sites.

And both proteins function inside cells, making them hard to fight with engineered antibodies, versions of the Y-shaped proteins our immune system uses to tag foreign invaders for destruction. Antibodies can’t easily get inside cells, so drugs based on them work best against cancer proteins that poke out from a tumor cell’s surface. But even though RAS and p53 stay inside tumor cells, the cell’s surface carries traces of them, snippets that can be sensed by the immune system. To target these fragments of mutant p53 and RAS, known as neoantigens, cancer geneticist Bert Vogelstein’s lab at Johns Hopkins University turned to bispecific antibodies. Standard antibodies have two identical arms, but bispecifics are crafted to have one arm that binds to immune soldiers called T cells, and another that links to a cancer cell surface protein, bridging the cells and activating the immune cell to attack its new, cancerous partner.

The challenge was that the bits of mutant p53 and RAS that an antibody could target are extremely scarce on tumor cells—fewer than 10 copies per cell, the researchers found. Discovering a bispecific antibody that would bind to them, but not to healthy cells, took Hopkins graduate students Emily Hsiue and Jacqueline Douglass and their team more than 5 years. First, they tested a library of antibody fragments to find those that stuck to the p53 and RAS neoantigens. Next, they converted these fragments into different bispecific antibody designs and tested which were best at coaxing T cells to kill cancer cells in a dish.

In the end, the team came up with a p53-targeting “diabody,” a compact, two-armed antibody lacking the stem of a typical Y-shaped antibody. In mice with tumor cells bearing a specific mutation in the p53 gene, this bispecific dramatically curbed tumor growth, the researchers report today in Science. Two separate RAS diabodies worked well on lines of cultured cells with two different cancer-promoting mutations and modestly slowed the growth of tumors in mice, the team writes today in Science Immunology. In a third study led by research fellow Suman Paul, the same type of pared-down, double-target antibody also worked in mice against a type of leukemia that involves T cells, they report in Science Translational Medicine. Its target was yet another hard-to-drug type of cancer.