PARL-deficient mice appear clinically normal until around 6 weeks of age, after which they develop a progressive necrotizing encephalomyelopathy resembling Leigh syndrome with death before the age of 8 weeks (Spinazzi et al., 2019). As previously described, these mice exhibit severe testis atrophy (Figure 1A; Spinazzi et al., 2019; Cipolat et al., 2006). We did not observe cryptorchidism. Upon closer examination, the testis weight of Parl^-/- mice at 5 weeks of age, when they do not show clinical signs of neurological impairment, is found to be nearly half of that in matched WT littermates (Figure 1A). This difference cannot be explained by concurrent body weight reduction (Figure 1A). Microscopic analysis reveals that the seminiferous tubules from Parl^-/- mice are smaller in diameter (Figure 1B; p=0.0002, Figure 1—figure supplement 1A) and contain approximately 40% fewer cells compared to WT littermates (p=0.0009, Figure 1—figure supplement 1B). Further investigation indicates that the seminiferous tubules in Parl^-/- mice are populated by immature germ cells exhibiting degenerative changes and prominent intraluminal exfoliation, often in the form of multinucleated syncytia (Figure 1B). Immunohistochemistry reveals that PARL deficiency leads to a complete meiotic prophase I arrest as the seminiferous tubules are predominantly populated by SCP-1-positive spermatocytes (Figure 1B; Figure 1—figure supplement 1C; Yang and Wang, 2009; p=0.0001) while spermatids and spermatozoa are completely absent (Figure 1B; Köhler, 2007). The epididymal ducts of Parl^-/- mice are also smaller in diameter and completely devoid of mature sperm (Figure 1B). Spermatogonia show a modest increase in number in Parl^-/- seminiferous tubules compared to WT littermates (p=0.01; Figure 1—figure supplement 1D). Additionally, the analysis of γH2AX expression pattern in Parl^-/- seminiferous tubules indicates specific meiotic prophase I arrest at the pachytene stage (Figure 1—figure supplement 1G). Supporting Sertoli cells appear to be slightly increased in number (p=0.0382; Figure 1—figure supplement 1E and F), and the distribution and morphology of Leydig cells appears normal.

To determine whether the observed testicular abnormalities are linked to neurodegeneration, mice with conditional deletion of Parl in the nervous system (Parl ^L/L::Nes^Cre) were studied. Surprisingly, despite developing severe Leigh-like encephalopathy, these mice exhibit normal testicular size, histology, and sperm production comparable to WT littermates (Figure 1B), demonstrating that the testicular disorder is not a consequence of neurodegeneration. As previously reported (Anand-Ivell et al., 2017; Jiang et al., 2014; Davidoff et al., 2004), Nes is also expressed in Leydig cells (Figure 1—figure supplement 2). Although PARL deficiency in situ could not be formally verified in the absence of specific PARL antibodies suitable for immunohistochemistry, Cre recombinase activation under the Nes promoter is predicted to effectively delete Parl in these cells as in the nervous system. Moreover, extensive morphological observations detailed in the following paragraph indicate that Leydig cells are structurally unaffected in the germline Parl^-/- testis (Figure 2—figure supplement 1B), suggesting that the spermatogenetic defect is not secondary to PARL deficiency in these cells.

Altogether, deficiency of PARL leads to a complete arrest of spermatogenesis at the level of primary spermatocytes, independent of the effects of PARL in the nervous system and in Leydig cells.

To gain insight into the possible pathological effects of PARL deficiency on germ cells, we conducted a detailed morphological analysis using semithin sections and electron microscopy.

In unaffected WT animals, germ cells undergo a maturation process, with less differentiated forms (spermatogonia and spermatocytes) in the abluminal layers, more differentiated spermatids in the adluminal compartment, and mature spermatozoa in the lumen of the seminiferous tubules (Figure 2A). Conversely, Parl^-/- mice exhibit severe vacuolar degeneration of arrested spermatocytes, leading to cell death, and this degeneration progressively worsens from the abluminal to the adluminal compartment (Figure 2A and B and Figure 2—figure supplement 1A). Analysis of spermatocyte ultrastructure showed a significant increase in the occurrence of degeneration/death in Parl^-/- spermatocytes compared to WT (18.9% degenerated spermatocytes out of 201 analyzed in Parl^-/- vs. 0% out of 79 WT spermatocytes analyzed; n = 3 for each genotype; p=0.0002 by two-sided Fisher’s exact test). Next, we assessed whether mitochondrial morphology was affected in PARL-deficient spermatocytes. Differentiation per se leads to important morphological adaptations of mitochondria that parallel increasing bioenergetic demands requiring a shift from more glycolytic to more oxidative metabolism (Varuzhanyan and Chan, 2020). To ensure accurate comparisons, we focused on primary spermatocytes showing fully assembled synaptonemal complexes, a characteristic feature during the zygotene and pachytene stages of meiotic prophase I (Figure 2C; Yang and Wang, 2009; Martins and Silva, 2001). Compared to the mitochondria of WT primary spermatocytes, which are typically small with dilated cristae and dense finely granular matrix, mitochondria of Parl^-/- spermatocytes appear consistently swollen with few thin irregular cristae and loss of normal matrix density (Figure 2C). Quantitative analysis of the mitochondrial ultrastructure in primary spermatocytes showed a dramatic increase of degenerating mitochondria in Parl^-/- compared to WT spermatocytes (92% of analyzed mitochondria in Parl^-/- were abnormal vs. 1.9% in WT; n = 3 for each phenotype; p=0.0002 by two-sided Fisher’s exact test). Importantly, abnormal mitochondrial morphology was the earliest ultrastructural change detected in PARL-deficient spermatocytes localized in the abluminal compartment, while adluminal germ cells exhibited additional abnormalities affecting other cell compartments, including the endoplasmic reticulum, Golgi apparatus, and nuclear envelope. Chromatin clumping and nuclear fragmentation were also evident (Figure 2B and Figure 2—figure supplement 1A).

In contrast, other cell types within the seminiferous tubules and surrounding interstitium, such as spermatogonia, Leydig, and Sertoli cells, displayed normal ultrastructural features with preserved mitochondrial morphology (Figure 2—figure supplement 1B). Altogether, these data indicate the presence of early mitochondrial ultrastructural abnormalities culminating in extensive degeneration and death of arrested PARL-deficient spermatocytes, without morphological evidence of spermatogonia, Leydig cells, or Sertoli cells involvement.

Next, we asked to what extent the severe spermatogenesis defect induced by PARL deficiency can be attributed to the misprocessing and altered maturation of PARL’s substrates. To answer this question, we first tested the testicular expression of established PARL substrates. Parl^-/- testis mitochondria exhibit remarkable accumulation of uncleaved PINK1 and PGAM5, as well as almost total lack of the mature form of TTC19 (Figure 3A). These findings were consistent with previous observations in the brain (Spinazzi et al., 2019) and cultured cells (Saita et al., 2017). Since other PARL substrates, such as DIABLO, STARD7, and CLPB, displayed only subtle misprocessing or expression changes, possibly due to compensatory proteolytic cleavage by alternative proteases, we focused our investigation on PINK1, PGAM5, and TTC19. We aimed to determine whether the genetic modulation of these substrates could either modify or reproduce the testicular phenotype observed in Parl^-/- mice. In particular, we assessed whether accumulation of uncleaved PINK1 and PGAM5, alone or in combination, or depletion of the cleaved form of PINK1, PGAM5, or TTC19 were the molecular mechanisms underlying the abnormalities documented in Parl^-/- testis. PINK1 and PGAM5 are known to play essential roles in maintaining mitochondrial integrity and homeostasis and have been linked to both Parkinson’s disease and defects of spermatogenesis (Valente et al., 2004; Lu et al., 2014; Agarwal et al., 2020; Deng et al., 2008). Similarly, TTC19 is a mitochondrial protein crucial for the catalytic activity of complex III, and pathogenic variants of TTC19 are associated with mitochondrial diseases in humans, including Leigh syndrome (Atwal, 2014).

To test this hypothesis, we analyzed testes from a series of genetically engineered mutant mouse lines, including single-gene knockouts such as Pink1^-/-, Pgam5^-/-, and Ttc19^-/-, as well as multiple gene knockouts including both Parl and Pink1 (Parl^-/-/Pink1^-/-); Parl and Pgam5 (Parl^-/-/Pgam5^-/-); Pink1 and Pgam5 (Pink1^-/-/Pgam5^-/-); and Parl, Pink1, and Pgam5 combined (Parl^-/-/Pink1^-/-/Pgam5^-/-). Remarkably, the severe testis phenotype resulting from PARL deficiency remained unaltered upon additional deletion of Pink1 or Pgam5 either individually or in combination (Figure 3B and Figure 3—figure supplement 1). In contrast, the single or combined knockouts of Pink1, Pgam5, and Ttc19 resulted in normal fertility and testis morphology, showing orderly and complete spermatogenesis (Figure 3B and Figure 3—figure supplement 1). In conclusion, these observations indicate that impaired spermatogenesis in PARL-deficient mice is not driven by altered proteolytic maturation of the substrates PINK1, PGAM5, and TTC19 despite their severely affected proteolytic processing, indicating that other pathogenetic mechanisms are responsible for the testis phenotype.

Spermatogenesis involves crucial metabolic adaptations, with mitochondrial function playing a critical role throughout germ cell maturation (Varuzhanyan and Chan, 2020). Given the interconnection between mitochondrial morphology and function, we investigated the impact of the structural abnormalities identified in the mitochondria of Parl^-/- spermatocytes by conducting a comprehensive functional analysis. Because of the previously reported role of PARL in mitochondrial biogenesis (Civitarese et al., 2010), we wondered whether mitochondrial mass is reduced in Parl^-/- testis. Expression of the outer mitochondrial membrane protein TOMM20 and of the inner membrane ATP synthase beta subunit (ATPB) were similar between WT and Parl^-/- testis, suggesting unaltered mitochondrial mass (Figure 4A, Figure 5B, and Figure 6B). Similarly, mitochondrial DNA abundance, often used as an indicator of mitochondrial mass, was not significantly different between the two groups (Figure 4B). Additionally, the expression of TFAM, a protein binding mitochondrial DNA in nucleoids (Ekstrand et al., 2004), did not show any significant difference between WT and Parl^-/- SCP-1-positive spermatocytes (Figure 4—figure supplement 1). Next, we examined whether mitochondrial respiratory chain complexes were appropriately assembled in Parl^-/- testis mitochondria. Blue native gel electrophoresis revealed severe assembly alterations in multiple respiratory chain complexes, including complex I, complex III, complex IV, and to a lesser extent complex V, as well as the supercomplex (Pérez-Pérez et al., 2016; Figure 4C). Since respiratory chain complexes’ supramolecular assembly is required for optimizing the efficiency of mitochondrial oxidative phosphorylation (OXPHOS), we then examined if PARL deficiency ultimately resulted in impaired mitochondrial respiration in testis mitochondria. To answer this question, we measured oxygen consumption by means of high-resolution respirometry in testis mitochondria supplied with substrates and specific inhibitors for complex I (CI), complex II (CII), and complex IV (CIV) as illustrated in Figure 4D. Basal mitochondrial respiration in presence of complex I substrates but no ADP (CI LEAK) was significantly increased in Parl^-/- testis compared to WT, suggesting pathological short-circuit of protons across the inner mitochondrial membrane. Conversely, both phosphorylating respiration, whether driven by complex I only (CI OXPHOS) or by both complex I and II together (CI + II OXPHOS), and maximal uncoupled respiration, whether driven by complex II (CII ET) or by both complex I and II (CI + II ET) were severely diminished in Parl^-/- testis mitochondria. Respiration driven by CIV was also decreased. These results localize the severe respiration defect at the level of electron transfer capacity (Figure 4E). However, the defects were not attributed to cytochrome c loss due to outer mitochondrial membrane permeabilization (Figure 4E; CIV+cytc graph). To gain cell-type insights into the observed electron transport defect, cytochrome c-oxidase activity staining was performed on frozen tissue sections. The enzyme function was significantly decreased in PARL-deficient seminiferous tubules but not in Leydig cells, highlighting the specific distribution of the defect (Figure 4F). The expression of the subunit 4 of cytochrome c-oxidase, COX4, was indeed severely decreased in Parl^-/- SCP-1-positive spermatocytes, confirming the defect in this cell type (Figure 5A; p=0.0027). This defect was again unrelated to changes in mitochondrial mass since TOMM20 expression was unmodified by PARL deficiency in SCP-1 spermatocytes (Figure 5B). Interestingly,e dramatic overexpression of the glucose intracellular transporter GLUT1 was observed in Parl^-/- spermatocytes suggesting increased glucose utilization as an adaptive response to disrupted OXPHOS (Figure 5C). In conclusion, PARL is crucial for maintaining the integrity of the mitochondrial electron transport chain. Its deficiency leads to severe respiratory chain defects and metabolic remodeling in arrested primary spermatocytes.

CoQ is a lipid essential for cellular functions, serving both as an electron carrier in the mitochondrial respiratory chain and as a lipophilic antioxidant, preventing lipid peroxidation (Gueguen et al., 2021). In mammalian mitochondria, CoQ is involved in multiple converging pathways for its reduction, including complex I, complex II, dehydro-orotate dehydrogenase, sulfide-quinone oxidoreductase, and electron transfer dehydrogenase, while complex III is responsible for its oxidation. CoQ plays a critical role in promoting testicular functions including the maturation of male germ cells by safeguarding against oxidative damage (Lin et al., 2021; Mancini and Balercia, 2011).

In previous studies, we showed that brain mitochondria from PARL-deficient mice have decreased CoQ concentration linked to impaired expression of the ubiquinone biosynthesis protein COQ4 homolog, mitochondrial COQ4 (Spinazzi et al., 2019), a protein required for the biosynthesis of CoQ (Wang and Hekimi, 2019). Additionally, we observed an increase in the reduced-to-oxidized CoQ ratio (CoQ red/ox) in neurons due to TTC19 deficiency, leading to complex III dysfunction (Spinazzi et al., 2019). Similarly, we found significantly decreased CoQ levels in Parl^-/- testis, accompanied by a dramatic increase in the CoQ red/ox (Figure 6A). This elevation in CoQ reduction can be attributed to impaired CoQH₂ oxidation, resulting from compromised complex III activity caused by TTC19 depletion (Figure 3A) and complex III assembly defects (Figure 4C). Notably, we also noticed a substantial decrease in COQ4 levels in Parl^-/- testis, as seen in the brain (Spinazzi et al., 2019). Western blotting (Figure 6B) and immunohistochemistry (Figure 6C) revealed a diffuse decrease in COQ4 expression in various cell types, including germ cells, Leydig cells, and Sertoli cells. The deficit was particularly pronounced in Parl^-/- arrested spermatocytes, even those with no or minimal degenerative changes, suggesting that the CoQ biosynthesis defect occurred upstream of the degenerative process. Quantitative immunofluorescence of COQ4 expression confirms severe deficiency of this protein in Parl^-/- SCP-1-positive spermatocytes compared to WT littermates (Figure 6—figure supplement 1). Collectively, our findings indicate that PARL plays a crucial role in maintaining CoQ biosynthesis and redox state.

To understand the specific mechanism responsible for the severe germ cell degeneration and death observed in PARL-deficient mice, we first considered apoptosis due to the characteristic ultrastructural features observed in arrested spermatocytes (i.e., chromatin clumping and nuclear fragmentation) and previous links of PARL to antiapoptotic properties in vitro (Cipolat et al., 2006). However, levels of caspase-3 activation in the seminiferous tubules of Parl^-/- mice were comparable to WT, suggesting that apoptosis was not significantly involved in this phenotype (Figure 7—figure supplement 1). Given the identification of decreased CoQ concentration and severe ultrastructural abnormalities involving mitochondria and other membranous cell compartments, we speculated about the possible role of ferroptosis. Ferroptosis is a programmed cell death modality characterized by lipid peroxidation of cell membranes (Stockwell et al., 2017; Santoro, 2020). Previous studies in cultured cells have shown the importance of CoQ producing mevalonate pathway (Shimada et al., 2016) and CoQ reducing pathways driven by FSP1 (Bersuker et al., 2019; Doll et al., 2019), DHODH (Mao et al., 2021), and GCH1 (Kraft et al., 2020) in ferroptosis. To test this hypothesis, we examined the expression of GPX4, a crucial antioxidant peroxidase that prevents ferroptosis by reducing phospholipid hydroperoxide in cell membranes using reduced glutathione as substrate (Chen et al., 2021; Seibt et al., 2019). Immunoblot analysis revealed a nearly complete absence of GPX4 expression in Parl^-/- testis (Figure 7A). Immunohistochemistry and immunofluorescence provided cell-type-specific insights, showing a dramatic decrease in GPX4 expression in Parl^-/- arrested spermatocytes (Figure 7C, top panels; Figure 8A, p=0.0013) but not in Leydig (Figure 7C, top panels, black arrowheads) or Sertoli cells (Figure 8—figure supplement 1; p=0.5313). The impact of PARL deficiency on GPX4 expression was not observed in other organs, indicating a specific effect on spermatocytes (Figure 7—figure supplement 2A). To rule out a possible effect of PARL proteolytic activity on GPX4 expression, we checked GPX4 expression in mouse embryonic fibroblasts with and without PARL expression, and knockouts rescued with proteolytically active or inactive PARL. The results do not show evidence of proteolytic misprocessing and do not indicate GPX4 as a direct substrate of PARL (Figure 7—figure supplement 2B). Further investigations demonstrated increased lipid peroxidation, as evidenced by significantly higher levels of 4-hydroxynonenal (HNE) adducts, the end-products of lipid peroxidation that defines ferroptosis, in Parl^-/- testis (Figure 7B, middle panel), but not in brain (Figure 7—figure supplement 2C). The accumulation of HNE adducts was particularly prominent in adluminal and exfoliated spermatocytes during the late stages of degeneration (Figure 7C, middle panels). We confirmed these data by quantitative immunofluorescence showing a dramatic increase in HNE signal in SCP-1 positive Parl^-/- spermatocytes (Figure 8B; p=0.0002), which is consistent with the specific loss of GPX4 expression in these cells.

Additional established biomarkers of ferroptosis, including cellular tumor antigen p53, a master regulator of both canonical and non-canonical ferroptosis pathways (Jiang et al., 2015; Liu and Gu, 2022), and transferrin receptor protein 1 (TfR1), which promotes the cellular uptake of iron via receptor-mediated endocytosis (Feng et al., 2020), were also investigated. Excessive intracellular iron can contribute to ferroptosis by triggering lipid peroxidation through Fenton's reaction. In normal postpubertal mice, expression of p53 levels in testis is very low (Beumer et al., 1998) while TfR1 is very high in spermatogonia and gradually decreases during germ cell maturation (Leichtmann-Bardoogo et al., 2012; Gao et al., 2021; Figure 7C, bottom panel). In contrast, PARL-deficient testis showed prominent nuclear expression of p53 in adluminal degenerating spermatocytes (Figure 7—figure supplement 3), while TfR1 exhibited persistent overexpression in arrested spermatocytes (Figure 7C, bottom panel), suggesting abnormally high iron uptake. We confirmed these data by quantitative immunofluorescence showing increased TfR1 expression in Parl^-/- SCP-1-positive spermatocytes compared to WT littermates (Figure 8—figure supplement 2; p=0.0229). These findings collectively indicate that ferroptosis is a cell-type-specific effect of PARL deficiency and the mechanism underlying the demise of Parl^-/- arrested spermatocytes.

Mice with full knockout germline deletion of Parl (Parl^-/-) (MGI:3693645), Pgam5 (Pgam5^-/-) (MGI:5882561), Pink1 (Pink1^-/-) (MGI:5436308), Ttc19 (Ttc19^-/-) (MGI:6276545), and conditional Parl ablation under the Nestin promoter (Parl ^L/L::Nes^Cre) (MGI:3526574, MGI:2176173) have been generated as previously described (Spinazzi et al., 2019; Cipolat et al., 2006). All mutant mouse lines were maintained on a C57BL/6J background. Mice were kept in a SPF facility and multiply housed in filter top polycarbonated cages enriched with wood‐wool and shavings as bedding. Standard rodent diet and acidified tap water were provided ad libitum. Animal rooms were maintained at 22°C ± 2°C with a 45 and 70% relative humidity range, 50 air changes per hour, and 12-hr light/dark cycles. Mice were included in a health-monitoring program developed in accordance with the guidelines of the Federation of European Laboratory Animal Science Associations (FELASA). All experiments were approved by the Ethical Committee on Animal Experimenting of the University of Leuven (IACUC protocol #072/2015) and the French Ministry (DUO-OGM 5769 3/2019).

Testes harvested from postpubertal mutant mice and WT matched controls were immersion-fixed in 10% neutral buffered formalin for 24–48 hr at room temperature (RT). Samples were then routinely processed for paraffin embedding, sectioned at 5 µm, and stained with hematoxylin and eosin (HE) for histopathological assessment. For immunohistochemistry (IHC), 5-µm-thick paraffin sections were mounted on ProbeOn slides (Thermo Fisher Scientific #15-188-51). Chromogenic immunohistochemistry (IHC) and multiplex immunofluorescence (IF) were performed as described elsewhere (Tarrant et al., 2021) using a Leica BOND RXm automated platform combined with the Bond Polymer Refine Detection kit (Leica #DS9800) for IHC or the OPAL Automation Multiplex IHC Detection Kit (Akoya Biosciences NEL830001KT) implemented onto a Leica BOND Research Detection System (DS9455) for IF. Briefly, after dewaxing and rehydration, sections were pretreated with the epitope retrieval BOND ER2 high pH buffer (Leica #AR9640) for 20 min at 98°C. Endogenous peroxidase was inactivated with 3% H₂O₂ for 10 min at RT. Nonspecific tissue–antibody interactions were blocked by incubating the sections for 30 min at RT with Leica PowerVision IHC/ISH Super Blocking solution (PV6122) for IHC or with the Akoya Biosciences Opal Antibody Diluent/Block solution (ARD1001EA) for IF. The same blocking solution also served as diluent for the primary antibodies. Primary antibodies were incubated on the sections for 45 min at RT. A biotin-free polymeric detection system consisting of HRP conjugated anti-rabbit IgG was then applied for 25 min at RT. For IHC, immunoreactivity was then revealed with the diaminobenzidine (DAB) chromogen reaction. Tissue sections were finally counterstained in hematoxylin, dehydrated in an ethanol series, cleared in xylene, and permanently mounted with a resinous mounting medium (Thermo Scientific ClearVue coverslipper). For IF, the sections were finally incubated with the Akoya Biosciences TSA reagents Opal 520 (OP-1001), 570 (OP-1002), and 690 (OP-1003) (working concentration 1/150) for 10 min at RTs followed by Spectral DAPI nuclear counterstain (Akoya Biosciences FP1490) and mounting with Fluoromount-G (SouthernBiotech 100-01). Negative controls were obtained by replacement of the primary antibodies with irrelevant isotype-matched rabbit antibodies. HE and IHC-stained slides were evaluated by two board-certified veterinary pathologists (ER and CAA) with extensive expertise in mouse pathology. Staging of the seminiferous tubules was performed according to well-established morphological criteria (Ahmed and de Rooij, 2009; Meistrich and Hess, 2013). The Aperio Versa 200 instrument was used for image acquisition. Digital image analysis for cell count and morphometry of seminiferous tubules as well as for normalized quantification of marker expression within the SCP-1-positive spermatocyte population was performed using FIJI/ImageJ open-source software (Schroeder et al., 2021; Arena et al., 2017; Schindelin et al., 2012). Values for the normalized quantification correspond to the average positive area per spermatocyte and are expressed in um (Aitken et al., 2022).

Total testis lysates were prepared by homogenization with a glass-to-glass potter homogenizer on ice in 20 mM HEPES, 100 NaCl, pH 7.4, supplemented with protease and phosphate inhibitors (ROCHE). The lysate was then transferred to a fresh tube, supplemented with Triton- X 1%, SDS 0.1%, and passed several times through a 26-gauge syringe. The samples were then centrifuged at 20,000 × g for 15 min at 4°C to remove insoluble material. Tissue extracts or enriched mitochondrial membranes were separated in reducing and denaturing conditions in NuPage gels (Invitrogen). Proteins were transferred to PVDF 0.45 µm membranes, blocked with milk 5% TRIS-buffered saline, Tween-20 0.1% (TTBS), and incubated with the indicated primary antibodies, washed in TTBS incubated for 1 hr at RT with horseradish peroxidase conjugated secondary antibodies in 5% milk-TTBS or Alexa Fluor conjugated secondary antibodies. Proteins were identified by chemiluminescence or by fluorescence according to the type of secondary antibody. A PARL carboxy-terminal antibody was generated in house as previously reported (Cipolat et al., 2006).

To prepare testis-enriched mitochondrial fractions for western blotting or blue native gel electrophoresis, freshly collected testis was homogenized with a motor-driven Teflon pestle set at 800 rpm in a glass potter containing ice-cold 20 mM HEPES, 225 mM sucrose, 75 mM mannitol, 1 mM EGTA pH 7.4, on ice. For mitochondrial respiration experiments, fresh testis was homogenized manually with a Teflon pestle in ice-cold 20 mM HEPES, 225 mM sucrose, 75 mM mannitol, 1 mM EGTA pH 7.4, on ice, then gently passed through a 22-gauge syringe. The homogenate was centrifuged at 700 × g for 10 min at 4°C to remove nuclei and unbroken debris. The supernatant (tissue homogenate) was then centrifuged at 10,000 × g for 10 min at 4°C to pellet mitochondrial enriched mitochondrial membranes. To prepare liver enriched mitochondrial fractions, freshly collected liver was thoroughly rinsed in homogenization buffer, then homogenized with a motor-driven Teflon pestle set at 800 rpm in a glass potter containing ice-cold 20 mM HEPES, 225 mM sucrose, 75 mM mannitol, 1 mM EGTA pH 7.4, on ice. The homogenate was centrifuged at 1000 × g for 10 min at 4°C to remove nuclei and unbroken debris. The supernatant (tissue homogenate) was then centrifuged at 6000 × g for 10 min at 4°C. Brain mitochondria were purified according to Sims’ method (Sims and Anderson, 2008).

Blue native gel electrophoresis of digitonin-solubilized mitochondria was performed as described (Jha et al., 2016). Then, 100 µg isolated mitochondria were solubilized with 600 µg digitonin in Invitrogen Native Page sample buffer on ice for 20 min, then centrifuged at 20,000 × g for 20 min at 4°C. 0.75% Coomassie G-250 was added to supernatants, which were loaded on a 3–12% gradient Invitrogen Native Page gel according to the instructions. After electrophoresis, mitochondrial complexes and super complexes were visualized by protein staining with InstantBlue Coomassie Protein Stain (ISB1L) (Abcam ab119211).

Mitochondrial respiration in testis mitochondria respiration was measured in Miro6 Buffer (Fasching et al., 2016) (20 mM HEPES, 110 mM sucrose, 10 mM KH₂PO₄, 20 mM taurine, 60 mM lactobionic acid, 3 mM MgCl₂, 0.5 EGTA, pH 7.1, 1 mg/ml fatty acid-free BSA, catalase 280 U/ml) at 37°C as previously described (Pesta and Gnaiger, 2012; Spinazzi et al., 2019). When needed H₂O₂ was added to reoxygenate the chambers by catalase mediated O₂ generation. Then, 150 µg of mitochondrial-enriched membranes were loaded into the Oroboros 2K oxygraph. A typical experiment is illustrated in Figure 4D. Oxygen consumption rates were measured before and after addition of the following sequence of substrates and specific inhibitors: (1) 2.5 mM pyruvate, 10 mM glutamate, and 1 mM malate to measure complex I-driven leak respiration (CI leak); (2) 2.5 mM ADP to determine complex I-driven phosphorylating respiration (CI OXPHOS). (3) 5 mM succinate to determine the phosphorylating respiration driven by simultaneous activation of complex I and II (CI + II OXPHOS); (4) titrating concentrations of the mitochondrial uncoupler CCCP to reach the maximal uncoupled respiration (CI + II electron transfer capacity, ET); (5) 200 nM rotenone to fully inhibit complex I-driven respiration and measure complex II-driven uncoupled respiration (CII electron transfer capacity, CII ET); (6) 0.5 µM antimycin A to block mitochondrial respiration at the level of complex III. Residual oxygen consumption was always negligible. (7) 2 mM ascorbate, 0.5 mM TMPD to measure cytochrome c oxidase (CIV)-driven respiration; (8) 125 µg/ml cytochrome c to evaluate mitochondrial outer membrane integrity and (9) 500 µM potassium cyanide (KCN) to specifically block cytochrome c oxidase activity and measure residual background oxygen consumption caused by chemical reaction between ascorbate and TMPD. Cytochrome c oxidase-driven respiration was calculated as the cyanide-sensitive oxygen consumption.

CoQ content and the ratio of the reduced vs. oxidized forms were measured as previously described (Rodríguez-Aguilera et al., 2017).

For mtDNA quantification, total DNA was isolated from 20 to 30 mg of testis tissues by using a DNeasy Blood and tissues kit (QIAGEN). qPCRs were performed in triplicate in 96-well reaction plates (Applied Biosystems). Each reaction (final volume 10 µl) contained 25 ng DNA, 5 µl of Power SYBR-Green PCR Master Mix (Applied Biosystems), and 0.5 µM of each forward and reverse primer. COX1, mitochondrial encoded gene, was amplified and β2 microglobulin (β2 m), nuclear encoded gene, was used as a normalizing control. Fold changes in mtDNA amount were calculated with the ΔΔCt method. The employed primers sequences were Cox1-Mus-F: TTTTCAGGCTTCACCCTAGATGA, Cox1-Mus-R: CCTACGAATATGATGGCGAAGTG, B2m-Mus-F: ATGGGAAGCCGAACATACTG, B2M-Mus-R:CAGTCTCAGTGGGGGTGAAT.

Testes of the indicated genotype were collected and immediately fixed with 2.5% glutaraldehyde, 2% paraformaldehyde in 0.1 M cacodylate buffer pH 7.4. Tissue was stored overnight at 4°C in the fixative solution, washed in 0.1 M cacodylate buffer, and post-fixed for 2 hr at RT with 1% OsO₄, 1.5% K₄Fe(CN)₆ in 0.1 M cacodylate buffer. Sections were rinsed, stained with 3% uranyl acetate for 1 hr at 4°C, and dehydrated in graded ethanol concentrations and propyleneoxide, followed by embedding in Epon Resin. Resin blocks were sectioned on a ultramicrotome. Post-staining was performed with 3% uranyl acetate followed by lead citrate staining. Semithin sections were collected on slides and stained with 1% Toluidine blue solution (Sigma-Aldrich). Ultrathin sections (60 nm) were mounted on copper grids and imaged using a JEOL transmission electron microscope.

Immortalized mouse embryonic fibroblasts (MEFs) derived from WT and Parl^-/- male mice were cultured in Dulbecco’s modified Eagle’s medium/F-12 (Gibco) containing 10% fetal bovine serum (Gibco). At 30–40% confluence, the MEFs were transduced using a replication-defective recombinant retroviral expression system (Clontech) with either wild-type (Parl WT) or catalytic inactive Parl S275A as previously described (Spinazzi et al., 2019). Cell lines stably expressing the desired proteins were selected based on their acquired resistance to 5 µg/ml puromycin. Cells were regularly tested to rule out Mycoplasma contamination.

Numerical data are expressed and illustrated in all graph bars as mean ± SD from biological replicates. No statistical tests were used to predetermine sample size. Replicates numbers were decided from experience of the techniques performed and practical considerations. Two-sided Student’s t-test was used to compare differences of all quantitative variables between two groups, and Fisher’s exact test was used for the analysis of contingency tables to compare the frequency distribution of ultrastructural abnormalities in two groups. Significance was calculated using GraphPad. Differences were considered statistically significant for p≤0.05. No data were excluded.