|Water mint (Mentha aquatica) in our frozen pond.|
Many animals have antifreeze chemicals in their blood and tissues e.g. cold-water fish such as the largely Antarctic notothenioids and the unrelated Arctic cod (Boreogadus saida). These have various types of glycoproteins that bind to small ice crystals, preventing them from growing and/or recrystallising (and in some cases there are systems allowing resistance to the damage caused by ice crystal formation). The precise mechanisms of the various classes of antifreeze protein aren't fully understood, but research has continued (e.g. Wierzbicki et al. 2007), with the shapes and orientations of the proteins being clearly important (as in enzymes) with effects seen at the ice-water interface where ice has many different faces for potential binding. So, what's the situation in plants? In broad terms it's quite similar - many overwintering plants produce antifreeze proteins that work in a similar way to those of fish, but there are some differences.
The first of these relates to the concept of 'thermal hysteresis'. This is the difference that antifreeze proteins create between the melting and freezing points, thus inhibiting ice formation and crystal growth. In fish this effect can reduce the freezing point by up to 1.5°C, but in plants the effect is weaker. Insect antifreeze proteins, by the way, have a much stronger hysteresis effect and are not all proteins (e.g. Walters et al. 2009 who report one comprised of saccharides and fatty acids, found in an Alaskan beetle) - maybe something for a future post...
Secondly, their function appears to be more to do with inhibiting the recrystallization of ice rather than preventing its formation (Griffith & Yaish 2004). They also have a different evolutionary origin, mostly having developed from proteins involved in tackling pathogens - indeed, some retain antifungal properties. In hindsight, maybe this isn't so surprising as a protein that can bind to a pathogen might plausably have a straightforward evolutionary path to be able to bind to somthing else.
Lastly, unlike the equivalent proteins found in fish and insects, plant antifreeze proteins have multiple ice-binding points (Griffith & Yaish 2004), though it is unknown (as far as I am aware) whether this is directly adaptive or simply a left-over from its pathogen-fighting evolutionary history.
So, next time you see a healthy looking plant in wintry conditions, these are the sorts of biochemical shenanigans going on inside - and a subject where a bit of extra biophysical amd molecular research could yield genuine breakthroughs.
Griffith, M. & Yaish, M.W. (2004). Antifreeze proteins in overwintering plants: a tale of two activities. Trends in Plant Science 9(8): 399–405.
Walters, K.R., Serianni, A.S., Sformo, T., Barnes, B.M. & Duman, J.G. (2009). A nonprotein thermal hysteresis-producing xylomannan antifreeze in the freeze-tolerant Alaskan beetle Upis ceramboides. Proceedings of the National Academy of Sciences of the USA 106(48): 20210–20215.
Wierzbicki, A., Dalal, P., Cheatham, T.E., Knickelbein, J.E., Haymet, A.D.J. & Madura, J.D. (2007). Antifreeze proteins at the ice/water interface: three calculated discriminating properties for orientation of Type I proteins. Biophysical Journal 93(5): 1442–1451.