An Antarctic toothfish eye lens protein resists thermal stress even when extensively deamidated

Crystallins are highly stable, soluble proteins that refract light and maintain transparency in the vertebrate eye lens.
They are not replaced after early development, making them an excellent system for studying protein stability and solubility in
crowded environments. To better understand the effects of deamidation on these ubiquitous vertebrate crystallins, we investigated
a particularly extreme example, a lens protein from the long-lived Antarctic toothfish (Dissostichus mawsoni), γS1 crystallin
(DmγS1). This protein remains soluble in the crowded fish lens, maintaining its transparency even at -2  C and at concentrations
more than two-fold that of humans (nearly 1000 mg/mL), and over a comparable timescale (1, 2). As the organism ages, crystallins
accumulate oxidative damage such as deamidation of Asn and Gln sidechains, leading to aggregation and cataract. Previous
studies of human γS crystallin (HγS) have shown that extensive deamidation reduces stability and increases aggregation
propensity. Here we present the biophysical characterization of wild-type DmγS1 and variants with 3, 5, and 7 deamidation sites.
In sharp contrast to results for human γS-crystallin, increasing the number of deamidations does not significantly change the
thermal stability of DmγS1. These proteins are startlingly resistant to thermal denaturation; despite their psychrophilic origin,
they have midpoint unfolding temperatures between 56-63 C. Extensive deamidation does make the protein more vulnerable
to chemical denaturation as well as aggregation below the unfolding temperature; however, all the variants resist aggregation
well above the fish’s physiological temperature. These proteins present a useful model system for aggregation resistance in
extreme environments – most studies of protein solubility focus on unusually aggregation-prone proteins, but understanding the
underlying biophysics also requires studying extremely soluble proteins.

All data collected and mentioned either in the main text or SI have been uploaded here. The table of contents lists the contents of each folder. The data for CD, DLS, DSF, SEC-MALS, and chemical unfolding fluorescence can be interpreted using any spreadsheet software (here we used Microsoft Excel). The mass spec data can be interpreted using mass spec software (here we used MassLynx) and NMR data is presented to be used in any NMR processing software (here we used TopSpin).