Where do we put all our data? This question is becoming increasingly urgent, now that the capacity of current data storage methods cannot keep up with the explosive production of data. Moreover, the conventional data carriers (discs, tapes) have a limited shelf life and the whole system is very energy consuming.   

Intact  

That explains the growing interest in developing organic molecules as data carriers, and particularly DNA seems very attractive in that respect. After all, the molecule is fully evolved for this function: all the genetic information of an organism is stored here. DNA has a very high information density and, under the right conditions, can remain intact for an extremely long time. ‘DNA data storage has everything to become an emergent technology,’ says Tom de Greef, professor of synthetic biology at the Dutch universities of Eindhoven and Nijmegen. ‘In the US and within the EU, large research programmes are already ongoing, but in the Netherlands, it is still a small field.’   

An important stimulus for DNA data storage research is that, in the meantime, large-scale synthesis of DNA (‘writing’) as well as rapid sequencing (‘reading’) have become commonplace. One of the big challenges right now concerns the steps between writing and reading. How to retrieve the desired and make sure you actually read the right piece of DNA? With the additional requirement that you can access that specific DNA files several times without compromising the quality of the information.   

‘DNA data storage has everything to become an emergent technology’ 

Copy  

These challenges centre boil to the copying of the stored DNA files. Sequencing requires sufficient material, which means that any DNA files first needs to be copied in high numbers, using the well-established PCR technique. But this has two major drawbacks. First, it allows you to access and read only one file at a time, because if you want to copy several DNA files at the same time, i.e. multiplex PCR, cross-reactions generate all kinds of by-products that contaminate the stored information. In addition, with PCR you always lose part of your original material and copying is never completely error-free, so the quality of the stored information quickly decreases. 

Mini-reactors   

In Nature Nanotechnology, De Greef’s group, together with researchers from Microsoft and US, British and Chinese universities, presents a potential solution in which they store DNA in temperature-sensitive, semi-permeable microcapsules. These capsules, called proteinosomes, consist of the protein bovine serum albumin (BSA) and poly(N-isopropylacrylamide (PNIPAm), a temperature-sensitive polymer widely used in adaptive hydrogel materials. At higher temperatures, required for PCR, the BSA-PNIPAm membrane will collapse and close,  thereby creating separate mini-reactors where each DNA file is copied without interfering with the other files. ‘So, you can now copy several DNA files, each in its own capsule, in parallel. You just throw in the desired primers, raise the temperature and then you have parallel PCR processes without cross-reactions,’ explains De Greef. ‘That is a lot more efficient than going through the copying process for each separate DNA file.’   

Fewer side products  

Once the copies are ready, you lower the temperature. The membranes of the capsules become permeable again and all the copies are released, after which sequencing can begin. ‘We compared this approach with bulk PCR, where copying is not done separately, and we clearly saw far fewer by-products with our approach,’ says De Greef. Thus, copying in the sealed capsules produces a much ‘cleaner’ final product.  This tackles one aspect of the challenge. But what about repeatedly reading out the DNA files? De Greef and co-workers have devised a solution for that too.   

The DNA strands — a 1Mb DNA file consists of about 65,000 double-stranded strings of DNA — are covalently bound in the capsule via a biotin linker. As a result, the original files remain intact and preserved. With each round of reading, you consult the original data again instead of a copy of a copy.   

’We were able to read out all files multiple times without loss of quality’

Original files   

‘We tried our method on 25 Mb of data, which means 25 separate DNA files, and we were able to read out all files multiple times without loss of quality. If you don’t leave the original files intact, you lose already 35 per cent of your information after three rounds of reading. We saw, by using capsules, a loss of only 0.3 per cent. So, repeated access is quite possible and that is an important condition.’ To easily extract the original files from the mixture of copied files, each capsule also contains a small magnetic particle. ‘That way we can easily just extract the original material.’ 

Was it all planned like this or did the opportunity present itself gradually? ‘The latter,’ says De Greef. ‘We had already published the microcapsules in 2019, also in Nature Nanotechnology, as an example of synthetic protocells that can communicate with each other. I then asked Bas Bögels, PhD student in my group, to find out whether we could also do PCR in such a capsule. That turned out to be possible, and when we subsequently learned that at the higher temperatures you need for PCR, the capsules started to close, we saw the idea emerging.’ So this was already the plan? ‘No, I was initially just curious about whether PCR was possible within the capsules, but there’s always a higher-order thought to such a question.’    

Colours  

The bigger question now concerns the scalability of the method. ‘We have demonstrated it for 25 files, but you would also like to know if it works for hundreds of files as well.’ Then targeted file selection also becomes important, because you probably don’t always want to copy an entire archive if you only need a few files. That’s why the team has already applied fluorescent coloured labels to the capsules, making it easy to select the ones you want. But the number of colours is limited, so De Greef will look for ways to further scale up that selection step as well. ‘I am already formulating several new research proposals’. 

Bas Bögels, et al., DNA storage in thermoresponsive microcapsules for repeated random multiplexed data access, Nature Nanotechnology (2023) [Open Access]