Freeze-challenged insects can invoke either of two strategies. They may resist freezing altogether, using an arsenal of molecules that inhibit ice formation; good freeze-avoiders remain liquid to below –30°C. Alternatively, insects may allow or even encourage ice formation, but with strict conditions on when and where crystals form. Unexpectedly, however, the obvious threat to freeze-tolerant insects – cell damage from ice crystals – is not the only threat. They must also cope with oxygen starvation. Oxygen moves at glacial speeds through ice at the tips of tracheoles. Moreover, tracheal ventilation usually depends on movement, which is impossible if muscles are partially frozen. How do frozen insects withstand hypoxia?

Pier Morin, David McMullen and Ken Storey examined this problem using the famous freeze-tolerant fly, Eurosta solidaginis. Larval Eurosta stimulate gall formation in goldenrods, and they live and feed in their galls throughout the summer. Full-grown larvae also overwinter in their galls – a strategy fraught with peril, as galls often are not insulated by snow and can reach very low temperatures. A mid-winter Eurosta may be mostly frozen.

The team focused on Eurosta's hypoxia-inducible factor 1α (HIF-1α). The molecule is known from studies of mammals (and a few insects and other invertebrates) to play a key role in triggering and coordinating gene expression in response to hypoxia. In short, high oxygen levels promote HIF-1α degradation, whereas low oxygen levels allow persistence. If it persists, it binds to another (constitutively expressed) molecule, HIF-1β, and the complex moves into the nucleus. There it acts as a transcription factor activating genes involved in the hypoxia response by binding to a conserved sequence known as the hypoxia response element.

The team quantified levels of hif-1α transcripts in larvae exposed to either anoxia or cold to see how the transcription factor responded to hypoxia and cold conditions. In the anoxia experiments, where the team subjected larvae to pure nitrogen for a day, the insects showed a 2.8-fold rise in transcript levels; and larvae given an extra day to recover had higher levels still. Looking at the larvae's responses to cold exposure, the team found that a chill also stimulated hif-1α transcript levels. Larvae cooled to 3°C for a day showed a 1.7-fold rise in hif-1α levels, as did those subsequently frozen for a day at– 16°C. The team also examined expression levels of HIF-1α protein, and these too were higher in 3°C larvae.

But larvae at 3°C were probably not oxygen-limited, as metabolic rates would have been low and tracheal systems unobstructed by ice. So why had their hif-1α levels risen? The authors suggest that chill-induced expression anticipates freeze-induced hypoxia, providing just-in-time preparation for the metabolic disruption of oxygen deprivation.

When Eurosta are cold, the master switch clearly sHIFts from off to on. What remains is to elucidate downstream consequences for gene expression and, ultimately, for the metabolic phenotype.