Since the time of the telegraphs and typewriters, rats have been a fixture of laboratory research.

And rats have been crucial  in understanding some of the basics of cancer, like  how cancers form, which cells hey target and which compounds  protect us from cancers. They’ve also helped us understand  social dynamics of addiction, and we’ve even used rat models to  help us develop treatments for HIV.

In fact, rats and mice are so burrowed into lab research that it’s difficult to  imagine a future without them. But, despite all the amazing science  built on the backs of lab rats, they’re not a perfect study animal. And there are some real ethical concerns about using living animals in our experiments.

So scientists are coming up  with high tech alternatives to using these furry creatures in experiments. Is the end of the lab rat era near? Let’s try and make heads, or tails, of it all. [♪INTRO] Rats started out as experimental  animals kind of by accident.

They were just sort of … there. Living in our homes, eating our  garbage and just doing rat things. That is, until the mid 1800s, when people started to selectively  breed rats for sport and for show.

These tamer rodents then slowly  made their way into labs. French scientist J. M.

Philippeaux  conducted one of the earliest experiments with rats when  he investigated the effect of removing the rodents’ adrenal  glands way back in 1856. But it wasn’t until half a century later that lab rats became an official thing. That’s when a neurologist named  Henry H.

Donaldson started a program to breed standardized  lab rats from albino rats. He wanted a reliable set of animals to study rat growth and development. Rats are relatively small,  fairly easy to take care of, have a short gestation period and grow up quickly, so you can get lots of individuals  in a short space of time.

All of which made them an ideal animal to study. In 1909 researchers at the  institution where Donaldson worked started inbreeding rats to  create the first ever rat strain, which is a group of rats that all  have pretty much the same genome. Today, researchers have taken that idea of breeding strains to the extreme.

There are now more than 500  different strains of rat, each one tailor-made to suit  different kinds of research. Having these little, near-identical  test subjects is important, because researchers want to know  that the thing they’re testing— whether it’s a new drug or  some funky lighting conditions— is actually the thing that led to the outcome. If the test subjects vary  by size, or age, or breed, those differences could be part of the variation, not the thing they’re investigating.

Although Donaldson initially  wanted to study the rats for their own sake, he soon  realized that these whiskered creatures could serve as stand ins for people. Basically, he could study a process  or a disease in rats and apply those findings to humans because,  biologically, we’re really similar. Okay, we might not spend our  days scurrying around a burrow.

But we’re really similar at a genetic level. And researchers are even able  to make comparisons between rat ages and human ones, so you  can study what happens to rats and know how those factors may  affect similarly-aged humans. Some rats are even bred to mimic human disorders like hypertension or multiple  sclerosis, making them even more similar to us for these research areas.

That ability to create little mini-mes  for scientific research is why rats have been used in more than  1.5 million biomedical studies. And they’ve been behind  pretty much any pharmaceutical you can think of developed in the 20th century. We just keep getting better at making rat models more and more like us.

Starting in the 1980s, researchers  have been making humanized rats. Scientists start by breeding  severely immunocompromised rats, then they inject them with immature versions of some human cell, like, say, a blood cell Over weeks or months, those human  cells grow inside the rat, and you end up with the body of a rat, but  with the immune system of a human! They can do this with just about  any kind of cell they want, so they can end up with a rat with a human-like  liver, or pancreas, or whatever.

One study even went so far as to  grow full-on human skin on a rat, complete with matching human immune system, so that they could study  bacterial skin infections. I’m not sure if I should  describe what that looks like… It might sound kind of Frankenstein, but this all comes back to the goal  of having models that resemble us. But despite more than a  century of work and some huge advances in gene editing, those  models still aren’t good enough.

For example, a study from  2014 found that treatments for Alzheimer’s disease that  looked promising in animal models failed in clinical  trials 99.6% of the time. And that’s setting aside the  valid ethical concerns that many people have about using living  animals as test subjects. Which are very valid!

Which is why researchers  have started talking about moving away from using lab rats,  and lab animals in general. But before I tell you more about that, a quick ad. Thanks to our Presidents of Science for making this SciShow video possible.

These incredible patrons help us  keep the stuff you’re watching right now totally free, so that  anyone can learn from them, because education is a human right! We make these videos to cultivate  curiosity and build knowledge in more people, in more ways and more places. And the Presidents of Science  keep that vision alive through generous donations at Patreon.com/SciShow.

To join McLaren Stanley,  Charlie Stanley, and TJ Steyn in supporting all that good  stuff, you can become a patron for $2 per month, or as much as  you want to responsibly give. But our biggest ask of you is that  you just keep watching SciShow. So, we should probably just  get back to the episode.

One alternative to the modern lab rat is to use computer simulations, which are called in silico. Now, computer simulations  aren’t a totally new idea. Researchers use them to study anything from how a river flows to how far a  nuclear blast might spread.

Which is why around 2007, the big group in the US in charge of how labs run— the National Research Council— started wondering if computer modelling might come in handy for toxicity testing too. That’s where you figure out how  much of a substance would have to eat or take or be exposed to in order to have negative effects on a body. Because, of course you’d want to  know that a new drain cleaner, pesticide or pharmaceutical drug is safe.

We needed an upgrade to the industry standard because the animal models we had weren’t always great at predicting what might happen in people. So researchers gathered all  the information on things like the chemistry of how specific molecules act, biological pathways, genetics  and records of when people have reactions, and plugged  it all into an algorithm. There are now close to 20 machine learning models for testing toxicity– from ones predicting the effects  different molecules have on the heart, to ones predicting how likely particular chemicals are to mutate cells and lead to cancer.

And overall these models are … okay. A study from 2023 looked at  how well these simulations performed compared to the  so-called Six Pack tox screenings. This is a set of six tests that  toxicologists use to see the different ways that the body would react  to different forms of a chemical.

The study found that the models  could correctly predict when a chemical would be toxic  between 67% and 78% of the time. So, not great. I mean, it is a  passing grade, but like barely.

The authors do describe a few technical ways to improve the models and see the screening more as a first step before other test tube studies. But one of the other big hurdles  for these in silico techniques— and actually for all of the lab rat  alternatives we’re talking about— is validation. For labs around the world  to jump on board with using computer models, researchers  first need to show that they can actually do what they’re supposed to do the same or better than current techniques.

And they need to do that over and over again. Even then, algorithms have other drawbacks. Like, you can’t tell how or why  a drug behaves the way it does, since these systems are basically black boxes.

That’s why there are other, more  biological alternatives to the lab rat. Take organoids for example. Organoids are 3D, lab-grown mini versions of human organs that live in a petri dish.

They’re basically simple organ  copies that mimic some of the functions and structures of a  full-size, living-in-a-person organ. The idea for organoids grew  from earlier experiments that were able to cultivate clumps  of human cells in petri dishes. But the idea really took off once  scientists understood stem cells— you know, those cells that can  develop into another type of cell.

There are a couple of ways of making an organoid, basically either using a little sample of whatever organ researchers want to  create, or by using stem cells. If researchers use an organ sample, they’ll take stem cells directly from that organ. Otherwise they’ll use pluripotent stem cells, which are the blank slate cells that can become any number of different types of cells.

They place those cells in a dish that mimics the cells’ natural environment. The right chemical and  biophysical conditions can be enough for those dormant cells to not only become, say, liver cells, but also arrange themselves in a 3D structure similar to a liver! What’s amazing about organoids is that they do a lot of the stuff that organs do— secrete mucus, allow molecules  to pass in and out of them, or mimic the way diseases like Alzheimer’s or even Zika virus show up in the body.

It’s kind of like those special  rat strains we mentioned, except without a live rat! And more  similar to the real human deal. Scientists can even make  organoid models of diseases where we don’t have an animal model yet, simply by taking cells directly from patients.

And when it comes to experimentation,  organoids are much easier to poke, prod or experiment on  in a dish than organs in a body— and also, kinder. But there are some downsides. Sometimes, that simple-sounding process of making organoids goes a little wonky, and researchers don’t always end up with  organoids that are all similar.

And so far, no organoid can do everything a full-on organ inside the body can do. Partly, that’s because these  organoids do just exist inside a dish, without any blood vessels running into them or a microbiome or any of that other bodily stuff. That might be ok if, say, the drug you’re testing targets one or a couple of specific functions, but not so great if you’re trying to study something complex like Alzheimer’s.

And since they don’t have blood  vessels constantly supplying them with fresh cells to regenerate,  some organoids last just a week. But organoids have only really  been around for about 15 years, so there’s still a lot of  potential for researchers to figure out how to make them more  repeatably-similar and complex. And they might be solving at least one of those problems with a similar  tech called organs on chips, which have been around since about 2010.

At first glance they look nothing like an organ. They’re more like a clear USB stick with colourful lines running through the center. The chip itself holds two channels of fluid— one full of human organ cells, and the other full of cells from blood vessels.

The scientists who developed them  describe organs on chips as living, 3D cross sections of the major parts of an organ. Like organoids, organs on chips are designed to replicate a specific  function or feature of an organ. But unlike organoids, they’re easier to replicate en masse since the organs are  built within a set structure.

And, thanks to that blood vessel, the chips last longer than organoids do. To run an experiment, just add  something to the blood vessel— pharmaceuticals, viruses, cigarette  smoke – and watch the result. Researchers can also build  their own body, so to speak.

Pop together a chip for a liver, a brain, and a heart, and you’ve got a multi-organ system. Some of the science that’s being done with these clear biotech wonders is pretty awesome. For example, the FDA is now  using lung organ chips to study the safety and usefulness of  COVID-19 therapies and vaccines, as well as other drugs they test.

And researchers have sent brain, bone, gut and other chips up to the  International Space Station to study the effects of space on human cells and help prepare for longer trips to Mars. The chips have been used to  study everything from toxicology, to the effects of radiation,  to infectious diseases. And based on research so far,  these chips have more predictable, replicable outcomes than organoids,  and even outperform lab rats.

But even though they may have a  few advantages, organs on chips are even more honed in on a  particular function than organoids. In that way, they’re even further away from a full-bodied, full-blooded test subject. So if a body is what you  need, nothing has been able to replace the super organized, tuned  architecture of a living thing.

Which brings us to the question we started with: Is this the end for lab rats? Right now, probably not. But there are already fewer rats in labs, which may be due in part to  institutions like the FDA using organ chips and researchers around the world choosing chips or organoids for their experiments.

And the problem of not having anything that can stand in for an entire organism? Well, scientists are slowly  making progress there too. Researchers in Waterloo,  Canada have spent more than a decade building the world’s  largest computerised human brain.

And other scientists are finding ways of running trials on people while still  being safe and ethical, like by giving them microdoses  of a particular drug and using high tech imaging  techniques to see the effects. So that scurry of little rat feet echoing in the lab might be getting a little bit quieter. [♪OUTRO]