Doing science can involve a variety of different experiments. Some experiments involve complex mechanisms that are tedious and costly to perform, perhaps with an added bonus of laboursome analysis. Others are conceptually very easy to grasp and can be done swiftly. For instance, it is not very hard to imagine that having a muscle-related disease may affect your ability to move fast. If you then take a stimulus that give a predictable response, you can do an experiment to test whether healthy individuals respond quicker than those with the illness. The answer to this may seem obvious, but perhaps the disease also makes the patients feel stressed, making them more alert. Maybe the disease mainly affects slow-twitch muscle fibres, or maybe you do not know how exactly it works yet. And if it is an age-dependent disease which becomes apparent as the patient become older, how early would you be able to detect a difference? Imagine now that you have some fish that tend to develop Parkinson’s Disease, and some that stay healthy, and that touching the head causes them to swim away. If you record their movements, you suddenly have an experiment which can be used to learn more about Parkinson’s Disease.
Parkinson’s Disease causes movement-related symptoms such as tremors, stiffness, and slowness, and often also affects both cognitive abilities and mood. The disease becomes apparent with ageing, becomes progressively worse, and is non-curable. It is caused by loss of dopaminergic neurons, in a region of the brain called substantia nigra. It is one of the most common neurodegenerative diseases in the world, affecting millions globally. Much research has therefore been done on this disease, and although most cases are sporadic, some are also hereditary. The gene park7 is one of the genes associated with parkinsonism, and it encodes the protein DJ-1. DJ-1 has multifaceted functions and is important in handling oxidative stress. Animal models in which DJ-1 is disrupted by techniques such as CRISPR/Cas9 are important for learning more about DJ-1, and DJ-1 deficient animals have been known to develop Parkinson related symptoms. The use of zebrafish for this purpose is common as they are easy to keep and breed, have see-through embryos and can be genetically modified.
Imagine now that you have one zebrafish line in which DJ-1 has been knocked out, and one wildtype line. For older individuals you have seen a variety of motor and non-motor effects, but now you are curious if you can see any changes already at the larval stage. Then there are some things to consider: Firstly, how old should the fish to be tested be? Secondly, how are you going to test them? Then, how are you going to analyse the results? These are the questions that I had to consider when planning this experiment.
Zebrafish develop quickly and will hatch already at 2 days post fertilisation (dpf), but the ‘eggshell’ (chorion) can also be removed earlier. This process is however stressful for the fish, and tedious to perform. After 5 dpf they start to feed and will thereafter require applying to the animal welfare authorities. So, in practice I could test them between 2 and 5 dpf. I first opted to test them at 4 dpf, before quickly realising that the constant swimming at that point would make things way too difficult. I then tried 3 dpf larvae, which worked fine as they mainly stay still, but can still move quickly when touched.
The next question was how to test them. I had been given a camera and a computer which I could use. To get the computer and camera to work took much longer than expected, partly because of the lack of internet on the computer, and because the camera was way fancier than I was used to. With some help, I finally got it working and got a bright screen to put under the fish. I then found a soft plastic needle I could use to gently touch the fish on the fish. Now everything was ready, and I was ready for my first test experiment. I recorded the response of some fish and soon discovered that the video files generated were enormous, and that the PC would not be able to handle many videos.
Although I did only get a few videos of larvae, they could still be used to plan how to track the larvae and analyse the data. I played around a bit with Python and ImageJ, until I discovered a far easier method: befriending an IT-genius. I sent some of the test videos to him, and voilà! there was now a program we could use. Also, we got a new hard disk which could store these massive files.
We were now ready do a proper experiment, and successfully poked 10 wildtype and 10 park7 knockout larvae and tracked their movements. The positional data was used to calculate maximum acceleration, and the results showed … no difference between the wildtypes and the knockouts. This was however not unexpected as the symptoms of Parkinson’s Disease appear as the patients age. Additionally, this experiment was done in parallel with a locomotion experiment, and with an experiment exploring the protein content of dopaminergic neurons. The fact that no changes were observed in the touch-evoked response experiment is useful as no matter what findings are done in the in the other experiments, we know that they can still produce rapid movements.
However, we still wondered if their response would differ if they were first exposed to oxidative stress, since DJ-1 is important in handling this. We therefore repeated the experiment but treated the larvae with the neurotoxin MPP+ 24 hours before doing the experiment. This has just been done, and now we have to patiently wait until the data is analysed.
A: 2 dpf zebrafish larvae. Only some have hatched. B: Adult zebrafish in their holding tanks. C: The experimental setup. A small petridish was filled with water and placed on a light pad, and under a camera. The camera recorded the movement as the fish were individually touched on the head with a soft plastic needle.
Simen Mannsåker – 12.12.2023