Can we fix hair cells?

Previously I spoke about how gene therapy could someday fix people's Connexin proteins. But that alone wouldn't restore hearing. That's because,  for most people affected by Connexin mutations, the hairs that detect noise vibrations are dead - and these hair cells don't grow back. This is why loud noises can cause permanent hearing loss - once the hair cells are damaged they cannot be replaced by new ones, unlike most cells in the body.

Hair Cells in the cochlea

Well, that's not quite true. They don't grow back in humans and other mammals, but they do in other vertebrates - birds and lizards, for example. So why is that? What are these creatures doing that we can't?
First, lets ask another question. All our cells contain the same DNA code, so why do they act differently? How does an eye cell develop differently to skin cell?
It's mostly because of Transcription Factors. These are chemicals that stick to the outside of your DNA without affecting the DNA code itself. By doing this they can switch on or off parts of your genetic code.
When you are conceived, you split off into a lump of cells called stem cells. As different transcription factors start working in these cells, they fulfill different roles, and your body starts to form. The Transcription Factor that leads to inner hair cells forming seems to be Atoh1. But after you are born, Atoh1 stops being created in your ear, so no new hair cells.
In birds it is different. When their hair cells die, the supporting cells surrounding them react. They start producing Atoh1 again, triggering DNA activity that changes these cells into hair cells, and the birds hearing gets repaired.
Scientists are looking at ways to replicate this. As with gene therapy (see previous post), the techniques are not yet being tested on humans. But some hopeful results have been seen in mice, where supporting cells have been turned into hair cells. This has been done by combining stem cells with Atoh1, and switching on some other chemicals.
I've skipped a lot of detail here - in reality there are many different factors interacting, and not all of these are fully understood. But it's good to know progress is being made!

Connexins and gene therapy

Last post, I looked quickly at how genetic treatments worked. It seems that restoring hearing through genetic therapy is still far away. But progress is being made.

From what I gather, there are 3 phases to testing gene therapies. First is in a test tube or lab dish (in vitro), next is in live creatures (in vivo); mice or other animals, and finally humans. I don't believe there are any human trials for hearing and gene therapy underway - at least I haven't found any publications. However there are a lot of animal trials out there.

Some of these use a virus called AAV to transfer genes to the ear. Viruses work by injecting their DNA into our cells. Our cells then make the proteins the virus wants. This video explains more about viruses if you are interested.

AAV is a virus which infects people but doesn't seem to do any damage, and your body doesn't seem to be bothered by it - it doesn't trigger much of an immune response. In fact there is a good chance you have already been infected by it.

AAV only needs some of its own DNA to spread itself. To use it in gene therapy, scientists keep this bit of DNA and replace the rest with whatever they want- such as GJB2, the gene for creating Connexin 26.

Testing this on mice and guinea pigs does lead to the cells creating good Connexin proteins, and restores the ear battery I described in a previous post. But this did not repair their hearing. The problem is is that many Cx26 and Cx30 mutations lead to the death of hair cells in the ear, and these don't grow back in mammals. Hair cells are the cells that detect sound and convert it to a nerve signal. So if you fix the mutation after the hair cells are dead, it's not going to make much difference.

This leads to a problem for human treatments. Most of the time, the hair cells are dead at birth, or soon after. Which would mean trying to carry out treatment when a baby is still in the womb; a lot more risky than treating adults or even children. But there is research being done on how to regenerate hair cells, meaning the treatment could work after birth.

So it looks like we wont be seeing genetic remedies for Connexin mutations any time soon. Perhaps the CRISPR technology I mentioned in my last post will change that - you can already buy CX26 CRISPR bacteria online, so I guess someone is doing research with them.

Finally, here's something not related to hearing but still interesting. Connexin 26 is actually being looked at in gene therapy to fight cancer. Scientists have come up with DNA which will kill a cancer cell. But they have difficulty getting it into them. For some reason, cancer cells don't produce much Connexin protein. Connexins are used to allow stuff to pass between cells, so if they aren't there its harder to get new DNA into the cancer cells. Using AAV, scientists stick in the GJB2 gene alongside the new cancer-killing DNA, and it has shown to be pretty effective in some cases.

However in other cases it has allowed the tumors to spread more quickly - so more research needed! Still, looks promising.

Genetic Treatments and CRISPR

My son's hearing has effectively been restored through the use of cochlear implants, and with great success so far. However I have often wondered if it would ever be possible to fix the underlying biological problems that cause the hearing loss. There is some research being done to see if gene therapy could be useful for this.

Gene therapy is where you go in and alter the DNA within a cell, to fix whatever ails it. This can either be 'knocking out' or switching off a faulty gene, replacing a gene with a healthy version, or introducing a new gene into the cell. As you can imagine, how this is done is pretty complicated and to date has been very difficult.

First off, it's no use changing the DNA of one cell. If we are trying to repair Connexin genes, you'll need to do it to a good chunk of cells in the ear - i'm guessing millions. And you want it focussed. There's no point in sending new Connexin genes into my shoulders, knees or bladder. You any as well be pissing them away, and you might actually do damage.

So you need a delivery vector. This is usually a modified bacteria or virus, which are very good at spreading themselves around specific parts of the body. But you don't want your immune system attacking the delivery vector, so you need to be careful about what you use. And you don't want the new gene going into the wrong part of your DNA - it might split another gene, causing major problems.

Because of this complexity, there are not many gene therapy treatments out there that are in use. That may be about to change however, due to something called CRISPR. This new technique allows us to neatly cut out bits of DNA we don't want, and replace them with ones we do. I'm not going to go into how it works exactly. I will however link to this snazzy (if somewhat lacking-in-content) video.

 One thing I find really fascinating about CRISPR is where it came from. The technique for cutting out specific bits of DNA evolved in bacteria. They used it to defend them against viruses - scientists just had to tweak it a bit.

Viruses don't reproduce by themselves. They get our cells to do it for them. They inject their DNA and it becomes part of that cells chromosome. Our cells own reproduction process then copies it, until the cell is full of virus DNA and literally bursts, releasing more viruses. Bacteria and viruses have been fighting against each other for billions of years (probably), and this evolutionary pressure has given some bacteria a useful tool. They are able to recognise 'foreign' DNA. They then copy it and strap it on to come chemicals. If the chemicals find other DNA that matches, they destroy it.

This was first noticed by yoghurt scientists, trying to see how bacteria used in fermentation defended themselves. Scientists have added to it by getting the chemicals to replace the destroyed DNA with DNA of their choice. I've no idea how.

The explanation above is a massive oversimplification, and as it may be clear I don't fully understand the technique. But there is tonnes of stuff online about it, and it all suggests that this is going to be a revolution for genetic treatments. It has already been used to edit DNA in some animals, and has actually been used to cure a rare genetic liver disease in mice. Some of the scientists behind it were tipped to win a Nobel prize in Chemistry in 2017, but it didn't happen. The law case over patents and who actually developed the technique probably didn't help.

Different mutations and their effects

In my last post I looked at a common mutation that causes deafness: the c.35DelG mutation. This type of mutation, where one of the DNA letters just goes missing, is called a 'frameshift deletion' mutation. Another one, 235DelC, is quite common among Asians and has pretty much the same effect.

Let's pretend that, instead of being a majestic molecule that encodes the very essence of life itself, DNA is a drab children's book. It still is only read as 3 letter words (remember codons). And it has the following line:

Pat and Ann ate ham and ran off

A frameshift mutation might delete a letter and give the following:

Pat and Ann ate ama ndr ano ffa

The sentence is now meaningless, just like the protein that such a mutation creates. But instead of deleting a letter, you could swap it with a different one:

Pat and Ann ate Pam and ran off

You know have a sentence with very different, and sinister, meaning. What are we subjecting our poor kids to? It's function has changed, and this is what happens with DNA - these mutations create proteins that do something, just not what they are meant to do.

Swapping one letter of DNA for another is called a missense mutation. When this happens with Connexin 26, you can get other problems as well as deafness. This is called Syndromic hearing loss.

The picture below shows the Connexin 26 molecule. Each of the coloured circles with a letter represents an amino acid. Scientists have noticed mutations that change a lot of different parts of the molecule. The ones that change the blue circles just cause deafness. But at the yellow parts, the mutations cause other problems too - mostly skin disorders, because that's where Cx26 is most active. And these tend to be missense mutations.

The skin problems can be severe, and can sometimes lead to loss of blood flow to fingers and toes, causing them to drop off, or to blindness.

But if these Cx26 mutations have such a big effect on skin, why does not having Cx26 leave your skin perfectly fine?

Scientists aren't sure, but they suspect it is because of 'gain of function'. The missense mutations mean that the Connexin 26 does things it isn't supposed to do. In this case it is likely it is letting more chemicals pass between cells, which might cause to much skin to grow (hyperkeratosis).

If there are no working Cx26, the skin doesn't seem to have any problems. They reckon this is because there are other Connexin proteins (like Cx30 or Cx43) that do more or less the same thing - you have redundancy there that isn't in the ear for some reason. It is possible that the missense Cx26 molecules also interfere with these other guys and stop them working properly. In genetics this is called Trans-dominance, where a mutated protein prevents healthy proteins from doing their job.

I wrote before about how Connexin molecules bunch together in groups of 6 to form the channels that connect cells. These aren't always the same Connexin molecules - Cx26 might combine with Cx30 or Cx43. These would give different channels, that let different molecules pass through. It might be that a mutated Cx26 is too eager to join up with these other Cx molecules. All the Cx43 and Cx30 proteins join up with the broken Cx26, instead of making their own channels that the skin needs.

So in this case, having a slightly-changed version of Cx26 can be worse than having none at all.

Most of this comes from a paper by Jack Lee & Thomas White, you can read it here if you fancy.
And if you want to know more about different mutation types here is a good guide.

More on mutations

So I know both of my son's GJB2 genes have mutations. What does this mean on a molecular level? Do his cells not make Connexin 26 at all? Or do they all come out wonky?

It seems that this depends on the mutation. And surprisingly enough, having wonky Cx 26 proteins might cause more problems than having none at all.

Genetics is a field of science which still holds a lot of mysteries, despite the fact that all DNA is made of just 4 chemicals. These are called A, G, C, T (for Adenine, Guanine, Cytosine, Thymine), and the whole genetic code for a person can be stored as a (rather large) collection of books with those 4 letters repeated again and again. 

Three of these letters together (called a codon) are enough to send a message to your cells. In fact that seems to be how the DNA is read, three letters at a time. Just like how a computer reads bytes, which is a group of eight 1s or 0s. There are 64 different 3-letter combos (4 x 4 x 4). 61 of these create molecules (amino acids) and the other 3 mark the end of a protein.

You can also think of it in terms of lego. Each codon (three letters) causes the cell to make a different brick, these all stick together in a specific way until you finish with your toy. Except the bricks are called amino acids and the toys are proteins. 

The c.35DelG that I carry is the most common mutation associated with hearing loss. Its ugly name actually describes a lot about it. The 'c' at the start means its a mutation in coding DNA  - this is DNA that is used specifically for making proteins. 'Del' means something has been deleted. In this case it's a Guanine, hence the 'G'. The 35 tells us where the deletion is - 35 codons in.
This turns the 35th codon into something called a stop codon - a 3 letter combo that basically tells your cell to stop making this protein. So it ends up just making the first part of Cx26, then stops.
This picture probably shows what normal Cx26 looks like:

And this one probably shows what my mutated Cx26 looks like:

Fairly shite isn't it? Looks like it does feck all. It is in fact completely useless. Mutations like this, that leave the protein doing nothing, are called 'knock out' mutations. Knock-out mutations in Cx26 seem to always cause hearing loss, with no other effects.
However different mutations can change how Cx26 works, and this leads to other issues. I'll look at this...... next time.


Terminology

• Nucleotide A molecule with a particular structure. The four building blocks of DNA (Adenine, Guanine, Cytosine, Thymine) are nucleotides.
• Codon A group of 3 nucleotides in a row on a DNA strand
• Amino Acid A molecule with a particular structure. They can be built through the use of codons and joined together to form proteins
• Protein One of the molecules constituting a large portion of the mass of every life form and necessary in the diet of all animals, composed of 20 or more amino acids linked into one or more long chains. Proteins include such specialized forms as collagen for supportive tissue, hemoglobin for transport, antibodies for immune defense, and enzymes for metabolism

Cx-Men: Rise of the mutants

Connexin 26 mutations are quite common in Caucasians. About 2% of the population carry them. As well as deafness, they can cause severe problems with skin, and in some cases blindness.  So why are they so common? Especially considering most of them are believed to be caused by a single person getting a mutation in the past, and passing it on (the Founder Effect).

Geneticists believe that for these mutations to be so common there must be some advantage to them. And they think they know of at least one. It's to do with the gap junctions; those paths that allow things to pass between cells, and are made from Connexin proteins. Turns out that sometimes it's better to keep the gates closed.

Ever since humans have been around, viruses and bacteria have been looking for new ways to attack us. Every single aspect of our physiology gets probed for weaknesses, or for some way them to get an advantage over our immune system. At least one of them, a diarrhea-causing bacteria called Shigella flexneri, manipulates gap junctions in our digestive tract to spread itself. And these gap junctions are made from Cx26.

It has been shown that people carrying Cx26 mutations are far more resistant to Shigellosis, the disease this bacteria causes. Shigellosis causes 700,000 deaths a year these days - imagine how bad it must have been before modern hygiene practices came along. Perhaps villages would be struck with a plague of Shigellosis , and only the carriers of mutated Cx26 genes would be left standing? A similar resistance has been seen for certain E. Coli infections, which cause similar intestinal problems and are extremely common. 

From a personal perspective, I rarely get any gastrointestinal illnesses. I had always assumed this was due to my parents' proclivity for foraging in the 'reduced to clear' aisles in Tesco's. The regular consumption of near-rancid meat had led to my steely constitution, or so I believed. Perhaps the true reason was my c.35delG GJB2 mutation.

Scientists have also noticed that people carrying Cx26 mutations tend to have thicker skin. It's not sure whether this provides an advantage, but skin is a barrier to infection, so who knows?

The theory that having one copy of seemingly bad genetic mutations can actually work out for you is called 'Heterozygote advantage'. For example, carriers of the mutation that causes Cystic Fibrosis are resistant to the effects of cholera, and carriers of the mutation that causes Sickle-cell anemia are resistant to Malaria. There's an article on it here if you want to know more.

Cx-Men: Origins

Mutations.
If it wasn't for mutations, we'd all still be bacteria, and my ability to type this would be severely hampered by my lack of appendages.

The GJB2 gene is the part of our DNA that creates the Connexin 26 protein. That's more or less what DNA is used for - instruction code for creating proteins. So how did my son ended up with a mutated version? Where did this mutation originate?

There's different versions of all the genes we have, which is why we are all different - for example, there are genes for blue eyes, brown eyes, grey eyes, also blonde hair, red hair, black hair, etc. Pretty much all of our physical characteristics are determined by genes, though environmental impacts have a role too, such as malnutrition, lack of exercise etc. As such there's no such thing as 'normal' and 'abnormal' genes - who decides what a 'normal' eye colour is? (some, such as Adolf Hitler, have tried). Geneticists refer to the more common functioning genes as 'wild' type, and the less common, possibly non-functional genes as mutants.

We inherit two versions of all of our genes, one from each parent. I have one 'wild' type GJB2 gene, and one with a specific mutation called c.35delG, which probably came from my mother. So does my wife. The c.35delG mutation is 'recessive'. That means you need two copies of it for it to have a notable impact. The single 'wild' GJB2 gene I have can create enough Cx26 for my hearing to work - though I could probably do with a bit more. So I carry the deafness-causing mutation, but I'm not actually deaf.

So where did this c.35delG mutation originate? There are over 300 different variations of the GJB2 gene, and most of these are considered mutations. Some are more common among Asians, some among Caucasians, some among specific Jewish populations. The c.35delG is one of the most common causes of deafness among Caucasians (such as my family), but isn't seen much in other ethnic groups. Around 1 in 50 white people carry it.

There are two reasons the mutation could be so common. The first is the 'hot-spot' theory. This proposes that the structure of the GJB2 gene has a weakness that means it keeps breaking the same way, like the ways zips always break on cheap trousers. But this wouldn't explain why it's only common among white people. If it kept breaking like that, wouldn't it be common among Africans and Asians too?

A more likely cause is what's known as the 'founder effect'. It basically says that all people carrying the mutation have a common ancestor - a single great-great-great....-great grandparent who somehow got a mutated copy of GJB2, had loads of kids and passed it on. Some geneticists think they know who here he or she was - or rather where they lived. Ancient Greece of all places, around 10,000 years ago. I think that's about 400 generations ago.

So if you have a c.35delG mutation, we are most likely related! Just like me and my wife. Though I guess everyone is related really when you think about it.

I knew very little about genetics when I started digging into this stuff. It's really fascinating, I found this book to be really good and accessible - The Gene: An Intimate History by Siddhartha Mukherjee. It's very accessible, I recommend a read of it if you have any interest.


Terminology
I'm going to start building a glossary in these blogs. I had to look up a lot of words when reading this stuff so I might save some of you the bother.

• Heterozygous This means you have one copy of a particular gene. Both me and my wife are heterozygous for the c.35delG mutation. 
• Homozygous This means you have two copies of a particular gene. My son is homozygous for the c.35delG mutation
• Genotype  The particular type of gene you have. My sons' genotype would be 35delG/35delG for the GJB2 gene
• Phenotype Your characteristic that are caused by a particular genotype. One phenotype of the 35delG/35delG mutated GJB2 genes is typically deafness
• Allele A particular form of a gene. The 35delG mutation is an allele of the GJB2 gene.[I often write gene where I should have written allele, mainly because most people don't know what an allele is.]