Cytoprotection by a naturally occurring variant of ATP5G1 in Arctic ground squirrel neural progenitor cells

  1. Neel S Singhal
  2. Meirong Bai
  3. Evan M Lee
  4. Shuo Luo
  5. Kayleigh R Cook
  6. Dengke K Ma  Is a corresponding author
  1. Department of Neurology, University of California-San Francisco, United States
  2. Cardiovascular Research Institute, University of California-San Francisco, United States
  3. Department of Physiology, University of California-San Francisco, United States
  4. Innovative Genomics Institute, United States

Abstract

Many organisms in nature have evolved mechanisms to tolerate severe hypoxia or ischemia, including the hibernation-capable Arctic ground squirrel (AGS). Although hypoxic or ischemia tolerance in AGS involves physiological adaptations, little is known about the critical cellular mechanisms underlying intrinsic AGS cell resilience to metabolic stress. Through cell survival-based cDNA expression screens in neural progenitor cells, we identify a genetic variant of AGS Atp5g1 that confers cell resilience to metabolic stress. Atp5g1 encodes a subunit of the mitochondrial ATP synthase. Ectopic expression in mouse cells and CRISPR/Cas9 base editing of endogenous AGS loci revealed causal roles of one AGS-specific amino acid substitution in mediating cytoprotection by AGS ATP5G1. AGS ATP5G1 promotes metabolic stress resilience by modulating mitochondrial morphological change and metabolic functions. Our results identify a naturally occurring variant of ATP5G1 from a mammalian hibernator that critically contributes to intrinsic cytoprotection against metabolic stress.

eLife digest

When animals hibernate, they lower their body temperature and metabolism to conserve the energy they need to withstand cold harsh winters. One such animal is the Arctic ground squirrel, an extreme hibernator that can drop its body temperatures to below 0°C. This hibernation ability means the cells of Arctic ground squirrels can survive severe shortages of blood and oxygen. But, it is unclear how their cells are able to endure this metabolic stress.

To answer this question, Singhal, Bai et al. studied the cells of Arctic ground squirrels for unique features that might make them more durable to stress. Examining the genetic code of these resilient cells revealed that Arctic ground squirrels may have a variant form of a protein called ATP5G1. This protein is found in a cellular compartment called the mitochondria, which is responsible for supplying energy to the rest of the cell and therefore plays an important role in metabolic processes.

Singhal, Bai et al. found that when this variant form of ATP5G1 was introduced into the cells of mice, their mitochondria was better at coping with stress conditions, such as low oxygen, low temperature and poisoning. Using a gene editing tool to selectively substitute some of the building blocks, also known as amino acids, that make up the ATP5G1 protein revealed that improvements to the mitochondria were caused by switching specific amino acids. However, swapping these amino acids, which presumably affects the role of ATP5G1, did not completely remove the cells’ resilience to stress. This suggests that variants of other genes and proteins may also be involved in providing protection.

These findings provide the first evidence of a protein variant that is responsible for protecting cells during the metabolic stress conditions caused by hibernation. The approach taken by Singhal, Bai et al. could be used to identify and study other proteins that increase resilience to metabolic stress. These findings could help develop new treatments for diseases caused by a limited blood supply to human organs, such as a stroke or heart attack.

Introduction

Arctic ground squirrels (AGS, Urocitellus parryii) survive harsh winter environmental conditions through hibernation. By virtue of their profound ability to suppress metabolism and core temperature, with body temperatures dropping below 0°C, AGS are known as ‘extreme’ hibernators (Barnes, 1989). Hibernation in AGS can last 7 months and is characterized by drastic (>90%) reductions in basal metabolic rate, heart rate, and cerebral blood flow (Buck and Barnes, 2000). Curiously, hibernation is interrupted periodically by interbout arousal (IBA) episodes in which temperature and cerebral blood flow normalize rapidly (Drew et al., 2004; Karpovich et al., 2009). Nonetheless, AGS suffer no ischemic injury during hibernation or reperfusion injury during an IBA. Hibernating ground squirrels are resistant to ischemic and reperfusion injuries in numerous models, including brain and heart tissues after cardiac arrest in vivo and hippocampal slice models derived from animals during an IBA (Dave et al., 2009; Quinones et al., 2016; Bhowmick et al., 2017; Bogren et al., 2014). This resilience to reperfusion injury does not depend on temperature of the animal or season (Bhowmick et al., 2017). In addition, AGS neural progenitor cells (NPCs) demonstrate resistance to oxygen and glucose deprivation ex vivo (Drew et al., 2016). Together, these studies suggest that in addition to physiological adaptations, AGS possess cell autonomous genetic mechanisms that contribute to intrinsic tolerance to metabolic stress or injury.

Proteomic and transcriptomic investigations have comprehensively catalogued the impact of season, torpor, and hibernation on cellular and metabolic pathways in several different tissues of hibernating ground squirrels, including the brain (Quinones et al., 2016; Ballinger et al., 2016; Chang et al., 2018; Gehrke et al., 2019; Hampton et al., 2013; Luan et al., 2018; Andrews, 2019; Hindle et al., 2014). Although the mechanisms underlying hibernating ground squirrel ischemia and hypothermia tolerance in the brain are not fully elucidated, studies suggest that post-translational modifications, regulation of cytoskeletal proteins, and upregulation of antioxidants play a prominent role (Bhowmick and Drew, 2017; Lee et al., 2007; Tessier et al., 2019). Gene expression profiling and bioinformatic analyses also indicate the cytoprotective contributions of mitochondrial and lysosomal pathways in adapting to hypothermia and hypoxia in ground squirrel and marmot species (Bai et al., 2019; Ou et al., 2018). In neurons differentiated from 13-lined ground squirrel (13LGS) induced pluripotent stem cells (iPSCs), Ou and colleagues found that hibernating ground squirrel microtubules retained stability upon exposure to hypothermia. The authors identified mitochondrial suppression of cold-induced reactive oxygen species (ROS) and preservation of lysosomal structure are key features of ground squirrel cytoprotection, and that pharmacological inhibition of ROS production or lysosomal proteases recapitulates the hypothermia-tolerant phenotype in human cells (Ou et al., 2018). Taken together, these studies provide important insights into pathways mediating AGS tolerance to metabolic stress. However, these studies have not focused on specific genes and proteins with cytoprotective effects uniquely evolved in hibernating ground squirrels. As such, we know very little about mechanistic details underlying genetic contribution to intrinsic stress resilience in ground squirrels.

Using a cDNA library expression-based genetic screen combined with phenotypic analyses of cell survival and mitochondrial responses to stress as compared in mouse versus AGS NPCs, we identified AGS transcripts imparting ex vivo cytoprotection against various metabolic stressors. We further use CRISPR/Cas9 DNA base editing (Koblan et al., 2018) to determine functional importance of amino acid substitutions uniquely evolved in AGS, and identified AGS ATP5G1L32 as a causal contributor to stress resilience in AGS, suggesting potential for targeting this component of ATP synthase for neuroprotective treatments.

Results

AGS neural cells exhibit marked resistance to metabolic stressors associated with improvements in mitochondrial function and morphology

When growing under identical cell culture conditions, AGS and mouse NPCs exhibit similar morphology, growth rates and expression of Nestin and Ki67, markers for proliferating NPCs (Figure 1A–B and Figure 1—figure supplement 1A-E). Although superficially indistinguishable, mouse and AGS NPCs demonstrate markedly different responses to metabolic stressors. When exposed to hypoxia (1% O2), hypothermia (31°C), or rotenone (30 µM), AGS NPCs exhibit profound resistance to cell death compared with mouse NPCs (Figure 1C), recapitulating resilient AGS phenotypes found in previous studies (Dave et al., 2009; Bhowmick et al., 2017; Bogren et al., 2014; Drew et al., 2016). Moreover, measurement of in vitro oxygen consumption of AGS NPCs after sequential exposure to mitochondrial toxins demonstrates strikingly higher ‘spare respiratory capacity’ in response to FCCP (Figure 1D and Figure 1—figure supplement 1F, G), indicating a greater metabolic reserve for stressors (Nicholls and Budd, 2000). Mitochondrial citrate synthase and oxidative phosphorylation (OXPHOS) enzymatic activities were similar between the two species, with the exception of complex IV (Figure 1—figure supplement 1H). Interestingly, functional improvements in mitochondrial function were also mirrored by changes in mitochondrial dynamic organization following exposure to FCCP at doses that lead to mitochondrial depolarization (Figure 1E). At baseline, mouse and AGS cells had similar mitochondrial organization as evidenced by similar mean branch length and number of cells with fragmented mitochondria (Figure 1F). Following FCCP treatment, mouse cells demonstrated marked increases in mitochondrial fission with concurrent decreases in mean branch length. By contrast, AGS cells appeared largely resistant to mitochondrial fission induced by FCCP (Figure 1G). Together, these results demonstrate intrinsic differential cell survival and mitochondrial responses to metabolic stresses between mouse and AGS NPCs.

Figure 1 with 1 supplement see all
Phenotypic characteristics of Mouse and AGS NPCs.

(A) Confocal image of mouse (top) and AGS (bottom) NPCs demonstrating similar morphology and expression of Nestin (red) and Ki-67 (teal) in nearly all cultured cells of both species. (B) Mouse and AGS NPCs have similar proliferation rates expressed as mean ± SEM of 3 independent experiments where 50,000 NPCs were seeded in a 24-well cell culture plate in triplicate and counted by automated cytometer on two subsequent days (C) AGS NPCs exhibit increased cell survival when exposed to hypoxia (1%, 24 hr), hypothermia (31°C, 24 hr), or rotenone (10 μM, 16 hr). Bar graphs represent the mean ± SEM of 3 independent experiments with three replicates/condition. (D) Seahorse XF analyzer assay of cultured mouse and AGS NPCs sequentially exposed to (i) oligomycin (1 μM), (ii) FCCP (2 μM), and (iii) rotenone/antimycin (0.5 μM) showing enhanced FCCP-stimulated oxygen consumption (spare respiratory capacity). Data represents the mean ± SEM of three independent experiments with 4–6 replicates/species. (E) Relative fluorescence ± SEM of three independent experiments in triplicate each of cultured mouse and AGS NPCs loaded with TMRE (50 nM) exposed to vehicle or FCCP (1 μM) (F) Representative confocal images of mouse (left) and AGS (right) NPCs expressing the mitochondrial marker mCherry-mito7 to demonstrate mitochondrial morphology at baseline (H) and one hour following treatment with 1 μM FCCP. Scale bar represents 10 μm. (G, I) Percent of mitochondria with fragmented morphology (left panel) and the mean branch length (right panel) of mitochondrial networks of NPCs expressing mCherry-mito7. Data obtained from 30 cells/species/condition. *p<0.05; ***p<0.001.

A cDNA library expression screen identifies AGS ATP5G1 as a cytoprotective factor

To identify cytoprotective genes expressed in AGS, we constructed a normalized cDNA expression library from AGS NPCs and introduced the library to mouse NPCs by nucleofection (Bertram et al., 2012; Figure 2—figure supplement 1A-B). Screening of inserts revealed the average library insert size was 2.4 kB. To minimize false negatives due to incorrect splice isoforms, we performed screens in triplicate and maintained representation at 1000 cells/open reading frame. Two days after AGS cDNA library nucleofection, we exposed cells to hypothermia (31°C) for 3 days, hypoxia (1%) for 2 days, or complex I inhibition (rotenone) for 3 days, respectively (Figure 2A). We then isolated plasmids from surviving cells, amplified cDNA insert sequences by PCR and used next-generation sequencing to identify a total of 378 putative cytoprotective genes, three of which (Ags Atp5g1, Ags Manf, and Ags Calm1) provided cytoprotection in all three examined metabolic stress conditions (see Figure 2B and Supplementary file 1).

Figure 2 with 1 supplement see all
AGS cDNA library survival screen identifies AGS ATP5G1 as a cytoprotective factor.

(A) AGS NPC cDNA was introduced into mouse NPCs by nucleofection. Cells were screened for survival after exposure to hypoxia (1%, 48 hr), hypothermia (31°C, 72 hr), or rotenone (20 μM, 48 hr) to identify AGS cytoprotective factors. (B) Venn-diagram demonstrating the number of cytoprotective proteins identified by next-generation sequencing of plasmids isolated from cells surviving each condition of the cDNA library screen. (C) Truncated sequence alignments demonstrating key GS AA substitutions (blue highlight) for ATP5G1, one of the three proteins imparting survival in all three screens. (D) Ground squirrel-unique amino acid substitutions are plotted as a function of BLOSUM62 score and Jensen-Shannon Divergence (JSD) score. Ground squirrel-unique AA substitutions with the highest probability of functional consequence are in the denoted red quadrant (top 1% scoring of high JSD values and low BLOSUM62 scores). The red dot represents the ATP5G1L32P substitution; orange dots represent two other ATP5G1 substitutions. (E–G) Mouse NPCs expressing human ATP5G1, AGS ATP5G1, AGS ATP5G1L32P, human ATP5G1 P32L, or empty vector (EV) and exposed to 24 hr of 1% O2 (E), 31°C (F), or 20 μM rotenone (G). Cell death was determined by flow cytometry for propidium iodide and experiments are mean ± SEM of three independent experiments with three replicates/genotype/condition, *p<0.05 or **p<0.01 vs EV; δ <0.05 vs human ATP5G1.

Since a portion of mouse NPCs survived metabolic stresses even without AGS cDNA library expression or as a result of protective secreted factors, we anticipated false positive hits without cell autonomous cytoprotective effects. Thus, in this study, we focused on characterizing the nuclear-encoded mitochondrial protein AGS ATP5G1 that conferred cytoprotective effects independently confirmed under all three metabolic stress conditions (Figure 2B,E–G). ATP5G1 is one of three ATP5G isoforms making up the C-subunit of mitochondrial ATP synthases, and is regulated distinctly from ATP5G2 or ATP5G3 (Gay and Walker, 1985De Grassi et al., 2006Wigington et al., 2016). As most identified genes do not appear to be differentially expressed between mouse and AGS NPCs (Ou et al., 2018), we hypothesized that resistance to metabolic stress may be related to uniquely evolved AGS proteins. Based on multiple sequence alignment of the ATP5G1 protein family in mammals, we observed three AGS-unique amino acid substitutions and two small insertions/deletions at the N-terminal region of AGS ATP5G1, whereas the C-terminal membrane-spanning segment is largely invariant (Figure 2C and Figure 2—figure supplement 1C).

We expanded the analysis of AGS-unique amino acid substitutions to other cytoprotective protein variants identified from the screen of the AGS cDNA library. In particular, we analyzed uniquely evolved AGS proteins by comparing sequence alignments of the screened cytoprotective candidates for two species of ground squirrels (AGS and the 13LGS, Ictidomys tridecemlineatus) against nine other reference species across mammalian subclasses. We calculated the Jensen- Shannon Divergence (JSD) score, which captures sequence conservation and difference from the background amino acid distribution, and average ground squirrel-versus-other mammalian block substitution matrix (BLOSUM)−62 scores for each unique residue (Capra and Singh, 2007). High JSD and low BLOSUM62 scores indicate chemically significant amino acid substitutions, and as such potentially important functional AGS adaptations. We found that the leucine-32 residue of AGS ATP5G1 in place of the otherwise highly conserved proline is unique to hibernating ground squirrels, and on conservation analysis scored among the highest of all AGS-unique amino acid substitutions in identified cytoprotective protein candidates from our screen (Andersson et al., 1997; Figure 2D and Supplementary file 2).

The N-terminal region of ATP5G proteins can undergo cleavage, but also modulate mitochondrial function directly, by unknown mechanisms (Vives-Bauza et al., 2010). Although the three C-subunit proteins are identical in sequence, they cannot substitute for one another and are all required to constitute a fully functional C-subunit (Vives-Bauza et al., 2010; Sangawa et al., 1997). To determine the relative levels of ATP5G1, −2, and −3 in mouse and AGS NPCs, we performed qRT-PCR analysis with species and transcript-specific primers. We found that in both mouse and AGS NPCs, expression of Atp5g3 or Atp5g2 is greater than that of Atp5g1, consistent with prior reports in human and mouse tissues (Gay and Walker, 1985; Vives-Bauza et al., 2010). We found that the relative abundance of the Atp5g1 isoform is elevated nearly twofold in AGS NPCs (Figure 2—figure supplement 1D). However, the relative abundance of the mature ATP5G (subunit C) protein or oligomycin sensitivity of complex V activity, is not different in mouse and AGS cells (Figure 2—figure supplement 1E-F).

Overexpression of the AGS variant of ATP5G1 in mouse NPCs confers cytoprotection in cells exposed to hypoxia, hypothermia, or rotenone (Figure 2E–G). We found that this protective response is not present in NPCs overexpressing ATP5G1L32P. Conversely, overexpression of the human ATP5G1P32L, which mimics the wild-type AGS ATP5G1 variant, leads to enhanced cytoprotection in these conditions of metabolic stress compared to that of human ATP5G1. The ATP5G1 substitutions did not alter the mitochondrial localization of ATP5G1 when expressed in either mouse or AGS NPCs (Figure 3—figure supplements 1,2). In addition, overexpression of the AGS variant of ATP5G1 recapitulated key features of the AGS resilient mitochondrial phenotype, including increasing spare respiratory capacity and reducing mitochondrial fission with reduced fragmentation and increased branch length of mitochondria in response to FCCP (Figure 3, Figure 3—figure supplement 3A). Interestingly, NPCs expressing the AGS ATP5G1L32P variant demonstrated reduced spare respiratory capacity and increased mitochondrial fragmentation compared to the AGS ATP5G1 over-expressing NPCs. Overexpression of human ATP5G1P32L improved survival to metabolic stressors and reduced mitochondrial fragmentation, but compared to AGS ATP5G1L32P spare respiratory capacity was not significantly improved. This may indicate that improving spare respiratory capacity itself is not the sole mechanism conferring resilience to metabolic stressors. Of note, expression of AGS ATP5G1 with two other identified AGS-unique amino acid substitutions (N34D, T39P) did not affect survival of mouse NPCs exposed to hypoxia, hypothermia, or rotenone (Figure 3—figure supplement 3B–D). Together, these results reveal cytoprotective effects of AGS Atp5g1 when ectopically expressed in metabolic stress-susceptible mouse NPCs, and identify functional importance of the leucine-32 residue of AGS ATP5G1 uniquely evolved in AGS.

Figure 3 with 3 supplements see all
Overexpression of AGS ATP5G1 in mouse NPCs recapitulates AGS metabolic phenotypes, which is dependent on the uniquely evolved leucine-32.

(A) Seahorse XF analyzer assay of cultured mouse NPCs expressing human ATP5G1, AGS ATP5G1, AGS ATP5G1L32P, human ATP5G1P32L, or empty vector and sequentially exposed to (i) oligomycin (1 μM), (ii) FCCP (2 μM), and (iii) rotenone/antimycin (0.5 μM) showing increased FCCP-stimulated oxygen consumption (spare respiratory capacity) with AGS ATP5G1. Substitution of the AGS leucine-32 results in reduced spare respiratory capacity. Data represents the mean + SEM of three independent experiments with 4–6 replicates/species. (B) Relative fluorescence ± SEM of three independent experiments in triplicate each of cultured mouse NPCs stably expressing the indicated ATP5G1 construct, loaded with TMRE (50 nM), and exposed to vehicle or FCCP (1 μM) (C) Percent ± SEM fragmented mitochondria and representative confocal images (D-G) of mitochondrial networks in mouse NPCs expressing human, AGS, and mutant forms of mCherry-ATP5G1 one hour following treatment with vehicle (top panel) or 1 μM FCCP (bottom panel). Data obtained from 30 cells/condition. *p<0.05; ***p<0.001 vs. human ATP5G1; δ <0.05 vs AGS ATP5G1. Scale bars represent 5 μm.

Knock-in of AGS ATP5G1L32P alters the resilient phenotype of AGS cells

Species-specific substitutions of amino acid residues at sites deeply conserved in mammals indicate either relaxed selective constrains at the sites during evolution or potentially adaptive significance functionally specific for that species. As ectopic expression may not fully reflect endogenous functions, precise manipulation of endogenous genetic loci is required to determine definitive causal contribution of ATP5G1L32 to the metabolic resilience of AGS. Using the recently reported adenine DNA base editor (ABEmax; 22), we successfully generated AGS cell lines homozygous for ATP5G1L32P by introducing a cytosine-to-thymine substitution in the (-) strand of Ags Atp5g1 (Figure 4A,B). We isolated three clonal AGS NPC lines harboring the desired knock-in mutation (ABE KI) and two clonal lines that underwent editing and remained homozygous for the wild-type allele (ABE WT). Compared to ABEmax-treated AGS cells without successful knock-in (Figure 4—figure supplement 1A), ABE KI cell lines did not demonstrate differences in Atp5g1 mRNA expression, protein abundance, or complex V activity (Figure 4—figure supplement 1B-C, Figure 4I). However, knock-in of the L32P residue resulted in markedly reduced survival of AGS NPCs following exposure to hypoxia, hypothermia, or rotenone (Figure 4C). In addition, we found the ABE KI AGS NPCs exhibited marked reduction in ‘spare respiratory capacity’ and altered mitochondrial dynamics in response to FCCP treatment (Figure 4D–H and Figure 4—figure supplement 1C). Although overall ATP5G protein abundance is unchanged (Figure 4—figure supplement 1D-E), we used clear-native gel electrophoresis (Kovalčíková et al., 2019; Wittig and Schägger, 2009) and identified a reduced presence of ATP synthase dimers relative to the total amount of ATP synthase in ABE KI cells (Figure 4J–K). Further biochemical experiments are necessary to delineate the specific mechanisms of how the AGS leucine-32 substitution affects the assembly or stability of ATP synthase complex proteins. Nonetheless, genetic evidence in our study based on ectopic expression and specific CRISPR base editing of endogenous loci demonstrates causal roles of the AGS leucine-32 substitution in cytoprotection. Collectively, these results identify a naturally occurring cytoprotective AGS variant that contributes to cytoprotection against various metabolic stresses likely by modulating mitochondrial function.

Figure 4 with 1 supplement see all
CRISPR base editing to generate ATP5G1L32P AGS NPCs results in a partial loss of AGS metabolic resilient phenotypes.

(A) AGS ATP5G1 CRISPR base editing strategy. To create AGS cells with the human amino acid substitution at leucine-32, AGS cells transiently expressing ABEmax were nucleofected with an sgRNA (blue underline) directed toward a PAM site (green underline) on the (-) strand to target conversion of adenine to guanine, which on the (+) strand is a cytosine-to-thymine (*). (B) Sequencing data from a successfully edited clonal AGS cell line demonstrating the cytosine-to- thymine base edit resulting in the desired leucine to proline knock-in cell line. (C) AGS ATP5G1L32P (ABE KI) NPCs exhibit decreased cell survival compared to unedited AGS cells (ABE WT) when exposed to hypoxia (1%, 24 hr), hypothermia (31°C, 72 hr), or rotenone (10 μM, 16 hr). Bar graphs represent the mean ± SEM of three independent experiments with three replicates/condition. (D) Seahorse XF analyzer assay of cultured ABE KI and WT cells sequentially exposed to (i) oligomycin (1 μM), (ii) FCCP (2 μM), and (iii) rotenone/antimycin (0.5 μM) showing enhanced FCCP-stimulated oxygen consumption (spare respiratory capacity). Data represents the mean ± SEM of 3 independent experiments with 4–6 replicates/species. (E and F) Representative confocal images of ABE WT (E) and ABE KI (F) NPCs expressing the mitochondrial marker mCherry-mito7 to demonstrate mitochondrial morphology one hour following treatment with FCCP. Scale bar represents 5 μm. (G) Percent of mitochondria ± SEM with fragmented morphology, data obtained from 50 to 60 cells/genotype. (H) Relative fluorescence ± SEM of 3 independent experiments in triplicate each of cultured ABE AGS NPCs loaded with TMRE (50 nM) and exposed to vehicle or FCCP (1 μM). (I) Complex V enzymatic activity normalized to protein content of ABE WT and KI mitochondrial extracts normalized to protein content and treated with vehicle or oligomycin (1 μM). Data are the mean ± SEM of 3 independent experiments expressed as a fraction of ABE WT enzymatic activity. (J) Representative immunoblots for ATP5G (left), ATP5A (right), or citrate synthase (CS, left, input control) of clear-native gel electrophoresis of mitochondrial extracts from ABE WT and ABE demonstrate ATP synthase dimers (D) and monomers (M). (K) Quantification of ATP5G demonstrates a reduction in ATP synthase dimers relative to total ATP synthase protein (D:(D+M) ratio) in ABE KI. Data are mean ± SEM of 3 independent blots. *p<0.05; **p<0.01.

Discussion

Previous studies have indicated that hibernating organisms evolved numerous physiological and cellular mechanisms enabling survival during the stressed metabolic conditions accompanying hibernation (Bai et al., 2019; Ou et al., 2018; Ballinger et al., 2017). However, we still know little about the mechanistic details of how AGS protein-coding genetic variants contribute to intrinsic cytoprotective functions. We show that ex vivo cultured AGS NPCs can recapitulate remarkable intrinsic resilience to hypoxia, hypothermia, and other metabolic stressors. Additionally, using an unbiased cDNA expression screening and bioinformatic strategy, we identified numerous AGS transcripts and uniquely evolved AGS amino acid substitutions potentially contributing to cytoprotection. We focused on discerning the protective effect of AGS ATP5G1, a nuclear-encoded mitochondrial protein, given that it was one of only three genes identified in all three metabolic stress paradigms and the prominent mitochondrial resilience phenotype of AGS NPCs. We hypothesize that analogous to amino acid substitutions in several human proteins providing adaptive benefits (Simonson et al., 2010; Song et al., 2014; Xiang et al., 2013; Yates and Sternberg, 2013), substitutions in AGS ATP5G1 may underlie AGS adaptive mechanisms contributing to its robust cytoprotective phenotype. Using the dCas9 ABE technology, we validated a unique AGS ATP5G1L32 amino acid substitution in the N-terminal region of ATP5G1 that leads to improvements in mitochondrial physiologic parameters. Thus, our study used CRISPR base editing in non-model organism hibernator cells, for the first time to our knowledge, to identify a naturally occurring cytoprotective protein variant from AGS. CRISPR edited ATP5G1L32P did not fully abolish the metabolic resilience phenotype in AGS NPCs, indicating that other gene variants may also be involved. The robust ex vivo paradigm of AGS phenotypes established from our study makes it tractable to investigate additional gene and protein variants that contribute to the metabolic resilience phenotype in AGS. Further understanding the gene variants and mechanisms responsible for the AGS phenotype has important implications for novel neuroprotective treatments in ischemic diseases as well as promoting survival of neural stem cell grafts (Bernstock et al., 2017).

Mitochondrial metabolic dysfunction is central to ischemia and reperfusion injury. Physiologic, transcriptomic, and proteomic studies have highlighted the importance of ketone and fatty acid metabolism in hibernating states (Brown and Staples, 2014; Xu et al., 2013) as well as pointed to a role for specific post-translational protein modifications in the differential regulation of metabolic pathways in hibernation (Ballinger et al., 2016; Chung et al., 2013; Herinckx et al., 2017). Specific variants of neuroprotective proteins have also been identified to be upregulated in ground squirrels during hibernation including s-humanin, however, the phenotypic or mechanistic consequences of these variants are not known (Szereszewski and Storey, 2019). We expanded this body of knowledge, by identifying altered mitochondrial dynamics and enhanced spare respiratory capacity in cells of AGS as potentially adaptive cellular mechanisms in hibernating animals. This mitochondrial phenotype is likely responsible for the broad resilience of AGS cells against a wide range of metabolic stressors.

Spare respiratory capacity, as measured by FCCP-stimulated oxygen consumption, represents a marker for cellular metabolic reserves, correlates with metabolic resilience (Nicholls and Budd, 2000) and is thought to be determined by the oxidative phosphorylation machinery (Pfleger et al., 2015; Yadava and Nicholls, 2007). Notably, human and mouse NPCs and neural cells have been reported to have diminished spare respiratory capacity as they may respire maximally at baseline (Khacho et al., 2016; Lorenz et al., 2017). However, AGS demonstrate marked elevations in spare respiratory capacity compared to mouse cells, which likely explains marked AGS NPC survival even under complex I inhibition by rotenone (Yadava and Nicholls, 2007). While the elevated spare respiratory capacity is likely the result of AGS adaptations in numerous metabolically active proteins, the importance of the ATP5G1 variant is highlighted by our experimental evidence demonstrating improvement in spare respiratory capacity in mouse NPCs over-expressing AGS ATP5G1 variants and decreased spare respiratory capacity in AGS NPCs with ATP5G1 L32P knock-in. A critical role for ATP5G1 in cellular energetics is also supported by recent work uncovering ATP5G1 as one of the major effectors of the transcription factor, BCL6, in regulating adipose tissue energetics as well as maintaining thermogenesis in response to hypothermia (Kutyavin and Chawla, 2019; Senagolage et al., 2018).

Mitochondrial fission and fusion are regulated by cellular metabolic state and a host of regulatory proteins, many of which have been implicated in cell survival response to stresses (Labbé et al., 2014). While metabolic stresses often lead to mitochondrial fission followed by apoptosis, mitochondrial fusion and resistance to fission in response to stress are anti-apoptotic (Abdelwahid et al., 2007; Chen et al., 2007). Fusion is hypothesized to allow for complementation of damaged and dysfunctional mitochondria, and in states of metabolic stress, hyperfusion of mitochondria helps maintain mitochondrial membrane potential and cell viability (Gomes et al., 2011). The increase in fusion and improvement in cell survival in mouse NPCs over-expressing AGS ATP5G1 and loss of resilient metabolic phenotypes in AGS cells carrying ATP5G1L32P underscore the importance of this pathway in altering mitochondrial morphologic response to metabolic stresses and increasing the metabolic oxidative capacity of cells.

In mammals, many of the approximately 1000 nuclear-encoded mitochondrial proteins contain a unique mitochondrial targeting sequence (MTS) providing a high degree of specificity in regulating mitochondrial import and sorting. These mitochondrial targeting and processing functions are regulated by the highly conserved mitochondrial membrane translocating protein complexes (TOM and TIM) and MTS cleaving proteins, mitochondrial processing peptide (MPP) and mitochondrial intermediate peptide (MIP). Processing of ATP5G1 and its incorporation of the mature peptide into oligomeric c-rings and Complex V-Fo appear to involve cleavage by MPP and stabilization by TMEM70 (Kovalčíková et al., 2019). We did not find evidence that the ATP5G1 MTS sequence variations from AGS and human/mouse affected the mitochondrial localization or cleavage of the immature protein. This is likely due to evolutionarily conserved mitochondrial import sequence motifs and the putative ATP5G1 MPP/MIP cleavage site (xRx↓(F/L/I)xx(S/T/G)xxxx↓; see Figure 2—figure supplement 1C; Gakh et al., 2002). Interestingly, under native gel electrophoresis conditions in ABE KI NPC mitochondria, we observed a reduction in ATP synthase dimers relative to total ATP synthase. Additional supporting evidence is required to understand the mechanistic basis of this effect. We speculate that the AGS variant alters ATP5G1 processing which subsequently affects downstream dimerization of ATP synthases. Suprastructual alterations in ATP synthase organization are known to be critical to mitochondrial morphology and formation of the mitochondrial permeability transition pore (MPTP) (Nesci and Pagliarani, 2019). Though the exact nature of the relationship between ATP5G1 and the MPTP is controversial, many studies demonstrate improved bioenergetic responses and cell survival with ATP synthase dimerization (Bonora et al., 2017; Daum et al., 2013; García-Aguilar and Cuezva, 2018). Others have postulated that the cleaved ATP5G1 N-terminal mitochondrial targeting sequence modulates mitochondrial function downstream of Complex IV distinct from the functionally active C-terminal protein (Vives-Bauza et al., 2010). Although increased abundance of the Atp5g1 transcript in AGS compared with mouse NPCs could contribute to the altered mitochondrial function as in prior investigations of regulation of ATP synthase in mouse brown adipose tissue (Andersson et al., 1997), the ABE ATP5G1L32P KI cells did not demonstrate a difference in Atp5g1 mRNA transcript abundance, further supporting the notion that AGS ATP5G1L32P contributes to cytoprotection likely via post-transcriptional processing of ATP5G1. Precise mechanisms of how AGS ATP5G1L32P affects mitochondrial function and metabolic stress resilience phenotypes await future investigations.

Further unraveling of the mechanisms underlying AGS mitochondrial and cellular resilience to metabolic stress or injuries holds the hope of finding novel cytoprotective strategies that may lead to improved treatments for human diseases. Systematic investigation of additional cytoprotective genes and amino acid substitutions identified from AGS should provide important insights into the mechanism and pathways underlying intrinsic stress resilience to metabolic stresses. The use of CRISPR gene editing technologies coupled with phenotypic analysis in AGS NPCs is a new and powerful approach to evaluate causal roles of genetic variants in conferring phenotypic traits of AGS traditionally intractable to study. Identification and analysis of such causal variants for stress resilience in AGS may help develop pharmacological, gene therapy, or CRISPR/genome editing-based therapeutic strategies to treat human ischemic disorders, including stroke and heart attack.

Materials and methods

Cell culture

Request a detailed protocol

AGS NPCs (Neuronascent, Gaithersburg, MD, USA) and mouse NPCs (gift of Song lab, Baltimore, MD) have been previously described (Drew et al., 2016; Ma et al., 2009). They were grown under standard conditions at 37°C and 5% CO2 with NeuroCult basal media (STEMCELL, Vancouver, BC, CA) with EGF (50 ng/ml, PeproTech, Inc, Rocky Hill, NJ, USA), FGF (100 ng/ml, PeproTech, Inc), heparin (0.002%), and proliferation supplements (STEMCELL). Early passage cultures (P2) were expanded and frozen and thawed in batches for use in experiments. These cultures contain cells ubiquitously expressing the NPC marker, Nestin, and the proliferation marker, Ki-67 (Figure 1—figure supplement 1). For in vitro modeling of metabolic stress, cells were exposed to either: (i) 1% hypoxia in a specialized incubator (Nuaire, Plymouth, MN, USA) saturated with Nitrogen/5% CO2; (ii) hypothermia in standard incubators maintained at lower temperatures; and (iii) complex I inhibition with the addition of rotenone to cell media. For cell proliferation determination, wells were seeded in triplicate with 50,000 cells. On subsequent consecutive days, cells were detached with Accutase (STEMCELL) and counted by automated cytometry (Nanoentek, Waltham, MA, USA).

DNA constructs and lentiviral transfection

Request a detailed protocol

The pHAGE-ATP5G plasmids were generated by direct PCR and PCR fusions; and the point mutation plasmids generated using Q5 site-directed mutagenesis (New England Biolabs, Beverly, MA, USA). For lentiviral transfection, the plasmids with packaging plasmids were co-transfected into HEK293FT (with a ratio of 2:1.5:1.5) using Turbofect reagent (Thermo Fisher Scientific Inc, Waltham, MA, USA) according to the manufacturer’s instructions. Lentivirus-containing medium was filtered from the post-transfection supernatant and used for transduction of HEK293T cells or mouse NPCs. All lentivirus-infected cells were cultured in the medium containing Polybrene (4 μg/ml; Sigma Aldrich, St. Louis, MO, USA) for 8 hr before changing media. Forty-eight hours after transduction, the cells were selected with 10 µg/ml Blastidicin S (Thermo Fisher Scientific Inc).

Generation of CRISPR base-edited AGS cells

Request a detailed protocol

ATP5G1L32P NPCs were generated using the dCas9 base editor, ABEmax (gift from David Liu, Addgene #112095), as previously described (Koblan et al., 2018). Briefly, a synthetic sgRNA (TCCTCTAGTCTATTCAGGAA) was selected by manual inspection of the AGS Atp5g1 sequence for a PAM (NGG) site near the desired edit on the (-) strand of the gene. AGS NPCs were nucleofected (Amaxa 4D, program DS113) in P3 solution (Lonza, Alpharetta, GA, USA) containing pCMV ABEmax (500 ng/200,000 cells). Following a 48 hr recovery period, the same cells were nucleofected with the synthetic sgRNA sequence above (100 pmol, Synthego, Menlo Park, CA, USA). Cells were expanded and then clonally plated. Clones were screened by PCR as the desired base edit also introduced a new BfaI restriction enzyme cutting site. Sanger sequencing was used to confirm the two WT and three KI clone sequences utilized. Potential off-target effects of CRISPR/Cas9 cleavage were analyzed by Sanger sequencing of the top 5 predicted off-target genomic locations [https://mit.crispr.edu], which demonstrated a lack of indels for all clones used in subsequent analysis.

Cell death assay

Request a detailed protocol

Mouse and AGS cells were plated in 24 or 96-well plates and grown to 70% confluence. Cells were exposed to metabolic stress paradigms as above, and detached and floating cells collected by centrifugation and washed with 1 ml PBS. The collected cells were resuspended with 200 μl PBS with addition of 0.2 μl Sytox blue (1 µM; Thermo Fisher Scientific) or propidium iodide (2 μg/ml) for an additional 5 min. The fluorescence intensity was measured for individual cells using automated cytometry (Nanoentek) or flow cytometry (BD Biosciences, San Jose, CA, USA) within 20 min of staining, and the percentage of cell death quantified using the FlowJo software.

cDNA Library generation, screening, and identification of AGS amino acid substitutions

Request a detailed protocol

RNA was isolated from AGS NPC cells grown under standard conditions. A normalized cDNA library was generated by a commercial research partner (Bio S and T, Montreal, QC, Canada) from RNA extracted from AGS NPCs. Library quality and normalization is shown in Figure 2—figure supplement 1A and B. For library screening, plates containing 1 × 107 mouse NPCs cells were grown in triplicate and nucleofected with 200,000 clones each. Plates were exposed to one of three metabolic stress conditions (hypoxia, hypothermia, or rotenone treatment) for 48–96 hr. Following this treatment, plasmid DNA was purified from surviving cells and PCR-amplified AGS cDNA inserts subjected to next-generation sequencing. Resulting fastq files were trimmed (Trim Galore!) and mapped to the Ictidomys Tridecemlineatus genome (SpeTri2.0) using HISAT2. Mapped reads were subjected a custom pipeline for analyzing amino acid substitutions (https://github.com/evanmlee/MaLab_spec_subs; copy archived at https://github.com/elifesciences-publications/MaLab_spec_subs; Singhal, 2020). Briefly, protein sequences of mapped genes were queried by gene symbol and downloaded from OrthoDBv10 for 10 species (13LGS, Mus musculus, Rattus norvegicus, Sorex araneus, Pongo abelii, Homo sapiens, Equus caballus, Bos taurus, Oryctolagus cuniculus, Sus scrofa). OrthoDB data was filtered by matching records against accepted GeneCards aliases for each gene (Kriventseva et al., 2019). Multiple records per species were resolved using maximum percent identity against the accepted human, mouse, and 13LGS sequences, such that only one record per species was used for alignment. AGS protein sequences were downloaded from the Entrez Protein database. Multiple AGS isoforms were resolved by best identity match to the OrthoDB sequence data. The final protein sequence set was aligned with KAlign 2.04 (Lassmann and Sonnhammer, 2005). From aligned sequences, GS-specific residue substitutions were defined as amino acid variants present in 13LGS and AGS sequences and present in no other included species. For each GS-specific residue, sequence weights, JSD, and average GS-versus-outgroup BLOSUM62 scores were calculated as described previously (Capra and Singh, 2007). BLOSUM62 scores were used instead of point-accepted mutation scores in order to prioritize protein sequence changes with higher probability of potential chemical and functional difference. JSD was used to capture sequence conservation and difference from the background amino acid distribution. BLOSUM62 scores were calculated for GS residues against all other mammalian species sequences and averaged to give GS vs Outgroup BLOSUM62. For the entire screened cytoprotective protein dataset, JSD and BLOSUM62 score were plotted for individual genes of interest against the remaining dataset.

Analysis of in vitro mitochondrial respiration

Request a detailed protocol

Analysis of mitochondrial respiratory potential was performed using a flux analyzer (Seahorse XFe96 Extracellular Flux Analyzer; Seahorse Bioscience, North Billerica, MA, USA) with a Seahorse XF Cell Mito Stress Test Kit according to the manufacturer’s instructions. Basal respiration and ATP production were calculated to evaluate mitochondrial respiratory function according to the manufacturer’s instructions. After the measurement, cells were harvested to count the cell number, and each plotted value was normalized relative to the number of cells used. Briefly, NPCs were seeded (25,000 cells/well) into each well of XFe96 cell culture plates and were maintained in standard culture media. After 2–3 days in culture, cells were equilibrated in unbuffered XFeassay medium (Seahorse Bioscience) supplemented with glucose (4.5 g/L), sodium pyruvate (25 mg/L) and transferred to a non-CO2 incubator for 1 hr before measurement. Oxygen consumption rate (OCR) was measured with sequential injections of oligomycin, FCCP, and rotenone/antimycin A.

Analysis of mitochondrial respiratory chain complex activity and mitochondrial potential

Request a detailed protocol

Analysis of mitochondrial respiratory chain complex I, II, and IV activity was measured in mitochondrial extracts using complex enzyme activity colorimetric or absorbance-based assays (ab109721, ab10908, ab109911; Abcam, Cambridge, MA). Complex V activity was measured with Complex V Mitocheck kit (Cayman Chemical, Ann Arbor, MI, USA) and citrate synthase activity with a Citrate Synthase Enzyme Assay (Detroit R and D, Detroit, MI). Mitochondrial extracts (50 μg) were obtained as previously described (Clayton and Shadel, 2014) and used to measure time-dependent absorbance alterations on a multi-well plate reader (SprectraMax, Molecular Devices, San Jose, CA, USA). Mitochondrial membrane potential was evaluated by loading 1 × 105 cells in triplicate with the lipophilic positively charged dye tetramethylrhodamine ethyl ester (TMRE, 50 nM). For depolarization control wells, 1 µM FCCP was added. Excitation and emission wavelengths (530 and 580 nm, respectively) were measured on a multi-well plate reader.

Mitochondrial ATP5G1 targeting and dynamic morphology assessment

Request a detailed protocol

Mitochondrial localization of ATP5G1 constructs as well as morphology and fission/fusion is assessed in mouse and AGS NPCs nucleofected with mCherry or mEmerald-mito7 (Gift from Michael Davidson, Addgene #55102, 54160) as a mitochondrial marker and grown on glass coverslips in standard media (Olenych et al., 2007). Cells are allowed to recover for 48 hr and then fixed with paraformaldehyde (4%) one hour following treatment with FCCP (1 μM) or DMSO. High magnification images of cells are captured by confocal microscopy (DM6, Leica, Wetzlar, Germany) and mitochondrial morphological characteristics were assessed with the Mitochondrial Network Analysis (MiNA) toolset in J-image as previously described (Valente et al., 2017; Martín-Maestro et al., 2017). Briefly, the plugin converts confocal images to binary pixel features and analyzes the spatial relationship between pixels. The parameters analyzed are: (i) individual mitochondrial structures; (ii) networked mitochondrial; and (iii) the average of length of rods/branches. Twenty randomly chosen fields containing 30–50 cells were used to quantify the morphological pattern and network branch lengths of mitochondria. We classify the mitochondrial morphology as fragmented when the appearance is completely dotted with branch lengths < 1.8 μm.

Electrophoresis and immunoblot analysis

Request a detailed protocol

For SDS-PAGE, Laemmli loading buffer (Bio-Rad Lab, Hercules, CA, USA) plus 5% β-mercaptoethanol was added to protein extracts from cell pellets reconstituted in cell lysis buffer (Cell Signaling Technology, Danvers, MA) before heating at 95°C for 5 min. Around 30 μg of whole cell protein lysate samples were separated on 4–15% mini-PROTEIN GTX precast gels, and transferred to nitrocellulose membranes (Bio-rad). For native electrophoresis, 20 μg of mitochondrial protein extracts were resuspended in buffer containing 50 mM NaCl, 2 mM 6‐aminohexanoic acid, 50 mM imidazole, 1 mM EDTA (pH 7), solubilized with digitonin (2 g/g protein) for 20 min on ice, and centrifuged for 20 min at 30,000 g to remove cell debris. Supernatants were removed and 10% glycerol and 0.01% Ponceau S were added as previously described (Kovalčíková et al., 2019; Wittig and Schägger, 2009). Samples along with a high molecular weight native marker (GE Healthcare Life Sciences, Marlborough, MA) were separated on 4–15% precast gels in 4°C with current limited to 15 mA and transferred to polyvinylidene difluoride membranes (Wittig and Schägger, 2009). Immunoblotting was performed after blocking in TBS (Tris-buffered saline) containing 5% non-fat milk and 0.1% Tween-20. Membranes were incubated overnight with primary antibodies diluted in blocking solution at 4°C, followed by incubation with secondary antibodies at room temperature for 1 hr. Immunoreactivity was visualized by the ECL chemiluminescence system (Bio-rad) on standard film. The antibodies were ATP5A (ab-14748, 1:1000, Abcam), ATP Synthase C-subunit (ab-181243, 1:1000, Abcam), and citrate synthase (#14309, 1:1000, Cell Signaling Technology).

Immunofluorescence and confocal microscopy

Request a detailed protocol

For immunocytochemistry of mammalian cells, AGS and mouse NSC/NPC cells were seeded on laminin-coated coverslips (Neuvitro, Vancouver, WA, USA) within 24-well plates. The cells were fixed with 4% paraformaldehyde in PBS, washed with PBS, and permeabilized with 0.02% Triton X-100 in PBS for 10 min. Blocking was done with 5% BSA in PBS for 1 hr, followed by incubation with antibodies against Nestin (MAB2736, 1:50, R and D Systems, Cambridge, MA, USA) or Ki-67 (NB600-1252, 1:500, Novus Biologicals, Littleton, CO, USA) in blocking buffer overnight at 4°C. The Nestin antibody was detected using goat anti-mouse AlexaFluor 488 or 647 (1:1000; Jackson ImmunoResearch Laboratories Inc, West Grove, PA, USA) and the Ki-67 antibody was detected using AlexaFluor 488 goat anti-rabbit (1:1000; Jackson Immunoresearch) or Cy3-conjuated donkey anti-rabbit (1:500; EMD Millipore, Burlington, MA, USA) in blocking buffer. Cells were washed with PBS after primary and secondary antibody staining. Stained cells were overlaid with Fluoroshield mounting medium with DAPI (Abcam) to label nucleus. Fluorescence microscopy was performed with a Leica confocal microscope using the following fluorescence filters: DAPI (405 nm excitation); Cy3 (551 nm excitation); AlexaFluor 647 (651 nm excitation); and GFP/AlexaFluor 488 (488 nm excitation). For comparison across conditions, identical light-exposure levels were used.

Quantitative RT-PCR

Request a detailed protocol

RNA was extracted from approximately 200,000 mouse or AGS NPCs per condition according to manufacturer instructions (Quick-RNA MiniPrep kit; Irvine, CA, USA). Total RNA was reverse transcribed into cDNA (Bimake, Houston, TX, USA), and real-time PCR was performed (LightCycler96, Roche, Basel, CHE) with SYBR Green (Thermo Fisher Scientific) as a dsDNA-specific binding dye. Quantitative RT-PCR conditions were 95°C for denaturation, followed by 45 cycles of: 10 s at 95°C, 10 s at 60°C, and 20 s at 72°C. Species-specific primers for each transcript were used (for list see Table 1). Melting curve analysis was performed after the final cycle to examine the specificity of primers in each reaction. Relative abundance of each Atp5g isoform as a fraction of total Atp5g was calculated by ∆∆CT method and normalized to Rpl27.

Table 1
Species-specific primers used in quantitative RT-PCR.
Mouse
Rpl27 ForwardATA AGA ATG CGG CCG CAA GC
Rpl27 ReverseATC GAT TCG CTC CTC AAA CTT
Atp5g1 ForwardTGC AGA CCA CCA AGG CAC TG
Atp5g1 ReverseGGC CTC TGG TCT GCT CAG GA
Atp5g2 ForwardCGT CTC TAC CCG CTC CCT GA
Atp5g2 ReverseCTG CAG ACA GCG GAC GAC TC
Atp5g3 ForwardGGG CCC AGA ATG GTG TGT GT
Atp5g3 ReverseTGC AGC ACC TGC ACC AAT GA
AGS
Rpl27 ForwardCTG CCA TGG GCA AGA AGA AA
Rpl27 ReverseAGC AGG GTC TCT GAA GAC AT
Atp5g1 ForwardTCC GGC TCT GAT CCG CTG TA
Atp5g1 ReverseGGG AGC TGC TGC TGT AGG AA
Atp5g2 ForwardTGC CTG CTC CAG GTT CCT CT
Atp5g2 ReverseGGG ACT GCC AAG CTG CTG AA
Atp5g3 ForwardTGA GGC CCA GAA TGG TGA ACG
Atp5g3 ReverseCAG CAC CAG AAC CAG CCA CT

Statistical analysis

Request a detailed protocol

Data were analyzed using GraphPad Prism Software (Graphpad, San Diego, CA) and presented as means ± S.E. unless otherwise specified, with P-values calculated by two-tailed unpaired Student’s t-tests or two-way ANOVA (comparisons across more than two groups) adjusted with the Bonferroni’s correction. No randomization or blinding was used and no power calculations were done to detect a pre-specified effect size.

Data availability

Data has been made available on Dryad (https://doi.org/10.7272/Q6MW2FCP) and code as been made available on GitHub (https://github.com/evanmlee/MaLab_spec_subs; copy archived at https://github.com/elifesciences-publications/MaLab_spec_subs).

The following data sets were generated
    1. Singhal N
    2. Ma D
    (2020) Dryad Digital Repository
    Data for: Cytoprotection by a naturally occurring variant of ATP5G1 in Arctic ground squirrel neural progenitor cells.
    https://doi.org/10.7272/Q6MW2FCP

References

Decision letter

  1. Diethard Tautz
    Senior Editor; Max-Planck Institute for Evolutionary Biology, Germany
  2. Agnieszka Chacinska
    Reviewing Editor; University of Warsaw, Poland
  3. Kelly Drew
    Reviewer
  4. Jean-Paul Di Rago
    Reviewer
  5. Michela Carraro
    Reviewer

In the interests of transparency, eLife publishes the most substantive revision requests and the accompanying author responses.

Acceptance summary:

This work demonstrates an interesting phenomenon that underlies intrinsic tolerance of squirrels to metabolic stressors during hibernation, such as severe hypoxia and low temperature. Through cell survival-based cDNA expression screens and comparative genomics, a variant of ATP5G1 gene, coding for the mitochondrial F-ATP synthase subunit c, is found to have a broad protective effect on the cellular level. Specifically, a single amino acid substitution in mitochondrial targeting sequence of ATP5G1, contributes to the cellular stress resilience by modulating mitochondrial metabolism.

Decision letter after peer review:

Thank you for submitting your article "Cytoprotection by a naturally occurring variant of ATP5G1 in Arctic ground squirrels" for consideration by eLife. Your article has been reviewed by Diethard Tautz as the Senior Editor, a Reviewing Editor, and three reviewers. The following individuals involved in review of your submission have agreed to reveal their identity: Kelly Drew (Reviewer #1); Jean-Paul di Rago (Reviewer #2); Michela Carraro (Reviewer #3).

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

As the editors have judged that your manuscript is of interest, but as described below that additional experiments are required before it is published, we would like to draw your attention to changes in our revision policy that we have made in response to COVID-19 (https://elifesciences.org/articles/57162). First, because many researchers have temporarily lost access to the labs, we will give authors as much time as they need to submit revised manuscripts. We are also offering, if you choose, to post the manuscript to bioRxiv (if it is not already there) along with this decision letter and a formal designation that the manuscript is 'in revision at eLife'. Please let us know if you would like to pursue this option. (If your work is more suitable for medRxiv, you will need to post the preprint yourself, as the mechanisms for us to do so are still in development.)

The manuscript reports on an extreme hypoxia resilience observed in the Arctic ground squirrels (AGS). These animals can tolerate hypoxic insults to survive during hibernation, and this phenomenon is maintained by the cultured cells. The work identifies variant ATP5G1 as a protein responsible for cytoprotection against hypoxia. The findings are overall very interesting and important. The work on non-model animals is a strength of this paper. At the same time, the referees realize limitations in experimental technology.

The points included in the specific comments of the reviewers should be addressed to strengthen the data in general and provide more precise explanations. The reviewers were overall not fully satisfied about the level of mechanistic understanding. Thus, the revisions should aim to provide some clarifications and/or discussions on the molecular and mechanistic basis of the ATP5G1 involvement in this resilience phenomenon.

Reviewer #1:

Some weaknesses are noted. A section should be added to the Discussion to identify weakness to be considered if additional information cannot be added to remedy the concern.

1) No randomization or blinding was used and no power calculations were done to detect a pre-specified effect size. Due to the nature of the initial unbiased approach I do not feel this is a fatal flow. However, the results would be more compelling if studies of the effect of amino acid substitutions had been blinded and guided by a prior power analysis. This weakness is clearly acknowledged in the text and should be noted in the discussion.

2) The study is based on NPCs which may not reflect properties of fully differentiated or adult neurons.

3) Further supplemental data or additional references are needed to allow others to replicate the approach used to identify unique amino acid substitutions. In particular what rationale was used to define the red quadrant (high JSD values and low BLOSUM62 scores) in Figure 2D?

4) Error bars should be defined in all figures or figure captions.

Neural progenitor cells should be noted in the title and the abstract.

Reviewer #2:

If I am correct the authors consider that the much higher oxygen consumption rate after FCCP addition in AGS versus mouse NPCs contributes to the differences in survival between these two types of cells when exposed to metabolic stress. Could the authors evoke a possible explanation for this? It would be interesting to known what is responsible for the large difference in uncoupled respiration between AGS and mouse NPCs. Is it because of differences in the levels of respiratory chain complexes? If this difference in respiratory capacity has a cytoprotective effect, how can we explain that after blocking the respiratory chain by rotenone during two days, AGS NPCs still have an enhanced rate of survival than mouse NPCs. This seems to this reviewer contradictory. This point should be clarified.

It is shown that the Pro to Leu variant in the MTS of ATP5G1 in AGS does not compromise import, accumulation and assembly of its protein product and activity of ATP synthase. Thus, this variant does apparently have no consequence on ATP synthase. It is puzzling how this variant could modulate the energy-transduction activity of mitochondria. The suggestion that the MTS with the Pro to Leu variant would remain associate after cleavage with ATP synthase, like the MTS of the Rieske protein in mammals, seems very unlikely in the light of recently published high resolution structures of ATP synthases from various origins. It would be interesting to test the cytoprotective properties of the ATP5G1's MTS, alone or fused to another protein unrelated to ATP synthase.

Reviewer #3:

The authors state that growth rate of AGS and mouse NPCs is not changed although this aspect is not appreciable from a single immunofluorescence image (Figure 1A). The authors also state to have measured cell death upon FCCP stimulation, even though data are not reported (Figure 1B). The Seahorse experiment in Figure 1C poses a serious concern. In general, I think it is hard to make comparisons among cell cultures deriving from different species (here AGS versus murine NPCs) and this would require a number of control experiments that have not been provided in this study. To my knowledge, the fact that mouse NPCs display no respiratory spare capacity is unusual. FCCP is used to push the respiratory chain to its maximal activity, however the amount of this compound should be carefully titrated in order to avoid secondary toxic effects that compromise mitochondrial function. The increased spare capacity visible in AGS NPCs is obtained with 2 μM FCCP, which appears to be ineffective and even detrimental for mouse NPCs. Did the authors test different concentrations of FCCP? Another concern regards the amount of mitochondria that could differ among species. From confocal analysis it emerges indeed that AGS NPCs contain probably fewer mitochondria than their murine counterparts. Another lacking information is the evaluation of the OXPHOS protein level which would help the interpretation of data. Then, the authors evaluated changes in mitochondrial morphology by using FCCP, although they never tested the mitochondrial depolarization upon FCCP stimulation and this is a crucial piece of information that should have been assessed. Anyway, in general I believe that the electron microscopy is a more reliable technique to evaluate mitochondrial morphology in the various species and conditions.

In Figure 2—figure supplement 1E, the authors show the level of ATP5G protein without reporting any other subunits of the F-ATP synthase. Since the generation of a single F-ATP synthase monomer is a multiple-step process which requires a fine-tuned incorporation of different subcomplexes, the alteration (over-expression or down-regulation) of any F-ATP synthase subunit could deeply alter the whole assembly. I believe that this aspect should have been carefully evaluated in the presented models. Moreover, for an accurate measurement of the mitochondrial ATPase hydrolytic activity, the addition of its selective inhibitor oligomycin is mandatory, and its effect should have been used as internal control. The overexpression of AGS and human variants and their mitochondrial localization should have been checked, since alterations in the mitochondrial targeting sequence might affect a correct targeting to the organelle. Figure 2—figure supplement 1F is not cited.

Figure 3—figure supplement 1(panels E and F are not present but are cited in the text) does not correspond to what the authors state (subsection “A cDNA library expression screen identifies AGS ATP5G1 as a cytoprotective factor”). The authors mention that AGS ATP5G1 substitution did not alter mitochondrial localization although this has never been tested. There is instead a single image showing the signal of wild-type ATP5G1 and Cox8 without calculating any colocalization index. The Seahorse experiment in Figure 3A is also puzzling. From the traces, it appears that the overexpression of human ATP5G1 can improve the spare capacity of murine NPCs per se and that the substitution of human ATP5G1 L32P did not further increase the FCCP-stimulated respiration but rather caused a slight decrease. This effect on mitochondrial respiration does not correlate with the resistance observed against cell death shown in Figure 2E-G. This means that the cytoprotection does not rely on improvement of mitochondrial function, thus the molecular explanation remains to be addressed. It is also not clear whether the decreased FCCP-stimulated respiration in AGS ATP5G1 L32P is statistically significant, since no details of number of experiments and statistics have been provided. The confocal experiments then lack the control images of untreated cells to appreciate changes upon FCCP stimulation (and again the mitochondrial membrane depolarization should have been checked).

Concerning the last part on ABE KI cells, I found differences of mean branch length marginal and of a questionable biological meaning. Again, electron microscopy images of mitochondrial morphology would have improved the analysis and the comparison between the two genotypes. Moreover, the correct targeting of the variant ATP5G1 needed to be addressed, since the only western blot showing ATP5G level has been carried out with cell lysates and appears over saturated.

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Thank you for re-submitting your article "Cytoprotection by a naturally occurring variant of ATP5G1 in Arctic ground squirrel neural progenitor cells" for consideration by eLife. Your article has been reviewed by Diethard Tautz as the Senior Editor, a Reviewing Editor, and three reviewers. The following individuals involved in review of your submission have agreed to reveal their identity: Kelly Drew (Reviewer #1); Jean-Paul di Rago (Reviewer #2); Michela Carraro (Reviewer #3).

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

In principle, the reviewers are satisfied with the additional experimental work, clarification and text revisions. There is one remaining issue specified below to be addressed prior to the acceptance.

Essential revisions:

Among the newly provided data, there is one aspect that should be carefully revised. The assembly of the F-ATP synthase has been evaluated by a clear-native PAGE and the quantification analysis shows that the D/M ratio is decreased in the ABE KI. The authors state that this is due to a decreased content of monomers, although this is not at all clear in Figure 4J (I found instead a higher level of monomers in ABE KI by looking at ATP5A signal) and it is anyway pretty in contradiction with a decreased D/M ratio. The difference among the two genotypes, which show a comparable amount of monomers, might be instead in the fraction of dimers which appears slightly decreased in the ABE KI (looking at ATP5G1 signal). I believe that a more informative analysis could be perhaps normalizing the fraction of monomers and dimers to the total amount of the F-ATP synthase, e.g. D/(M+D). The authors should revise Results section and Discussion section of this part.

https://doi.org/10.7554/eLife.55578.sa1

Author response

Reviewer #1:

Some weaknesses are noted. A section should be added to the Discussion to identify weakness to be considered if additional information cannot be added to remedy the concern.

1) No randomization or blinding was used and no power calculations were done to detect a pre-specified effect size. Due to the nature of the initial unbiased approach I do not feel this is a fatal flow. However, the results would be more compelling if studies of the effect of amino acid substitutions had been blinded and guided by a prior power analysis. This weakness is clearly acknowledged in the text and should be noted in the discussion.

Thank you for your helpful suggestions. We have added additional points to clarify sources of potential false positive and negative candidates from the genetic screen approach in subsection “A cDNA library expression screen identifies AGS ATP5G1 as a cytoprotective factor”. We acknowledge that no blinding or a priori power analysis was performed, and we plan to perform power analyses/blinding procedures as key steps in performing future preclinical in vivo work.

2) The study is based on NPCs which may not reflect properties of fully differentiated or adult neurons.

This is certainly accurate. We have noted this in the Title and Abstract.

3) Further supplemental data or additional references are needed to allow others to replicate the approach used to identify unique amino acid substitutions. In particular what rationale was used to define the red quadrant (high JSD values and low BLOSUM62 scores) in Figure 2D?

We have uploaded the code for the algorithm we produced to Github and indicated this in the data transparency section. This script will allow users to replicate our findings, and in addition, modify species inputs for the identification of unique amino acid substitutions in other organisms of interest. The red quadrant reflects the top 1% of combined JSD and BLOSUM scores, which we clarified in the legend.

4) Error bars should be defined in all figures or figure captions.

Neural progenitor cells should be noted in the title and the abstract.

Thank you for noticing this. We have now defined the standard error mean bars in all the captions.

Reviewer #2:

If I am correct the authors consider that the much higher oxygen consumption rate after FCCP addition in AGS versus mouse NPCs contributes to the differences in survival between these two types of cells when exposed to metabolic stress. Could the authors evoke a possible explanation for this? It would be interesting to known what is responsible for the large difference in uncoupled respiration between AGS and mouse NPCs. Is it because of differences in the levels of respiratory chain complexes?

This is an important question in the field as the specific OXPHOS proteins and associated factors contributing to cellular reserve respiratory capacity are not completely unknown. The specific mechanism for AGS cells will require additional research, but we have added data that begins to address this and is also related to reviewer #3 concerns (Figure 1—figure supplement 1H). The literature has suggested that complex II OXPHOS proteins are crucial to this response (Pfleger et al., 2015), however, as shown in Figure 1—figure supplement 1H, there is no appreciable difference in complex II activity between AGS/mouse NPCs. Citrate synthase (a marker of mitochondrial content) or complex I activity are also similar between AGS and mouse NPC mitochondria. We did find a decrease in the activity of Complex IV. The contributions of CIV and its supramolecular organization as a mediator of differences between species warrants follow-up and is part of another collaborative manuscript currently in revision.

If this difference in respiratory capacity has a cytoprotective effect, how can we explain that after blocking the respiratory chain by rotenone during two days, AGS NPCs still have an enhanced rate of survival than mouse NPCs. This seems to this reviewer contradictory. This point should be clarified.

This is also an important point. Rotenone results in cell death via oxidative stress and ATP deficiency (Yadava and Nicholls, 2007). AGS likely possess numerous adaptations in these pathways conferring a survival advantage. Our over-expression screen and follow-up ATP5G1 studies suggest that ATP5G1 is one component contributing to this metabolic resilience, but it is likely only one adaptation among many. Other AGS OXPHOS adaptations may also confer survival advantages which help cells effectively bypass complex I and cope with rotenone-induced energy failure. Previous work on hibernating species has also implicated robust anti-oxidant adaptations (Bhowmick and Drew, 2017) in conferring cytoprotection, which is likely also contributing to the AGS phenotype. We have elaborated these points (Discussion section).

It is shown that the Pro to Leu variant in the MTS of ATP5G1 in AGS does not compromise import, accumulation and assembly of its protein product and activity of ATP synthase. Thus, this variant does apparently have no consequence on ATP synthase. It is puzzling how this variant could modulate the energy-transduction activity of mitochondria. The suggestion that the MTS with the Pro to Leu variant would remain associate after cleavage with ATP synthase, like the MTS of the Rieske protein in mammals, seems very unlikely in the light of recently published high resolution structures of ATP synthases from various origins. It would be interesting to test the cytoprotective properties of the ATP5G1's MTS, alone or fused to another protein unrelated to ATP synthase.

We agree that it is unlikely the ATP5G1 MTS remains a part of ATP synthase. We further probed the activity of ATP synthase and along with suggestions from reviewer 3 have added new data demonstrating that ABE KI NPCs have reduced abundance of the monomeric form of ATP synthase under native gel conditions (Figure 4J-K). This indicates that the P32L variant in the MTS actually does compromise the normal assembly of ATP synthase complex. As noted by Vives-Bauza et al., 20170 the ATP5G1 MTS is protective independently of fusion to the mature ATP5G1 peptide, and the possibility that this contributes to the observed phenotypes remains (elaborated on in the Discussion section). Finally, we have preliminarily found binding partners of the N-terminal sequence using mass spectrometry and co-immunoprecipitation. These direct protein interactions likely regulate ATP5G1 stability and processing, but will require further functional validation that we plan to include in follow-up work.

Reviewer #3:

The authors state that growth rate of AGS and mouse NPCs is not changed although this aspect is not appreciable from a single immunofluorescence image (Figure 1A).

Thank you for suggesting this. We now show the proliferation rate data in Figure 1B.

The authors also state to have measured cell death upon FCCP stimulation, even though data are not reported (Figure 1B).

Thank you for pointing this out. We’ve removed this incorrect reference.

The Seahorse experiment in Figure 1C poses a serious concern. In general, I think it is hard to make comparisons among cell cultures deriving from different species (here AGS versus murine NPCs) and this would require a number of control experiments that have not been provided in this study. To my knowledge, the fact that mouse NPCs display no respiratory spare capacity is unusual. FCCP is used to push the respiratory chain to its maximal activity, however the amount of this compound should be carefully titrated in order to avoid secondary toxic effects that compromise mitochondrial function. The increased spare capacity visible in AGS NPCs is obtained with 2 μM FCCP, which appears to be ineffective and even detrimental for mouse NPCs. Did the authors test different concentrations of FCCP?

We agree that comparisons between cell type and species types require careful control and interpretation. We previously performed a FCCP dose titration which is now included in Figure 1—figure supplement 1F-G. Also note, we have added the following sentence and references to the Discussion section to further clarify the limited mouse NPC spare respiratory capacity, “…human and mouse NPCs and neural cells have diminished spare respiratory capacity as they may respire maximally at baseline.” We refer to prior reports corroborating this in human and mouse cells: Lorenz et al., 2017, Figure 2C, E; Khacho et al., 2016, Figure 4.

Another concern regards the amount of mitochondria that could differ among species. From confocal analysis it emerges indeed that AGS NPCs contain probably fewer mitochondria than their murine counterparts. Another lacking information is the evaluation of the OXPHOS protein level which would help the interpretation of data.

We also agree that interspecies comparisons are difficult to standardize. The amount of mitochondria and specific make-up of subunits are important to OXPHOS enzymatic activities, and as such we provide additional information regarding mitochondrial and OXPHOS enzymatic function with citrate synthase and individual OXPHOS complex enzymatic activity (Figure 1—figure supplement 1H).

Then, the authors evaluated changes in mitochondrial morphology by using FCCP, although they never tested the mitochondrial depolarization upon FCCP stimulation and this is a crucial piece of information that should have been assessed.

Thank you for suggesting including this important control. We now show previously obtained control data on the effect of FCCP on mitochondrial potential as measured by TMRE (Figure 1E).

Anyway, in general I believe that the electron microscopy is a more reliable technique to evaluate mitochondrial morphology in the various species and conditions.

This will indeed be an extremely valuable contribution. We have embarked on a collaboration to obtain EM data, however, we hope this will be part of a follow-up manuscript related to additional mechanistic details of ATP5G1.

In Figure 2—figure supplement 1E, the authors show the level of ATP5G protein without reporting any other subunits of the F-ATP synthase. Since the generation of a single F-ATP synthase monomer is a multiple-step process which requires a fine-tuned incorporation of different subcomplexes, the alteration (over-expression or down-regulation) of any F-ATP synthase subunit could deeply alter the whole assembly. I believe that this aspect should have been carefully evaluated in the presented models. Moreover, for an accurate measurement of the mitochondrial ATPase hydrolytic activity, the addition of its selective inhibitor oligomycin is mandatory, and its effect should have been used as internal control. The overexpression of AGS and human variants and their mitochondrial localization should have been checked, since alterations in the mitochondrial targeting sequence might affect a correct targeting to the organelle. Figure 2—figure supplement 1F is not cited.

Thanks for this very helpful suggestion. Interestingly, although we found AGS ATP synthase activity to have a similar degree of sensitivity to oligomycin compared to mouse and ABE KI cells, we followed up this finding with clear native electrophoresis and immunoblots demonstrating that ABE WT have decreased abundance of ATP Synthase monomers (Figure 4J-K). Given the extreme complexity of ATP synthase biogenesis and assembly, additional work in the future is still needed to understand how the AGS amino acid substitution alters ATP synthase assembly. Recent work by others and our own preliminary data, which needs further validation, suggest that the N-terminal may be regulated by ubiquitination and also bind to the membrane assembly factor, TMEM70.

Figure 3—figure supplement 1(panels E and F are not even present but are cited in the text) does not correspond to what the authors state (subsection “A cDNA library expression screen identifies AGS ATP5G1 as a cytoprotective factor”). The authors mention that AGS ATP5G1 substitution did not alter mitochondrial localization although this has never been tested. There is instead a single image showing the signal of wild-type ATP5G1 and Cox8 without calculating any colocalization index.

In Figure 3—figure supplement 1, we now provide confocal images of mouse and AGS NPCs expressing all versions of ATP5G constructs. The images demonstrate proper targeting of the AGS and human ATP5G1 constructs to the mouse and AGS NPC mitochondria.

The Seahorse experiment in Figure 3A is also puzzling. From the traces, it appears that the overexpression of human ATP5G1 can improve the spare capacity of murine NPCs per se and that the substitution of human ATP5G1 L32P did not further increase the FCCP-stimulated respiration but rather caused a slight decrease. This effect on mitochondrial respiration does not correlate with the resistance observed against cell death shown in Figure 2E-G. This means that the cytoprotection does not rely on improvement of mitochondrial function, thus the molecular explanation remains to be addressed. It is also not clear whether the decreased FCCP-stimulated respiration in AGS ATP5G1 L32P is statistically significant, since no details of number of experiments and statistics have been provided. The confocal experiments then lack the control images of untreated cells to appreciate changes upon FCCP stimulation (and again the mitochondrial membrane depolarization should have been checked).

We clarified the statistical details in the text and figure caption. Statistically there was no difference between the spare respiratory capacity in mouse NPCs expressing human ATP5G1, agsATP5G1-L32P or the human ATP5G1-P32L. Thus, it does appear that the survival to metabolic stressors does not correlate perfectly with the spare respiratory capacity. Cell survival from metabolic stresses likely engages additional cellular mechanisms as it is different from the acute changes measured in the Seahorse assay. We now discuss that additional mechanisms beyond spare respiratory capacity could be involved on subsection “A cDNA library expression screen identifies AGS ATP5G1 as a cytoprotective factor”. In addition, with the added experiments in response to reviewer 2 comments, we provide a discussion of additional mechanisms contributing to the survival phenotype in the Discussion section. We have added representative control confocal images to the top panels of Figure 3D-G and left panels of Figure 4E,F.

Concerning the last part on ABE KI cells, I found differences of mean branch length marginal and of a questionable biological meaning. Again, electron microscopy images of mitochondrial morphology would have improved the analysis and the comparison between the two genotypes.

The mean mitochondrial network branch length calculated using automated software from confocal images has been used to assess mitochondrial dynamics on populations of cells in response to treatments and helps corroborate other aspects of mitochondrial morphology (fragmented vs elongated mitochondria). Mutations in master regulators of fission and fusion cause larger changes in mitochondrial length in cells at baseline (up to 0.5-1 mm in some reports), however, we feel that the population-based morphological changes in response to FCCP as well as the mean branch length changes present in our experiments (<0.5 mm) are still an important reflection of the mitochondrial phenotype as related to cell genotype and treatment. Nonetheless, we have moved some of the mean branch length data to the supplemental figures in favor or new figures above. As above we whole-heartedly agree with the utility of EM data and plan to obtain this for follow-up reports.

Moreover, the correct targeting of the variant ATP5G1 needed to be addressed, since the only western blot showing ATP5G level has been carried out with cell lysates and appears over saturated.

We now provide additional mitochondrial localization data in Figure 3—figure supplement 1 showing correct targeting of various ATP5G1 constructs to the mouse NPC mitochondria in AGS NPCs expressing co-localized ATP5G1 variants and the Cox8 mitochondrial marker.

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Reviewer comments

Among the newly provided data, there is one aspect that should be carefully revised. The assembly of the F-ATP synthase has been evaluated by a clear native PAGE and the quantification analysis shows that the D/M ratio is decreased in the ABE KI. The authors state that this is due to a decreased content of monomers, although this is not at all clear in Figure 4J (I found instead a higher level of monomers in ABE KI by looking at ATP5A signal) and it is anyway pretty in contradiction with a decreased D/M ratio. The difference among the two genotypes, which show a comparable amount of monomers, might be instead in the fraction of dimers which appears slightly decreased in the ABE KI (looking at ATP5G1 signal). I believe that a more informative analysis could be perhaps normalizing the fraction of monomers and dimers to the total amount of the F-ATP synthase, e.g. D/(M+D). The authors should revise Results section and Discussion section of this part.

Thank you for drawing our attention to this important oversight. We also agree that our text was incongruent with the data presented and have modified both as suggested. The clear native PAGE data demonstrates a reduction in the dimerization in the ABE KI cell line which we clarify in subsection “Knock-in of AGS ATP5G1L32P alters the resilient phenotype of AGS cells” and the Discussion section. We have also altered the quantification chart (Figure 4K) to reflect Dimer/Total ATP Synthase (Monomer + Dimer), which also supports the conclusion that dimerization is reduced in the ABE KI cell line.

We have also explicitly stated that conclusions relevant to the mechanistic details of AGS ATP5G1 cytoprotection require additional supporting data (subsection “Knock-in of AGS ATP5G1L32P alters the resilient phenotype of AGS cells”; Discussion section).

https://doi.org/10.7554/eLife.55578.sa2

Article and author information

Author details

  1. Neel S Singhal

    Department of Neurology, University of California-San Francisco, San Francisco, United States
    Contribution
    Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Writing - original draft
    Contributed equally with
    Meirong Bai
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0003-1605-4444
  2. Meirong Bai

    1. Cardiovascular Research Institute, University of California-San Francisco, San Francisco, United States
    2. Department of Physiology, University of California-San Francisco, San Francisco, United States
    Contribution
    Data curation, Formal analysis, Investigation, Methodology
    Contributed equally with
    Neel S Singhal
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-5919-7464
  3. Evan M Lee

    1. Cardiovascular Research Institute, University of California-San Francisco, San Francisco, United States
    2. Department of Physiology, University of California-San Francisco, San Francisco, United States
    Contribution
    Data curation, Formal analysis, Investigation, Methodology
    Competing interests
    No competing interests declared
  4. Shuo Luo

    1. Cardiovascular Research Institute, University of California-San Francisco, San Francisco, United States
    2. Department of Physiology, University of California-San Francisco, San Francisco, United States
    Contribution
    Investigation, Methodology
    Competing interests
    No competing interests declared
  5. Kayleigh R Cook

    1. Cardiovascular Research Institute, University of California-San Francisco, San Francisco, United States
    2. Department of Physiology, University of California-San Francisco, San Francisco, United States
    Contribution
    Investigation, Methodology
    Competing interests
    No competing interests declared
  6. Dengke K Ma

    1. Cardiovascular Research Institute, University of California-San Francisco, San Francisco, United States
    2. Department of Physiology, University of California-San Francisco, San Francisco, United States
    3. Innovative Genomics Institute, Berkeley, United States
    Contribution
    Conceptualization, Resources, Data curation, Supervision, Funding acquisition, Investigation, Visualization, Project administration, Writing - review and editing
    For correspondence
    Dengke.Ma@ucsf.edu
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-5619-7485

Funding

National Institute of General Medical Sciences (R01GM117461)

  • Dengke K Ma

Pew Charitable Trusts (Pew Scholar Award)

  • Dengke K Ma

David and Lucile Packard Foundation (Fellowship)

  • Dengke K Ma

Innovative Genomics Institute (Curci Scholar Award)

  • Dengke K Ma

American Heart Association (18CDA34030443)

  • Neel S Singhal
  • Meirong Bai

American Heart Association (19POST34381071)

  • Neel S Singhal
  • Meirong Bai

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

NSS and MB receive support from the American Heart Association, 18CDA34030443 and 19POST34381071, respectively. DKM receives support from National Institutes of Health grant R01GM117461, Pew Scholar Award, Curci Faculty Scholar Award from the Innovative Genomics Institute, and Packard Fellowship in Science and Engineering. We thank Dr. Judith Kelleher of Neuronascent for AGS NPCs and acknowledge use of sponsored core facilities including the UCSF Laboratory for Cell Analysis (P30CA082103) and the Histology and Light Microscopy Core at the Gladstone Institutes.

Senior Editor

  1. Diethard Tautz, Max-Planck Institute for Evolutionary Biology, Germany

Reviewing Editor

  1. Agnieszka Chacinska, University of Warsaw, Poland

Reviewers

  1. Kelly Drew
  2. Jean-Paul Di Rago
  3. Michela Carraro

Version history

  1. Received: January 29, 2020
  2. Accepted: October 8, 2020
  3. Accepted Manuscript published: October 14, 2020 (version 1)
  4. Version of Record published: November 17, 2020 (version 2)

Copyright

© 2020, Singhal et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 2,050
    Page views
  • 263
    Downloads
  • 7
    Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Neel S Singhal
  2. Meirong Bai
  3. Evan M Lee
  4. Shuo Luo
  5. Kayleigh R Cook
  6. Dengke K Ma
(2020)
Cytoprotection by a naturally occurring variant of ATP5G1 in Arctic ground squirrel neural progenitor cells
eLife 9:e55578.
https://doi.org/10.7554/eLife.55578

Further reading

    1. Biochemistry and Chemical Biology
    2. Cell Biology
    Maria Körner, Susanne R Meyer ... Alexander Buchberger
    Research Article Updated

    The ATPase p97 (also known as VCP, Cdc48) has crucial functions in a variety of important cellular processes such as protein quality control, organellar homeostasis, and DNA damage repair, and its de-regulation is linked to neuromuscular diseases and cancer. p97 is tightly controlled by numerous regulatory cofactors, but the full range and function of the p97–cofactor network is unknown. Here, we identify the hitherto uncharacterized FAM104 proteins as a conserved family of p97 interactors. The two human family members VCP nuclear cofactor family member 1 and 2 (VCF1/2) bind p97 directly via a novel, alpha-helical motif and associate with p97-UFD1-NPL4 and p97-UBXN2B complexes in cells. VCF1/2 localize to the nucleus and promote the nuclear import of p97. Loss of VCF1/2 results in reduced nuclear p97 levels, slow growth, and hypersensitivity to chemical inhibition of p97 in the absence and presence of DNA damage, suggesting that FAM104 proteins are critical regulators of nuclear p97 functions.

    1. Cell Biology
    2. Neuroscience
    Elisabeth Jongsma, Anita Goyala ... Collin Yvès Ewald
    Research Article Updated

    The amyloid beta (Aβ) plaques found in Alzheimer’s disease (AD) patients’ brains contain collagens and are embedded extracellularly. Several collagens have been proposed to influence Aβ aggregate formation, yet their role in clearance is unknown. To investigate the potential role of collagens in forming and clearance of extracellular aggregates in vivo, we created a transgenic Caenorhabditis elegans strain that expresses and secretes human Aβ1-42. This secreted Aβ forms aggregates in two distinct places within the extracellular matrix. In a screen for extracellular human Aβ aggregation regulators, we identified different collagens to ameliorate or potentiate Aβ aggregation. We show that a disintegrin and metalloprotease a disintegrin and metalloprotease 2 (ADM-2), an ortholog of ADAM9, reduces the load of extracellular Aβ aggregates. ADM-2 is required and sufficient to remove the extracellular Aβ aggregates. Thus, we provide in vivo evidence of collagens essential for aggregate formation and metalloprotease participating in extracellular Aβ aggregate removal.