What sparked the idea to create a light-activated diabetes drug?
We've known about chemicals that can be light-activated for about five to ten years now. They’ve mainly all been applied to neurons and, more specifically, the retina. Nobody has ever really looked at any tissues outside of the nervous system.
We thought that type 2 diabetes would be a good disease state for this application because diabetes often requires medical treatment. We thought that anything that could refine the treatment, would be an excellent target.
In my lab in particular we use a lot of tissue from human donors. Although we look at quite basic things such as how cells work together and the specific pathways involved, we can't use transgenic technologies such as those used in a mouse, so we're always looking at research tools where we can yield control over cells or genetic processes in a more chemical way.
How did you adapt the existing drug to be inactive during normal conditions and active under blue light?
We used a molecule called azobenzene, which has been known to have unique light properties for 200 years now and has been used more recently as what we call a photoresponsive element. We installed one of these azobenzene rings onto one of the groups in a sulfonylurea, a type of drug used to treat diabetes.
This azobenzene changes the shape of the sulfonylurea molecule in response to light, which affects how the drug interacts with its receptor to stimulate activity. In this case, the drug interacts with ATP-sensitive potassium channels in the pancreas to stimulate electrical activity and insulin release. When the light is off, the drug does not interact with the receptor but when the light is on, it activates the receptor well.
Why did you choose blue light to activate the drug?
The cells are generally sensitive to the UV range (between 300 and 400 nm), which is a sunbed kind of color. However, this isn't ideal for using in human tissue because the lower the wavelength, the harder it is for the light to penetrate tissue. Higher wavelengths are needed because the photons get scattered.
Making the therapy responsive at 450 nm rather than 340 nm was quite a challenge. We had to manipulate the azobenzene by modifying it slightly, so it would become active at higher wavelengths.
At the moment, we're creating new variants which are hopefully going to be active at even higher wavelengths. If we can get up to the 600 nm barrier, then it becomes reasonably straightforward to get light into the body.
What would be the main benefits of having a diabetes drug that could be switched on and off via light?
In the UK, around 10% of the population suffer from type 2 diabetes and it's hugely expensive to treat. The disease shortens lifespan because patients are unable to control their insulin production properly.
Insulin resistance also develops and the high circulation of glucose affects the heart, the retina and the nerves. Type 2 diabetes is a syndrome and has wide ranging effects, where secondary complications dictate how the disease affects people.
Diabetes can be problematic in terms of people not viewing it as an illness they have to do something about, like cancer, for example. It's only during the later stages that diabetes causes quite serious problems. Having said that, very few people will end up insulin dependent because you can normally manage the condition with diet, exercise, and oral medications.
Patients usually take about three oral medications, which in the UK, usually begins with a sulfonylurea because they're very effective and very cheap. Sulfonylureas increase secretion of insulin from the pancreas. A patient may then take another drug to sensitize their tissue to insulin and a further drug to help expel excess glucose from the kidney.
However, despite almost sixty years of research, we've only got maybe five or six anti-diabetic drugs in total. They all have side effects because the receptors for these drugs are not usually specific to one organ. Activation of these receptors might do different things to different organs.
When you consider each diabetes drug, there is a risk of side effects, just as there is with any other drug. We wanted to create a treatment for the pancreas where the drug is light-activated and could be switched on when it is needed and switched off when it isn’t.
In theory, you could take the sulfonylurea and it would remain inactive in the body, not binding anywhere in the brain, heart or anywhere else where it could cause side effects.
After eating a meal, you could then illuminate the drug in a specific location, with it becoming activated only there, for example in the abdomen where the pancreas is. This is an important refinement because usually the drug just circulates around the body doing whatever
The second point is that currently, a sulfonylurea stays active once taken, meaning insulin is released all the time. It is quite convenient being able to take one capsule a day, but having a lot of insulin circulating around continuously is not good. Insulin is a potent anabolic and its continuous circulation can lead to weight gain.
Also, hormones are not normally released continuously, but in response to demand and they are more effective when they are only released in this way. Furthermore, it’s not good to make the beta cells in the pancreas work all the time, because this can exhaust them.
We are aiming to address all of these issues by targeting the drug, targeting demand and controling insulin release more finely.
In theory, how would the patient activate the drug? Would the blue light penetrate through the skin?
At the moment, we're a long way off. We don't know how feasible it's going to be to illuminate the drug in the body. Theoretically, one of the advantages of this approach is that the photoresponsive element is very light sensitive and only a few photons need to enter the body for it to become activated.
Obviously, if you're just walking around or going to the beach, it won’t be activated because you need a specific wavelength, which is where the wavelength tuning aspect comes in.
We think it's possible, but there are concerns about how to get the light in there. The best way would probably be to use high-intensity LEDs because they don't produce any heat and they are very bright. These could be mounted on a patch that could be applied to the abdominal region where the pancreas lies.
How would the patient know when to turn the light and thereby the drug off?
I think there are a couple of ways you could check that. You could have small, very minimally invasive patches that would monitor the blood glucose in real time and control when the light goes on or off.
Alternatively, you could just take the drug when you need it, after a meal, and shine the light for half an hour when you know the blood glucose is going to go up. The rest of the time, the body would regulate glucose quite well so the therapy would only really need activating after a meal.
The theory is that tailoring treatment to meal times in this way would achieve more precise control of the blood glucose level.
What more needs to be done before a therapy like this is available to patients?
Quite a lot more needs doing in terms of in vivo work. Firstly, we know the therapy works in human tissue and animal tissue, but we need to show that this is going to work in a living animal.
Secondly, we need to make sure the drug has all the same pharmacokinetics and pharmacodynamics as a conventional sulfonylurea and is not excreted through different pathways. If we administer the drug and it just goes through first pass metabolism, then it's effectively useless.
We also need to make sure it's not toxic. Initial screening suggests it isn’t toxic . However, robust toxicological screening is needed to check whether it has any mutagenic effects, for example.
What complicates testing further, is that the therapy would be classified as completely novel for these purposes. It would therefore have to go through phase 1, right through to phase 3 trials. Just that in itself is hugely expensive and risky.
What we're trying to do is show people that when they're designing drugs, this is something they should seriously think about. It could be a good way of refining the activity of drugs.
How important do you think photoswitchable drugs will be in the future of diabetes management?
That all depends on how enthusiastic companies are to pursue it as a viable option.
At the moment, a new group of drugs has just been introduced for the treatment of diabetes – the incretin mimetics. Already, side effects on the gut and brain have been reported, although the latter can be beneficial because they actually make you eat less.
Nevertheless, we believe photoswitchable drugs would provide a good way of refining drug activity, a fact that should always be considered in the design of new therapies. From that point of view, we think they are quite important.
What other diseases could photoswitchable drugs be used for?
During this process, we have collaborated with a super talented chemical biologist called Dirk Trauner who is an expert in optochemistry. He currently has a vision restoration program in place and has shown that photopharmacology can be used to rescue vision in mouse models of retinitis pigmentosa. For that particular disease, such therapies are massively applicable because you could just wear a pair of glasses that shine light into the pupil, for example.
Also, Ben Feringa’s group in the Netherlands have produced photoswitchable antibiotics. This could have huge implications for bacterial resistance, some of which stems from residual drug action in the environment.
I think this collaboration between a chemist and a biologist highlights an important point in terms of what can be achieved when you bring networks within Europe together. Neither Dirk or I could have achieved this without the other’s input, so it was really a very synergistic relationship and shows how thinking outside of the box and networking with people outside of your subject can yield surprisingly good results.
Also, as basic researchers, we're equipped to look at the pancreas when it is in a pathological state, but there's a lot we don't know about how the pancreas works under normal conditions, which is a real stumbling block. We're hoping these tools will allow people to get a handle on that by helping them to manipulate how the cells signal, especially in human tissue.
We would be interested in finding more pharmaceutical partners because as basic scientists, we have no idea what would be required to drive this forward. We are eager to make those contacts in the industrial setting at the moment.
Where can readers find more information?
Here are links to more info:
Here’s a nice article summarising the field by one of the founders:
Here’s the Twitter site:
About Dr David Hodson
David Hodson is currently an RD Lawrence Research Fellow and Lecturer at Imperial College London. Previous to this, he studied Veterinary Medicine followed by a PhD at the University of Bristol, before embarking upon postdoctoral studies in the laboratory of Patrice Mollard at the CNRS, Montpellier, France.
David’s research revolves around the use of imaging, combined with genetics and chemical biology, to interrogate and decipher the population basis for hormone secretion from endocrine tissues both in vivo and in vitro.
On-going work concentrates on the role of genetics in dictating beta cell dynamics, and the design and validation of tools with which to specifically manipulate endocrine cell function (i.e. optochemicals and optogenetics).