Tether mutations that restore function and suppresspleiotropic phenotypes of the C. elegans isp-1(qm150)Rieske iron–sulfur proteinGholamali Jafaria, Brian M. Waskoa, Ashley Tongea, Nathan Schurmana, Cindy Donga, Zhongyu Lia, Rebecca Petersa,Ernst-Bernhard Kayserb, Jason N. Pitta, Phil G. Morganb, Margaret M. Sedenskyb, Antony R. Croftsc,and Matt Kaeberleina,1

aDepartment of Pathology, University of Washington, Seattle, WA 98195; bCenter for Developmental Therapeutics, Seattle Children’s Research Institute,Seattle, WA 98101; and cDepartment of Biochemistry, University of Illinois at Urbana–Champaign, Urbana, IL 61801

Edited by Gary Ruvkun, Massachusetts General Hospital, Boston, MA, and approved September 1, 2015 (received for review May 14, 2015)

Mitochondria play an important role in numerous diseases as wellas normative aging. Severe reduction in mitochondrial functioncontributes to childhood disorders such as Leigh Syndrome, whereasmild disruption can extend the lifespan of model organisms. TheCaenorhabditis elegans isp-1 gene encodes the Rieske iron–sulfurprotein subunit of cytochrome c oxidoreductase (complex III ofthe electron transport chain). The partial loss of function allele,isp-1(qm150), leads to several pleiotropic phenotypes. To better un-derstand the molecular mechanisms of ISP-1 function, we soughtto identify genetic suppressors of the delayed development ofisp-1(qm150) animals. Here we report a series of intragenic suppres-sors, all located within a highly conserved six amino acid tetherregion of ISP-1. These intragenic mutations suppress all of theevaluated isp-1(qm150) phenotypes, including developmental rate,pharyngeal pumping rate, brood size, body movement, activation ofthe mitochondrial unfolded protein response reporter, CO2 produc-tion, mitochondrial oxidative phosphorylation, and lifespan exten-sion. Furthermore, analogous mutations show a similar effect whenengineered into the budding yeast Rieske iron–sulfur protein Rip1,revealing remarkable conservation of the structure–function rela-tionship of these residues across highly divergent species. The focuson a single subunit as causal both in generation and in suppression ofdiverse pleiotropic phenotypes points to a common underlying mo-lecular mechanism, for which we propose a “spring-loaded” model.These observations provide insights into how gating and control pro-cesses influence the function of ISP-1 in mediating pleiotropic pheno-types including developmental rate, movement, sensitivity to stress,and longevity.

mitochondrial iron–sulfur protein | aging | Rip1 | isp-1 | complex III

Mitochondria are sites for adenosine 5′-triphosphate (ATP)production by oxidative phosphorylation, cellular calcium

buffering, iron–sulfur cluster biogenesis, reactive oxygen species(ROS) formation, and regulation of apoptosis. Although inheriteddefects in mitochondrial function are most often associated withsevere childhood disorders, a large number of age-related diseasessuch as heart disease, cancer, diabetes, obesity, and neuro-degeneration have also been linked to mitochondrial dysfunc-tion (1, 2).In Caenorhabditis elegans, multiple studies have demonstrated

that reduced electron transport chain activity (ETC) can lead toincreased lifespan. These include mutations in the coenzyme Qbiosynthetic gene clk-1, the pyrophosphokinase gene tpk-1, andthe Rieske iron–sulfur protein isp-1 (3–7). Following RNAiknockdown of ETC components, several other proteins havebeen implicated in lifespan extension, including HIF-1, GCN-2,CEP-1, CEH-23, TAF-4, AHA-1, CEH-18, JUN-1, NHR-27, andNHR-49 (8-12). In addition, it was proposed that the mito-chondrial unfolded protein response (mtUPR) directly mediatedlifespan extension from ETC inhibition (13); however, more re-cent work has suggested that induction of the mtUPR is neither

necessary nor sufficient to extend lifespan in worms (14). A self-consistent model is emerging suggesting that isp-1(qm150) ani-mals have increased levels of ROS, which induces activation ofthe intrinsic apoptotic pathway to extend lifespan (15).ISP-1 is an evolutionary conserved, nuclear-encoded iron–

sulfur (2Fe-2S) protein that functions within complex III of theelectron transport chain (16). The isp-1(qm150) allele, whichresults in a proline to serine substitution, has been particularlywell studied due to its robust positive effect on lifespan (6). Inthis context, we set out to further explore the biochemical andmolecular mechanisms by which the isp-1(qm150) mutationcauses pleiotropic phenotypes including delayed developmentand increased lifespan. Here we report the identification of in-tragenic suppressors of isp-1(qm150) all located in the highlyconserved six amino acid tether (also sometimes referred to as a“hinge”) region of ISP-1. These mutations suppressed all of thephenotypes associated with isp-1(qm150) examined, including apreviously unreported sensitivity to hyperoxia. In addition, weshow a similar relationship between the isp-1(qm150) mutationand two of the isolated suppressors in the budding yeast Rieskeiron–sulfur protein Rip1, demonstrating a striking conservationof the structure–function relationship across widely divergentphyla. Analysis of the extensive literature on physicochemicalparameters involving the role of ISP reveals a “spring-loaded”mechanism of action, summarized in the discussion and furthersupported in the SI Appendix.

Significance

Mitochondrial function is critical for health and longevity.Mutation of the highly conserved Rieske iron–sulfur subunit(ISP-1) of complex III in the respiratory chain results in pleio-tropic phenotypes in Caenorhabditis elegans, including delayeddevelopment and increased lifespan. We identified intragenicmutations within a conserved 6-aa tether region of ISP-1. Thesesuppressors are capable of suppressing all of the phenotypesassociated with the isp-1(qm150) mutation. We further demon-strated that this mutation/suppressor relationship is conserved inthe Rieske iron–sulfur protein (Rip1) of yeast complex III. Thesefindings provide insights into conserved features of the structureand function of this protein, and allow us to propose a unique“spring-loaded” mechanism to account for these effects, sup-ported by empirical physicochemical data.

Author contributions: G.J., B.M.W., and M.K. designed research; G.J., B.M.W., A.T., N.S.,C.D., Z.L., and R.P. performed research; E.-B.K., J.N.P., P.G.M., M.M.S., A.R.C., and M.K.contributed new reagents/analytic tools; G.J., B.M.W., E.-B.K., J.N.P., P.G.M., M.M.S., A.R.C.,and M.K. analyzed data; and G.J., B.M.W., A.R.C., and M.K. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.1To whom correspondence should be addressed. Email: kaeber@uw.edu.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1509416112/-/DCSupplemental.

E6148–E6157 | PNAS | Published online October 26, 2015 www.pnas.org/cgi/doi/10.1073/pnas.1509416112

Dow

nloa

ded

by g

uest

on

Mar

ch 1

0, 2

021

ResultsIsolation of Mutations That Suppress C. elegans isp-1(qm150) SlowDevelopment. We performed a forward genetic screen to iden-tify suppressors of the isp-1(qm150) slow development pheno-type (Fig. 1A). At 25 °C isp-1(qm150) homozygotes require about5 d to reach adulthood, whereas the N2 (wild type) and the sup-pressor mutants develop to adulthood in 2–3 d. Using ethyl meth-anesulfonate/N-ethyl-N-nitrosourea mutagenesis, ∼50,000 haploidgenomes were screened and eight unique suppressor mutationswithin ISP-1 were isolated (Table 1 and SI Appendix, Table S1).These strains were serially backcrossed to isp-1(qm150) threetimes and then reassessed for their phenotypes.

To determine the relative effect of each suppressor on de-velopmental growth rate, we quantified the size of N2, isp-1(qm150),and suppressor strains at 25 °C using flow cytometry (Fig. 1B). L1synchronized N2 animals reach adulthood about 40 h after platingon food, whereas isp-1(qm150) animals need about 100 h to developto adulthood. The suppressor mutants were each able to reach toadult stage in 55–75 h after L1 release.Seven out of the eight unique intragenic suppressor mutations

were clustered in a six amino acid region (DQRALA) tetherregion of ISP-1 located 73 residues upstream of the P-to-S mu-tation in the isp-1(qm150) strain (Fig. 1C). Mutations at position149 (alanine) were the most prevalent intragenic suppressormutation found in our screen (Fig. 1C and SI Appendix, Table S1).

Fig. 1. Intragenic suppressors of isp-1(qm150) suppress the slow development and are located in the tether region. (A) Development of N2, isp-1(qm150), and oneof the suppressors, 48 h after release of synchronized L1 at 25 °C. (B) Complex Object Parametric Analyzer and Sorter (COPAS) assessment of the size of the in-dividuals in the population cultivated at 25 °C for 48 h for N2, or 60 h for isp-1(qm150), and the suppressor strains (n = 6,000 per strain). N2 were assayed at 48 h toprevent substantial contamination with progeny. (C) Structure of the isp-1 gene showing the positions, residues, and the incidence of the suppressor mutations ingreen and the P > S mutation in isp-1(qm150) in red. (D) Protein sequences of ISP from S. cerevisiae, C. elegans, Gallus gallus, and Homo sapiens, were aligned usingClustalW and represented with Boxshade. Amino acid numbers corresponding to the first shown residue are to the left of each segment. Functionally identicalresidues are indicatedwith black background. The proline that is mutated to serine in isp-1(qm150) is highlighted in red, and suppressor mutations within the tetherregion are in green background. Arrows in the ribbon structure of the cytochrome bc1 complex show the tether and P > S mutation locations.

Jafari et al. PNAS | Published online October 26, 2015 | E6149

CELL

BIOLO

GY

PNASPL

US

Dow

nloa

ded

by g

uest

on

Mar

ch 1

0, 2

021

The only intragenic suppressor outside the tether was located 20amino acids downstream of the tether region, sea8, R171 (Fig. 1 Cand D). The tether region of the ISP protein is highly conserved(17) (Fig. 1D and SI Appendix, Fig. S1). Because alanine residuesin the ISP-1 tether region are highly conserved, and because allof the suppressor mutants grew at relatively similar rates, wefocused our further analyses on A149T/V (sea4 and sea5 forcomparison in the same residue) and A151T (sea7 for compar-ison in another residue).

Intragenic Suppressors of the isp-1(qm150) Developmental Delay alsoSuppress Defects in Fecundity, Motility, and Pumping. The isp-1(qm150)mutation causes a dramatic decrease in fecundity (brood size)compared with wild type (6). We measured the brood size of isp-1(qm150) animals carrying the sea4, sea5, or sea7 suppressor mu-tations at 20 °C and 25 °C, and found that all three suppressorstrains produced significantly more progeny than isp-1(qm150)animals (Fig. 2 A and B and SI Appendix, Table S2). Likewise,increased motility was observed in sup mutants, as assessed by a“thrashing” (lateral swimming) assay (Fig. 2C and SI Appendix,Table S2). The isp-1(qm150) mutation also results in dramaticreductions in pharyngeal pumping rate and motility, relative towild-type C. elegans (15). In every case examined, the suppressormutations resulted in increased pharyngeal pumping rate com-pared with the isp-1(qm150) animals (Fig. 2D, SI Appendix, andMovies S1–S5).

Intragenic Suppressors of the isp-1(qm150) Developmental Delay alsoSuppress Lifespan Extension and Activation of the MitochondrialUnfolded Protein Response. We determined whether suppressionof the slow development of the isp-1(qm150) mutant is asso-ciated with suppression of the lifespan extension. Animals car-rying the developmental suppressor alleles also had reduced lifespan extension, with the degree of suppression dependentupon temperature. At 25 °C, the lifespan of isp-1(qm150sea5)and isp-1(qm150sea7) were similar to N2, and isp-1(qm150sea4)was short lived; whereas, at 15 °C, all sup mutations tested par-tially suppressed the lifespan extension of isp-1(qm150) (Fig. 2Eand SI Appendix, Table S3).The mitochondrial unfolded protein response (mtUPR) is

induced by mitochondrial stress in C. elegans, as assessed byexpression of the hsp-6p::gfp reporter (18). We outcrossed thehsp-6p::gfp strain with the isp-1 strain and three of the suppressorstrains. All three isp-1(qm150) suppressors reduced hsp-6p::gfpfluorescence in the isp-1(qm150); hsp-6p::gfp strain, indicatingthat activation of the mtUPR is also attenuated (Fig. 3 A and B).

The isp-1(qm150) Mutation Causes Sensitivity to High and LowOxygen That Is Suppressed by Mutations in the Tether Region. Ex-posure to oxygen concentrations higher or lower than atmo-spheric levels (21%) can increase ROS production (19, 20). Aspreviously reported (21), we observed that development andreproduction of wild-type animals were not significantly affectedby hyperoxia (100% oxygen, 90 kPa). In contrast, a previouslyunreported defect was observed in hyperoxia when isp-1(qm150)showed a profound sensitivity to hyperoxia and failed to develop,ultimately dying as L1/2 larvae after 3–5 d. The tether region

mutations conferred partial suppression of this phenotype, asanimals carrying isp-1(qm150) suppressors developed to the L4stage; however, they never reached the adult reproductive stage(Fig. 3 C and D, Table 2, and SI Appendix, Fig. S2).In addition, hypoxia (1% oxygen; ref. 22) also can exacerbate ROS

production in respiratory deficient mutants such as isp-1(qm150) (9,23), and we observed that isp-1(qm150) animals arrested at the L1/2developmental stage and died in hypoxia at 25 °C (Fig. 3E andSI Appendix, Fig. S3). Under this condition, tether region mutationsallow isp-1(qm150) animals to develop further.

Mutations in the isp-1(qm150) Tether Increase Mitochondrial Functionand Stabilize Respirasomes. To directly assess the effect of theintragenic suppressors on the capacity for oxidative ATP pro-duction, we measured the state 3 (ADP-stimulated) respirationof intact isolated mitochondria. We compared the rates frommitochondria isolated from N2, isp-1(qm150), and one of thesuppressors, isp-1(qm150sea4). Mitochondria were isolated fromday 1 adults grown in normoxia at 21 ± 2 °C. We found that state3 respiration in isp-1(qm150) mitochondria is impaired whenelectrons are donated from either malate or succinate substratesto the respiratory chain upstream of complex III (Fig. 4 A–F), ashas been reported by others (24). In contrast, respiration is notaffected when electrons are donated via ascorbate plus N,N,N′,N′-tetramethyl-p-phenylenediamine directly to cytochrome c (cyt c)downstream of complex III (Fig. 4 C and D, red asterisks areafter adding TMPD/Ascorbate to bypass complex III). Intact res-piration of isp-1(qm150) isolated mitochondria, through TMPD/Ascorbate confirms an electron transport defect in isp-1(qm150) atcomplex III. In isp-1(qm150sea4), state 3 respiration rates are sig-nificantly higher, approaching wild-type function, indicating a rescueof complex III (Fig. 4 A–H and SI Appendix, Fig. S4). In addition,isp-1(qm150) has been reported to affect complex I dependentrespiration by destabilizing the respirasome supercomplex formedby complexes I, III, and IV (24). In Blue Native Gels, we observeddiaphorase activity, commonly attributed to the presence of com-plex I, in I,III and I,III,IV supercomplex bands from all N2 andall isp-1(qm150; sea4) samples but not in isp-1(qm150) samples(Fig. 4J). We also analyzed the CO2 output of three suppressors(sea4, sea5, and sea7) and compared their CO2 output with N2and isp-1(qm150). We observed that the isp-1(qm150) mutantproduced significantly lower CO2 compared with N2, and theisp-1(qm150sea4) mutant partially restored the CO2 output (Fig.4K and SI Appendix, Fig. S5 and Table S4).

Yeast Expressing the Rip1(P166S) Mutant Protein Display a RespiratoryDefect That Is Rescued by Mutations in the Tether Region. To deter-mine if the structure–function relationship of the isp-1(qm150)mutation and suppressors is conserved across phyla, we introducedanalogous mutations into the Saccharomyces cerevisiae Rieskeiron–sulfur protein, Rip1. Yeast rip1Δ mutants are capable of fer-mentative growth (Fig. 5A), but are respiratory deficient (25), asevidenced by a lack of growth on media containing nonfermentableglycerol as the only carbon source (Fig. 5B). To assess the effects ofexpression of Rip1 with and without suppressors, plasmids con-taining mutated RIP1 were introduced into haploid yeast lackingthe RIP1 gene (rip1Δ). Reintroduction of a wild-type copy of the

Table 1. Suppressor mutations in C. elegans MQ887 isp-1(qm150) background

Supressor Supressor mutations Nucleotide change Mutation location

isp-1(qm150sea1) D146N g > a Tetherisp-1(qm150sea2) R148C c > t Tetherisp-1(qm150sea3) R148H g > a Tetherisp-1(qm150sea4) A149T g > a Tetherisp-1(qm150sea5) A149V c > t Tetherisp-1(qm150sea6) A151D c > a Tetherisp-1(qm150sea7) A151T g > a Tetherisp-1(qm150sea8) R171C c > t Close to tether

E6150 | www.pnas.org/cgi/doi/10.1073/pnas.1509416112 Jafari et al.

Dow

nloa

ded

by g

uest

on

Mar

ch 1

0, 2

021

RIP1 gene rescues the rip1Δ respiratory defect, as expected. Yeastexpressing Rip1 containing the P166S mutation, analogous to theworm isp-1(qm150) mutation, exhibited slowed growth under re-spiratory conditions, indicating this proline to serine mutationstrongly diminishes mitochondrial respiratory function in yeast aswell as worms. A D87N mutation partially suppressed the re-spiratory defect caused by the P166S mutation, whereas an A90Tmutation rescued this defect more strongly (Fig. 5 A and B).Growth was also assayed under liquid culture conditions, whereyeast switch from fermentative to respiratory metabolism uponexhaustion of glucose during a growth phase called diauxic shift(about 12 h of growth in Fig. 5C). Yeast expressing the P166Smutation grew poorly during the diauxic shift, and the suppressingmutations improved this defect. These findings mirror our resultsin C. elegans, demonstrating that the effects of the isp-1(qm150)mutation and our isolated suppressors are conserved (Fig. 5D).

DiscussionIn this study, we identified several intragenic suppressors ofisp-1(qm150) that not only enhanced the slow developmental rateof isp-1(qm150) animals, but also suppressed other phenotypesincluding pharyngeal pumping rate, brood size, body movement,mitochondrial unfolded protein response, CO2 production, mito-chondrial oxidative phosphorylation (O2 consumption), lifespanextension, and a previously unreported sensitivity to hyperoxia(100% O2). Nearly all of the intragenic isp-1(qm150) suppressormutations occurred in a small highly conserved tether region of theISP protein composed of six amino acids DQRALA (D__A_A areconserved in α-proteobacteria, yeast, worm, chicken, and human).Crystal structures of complex III have revealed that the extrinsicmobile domain of ISP containing the 2Fe-2S iron–sulfur clustercan exist either near the electron donor site (Qo-site, catalytic sitefor ubihydroquinone oxidation) (26), or proximal to the electronacceptor site (cyt c1 position). Because the gap between theseelectron donors and the acceptor sites is too large to allowthe observed rate of electron transfer (17, 27–31), movement ofthe head unit of ISP was postulated. Multiple studies support thenotion that the tether region is a flexible element important inmovement of the head region of ISP, with two (17, 32–34) orthree (35) main locations for the mobile ISP head domain, eitherat the electron donor or at the acceptor site, or intermediate.Because the oxidized ISP (ISPox) is a substrate for the rate de-termining bifurcated reaction that determines the flux throughthe complex, the pleiotropic phenotypes of the isp-1(qm150)allele and their suppression likely reflect this role.Previous efforts to identify suppressors of the isp-1(qm150)

delayed development phenotype identified a mutation (A170V)in ctb-1, the mitochondrial encoded cytochrome b gene (6). Onephenotypic difference between the identified ctb-1 suppressorqm189 and our tether region suppressors is that the lifespanextension of isp-1(qm150) was reported as unchanged byctb-1(qm189), A170V (6). The reason for this difference is un-clear, but may arise from the different subunit locations andfrom differential effects of these suppressors on complex III andits partner complexes within the respirasome. Suthammaraket al. (24) reported isp-1(qm150) strongly decreased not onlycomplex III but also complex I activity measured in solubilizedmitochondria. Suppressor ctb-1(qm189) did not rescue complexIII activity but surprisingly improved complex I activity which sug-gested a rescue mechanism via allosteric effects within the super-complex. By contrast, our tether region suppressor qm150sea4restored oxidative phosphorylation of structurally intact mito-chondria regardless whether the electrons entered the ETC viacomplex I or II, indicating rescue of complex III activity. Moreover,normal complex I-dependent respiration in qm150sea4mitochondria

Fig. 2. Physiological parameters of the isp-1(qm150) suppressors. (A and B)Brood size of N2, isp-1(qm150), and one of the suppressors at two differenttemperatures (20 °C and 25 °C), n = 25 for each strain/condition.(C) Thrashing in the course of 20 min. (D) Pharyngeal pumping rates, n = 30per strain, the two-tailed heteroscedastic t test was performed on two strainsat a time in all 10 combinations to establish which differences in pumpingrate were statistically significant. (E) Lifespan of the suppressors, at 25 °C.The depicted result is a single typical trial. N2 (n = 90) isp-1(qm150) (n = 65),isp-1(qm150sea4) (n = 53), isp-1(qm150sea5) (n = 158), isp-1(qm150sea7) (n =

198). The graph indicates animals that died during the course of the ex-periment. Animals that left the agar surface are not considered. No othercensoring was performed. Error bars: SEM for highlighted experiments, *P <0.05; **P < 0.01; ***P < 0.001.

Jafari et al. PNAS | Published online October 26, 2015 | E6151

CELL

BIOLO

GY

PNASPL

US

Dow

nloa

ded

by g

uest

on

Mar

ch 1

0, 2

021

suggested if qm150 caused an additional complex I deficiency, thiswas rescued too. (Our polarography method did not allow to verifywhether a combined complex III and I deficiency actually existedin qm150.)It is noteworthy that Saccharomyces yeast does not have a true

complex I, but instead relies on three single enzyme NADHdehydrogenases (Nde1, Nde2, and Ndi1) to perform this activity.Because the yeast D87N (sea1) and A90T (sea4) mutations weresufficient to rescue the respiratory defect caused by the P166Smutation in Rip1, this further suggests that these tether regionmutations are sufficient to restore complex III activity. By homologywith the yeast structure, the ctb-1 A170 residue is located in aclamp, contributed by spans from cyt b (both subunits of thedimer) and cyt c1, that holds ISP as it emerges from the mem-

brane at the C-terminal end of the anchoring helix (27). In yeast,the equivalent residue, P174, is adjacent to the ISP tether span-TADV-. Because the clamp constrains the tether, changes mayalter control of ISP head movement (6, 35), which could there-fore be taken as supporting a model in which a slow conforma-tional change of the isp-1(qm150) head region may account forthe reduced ETC function and observed phenotypes (slow de-velopment, slow defecation, slow locomotion, etc.).In yeast, our data illustrate that mutation of proline 166 to

serine results in drastically diminished respiratory function, andthat secondary intragenic mutations within the tether region(D87N and A90T) can rescue this defect. Another mutation,G175S, located within the same intervening region as P166, hasalso been reported to reduce the respiratory capacity of yeast,

Fig. 3. Effects of intragenic suppressors on isp-1(qm150) induction of the mitochondrial unfolded protein response, sensitivity to hyperoxia, and sensitivity tohypoxia. (A) Suppressors attenuated the hsp-6p::gfp expression (mtUPR response). Epifluorescence and DIC Nomarski images of the worms expressing hsp-6p::gfp.(B) hsp-6p::gfp fluorescence intensity of the strains were compared by ImageJ software (50). (C) Intragenic suppressors partially rescued isp-1(qm150)hyperoxia lethality. N2 at 100% O2 produced next generation. isp-1(qm150) at 21% O2 produced next generation. isp-1(qm150) died at L1 stage at 100% O2.Intragenic suppressors reached to L3/4 stage under 100% O2. Images are from day eight from synchronized L1 seeded onto NGM-OP50 at 25 °C. (D and E)COPAS sorted worms in hyperoxia and hypoxia. N2 worms were sorted at day two and other strains at day five at 25 °C. Please note that L1 in N2 are from thenext generation. Error bars: SEM for highlighted experiments, *P < 0.05 **P < 0.01 ***P < 0.001.

Table 2. Hyperoxia effect on development at 25 °C

N2* isp-1(qm150) isp-1(qm150sea4) isp-1(qm150sea5) isp-1(qm150sea7)

Stage Worms, # Worms, % Worms, # Worms, % Worms, # Worms, % Worms, # Worms, % Worms, # Worms, %

L1 113 11.3 300 100 671 67.1 772 77.2 745 74.5L2-4 254 25.4 0 0 329 32.9 228 22.8 255 25.5Adult 633 63.3 0 0 0 0 0 0 0 0Total 1,000 100 300 100 1,000 100 1,000 100 1,000 100

*N2 were sorted at day 2 (to avoid F1), and other strains were sorted at day 5.

E6152 | www.pnas.org/cgi/doi/10.1073/pnas.1509416112 Jafari et al.

Dow

nloa

ded

by g

uest

on

Mar

ch 1

0, 2

021

Fig. 4. ADP-stimulated respiration of intact isolated mitochondria. Oxygen consumption traces are shown in N2, isp-1(qm150), and isp-1(qm150sea4).(A) State 3 of ADP-stimulation rates (arrow) in N2 is measured by using malate substrate as an upstream (complex I) electron donor. State 3 respiration with2 mM ADP represents the maximum capacity for oxidative phosphorylation of the mitochondria in N2. (B) Oxygen consumption traces in N2 by using succinatesubstrate as an upstream (complex II) electron donor. (C–F) comparison of state 3 of ADP-stimulation rates (arrows) in isp-1(qm150) and isp-1(qm150sea4) byusing complex III upstream substrates (malate and succinate). In C and D, asterisks (*) show the oxygen consumption in isp-1(qm150) after adding TMPD/ascorbate (electron donor to cytochrome c, downstream of complex III). (G and H) Quantitative results for state 3 measurements of 3 biological replicatesare depicted. P values are compared with isp-1(qm150) triplicates. P values in H: N2: 0.000204, isp-1(qm150sea4): 0.000707. P values in I: N2: 0.002405,isp-1(qm150sea4): 0.009303. Respiration of intact isolated mitochondria was measured with a Clark electrode. The traditional unit for the respiration rate,AO/min/mgprotein is equivalent to nmol 1/2O2/min/mgprotein (52). (I) A schematic cartoon depicting electron flow from the electron donor substrates (malate,succinate, or TMPD/ascorbate) through the mitochondrial ETC complexes. ADP stimulates respiration (state 3) because its phosphorylation to ATP by complexV allows proton flow, through complexes I, III, IV (out), and V (in). Note that electron flow from TMPD/ascorbate is independent of functional complex III.(J) Blue Native Gel. Complex I is present in supercomplex bands of N2 and isp-1(qm150sea4) but not in isp-1(qm150). (K) isp-1(qm150) CO2 output comparedwith the suppressors and N2 (for each strain: n = 1,000 COPAS sorted synchronized young adult animals of identical size (and likely weight) in three technicalreplicates (3000 worms/strain/condition). Error bars: SEM for highlighted experiments, *P < 0.05 **P < 0.01 ***P < 0.001.

Jafari et al. PNAS | Published online October 26, 2015 | E6153

CELL

BIOLO

GY

PNASPL

US

Dow

nloa

ded

by g

uest

on

Mar

ch 1

0, 2

021

though in a temperature-sensitive manner (36). Due to the lo-cation of P166 and the structural rigidity that proline provides,this residue may be important for maintaining the tertiary struc-ture of the loop. Notably, mutation of this proline residue toleucine has previously been found to lower the midpoint redoxpotential of the iron–sulfur cluster (37), and mutations of thetether region have been shown to raise the midpoint potential ofthe iron–sulfur cluster (38). However, although these opposingeffects correlate with the longevity and suppressor phenomena,studies of the molecular basis of the changes in redox potential,especially from tether mutations, show that they do not involve adirect effect on the cluster, but are mediated by structure–functioninteractions.Physicochemical parameters from a wider literature prompted

us to examine the interplay between relatively weak binding ofthe ISP head domain at the Qo-site in the wild type, and forcesassociated with conformational changes in the tether region,which support an alternative spring-loaded mechanism (39). Inbrief, we suggest that when complexes of ISP form with Qo siteoccupants, two sets of counteracting forces are in play, with oneset (the forces associated with formation of the complex), pullingon the cluster domain, and the other set (the “spring”), pullingon the tether at the other end of the extrinsic domain. Complexformation is in concert with changes between extended and re-

laxed conformations of the tether-span spring, with all forcesmediated by H-bonding and van der Waals interactions, whichare differentially modified on mutation in these regions to givethe effects observed.The ISP head domain is involved in formation of complexes

with a variety of occupants of the Qo-site. These include (i) theenzyme–substrate complex (ES-complex) between QH2 andISPox (the 2Fe-2S cluster oxidized) that initiates the bifurcatedreaction; (ii) an enzyme-product (EP-) complex between ISPH(the cluster reduced) and Q; and (iii) complexes betweenISPH and inhibitors, exemplified by UHDBT and stigmatellin.(iv) Structures have also been solved for a different type of in-hibitors, exemplified by myxothiazol and MOA-type inhibitors,which bind deeper in the Qo-site and do not interact with ISP.Complexes ii and iii involving ISPH can be assayed through theirEPR spectra because the reduced cluster is paramagnetic, andinteraction induces a new line in the spectrum. On this basis,physicochemical properties involved in formation of these com-plexes can be determined, as detailed in the SI Appendix. In par-ticular, because formation of a complex pulls the reduced form(ISPH) out of the redox mix, in each of these cases, the redoxpotential, Em, is raised, and the extent provides a measure of thestrength of binding, ΔGapp

KA(SI Appendix, Table S5). The binding of

substrates to form the ES-complex can also be assayed, but be-cause the state is metastable, only through kinetic studies. In eachof these complexes, a main contribution to the binding energycomes from formation of an H-bond between ISP and the siteoccupant. The histidine, which ligands the cluster through Nδ,can serve as H-acceptor (-Ne—HO-) when ISPox is involved or asH-donor (-NeH—O=) when ISPH is involved.When a complex is formed (17, 27–29, 40), structures show

that the ISP head is tightly packed at an interface with cytb. Otherwise, the ISP head is rotated away, and the histidine makesno H-bond (Fig. 6). These different positions are correlated tochanges in the tether span connecting the extrinsic domain to itsanchoring helix. When ISP is bound to a Qo-site occupant, thetether is pulled into an extended configuration, but when not, thetether collapses to a helical configuration. From these and sim-ilar studies in mutant strains (30, 41, 42), a possible mechanismbecomes obvious (39). In normal complex formation, the bind-ing is balanced against a restraining force, mainly from H-bondsin the helix, broken in the extension (SI Appendix, Table S6).When additional residues are inserted to introduce slack, orwhen the helical configuration is discouraged by replacing res-idues by helical breakers (glycine or proline), the tether can beextended with less effort, and the apparent strength of theH-bond stabilizing the complex, ΔGapp

KA, increases. When the

tether is so modified that the spring loses its pull, the value ofΔGapp

KAis equal to the loss of the balancing force. The physico-

chemical underpinning of this hypothesis is discussed at lengthin the SI Appendix.Similar balancing forces will be in play in formation of the ES-

complex, and these are more pertinent to the present discussionsince this initiates forward electron transfer. Mutations in thehead domain at the interface with cyt b likely inhibit formation ofthe ES-complex by steric hindrance, decreasing the binding en-ergy available from the H-bond by unfavorable VDW intera-ctions, and thus changing the balance of spring-loading. Bylimiting ES-complex formation, such mutations, includingisp-1(qm150), would restrict the input flux to the Qo-site, andtherefore the occupancy of SQo, without affecting the reactions forits removal. This educed occupancy would lower ROS production,perhaps contributing to the longevity-associated effects. In addi-tion, the reduced flux would propagate to the energy metabolismof the cell, lowering the proton gradient, and hence the poise ofthe ATP synthase, leading to substantial metabolic and physio-logical effects, which could plausibly account for many pleiotropicphenotypes. Then, all of the suppressor function from tether mu-tations can be readily explained by an easing of the tension in thespring so that the ES-complex could be formed more easily, and theinput flux restored. Additional support for this simple mechanism

Fig. 5. Tether mutations rescue the respiratory defect of S. cerevisiae Rip1(P166S). Yeast lacking the entire RIP1 gene ORF (rip1Δ) were transformedwith a plasmid containing empty vector, or plasmids expressing wild-typeRip1(WT), Rip1(P166S), or Rip1(P166S) with D87N (Sea1), or Rip1(P166S) withA90T(Sea4). Respiratory growth of yeast expressing Rip1(P166S) was dra-matically impaired, which is rescued by the secondary mutations within thetether region, D87N or A90T. (A and B) Growth was assessed on solid syn-thetic media containing 2% (wt/vol) glucose (fermentative conditions; A) or3% (wt/vol) glycerol (respiratory conditions; B). (C) Growth curves weregenerated from Bioscreen liquid culture experiments in rich media con-taining 2% (wt/vol) glucose. (D) Comparison of Rip1 conservation with isp-1.

E6154 | www.pnas.org/cgi/doi/10.1073/pnas.1509416112 Jafari et al.

Dow

nloa

ded

by g

uest

on

Mar

ch 1

0, 2

021

Fig. 6. Three configurations of the extrinsic domain of ISP show the spring-loaded mechanism in action. (Top) ISP docked at cyt b interface, H-bonded to theQo-site occupant [Protein Data Bank (PDB) ID 3H1K, avian complex, stigmatellin bound]. The main features of interest are labeled. (Middle) ISP in relaxedconfiguration [PDB ID 3L71, avian complex, azoxystrobin (MOA-inhibitor) bound]. (Bottom) ISP H-bonding with heme c1 at cyt c1 interface (PDB ID 1BE3,bovine complex, no Qo-site occupant indicated). In the relaxed configuration (Middle), the cluster ligand (H161 in bovine) makes no H-bonding contacts, sothe configuration in the tether region is not under tension. When the H161 H-bonds to the Qo-site occupant (Top), the tether is stretched due to applicationof a force (mainly from the strength of the H-bond to the occupant) to lengthen the tether and rotate the head. The helical region is pulled into an extendedconfiguration, leading to the breaking of several intra- and interspan H-bonds. When H161 H-bonds to heme c1, the displacements are much smaller, and thetether region retains a H-bonding configuration similar to that of the relaxed state. SI Appendix, Fig. S6 and Table S6 shows details of H-bonding changesbetween Qo-site docked (stigmatellin) and relaxed (azoxystrobin) configurations. Chains are colored as follows: chain c, cyt b, pink; chain d, cyt c1, yellow;chain e, ISP, orange; chain j, 7.2-kDa protein, cyan; chain p, cyt b, green (or blue-green in 1BE3); chain q, cyt c1, olive (or light green in 1BE3). Gray spans havethe cartoon colored by the CPK coloring of the C-atom or C-C bond. Atoms within 7 Å of a neighboring chain are shown as space-filled spheres of 0.2-Å radius.Atoms within 3.5 Å of a neighboring chain are shown as space-filled spheres of 0.4-Å radius. Selected atoms are CPK colored (by atom type). The threestructures are shown with the protein scaffold in the same orientation, so that the main difference comes from the configurational changes of the ISP tetherand extrinsic domains. Stereoscopic images for crossed-eye viewing.

Jafari et al. PNAS | Published online October 26, 2015 | E6155

CELL

BIOLO

GY

PNASPL

US

Dow

nloa

ded

by g

uest

on

Mar

ch 1

0, 2

021

from earlier mutational studies and a possible role for spring-loading in control and gating are discussed in the SI Appendix.In humans, specific mutations of cyt b have been reported in

patients with severe cardiomyopathy (43) and exercise intolerance(44). Introduction of human counterpart mutations into yeast cyt bresults in respiratory defects (45). Multiple suppressors of the re-spiratory deficient phenotype from cyt b S152P and G291D mu-tations (both at the docking interface for ISP) were identifiedwithin in the tether region of Rip1 or in the region within cyt blocated near the Rip1 tether (45). Similar to our findings, many ofthese suppressors were in alanine residues of the tether, whichmay be important for flexibility of the tether region. Interestingly,mutation of A90T was among the stronger suppressors of the cyt bG291D respiratory defect, and our data indicate that this mutation(sea4, A149T) was the most potent suppressor identified from ourscreen in C. elegans. Collectively, these findings highlight the im-portance of the tether region in modulating the interaction of ISPwith cyt b within complex III.In this study, we have shown that the functional consequences

and interactions of mutations in the Rieske iron–sulfur protein ofmitochondrial complex III of C. elegans are replicated in the highlyevolutionarily diverged unicellular model organism, S. cerevisiae.Our data provide more evidence that intragenic mutations withinthe tether region can rescue respiratory defects in different or-ganisms. We have provided, to our knowledge for the first time, aframework for discussion of the mechanistic features underlyingthese effects, which brings a molecular understanding of the role ofcomplex III in mediating effects on metabolic rate, embryonic andpost embryonic lethality, developmental delay, defecation cycle,fecundity, stress resistance, and longevity.Structural mechanistic analysis and discussion on the proposed

spring-loaded model is available in the SI Appendix.

Experimental ProceduresStrains and Growth Conditions. Standard C. elegans procedures were used asdescribed (46–48). Worms were maintained on solid Nematode GrowthMedium (NGM) and grown at 20 °C unless otherwise stated. To have asynchronized L1 population, eggs were hatched at room temperature (∼20–22 °C), in M9 overnight. The postembryonic development assay screen wasdone at 25 °C on solid NGM, seeded with UV arrested OP50 (49). Strains aredescribed in SI Appendix.

EMS/ENU Mutagenesis Screen. Ten thousand P0 isp-1(qm150) L4 were muta-genized with 50 mM EMS (Sigma M0880-5G) + 1 mM ENU (Sigma N3385-1G)for 6 h. Then the P0 animals were distributed on 10 individual 15-cm plates.F1 animals were grown to adulthood, and their eggs (F2) were obtained bythe bleach method. For each plate, 3,000 F2 eggs were seeded onto eachnew NGM plate. After 72 h at 25 °C, the plates were screened for any fastgrowing animals. From each P0 population, only one fast growing wormwas selected.

Lifespan Analyses. Lifespan analyses were performed as described (49)(SI Appendix).

Microscopy. Pictures were taken immediately after slide preparation using aNikon Eclipse E600 microscope using a QImaging Retigia EX CCD Camera andthe fluorescent intensity was analyzed using ImageJ (50).

Hypoxia (0.5% O2) and Hyperoxia (100% O2).Worms were synchronized by egg-bleach preparation (51) and placed onto NGM into atmospheric chambers at25 °C in an incubator and raised. Atmospheres were humidified and gas wascontinually flowed at a rate of one chamber volume per minute. Hypoxicand hyperoxic atmospheres were generated using combinations of com-

pressed air, 100% oxygen and 100% nitrogen mixed with mass flow con-trollers and analyzed with a paramagnetic oxygen analyzer that wasspanned using oxygen standards. Animals were scored dead if they couldnot move after evoked touch (SI Appendix).

CO2 Production Assay. One thousand synchronized young adults were sortedby COPAS Biosort (Union Biometrica) in three technical replicates (3,000worms in total) and placed on agar pads in atmospheric chambers in theabsence of food. The atmospheric chambers were flushed with CO2 free airfor 15 min then sealed for 15 min. The expired CO2 was measured by LI-COR6262 CO2 analyzer (SI Appendix).

Brood Size Assay. Brood size assays were performed at 20 °C by picking 25 L4sonto fresh NGM plates (5 L4 per plate) and transferring them to new platesevery 24 h for 4–7 d. Progeny plates were incubated at 20 °C for about 2 dand then scored.

Thrashing Assay. For each strain, 10 L4s were placed onto an NGM plateseeded with OP50. The thrashing assay was performed by transferring 1 L4 toan unseeded plate containing 1.6 mL of M9 buffer and thrashes were scoredwith the intervals of 1 min, 5 min, 10 min, 15 min, and 20min for the durationof 1 min each.

Pharyngeal Pumping Assay. L4 nematodes were acclimated 20 h on NGM-OP50.The assay was performed within a single day at ∼22 °C (room temperature).Pharyngeal contractions of 30 nematodes for each strain were counted over10-s intervals in three cycles of 10:10:10:10:10 with and without the aid ofreduced speed video, and these measurements were extrapolated into pumpsper minute.

Worm Sorting. The COPAS BIOSORT (Union Biometrica) was used to differ-entiate the different animals based on size. Time of Flight vs. Extinction wereused as gating parameters. Thewormswere gated to exclude eggs and debrisaccording to COPAS Application Note B-03.

Mitochondrial Fraction Isolation and Respiratory Capacity Assessment. Mea-surements of oxidative phosphorylation capacity on freshly isolated mito-chondria and Blue Native gel assays were performed as published (52–54)with the modification that animals were grown on 8P plates (55), and nylonwoven mesh sheet 30, 40, and 110 microns were used to clean and sortworms based on size.

Yeast Assays. S. cerevisiae haploid strains were all derived from isogenicBY4741 MATa or BY4742 MATα backgrounds. Yeast lacking rip1 (strainBW998) were obtained from the yeast deletion collection and the presenceof the deletion was confirmed by PCR (Open Biosystems). Solid media wasCSM-URA supplemented with 2% (wt/vol) glucose and cell number was nor-malized by equalizing the optical density of each sample. Cells were fivefoldserially diluted in water, and 4 μL of each dilution was pipetted onto the in-dicated solid media containing 2% (wt/vol) agar. For liquid growth experi-ments, overnight cultures were grown in CSM-URA 2% (wt/vol) glucose liquidmedia and 5 μL of these overnight cultures were used to inoculate 145 μL offresh media consisting of 1% (wt/vol) yeast extract, 2% (wt/vol) peptone, and2% (wt/vol) glucose. Growth rate was assayed using a Bioscreen C MBR in-strument (Growth Curves) as described (56, 57).

ACKNOWLEDGMENTS. Some strains were provided by the CaenorhabditisGenetics Center, which is funded by NIH Office of Research InfrastructurePrograms (P40 OD010440). B.M.W. was supported by NIH Training GrantT32ES007032. J.N.P. was supported by NIH Training Grant T32AG000057.This work was supported by NIH Grant R01AG039390 (to M.K.), the Universityof Washington Nathan Shock Center of Excellence in the Basic Biology ofAging (NIH P30AG013280), the Healthy Aging and Longevity Research In-stitute, and an award from the Murdock Charitable Trust (to M.K.).

1. Wallace DC (2010) Mitochondrial DNA mutations in disease and aging. Environ Mol

Mutagen 51(5):440–450.2. Bratic A, Larsson NG (2013) The role of mitochondria in aging. J Clin Invest 123(3):

951–957.3. Butler JA, Mishur RJ, Bhaskaran S, Rea SL (2013) A metabolic signature for long life in

the Caenorhabditis elegans Mit mutants. Aging Cell 12(1):130–138.4. de Jong L, Meng Y, Dent J, Hekimi S (2004) Thiamine pyrophosphate biosynthesis and

transport in the nematode Caenorhabditis elegans. Genetics 168(2):845–854.

5. Felkai S, et al. (1999) CLK-1 controls respiration, behavior and aging in the nematode

Caenorhabditis elegans. EMBO J 18(7):1783–1792.6. Feng J, Bussière F, Hekimi S (2001) Mitochondrial electron transport is a key de-

terminant of life span in Caenorhabditis elegans. Dev Cell 1(5):633–644.7. Lakowski B, Hekimi S (1996) Determination of life-span in Caenorhabditis elegans by

four clock genes. Science 272(5264):1010–1013.8. Ventura N, et al. (2009) p53/CEP-1 increases or decreases lifespan, depending on level

of mitochondrial bioenergetic stress. Aging Cell 8(4):380–393.

E6156 | www.pnas.org/cgi/doi/10.1073/pnas.1509416112 Jafari et al.

Dow

nloa

ded

by g

uest

on

Mar

ch 1

0, 2

021

9. Lee SJ, Hwang AB, Kenyon C (2010) Inhibition of respiration extends C. elegans lifespan via reactive oxygen species that increase HIF-1 activity. Curr Biol 20(23):2131–2136.

10. Khan MH, et al. (2013) TAF-4 is required for the life extension of isp-1, clk-1 and tpk-1Mit mutants. Aging (Albany, NY) 5(10):741–758.

11. Walter L, Baruah A, Chang HW, Pace HM, Lee SS (2011) The homeobox protein CEH-23 mediates prolonged longevity in response to impaired mitochondrial electrontransport chain in C. elegans. PLoS Biol 9(6):e1001084.

12. Baker BM, Nargund AM, Sun T, Haynes CM (2012) Protective coupling of mitochon-drial function and protein synthesis via the eIF2α kinase GCN-2. PLoS Genet 8(6):e1002760.

13. Durieux J, Wolff S, Dillin A (2011) The cell-non-autonomous nature of electrontransport chain-mediated longevity. Cell 144(1):79–91.

14. Bennett CF, et al. (2014) Activation of the mitochondrial unfolded protein responsedoes not predict longevity in Caenorhabditis elegans. Nat Commun 5:3483.

15. Yee C, Yang W, Hekimi S (2014) The intrinsic apoptosis pathway mediates the pro-longevity response to mitochondrial ROS in C. elegans. Cell 157(4):897–909.

16. Iwata S, Saynovits M, Link TA, Michel H (1996) Structure of a water soluble fragmentof the ‘Rieske’ iron-sulfur protein of the bovine heart mitochondrial cytochrome bc1complex determined by MAD phasing at 1.5 A resolution. Structure 4(5):567–579.

17. Zhang Z, et al. (1998) Electron transfer by domain movement in cytochrome bc1.Nature 392(6677):677–684.

18. Yoneda T, et al. (2004) Compartment-specific perturbation of protein handling acti-vates genes encoding mitochondrial chaperones. J Cell Sci 117(Pt 18):4055–4066.

19. Mansfield KD, et al. (2005) Mitochondrial dysfunction resulting from loss of cyto-chrome c impairs cellular oxygen sensing and hypoxic HIF-alpha activation. Cell Metab1(6):393–399.

20. Jamieson D (1989) Oxygen toxicity and reactive oxygen metabolites in mammals. FreeRadic Biol Med 7(1):87–108.

21. Van Voorhies WA, Ward S (2000) Broad oxygen tolerance in the nematode Caeno-rhabditis elegans. J Exp Biol 203(Pt 16):2467–2478.

22. Guzy RD, et al. (2005) Mitochondrial complex III is required for hypoxia-induced ROSproduction and cellular oxygen sensing. Cell Metab 1(6):401–408.

23. Guzy RD, Schumacker PT (2006) Oxygen sensing by mitochondria at complex III: Theparadox of increased reactive oxygen species during hypoxia. Exp Physiol 91(5):807–819.

24. Suthammarak W, Morgan PG, Sedensky MM (2010) Mutations in mitochondrialcomplex III uniquely affect complex I in Caenorhabditis elegans. J Biol Chem 285(52):40724–40731.

25. Beckmann JD, Ljungdahl PO, Lopez JL, Trumpower BL (1987) Isolation and charac-terization of the nuclear gene encoding the Rieske iron-sulfur protein (RIP1) fromSaccharomyces cerevisiae. J Biol Chem 262(18):8901–8909.

26. Brandt U, Haase U, Schägger H, von Jagow G (1991) Significance of the “Rieske” iron-sulfur protein for formation and function of the ubiquinol-oxidation pocket of mi-tochondrial cytochrome c reductase (bc1 complex). J Biol Chem 266(30):19958–19964.

27. Crofts AR, et al. (1999) Mechanism of ubiquinol oxidation by the bc(1) complex: Roleof the iron sulfur protein and its mobility. Biochemistry 38(48):15791–15806.

28. Crofts AR, et al. (1999) Mechanism of ubiquinol oxidation by the bc(1) complex:Different domains of the quinol binding pocket and their role in the mechanism andbinding of inhibitors. Biochemistry 38(48):15807–15826.

29. Crofts AR, Hong S, Zhang Z, Berry EA (1999) Physicochemical aspects of the movementof the rieske iron sulfur protein during quinol oxidation by the bc(1) complex frommitochondria and photosynthetic bacteria. Biochemistry 38(48):15827–15839.

30. Darrouzet E, Valkova-Valchanova M, Moser CC, Dutton PL, Daldal F (2000) Uncoveringthe [2Fe2S] domain movement in cytochrome bc1 and its implications for energyconversion. Proc Natl Acad Sci USA 97(9):4567–4572.

31. Moser CC, Keske JM, Warncke K, Farid RS, Dutton PL (1992) Nature of biologicalelectron transfer. Nature 355(6363):796–802.

32. von Jagow G, Ohnishi T (1985) The chromone inhibitor stigmatellin–binding to theubiquinol oxidation center at the C-side of the mitochondrial membrane. FEBS Lett185(2):311–315.

33. Matsuura K, Bowyer JR, Ohnishi T, Dutton PL (1983) Inhibition of electron transfer by3-alkyl-2-hydroxy-1,4-naphthoquinones in the ubiquinol-cytochrome c oxidoreduc-tases of Rhodopseudomonas sphaeroides and mammalian mitochondria. Interaction

with a ubiquinone-binding site and the Rieske iron-sulfur cluster. J Biol Chem 258(3):1571–1579.

34. Bowyer JR, Dutton PL, Prince RC, Crofts AR (1980) The role of the Rieske iron-sulfurcenter as the electron donor to ferricytochrome c2 in Rhodopseudomonas sphaer-oides. Biochim Biophys Acta Bioenergetics 592(3):445–460.

35. Iwata S, et al. (1998) Complete structure of the 11-subunit bovine mitochondrial cy-tochrome bc1 complex. Science 281(5373):64–71.

36. Beckmann JD, Ljungdahl PO, Trumpower BL (1989) Mutational analysis of the mito-chondrial Rieske iron-sulfur protein of Saccharomyces cerevisiae. I. Construction of aRIP1 deletion strain and isolation of temperature-sensitive mutants. J Biol Chem264(7):3713–3722.

37. Gatti DL, Meinhardt SW, Ohnishi T, Tzagoloff A (1989) Structure and function of themitochondrial bc1 complex. A mutational analysis of the yeast Rieske iron-sulfurprotein. J Mol Biol 205(2):421–435.

38. Darrouzet E, Daldal F (2002) Movement of the iron-sulfur subunit beyond the ef loopof cytochrome b is required for multiple turnovers of the bc1 complex but not forsingle turnover Qo site catalysis. J Biol Chem 277(5):3471–3476.

39. Crofts AR, Shinkarev VP, Dikanov SA, Samoilova RI, Kolling D (2002) Interactions ofquinone with the iron-sulfur protein of the bc(1) complex: Is the mechanism spring-loaded? Biochim Biophys Acta 1555(1-3):48–53.

40. Esser L, et al. (2004) Crystallographic studies of quinol oxidation site inhibitors: Amodified classification of inhibitors for the cytochrome bc(1) complex. J Mol Biol341(1):281–302.

41. Darrouzet E, Daldal F (2003) Protein-protein interactions between cytochrome b andthe Fe-S protein subunits during QH2 oxidation and large-scale domain movement inthe bc1 complex. Biochemistry 42(6):1499–1507.

42. Darrouzet E, Valkova-Valchanova M, Daldal F (2002) The [2Fe-2S] cluster Em as anindicator of the iron-sulfur subunit position in the ubihydroquinone oxidation site ofthe cytochrome bc1 complex. J Biol Chem 277(5):3464–3470.

43. Andreu AL, Checcarelli N, Iwata S, Shanske S, DiMauro S (2000) A missense mutationin the mitochondrial cytochrome b gene in a revisited case with histiocytoid cardio-myopathy. Pediatr Res 48(3):311–314.

44. Dumoulin R, et al. (1996) A novel gly290asp mitochondrial cytochrome b mutationlinked to a complex III deficiency in progressive exercise intolerance. Mol Cell Probes10(5):389–391.

45. Fisher N, et al. (2004) Human disease-related mutations in cytochrome b studied inyeast. J Biol Chem 279(13):12951–12958.

46. Kaeberlein TL, et al. (2006) Lifespan extension in Caenorhabditis elegans by completeremoval of food. Aging Cell 5(6):487–494.

47. Smith ED, et al. (2008) Age- and calorie-independent life span extension from dietaryrestriction by bacterial deprivation in Caenorhabditis elegans. BMC Dev Biol 8:49.

48. Sutphin GL, Kaeberlein M (2008) Dietary restriction by bacterial deprivation increaseslife span in wild-derived nematodes. Exp Gerontol 43(3):130–135.

49. Sutphin GL, Kaeberlein M (2009) Measuring Caenorhabditis elegans life span on solidmedia. J Vis Exp (27):1152.

50. Schneider CA, Rasband WS, Eliceiri KW (2012) NIH Image to ImageJ: 25 years of imageanalysis. Nat Methods 9(7):671–675.

51. Stiernagle T (2006) Maintenance of C. elegans. WormBook. Available at www.ncbi.nlm.nih.gov/books/NBK19649/. Accessed February 15, 2014.

52. Falk MJ, Kayser EB, Morgan PG, Sedensky MM (2006) Mitochondrial complex Ifunction modulates volatile anesthetic sensitivity in C. elegans. Curr Biol 16(16):1641–1645.

53. Kayser EB, Hoppel CL, Morgan PG, Sedensky MM (2003) A mutation in mitochondrialcomplex I increases ethanol sensitivity in Caenorhabditis elegans. Alcohol Clin Exp Res27(4):584–592.

54. Johnson SC, et al. (2013) mTOR inhibition alleviates mitochondrial disease in a mousemodel of Leigh syndrome. Science 342(6165):1524–1528.

55. Bianchi L, Driscoll M (2006) Culture of embryonic C. elegans cells for electrophysio-logical and pharmacological analyses. WormBook. Available at www.ncbi.nlm.nih.gov/books/NBK19713/. Accessed February 15, 2014.

56. Olsen B, Murakami CJ, Kaeberlein M (2010) YODA: Software to facilitate high-throughput analysis of chronological life span, growth rate, and survival in buddingyeast. BMC Bioinformatics 11:141.

57. Murakami C, Kaeberlein M (2009) Quantifying yeast chronological life span by out-growth of aged cells. J Vis Exp (27):1156.

Jafari et al. PNAS | Published online October 26, 2015 | E6157

CELL

BIOLO

GY

PNASPL

US

Dow

nloa

ded

by g

uest

on

Mar

ch 1

0, 2

021