CRISPR activation in lymphoma reveals a new route to drug resistance

CRISPR activation in lymphoma turns a familiar functional-genomics experiment on its head. Instead of systematically disabling genes to expose a tumour's vulnerabilities, researchers activated genes across the genome and asked which ones could help lymphoma cells survive treatment or become more aggressive.

One of the clearest examples was Irx5. Increasing its activity made cells from a mouse model of aggressive lymphoma more resistant to venetoclax, a drug that inhibits the pro-survival protein BCL-2. The Irx5-activated cells increased another survival protein, providing an alternative route around BCL-2 inhibition.

The finding came from a broader study published in Science Advances that introduces Partita, a high-density genome-wide CRISPR activation platform for mice.

Turning genes up rather than knocking them out

Many CRISPR screens use loss of function. A gene is disrupted, and researchers measure what happens to the cell. This is extremely useful for answering questions such as: Which genes does a cancer cell depend on?

But cancer can also change because a gene becomes too active rather than because it is lost. Increased expression may accelerate tumour growth, alter cell identity or create drug resistance. A knockout screen is not designed to reproduce that kind of change.

CRISPR activation, or CRISPRa, addresses the opposite side of the problem. Cas9 is used as a targeting system rather than as a DNA-cutting enzyme, bringing transcriptional activators to the regulatory region of a selected gene and increasing its endogenous expression.

The researchers built Partita as a mouse whole-genome CRISPRa library. It uses 10 guide RNAs for each transcription start site and is organised into five sub-libraries. Using multiple guides per target increases the chance of achieving robust activation and makes genome-scale screening more reliable.

These experiments are pooled rather than performed one gene at a time in separate dishes. Different cells receive different guides, and researchers later determine which guides have become enriched or depleted. A guide that becomes much more common after treatment can point to a gene that gives cells a selective advantage.

An escape route from venetoclax

The team applied Partita to a mouse model of double-hit lymphoma, an aggressive lymphoma driven by high MYC and BCL-2 activity. Genome-wide activation screens were performed under treatment with three pro-apoptotic drugs: venetoclax, nutlin-3a and etoposide.

The logic was straightforward. If activating a particular gene allowed cells to survive a drug that killed most of the population, guides targeting that gene would become enriched among the survivors.

The screens recovered expected components of known survival pathways but also produced less obvious candidates. Irx5 was particularly notable in the venetoclax screen.

Venetoclax works by blocking BCL-2, a protein that prevents apoptosis in many blood cancer cells. Yet cellular survival is controlled by a network rather than by a single protein. When Irx5 was activated, lymphoma cells increased another pro-survival protein and became less dependent on the BCL-2 pathway targeted by venetoclax.

This illustrates an important route to drug resistance. A tumour cell does not necessarily need to alter the drug target itself. It may instead rewire the surrounding survival network and bypass the blocked pathway.

Screening tumour acceleration in living mice

The researchers also used Partita in vivo to search for genes whose activation could accelerate MYC-driven lymphomagenesis.

Runx2, Runx3 and Csf1r emerged among the prominent hits. Activating these genes gave the corresponding tumour cells a selective advantage during lymphoma development, demonstrating that the platform can identify not only drug-resistance mechanisms but also gain-of-function changes that promote cancer progression.

The study does not establish IRX5, RUNX2, RUNX3 or CSF1R as validated therapeutic targets in patients. The experiments were performed primarily in mouse cells and mouse cancer models, and experimentally enforced CRISPR activation may produce expression levels that do not precisely match those found naturally in human tumours. The relevance of individual hits therefore has to be tested in human disease.

The conceptual contribution is nevertheless distinct. Loss-of-function screens ask what a tumour cannot live without. Partita allows researchers to ask the complementary question: what is sufficient to make a tumour grow faster or survive a drug?