Recognizing the invaders – recognition of hybrids by the innate immune system

Arun Mankan works on DNA:RNA hybrids (photo by A. Mankan)

Within their publication “Cytosolic RNA:DNA hybrids activate the cGAS – STING axis“ (EMBO Journal 33, 2937-46 (2014)) Mankan et al. showed that hybrids of RNA and DNA are able to activate the cytosolic DNA sensor cGAS and thus induce a type I Interferon response.

We interviewed Arun Mankan, who is a Postdoc at the Institute of Molecular Medicine, about his work on RNA:DNA hybrids.


Arun, recently you showed that not only cytosolic DNA or RNA can be sensed by innate immune receptors, but also that DNA:RNA hybrids are able to induce an antiviral immune response. How did you come up with the idea that such nucleic acid hybrid molecules could also be sensed by our immune system?

It was many things coming together at the same time. First, Prof Veit Hornung (Director of the Institute of Molecular Medicine) was involved in the identification of cytosolic DNA and RNA sensors. RNA:DNA hybrids were known to exist for a long time. But the question whether an immune response is induced by them had remained unanswered.

Secondly, at the time I started my Postdoc in Bonn, a publication in Cell showed that in HIV infection not all cells were productively infected with the virus.

In some cells, replication is actually blocked, and nucleic acids could accumulate and induce a pro-inflammatory response in these non-productively infected cells.

When we think about retroviruses, we associate RNA:DNA hybrids with them. So, we wondered if RNA:DNA hybrids had a role in this scenario. Finally, there is a fatal genetic disease called Aicardi-Goutières syndrome (AGS). One of the genes linked with this disease is RNASEH2, which is a ribonuclease that degrades the RNA part of RNA:DNA hybrids. AGS patients show very high levels of interferon. So it appeared suspicious that RNA:DNA hybrids might have a role in this disease.

Since we were curious about these questions, we decided to study the immune stimulatory capacity of cytoplasmic RNA:DNA hybrids.


What were the bottlenecks and problems you faced when solving this question?

Initially, the project started very smoothly. However, we soon encountered a very critical technical problem. We needed the purest form of RNA:DNA hybrids because single stranded DNA or single stranded RNA can also induce immune responses. And more crucially contaminating double-stranded DNA or RNA can also be recognized by the cell.

Initially, we started with in vitro reverse transcribed RNA:DNA hybrids. But in control experiments with different nucleases we realized that our sample was not 100% pure and contaminated with double-stranded DNA, which very potently could activate the cytosolic sensor cGAS.

After trying several different approaches, we finally used commercial homopolymers of polyrA and polydT. When annealing them, we observed homogenous and stable RNA:DNA hybrids.


How exactly are RNA:DNA hybrids recognized by the innate immune system?

By generating knockout cells for key adaptor proteins, involved in dsDNA and dsRNA recognition, we identified cyclic GMP-AMP synthase (cGAS) as the intracellular receptor for RNA:DNA hybrids. Further, synthetic hybrids interact with recombinant cGAS in vitro and produce cyclic GMP-AMP (cGAMP), the second messenger synthesized by cGAS upon activation. Structural analysis of RNA:DNA hybrids, based on published structures of single-stranded RNA and DNA in silico showed that the structure of RNA:DNA hybrids rather resembles dsDNA than dsRNA. Recognition of double-stranded DNA by cGAS is known to be sequence independent and is just based on the 3-dimensional structure of double-stranded DNA.

In addition to cGAS, dsDNA within the cytoplasm is also recognized in a sequence-independent manner by the immune receptor AIM2. Activation of the AIM2 inflammasome results in the secretion of processed IL-1b. In a follow-up study we could show that RNA:DNA hybrids are also recognized by AIM2 and induce a pro-inflammatory immune response via IL-1b.

Other published data shows that RNA:DNA hybrids are recognized by the innate immune receptors TLR9 and NLRP3. This shows that other receptors also identify the presence of cytoplasmic RNA:DNA hybrids.


What relevance and impact do you think such DNA:RNA hybrid molecules have on the immune response and in which circumstances could they play an important role?

When we think of RNA:DNA hybrids we think of infection with retroviruses, and other viruses or bacteria which form RNA:DNA hybrids or diseases such as Aicardi-Goutières syndrome.

From a pathogenic point of view, when a pathogen infects a host cell there is the question how much of RNA:DNA hybrid is present compared to, for example, double stranded DNA. We know that DNA and RNA are very potent stimuli for the innate immune system. Thus, the immune response to hybrids will depend on their relative abundance.

As already mentioned, RNA:DNA hybrids might accumulate in non-productively HIV infected cells, but so far, this still remains to be shown. However, for enterohemorhagic E.coli (the bacteria we all remember well from the broad media coverage in 2011) one study reveals RNA:DNA hybrids as potent inducers of an immune response.

In Aicardi-Goutières syndrome resulting from RNASEH2 mutation, it has been suggested that RNA:DNA hybrids must accumulate in these patients and induce an immune response. But, so far no one has been able to prove that. Nevertheless, this is a disease were RNA:DNA hybrids may have a causative role.

Overall, we demonstrated that RNA:DNA hybrids are potent inducers of anti-viral responses. But this also reveals that there are several elements of pathogens and probably some we still do not know much about that also induce immune responses. In the end, it is always important to delineate the range of ligands that can in principle be detected by the immune system. As experience has taught us, it might take another couple of years to identify the corresponding pathogens. In this sense, we jumped a little ahead using synthetic ligands instead of pathogens to study this phenomenon.


Author: Klara Höning