The immune system is well-known as the body’s first line of defence against harmful pathogens. It is also responsible for cleaning up dead or dying cells in our own body. To do these tasks, the immune system has evolved intricate mechanisms for detecting potential threats. Antigens like the sugar coats of bacteria cells, or the spike protein of the SARS-CoV-2 virus, serve as red flags that not only allow the immune system to launch targeted attacks, but also to remember past threats.

However, the mechanisms by which immune cells detect and remember specific antigens are not always perfect. Some molecules produced by human cells closely resemble antigens introduced by foreign invaders. Double-stranded RNA (dsRNA) is a prime example. Structurally, it resembles the genetic material of viruses. This means that just as virus-derived dsRNAs stimulate antiviral responses, human-derived dsRNAs could also activate analogous immune pathways.

The challenge for the body is to distinguish its own RNA from a genuine viral threat.

One of the key checkpoints protecting cells against its own dsRNA involves adenosine-to-inosine (A-to-I) RNA editing by the protein ADAR1, which was long-believed to mask dsRNA by destabilising its structure to keep it off the immune system’s radar.

However, recent research led by Associate Professor Polly Chen from the Cancer Science Institute of Singapore, published in Science Immunology in June 2026, has shown that this is only partially accurate.

DDX6 keeps the body’s own RNA off the immune radar

In dissecting the role of ADAR1, the team discovered that A-to-I RNA editing was not guaranteed to destabilise dsRNA. Destabilisation of dsRNA regions occur only when RNA editing occurs at adenosine-uracil (A-U) base-pairs, converting it to the less stable inosine-uracil (I-U) wobble pairs. When ADAR1 targeted adenosine-cytosine (A:C) mismatches, which are commonly found in dsRNA, its editing function reinforced the dsRNA structure by creating more stable inosine-cytosine (I-C) pairs. Since A:C mismatches are preferred editing targets of ADAR1, it means that ADAR1, if working alone, will effectively stabilise dsRNA and enhance its recognition by immune sensor proteins such as MDA5. This will eventually trigger the activation of interferon signalling pathway, a fundamental cellular immune response that may signal immune cells to launch an attack.

But proteins rarely work alone. With numerous protein-based checks and balances in place to regulate our genome and its function, it is not surprising that ADAR1 also works together with other proteins when carrying out its duties.

Seeking to identify which proteins were working with ADAR1, the team turned to the DEAD-box family of RNA helicases. They identified that one RNA helicase, DDX6, works alongside ADAR1 to tip the balance towards destabilising dsRNA.

DDX6 was found to preferentially bind to dsRNA within the cytoplasm, where it repressed ADAR1-mediated editing of A:C mismatches. It did this by directly interfering with the editing process, rather than non-selectively unwinding the dsRNA, or acting through its other established pathways. This meant that DDX6 prevents ADAR1 from stabilising dsRNA, but only weakly hinders the A-to-I editing of A-U pairs. Once these A-U pairs are edited by ADAR1, the dsRNA becomes effectively destabilised and remains hidden from MDA5. This means that ADAR1 and DDX6 cooperatively destabilise endogenous dsRNA structures to dampen interferon response.

DDX6: the cloak that keeps cancer hidden

Due to several reasons, cancer cells frequently accumulate high levels of endogenous dsRNA. To cope with the elevated dsRNA load, cancer cells have adopted several mechanisms, such as increasing ADAR1 levels, to ensure that these dsRNA remain “hidden”. Consequently, a hallmark of many cancers is the suppression of interferon signalling and the evasion of immune surveillance.

Like ADAR1, the team found that DDX6 is also elevated across several cancer types. Further examinations of DDX6 levels and interferon-stimulated genes (ISGs) in tumour cells of liver, oesophageal and colorectal cancer patients hinted that DDX6 may play a role in suppressing interferon signalling in cancer. Importantly, removal of DDX6 from tumour cells in in vivo models triggered dsRNA accumulation and MDA5 sensing, subsequently leading to interferon activation, increased immune cells in tumours, and reduced tumour growth. This confirms that DDX6 indeed plays a role in hiding tumours from the immune system, and that depleting DDX6 could make these tumours more ‘visible’ to the immune system.

A promising, if early, target

Efforts to exploit the body’s own immune system so that it can detect and remove cancer cells is an ongoing pursuit in laboratories around the world. New targets and therapies are always desired, and ultimately, DDX6’s repression of ADAR1 dsRNA editing represents a promising mechanism that could targeted by immunotherapies to potentially unmask cancer cells and leave them vulnerable to an immune system attack.