Writing code with cells and genes
When Chinese geneticist He Jiankui announced his genomic editing of embryos in November 2018, he set off a firestorm of ethical and legal arguments against modifying genes to determine a person’s characteristics or even health outcomes. The issue of editing an embryo’s DNA, said a multitude of scientists and ethics experts at the time, had hardly begun to be debated, and it was clear that the science of genome editing was getting well ahead of discussions about its consequences. New advances in biomedical engineering, however, may offer some benefits of genome editing, without touching the genome.
Just the idea of genome editing is disturbing to a lot of people, because altering DNA seems like messing with something that’s immutable, our genetic code. After all, when we talk about our basic principles or ideals, we talk out them “being in our DNA.” For many of us then, modifying DNA, even to treat or prevent serious disease, is something we should do sparingly, if at all.
To biologists, however, our genetic code is a collection of nucleic acids, simple chemicals in our bodies arranged in a certain sequence forming the basis of proteins that instruct our cells what to do and when to do it. With editing techniques like Crispr — short for clustered, regularly interspaced short palindromic repeats — altering those chemicals becomes much easier and more certain. And as a result, what we once considered difficult, impossible, or even unimaginable, is now quite doable.
Nonetheless, achieving more predictability in health outcomes is desirable, but up to now it’s a largely illusive goal. Two scientific papers published this month suggest that may be changing. The papers describe techniques to achieve push-button biological outcomes, by programming the actions of genes and cells much like computer programmers write software. Only now the results are biological instructions, not electronic signals.
From yeast cells to circuits
In the 18 April issue of the journal Science (paid subscription required), a biomedical engineering team from the labs of Caleb Bashor at Rice University in Houston and Boston University’s Ahmad Khalil describe their process for constructing circuits from simple yeast cells. The Rice-Boston researchers genetically engineer yeast cells to activate and respond to specific chemicals in highly certain ways, comparable to binary circuits in electronics.
Disclosure: I was managing editor of Science magazine’s careers section for a number of years.
Thus the technique can predict and control the yeast’s production of a specific chemical, with inputs of another chemical. The chemicals, in this case, are transcription factors, proteins that regulate the translation of genetic codes in DNA to ribonucleic acid, or RNA, providing instructions to cells.
These simple yeast cells have another key property, however: they bind to each other. For this purpose, the ties that bind are weak, but they still have enough power to enable yeast cells to assemble into more complex networks. The researchers found they could program the assembly of these yeast cells into circuit-like networks performing, for example, Boolean math operations like digital circuits.
And to prove the concept, the Rice-Boston team built these yeast networks into more complex signal processing circuits, including low-pass filters that allow for low-frequency electronic signals to move on, and band-stop filters that allow signals higher than a certain frequency to pass through.
Coding with Crispr
About a week before the Rice-Boston paper appeared, researchers from ETH Zurich, a science and engineering university in Switzerland, described its methods for biological programming in the Proceedings of the National Academy of Sciences. A team from the bioengineering lab of Martin Fussenegger uses Crispr for gene editing, but not to alter a genome.
In this case, Fussenegger and colleagues use Crispr to alter genes taken out of the genome to express a specific protein. And the presence or absence of that protein acts like a binary on-off circuit. Like the Rice-Boston team, the ETH Zurich researchers put together these simple components into circuit-like assemblies to carry out Boolean operations.
To prove their concept, the team combined these Crispr-edited genes to function like a half-adder circuit, with logic gates that perform simple addition. A half-adder adds 2 binary digits, resulting in sum and carry outputs. and 2 half-adder circuits together perform full addition functions.
And also like the Rice-Boston team, the ETH researchers took their biological programming a step further. Fussenegger’s group configured their biological circuits to run like a dual-core central processing unit. Their Crispr-CPU as they call it, has 2 independent processing components. The team edited genomes from 2 different bacteria to produce the Crispr-CPU, and tested it inside various cells, including adult stem cells taken from human bone marrow.
On a practical level, the Rice-Boston researchers believe their engineered circuits could be useful in engineered tissue for wound healing, for example, to watch for signs of infection or other problems. Likewise, the half-adder circuits designed by the ETH Zurich team could monitor patients for problems, with the outputs programmed to return either a detection signal or trigger a therapeutic response.
But Fussenegger believes their highly efficient Crispr-CPUs could open up an entirely new type of biological computing. “Imagine a microtissue with billions of cells, each equipped with its own dual-core processor,” says Fussenegger in a university statement. “Such ‘computational organs’ could theoretically attain computing power that far outstrips that of a digital supercomputer, and using just a fraction of the energy.”