Fish that live in polar and subpolar seas survive temperatures that should freeze their blood by producing specialized proteins that attach to ice crystals and stop them from growing. These molecules, broadly called antifreeze proteins (AFPs) and antifreeze glycoproteins (AFGPs), lower the actual freezing point of a fish’s body fluids below the freezing point of seawater without changing the melting point. The gap between those two temperatures is what keeps the fish alive. But the story extends well beyond a single molecular trick: different fish lineages invented their antifreeze solutions independently, some species cycle production with the seasons while others run it year-round, and the very proteins that prevent freezing can introduce problems of their own.
How Antifreeze Proteins Stop Ice From Growing
The core mechanism is deceptively simple. Antifreeze proteins bind directly to the surface of tiny ice crystals that form inside a fish’s body. Once attached, they force the ice to grow only in the narrow gaps between adjacent protein molecules. That constrained growth creates curved, convex ice fronts, and curved surfaces are thermodynamically harder to expand. The result is that the ice crystal cannot grow further unless the temperature drops even lower.
1Cryobiology. The mechanism by which fish antifreeze proteins cause thermal hysteresisThis effect is called thermal hysteresis: a separation between the melting point and the freezing point. In plain terms, the melting point stays the same, but the temperature at which ice can actually grow is pushed lower. For Antarctic notothenioid fish, that gap is enough to handle seawater temperatures hovering around −1.9 °C, which is colder than the normal freezing point of their blood.
The binding itself is remarkably precise. The winter flounder’s antifreeze protein, for example, is a single helical molecule with four repeated ice-binding motifs. X-ray crystallography shows that the side chains in those motifs are either naturally rigid or locked in place by interactions with neighboring side chains, creating a flat surface that matches specific planes on an ice crystal.
2Nature. Ice-binding structure and mechanism of an antifreeze protein from winter flounderFour Distinct Molecular Solutions
Not all fish antifreeze molecules look alike. Researchers have identified at least four structurally different types, each found in different fish families. This diversity was initially surprising, because the proteins all do the same job.
Antifreeze glycoproteins (AFGPs) are the largest group by abundance in Antarctic waters. They consist of a repeating three-amino-acid backbone with a sugar molecule attached to every third residue. Type I AFPs, found in flounders and some sculpins, are small helical proteins rich in the amino acid alanine. Type II AFPs are larger, contain multiple internal crosslinks called disulfide bridges, and appear in species like sea raven and herring. Type III AFPs are mid-sized and lack any standout structural features, making them something of an oddball in the family.
3PubMed. Biochemistry of fish antifreeze proteinsThe practical takeaway is that nature has not settled on one design. Each type binds to different faces of an ice crystal, with different levels of thermal hysteresis activity. This structural diversity hints at how separately these solutions arose, a theme the evolutionary record confirms.
An Evolutionary Accident That Saved a Lineage
The origin story of AFGPs in Antarctic notothenioid fish reads like a case study in genetic improvisation. These proteins evolved from a completely unrelated gene: one that codes for a digestive enzyme called trypsinogen. The ancestral trypsinogen gene donated its front and back ends, which provided the signals for the protein to be secreted into the bloodstream. In between, a tiny nine-nucleotide sequence that coded for the three-amino-acid repeat of the antifreeze backbone was amplified over and over, creating an entirely new protein-coding region.
4PubMed. Evolution of antifreeze glycoprotein gene from a trypsinogen gene in Antarctic notothenioid fishResearchers later found chimeric genes that still encode both the original protease and the antifreeze polyprotein on the same transcript, essentially a molecular fossil capturing the transition in progress. These intermediates confirm that the AFGP gene was not imported from another organism or created from scratch but was cobbled together from spare parts of a pre-existing gene.
5Nature. Evolution of an antifreeze glycoproteinThe timing matters, too. The Southern Ocean began cooling dramatically around 10 to 14 million years ago as the Antarctic ice sheet expanded. Notothenioids that happened to produce even weak antifreeze activity gained a survival edge, and natural selection did the rest. Today, notothenioids dominate Antarctic fish diversity, making up the vast majority of species in the Southern Ocean’s coastal waters.
Same Answer, Different Origins
One of the more striking findings in antifreeze biology is that Arctic and Antarctic fish arrived at nearly identical antifreeze glycoproteins through completely independent evolutionary paths. Arctic cod produce AFGPs that are functionally and chemically very similar to those of Antarctic notothenioids, yet the two groups are not closely related and separated by tens of millions of years of evolutionary history. Molecular, morphological, and fossil evidence all point to separate ancestry.
6PubMed. Convergent evolution of antifreeze glycoproteins in Antarctic notothenioid fish and Arctic codConvergence shows up in the Type I AFPs as well. Winter flounder, grubby sculpin, and cunner all produce Type I antifreeze proteins from different genomic origins, yet the proteins share strikingly similar features: alanine-rich repeats with evenly spaced threonine residues, and even identical short motifs at the beginnings or ends of their sequences. These shared features are not inherited from a common ancestor but instead reflect the constraints ice itself imposes on any protein trying to bind to it.
7Molecular Biology and Evolution. Diverse Origins of Near-Identical Antifreeze Proteins in Unrelated Fish Lineages Provide Insights Into Evolutionary Mechanisms of New Gene Birth and Protein Sequence ConvergenceIn other words, the physical geometry of an ice crystal surface is so demanding that evolution, working independently in multiple lineages, keeps arriving at the same molecular shape. It is one of the cleanest examples of convergent evolution at the protein level in any vertebrate system.
How Fish Keep Antifreeze Proteins in Their Blood
Producing antifreeze proteins would be pointless if the kidneys immediately flushed them out. Many AFPs and AFGPs are small enough that they would pass through a normal kidney’s filtration system. Antarctic notothenioid fish solve this problem with an unusual kidney architecture: their kidneys lack the filtration units, called glomeruli, found in most other fish. Instead of filtering blood and then reabsorbing what the body needs, these aglomerular kidneys form urine entirely by secretion. Small antifreeze molecules stay in the bloodstream because they never get filtered out in the first place.
8PubMed. Renal conservation of antifreeze peptide in Antarctic eelpout, Rhigophila dearborniThe proteins do not work alone, either. Dissolved salts in the blood act synergistically with antifreeze proteins. When AFPs are dissolved in salt solutions at concentrations similar to what fish blood naturally contains, the thermal hysteresis effect is enhanced beyond what either component achieves on its own. The ions draw water molecules into their own hydration shells, reducing the amount of water available to form ice and effectively boosting the antifreeze proteins’ performance.
9Comparative Biochemistry and Physiology Part A: Molecular & Integrative Physiology. The importance of dissolved salts to the in vivo efficacy of antifreeze proteinsSeasonal Switches in Temperate Fish
Not every antifreeze fish runs its defenses year-round. Winter flounder on the coast of Long Island, for instance, ramp up production of antifreeze protein messenger RNA in fall and winter, when water temperatures and daylight hours drop, and shut it down to undetectable levels by summer.
10PubMed. Seasonal cycle and regulation by temperature of antifreeze protein mRNA in a Long Island population of winter flounderThe switch is hormonal. During the long days of summer, the pituitary gland releases growth hormone, which blocks antifreeze gene activity in the liver. As days shorten in autumn, the central nervous system suppresses growth hormone release, and the antifreeze genes turn on. Come spring, growth hormone secretion resumes and antifreeze production stops.
11PubMed. Hormonal regulation of antifreeze protein gene expression in winter flounder 12General and Comparative Endocrinology. Regulation of antifreeze protein production in winter flounder: A unique function for growth hormone
This is a clever energy-saving arrangement: making antifreeze proteins costs metabolic resources, and there is no reason to maintain them when the ocean is warm. Antarctic notothenioids, by contrast, live in water that stays near −1.9 °C year-round, so they produce antifreeze continuously. Some species in intermediate latitudes adjust their antifreeze concentrations based on depth and ambient temperature rather than season.
13Marine Ecology Progress Series. Antifreeze glycopeptides and peptides in Antarctic fish species from the Weddell Sea and the Lazarev SeaThe Hidden Cost of Never-Melting Ice
Antifreeze proteins are not a perfect defense, and their most troubling side effect was only recently understood. The same mechanism that prevents ice crystals from growing also prevents them from melting. Researchers found that AFP-stabilized ice inside Antarctic notothenioid fish resisted melting at temperatures more than 1 °C above the expected melting point, and that ice persisted inside live fish for at least 24 hours at elevated temperatures and for several days at temperatures closer to the normal range.
14PubMed Central. Antifreeze protein-induced superheating of ice inside Antarctic notothenioid fishes inhibits melting during summer warmingField measurements confirmed that seawater temperatures in the Antarctic do rise above the fish’s equilibrium melting point during most summers, but never high enough to overcome the superheating effect. The upshot is that tiny ice crystals formed during winter may never fully melt, even when conditions warm. These persistent crystals accumulate over a fish’s lifetime, and the spleen appears to collect them as if they were foreign particles. Since Antarctic notothenioids can live 20 years or longer, the cumulative ice burden may become significant.
15Journal of Experimental Biology. Animal ice-binding (antifreeze) proteins and glycolipids: an overview with emphasis on physiological functionThis amounts to a trade-off built into the system: the proteins prevent lethal freezing but also lock in small amounts of ice that the body cannot clear. It is a form of antagonistic pleiotropy, where the same trait that helps survival in one context creates a cost in another.
Cold Adaptation Beyond Antifreeze
Antifreeze proteins are the headline act, but fish in polar waters rely on a suite of additional adaptations that work alongside them. One of the most fundamental involves cell membranes. At cold temperatures, the fat molecules in a membrane become rigid and stop functioning properly. Cold-water fish compensate by incorporating more unsaturated fatty acids into their membranes, which keeps the membrane fluid at low temperatures. This process, sometimes called homeoviscous adaptation, involves specific enzymes that ramp up production of polyunsaturated fatty acids after a temperature drop.
16PubMed. Effects of temperature on the structure and metabolism of cell membranes in fishAntarctic notothenioids have gone even further. Their hemoglobin, the oxygen-carrying protein in blood, has evolved in parallel with their cold habitat. These fish carry fewer hemoglobin variants than temperate relatives, and each variant binds oxygen less tightly. In water near 0 °C, oxygen dissolves more readily, so there is less pressure to have highly efficient hemoglobin. The most extreme case is the white-blooded icefishes, a family within the notothenioids that lost hemoglobin genes entirely. They are the only vertebrates on the planet that live without functional hemoglobin. To compensate, they have enlarged hearts, greater blood volume, and lower blood cell density, all of which increase the rate of dissolved oxygen delivery.
17PubMed Central. Cold-Driven Hemoglobin Evolution in Antarctic Notothenioid Fishes Prior to Hemoglobin Gene Loss in White-Blooded Icefishes 18PubMed Central. Gene loss in Antarctic icefish: evolutionary adaptations mimicking Fanconi Anemia?
These supplementary adaptations underscore that surviving in icy water is not about one protein doing one job. It is a whole-body remodeling effort shaped by millions of years of relentless cold.
What Warming Oceans Mean for Cold Specialists
The flip side of extreme cold adaptation is extreme vulnerability to warming. Antarctic notothenioids have evolved such narrow temperature tolerances that some species die when water temperatures rise just a few degrees above their normal habitat range.
19Journal of Experimental Biology. Antarctic notothenioid fish: what are the future consequences of ‘losses’ and ‘gains’ acquired during long-term evolution at cold and stable temperatures?Their specialized traits, which were advantages in stable cold, become liabilities in a warming world. Reduced hemoglobin diversity means less flexibility to adjust oxygen transport if metabolic demands change. The inability to fully melt internal ice means that any increase in ice entry could be permanent. And the enzymes and membranes tuned for sub-zero performance may lose efficiency at higher temperatures. Researchers have noted that these fish evolved under stable, extreme cold for millions of years, and the resulting phenotypic specialization makes them particularly susceptible to ocean warming and acidification.
20Integrative and Comparative Biology. Understanding the Metabolic Capacity of Antarctic Fishes to Acclimate to Future Ocean ConditionsFor species that can regulate antifreeze seasonally, like winter flounder, a warmer ocean might simply shift the timing of antifreeze production. For year-round producers locked into the coldest waters on the planet, the margin for adaptation is razor thin.
From Fish Blood to Frozen Carrots and Anti-Icing Coatings
The properties that keep polar fish alive have attracted attention from engineers and food scientists. In the food industry, antifreeze proteins have been tested as additives to limit ice crystal growth during freezing and thawing. Large ice crystals are the main cause of texture damage, moisture loss, and flavor degradation in frozen foods. By controlling recrystallization, AFPs help preserve texture, reduce drip loss on thawing, and maintain volatile flavor compounds.
21PubMed Central. Effect of antifreeze proteins on the freeze-thaw cycle of foods: fundamentals, mechanisms of action, current challenges and recommendations for future workExperiments with synthetic antifreeze peptides applied to frozen carrots, for example, showed smaller ice crystals, less drip loss, and better retention of color, texture, and volatile compounds compared to untreated samples.
22PubMed. Effect of Antifreeze Peptide Pretreatment on Ice Crystal Size, Drip Loss, Texture, and Volatile Compounds of Frozen CarrotsMedical researchers are exploring whether the same principles could improve organ and tissue preservation. Currently, donated organs must be transplanted within hours because ice damage during cold storage degrades tissue quality rapidly. If antifreeze proteins can protect fish cells from ice damage in a living body, the thinking goes, they might extend the viability window for human organs in transit.
23PubMed Central. From the freezer to the clinic: Antifreeze proteins in the preservation of cells, tissues, and organsMaterials scientists have taken a different angle entirely. Rather than using the proteins themselves, they have designed synthetic coatings that mimic the structure of antifreeze proteins, with distinct ice-binding and non-ice-binding regions. One such coating, inspired by AFPs, pushed the ice nucleation temperature below −29 °C, slowed ice spread to a near standstill, and reduced ice adhesion to very low levels, all while being able to self-heal at −20 °C.
24PubMed Central. Inhibition of Defect-Induced Ice Nucleation, Propagation, and Adhesion by Bioinspired Self-Healing Anti-Icing CoatingsOther teams have combined the AFP-inspired approach with graphene oxide nanosheets and photothermal effects, creating coatings that not only resist icing but can actively de-ice when exposed to light.
25PubMed. Antifreeze Protein-Inspired Zwitterionic Graphene Oxide Nanosheets for a Photothermal Anti-icing CoatingThe irony is worth noting. Fish evolved these proteins under survival pressure in some of the coldest water on Earth. Humans are now borrowing the design to keep airplane wings clear of ice and frozen peas from turning to mush. The molecules were shaped by millions of years of natural selection, but their utility extends into problems their original owners never had to face.