Despite its popularity, Botox - or botulinum toxin - is one of the deadliest substances on Earth, with just 160 nanograms enough, on average, to kill an 80-kg human.
That's not a problem for people who use it therapeutically in small doses - assuming it stays where it's meant to. But new research has shown for the first time that the toxin is capable of spreading between nerve cells, and moving from its injection site.
Before you freak out, in the tiny doses doctors currently use on patients, that's nothing to be worried about - there's 0.73 nanograms per 100 unit vial, and most people get around 10 units per session.
Research has shown that Botox is safe for the majority of patients, and it also has some pretty important medical benefits.
But evidence that Botox can spread along nerve cells is worthy of further investigation, seeing as the drugs only entered the market based on the idea that they couldn't spread.
"The idea was that they are safe to use, they stay where they are injected, and you don't have to worry about toxin going to the central nervous system and causing weird effects," said team leader Edwin Chapman from the University of Wisconsin-Madison.
Now, Chapman and his team have "shown unambiguously the existence of a second entry pathway that takes some of the toxin molecules to other neurons", he says.
This isn't the first time researchers have suspected that Botox can spread.
Back in 2009, the US Food and Drug Administration (FDA) added a warning to the drug information that stated, "Botulinum toxin may spread from the area of injection to produce symptoms consistent with botulism," - which is the name given to the condition that results from being exposed to too much of the toxin, or the Clostridium botulinum bacteria that produces it.
"Understand that swallowing and breathing difficulties can be life-threatening and there have been reports of deaths related to the effects of spread of botulinum toxin," the FDA cautioned.
Doctors have also seen some strange results that suggest Botox is capable of affecting other parts of the body.
"In many cases, after an injection for a disabling spasm of neck muscles called cervical dystonia, there is no change in muscle tone but the patient finds relief and is perfectly happy," said one of the researchers, Ewa Bomba-Warczak. "That result can't be explained by the local effects."
But despite this speculation, no one had ever actually caught Botox in the act of spreading between nerve cells - until now.
To figure this out, the researchers grew mouse neurons in the lab, and kept each neuron in a dish connected by small channels that allowed them to communicate via their long 'tails' called axons.
The researchers then injected Botox into one of the nerve cells, the way doctors would clinically, and watched on a molecular level where it travelled.
They watched as the botulinum toxin stopped the nerve cells from communicating with muscles and caused local paralysis. But then they also saw the toxin move to nerve cells in neighbouring wells through the axons.
"Every time one fraction of the toxin acts locally (on the first nerve cell it contacts), another fraction acts at a distance," said Chapman. "It's unknown how far they travel, which likely depends on the dose of toxin and other factors."
A lot more research needs to be done to find out exactly how and why this is happening, and to verify that it's actually happening in humans, and not just in the lab. So don't give up your Botox injections just yet.
But the good news is that if we can better understand how the toxin works on the molecular level, we might be able to find a way to stop it from spreading, while maintaining its therapeutic effects - something Chapman suggests would make the drug even more appealing to doctors.
"I have a hard time imagining that any physician is going to want to inject something they know can move about when they have an option to use something that stays put," said Chapman. "It's an exciting prospect, supplanting a $2 billion drug with a safer drug."
The research has been published in Cell Reports.