D'Amico et al. 2002

D’Amico S, Claverie P, Collins T, Georlette D, Gratia E, Hoyoux A, Meuwis M-A, Feller G, Gerday C. 2002. Molecular basis of cold adaptation. Proceedings of the Royal Society of London B 257: 917-925.

These authors review the relationships between enzyme parameters and temperature. A trade-off exists between enzyme activity at low temperatures and thermal stability. This trade-off is driven by enzyme flexibility: more flexible enzymes are more active (lower activation energy) at low temperatures, but become denatured at lower temperatures than less flexible, more stable enzymes. Enzyme flexibility is related to the strengths and frequencies of bonds that hold enzymes in their three-dimensional shapes; greater flexibility, with weaker and/or fewer such bonds can interact with their substrates more readily, especially at low temperatures, but are denatured easily by increasing temperatures.

The flexible portion of a cold-adapted enzyme has always been found in the domain of the active site. This argues strongly against some proposed explanations for enzymes structures and temperature that rely on relaxed selection pressure and the predominance of drift at low temperatures.

High enzyme flexibility may also allow reversible denaturation, though this point is barely explored by these authors. This suggests to me that some cold-adapted organisms may be able to tolerate temporary high temperatures more readily than “mesothermic” organisms could tolerate an increase of a similar magnitude.
Enzymes that break the trade-off and have high stability, high flexibility, and high low-temperature activity may be possible in the laboratory, especially when using particular types of artificial substrates, but have so far never been found in nature.