DNA computing: researchers combine memory and computer in one system

Contents

A team from North Carolina State University and Johns Hopkins University claims to have developed the first DNA-based method that can store, retrieve, calculate, delete and repeatedly overwrite data. The scientists write in the journal Nature Nanotechnology that such functions have not previously been available based on DNA technology.

Unlike previous approaches to DNA computing, the structure chosen by the team led by Albert Keung, Associate Professor of Chemical and Biomolecular Engineering at North Carolina State University, does not destroy the DNA molecules during computation. Instead, the information encoded in DNA is read by an enzyme (RNA polymerase) as in living organisms and synthesized into a complementary RNA strand. Its sequence is then determined by conventional nanopore sequencing. This preserves the original data – comparable to copying a document before editing.

Nanopore sequencing

Nanopore sequencing is a method for DNA/RNA analysis. It uses tiny pores in a membrane through which individual nucleic acid strands are drawn. Characteristic electrical current changes are measured for each base. This allows long strands to be sequenced in real time with minimal sample preparation. In the DNA computer system described, it enables direct readout of the data transcribed into RNA.

Solving simple chess and Sudoku puzzles

To demonstrate the capabilities of their system, the team encoded three JPEG images in 2,775 different strands of DNA, each 243 nucleotides long, and read each strand ten times. The 243 nucleotides can be imagined as blocks of data which, when read in the correct order, form a file.

The scientists also used the system to solve simple chess and Sudoku problems on a 3×3 grid. To do this, they loaded each of the approximately 1,000 possible board positions into DNA microparticles and transcribed them into RNA. They then used an enzyme to eliminate all results that violate the rules of chess or Sudoku, leaving only valid solutions.

To solve a simple 3×3 Sudoku, for example, the scientists first encode the different states of the Sudoku puzzle as a sequence of DNA nucleotides:

Position 1: ACGT (encodes number 1)
Position 2: TGCA (encodes number 2)
...
Position 9: GTCA (coded number 3)

These DNA pieces are then synthesized in the laboratory.

The synthesized DNA pieces are then applied to a substrate. This stabilizes the DNA and forms a functional environment for molecular reactions (for more on this, see below). The copying process from DNA to RNA then takes place on the substrate (transcription using RNA polymerase). All synthesized RNA strands together represent all possible and impossible solutions of the Sudoku puzzle. In order to delete the invalid solutions, the researchers delete these fragments in two steps: First, they mark the unwanted fragments using complementary DNA strands. This turns the unwanted elements into food for the enzyme RNase H: it binds specifically to combined strands of RNA and DNA and breaks down the RNA strand into single nucleotides. In the final step, only the sequence of the unlabeled RNA strands is determined, i.e. the valid Sudoku solutions.

Cellulose acetate as data storage

All of the above processes could actually take place in a liquid solution. However, this considerably reduces the durability of the DNA because it is exposed to shear forces in solution. The team therefore looked for a carrier material for DNA and examined various macromolecules introduced into aqueous solutions. In the end, cellulose acetate, which forms branched colloids in aqueous solution, came out on top; it therefore belongs to the dendricolloid class of substances.

In aqueous solution, it forms highly porous colloid particles about 50 micrometers in size with a hierarchically branched structure and an enormous surface area of 200 cm² per milligram. One such particle alone binds up to one trillion DNA oligonucleotides, which corresponds to a storage capacity of 10 terabytes. The storage density is an impressive 10,000 terabytes per cubic centimeter, according to the paper. "You could put a thousand laptops’ worth of data into DNA-based storage that’s the same size as a pencil eraser," comments Albert Keung.

Long-lasting data storage

The memory made from DNA on cellulose acetate is also extremely durable. In accelerated ageing tests, the research team determined half-lives of around 6,000 years at a storage temperature of 4°C and 2 million years at -18°C.

For long-term storage, DNA is normally freeze-dried (lyophylized) and then returned to an aqueous solution (rehydrated) for reading. As this again causes shear forces under which the DNA strands break, the process cannot be repeated as often as required. Pure DNA no longer provides correct data after 60 such processes. However, cellulose acetate stabilizes it so well that it can be freeze-dried and rehydrated over 170 times.

(vza)