Repeated freezing induces oxidative stress and reduces survival in the freeze-tolerant goldenrod gall fly,Eurosta solidaginis

Freeze tolerant insects must not only survive extracellular ice formation but also the generation of reactive oxygen species (ROS) during oxygen reperfusion upon thawing. Furthermore, diurnal fluctuations in temperature place temperate insects at risk of being exposed to multiple freeze–thaw cycles, yet few studies have examined metrics of survival and oxidative stress in freeze-tolerant insects subjected to successive freezing events. To address this, we assessed survival in larvae of the goldenrod gall fly Eurosta solidaginis, after being subjected to 0, 5, 10, 20, or 30 diurnally repeated cold exposures (RCE) to ?18 °C or a single freeze to ?18 °C for 20 days. In addition, we measured indicators of oxidative stress, levels of cryoprotectants, and total aqueous antioxidant capacity in animals exposed to the above treatments at 8, 32, or 80 h after their final thaw. Repeated freezing and thawing, rather than time spent frozen, reduced survival as only 30% of larvae subjected to 20 or 30 RCE successfully pupated, compared to those subjected to fewer RCE or a single 20 d freeze, of which 82% pupated. RCE had little effect on the concentration of the cryoprotectant glycerol (4.26 ± 0.66 μg glycerol·ng protein^?1 for all treatments and time points) or sorbitol (18.8 ± 2.9 μg sorbitol·mg protein^?1 for all treatments and time points); however, sorbitol concentrations were more than twofold higher than controls (16.3 ± 2.2 μg sorbitol·mg protein^?1) initially after a thaw in larvae subjected to a single extended freeze, but levels returned to values similar to controls at 80 h after thaw. Thawing likely produced ROS as total aqueous antioxidant capacities peaked at 1.8-fold higher than controls (14.7 ± 1.6 mmol trolox·ng protein^?1) in animals exposed to 5, 10, or 20 RCE. By contrast, aqueous antioxidant capacities were similar to controls in larvae subjected to 30 RCE or the single 20 d freeze regardless of time post final thaw, indicating these animals may have had an impaired ability to produce primary antioxidants. Larvae lacking an antioxidant response also had elevated levels of oxidized proteins, nearly twice that of controls (21.8 ± 3.2 mmol chloramine-T·mg protein^?1). Repeated freezing also lead to substantial oxidative damage to lipids that was independent of aqueous antioxidant capacity; peroxides were, on average, 5.6-fold higher in larvae subjected to 10, 20 or 30 RCE compared to controls (29.1 ± 7.3 mmol TMOP·μg protein^?1). These data suggest that oxidative stress due to repeated freeze–thaw cycles reduces the capacity of E. solidaginis larvae to survive freezing.