RESEARCH ARTICLE

Naked mole-rats reduce the expression of ATP-dependent but notATP-independent heat shock proteins in acute hypoxiaVu Chau Nguyen1, Courtney A. Deck1,2 and Matthew E. Pamenter1,3,*

ABSTRACTNaked mole-rats are one of the most hypoxia-tolerant mammalsidentified, and putatively experience intermittent and severe hypoxia intheir underground burrows. Systemicphysiological adaptions to hypoxiahave begun to be investigated in this species; however, the cellularadaptations that underlie this tolerance remain poorly understood.Hypoxia compromises cellular energy production, and themaintenanceof protein integrity when ATP generation is limited poses a majorchallenge. Heat shock proteins (HSPs) are cellular chaperones that arecytoprotective during hypoxia, and we hypothesized that theirexpression would increase during acute hypoxia in naked mole-rats.To test this hypothesis, we used qPCR and western blot approaches tomeasure changes in gene and protein expression, respectively, ofHSP27, HSP40, HSP70 and HSP90 in the brain, heart, liver andtemporalismuscle fromnakedmole-rats following exposure to normoxia(21% O2) or hypoxia (7% O2 for 4, 12 or 24 h). Contrary to ourexpectations, we observed significant global reductions of ATP-dependent HSP70 and HSP90 (83% and 78%, respectively) after 24 hof hypoxia. Conversely, the expression ofATP-independentHSP27 andHSP40 proteins remained constant throughout the 24-h hypoxictreatment in brain, heart and muscle. However, with prolonged hypoxia(24 h), the expression ofHsp27 andHsp40 genes in these tissues wasalso reduced, suggesting that the protein expression of thesechaperones may also eventually decrease in hypoxia. These resultssuggest that energy conservation is prioritized over cytoprotectiveproteinchaperoning innakedmole-rat tissuesduringacutehypoxia.Thisunique adaptation may help naked mole-rats to minimize energyexpenditure while still maintaining proteostasis in hypoxia.

KEY WORDS: Molecular chaperone, Proteostasis, Metabolic ratesuppression, Unfolded protein response

INTRODUCTIONHypoxic environments are common in nature, and for terrestrialmammals, these environments include life at high altitude and inunderground burrows. In hypoxia, cellular aerobic respiration canbecome severely compromised, which can uncouple ATP supplyfrom demand and disrupt cellular energy balance (Buck andPamenter, 2006; Hochachka, 1986; Land et al., 1993). In hypoxia-intolerant mammals, prolonged hypoxic exposure can lead to organfailure and, ultimately, death. Conversely, some animals have evolved

systemic and cellular adaptations to hypoxia, which enables thesehypoxia-tolerant species to survive in hypoxic environments (Bicklerand Buck, 2007; McClelland and Scott, 2019).

In addition to disrupting energy balance, hypoxia can alsocontribute to a loss of protein integrity via disruptions of redoxbalance, which can alter protein structure, and lead to proteindamage, aggregation or misfolding, especially of mitochondrialproteins (Kaufman and Crowder, 2015; Kaufman et al., 2017).Impaired proteostasis during hypoxia may result in loss of cellularfunction and tissue damage in hypoxia-intolerant species. To avoidthis, a key to surviving hypoxia is to maintain good proteostasis viathe unfolded protein response (UPR) (Hetz et al., 2015). The twoprimary components of the UPR are (1) the ubiquitin–proteasomesystem, which limits the accumulation of damaged proteins bydegrading denatured non-functional proteins and preventing theiraccumulations from affecting healthy proteins and the cell, and(2) molecular chaperones, also known as heat shock proteins(HSPs). HSPs are classified based on their function as eitherholdases, foldases or disaggregases (Díaz-Villanueva et al., 2015).Holdases are ATP independent and bind to and passively stabilizedenatured proteins and prevent aggregation. Foldases are ATPdependent and actively refold denatured proteins. Disaggregases arealso ATP dependent and bind to and disaggregate protein clumps.HSPs can also be classified according to their molecular weight:HSP10, 27, 40, 60, 70, 90, 110, etc.

HSPs are known to be upregulated in hypoxia-intolerant speciesduring hypoxia, including in mice, rats, rabbits, piglets, flies,nematodes and estuarine fishes (Baird et al., 2006; David et al.,2006; Liu et al., 2013; Mestril et al., 1994; Shen et al., 2005; Tiedkeet al., 2014; Tokyol et al., 2005). However, few studies have examinedthe HSP response to acute hypoxia in hypoxia-tolerant organisms.This is a significant gap in our knowledge as it is important tounderstand the role of HSPs in cytoprotective adaptations to hypoxiain animals for which hypoxia is not a stress per se. One study of theanoxia-tolerant western painted turtle found that HSP70 becamesignificantly elevated in the brain (1.9-fold), heart (3.5-fold), liver(1.7-fold) andmuscle (5.6-fold) after a 30 h forced dive (Ramaglia andBuck, 2004). In addition, HSP90 increased significantly in the brain(5.6-fold), liver (2.1-fold) and muscle (2.4-fold) after a 24 h forceddive (Ramaglia and Buck, 2004). Similarly, in the anoxia-tolerantcrucian carp, HSP70 expression increased drastically in the brain andheart after 7 days of anoxia, while HSP90 was significantly elevatedin the heart only after 1 day of anoxia (Stensløkken et al., 2010).Recently, Wu et al. (2018) observed that HSP60 was upregulated1.73-fold and 1.59-fold after 4 and 24 h of anoxia, respectively, inwood frog brain. Conversely, no studies have addressed HSPresponses to hypoxia in any hypoxia-tolerant mammalian model.

One of the most hypoxia-tolerant mammals is the naked mole-rat(Heterocephalus glaber). Studies in this species have revealed severalphysiological adaptations that are protective against hypoxiaincluding profound depression in metabolism during acute hypoxiaReceived 6 August 2019; Accepted 18 October 2019

1Department of Biology, University of Ottawa, Ottawa, ON K1N 9N1, Canada.2Department of Biological Sciences, North Carolina State University, Raleigh,NC 27607, USA. 3University of Ottawa Brain and Mind Research Institute,Faculty of Medicine, University of Ottawa, Ottawa, ON K1H 8M5, Canada.

*Author for correspondence (mpamenter@uottawa.ca)

M.E.P., 0000-0003-4035-9555

1

© 2019. Published by The Company of Biologists Ltd | Journal of Experimental Biology (2019) 222, jeb211243. doi:10.1242/jeb.211243

Journal

ofEx

perim

entalB

iology

and metabolic plasticity following prolonged hypoxia (Chung et al.,2016; Pamenter et al., 2015). A key finding was that naked mole-ratscan suppress their metabolism by up to 85% in acute hypoxia(Pamenter et al., 2018). Marked metabolic rate suppression is veryimportant in hypoxia, as it reducesATPdemand tomatch limitedATPsupply (Buck and Pamenter, 2006). Hypoxic metabolic suppressionin this species is achieved in part by decreasing body temperature tonear-ambient levels and reducing physical activity (Houlahan et al.,2018; Ilacqua et al., 2017; Kirby et al., 2018). However, theunderlying cellular mechanisms of metabolic suppression are stillpoorly understood. Naked mole-rats can also use fructose as analternative fuel source during O2 deprivation (Park et al., 2017). Inaddition, naked mole-rats have an enhanced blood O2 carryingcapacity in hypoxia owing to the expression of a high-affinityhemoglobin (Johansen et al., 1976).Although systemic physiological and behavioral adaptions to

acute hypoxia have begun to be investigated in this fascinatingmammalian model, few studies have explored the cellularmechanisms of hypoxia tolerance in naked mole-rats. Specifically,little is known regarding how naked mole-rats maintain proteinintegrity in hypoxia. Importantly, naked mole-rats do expressfunctional HSPs, and the basal and heat-shock-induced expressionsof most HSPs are significantly higher in fibroblasts of nakedmole-rat compared with mouse, suggesting that naked mole-ratHSPs are endogenously primed for heat stress (Pride et al., 2015),and presumably also for hypoxia. Given this knowledge, weexplored changes in HSPs in a tissue-specific manner during acutehypoxia exposure. Because HSPs are putatively cytoprotective andnaked mole-rats are hypoxia tolerant, we hypothesized that HSPgene and protein expression would be upregulated in naked mole-rattissues to maintain cellular integrity and proteostasis duringhypoxia.

MATERIALS AND METHODSAnimalsNaked mole-rats (Heterocephalus glaber Rüppell 1842) were group-housed in interconnected multi-cage systems at 30°C and 21% O2 in50% humidity with a 12 h:12 h light:dark cycle. Animals were fedfresh tubers, vegetables, fruit and Pronutro cereal supplement adlibitum. Animals were not fasted prior to experimental trials. Allexperimental procedures were approved by the University of OttawaAnimal Care Committee in accordancewith theAnimals for ResearchAct and by the Canadian Council on Animal Care. Non-breeding(subordinate) naked mole-rats do not undergo sexual development orexpress sexual hormones, and thus we did not take sex intoconsideration when evaluating our results (Holmes et al., 2009).

Experimental designSeventeen naked mole-rats were randomly divided into fourexperimental treatment groups: a control normoxia group (21%O2; n=5), and three hypoxia treatment groups (7% O2) of 4, 12 and24 h duration (n=4 each). Naked mole-rats were treated in theircolony groups and home cages, and were supplied with fresh fruitsand vegetables during treatment. After treatment, naked mole-ratswere removed from the chamber and immediately killed by cervicaldislocation followed by rapid decapitation within 15 s. Brain, heart,liver and temporalis muscle were rapidly extracted over ice, flash-frozen in liquid N2 and then stored at −80°C until analysis.

Western blottingFrozen tissues were ground into a fine powder under liquid N2 usinga mortar and pestle. For western blot analysis, 50–100 mg of frozen

powdered tissue was homogenized on ice in 1 ml of homogenizationbuffer [3 mmol l−1 sucrose, 3 mmol l−1 dithiothreitol (DTT),1 mmol l−1 EDTA and protease inhibitor cocktail; Sigma-Aldrich,Oakville, ON,Canada] for 20–30 s and then sonicated three times for10 s each. Cell lysates were then centrifuged at 4000 g for 10 min at4°C. The resulting pellet was discarded and supernatants werecentrifuged again at 13,000 g for 45 min. The resulting pellet wasre-suspended in homogenization buffer and stored at −20°C. Proteinconcentrations were determined using Bradford assays (Bradford,1976) (Sigma-Aldrich).

For gel loading, loading buffer [65.8 mmol l−1 Tris-HCl withpH 6.8, 26.3% (w/v) glycerol, 2.1% SDS, 0.05% 2-mercaptoethanol,0.01% Bromophenol Blue; Bio-Rad Laboratories, Mississauga, ON,Canada] was added to the protein sample in 1:1 v/v ratio and boiled at95°C for 5 min. Protein samples and an unstained molecular marker(Bio-Rad) were loaded onto 10% SDS-polyacrylamide stain-freemini-PROTEAN® gels (Bio-Rad) and subjected to electrophoresis inrunning buffer (Bio-Rad) at 200 V for 45 min (using Bio-RadPowerPac™). Precision Plus Protein™ Unstained Protein Standards(Bio-Rad) were used to determine the migration and size of proteinson the gel. The proteins were then electrophoretically transferredonto an Immun-Blot® PVDF membrane (Bio-Rad) in transfer buffer(Bio-Rad) at 20 V for 90 min (using Bio-Rad Trans-Blot® SDCell). The membranes were blocked with TBS-T (20 mmol l−1 Tris,150 mmol l−1 NaCl, pH 7.5, 0.1% Tween20; Thermo FisherScientific, Nepean, ON, Canada) with 5% skim milk (Bio-Rad) for1 h at room temperature. Membranes were then washed three timeseach for 5 min in TBS-T and then incubated in primary HSP antibodyovernight at 4°C. Membranes were then washed again three timeseach for 5 min in TBS-T and then incubated in secondary horseradishperoxidase (HRP)-linked antibody for 1 h at room temperature.Then, membranes were again washed three times each for 5 min inTBS-T. Immunoreactivity was visualized using a chemiluminescenttechnique by applying the detecting agent Clarity™ ECL substrates(Bio-Rad) for 5 min and imaged (using Bio-Rad Universal Hood IIand ChemiDoc™XRS+). Finally, band intensity was analyzed usingImage Lab Software (Bio-Rad) for relative HSP levels using totalprotein as loading control.

AntibodiesThe following antibodies were used: Hsp27 Ab (Santa CruzBiotechnology, Santa Cruz, CA, USA), catalog no. sc-13132, lotL1015–1:1000 dilution; Hsp40 Ab (Cell Signaling Technology,Danvers, MA, USA), catalog no. 4871S, lot 2–1:1000 dilution;Hsp70 Ab (Santa Cruz Biotechnology), catalog no. sc-373867, lotD1114–1:350 dilution; Hsp90 Ab (Cell Signaling Technology),catalog no. 4877S, lot 3–1:1000 dilution; anti-rabbit IgG HRP-linked goat Ab (Cell Signaling Technology), catalog no. 7074S, lot26–1:1600 dilution; and anti-mouse IgGK HRP-linked Ab (SantaCruz Biotechnology), catalog no. sc-516102, lots L2017 andB0217–1:1000 dilution.

Quantitative PCR (qPCR)Total RNAwas extracted from 50–100 mg of frozen powdered tissueusing TRIzol® LS reagent (Thermo Fisher Scientific) according to themanufacturer’s instructions. The RNA concentrations and puritieswere determined (NanoDrop® ND-1000 Spectrophotometer) andstored at −80°C. To ensure the RNA samples were not contaminatedwith genomic DNA, samples were incubated with DNaseI (ThermoFisher Scientific; 2 µg RNA in 5 µl nuclease-free H2O, 1 µl 10×DNaseI reaction buffer, 1 µl DNaseI, 3 µl nuclease-free H2O) at roomtemperature for 12 min. This was followed by 10 min incubation with

2

RESEARCH ARTICLE Journal of Experimental Biology (2019) 222, jeb211243. doi:10.1242/jeb.211243

Journal

ofEx

perim

entalB

iology

1 µl EDTA at 65°C to inactivate the DNaseI. Next, the RNA templatewas combined with 375 ng random hexamers/125 ng oligo dTs(Integrated DNA Technology, Kanata, ON, Canada) and dNTPs(Thermo Fisher Scientific; 0.5 mmol l−1 final concentration) andincubated at 65°C for 5 min. First-strand reaction buffer, RNaseOut™and DTT (Thermo Fisher Scientific; 4 µl FS buffer, 1 µl DTT, 0.5 µlRNaseOut™, 0.5 µl nuclease-free H2O) were added and the mixturewas incubated at 42°C for 2 min to inhibit ribonuclease activity and

preserve the RNA for reverse transcription. Lastly, 1 µl SuperScript™II reverse transcriptase (Thermo Fisher Scientific) was added for a totalvolume of 20 µl and incubated at 42°C for 50 min followed byinactivation at 75°C for 15 min. The cDNAsampleswere then stored at−20°C.

qPCR was performed on these cDNA samples using MaximaSYBR Green qPCR master mix (Thermo Fisher Scientific) [6.25 µlSYBR Green master mix, 200 nmol l−1 each gene-specific sense andantisense primers (Table 1), 1.0 µl cDNA sample, 4.75 µl nuclease-free H2O]. qPCR was performed on 40 cycles of: melting at 95°C for15 s, followed byannealing, elongation and acquiring steps at 60°C for60 s (usingQiagen, Rotor-GeneQ). Standard curves were constructedbypooling 10 µl of all the samples together; 1×, 4×, 16× and 64× serialdilutions were used. All samples were run in duplicate. Non-templatecontrols and melting curves were used to ensure that the primerswere specific and there was no contamination. Two housekeepinggenes, actin and EEF2 (encoding the eukaryotic elongation factor 2protein), were used as internal controls. Gene expression analysis wasperformed using the NORMA-Gene normalization method(Heckmann et al., 2011).

Statistical analysisFor detection of differences in HSP protein or gene expression, two-way ANOVA (with organ and hypoxia treatment as independentvariables and protein level or gene expression as the dependentvariable) was performed using GraphPad Prism software. Dunnett’spost hoc analysis was conducted for specific statistically significantdifferences with respect to the control group within an organ, with a

Brain

Heart

Muscle

Liver

HSP27 protein

N H4 H12 H24

C

BA

0

N H4

H12

H24 N H

4H

12H

24 N H4

H12

H24 N H

4H

12H

24 N H4

H12

H24 N H

4H

12H

24 N H4

H12

H24 N H

4H

12H

24

0.5

1.0

1.5

2.0

Fold

-cha

nge

in m

RN

A ex

pres

sion

Fold

-cha

nge

in p

rote

in le

vel

(rel

ativ

e to

tota

l pro

tein

)

2.5Hsp27 gene HSP27 protein Brain

Heart

Liver

Muscle

0

0.5

1.0

1.5

2.0

2.5**

** ** * *

Fig. 1. Hypoxia-inducedHSP27 protein and gene responses throughout 24 h of hypoxia.Nakedmole-rats were randomly divided into four treatment groups:normoxic control of 21% O2 (N; n=5), or acute hypoxia (7% O2) for 4 h (H4; n=4), 12 h (H12; n=4) or 24 h (H24; n=4). Homogenates of the brain (blue),heart (red), liver (green) and temporalis muscle (purple) tissues were analyzed for mRNA (A) and protein expression (B). (C) Representative western blots ofHSP27 protein expression in four tissues of treated naked mole-rats as indicated above. Data are expressed as fold-change with respect to normoxic controlsand are presented as means±s.e.m. Asterisks denote statistically significant differences from respective normoxic controls (two-way, two-tailed ANOVA withDunnett’s post hoc multiple comparison; *P<0.05, **P<0.01).

Table 1. DNA primer sequences of the four heat shock proteins (Hsp27,Hsp40, Hsp70 and Hsp90) and two housekeeping proteins (actin andeukaryotic elongation factor 2) used in this study

Primer Sequence

HSP27 F: TTCCCAAGGTCACACAGTCAR: CTCCAGACTGCTCAGGCTTC

HSP40 (liver) F: GTTCAAGGAGATCGCTGAGGR: TCCGTGGAATGTGTAGCTGA

HSP40 (brain, heart, muscle) F: GGAGACCAGACCTCCAACAAR: TTCACCGTACAGCCACAGAG

HSP70 (liver) F: TGCTGATCCAGGTGTACGAGR: TGGTGATCTTGTTGGCTTTG

HSP70 (brain, heart, muscle) F: AGCAAGGAGGAGATCGAACAR: GCCCTTCAAACCATCATCAC

HSP90 F: AGCAAAGAAGCACCTGGAGAR: CCAGACTGAAGCCAGAGGAC

Actin F: GGCTACAGCTTCACCACCACR: GGGCAGCTCGTAGCTCTTCT

Eukaryotic elongation factor 2 F: CTGCCAGCTCATCCTAGACCR: CTTGTCCTTGTCCTCGCTGT

3

RESEARCH ARTICLE Journal of Experimental Biology (2019) 222, jeb211243. doi:10.1242/jeb.211243

Journal

ofEx

perim

entalB

iology

multiple comparison corrected α-level of 0.05. All data were thenpresented as means±s.e.m.

RESULTSHSP27 was unchanged in most organs throughout 24 hhypoxic treatmentThe gene profile of Hsp27 was consistent globally. Hsp27 mRNAlevels stayed constant for the first 12 h of acute hypoxia in all organs(Fig. 1A). Interestingly, at the 24 h mark, we observed significantglobal reductions in Hsp27 mRNA across all organs tested: 74% inthe brain, 74% in the heart, 87% in the liver and 83% in the muscle(Fig. 1A). By contrast, HSP27 protein levels remained unchangedthroughout the 24 h hypoxic treatment in the brain, heart and muscle(Fig. 1B, blue, red and purple bars, respectively). However, in theliver, the HSP27 protein level significantly increased by 90% after4 h of acute hypoxia (Fig. 1B, green bars). Conversely, after 12 h,liver HSP27 was significantly reduced by 66% compared with thenormoxic control.

HSP40 was unchanged in most organs throughout 24 hhypoxic treatmentThe gene profile of Hsp40 was similar to that observed for Hsp27.Specifically, the Hsp40 mRNA levels remained unchanged for thefirst 12 h of acute hypoxia in most organs, except in the liver(Fig. 2A). Then, at the 24 h mark, Hsp40 was significantly reducedby 79% in the brain, 79% in the heart and 81% in the muscle(Fig. 2A, blue, red and purple bars, respectively). Unlike in otherorgans, liver Hsp40 mRNAwas significantly upregulated by 317%at 12 h and 293% at 24 h of hypoxic treatment compared with the

control group (Fig. 2A, green bars). By contrast, HSP40 proteinlevels remained at control levels throughout the 24 h hypoxictreatment in the brain, heart and muscle (Fig. 2B blue, red andpurple bars, respectively). Consistent with the gene data, in the liver,HSP40 protein levels were significantly increased by 72% after both12 and 24 h of acute hypoxia treatment with respect to the controlgroup (Fig. 2B, green bars).

HSP70 fluctuated during the first 12 h of acute hypoxia, butwas significantly reduced in all organs by 24 hHypoxia-induced changes in Hsp70 gene expression varied indirection, magnitude and time across different tissues.Hsp70mRNAshared similar patterns in the brain and the heart, wherein both wereunchanged at 4 h, significantly upregulated to 199% of that of thecontrol at 12 h, but then significantly downregulated by 55%, withrespect to normoxic controls, at 24 h of hypoxia (Fig. 3A, blue andred bars, respectively). In the liver, Hsp70 was significantly reducedby 80% at 12 h and by 93% at 24 h (Fig. 3A, green bars). Finally, inmuscle, Hsp70 remained unchanged after 4 and 12 h and was thensignificantly reduced by 70% at the 24-h mark compared withcontrols (Fig. 3A, purple bars). Similarly, hypoxia-induced HSP70responses were also different in different tissues. In the brain, HSP70protein expression remained unchanged after 4 and 12 h of hypoxiaand was then significantly reduced by 57% at 24 h comparedwith controls (Fig. 3B, blue bars). In the heart, after just 4 h,HSP70 protein expression increased significantly to 217% of that ofthe control (Fig. 3B, red bars). However, at 12 h, HSP70 wassignificantly reduced by 74% of controls, but, by 24 h, this reductionbecame non-significant at 46% with respect to normoxic controls

Brain

Heart

Muscle

Liver

HSP40 protein

N H4 H12 H24

C

BA

0

1

2

3

4Hsp40 gene HSP40 protein

0

1

2

3

4

N H4

H12

H24 N H

4H

12H

24 N H4

H12

H24 N H

4H

12H

24 N H4

H12

H24 N H

4H

12H

24 N H4

H12

H24 N H

4H

12H

24

Fold

-cha

nge

in m

RN

A ex

pres

sion

Fold

-cha

nge

in p

rote

in le

vel

(rel

ativ

e to

tota

l pro

tein

)

Brain

Heart

Liver

Muscle

* *

* **

*** ***

Fig. 2. Hypoxia-induced HSP40 protein and gene responses throughout 24 h of hypoxia. Naked mole-rats were randomly divided into four treatmentgroups: normoxic control (21%O2) (N; n=5), or acute hypoxia (7%O2) for 4 h (H4, n=4), 12 h (H12; n=4) or 24 h (H24; n=4). Homogenates of the brain (blue), heart(red), liver (green) and temporalis muscle (purple) tissues were analyzed for mRNA (A) and protein expression (B). (C) Representative western blots ofHSP40 protein expression in four tissues of treated nakedmole-rats as indicated above. Data are expressed as fold-changewith respect to normoxic controls andare presented as means±s.e.m. Asterisks denote statistically significant differences from respective normoxic controls (two-way, two-tailed ANOVA withDunnett’s post hoc multiple comparison; *P<0.05, ***P<0.001).

4

RESEARCH ARTICLE Journal of Experimental Biology (2019) 222, jeb211243. doi:10.1242/jeb.211243

Journal

ofEx

perim

entalB

iology

(Fig. 3B, red bars). In the liver, HSP70 remained unchanged after 4and 12 h of hypoxia andwas then significantly reduced by 54% at the24-h mark with respect to normoxic controls (Fig. 3B, green bars).Finally, in the muscle, HSP70 levels were significantly reduced by86% and 83% after 12 and 24 h, respectively (Fig. 3B, purple bars).

HSP90 globally reduced in all organs throughout 24 h ofacute hypoxiaUnlike Hsp70, Hsp90 gene expression was globally decreasedduring hypoxia across all organs tested. In both the brain and theheart,Hsp90mRNA decreased by 24% at 4 h, 58% at 12 h and 87%at 24 h of hypoxic treatment (Fig. 4A, blue and red bars,respectively). In the liver, Hsp90 mRNA decreased significantlyby 68% at 4 h, 64% at 12 h and 99% after 24 h of hypoxia (Fig. 4A,green bars). Finally, in the muscle,Hsp90 remained unchanged after4 h but was then significantly reduced by 70% at 12 h and by 84% at24 h of hypoxia (Fig. 4A, purple bars). The HSP90 protein responsehad a similar pattern to that of the Hsp90 gene profile. In the brain,the HSP90 protein level was significantly reduced by 20% at 4 h,35% at 12 h and 23% at 24 h of hypoxia (Fig. 4B, blue bars).Similarly, in the heart, HSP90 decreased significantly by 24% at4 h, 41% at 12 h and 62% at 24 h of hypoxia (Fig. 4B, red bars). Inthe liver, HSP90 remained unchanged after 4 h, but wassignificantly reduced by 45% at 12 h and 61% at 24 h (Fig. 4B,green bars). In the muscle, HSP90 also remained unchanged after4 h, and was also significantly reduced by 83% at 12 h and 78% at24 h in hypoxia (Fig. 4B, purple bars).

DISCUSSIONIn the present study, we examined the HSP response during 24 h ofhypoxic exposure in the hypoxia-tolerant naked mole-rat for the firsttime. Our study produced two salient findings. First, HSP27 andHSP40 expression remained largely unchanged throughout thehypoxic treatment. Second, HSP70 and HSP90 expression wereglobally reduced in all tissues during the hypoxic treatment. Theseresults contradict our initial hypothesis that HSP gene and proteinexpression would be upregulated in naked mole-rat tissues tomaintain cellular integrity and proteostasis during hypoxia.Nonetheless, closer examination of the ATP dependence of theseHSPs suggests that naked mole-rats prioritize energy savings duringhypoxia and regulate their HSP response accordingly to providesome degree of proteostasis at a minimal energetic cost.

Divergent HSP response patterns during hypoxia correspondto the ATP dependency of individual HSPsDespite refuting our hypothesis, the divergent changes in HSPsin our study are intriguing. The smaller HSPs, HSP27 and HSP40,remained unchanged in hypoxia, whereas the larger HSPs, HSP70and HSP90, were globally reduced in hypoxia. These divergentpatterns may be explained by the functional classification and ATPdependency of these HSPs. Specifically, the large HSPs (HSP70and HSP90) are classified as foldases and are ATP-dependentchaperones that actively refold unfolded proteins via ATPhydrolysis (Díaz-Villanueva et al., 2015). Because these largerfoldases require ATP to function, their downregulation in an

*

Brain

Heart

Muscle

Liver

N H4 H12 H24

C

BA

0

0.5

1.0

1.5

2.0

Fold

-cha

nge

in m

RN

A ex

pres

sion

2.5

Fold

-cha

nge

in p

rote

in le

vel

(rel

ativ

e to

tota

l pro

tein

)

0

0.5

1.0

1.5

2.0

2.5

N H4

H12

H24 N H

4H

12H

24 N H4

H12

H24 N H

4H

12H

24 N H4

H12

H24 N H

4H

12H

24 N H4

H12

H24 N H

4H

12H

24

Hsp70 gene HSP70 protein Brain

Heart

Liver

Muscle

**

****

***

*** *****

***

*** **

*

HSP70 protein

Fig. 3. Hypoxia-induced HSP70 protein and gene responses throughout 24 h hypoxia. Naked mole-rats were randomly divided into four treatment groups:normoxic control (21% O2) (N; n=5), or acute hypoxia (7% O2) for 4 h (H4; n=4), 12 h (H12; n=4) or 24 h (H24; n=4). Homogenates of the brain (blue), heart(red), liver (green) and temporalis muscle (purple) tissues were analyzed for mRNA (A) and protein expression (B). (C) Representative western blots ofHSP70 protein expression in four tissues of treated naked mole-rats as indicated above. Data are expressed as fold-change with respect to normoxic controlsand are presented as means±s.e.m. Asterisks denote statistically significant differences from respective normoxic controls (two-way, two-tailed ANOVA withDunnett’s post hoc multiple comparison; *P<0.05, **P<0.01, ***P<0.001).

5

RESEARCH ARTICLE Journal of Experimental Biology (2019) 222, jeb211243. doi:10.1242/jeb.211243

Journal

ofEx

perim

entalB

iology

ATP-deficient environment such as hypoxia would help to conserveenergy for other more important cellular processes. Interestingly,the downregulation of these HSPs is delayed such that in the first12 h of hypoxia, when cellular ATP is presumably less depleted,there is an activation of these ATP-dependent HSPs forcytoprotection of hypoxia-induced protein damage. Indeed, weeven observed an upregulation ofHsp70mRNA in the brain and theheart at 12 h. However, during longer periods of hypoxia (24 h),when ATP levels are presumably compromised and/or metabolicrate suppression is enhanced, these larger HSPs are downregulatedat both the mRNA and protein level. Conversely, the small HSPs(HSP27 and HSP40) function as holdases: they are ATP-independent co-chaperones that recognize and stabilize unfoldedproteins and bring them to the foldases for re-folding (Díaz-Villanueva et al., 2015). Because these smaller holdases do notrequire ATP, their function is presumably maintained duringprolonged hypoxia despite any hypoxia-related ATP deficit. Insummary, the initial trigger of hypoxia results in the activation ofATP-dependent HSPs; however, over time there appears to be aswitch from relying on ATP-dependent to ATP-independent HSPactivity in hypoxia. This would presumably help conserve energyduring long-term hypoxia, when ATP levels may be compromised.Although we do not see global downregulation of HSP

expression during hypoxia, Hsp gene transcription exhibited avery consistent global trend such that all four Hsp genes examinedwere significantly downregulated after 24 h of hypoxia in all fourorgans tested. Because gene expression is the precursor to proteinexpression, this suggests that a longer hypoxic exposure would

cause a global downregulation in the expression of all four HSPproteins. Indeed, although HSP27 and HSP40 do not require ATPfor their normal cellular function, ATP is nonetheless required forprotein synthesis to maintain the expression of these HSPs. Thus, itwould become increasingly energetically expensive for the cell tomaintain HSP protein levels during prolonged hypoxia, given thereduced energy production when O2 is limited.

Energy conservation is prioritized over cytoprotectiveprotein chaperoning during hypoxiaTaken together, our gene and protein expression results suggest thatenergy conservation is prioritized over cytoprotective proteinchaperoning in naked mole-rats during hypoxia. Indeed, proteinsynthesis is one of the most energetically expensive cellularprocesses (Stouthamer, 1973), and many studies have demonstratedthat translational arrest of protein synthesis is a hallmark of hypoxiatolerance. For example, in the heart of the anoxia-tolerant red-earedslider turtle, protein synthesis is reduced 3-fold after both 2 and 3 hof anoxia, and mitochondrial protein expression is particularlysuppressed (Bailey and Driedzic, 1996). Similarly, the fractionalrate of protein synthesis is reduced by 92% in hepatocytes ofwestern painted turtles after 12 h of anoxic treatment (Land et al.,1993). In the liver of hypoxia-treated anoxia-tolerant goldfish,protein synthesis drops by 94% compared with normoxic controls,along with a 5.5-fold increase in AMPK activity (Jibb and Richards,2008). AMPK is a molecular switch that inhibits anabolic processeswhile promoting catabolic processes to maintain energy balance in alow-energy situation, including protein synthesis (Hardie, 2007).

Brain

Heart

Muscle

Liver

HSP90 proteinC

BA

N H4 H12 H24

0

0.5

1.0

1.5

2.0

Fold

-cha

nge

in m

RN

A ex

pres

sion

2.5

N H4

H12

H24 N H

4H

12H

24 N H4

H12

H24 N H

4H

12H

24 N H4

H12

H24 N H

4H

12H

24 N H4

H12

H24 N H

4H

12H

24

Fold

-cha

nge

in p

rote

in le

vel

(rel

ativ

e to

tota

l pro

tein

)

0

0.5

1.0

1.5

2.0

2.5

HSP90 protein Brain

Heart

Liver

Muscle

* ******

*** ***

*** *** ******

***

*********

***** ** ** **

Hsp90 gene

Fig. 4. Hypoxia-induced HSP90 protein and gene responses throughout 24 h hypoxia. Naked mole-rats were randomly divided into four treatment groups:normoxic control (21% O2) (N; n=5), or acute hypoxia (7% O2) for 4 h (H4; n=4), 12 h (H12; n=4) or 24 h (H24; n=4). Homogenates of the brain (blue), heart(red), liver (green) and temporalis muscle (purple) tissues were analyzed for mRNA (A) and protein expression (B). (C) Representative western blots ofHSP90 protein expression in four tissues of treated naked mole-rats as indicated above. Data are expressed as fold-change with respect to normoxic controlsand are presented as means±s.e.m. Asterisks denote statistically significant differences from respective normoxic controls (two-way, two-tailed ANOVA withDunnett’s post hoc multiple comparison; *P<0.05, **P<0.01, ***P<0.001).

6

RESEARCH ARTICLE Journal of Experimental Biology (2019) 222, jeb211243. doi:10.1242/jeb.211243

Journal

ofEx

perim

entalB

iology

Finally, a reduction in protein synthesis is also observed in the liver,heart and muscle of the anoxia-tolerant crucian carp in hypoxia(Smith et al., 1996).As briefly discussed in the Introduction, an alternative means of

cytoprotection and maintaining good proteostasis in hypoxia isthe selective degradation of damaged proteins. Besides beingnon-functional, damaged proteins can accumulate and aggregatewith each other and interfere with other normal cellular functions.Many studies have demonstrated that proteolytic capacity is oftenupregulated in hypoxia. For example, LON (a quality-controlmitochondrial protease) mRNA expression is significantly inducedin the heart, lung and liver of mice exposed to 10% O2 for 3 weeksby 4.7-fold, 3.3-fold and 2.9-fold, respectively, compared withnormoxic controls (Fukuda et al., 2007). Another study found thathypoxia-tolerant hard clams and oysters are able to maintain steady-state activity of ATP-dependent and -independent mitochondrialproteases under anoxia and 5%O2 (Ivanina and Sokolova, 2016). Incontrast, anoxia and hypoxia strongly suppress ATP-dependentmitochondrial protease activity in hypoxia-intolerant scallops; thissuppression is also associated with the accumulation of oxidativedamage to mitochondrial proteins (Ivanina and Sokolova, 2016).Similarly, naked mole-rats have a unique ubiquitin–proteasomesystem that gives them generalized resilience against a variety ofstressors. The proteasome of naked mole-rats has significantlyhigher chemotrypsin- and trypsin-like catalytic activities relative tomice (Rodriguez et al., 2012). This means that naked mole-rats aremore efficient at removing damaged or unfolded proteins to preventtheir aggregation from affecting healthy proteins or harming the cellduring stress. More interestingly, the naked mole-rat proteasome ismore resistant to inhibition, both stress and drug induced, throughan association with small molecular factors HSP40 and HSP70(Rodriguez et al., 2014). These molecular chaperones both enhancethe naked mole-rat proteasome activities and preserve its functionduring stressful conditions, although the role of these pathways inhypoxia has yet to be investigated.

Study limitationsIn evaluating our results, it is important to consider that naked mole-rats can tolerate O2 levels as low as 0% for 18 min, 3% for hours,and 8% for days to weeks in a laboratory setting (Chung et al., 2016;Pamenter et al., 2015; Park et al., 2017). It is therefore possible that7% O2 is not sufficiently stressful to trigger the HSP response innaked mole-rats. Thus, examining the HSP response at a lower O2

concentration might provide a better understanding of its role inhypoxia tolerance. However, we consider this unlikely as we didobserve changes in both gene and protein expression at this level ofhypoxia, suggesting active regulation of these chaperones. Inaddition, HSPs are regulated via phosphorylation; however, ourprimary antibodies were not specific and bind to both isoforms.Thus, measuring the phosphorylated form of these HSPs, usingphosphorylation-specific primary antibodies, would provide a betterunderstanding of the regulation of HSP activity during hypoxia inthis species. Nonetheless, this study is the first to address hypoxia-mediated regulation of HSPs in a hypoxia-tolerant mammal andrepresents a useful first step in this research area.

Conclusions and significanceIn summary, and contrary to results in other hypoxia- and anoxia-tolerant vertebrates, hypoxia exposure of 7% O2 for 24 h did notresult in the upregulation of four molecular chaperones (HSP27,HSP40, HSP70 and HSP90) in naked mole-rats. Interestingly, wefound that ATP-independent HSP27 and HSP40 remained

unchanged (except the liver), whereas ATP-dependent HSP70 andHSP90were globally reduced by hypoxia. Based on these results, wepropose that energy conservation is prioritized over cytoprotectiveprotein chaperoning in naked mole-rats during hypoxia. Wespeculate that these HSP responses are a unique adaptation in thisspecies that enables them tominimize energy expenditure tomaintaina healthy proteostasis, allowing them to survive in severely butintermittently hypoxic underground burrows.

AcknowledgementsWe would like to thank the University of Ottawa animal care and veterinary servicesteam for their assistance in animal handling and husbandry.

Competing interestsThe authors declare no competing or financial interests.

Author contributionsConceptualization: V.C.N., M.E.P.; Methodology: V.C.N., C.A.D., M.E.P.;Validation: V.C.N.; Formal analysis: V.C.N.; Investigation: V.C.N.; Resources:M.E.P.; Writing - original draft: V.C.N.; Writing - review & editing: C.A.D., M.E.P.;Supervision: M.E.P.; Project administration: M.E.P.; Funding acquisition: M.E.P.

FundingThis work was supported by a Natural Sciences and Engineering Research Councilof Canada Discovery Grant and a Canada Research Chair awarded to M.E.P.

ReferencesBailey, J. R. and Driedzic, W. R. (1996). Decreased total ventricular and

mitochondrial protein synthesis during extended anoxia in turtle heart.Am. J. Physiol. 271, R1660-R1667. doi:10.1152/ajpregu.1996.271.6.R1660

Baird, N. A., Turnbull, D. W. and Johnson, E. A. (2006). Induction of the heatshock pathway during hypoxia requires regulation of heat shock factor by hypoxia-inducible factor-1. J. Biol. Chem. 281, 38675-38681. doi:10.1074/jbc.M608013200

Bickler, P. E. andBuck, L. T. (2007). Hypoxia tolerance in reptiles, amphibians, andfishes: life with variable oxygen availability. Annu. Rev. Physiol. 69, 145-170.doi:10.1146/annurev.physiol.69.031905.162529

Bradford, M. M. (1976). A rapid and sensitive method for the quantitation ofmicrogram quantities of protein utilizing the principle of protein-dye binding. Anal.Biochem. 72, 248-254. doi:10.1016/0003-2697(76)90527-3

Buck, L. T. and Pamenter, M. E. (2006). Adaptive responses of vertebrate neuronsto anoxia–matching supply to demand. Respir. Physiol. Neurobiol. 154, 226-240.doi:10.1016/j.resp.2006.03.004

Chung, D., Dzal, Y. A., Seow, A., Milsom, W. K. and Pamenter, M. E. (2016).Naked mole rats exhibit metabolic but not ventilatory plasticity following chronicsustained hypoxia. Proc. R. Soc. B 283. doi:10.1098/rspb.2016.0216

David, J.-C., Boelens, W. C. and Grongnet, J.-F. (2006). Up-regulation of heatshock protein HSP 20 in the hippocampus as an early response to hypoxia of thenewborn. J. Neurochem. 99, 570-581. doi:10.1111/j.1471-4159.2006.04071.x

Dıaz-Villanueva, J. F., Dıaz-Molina, R. and Garcıa-Gonzalez, V. (2015). Proteinfolding and mechanisms of proteostasis. Int. J. Mol. Sci. 16, 17193-17230. doi:10.3390/ijms160817193

Fukuda, R., Zhang, H., Kim, J. –W., Shimoda, L., Dang, C. V. and Semenza, G. L.(2007). HIF-1 regulates cytochrome oxidase subunits to optimize efficiency ofrespiration in hypoxic cells. Cell 129, 111-122. doi:10.1016/j.cell.2007.01.047

Hardie, G. D. (2007). Biochemistry. Balancing cellular energy. Science 315,1671-1672.

Heckmann, L.-H., Sørensen, P. B., Krogh, P. H. and Sørensen, J. G. (2011).NORMA-Gene: a simple and robust method for qPCR normalization based ontarget gene data. BMC Bioinformatics 12, 250. doi:10.1186/1471-2105-12-250

Hetz, C., Chevet, E. and Oakes, S. A. (2015). Proteostasis control by the unfoldedprotein response. Nat. Cell Biol. 17, 829-838. doi:10.1038/ncb3184

Hochachka, P. W. (1986). Defense strategies against hypoxia and hypothermia.Science 231, 234-241. doi:10.1126/science.2417316

Holmes, M. M., Goldman, B. D., Goldman, S. L., Seney, M. L. and Forger, N. G.(2009). Neuroendocrinology and sexual differentiation in eusocial mammals.Front. Neuroendocrinol. 30, 519-533. doi:10.1016/j.yfrne.2009.04.010

Houlahan, C. R., Kirby, A. M., Dzal, Y. A., Fairman, G. D. and Pamenter, M. E.(2018). Divergent behavioural responses to acute hypoxia between individualsand groups of naked mole rats. Comp. Biochem. Physiol. B Biochem. Mol. Biol.224, 38-44. doi:10.1016/j.cbpb.2018.01.004

Ilacqua, A. N., Kirby, A. M. and Pamenter, M. E. (2017). Behavioural responses ofnaked mole rats to acute hypoxia and anoxia. Biol. Lett. 13, 20170545. doi:10.1098/rsbl.2017.0545

7

RESEARCH ARTICLE Journal of Experimental Biology (2019) 222, jeb211243. doi:10.1242/jeb.211243

Journal

ofEx

perim

entalB

iology

Ivanina, A. V. and Sokolova, I. M. (2016). Effects of intermittent hypoxia onoxidative stress and protein degradation in molluscan mitochondria. J. Exp. Biol.219, 3794-3802. doi:10.1242/jeb.146209

Jibb, L. A. and Richards, J. G. (2008). AMP-activated protein kinase activity duringmetabolic rate depression in the hypoxic goldfish,Carassius auratus. J. Exp. Biol.,211, 3111-3122. doi:10.1242/jeb.019117

Johansen, K., Lykkeboe, G., Weber, R. E. and Maloiy, G. M. O. (1976). Bloodrespiratory properties in the naked mole rat Heterocephalus glaber, a mammal oflow body temperature. Respir. Physiol. 28, 303-314. doi:10.1016/0034-5687(76)90025-6

Kaufman, D. M. and Crowder, C. M. (2015). Mitochondrial proteostatic collapseleads to hypoxic injury. Curr. Biol. 25, 2171-2176. doi:10.1016/j.cub.2015.06.062

Kaufman, D.M.,Wu, X., Scott, B. A., Itani, O. A., VanGilst, M. R., Bruce, J. E. andCrowder, C. M. (2017). Ageing and hypoxia cause protein aggregation inmitochondria. Cell Death Differ. 24, 1730-1738. doi:10.1038/cdd.2017.101

Kirby, A. M., Fairman, G. D. and Pamenter, M. E. (2018). Atypical behavioural,metabolic and thermoregulatory responses to hypoxia in the naked mole rat(Heterocephalus glaber). J. Zool. 305, 106-115. doi:10.1111/jzo.12542

Land, S. C., Buck, L. T. and Hochachka, P. W. (1993). Response of proteinsynthesis to anoxia and recovery in anoxia-tolerant hepatocytes. Am. J. Physiol.265, R41-R48. doi:10.1152/ajpregu.1993.265.1.R41

Liu, G., Roy, J. and Johnson, E. A. (2013). Identification and function of hypoxia-response genes in Drosophila melanogaster. Physiol. Genomics 25, 134-141.doi:10.1152/physiolgenomics.00262.2005

McClelland, G. B. and Scott, G. R. (2019). Evolved mechanisms of aerobicperformance and hypoxia resistance in high-altitude natives. Annu. Rev. Physiol.81, 561-583. doi:10.1146/annurev-physiol-021317-121527

Mestril, R., Chi, S. H., Sayen, M. R. and Dillmann,W. H. (1994). Isolation of a novelinducible rat heat-shock protein (HSP70) gene and its expression duringischaemia/hypoxia and heat shock. Biochem. J. 298, 561-569. doi:10.1042/bj2980561

Pamenter, M. E., Dzal, Y. A. and Milsom, W. K. (2015). Adenosine receptorsmediate the hypoxic ventilatory response but not the hypoxic metabolic responsein the naked mole rat during acute hypoxia. Proc. R. Soc. B Biol. Sci. 282,20141722. doi:10.1098/rspb.2014.1722

Pamenter, M. E., Lau, G. Y., Richards, J. G. and Milsom, W. K. (2018). Nakedmole rat brain mitochondria electron transport system flux and H+ leak are reducedduring acute hypoxia. J. Exp. Biol. 221, jeb171397. doi:10.1242/jeb.171397

Park, T. J., Reznick, J., Peterson, B. L., Blass, G., Omerbasic, D., Bennett, N. C.,Kuich, P. H. J. L., Zasada, C., Browe, B. M., Hamann,W. et al. (2017). Fructose-

driven glycolysis supports anoxia resistance in the naked mole-rat. Science 356,307-311. doi:10.1126/science.aab3896

Pride, H., Yu, Z., Sunchu, B., Mochnick, J., Coles, A., Zhang, Y., Buffenstein, R.,Hornsby, P. J. Austad, S. N. Perez, V. I. (2015). Long-lived species haveimproved proteostasis compared to phylogenetically-related shorter-livedspecies. Biochem. Biophys. Res. Commun. 457, 669-675. doi:10.1016/j.bbrc.2015.01.046

Ramaglia, V. and Buck, L. T. (2004). Time-dependent expression of heat shockproteins 70 and 90 in tissues of the anoxic western painted turtle. J. Exp. Biol. 207,3775-3784. doi:10.1242/jeb.01211

Rodriguez, K. A., Edrey, Y. H., Osmulski, P., Gaczynska, M. and Buffenstein, R.(2012). Altered composition of liver proteasome assemblies contributes toenhanced proteasome activity in the exceptionally long-lived naked mole-rat.PLoS ONE 7, e35890. doi:10.1371/journal.pone.0035890

Rodriguez, K. A., Osmulski, P. A., Pierce, A., Weintraub, S. T., Gaczynska, M.and Buffenstein, R. (2014). A cytosolic protein factor from the naked mole-ratactivates proteasomes of other species and protects these from inhibition.Biochim. Biophys. Acta 1842, 2060-2072. doi:10.1016/j.bbadis.2014.07.005

Shen, C., Nettleton, D., Jiang, M., Kim, S. K. and Powell-Coffman, J. A. (2005).Roles of the HIF-1 hypoxia-inducible factor during hypoxia response inCaenorhabditis elegans. J. Biol. Chem. 280, 20580-20588. doi:10.1074/jbc.M501894200

Smith, R. W., Houlihan, D. F., Nilsson, G. E. and Brechin, J. G. (1996). Tissue-specific changes in protein synthesis rates in vivo during anoxia in crucian carp.Am. J. Physiol. 271, R897-R904. doi:10.1152/ajpregu.1996.271.4.R897

Stensløkken, K.-O., Ellefsen, S., Larsen, H. K., Vaage, J. and Nilsson, G. E.(2010). Expression of heat shock proteins in anoxic crucian carp (Carassiuscarassius): support for cold as a preparatory cue for anoxia. Am. J. Physiol. Regul.Integr. Comp. Physiol. 298, R1499-R1508. doi:10.1152/ajpregu.00675.2009

Stouthamer, A. H. (1973). A theoretical study on the amount of ATP required forsynthesis of microbial cell material. Antonie Van Leeuwenhoek 39, 545-565.doi:10.1007/BF02578899

Tiedke, J., Thiel, R. andBurmester, T. (2014). Molecular response of estuarine fishto hypoxia: a comparative study with ruffe and flounder from field and laboratory.PLoS ONE 9, e90778. doi:10.1371/journal.pone.0090778

Tokyol, C., Karaorman, G. and Bastug, M. (2005). Effects of acute and adaptivehypoxia on heat shock protein expression in hepatic tissue. High Alt. Med. Biol. 6,247-255. doi:10.1089/ham.2005.6.247

Wu, C.-W., Tessier, S. N. and Storey, K. B. (2018). Stress-induced antioxidantdefense and protein chaperone response in the freeze-tolerant wood frog Ranasylvatica.CellStressChaperones23, 1205-1217. doi:10.1007/s12192-018-0926-x

8

RESEARCH ARTICLE Journal of Experimental Biology (2019) 222, jeb211243. doi:10.1242/jeb.211243

Journal

ofEx

perim

entalB

iology