The case of the edited boy

The promises and challenges of gene therapy

In February of this year, 6-month-old KJ Muldoon received his first dose of CRISPR gene editing to treat a rare liver disease. This marks the first application of gene therapy in people. KJ had a metabolic disorder called carbamoyl phosphate synthetase 1 (CPS1) deficiency. A single nucleotide mutation in his CPS1 gene meant that ammonia, a normal byproduct of metabolism, was building up in his body and, if left untreated, would cause severe brain damage and death. To treat (and potentially cure) KJ’s disease, researchers engineered a gene therapy composed of three components: mRNA to code for an editor to make the gene edit, guide RNA to direct this editor, and lipid nanoparticles to deliver it all to the cells in the liver. After the guide and mRNA were transported into the liver cells, the editor was assembled by cellular machinery and directed by the guide RNA to the faulty DNA where it could correct the mutation. KJ’s cells that received this edit are now able to produce functional CPS1 protein, allowing his liver to properly degrade and dispose of ammonia.

Why should we care about the treatment of a boy afflicted with a disease that affects less than one in a million babies? Because the conventional wisdom with genes is that what you’re born with is what you get. We may be able to treat a man’s symptoms, but until now, we’ve never been able to actually fix the genetic causes of his disease.

KJ’s case was a unique one that lent itself particularly well to treatment. He only needed to edit a few of his liver cells (around 10%) to produce enough CPS1 to break down the ammonia. KJ’s problematic cells were in the liver, which is easily targetable by lipid nanoparticles that delivered the editing machinery. He was diagnosed early in his life, before ammonia had a chance to build up to toxic levels and cause irreversible brain damage. Most importantly, KJ carried only one mutation in one gene.

If KJ’s disease was an orchestra, his mutated CPS1 gene was the drunk violinist, who’s loud playing disrupted the music of the entire group. Replace the violin player and you’d restore the harmony. However, most common diseases that we struggle with like diabetes, heart disease, and schizophrenia are polygenic, meaning thousands of genes each play small roles in their cause. Polygenic diseases are like a jet-lagged orchestra who’ve just flown in on the red-eye for a show. The violins are too sharp, the percussion is off-tempo, and the horns come in late. Individually these musicians wouldn’t spoil the show but the sum of their small mistakes produce a failed performance.

Polygenic diseases are not amenable to gene therapies. First, we often don’t know what genes are truly driving the disease, resulting in a game of biological whack-a-mole. Second, fixing hundreds of genes across millions to billions of cells is beyond the capabilities of existing treatments. Many genes are secretly pleiotropic, meaning they affect multiple traits (or phenotypes from post 1). If we go around editing dozens of genes driving schizophrenia, we may be unknowingly disrupting the same genes that regulate brain development, depression, or bipolar disorder. This kind of editing may carry the tradeoff of the lobotomy: while a man’s schizophrenia symptoms might be cured, driving a sharp stake through his eyes and into the frontal lobe of his brain may also cause some unforeseen consequences. These may include leaving him in a vegetative state. I’m exaggerating here but I want to emphasise that our understanding of the gene → phenotype connection just isn’t there yet and our ability to play with single genes to influence single phenotypes is barely in its infancy (as demonstrated by the excitement around KJ’s case).

Nevertheless, KJ’s successful treatment, especially if he’s able to continually metabolise ammonia properly throughout his life, presents a paradigm shift in how we think about treating and curing disease. We can now conceptualise treating monogenic (single gene) diseases such as cystic fibrosis, sickle cell anemia, and Huntington’s disease, and can dream of eventually identifying and correcting genes driving polygenic diseases like cancer or Alzheimer’s.

As a caveat, there’s all sorts of complicating factors that I won’t delve into in this post that make implementing gene therapies incredibly difficult. These include knowing what genes to target, knowing what tissues to target them in, knowing what developmental stage to treat at, successfully delivering the editing machinery into the relevant cells (certain brain cell types are notoriously hard to penetrate), and ensuring the correct edits are made without introducing new mutations into the DNA. To provide an irrelevant but interesting analogy: there are approximately three million letters in the bible, a thousandth of the number of basepairs in the average human’s DNA. We’re hoping to scratch out and correct one letter on line 453 of page 34,352, in a book one thousand times as long as the bible, across millions to billions of bibles. The chances of flipping a few pages too far and introducing a potentially catastrophic edit are quite real (check out the Wicked Bible, whose publishers lost their license due to a simple but critical mistake). An incorrect edit would most likely be innocuous but may also cause the misfolding of proteins or eventual development of cancer. Hopefully I’m making it clear here that fixing the G side of our GxE equation is incredibly difficult, but not impossible to do.

How successful gene therapies might reshape fundamental aspects of human biology such as longevity and anatomy is unclear. My immediate bet is that certain people with access may start living measurably longer relatively soon. Even finding and correcting a handful of mutations driving cancer, inflammation, or senescence (cellular aging) could increase a person’s lifespan by several years. Knowing the genetics underlying muscle mass, obesity, metabolism, and eye color present the possibility of developing tailored gene therapies to optimise these characteristics depending on a patient’s medical needs or even tastes. Think about how steroids and lip injections have shifted body types outside the normal distribution of anatomical proportions and now imagine turning on genes that grow muscles or minimise excess fat. That being said, I think these applications are further off. Gene therapies are incredibly expensive to develop due to their personalised nature and need for rigorous safety testing. While some men may be willing to risk baldness, aggression, and increased risk of heart attack when taking supplemental testosterone, I’m guessing far fewer would be willing to take muscle growing gene therapies that we are unsure how to turn off.

I’m going to leave questions regarding cost, access, and societal repercussions of gene therapies for another post. So far, I also haven’t addressed in vitro fertilisation, or IVF, a much more commonly practiced and accepted form of gene selection. By screening for monogenic diseases, chromosomal abnormalities, or the strongest risk markers for polygenic ailments, parents are essentially bumping the gene therapies KJ was treated with a few months earlier in the process. By making selections for or against certain genes in the embryo, researchers are able to confirm that each daughter cell of that little clump of precursors that becomes a person will all have the same genes, negating questions regarding gene delivery, editing penetrance, and off-target changes. Given how parents are increasingly having fewer children, I predict that this sort of preimplantation screening through IVF will become more and more prevalent, nipping the monogenic disease problem in the bud before children with diseases like CPS1 deficiency are born.

One unique aspect of KJ’s case is that because only his liver cells were edited, KJ could still give birth to a child with CPS1 deficiency. Gene therapies like KJ’s edit his somatic cells (everything but sex cells like sperm and eggs), meaning that these edits are non-heritable and disease-free parents can give birth to diseased offspring. While this may sound bad, as a scientific community we’ve generally shied away from editing people in heritable ways which evoke the troubled history of eugenics. However, the flimsy moral boundaries (and correspondingly weak lab regulations) restricting the heritable editing of humans was blown open in 2018 with the birth of Chinese twins carrying gene edits that they may eventually pass on to their children.

Regardless, KJ’s case is an amazing one, where a boy who most likely would have died due to the misprinting of a single nucleotide has been saved and gifted a normal, disease-free life. How these therapies shift (and hopefully increase) in affordability and access will play a big role in how widely they are implemented. KJ’s treatment presents a more ethical path forward for gene editing and lays the groundwork for the eventual wide-scale implementation of gene therapies to treat diseased children.

Enjoy some more examples of environments that fit my GxE equation.