CDKL5 deficiency is biologically reversible, bringing much hope and creating a sense of urgency for the development of restorative treatments for CDKL5 deficiency disorder (CDD).
These are the conclusions of a publication from Zhaolan (Joe) Zhou’s lab from the University of Pennsylvania. The study was just published in the Journal of Clinical Investigation, and it is poised to become a landmark study in the field.
It has been almost 15 years since a similar publication from Adrian Bird in mice with Rett syndrome due to mutations in MECP2. Bird’s lab showed that Rett syndrome was reversible if the gene was returned to adult mice who had developed without MECP2. The study told us that it is likely never too late to rescue MECP2 deficiency, and it encouraged research into the genetic rescue of Rett syndrome in patients. The experiment (and the investigator) became so famous, that gene therapy researchers from universities and industry have been asking for the last 15 years “has anyone already done the Adrian Bird experiment for this gene?” for any genetic syndrome that they consider working on.
Joe Zhou and his colleagues just did that for CDKL5. And the response is affirmative: CDKL5 deficiency is biologically reversible.
Here is a summary of these findings, and what they mean in the context of developing curative treatments for CDD. There is also a press release from Penn here.
WHAT THE PUBLICATION SHOWS, IN LAYMAN’S TERMS
I love the elegance of this study, and I am impressed by the amount of work and exquisite diligence that went into creating this publication. This is why I feel so strongly about the importance of these results.
Terzic et al (2021) creates two types of genetically modified mice. One type is mice that are born with normal CDKL5 expression, but that can stop expressing CDKL5 when the investigator decides. The other type are born with CDKL5 deficiency, but can produce CDKL5 when the investigator decides. Essentially there is one mouse strain where scientists can decide when the disease will start, and one where scientists can decide when it will stop. And they use these mice to answer key questions about the biology behind CDKL5 deficiency.
Because the protein CDKL5 is expressed in the brain throughout life, the scientists hypothesized that perhaps it is important also throughout life, not just in early development. This is not trivial. People born with a bad copy of CDKL5 develop a disease called CDKL5 deficiency disorder (CDD) that starts very very early in life, leading to global developmental delay in addition to seizures and other complications. Because it starts so early, it looks like having enough CDKL5 is important for very early development, and maybe if you make it past all those critical early stages of post-natal development with normal expression of CDKL5 then you will always be ok even if you stop producing enough CDKL5 later in life. Short answer: you are not.
It is never too late to develop CDD. That is why the protein CDKL5 is expressed in the brain throughout life, because it is always needed.
This is what the publication shows, by letting mice develop normally and then turning off their CDKL5 gene when they are young adults. By that age, mice that are born with CDKL5 deficiency already have all their symptoms, which include cognitive, motor and behavioral deficits, as well as alterations of neuronal excitability and synapses. Same as for people with CDD. And the first important finding of the publication is that if you let mice grow up to that age completely normal, and then turn off their CDKL5 gene, they will progressively develop almost all the same symptoms of the disease (except for less motor symptoms). Pretty much as if they had been born with CDKL5 deficiency.
And that is the first conclusion: CDKL5 deficiency looks like a disease of neuronal maintenance, not so much of neuronal development. We talked about that at the 2020 CDKL5 Forum (read it here).
The second part of the study gets better: if CDKL5 is so important for the mature brain, what happens if we bring it back at that point? How much can we get back?
For this the Penn team uses the second type of mice, the ones in which they can stop CDKL5 deficiency at any age they want. And they let them grow through all those key developmental ages with CDKL5 deficiency, and they let them display all of the disease symptoms. And then, when mice are early adults, they turned on the gene so they no longer lacked CDKL5. This is the mouse version of what would happen if we already had the perfect gene therapy for CDD and we treated patients that are no longer children.
And the results are beautiful and make this study a landmark in the development of curative treatments for CDD. Barbara Terzic and Joe and their colleagues show how later-age restoration of CDKL5 expression rescues most CDD deficits. They re-start CDKL5 expression in mice that are of young adult age, that already have developed all their CDD symptoms, and then see reduction or full rescue in many motor, cognitive and behavioral readouts. Mice with CDD don’t develop easy-to-measure seizures, which is a difference with how the disease presents in people, but the Penn group was able to interrogate brain circuit hyperexcitability using an excitatory neurotransmitter similar to glutamate to “poke” neurons, and saw how the hyperexcitability that comes with CDKL5 deficiency was also rescued by reintroduction of CDKL5 expression. They also used this neurotransmitter to induce seizures in the CDKL5 deficient mice since they are prone to epilepsy, and saw how this was also rescued by reintroduction of CDKL5 expression. They even looked at the presence of receptors in synapses, which is altered by CDKL5 deficiency, and this was also restored in the adult-rescued mice. So I want to highlight again how thorough and well-done this study is: it is very clear from this study that reintroducing CDKL5 expression later in life is able to correct functional and behavioral deficits across the board, not just a few symptoms.
And that is the big beautiful conclusion of the study: the promise of disease reversal in CDD, across a broader-than-expected time window.
WAS THIS NOT OBVIOUS OR KNOWN BEFORE?
No it wasn’t.
During brain development neurons need to be born, then need to migrate to where they are needed in the brain, grow connections to the parts of the brain that they need to be connected to, and then do their job properly remaining in place until we get old and neurons start dying. In many neurodevelopmental diseases neurons might fail to develop properly, or might migrate to the wrong places, or not connect with who they should, or suffer from neuronal loss. And for all those it is too late to come with the gene after those problems have happened.
We knew that CDD looked quite good for a potential genetic rescue in older individuals. We knew the neurons where there, in the right location, that brain connections looked good, and there were no signs of neurodegeneration. At the biological level CDD looks more like a channelopathy, like Dravet syndrome (Nav1.1 channel), than a proper developmental disease.
But from thinking “the odds are good” to being able to say “at least in mice this disease is reversible” there is a long stretch. That is why this study is such a landmark.
The results from the Adrian Bird study in Rett syndrome are actually an exception. Similar studies in related diseases have failed to show such an across-the-board rescue. For example in Phelan-McDermid mice, reintroduction of the SHANK3 gene in adult mice only rescues some symptoms, and early postnatal intervention is needed for across-the-board rescue (link). And in the Angelman mice (Ube3a gene), you can only rescue many of the symptoms if you restore the gene during early development, and the authors concluded that “adult reactivation of Ube3a is only minimally efficacious as a therapeutic intervention” (link).
So with CDD starting so early in patients, and with the clinical symptoms being so severe, it was not obvious at all that CDD would look so promising for rescue in older individuals and not more similar to the mouse study for Angelman.
WHAT IT MEANS FOR GENE THERAPIES
This study means that we now know that CDKL5 deficiency is biologically reversible. Now we need to get the tools to do that in the clinic.
What these genetic rescue studies in mice have in common is that they use mice that are genetically re-programmable, where we as scientists can choose to turn on or off a particular gene defect. People is not born like that, we cannot choose to turn off their genetic disease. We need therapies for that. These could be therapies like gene therapy, where we bring a new copy of the CDKL5 gene to their neurons, or other therapies like gene editing that fixes the mutation that they have, or even protein therapies if we can supply enough CDKL5 protein to their brain.
So there are many technical possibilities to restore CDKL5 in patients, but there are also many challenges to do this. This is why developing this type of treatment takes so long. One of the main challenges is getting to all (or most) neurons in the brain. CDD is not a disorder that impacts a small brain region, and where we only need to fix it there with an injection. It actually involves the entire brain, and we would need to make sure all (or many) of those neurons have sufficient CDKL5 as a result of a treatment. This sounds simple but is tremendously difficult. Understanding how many neurons need to be rescued to have enough efficacy, and how much CDKL5 exactly do they need (is it ok with half of the amount?), and having the right therapy that can do that in a safe way… it is not easy.
But the good news is that this is a challenge for ALL neurological diseases, so scientists and companies around the world are trying to make sure we know how to make treatments that can put back a gene (or fix it) throughout the brain. This will benefit CDD and many others. And it is also good news that we already have several of these scientists and companies working on developing these treatments for CDD, and we have already seen some early results with two of these gene therapies in mice with CDD (see here).
It will only get better from here. I expect this study to attract more attention, more research and more resources, to developing curative treatments for CDD now that we know biology is on our side.
In summary:
The part of the puzzle that we could do nothing about, whether the brain has a narrow temporal window where it needs CDKL5 or not, is the one where we got very lucky. CDKL5 deficiency is biologically reversible.
The second part of the puzzle is to build those therapies, and that one is not about luck, but about science and time and money. We can do much about it.
This is like the difficulty of building a rocket to go to the moon: it is hard, but you know that your destination is waiting at the other side.
Now that we know that moon exists, that the disease is biologically reversible, all hands on deck to build that rocket.
PS1: Thank you to the Penn teams for working so many years in this study. This study had financial support from the Loulou Foundation, IFCR and several NIH grants. Science is not easy or cheap. And thank you for making the publication open access!
PS2: The 2021 edition of the CDKL5 Forum is coming up in November 1-2, and this study is only one of the several good news that will be presented at the meeting.
Ana Mingorance, PhD