PINK1‐mediated mitophagy maintains pluripotency through optineurin

Abstract Objectives Dysfunction of autophagy results in accumulation of depolarized mitochondria and breakdown of self‐renewal and pluripotency in ESCs. However, the regulators that control how mitochondria are degraded by autophagy for pluripotency regulation remains largely unknown. This study aims to dissect the molecular mechanisms that regulate mitochondrial homeostasis for pluripotency regulation in mouse ESCs. Materials and methods Parkin+/+ and parkin −/− ESCs were established from E3.5 blastocysts of parkin+/− x parkin+/− mating mice. The pink1 −/−, optn −/− and ndp52 −/− ESCs were generated by CRISPR‐Cas9. shRNAs were used for function loss assay of target genes. Mito‐Keima, ROS and ATP detection were used to investigate the mitophagy and mitochondrial function. Western blot, Q‐PCR, AP staining and teratoma formation assay were performed to evaluate the PSC stemness. Results PINK1 or OPTN depletion impairs the degradation of dysfunctional mitochondria during reprogramming, and reduces the reprogramming efficiency and quality. In ESCs, PINK1 or OPTN deficiency leads to accumulation of dysfunctional mitochondria and compromised pluripotency. The defective mitochondrial homeostasis and pluripotency in pink1 −/− ESCs can be compensated by gain expression of phosphomimetic Ubiquitin (Ub‐S65D) together with WT or a constitutively active phosphomimetic OPTN mutant (S187D, S476D, S517D), rather than constitutively inactive OPTN (S187A, S476A, S517A) or a Ub‐binding dead OPTN mutant (D477N). Conclusions The mitophagy receptor OPTN guards ESC mitochondrial homeostasis and pluripotency by scavenging damaged mitochondria through TBK1‐activated OPTN binding of PINK1‐phosphorylated Ubiquitin.


| INTRODUC TI ON
Autophagy is a cellular degradation process that sequesters portions of cytoplasm into autophagosomes for degradation and recycling. [1][2][3] In contrast to proteasome-mediated degradation of small and shortlived proteins, large protein aggregates and damaged organelles are degraded by autophagy. 4 At first, autophagy was proposed to carry out bulk degradation of protein aggregates and organelles under stress conditions; however, more and more studies have provided evidence supporting the idea that autophagy is a selective process. 5 Selective removal of mitochondria by autophagy, named mitophagy, has been extensively demonstrated. Mitophagy receptors bridge the damaged mitochondria with MAP1LC3B (LC3-II) on autophagic membranes, leading to the cargo engulfment. 6 Currently, two types of mitophagy receptors have been identified. One type, like p62/ SQSTM1 and optineurin (OPTN), contain an ubiquitin-binding domain that localizes them to ubiquitin-tagged mitochondria. The other type, like BNIP3 and NIX, harbor a LC3-II-interacting region (LIR), and are located on mitochondria where they can be directly recognized by isolation membranes, which are precursors of autophagosomes. 4 Pluripotent stem cells (PSCs), including embryonic stem cells (ESCs) and induced pluripotent stem cells (iPSCs), can undergo unlimited self-renewal and give rise to any cells of the three germ layers; thus, they hold great promise for regenerative medicine. [7][8][9][10][11][12][13] PSCs have distinct cellular components and organelles compared with somatic cells. Recent studies have demonstrated that mitochondria are remodeled by autophagy during reprogramming of somatic cells. 14,15 In ESCs, dysfunction of autophagy leads to accumulation of damaged mitochondria which thereafter inhibits self-renewal and pluripotency. 1,14,[16][17][18] These studies indicate that mitochondria are strictly regulated by autophagy to maintain their homeostasis in PSCs.
However, the molecules which are directly responsible for recognition of mitochondria by autophagy in ESCs have not been identified.
In this study, we showed that OPTN is a mitophagy receptor that regulates mitochondrial homeostasis in ESCs. OPTN deficiency, like PINK1 deficiency, leads to abnormal self-renewal and reduced pluripotency of ESCs. PINK1 and OPTN are linked via PINK1-mediated phosphorylation of Ub and TBK1-mediated activation of the Ubbinding function of OPTN in a pathway that is independent of the E3 ligase PARKIN. Yamanaka's lab. Pink1, Pink1 mutants, 19,20 Parkin, Optn, and Optn mutants were cloned into pMXs, pCDH-CAG-PURO and pCDH-CAG-RFP lenti-vectors as described previously. ShRNAs were designed and cloned into pSicor-GFP and pSicor-Vector.

| ESC isolation and knockout ESC generation
Parkin +/+ and parkin −/− ESCs were isolated at day E3.5. The inner cell mass was picked and cultured in 2i medium. The earlypassage ESCs were cultured in 2i medium, and the later passages were maintained in ESC medium. Medium was used as described previously. 14 Pink1, Optn and Ndp52 knockout ESCs were generated by CRISPR-Cas9. We designed the gRNAs to target Pink1, Optn and Ndp52, and transfected the gRNA vector into ESCs using a Gene Pulser Xcell II (Bio-Rad) according to the manufacturer's protocols.
Knockout ESCs were identified by sequencing and western blotting.

| Somatic cell reprogramming
Around 50 000 MEFs per well were seeded into 6-well plate and infected with reprogramming vector cocktails (Pou5f1, Sox2, Klf4 and c-Myc) as previously described. 9 24 hours after infection, the medium were changed with fresh ESC medium and replaced every day. Colonies occurred approximately 12 days after reprogramming.

| Western blot analysis
Cell samples were lysed in RIPA buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0.5% sodium deoxycholate, 1% Nonidet P-40, 5 mM EGTA, 2 mM EDTA, 10 mM NaF) with protease inhibitor cocktail (Roche, 04693116001). Equal quantities of protein were loaded onto gels for SDS-PAGE, and transferred to nitrocellulose membranes (Millipore). The first antibody was used at the indicated concentration to incubate the membrane. Then the membrane was incubated with an appropriate HRP-conjugated secondary antibody (Santa Cruz). A Luminata Forte Western HRP Substrate Kit (Millipore, WBLUF0100) was used to detect the immunoreactive bands.

| Mito-Keima assay
The Mito-Keima constructs were employed to monitor mitophagosome formation as previously described. 21 The cells under normal conditions or treated with FCCP were detected by a confocal laser scanning microscopy for Mito-Keima imaging.  Interestingly, the defective somatic cell reprogramming in pink1 knockdown MEFs were rescued by gain expression of WT-but not kinase activity dead (A168P or G385A) mutant-PINK1, indicating the kinase activity of PINK1 is required for pluripotency acquisition ( Figure 1G; Figure S1A). In addition, while knockout of Parkin did not change reprogramming efficiency, knockdown of Pink1 expression in parkin −/− MEFs during reprogramming inhibited mitophagosome formation and significantly decreased reprogramming efficiency ( Figure 1H; Figure S1B). Together, these data suggest that PINK1-mediated mitophagy regulates mitochondrial remodeling and somatic cell reprogramming independently of

PARKIN.
Furthermore, we established iPSCs with decreased Pink1 expression ( Figure S2A). We found inhibition expression of Pink1 results in increased Mito-mass, decreased mitochondrial membrane potential, elevated intracellular reactive oxygen species (ROS) and deteriorated ATP generation in established iPSCs ( Figure S2B-E).
Correspondingly, the expression of pluripotency marker genes in Pink1 knockdown iPSCs was significantly decreased compared to wild type (WT) iPSCs ( Figure S6F). These data indicate inhibition of Pink1 not only compromises reprogramming efficiency but also deteriorates iPSC quality.

| PINK1 guards ESC identity
We next asked whether the PINK1/PARKIN pathway contributes to pluripotency regulation in ESCs. Parkin +/+ and parkin −/− ESCs were isolated from mouse blastocysts at embryonic d 3.5 and were shown to have a normal karyotype ( Figure S3A). Colony formation assays showed that the absence of Parkin did not affect ESC self-renewal ability ( Figure S3B). Furthermore, lack of Parkin did not change the expression of pluripotency genes, which indicates that PARKIN is dispensable for pluripotency maintenance of ESCs ( Figure S3C).

We then knocked out Pink1 in ESCs by CRISPR/Cas9
( Figure 2A; Figure S4A-C). We found that knockout of Pink1 significantly inhibited the self-renewal and pluripotency of ESCs F I G U R E 1 PINK1 but not PARKIN is essential for somatic cell reprogramming. A, Total genomic DNA extracted from Parkin +/+ and parkin −/− MEF cells was amplified by PCR to confirm the knockout mutation in Parkin. BLA, H 2 O. B, PARKIN is dispensable for mitochondrial removal during reprogramming of somatic cells (MEFs). Cells stained with Mito-tracker and anti-SSEA-1 antibody were detected by a FACS at reprogramming day 10. SSEA-1-positive cells were used to measure the Mito-mass. Data are shown as mean ± SD, n = 3; NS, not significant; Student's t test. C, PARKIN is not required for somatic cell reprogramming. Colonies were stained with alkaline phosphatase (AP) on day 12 of reprogramming. Data are shown as mean ± SD, n = 3; NS, not significant; Student's t test. D, Cells were transfected with scramble short RNA or small interfering RNA targeting Pink1, then harvested and subjected to RT-PCR analysis to detect Pink1 mRNA expression. E, Inhibition of Pink1 expression leads to mitochondrial accumulation during reprogramming. The Mito-mass was determined in SSEA-1-positive cells using Mito-RFP (which labels mitochondria with RFP) at reprogramming day 10. Data are shown as mean ± SD, n = 3; *P < .05; **P < .01; Student's t test. F, Knockdown of Pink1 decreases the reprogramming efficiency. iPSC colonies were stained by AP at reprogramming day 12. Data are shown as mean ± SD, n = 3; **P < .01; Student's t test. G, The decreased reprogramming efficiency by Pink1 knockdown was rescued by gain of expression of wild-type but not kinase dead mutant (A168P or G385A) Pink1. Colonies were stained with alkaline phosphatase (AP) on day 12 of reprogramming. Data are shown as mean ± SD, n = 3; *P < .05; **P < .01; NS, not significant; Student's t test. H, While knockout of Parkin did not affect reprogramming efficiency, knockdown Pink1 in both Parkin +/+ and parkin −/− MEF cells significantly reduced reprogramming efficiency. Colonies were stained with alkaline phosphatase (AP) on day 12 of reprogramming. Data are shown as mean ± SD, n = 3; **P < .01; NS, not significant; Student's t test Parkin Figure 2B, C). Reduced colony formation of pink1 −/− ESCs was not caused by abnormal cellular proliferation or apoptosis since Pink1 knockout did not affect proliferation and viability ( Figure S4D, E).
Furthermore, pink1 −/− ESCs showed significantly decreased expression of pluripotency genes, which suggests that knockout of Pink1 leads to the compromised pluripotency in ESCs ( Figure 2D).
In support of this idea, pink1 −/− ESCs showed abnormal embryonic body (EB) differentiation, characterized by delayed expression of certain endoderm marker genes and advanced expression of certain ectoderm genes ( Figure 2E). In addition, the teratoma formation assay was employed to test the contribution of Pink1 to ESC differentiation. While both Pink1 +/+ and pink1 −/− ESCs formed teratomas, the average weight of teratomas formed by Pink1 +/+ ESCs is significantly higher than those formed by pink1 −/− ESCs, which supports the notion that PINK1 is pivotal for differentiation of ESCs ( Figure 2F). Taken together, these data indicate that PINK1 is essential for ESC self-renewal, pluripotency and differentiation.

| PINK1 is essential for maintaining mitochondrial homeostasis in ESCs
Since PINK1 is a critical regulator of mitochondrial autophagy in somatic cells, we next asked whether lack of PINK1 affects the mitochondrial homeostasis in ESCs. The mitochondrial mass was determined by a FACS using Mito-tracker staining. We found that mitochondrial accumulation was higher in pink1 −/− ESCs than in Pink1 +/+ ESCs ( Figure 3A). Further investigation identified that de- Pink1 +/+ pink1 -/-1 pink1 -/-2 oxygen species (ROS) ( Figure 3B-D). These data indicate that Pink1 knockout leads to accumulation of damaged mitochondria in ESCs.
As a result, pink1 −/− ESCs generate less ATP than Pink1 +/+ ESCs ( Figure 3E). Taken together, these data provide evidences that PINK1 regulates mitochondrial homeostasis to maintain normal physiological function in ESCs. Optn knockout did not affect cellular proliferation and viability ( Figure S6D, E).

| Regulation of stemness by OPTN depends on TBK1 phosphorylation and ubiquitin binding activity
In human somatic cells, mitochondrial depolarization activates the kinase TBK1 to phosphorylate the autophagy receptor OPTN at Serine 177 (S177), Serine 473 (S473) and Serine 513 (S513); this promotes ubiquitin chain binding by OPTN, mitochondrial retention of OPTN and efficient mitophagy. 23,26 To dissect how OPTN is involved in mitochondrial homeostasis and pluripotency regulation in ESCs, a gain-of-function assay was performed by introducing  (Figure S9A, B). The results showed that reduced mitophagosomes, accumulation of abnormal mitochondria, decreased mitochondrial membrane potential, elevated ROS and compromised ATP generation were restored in optn −/− stable ESC lines carrying WT or △3D Optn but not in those with the empty vector, △3A or D477N Optn ( Figure 5A-E). These data support the idea that TBK1 phosphorylation and ubiquitin binding are critical for OPTN-mediated mitophagy in ESCs. Correspondingly, the defective colony formation ability was rescued in optn −/− stable ESC lines carrying wild-type and △3D Optn ( Figure 5F). Collectively, these data demonstrate that OPTN regulates ESC stemness in a manner that depends on TBK1 phosphorylation and ubiquitin-binding activity.

| PINK1 mediates mitochondrial homeostasis and stemness through OPTN
We next investigated whether PINK1 regulates mitochondrial homeostasis and pluripotency through OPTN. In addition to phosphorylating PARKIN at Serine 65 (S65) to activate its ubiquitin ligase is still controversial. An early study showed that inhibition of mitochondrial respiration promotes pluripotency, 33 while another report suggested that normal mitochondria control proliferation of ESCs but have no effect on ESC pluripotency. 34  kinase. 40 The activated TBK1 in human cells phosphorylates OPTN at S177, S473 and S513, thus promoting efficient mitophagy. 23,25,26 In contrast, by using Pink1/Parkin knockout and mutation of the TBK1 phosphorylation sites on OPTN (S187, S476 and S517) in mouse ESCs, we demonstrated that phosphorylation of OPTN by TBK1 is required for mitophagy, which is independent of PARKIN in ESCs (Figure 7). PINK1 mediated mitochondrial degradation is currently defined to depend on recruitment of ubiquitin to damaged mitochondria by the E3 ligases like PARKIN. 5 In this regard, the fact that PARKIN is not required for mitochondrial homeostasis and pluripotency regulation indicates the existing undefined E3 ligases in ESCs to build the ubiquitin chains to link depolarized mitochondria with OPTN ( Figure 7).
This area needs further investigation.
In conclusion, we identified that PINK1/OPTN-mediated PARKIN-independent mitophagy serves as a scavenger to de-

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.