Mild heat induces a distinct “eustress” response in Chinese Hamster Ovary cells but does not induce heat shock protein synthesis

The current research on cellular heat stress management focuses on the roles of heat shock proteins (HSPs) and the proteostasis network under severe stress conditions. The mild, fever-type stress and the maintenance of membrane homeostasis are less well understood. Herein, we characterized the acute effect of mild, fever-range heat shock on membrane organization, and HSP synthesis and localization in two mammalian cell lines, to delineate the role of membranes in the sensing and adaptation to heat. A multidisciplinary approach combining ultrasensitive fluorescence microscopy and lipidomics revealed the molecular details of novel cellular “eustress”, when cells adapt to mild heat by maintaining membrane homeostasis, activating lipid remodeling, and redistributing chaperone proteins. Notably, this leads to acquired thermotolerance in the complete absence of the induction of HSPs. At higher temperatures, additional defense mechanisms are activated, including elevated expression of molecular chaperones, contributing to an extended stress memory and acquired thermotolerance.


Supplementary
: The effect of heat treatment on the phosphorylation status of HSP25. CHO cells harboring the GPI-mGFP construct were incubated at a given temperature for 20 min before sample preparation for western blotting. See Supplementary Figure S9 for full length blots. Figure S3: HSP25 distribution after a second heat treatment of MEF cells. Representative images of HSP25 distribution in MEF cells at 37˚C; after a 20 min heat treatment at specified temperatures (1st heat); followed by 6 h recovery at 37˚C (rec 37˚C); and after a second, 20 min, heat treatment at specified temperatures (2nd heat).

ImFCS data analysis
The autocorrelation functions (ACFs) for every pixel were calculated using a multi-tau correlation scheme 91 . An exponential of polynomial bleach correction was used to correct data before fitting. To obtain the diffusion coefficient (D) for all pixels ACFs were fitted according to the equation below.
Where is the ACF, τ is the correlation time, N is the number of detected particles, a is the pixel size, and 5 is the 1/e 2 radius of the Gaussian approximation of the point spread function. To identify and describe the mode of membrane organization by investigating the size-dependency of diffusion coefficient, we used the Imaging FCS type of FCS diffusion law 33 . According to that, the diffusion time (τ = ) of the fluorescent probe depends on the observation area ( ?@@ ), as described by τ = ?@@ = τ 5 + ?@@ where ?@@ is the area of the membrane in which the labeled particle travels across, and is calculated by the convolution of the detection area with the point spread function. τ 5 is the intercept of the diffusion law plot on the y-axis of ?@@ /D vs. ?@@ . This parameter provides information about the diffusion confinement. The diffusion law can be plotted by using different ?@@ values that are calculated by postacquisition binning of pixels. In the case of free diffusion, D is constant regardless of ?@@ so τ = ?@@ is a straight line passing through the origin (±0.1 s) in the diffusion law graph. A heterogeneous system, where membrane domains or meshwork are present, however, allows spatial scale-dependent diffusion, which results in a remarkably different diffusion law plot with positive or negative intercepts for domain partitioning or meshwork diffusion, respectively 33 .

Lipidomics methods
Lipidomics analyses were performed on an LTQ-Orbitrap Elite instrument (Thermo Fisher Scientific, Bremen, Germany) equipped with a robotic nanoflow ion source TriVersa NanoMate (Advion BioSciences, Ithaca, NY), using chips with 5.5-µm diameter spraying nozzles. The ion source was controlled by Chipsoft 8.3.1 software (Advion). The ionization voltages were +1.3 kV and −1.9 kV in the positive and negative modes, respectively, and the backpressure was set at 1 psi in both modes. The temperature of the ion transfer capillary was 330°C. Data acquisition was performed at the mass resolution R m/z 400=240,000.
The sum of absolute mol% difference relative to control (SoamD score) values were calculated based on Tarasov et al, (2014), as follows: where mol% i,T is the mol% of membrane lipids value for lipid species i at stress temperature T, and mol% i,37 is the mol% of membrane lipids value for lipid species i at control temperature (37°C).