Cold comfort? Of excruciating eyeballs and frozen frogs.

7 March 2014 by Malcolm Campbell, posted in Biology, Evolution

They tell me this is cold. I don't know what cold is, because I don't freeze.” From Northern Lights by Philip Pullman (1946- )

Believe it or not, it is possible to freeze your eyeball solid.

This said, freezing your eyeball works best under one of two conditions. First, when your entire head is frozen solid. Second, when the eyeball is actually removed from its eye socket and then exposed to sub zero temperatures. Needless to say, neither circumstance is conducive to continued existence - for either your eyeball or you.

On the other hand, it is possible to freeze a part of your eye, and then continue life fairly normally.

This week, I did the experiment.

Half way through a run, at -20oC, I turned for home. Into the wind. Before long, it was hard to blink my left eye, the eye-socket was numb, and any tears froze almost instantly. With just 10 km to go, there was nothing to do but grimace and bear it.

At home, the view from the left eye was like looking through a foggy lens. Or like someone had smeared petroleum jelly on your eye. eye1

After a few minutes of warming, the pain set in.


The pain was on the surface of the eye. More intense than stinging. It made it impossible to open my eyelid.

It might be an overstatement to say that the pain was excruciating. It was definitely intense.

And no small surprise. One of the most pain-sensitive parts of the body had been frozen – the cornea.

The cornea is the part of the eye that covers the iris, the pupil, and a region called the anterior chamber. The cornea is a transparent covering that accounts for a significant portion of the eye’s optical power – almost 66% of it in fact. The cornea is important for focusing light onto the retina at the back of the eye, but it has a fixed focus.

As the outer sheath of the eyeball, the cornea also plays an important role in protecting the underlying structure. In keeping with this, the cornea is loaded with nerves to detect a variety of stimuli – touch, temperature, and chemicals. When the cornea's nerves detect some of these stimuli, particularly touch, they will invoke an involuntary closure of the eyelid – thereby protecting the eye.

So as to keep it transparent, the cornea has no blood vessels, but it is incredibly  innervated – chock full of nerves, neurons. The cornea’s nerves originate from the ophthalmic division of the trigeminal nerve. They are connected to the trigeminal nerve via 70–80 long ciliary nerves and short ciliary nerves.

Crucially, the cornea is innervated with a remarkable density of neurons that sense pain – nociceptors. Nociceptors respond to noxious, potentially damaging stimuli by sending signals to the brain. They are responsible for providing you with the sensation of pain.

The cornea has a very high density of nociceptors. It is thought that the density of pain receptors in the cornea may be 300-600 times greater than that of normal skin. If you thought that a toothache was bad, the cornea has 20-40 times higher density of nociceptors than teeth. Needless to say, the cornea is well suited to let you know if you have provided it with potentially harmful stimuli.

Amongst the stimuli that the cornea’s nerves can detect is cold.

Cold pain detected by the cornea can be debilitating. For example, four-time winner of the Iditarod Trail Sled-Dog Race, Doug Swingley, was forced out of the race in 2004 when his corneas were frozen – all as a consequence of removing his goggles during one segment of the race. eye2

While humans can detect reduction in temperatures that are just a single degree less than body temperature, changes in temperature are normally not considered noxious until they descend to less than 15oC. For humans, at temperatures less than 15oC, around 20-30 percent of cold-perceiving nociceptors are activated. At temperatures less than 0oC, all cold-responsive nociceptors are thought to be activated – and the pain of cold is intense.

Nociceptors perceive noxious temperatures by virtue of a class of proteins known as transient receptor potential cation channels, also known as TRP channels. TRP channels are embedded in the membrane that defines each neuron. As their name implies, TRP channels function to enable passage of positively charged ions, cations, from one side of the membrane to the other.

Under normal circumstances, TRP channels are closed. The channels are opened in response to an appropriate stimulus. The stimulus induces a conformational change in the proteins that make up the channels, causing the channels to open and ions the pass from outside of the neuron, through the channel, and into the interior of the neuron. Specifically, when they are stimulated, TRP channels allow calcium ions, Ca2+, and sodium ions, Na+, to enter the neuron. This influx of cations creates a signal, an action potential, in the neuron. This action potential is transmitted from neuron to neuron, so that eventually the perception of the stimulus is conveyed to the brain.

There are different kinds of TRP channels. A subset of TRP channels detect noxious temperatures. Some detect heat, while others detect cold. Two TRP channels have been implicated in the perception of cold, TRPM8 and TRPA1.

TRPM8 has the clearest role in cold perception. In humans, TRPM8 channels are opened in response to temperatures from 26oC down to 8oC. Importantly, the extent of activation of the channels increases as the temperature gets lower, thereby intensifying the signal that is perceived, and thus conveying more pain to the brain.

The role for TRPA1 as a generic cold pain receptor is less clear. The reason for this is that TRPA1 functions to condition cold sensitivity in rodents, like rats and mice, but is not activated by cold in primates, like monkeys and humans. Remarkably, a single amino acid change between the primate and rodent TRPA1 proteins accounts for this difference. Just one amino acid is all that it takes to covert the channel into one that is cold sensitive, from one that is not. What’s more, in non-mammalian species, including flies, frogs, and reptiles, TRPA1 functions to detect uncomfortably elevated temperatures. Evolution has captured the basic channel function of TRPA1 and used it as the building block for different kinds of temperature sensing depending on the organism in which is used.

The same sort of differences in context-dependent activity are apparent with TRPM8, but in a much more subtle manner.

So far, TRPM8 appears to function in cold perception across all species tested. However, the temperature ranges that invoke a TRPM8 response vary, contingent on the species in which the channel is functioning. For example, chicken TRPM8 channels are activated at temperatures as high as 35oC, and reach maximum activity at 18oC. By contrast, rat TRPM8 channels require temperatures less than 29oC to be activated, and reach their activity maximum at 10oC.

Both rats and chickens are homeotherms (also referred to as endotherms) – they maintain a constant, warm temperature – but they have had very different evolutionary trajectories, and occupy different niches. Their TRPM8-based sensitivities to cold reflect this – the sub-tropical origin of chickens versus the temperate origin of rats. The cold-induced pain that each of these animals perceives would invoke responses appropriate to their species and environment. Discomforting cold might cause chickens to migrate to a new roosting location, or to roost with a larger flock. Importantly, the instances when such discomfort would occur, over their natural range, would be infrequent. By contrast, rats, able to live in temperate zones, require a lower cold temperature to be discomforted enough to do something about it. In this way the differences in TRPM8 activity are appropriate for the species in its niche. eye3

But these differences are small, relative to those between these two homeotherms, and a species like a frog.

Frogs are ectotherms (also known as poikilotherms) – their body temperature fluctuates in accordance with the prevailing external environment. Sometimes these temperatures are very cold. Consequently, it would be disadvantageous for frogs to feel the pain of cold at too high a temperature.

In keeping with making their homes in cold climates, frog TRPM8 channels only respond to cold at much lower temperatures than homeotherms like rats and chickens. For example, even for the African clawed frog, Xenopus laevis, TRPM8 is not appreciably active until 22oC, and only reaches maximum activity at 0oC.

While it has not yet been investigated, it is tempting to speculate that some frogs may not feel pain in response to cold at all. For example, the wood frog, Rana sylvatica, has a geographical range that means that it must survive sub-zero temperatures throughout the winter months. To do so, it embraces the cold. When the temperature plummets, the wood frog freezes solid.

The wood frog is able to freeze solid using an evolutionary innovation employed by other northern denizens, including birch trees. The wood frog, like birch trees, draws water out of cells, placing it outside the cell membrane, and simultaneously elevates sugar content within the cells. This accomplishes two things: it encourages water crystals to form only on the outside of cells, where they cannot damage cellular function; while the sugars in the cells function as cryoprotectants – to both inhibit ice crystal formation, and to protect cellular structures from cold damage.

As wood frogs can go through multiple freeze-thaw cycles during the winter months, there would be a clear advantage if they were unable to feel the pain of the cold. It is tempting to speculate that wood frogs may have lost TRPM8 activity altogether – that their TRPM8-encoding genes are no longer functional for example. Alternatively, as is the case with TRPA1 in rodents, it may be that TRPM8 has been co-opted for another function in wood frogs. Another attractive possibility may be that TRPM8 still functions to perceive cold in wood frogs, but that the neurons containing TRPM8 channels are no longer nociceptors per se, but instead feed into neural circuits that induce winter-appropriate behaviour for the frogs – activity geared for hibernation, for example. These intriguing possibilities will be resolved by future research.

Even without a complete picture of TRPM8 function in wood frogs, both TRPM8 and TRPA1 provide wonderful examples of the ways in which evolution hones molecular machinery for specific functions. Whether it involves re-purposing such machinery to serve alternative functions, or that it involves tuning the machinery so that it is well matched for a specific set of circumstances, evolution crafts remarkable tools for species to perceive and respond to their environment. Importantly, TRPA1 and TRPM8 underscore the fact that it is overly facile to imagine a molecular machine as merely doing the same thing across multiple species. Rather, it is better to think of evolutionary innovation as being context dependent – with genes and their products being in each species to fulfill a specific function in the milieu of that organism.

This said, for those species where TRPM8 is functioning to let us know that cold is painful, including human runners, we would do well to heed the warning they are providing us. For some of us, our corneas will be most grateful. eye4

Images: All photographs by Malcolm M. Campbell.


Bautista DM, Siemens J, Glazer JM, Tsuruda PR, Basbaum AI, Stucky CL, Jordt S-E, & Julius D (2007) The menthol receptor TRPM8 is the principal detector of environmental cold. Nature 448: 204-208

Byard S, Wisniewski M, Li J, & Karlson D (2010) Interspecific analysis of xylem freezing responses in Acer and Betula. HortScience 45: 165-168

Chen J, Kang D, Xu J, Lake M, Hogan JO, Sun C, Walter K, Yao B, & Kim D (2013) Species differences and molecular determinant of TRPA1 cold sensitivity. Nature Communications 4: 2501 doi:10.1038/ncomms3501

McKemy DD (2007) TRPM8: The Cold and Menthol Receptor. In: Liedtke WB, Heller S, editors. TRP Ion Channel Function in Sensory Transduction and Cellular Signaling Cascades. Boca Raton (FL): CRC Press

Myers BR, Sigal YM, & Julius D (2009) Evolution of thermal response properties in a cold-activated TRP channel. PLOS ONE 4(5): e5741

Sinclair BJ, Stinziano JR, Williams CM, MacMillan HA, Marshall KE, & Storey KB (2013) Real-time measurement of metabolic rate during freezing and thawing of the wood frog, Rana sylvatica: implications for overwinter energy use. The Journal of Experimental Biology 216: 292-302


Leave a Reply

eight + 9 =