A scalable open platform for longevity intervention testing in Daphnia
Smart Tanks is a platform by Leon Peshkin which aims to crowdsource the study of aging and pharmaco-biology using small crustaceans - Daphnia - in special standard aquariums. The project implies radically open science: the knowledge instantly becomes public. We need funding to get this off the ground.
Summary:
There is a drastic shortage of high quality standardized data on life- and health-span effects of interventions in model organisms. We aim at setting up a self-sustained de-centralized platform which will address this need and drastically accelerate anti-aging discovery.
The platform is based on Daphnia which live in a standardized aquariums built according to a blueprint -- so that all research teams use the same conditions, apply various drugs and other interventions which potentially extend healthy life and video record responses to behavioral tests.
Based on Daphnia's behavior, one can judge whether the drug is working, while the pharmacology profile of respective drugs will reveal why it does. The recorded data instantly gets to the central server for processing.
Every experiment is reproduced many times by various groups - and voila, we collect high-quality standardized, reproducible, meaningful data on the pharmaco-biology of aging.
Get Daphnia and food (in a pet store or a standard clone in a vile by mail).
Assemble 1-liter tanks and imaging setup. Place animal colonies in tanks.
Feed the colonies, apply drugs/diet/intervention. Observe, test, record.
The ways Daphnia swim and react to light reflect their well-being. Upload videos to a Web server.
Live points on the chart
The video shows an example of how Daphnia dance. In the absence of stimuli, they wander around. Once the light is ON at the top - the animals swim towards it. These are young animals, the old and the weak ones would ignore the light. Do they not see? Do they choose not to move? This remains to be seen. It's possible to count and track daphnids automatically, to distinguish males from females, to learn to evaluate nuances of their motion. The very fact that we have a convenient visual marker allows us to automate and, therefore, scale this up using Computer Vision and Machine Learning. The result will look like this (daphnids getting drunk on 3% ethanol a few minutes after treatment).
Why is there lack of quality data on lifespan interventions ?
There are not enough data on pharmaco-biology of aging in principle. Since there is little research done. Drugs that could have been tested a long time ago are not tested for many reasons (ethical, economic). The number of anti-aging laboratories is disproportionately small compared to the complexity of the task and the amount of work that needs to be done. Search DrugAge for the effect of rapamycin in mice, or caffeine in worms to see examples of dramatic discordance in the present knowledge. Plus, there is the problem of irreproducibility: many studies in biology and medicine are not credible, since their results cannot be repeated.
There is not enough data to unleash Artificial Intelligence at the search for a cure. All the advances in AI that we see today - e.g., translation, navigation, pattern and speech recognition - were enabled by existance of huge high-quality well-annotated training data. Alas, there are no such data in biology of aging. We are fixing that.
What are Daphnia good for?
Freshwater crustaceans, widely used in toxicological testing due to their permeability to small molecules!
Save time: Conveniently, Daphnia live on average only 30 days. Moreover, for some interventions there is a chance to reduce the duration of the experiment to a couple of hours: apply the drug at a high dosage and immediately observe the behavioral change. But in order to judge the accuracy of accelerated approach, it is necessary to conduct a number of preliminary studies.
Convenient features: Sensitive to drugs. Transparent. Visible heart, macrophages. Clonal. Large progeny.
Complete genome: A small 250Mb genome is sequenced, wel annotated and CRISPR-editable.
Aging model: First, daphnia also age (the probability of their death increases with age). Secondly, at least one mechanism associated with aging in daphnia overlaps with other model organisms and even, probably, with humans: a moderate reduction in calories prolongs their life.
Convenient to dose the drug: Studies of drugs in worms use unphysiologically high doses to overcome natural xenobiotic pathways. Flies mainly feed as larva, it is impossible to control the amount of drugs that gets into them. In contrast, Daphnia is sensitive to small concentrations.
Save money: A one liter aquarium can comfortably accommodate 100 daphnia. For comparison, we would place only one fish in the same volume. In addition, very low concentrations of the test substance are sufficient for Daphnia (important for testing expensive drugs, which are typically the new drugs). A bunch of cheap Daphnia in a small volume of drug solution makes the experimentation very economical.
Convenient to evaluate the results: It is enough for us to observe the Daphnia, i.e., their movements (no need to open them up or make molecular measurements). This enables automation and scaling up, using video cameras and image analysis of videos from many tanks in the hands of different teams.
... also: It's a new idea, a new model organism for aging research. This has not been tried yet. Setting up an organism that is perfectly suited for the rapid testing of many drugs and their combinations is one of the primary aims in prolonging human life.
How does Daphnia compare to other organisms ?
Do we expect with the help of Daphnia to find a remedy that prolongs human life?
Yes and No. The ultimate goal is exactly that. But it's not that simple. It is unlikely that during the testing process we will directly find a "cure for old age" intervention. The plan is something like this:
1. Double-check the ITP and DrugAge results. Since some of the data is based on non-reproducible studies, the list of positive hits is likely to be reduced. Separate the wheat from the chaff but most importantly augment lifespan with rich health-span data. In parallel, look at the new drugs and more than drugs.
2. Process uniform, high-quality health-span data using Machine Learning and Computer Vision.
3. We expect to identify and characterize some moderately yet convincingly working interventions which are not the end goal. Rather they will implicate the critical molecular pathways and allow formulating specifications for the actual interventions and drug targets.
Transferring the result into human
Daphnia has many genes and cell types that are homologous to humans. Unlike C.elegans worms, Daphnia have complex internal organs, and some of the tissues under a microscope are virtually indistinguishable from, e.g., mice. Also, Daphnia reacts to many human drugs. The problem of translating the results obtained in laboratory animals to people is not unique to Daphnia and is typical for the pharmaceutical industry in principle. Any preclinical studies are carried out in animals or cell cultures. There are many types of such laboratory animals. For example, geroprotectors have often been tested on flies, worms, and mice, and a similar question arises for all of these animals. In fact, Daphnia have several competitive advantages hover all of them (see the table above). Here is a great writeup on this subject: Why drugs that work in mice don't work in humans
What are we testing?
The literature describes about 1500 statistically significant results of life extension using more than 560 different drugs in animals of 30 species, mainly worms and flies. Accordingly, there is the DrugAge database - a list of potential geroprotectors (that is, substances that can slow down the aging process). This entire database needs to be rechecked. We propose to start with this list. Interventions other than drugs, e.g., diets and light cycle. Specific library tests can be ordered and paid for using this platform. Any form of volunteering, collaborations and subcontracting is possible.
Bite-size tasks for volunteers
