ABSTRACT
Diapause exhibited by embryos of Artemia franciscana is accompanied by severe arrest of respiration. A large fraction of this depression is attributable to downregulation of trehalose catabolism that ultimately restricts fuel to mitochondria. This study now extends knowledge on the mechanism by revealing metabolic depression is heightened by inhibitions within mitochondria. Compared with that in embryo lysates during post-diapause, oxidative phosphorylation (OXPHOS) capacity P is depressed during diapause when either NADH-linked substrates (pyruvate and malate) for electron transfer (electron transfer capacity, E) through respiratory Complex I or the Complex II substrate succinate are used. When pyruvate, malate and succinate were combined, respiratory inhibition by the phosphorylation system in diapause lysates was discovered as judged by P/E flux control ratios (two-way ANOVA; F1,24=38.78; P<0.0001). Inhibition was eliminated as the diapause extract was diluted (significant interaction term; F2,24=9.866; P=0.0007), consistent with the presence of a diffusible inhibitor. One candidate is long-chain acyl-CoA esters known to inhibit the adenine nucleotide translocator. Addition of oleoyl-CoA to post-diapause lysates markedly decreased the P/E ratio to 0.40±0.07 (mean±s.d.; P=0.002) compared with 0.79±0.11 without oleoyl-CoA. Oleoyl-CoA inhibits the phosphorylation system and may be responsible for the depressed P/E in lysates from diapause embryos. With isolated mitochondria, depression of P/E by oleoyl-CoA was fully reversed by addition of l-carnitine (control versus recovery with l-carnitine, P=0.338), which facilitates oleoyl-CoA transport into the matrix and elimination by β-oxidation. In conclusion, severe metabolic arrest during diapause promoted by restricting glycolytic carbon to mitochondria is reinforced by depression of OXPHOS capacity and the phosphorylation system.
INTRODUCTION
The diapause program is a developmental pathway that delays direct morphogenesis and is characterized by ontogenetic arrest for a period spanning weeks to months (Koštál, 2006; Hand et al., 2016). Diapause is genetically programmed and triggered by endogenous physiological factors in response to environmental cues; typically, entry into this state precedes the onset of adverse environmental conditions and thus can serve to prepare the organism for ensuing stresses (Denlinger, 2002; Denlinger et al., 2012; Koštál, 2006). In contrast, quiescence is a dormant state directly imposed on the organism by an unfavorable and acute change in environmental conditions (e.g. anoxia, desiccation). Initiation of diapause involves differential gene expression that can control a number of features of the diapause phenotype (Hahn and Denlinger, 2011; Koštál et al., 2017; Qiu and MacRae, 2010; Reynolds, 2017). A growing body of work on insect diapause has shed some light on possible mechanisms of hormonal control of the endogenous triggers such as diapause hormone and ecdysteroids (for reviews, see Hahn and Denlinger, 2011; Denlinger et al., 2012). Developmental arrest may be accompanied by metabolic arrest depending on the developmental stage and the species in which diapause occurs (Denlinger, 2002; Reynolds and Hand, 2009; Hahn and Denlinger, 2011; Hand et al., 2011; Denlinger et al., 2012; Podrabsky and Hand, 2015; Popović et al., 2021; Lebenzon et al., 2022). In some species, metabolic depression is not constantly maintained, and bouts of metabolic activity occur, presumably to replenish fuel (Hahn and Denlinger, 2011; Chen et al., 2021). In contrast, embryos of the brine shrimp Artemia franciscana show arguably one of the deepest metabolic arrests associated with diapause (Clegg et al., 1996; Reynolds and Hand, 2004; Patil et al., 2013; Podrabsky and Hand, 2015). Respiration is depressed by over 99%, and part of the arrest is attributable to restriction of metabolic fuel to the mitochondrion (Patil et al., 2013). In this study, we report evidence that there is also diminished oxidative phosphorylation (OXPHOS) capacity and inhibition of the phosphorylation system as measured in concentrated lysates of diapause embryos of A. franciscana.
Adult A. franciscana females switch from ovoviviparous reproduction (release of free-swimming larvae) to oviparous reproduction (release of embryos destined for diapause) in response to environmental cues such as day length (Dutrieu, 1960; Berthelemy-Okazaki and Hedgecock, 1987; Drinkwater and Clegg, 1991; for life cycle diagram, see Podrabsky and Hand, 2015). The embryos thus formed are encysted and released into the water column where over a period of days they exhibit respiratory depression (Clegg et al., 1996; Patil et al., 2013). It should be noted that the metabolic depression is not synchronized with developmental arrest. Development is halted at the gastrula stage before the encysted embryos are released (Clegg and Conte, 1980; Berthelemy-Okazaki and Hedgecock, 1987). When released from the adult female, the diapause-destined embryos have respiration rates that are close to those of actively metabolizing embryos. However, over a period of about 5 days post-release, respiration decreases dramatically to barely detectable levels (Clegg et al., 1996; Patil et al., 2013). While known to be a contributor to metabolic depression during anoxia-induced quiescence in A. franciscana embryos (e.g. Busa and Crowe, 1983; Carpenter and Hand, 1986; Hand and Gnaiger, 1988), intracellular acidification apparently is not required for maintaining diapause in A. franciscana, and alkalinization of diapause embryos does not enable these cysts to hatch (Drinkwater and Crowe, 1987), which supports the notion that mechanisms regulating diapause and quiescence are different. Indeed, inhibition sites within the glycolytic pathway differ between diapause and quiescence (Carpenter and Hand, 1986; Patil et al., 2013). Whether or not acidification of pHi has a role in the metabolic arrest during diapause is still debated (Clegg, 2011).
Reynolds and Hand (2004) investigated mitochondrial bioenergetics of diapause embryos of A. franciscana and found that, at least in isolated mitochondrial preparations, there were no major differences in mitochondrial function between diapause embryos and post-diapause embryos (ones for which diapause has been terminated). However, the results with isolated mitochondria do not rule out the possibility that there could have been removal of an endogenous inhibitor during the purification process or a change in covalent modification of mitochondrial enzymes during the time required for isolation. For example, by homogenizing embryos in an SDS denaturing buffer, Patil et al. (2013) were able to demonstrate that phosphorylation of pyruvate dehydrogenase (which is associated with inhibition of the enzyme) occurred during entry into diapause.
Here, we report evidence that supports an inhibition of OXPHOS during diapause by using lysates prepared from whole embryos of A. franciscana. The potential presence of inhibitors in diapause lysates was interrogated by evaluating oxygen consumption at multiple dilutions of lysates. ADP-stimulated respiration rates were measured along with chemically non-coupled respiration to evaluate the capacity of OXPHOS (P) and of electron transfer (E) of the mitochondria with NADH-linked substrates (pyruvate and malate) for electron transfer through respiratory Complex I, with the Complex II substrate succinate, or with pyruvate, malate and succinate in combination. Evidence shows that respiration fueled by pyruvate+malate or by succinate is substantially diminished in lysates from diapause embryos. Further, P/E flux control ratios (Gnaiger, 2020) indicate that the phosphorylation system is inhibited in diapause and that this inhibition is eliminated as the extract is diluted, a result consistent with the presence of an inhibitor in diapause embryos. Previous work with mammalian hibernation suggested inhibition of mitochondrial respiration may occur as a result of inhibition of the adenine nucleotide transporter (ANT) (Lerner et al., 1972) via the action of long-chain acyl-CoA esters (Chua and Shrago, 1977; Lerner et al., 1972; Pande and Blanchaer, 1971; Soboll et al., 1984). Because the ANT is a component of the phosphorylation system, we evaluated the effect of oleoyl-CoA and found a depressive impact of oleoyl-CoA on P/E flux control ratios and its reversal by l-carnitine, which provides proof of principle that long-chain acyl-CoA esters are viable candidates for such an inhibitor, operative during diapause.
MATERIALS AND METHODS
Reagents
Trehalose (product number T-104-4) was purchased from Pfanstiehl (Waukegan, IL, USA), sucrose was from J. T. Baker (Paris, KY, USA), oleoyl-CoA (product number 870719P) was from Avanti Polar Lipids (Birmingham, AL, USA) and bovine serum albumin (BSA, product number A6003, fatty acid free, Fraction V) was from Sigma-Aldrich (St Louis, MO, USA). All other reagents used were of the highest quality available and were purchased from either Fisher Scientific (Pittsburgh, PA, USA) or Sigma-Aldrich.
Processing of A. franciscana embryos
For this study, hydrated diapause embryos of Artemia franciscana Kellogg 1906 were collected from the surface of the Great Salt Lake, UT, USA, across multiple seasons. Dehydration terminates the diapause state so must be avoided during handling of embryos at all points after collection (for handling of diapause embryos, see Clegg et al., 1996). Year to year differences in mitochondrial properties have been previously documented (Reynolds and Hand, 2004). The embryos were washed and stored in 1.25 mol l−1 NaCl with kanamycin (50 μg ml−1), penicillin–streptomycin (50 μg ml−1) and nystatin (200 units ml−1) at room temperature in the dark. Prior to each assay, diapause embryos were incubated in 35 PSU artificial seawater (Instant Ocean™, Aquarium Systems, Mentor, OH, USA) with orbital shaking (110 rpm) at room temperature for 4 days in the dark to allow non-diapause embryos to hatch; larvae and empty shells were removed using a separatory funnel. The viability of diapause embryos was confirmed periodically following the method of Reynolds and Hand (2004) with minor modifications (LeBlanc et al., 2019). Briefly, unhatched cysts were dried at ambient temperature and humidity for up to 2 weeks. The dried cysts were treated with 3% hydrogen peroxide prepared in 0.25 mol l−1 NaCl for approximately 30 min. The treatment with hydrogen peroxide is a standard technique for terminating the diapause state in these embryos (cf. Lavens et al., 1986; Clegg et al., 1996). After rinsing thoroughly with 0.25 mol l−1 NaCl, the cysts were incubated as described above (35 PSU artificial seawater, 4 days). After 4 days, the percentage hatching was determined. The number of nauplii divided by the total number counted (i.e. unhatched cysts+nauplii) was used to calculate percentage viability. Cyst batches with a hatching percentage above 70% were considered to be suitable for study. It should be noted that some cysts are exceptionally refractory to diapause termination with the above technique depending on the depth of diapause.
Dehydrated post-diapause embryos (Grade: laboratory reference standard) from the Great Salt Lake were obtained from Great Salt Lake Artemia (Dr Brad Marden, Research Division, Ogden, UT, USA). Experiments used post-diapause embryos collected and processed from different years. The dried embryos were stored at −20°C. In preparation for use in experiments, these embryos were hydrated in 0.25 mol l−1 NaCl at 0°C overnight. The embryos were then allowed to develop under normoxic conditions in 0.25 mol l−1 NaCl at room temperature for 6–8 h depending on the commercial batch. The hatching percentage for post-diapause embryos, as determined above in 35 PSU artificial seawater, was above 90%.
Preparation of embryo lysates and mitochondria for respiration measurements
Prior to homogenization, the hydrated cysts (10–20 g) were dechorionated by treatment with antiformin solution (1% hypochlorite, 0.4 mol l−1 sodium hydroxide and 60 mmol l−1 sodium carbonate) for 10–15 min at room temperature, followed by three rinses with ice-cold 0.25 mol l−1 NaCl. Embryos were then incubated in ice-cold 1% (w/v) sodium thiosulfate for 5 min and were rinsed two times with cold 0.25 mol l−1 NaCl. Finally, embryos were incubated for 3–5 min in cold 40 mmol l−1 hydrochloric acid prepared in 0.25 mol l−1 sodium chloride, followed by three rinses with 0.25 mol l−1 NaCl. Hypochlorite treatment of diapause embryos (2.5%, 2 min) is a routine way to aseptically clean embryos prior to storing them in the diapause state (Clegg et al., 1996). Extending the incubation in 1% hypochlorite for several minutes at room temperature removes the chorion layer, followed by immediate chemical inactivation of the hypochlorite with thiosulfate at 0°C, allowing gentle homogenization of embryos at 0°C, which retains mitochondrial integrity. Dechorionation of A. franciscana embryos may improve the hatching percentage after 4 days.
After dechorionation, 10 g of embryos (approximately 100,000 embryos g−1; Stocco et al., 1972) was homogenized 1:1 (w/v) with a Teflon-glass Potter-Elvehjem tissue homogenizer (Thomas Scientific, Swedesboro, NJ, USA) in ice-cold medium composed of 0.5 mol l−1 sucrose (or trehalose), 150 mmol l−1 KCl, 2 mmol l−1 MgCl2, 10 mmol l−1 KH2PO4, 1 mmol l−1 EGTA, 0.5% (w/v) bovine serum albumin (fatty acid free, fraction V) and 20 mmol l−1 Hepes, titrated to pH 7.5 with 1.0 mol l−1 KOH. Sucrose and trehalose are equally effective as an osmotic agent in this context. The resulting homogenate was centrifuged at 1000 g for 10 min at 4°C. The semi-transparent supernatant was collected with a 10 ml pipette, taking care to exclude the entire uppermost layer of lipoprotein and any material from the pellet fraction, and then stored on ice. Generally, about 6.8 ml of this concentrated supernatant can be prepared from 10 g of hydrated embryos. Protein content was measured using either the Coomassie Plus (Bradford) assay kit (Thermo Fisher Scientific Inc., Rockford, IL, USA) or a modified Lowry assay as described by Peterson (1977).
Mitochondria were isolated from post-diapause embryos of A. franciscana that were incubated at 23°C for 6 h and then dechorionated, as previously described (Kwast and Hand, 1993; Reynolds and Hand, 2004), with minor modifications; BSA was omitted from the homogenization medium. These mitochondria were used for experiments on the reversibility of the effects of oleoyl-CoA on P/E flux control ratios as described in Results.
Oxygen flux measurements
Respiration measurements on embryo lysates were made at 25°C with an Oxygraph O2k (Oroboros Instruments, Innsbruck, Austria) with polarographic oxygen sensors (Gnaiger, 2008). Oxygen concentration and its time derivative automatically corrected for instrumental background, oxygen flux, were digitally recorded with Oroboros Datlab 7 software. Oxygen sensors were calibrated routinely at air saturation and in oxygen-depleted media. The 2 ml respiration chamber was filled with undiluted lysate, or lysate was added to homogenization medium in the chamber to yield final dilutions of 1/4, 1/10 and 1/100. Because of the high oxygen fluxes measured for the undiluted lysate and the 1/4 diluted lysate, oxygen concentration of the respiration chamber was elevated by raising the stopper partially, flushing the gas phase with pure O2, allowing oxygenation of the liquid phase with stirring for the time necessary to reach 500–600 nmol O2 ml−1, and then closing the chamber.
As depicted in the representative respirometric trace shown in Fig. 1, steady-state oxygen flux was measured first without any additions, which corresponded to respiration with endogenous substrates and endogenous ADP (denoted as e). Next, saturating ADP (final concentration 2.5 mmol l−1) was added to the chamber and respiration was recorded as eP, i.e. respiration with endogenous substrates and saturating, exogenous ADP to measure OXPHOS capacity, P, limited by the availability of endogenous substrates (at high dilutions of lysate, ADP concentration was lowered in some cases because of very low flux values). ADP-stimulated respiration with saturating NADH-linked substrates was initiated by adding 15 mmol l−1 pyruvate plus 2 mmol l−1 malate (ePMP). ADP-stimulated respiration with saturating Complex II substrate was initiated by adding 10 mmol l−1 succinate (eSP), often performed in separate trials. It is appropriate to note that all assays were performed without rotenone (specific inhibitor of Complex I). Previous experiments have verified that only modest differences are observed in succinate-stimulated respiration in the presence or absence of rotenone for these mitochondria (Hand and Gnaiger, 2014); any inhibition of succinate dehydrogenase due to the anticipated oxaloacetate accumulation is small. Moreover, malate alone does not support significant respiration (S.C.H., unpublished), and thus its generation from succinate is inconsequential presumably because of the low malate dismutation to pyruvate (i.e. low mitochondrial malic enzyme activity). Additionally, endogenous substrates are present that may support further modification of oxaloacetate by citrate synthase or transamination (Gnaiger, 2020), which provides a rationale to eliminate rotenone in the present protocol with lysates.
ADP-stimulated respiration with saturating NADH-linked substrates plus succinate was initiated by adding 15 mmol l−1 pyruvate, 2 mmol l−1 malate and 10 mmol l−1 succinate (ePMSP). In some cases, ePMP and ePMSP respiration were measured in separate trials. Next, the respiration mixture was titrated with 2–4 µmol l−1 carbonyl cyanide 4-(trifluoromethoxy)phenylhydrazone (FCCP; typically, 0.5 µmol l−1 increments) to determine the maximum electron transfer capacity (E) at optimal uncoupler concentration; for more concentrated extracts, an initial addition of FCCP was made followed by 0.5 µmol l−1 increments to reach the maximum non-coupled rate, in order to constrain the time and oxygen utilization from the chamber. Finally, residual oxygen consumption (Rox) was measured by addition of 2.5 µmol l−1 antimycin A and used as a baseline correction of mitochondrial activity. During respirometry runs, the oxygen in the chamber was periodically replenished to the starting values with the procedure described above. The extremely high volume-specific oxygen fluxes in the lysates at low dilution (i.e. 0.25 dilution or undiluted) could be resolved only by using a minimum data sampling interval of 0.2 s (with 100 signals averaged for each data point) and calculating the slope of oxygen concentration over time from a spline fit function over only 5 data points. For presentation and comparison across lysate dilutions, oxygen flux was expressed per ml undiluted lysate or per mg lysate protein with the appropriate conversion factor. In this way, i.e. by converting flux back to values reflective of mitochondrial contents in original (undiluted) lysates, it is easier to detect potential inhibitors (or other effectors) whose impact should intensify for assays conducted at higher concentrations of lysate.
To test for oleoyl-CoA inhibition of respiration with post-diapause lysates at 0.25 dilution, 10 mmol l−1 succinate was added and then 30 μmol l−1 oleoyl-CoA, followed by either 1.0 mmol l−1 or 2.5 mmol l−1 ADP (because of the competition between ADP and long-chain acyl-CoA ester for the ANT binding site). This concentration of oleoyl-CoA (and lower) has been used previously for ANT inhibition with mammalian tissue preparations and mitochondria (Pande and Blanchaer, 1971; Chua and Shrago, 1977). For some experiments with oleoyl-CoA, the 0.5% BSA was omitted from the respiration medium (because of its known ability to chelate long-chain acyl-CoAs). FCCP titration was performed as before. Finally, 2.5 μmol l−1 antimycin A was added to quantify Rox.
For isolated mitochondria, oxygen consumption was measured in respiration medium consisting of 500 mmol l−1 trehalose, 150 mmol l−1 KCl, 20 mmol l−1 Hepes, 10 mmol l−1 KH2PO4, 2 mmol l−1 MgCl2 and 1 mmol l−1 EGTA, without BSA, pH 7.5. Based on preliminary results for oleoyl-CoA titrations, 3 µmol l−1 oleoyl-CoA was chosen to quantify the inhibitory action of the compound with isolated mitochondria. The sequence of additions was mitochondria (0.2–0.3 mg protein), 10 mmol l−1 succinate, 3 µmol l−1 oleoyl-CoA (omitted for controls), 0.375 mmol l−1 ADP and FCCP (serial titrations). To test for the reversal of oleoyl-CoA inhibition by l-carnitine, the additions were mitochondria (0.2–0.3 mg protein), 3 µmol l−1 oleoyl-CoA, 1 mmol l−1 malate plus 5 mmol l−1l-carnitine, 0.375 mmol l−1 ADP, a 15 min pause to permit oleoyl-CoA transport and oxidation, then 10 mmol l−1 succinate, 0.375 mmol l−1 ADP, and FCCP titration. Oxygen in the chamber was periodically replenished to the starting values.
Statistical analyses
One-way and two-way ANOVA were performed with GraphPad Prism software (v.10.1.0, GraphPad Software, Boston, MA, USA). Post hoc pairwise comparisons were made using Dunnett's multiple comparisons tests (one-way ANOVA) and unpaired two-tailed t-tests with Holm–Šidák adjustments for multiple comparisons (two-way ANOVA). P≤0.05 was taken to indicate statistical significance.
RESULTS
The quantity of mitochondria in diapause embryos is statistically indistinguishable from that in post-diapause embryos, based on the mitochondrial protein extractable from embryos (Reynolds and Hand, 2004). Thus, decreased respiration in diapause versus post-diapause lysates reported below was not due to a difference in mitochondrial density. It is appropriate to note that when respirometry data were collected across dilutions of lysate, the final values presented for oxygen flux were converted to per ml undiluted lysate or per mg lysate protein. In this way, the presence of potential inhibitors (or other effectors), whose impact should intensify at the higher concentrations of lysate, is more readily detectable.
OXPHOS capacity with NADH-linked substrates is reduced during diapause
Oxygen flux measured for post-diapause and diapause lysates is presented as box plots for four different respiratory treatments at four different lysate dilutions in Fig. 2. Oxygen consumption supported by endogenous fuel alone (e) is given first and followed by values after exogenous ADP was added (eP). At high lysate dilution (0.01, 0.1), the addition of ADP had a modest effect because of the limited supply of endogenous carbon substrate, but generated larger increases in most cases for the more concentrated lysates (0.25, 1.0). ePMP was measured next by addition of exogenous pyruvate and malate in the presence of ADP. Addition of exogenous pyruvate and malate to undiluted (1.0) lysates did not increase oxygen consumption, which suggests the presence of significant concentrations of endogenous fuel. Titration with the chemical uncoupler FCCP was used to stimulate non-coupled respiration and provide the maximum electron transfer capacity obtainable with NADH-linked substrates (ePME). Generally, the values for eP, ePMP and ePME were lower overall for diapause lysates than for post-diapause at all dilutions. Depressed respiration in lysates from diapause embryos fueled by endogenous substrates is consistent with generally lower levels of substrate and the blockage of carbohydrate supply to the mitochondrion during diapause (Patil et al., 2013) when compared with post-diapause embryos.
To allow more quantitative comparisons between post-diapause and diapause respiration, selected data have been re-plotted in Fig. 3A. Two-way ANOVA showed that both diapause status (F1,28=612.7; P<0.0001) and dilution (F3,28=360.6; P<0.0001) had significant impacts on eP values. Post hoc pairwise comparisons indicated the eP values were lower for diapause versus post-diapause lysates at all dilutions (all P≤0.0002). Likewise, in the case of ePMP values, a second two-way ANOVA indicated that both diapause (F1,28=2877; P<0.0001) and dilution (F3,28=239.2; P<0.0001) had significant impacts. Post hoc comparisons confirmed that ePMP values were lower for diapause versus post-diapause lysates at all dilutions (all P≤0.00003).
OXPHOS capacity with the Complex II substrate succinate is strongly depressed in diapause
Succinate was used to evaluate OXPHOS when electrons enter the electron transport system (ETS) via Complex II (Fig. 3B). Two-way ANOVA indicated that both diapause status (F1,63=217.5; P<0.0001) and dilution (F3,63=333.1; P<0.0001) had significant impacts on eP values. The interaction term was also significant (F3,63=54.91; P<0.0001). Post hoc pairwise comparisons show that the eP values were lower for diapause versus post-diapause lysates at 0.1, 0.25 and 1.0 dilution (all P<0.000001). In the case of eSP values, a second two-way ANOVA indicated that both diapause status (F1,51=34.49; P<0.001) and dilution (F3,51=7.998; P<0.0002) had significant impacts. Post hoc pairwise comparisons were made, which confirmed that eSP values were lower for diapause versus post-diapause lysates at 0.1, 0.25 and 1.0 dilution (all P≤0.04).
OXPHOS capacity with dual input to the Q-junction is not strongly additive
For addition of the substrate combination pyruvate, malate and succinate (Fig. 3C), two-way ANOVA showed that both diapause (F1,32=62.34; P<0.0001) and dilution (F3,32=338.9; P<0.0001) had significant impacts on ePMSP. The interaction term was significant (F3,32=32.46; P<0.0001), and thus post hoc pairwise comparisons were conducted. Values for ePMSP differed for diapause versus post-diapause lysates at 0.01 and 0.25 dilution (both P≤0.03). These results indicated that when NADH-linked substrates and succinate were provided in combination, the oxygen fluxes generated were not strongly additive as judged by total flux values. While the result is not unusual, supply of pyruvate, malate and succinate simultaneously does offer the potential in some systems to further increase electron flow through the ETS by convergent entry into the ubiquinone pool (i.e. the Q-junction) because Complexes I and II operate simultaneously (Gnaiger, 2020). An unanticipated aspect with pyruvate and malate and succinate was the absence of inhibition in undiluted diapause lysate (P=0.128) observed previously with succinate alone or pyruvate and malate alone.
P/E flux control ratios reveal inhibition of the phosphorylation system
Results for 1/100-diluted lysates (0.01) were eliminated from the evaluation of P/E flux control ratios when NADH-linked substrates were involved, because for both diapause and post-diapause embryos, the ratios were above 1.0 as calculated from Fig. 2, even at the low concentrations of FCCP typically used in titrations (0.05–0.15 µmol l−1). Thus, it appears that at 0.01 dilution of lysates, NADH-linked respiration was very sensitive to FCCP inhibition. Accordingly, we evaluated this issue further using an even lower concentration range of FCCP (0.001–0.01 µmol l−1); inhibition of oxygen flux was observed at these FCCP concentrations as well. Consequently, it was not possible to demonstrate stimulation by FCCP when 0.01 dilution lysates were driven by NADH-linked substrates. Perhaps evaluating alternative chemical uncouplers could be useful in the future. For example, BAM15 has emerged as an effective mitochondrial uncoupler that is less toxic than FCCP and, importantly, exhibits only slight inhibit of mitochondrial respiration even at high concentrations compared with FCCP (Alexopoulos et al., 2020; Axelrod et al., 2020; Kenwood et al., 2014).
As shown in Fig. 4A for the NADH-linked substrates pyruvate and malate, two-way ANOVA indicated that dilution (F2,18=38.87; P<0.0001) did have an effect on P/E ratios, but that diapause status did not (F1,18=1.756; P=0.2017). The interaction term was significant (F2,18=4.355; P=0.0287). Accordingly, post hoc pairwise comparisons using unpaired t-tests showed that P/E ratios differed only at 0.1 dilution (P=0.02). A P/E ratio below 1.0 indicates that the maximum coupled respiration is limited by the phosphorylation system (i.e. the F1Fo ATP synthase, ANT and phosphate transporter), and thus there is apparent excess capacity of the ETS that is not utilized. In contrast to the pyruvate and malate results, evaluation by two-way ANOVA for succinate (Fig. 4B) showed that both diapause (F1,32=120.1; P<0.0001) and dilution (F3,32=16.76; P<0.0001) had a significant impact on P/E ratios. The interaction term was not significant (F3,32=0.6782; P=0.5718), indicating that the effect of dilution did not depend upon the diapause status of the embryos. Importantly, post hoc pairwise comparisons indicated that P/E ratios for diapause lysates were significantly lower than the corresponding values for post-diapause lysates at all dilutions (all P≤0.01). The higher overall electron transfer capacity (eSE) typically observed with the Complex II substrate compared with NADH-linked substrates in A. franciscana (Fig. 3A versus B) likely contributed to the stronger limitation by the phosphorylation system in diapause lysates.
P/E flux control ratios obtained with pyruvate and malate and succinate combined for diapause and post-diapause embryo lysates are depicted in Fig. 4C. Again, in this case, the patterns exhibited for post-diapause versus diapause were quite different. Two-way ANOVA indicated that both diapause status (F1,24=38.78; P<0.0001) and dilution (F2,24=119.5; P<0.0001) had significant impacts on P/E ratios. A significant interaction (F2,24=9.866; P=0.0007) indicated that the effect of lysate dilution depended upon the diapause status of embryos, with a stronger decrease in P/E ratio in undiluted extracts from diapause embryos. Post hoc pairwise comparisons confirmed that P/E ratios for diapause lysates when measured at 0.25 dilution and without dilution (1.0) were significantly lower than the corresponding values for post-diapause lysates (both P≤0.005). These data showed that during diapause, the phosphorylation system was limiting overall oxidative phosphorylation to a greater degree than during post-diapause. Further, the inhibition of the phosphorylation system in diapause lysates markedly diminished as a function of dilution, such that the P/E ratio became statistically identical to post-diapause values at the 0.1 dilution. These data provide strong support for the presence of a diffusible inhibitor that may exist at a higher titer during diapause. In other words, the inhibitory influence of a lysate component on the phosphorylation system in diapause embryos was eliminated by dilution of the lysate.
Long-chain acyl-CoA esters reversibly inhibit P/E flux control ratio
Lysates of post-diapause embryos (0.25 dilution) were used to evaluate possible inhibition of the phosphorylation system by long-chain acyl-CoA esters as previously shown for mammalian mitochondria, e.g. during hibernation, where the ANT can be inhibited by these esters in a competitive fashion against ADP. Upon addition of 30 µmol l−1 oleoyl-CoA to lysates, strong inhibition of the phosphorylation system occurred when BSA was omitted and ADP was reduced from 2.5 mmol l−1 to 1.0 mmol l−1 (compare Fig. 5A versus B); the P/E flux control ratio was significantly decreased from 0.79±0.11 (mean±s.d., N=3) to 0.40±0.07 (N=4) (P=0.002). Succinate was used as the substrate. To evaluate the reversibility of such inhibition, isolated mitochondria from post-diapause embryos of A. franciscana were prepared to provide a cleaner preparation with far lower contamination by endogenous substrates, fatty acid species and soluble proteins. In the absence of BSA, one-way ANOVA (Fig. 5C) shows highly significant differences exist among treatments (F2,9=119.7, P=0.0001). Dunnett's post hoc multiple comparisons tests indicate a significant decrease in P/E flux control ratio with succinate between control versus addition of 3 µmol l−1 oleoyl-CoA (P<0.0001) and no significant difference between control and recovery with l-carnitine addition, an indication of full reversibility of the effect (P=0.338). l-Carnitine accepts the acyl group from long-chain acyl-CoAs through the action of carnitine palmitoyl transferase I (CPT I), thereby forming acylcarnitine, which enters the matrix via the carnitine–acylcarnitine translocase and is eliminated by β-oxidation (Fig. 5D). These data taken together provide proof of principle that long-chain acyl-CoAs such as oleoyl-CoA could be viable candidates for inhibition of the phosphorylation system during diapause.
DISCUSSION
This study provides evidence for a decrease in respiration as assessed with NADH-linked substrates and succinate, and an inhibition of the phosphorylation system in A. franciscana embryos during diapause. The diminished activity with pyruvate and malate may be explained by compromised activity of Complex I, one or more substrate carriers, pyruvate dehydrogenase complex (PDHC), or combinations thereof. Inhibition of PDHC during diapause as a result of phosphorylation, reported by Patil et al. (2013), in principle could contribute to the above observation. Shifting the phosphorylation state of PDHC (i.e. pyruvate dehydrogenase complex, E1, site 1) maximally from dephosphorylated to phosphorylated promotes a 16.5-fold change in PDHC activity in mitochondria isolated from embryos of A. franciscana (D. Arabie and S.C.H., in preparation). Indeed, activation of PDHC by dephosphorylation is associated with periodic arousal from pupal diapause in the flesh fly (Chen et al., 2021). In terms of the reduction in succinate-based respiration, the SDHA subunit of Complex II can be physiologically regulated by phosphorylation–dephosphorylation, as well as by acetylation–deacetylation (Bezawork-Geleta et al., 2017). Because the inhibition observed for the phosphorylation system is diminished as the lysate is diluted, this result is consistent with a diffusible inhibitor molecule operating on the mitochondrion during diapause. Finally, evidence is presented that long-chain acyl-CoA esters are viable candidates for such an inhibitor operating during diapause. Together with the restriction of fuel to the mitochondrion in vivo (Patil et al., 2013), these features likely explain the majority of the metabolic arrest during diapause.
Depression of respiration during diapause
Levin et al. (2003) suggested that during diapause in insect larvae, a decrease in mitochondrial content could be responsible for decreased oxygen consumption. Elegant experimental work has recently supported this concept in diapausing Colorado potato beetles (Lebenzon et al., 2022), where reversible mitophagy is demonstrated with insect flight muscle. In contrast, hibernating ground squirrels are known to retain muscle mitochondria during hibernation as reviewed by Cotton (2016). Work by Reynolds and Hand (2004) on A. franciscana embryos indicates that the amount of mitochondrial marker protein does not differ between mitochondria from diapause and post-diapause embryos, suggesting that the observed decrease in embryonic respiration during diapause may be an active downregulation of mitochondrial function. Reynolds and Hand (2004) have reported that the bioenergetic features of mitochondria isolated from diapause and post-diapause embryos are similar. It is, however, possible that the mitochondria may display different behavior in situ. The current study used lysates prepared from A. franciscana embryos in an attempt to preserve some of the regulatory features that might have been lost during purification of isolated mitochondria.
Oxygen consumption in diapause lysates fueled by exogenously added NADH-linked substrates and succinate was substantially lower than that measured for post-diapause lysates at all dilutions. Because of the brief time needed for preparation and assay of the lysates, it is probable that covalent modifications of macromolecules existing in vivo were retained to some degree in these preparations. In addition to phosphorylation of PDHC mentioned above, kinase-mediated phosphorylation of dozens of proteins associated with all four respiratory complexes and the F1FO-ATPase has been identified (Phillips et al., 2011, and references therein). For example, phosphorylation of selected subunits of Complex I has been documented, but in vivo physiological roles for these modifications await verification (see Chen et al., 2004; Palmisano et al., 2007). S-Nitrosation of cysteine is prominently involved in downregulating Complex I during ischemia–reperfusion to reduce reactive oxygen species (ROS) generation (Chouchani et al., 2013). For Complex II, activity is increased through tyrosine phosphorylation of the flavoprotein catalytic subunit (FpSDH, SDHA) of succinate dehydrogenase and is mediated by the tyrosine kinase Fgr (Acín-Pérez et al., 2014). The activation of Complex II is a quick response (promoted by ROS) and does not involve gene expression or mitochondrial biogenesis (Acín-Pérez et al., 2014, and references therein). A second post-translational modification, acetylation–deacetylation, also impacts the activity of Complex II through the action of the NAD-dependent deacetylase SIRT3 (Cimen et al., 2010; Bezawork-Geleta et al., 2017).
Pyruvate import into the mitochondrion is mediated by a family of mitochondrial pyruvate carrier (MPC) proteins (Hildyard and Halestrap, 2003; Herzig et al., 2012). While the MPC proteins are considered vital to the task of transporting pyruvate into the mitochondrial matrix (Herzig et al., 2012; Bricker et al., 2012), there are no published reports of this carrier being physiologically regulated, other than its dependence on mitochondrial ΔpH. Thus, it appears that there are no regulatory features that would suggest that it could play a role in reducing the pyruvate and malate-dependent respiration observed in diapause lysates, although alterations in MPC protein expression could presumably contribute. Because pyruvate is decarboxylated by PDHC and yields acetyl-CoA, which must condense with oxaloacetate to form citrate, malate is a source of oxaloacetate that is needed by isolated mitochondria to permit pyruvate respiration. Malate can be transported into the matrix by three different carriers (tricarboxylate, dicarboxylate and 2-oxoglutarate carriers), and downregulation of one or more of these carriers could impede respiration on pyruvate and malate. Similarly, a downregulation of the dicarboxylate carrier could impact succinate/phosphate exchange.
When NADH-linked substrates and succinate were provided in combination with embryo lysates, complete additivity of respiratory flux was not observed, when compared with the separate additions of NADH-linked substrates and succinate (Figs 1 and 3). Lack of complete additivity of respiratory flux may be due to multiple factors, including regulatory mechanisms in the tricarboxylic acid (TCA) cycle, substrate competition for transport across the mitochondrial inner membrane, flux control by enzyme capacities downstream of the Q-junction, and the degree of channeling through supercomplexes (Bianchi et al., 2004; Cogliati et al., 2016; Haslam and Krebs, 1963; Komlódi et al., 2021; Lemieux et al., 2017; Schägger, 2002). Controversy remains regarding the existence of substrate channeling by supercomplexes (e.g. Fedor and Hirst, 2018; for review, see Vercellino and Sazanov, 2022). Expanding biochemical evidence does not support non-interacting quinone pools (e.g. one sequestered in supercomplexes such as CICIII2CIV, and another for the inner membrane-imbedded quinone). Nevertheless, the degree of additivity for flux observed with convergent electron transfer at the Q-junction varies widely among species, tissues and physiological states (Gnaiger, 2020), and, as previously pointed out, gives rise to a diversity of mitochondrial respiratory control patterns (Gnaiger, 2009).
Inhibition of the phosphorylation system
By increasing the input of electrons to the Q-junction by simultaneous addition of NADH-linked substrates and succinate, a limitation of oxidative phosphorylation by the phosphorylation system was detectable in diapause lysates. Limitation by the phosphorylation system in diapause lysates compared with post-diapause lysates was also observed with succinate, which supported high ETS activity with the single electron input into the Q-junction. As the diapause lysate was diluted, the lower P/E flux control ratios measured with pyruvate and malate and succinate returned to the values observed for post-diapause preparations (Fig. 4C). This result is consistent with the presence of an inhibitor specific to the diapause state. The identity of this compound is unknown, although there is precedence for metabolically derived inhibitors that can specifically impede the function of the ANT and thus restrict phosphorylation of ADP to ATP.
One example of such an inhibitor is long-chain acyl-CoA esters that are known to inhibit the ANT (e.g. oleoyl-CoA, palmitoyl-CoA, myristoyl-CoA) (Chua and Shrago, 1977; Lerner et al., 1972; Pande and Blanchaer, 1971; Soboll et al., 1984). Inhibition of mitochondrial respiration during mammalian hibernation has been suggested to occur as a result of inhibitors that specifically block the action of the ANT (Lerner et al., 1972). These compounds can inhibit from either side of the mitochondrial inner membrane (Chua and Shrago, 1977) and are modulated in response to various metabolic states (Prentki et al., 1992). Long-chain acyl-CoA esters are competitive inhibitors for the ADP binding site of the ANT, as elevation of ADP progressively reverses the inhibition (Pande and Blanchaer, 1971). Oleoyl-CoA at 30 μmol l−1 inhibited the phosphorylation system in lysates of post-diapause embryos (Fig. 5B) as judged by the decrease in P/E coupling control ratio. This depression was observed when ADP was lowered from 2.5 mmol l−1 to 1.0 mmol l−1 and in the absence of added BSA. The in vivo concentrations of ADP in diapause and post-diapause embryos of A. franciscana are approximately 0.13 mmol l−1 and 0.11 mmol l−1, respectively (Patil et al., 2013). In this study, the added ADP exceeded the in vivo concentration in order to support the kinetically ADP-saturated respiratory capacity of the lysates for extended periods. Nevertheless, the inhibition by oleoyl-CoA was observable even with ADP maintained 8-fold higher than measured in vivo. It is not surprising that the presence of BSA obscured the oleoyl-CoA effect because of the chelation by BSA of long-chain acyl-CoA molecules. BSA binds free fatty acids with high affinity (Spector et al., 1969); its affinity for long-chain acyl-CoA esters is 5–10 times lower, yet is still quite impressive (Kd for palmitoyl-, stearoyl- and oleoyl-CoA is 0.2–0.4 µmol l−1; Demant and Nystrøm, 2001). Because inhibition of the phosphorylation system in diapause lysates was observed in the presence of BSA (Figs 4B,C and 5), one would predict that in vivo concentrations of oleoyl-CoA during diapause would be higher than the 30 μmol l−1 used in Fig. 5B. This experimental concentration of oleoyl-CoA was chosen because of its potency in mitochondrial studies (Lerner et al., 1972; Chua and Shrago, 1977) and the prohibitive expense of using higher concentrations. In isolated mitochondria, addition of 5 mmol l−1 of l-carnitine reversed the inhibition of the ANT (Fig. 5C) by combining with long-chain acyl-CoA esters, decreasing their free concentrations, and promoting elimination by β-oxidation (Fig. 5D; also see Pande and Blanchaer, 1971; Lerner et al., 1972; Chua and Shrago, 1977).
In summary, this study underscores the importance of conducting bioenergetic analyses at multiple levels of biological complexity. Experiments with embryo lysates reveal new features contributing to the arrest of metabolism during diapause in A. franciscana that are not detectable with isolated mitochondria removed from the intracellular milieu, nor presumably with permeabilized cells or tissues. In order to further explore the possibility of an endogenous inhibitor during diapause, future experiments will quantify long-chain acyl-CoA esters in diapause and post-diapause embryos. Tractable methods exist for quantitative isolation and purification of short-, medium- and long-chain acyl-CoA esters from tissue extracts (Minkler et al., 2008).
Acknowledgements
The technical assistance of Drs Leaf Boswell, Apu Borcar and Daniel Moore during fieldwork is gratefully acknowledged. Logistics for fieldwork on the Great Salt Lake were provided by Dr Brad Marden and Great Salt Lake Artemia, Ogden, UT, USA. Appreciation is extended to Dr Barney Rees (University of New Orleans) for helpful comments and discussions regarding the study. Parts of the results in this article were previously published in the PhD thesis of Y.N.P. (Patil, 2012) and the honors thesis of A.P.L. (Landry, 2021).
Footnotes
Author contributions
Conceptualization: Y.N.P., E.G., S.C.H.; Methodology: E.G., S.C.H.; Validation: Y.N.P., E.G., S.C.H.; Formal analysis: Y.N.P., E.G., A.P.L., Z.J.L., S.C.H.; Investigation: Y.N.P., E.G., A.P.L., Z.J.L., S.C.H.; Writing - original draft: Y.N.P., S.C.H.; Writing - review & editing: Y.N.P., E.G., S.C.H.; Funding acquisition: S.C.H.
Funding
This work was supported by the National Science Foundation [IOS-1457061/IOS-1456809 to S.C.H.].
Data availability
All relevant data can be found within the article.
References
Competing interests
E.G. is founder and CEO of Oroboros Instruments (Innsbruck, Austria).