If you’re from the Northeast or the Midwest (like I am) then you probably know that children enjoy playing in the snow. Snowball fights, snow angels, snow forts, and catching snowflakes on their tongues, children revel in the soft, downy, cold crystals. As scientists, our idea of fun has changed—following a loss of innocence, there are adults who are more interested in how those flakes of ice form than in playing in them. In 2008, a study published in Science showed snow samples collected in areas around the world (including Antarctica) have DNA containing cells in the center of some of the flakes, and many of these cells have ice nucleating proteins1. Why are there bacteria in snowflakes? What are these ice-nucleating proteins? But most importantly, what will overprotective mothers who learn that their children are being exposed to bacteria in the snow do with this new knowledge (just kidding)?

Perplexed by these questions, I dug through the literature and found a wonderful review article which discussed the “Bioprecipitation Cycle”2. It turns out that plant biologists were the first to look into this phenomenon. Normally, pure water won’t turn into ice until about -40 °C (for those of you on the English system, that happens to be -40 °F). Even though water “freezes” at 0°C (32 °F), it can’t freeze unless there is some small particle (dust, protein, etc.) that will force water particles into the correct formation to form a crystal—these alien particles are called nucleators, because they help cause formation of (or “nucleate”) the ice crystal. With the presence of nucleators, ice can form at temperatures as balmy as -1°C for some ice nucleating proteins, and warm as -5 °C to -11 °C for other nucleators, such as pollen or inorganic molecules.

Because these bacteria have proteins that are more effective than any other molecule at making water turn to ice at cool temperatures, they catalyze formation of frost. This frost breaks open cells on the surface of the leaf, and can then feast on the released proteins and sugars. This adaptation has made several species of plant bacteria, notably Pseudomonas syringae, able to banquet on plants. If causing frost and ice formation to get at the juicy insides of plants was all the bacteria could do, that would be cool enough.
So where does snowfall and rain come in?

It turns out that not only are bacteria on plant leaves and in the soil, but they float through the air—on a windy day bacteria are whisked up into the upper atmosphere. In fact, during the warm daylight hours (10am-2pm to be precise)2, there is a net upward movement of bacteria in the atmosphere. At these altitudes bacteria are exposed to extreme weather conditions and are removed from their terrestrial food sources. Further studies have shown that there is a net downward movement of bacteria during precipitation such as rainfall. Therefore, it is in the bacteria’s best interest to somehow cause rain. The few bacteria with ice nucleating proteins are uniquely situated to seed precipitation. In the cold atmosphere they can get into a cloud and start to freeze the droplets, causing snowflakes. If this process occurs in the summer, the snowflakes will melt into rain, or during winter, the snowflakes will gracefully dance into some small child’s upturned and gaping mouth. There is plenty of evidence for this theory, as there is a reported large downward flux of bacteria with ice nucleating proteins during rainfall, and rainfall is more plentiful over irrigated and farmed areas (as more bacteria are thrown into the air due to farming activities).

Not all snowflakes are nucleated by bacteria, and there are very few species of bacteria capable of this amazing feat. But it is a mind-boggling idea to think about how versatile one protein can be—ripping open plants for food, and contributing to the precipitation cycle in an important way by causing snow and rain. It leaves hope and wonder that there are many more fantastic proteins and species waiting to be discovered.

For Further Reference:
Listen to the Radiolab Podcast on NPR.

1. Christner, B. C., Morris, C. E., Foreman, C. M., Cai, R. & Sands, D. C. Ubiquity
of biological ice nucleators in snowfall. Science 319, 1214 (2008).

2. Morris, C. E., Georgakopoulos, D. G. & Sands, D. C. Ice nucleation active bacteria
and their potential role in precipitation. J. Phys. IV 121, 87–103 (2004).


David is a graduate student in chemistry in the Alivisatos Group. He studies DNA mediated self-assembly of nanoparticles, because it’s interesting, but also because it sounds awesome. He enjoys hiking, archery, and origami (both with paper and with DNA).