AUTOPHAGY 2016, VOL. 12, NO. 9, 1675–1676 http://dx.doi.org/10.1080/15548627.2016.1193656
AUTOPHAGIC PUNCTUM
FUNDC1 is a novel mitochondrial-associated-membrane (MAM) protein required for hypoxia-induced mitochondrial fission and mitophagy Wenxian Wua, Wen Lia, Hao Chena, Lei Jiangb, Runzhi Zhua, and Du Fenga,c a
Guangdong Key Laboratory of Age-related Cardiac-cerebral Vascular Disease, Institute of Neurology, Department of Neurology, Affiliated Hospital of Guangdong Medical University, Guangdong Medical University, Zhanjiang, Guangdong China; bDepartment of Cardiology, Guangdong Cardiovascular Institute, Guangdong Provincial Key Laboratory of Coronary Disease, Guangdong General Hospital, Guangdong Academy of Medical Sciences, Southern Medical University, Guangzhou China; cDepartment of Developmental Biology, Harvard School of Dental Medicine, Harvard Medical School, Boston, MA USA
ABSTRACT
ARTICLE HISTORY
Mitochondria need to be fragmented prior to engulfment by phagophores, the precursors to autophagosomes. However, how these 2 processes are finely regulated and integrated is poorly understood. We have shown that the outer mitochondrial membrane protein FUNDC1 is a novel mitochondrial-associated membrane (MAM) protein, enriched at the MAM by interacting with the ER resident protein CANX (calnexin) under hypoxia. As mitophagy proceeds, it dissociates from CANX and preferably recruits DNM1L/DRP1 to drive mitochondrial fission in response to hypoxic stress. In addition, knocking down of FUNDC1, DNM1L or CANX in hypoxic cells increases the number of elongated mitochondria and also reduces the colocalization of autophagosome and mitochondria, thus preventing mitophagy. These findings identify FUNDC1 as a molecular hub integrating mitochondrial fission and mitophagy at the MAM in response to hypoxia.
Received 13 May 2016 Revised 16 May 2016 Accepted 17 May 2016
Stresses such as hypoxia or membrane potential uncouplers, etc. will send at least 2 signals to cells; one is to trigger mitochondrial fission, and the other is to initiate autophagosome formation to finely control the quality and quantity of a healthy mitochondrial pool. The MAM has been recently shown to be not only the site for selective autophagosome (mitophagosome) formation but also the spot for mitochondrial fission. In light of this, we speculated that there should exist some key molecules at the MAM controlling both processes. In search of key proteins at the MAM, we first needed to get highly purified MAM fractions from hypoxic cells. Twenty (20 cm) culture dishes were used to homogenize cells and the lysates were subjected to Percoll density-gradient centrifugation for enriching highly purified MAM; the purity was assessed by different organelle and MAM markers. Proteomics and mass spectrometry were performed to identify the proteins in the MAM fraction. The experiments were repeated 3 times and only the overlapping proteins were counted. Finally, we obtained 45 highly abundant proteins and then eliminated nonmitochondrially localized proteins, leaving »10 proteins for further analysis. Similar to the previous reports, ACSL4/ FACL4, MFN2, and VDAC, etc. are enriched in this fraction but we also found some new MAM proteins (manuscript in preparation). Interestingly, the newly identified mitophagy receptor FUNDC1 is in this fraction. We then tested the fission and mitophagy induction roles of the remaining 10 proteins. Among them, only FUNDC1 shows the ability to both promote
calnexin; DRP1; ER; FUNDC1; mitochondrial-associated membranes; mitophagy
mitochondrial fragmentation and LC3 puncta formation. So, we focused our study on FUNDC1. To prove that FUNDC1 is indeed a MAM protein, we compared its subcellular localization either under normoxia or hypoxia conditions. FUNDC1 substantially accumulates at the MAM in response to hypoxia, although a small amount of FUNDC1 can be also found in the ER-mitochondria contact sites under normoxic conditions. Interestingly, the protein level of DNM1L at the MAM is also prominently increased under hypoxia and colocalizes with FUNDC1 at these contact interfaces. Immunogold labeling confirmed that FUNDC1 indeed accumulates at the MAM together with the MAM marker protein ACSL4. The latter is enriched at this site in response to hypoxia. Notably, the number of MAMs increases under hypoxia conditions, implying that MAM may have a functional role during mitochondrial fission. The reason why FUNDC1 accumulates at MAM is still not very clear. We hypothesized that some protein might pull FUNDC1 to this site. Thus, we tested whether FUNDC1 could be immunoprecipitated by different reported MAM proteins, including STX17, ACSL4, ITPR/IP3Rs and CANX, etc., and found that the N terminus of CANX associates with the hydrophilic domain of FUNDC1 (96–139) at the MAM. The binding is indirect because the N terminus of CANX is in the lumen of the ER. It is unlikely that the hydrophilic domain of FUNDC1 (96–139) can bind to the lumenal part of CANX. Accordingly, there should be intermediate protein(s) between FUNDC1 and CANX, which need to be further clarified. Depletion of CANX
57#, People’s road, Zhanjiang, China, 524001; Runzhi Zhu CONTACT Du Feng [email protected] No. 96 Dongchuan Road, Guangzhou, China, 510000 jiang, China, 524001; Lei Jiang [email protected] Color versions of one or more of the figures in the article can be found online at www.tandfonline.com/kaup. © 2016 Taylor & Francis
KEYWORDS
[email protected]
57#, People’s road, Zhan-
1676
W. WU ET AL.
Figure 1. The mitophagy receptor FUNDC1 accumulates at the MAM in the early stage of hypoxia, which might be driven by association with the ER-resident protein CANX through binding to an unknown intermediate protein by an unclear mechanism. As mitophagy proceeds, the binding between CANX and FUNDC1 weakens and much of the cytosolic DNM1L is recruited by FUNDC1 for the final scission of mitochondria.
in hypoxic cells prevents the accumulation of FUNDC1 at the MAM, suggesting that this resident ER protein might drive the accumulation of FUNDC1. Several pieces of data suggest that hypoxia induces fission of mitochondria but the mechanism of action remains unclear. MFF, FIS1, or MIEF1/MID51-MIEF2/MID49 is involved in recruiting DNM1L to mediate mitochondrial scission in many physiological or stress conditions. But whether these proteins are involved in mitochondrial fission and DNM1L recruitment in response to hypoxia is less studied. Silencing of MFF, FIS1, or MIEF1-MIEF2 does not prevent mitochondrial translocation of DNM1L and subsequent mitochondrial fragmentation in hypoxic cells, although the effect of MFF depletion is less severe. However, in FUNDC1-depleted cells, the number of translocated DNM1L is substantially reduced and the mitochondria are dramatically elongated under hypoxia treatment. Interestingly, when FUNDC1 was overexpressed, most DNM1L colocalized with the fragmented mitochondria. In addition, a double-immunogold labeling experiment indicates they both accumulate at the MAM. Based on the above strong phenotype, we tested if they have a physical interaction. Immunoprecipitation and GST affinity isolation assays both demonstrate that they have direct binding. Similar to CANX, DNM1L also interacts with the cytosolic loop of FUNDC1 (96–139). Furthermore, the truncated FUNDC1 that lacks binding to DNM1L cannot induce mitochondrial fission. Notably, FUNDC1 binds to CANX at a very early stage of hypoxia and then dissociates from CANX and preferentially interacts with DNM1L instead when the length of time under hypoxic conditions is prolonged. This is possible because both proteins bind to the same region of FUNDC1 and they may compete in
different situations by mechanisms yet to be determined. An enhanced binding between LC3 and FUNDC1 at later stages of hypoxia is observed, and FUNDC1 (AA 1–50) is essential for its binding with LC3. In addition to the effects of FUNDC1, DNM1L, CANX, MFF or FIS1 in mitochondrial fission, we also compared their role in mitophagy. Silencing of FUNDC1, CANX or DNM1L causes mitochondrial elongation and reduces mitophagy in response to hypoxia. In contrast, mitochondrial fragmentation and mitophagy frequently occur in hypoxic MFF or FIS1 knockdown cells, or control cells treated with scramble siRNA. Together, we concluded that the mitophagy receptor FUNDC1 is not only a novel MAM protein, but also a new mitochondrial receptor for DNM1L to facilitate mitochondrial fragmentation and mitophagy in response to hypoxia (Fig. 1).
Disclosure of potential conflicts of interest No potential conflicts of interest were disclosed.
Funding This work is supported by the NSFC (No. 31401182), by the National Basic Research Program of China (2013CB910100), by the Natural Science Foundation of Guangdong Province, China (2014A030313533), by the Science and Technology Planning Project of Guangdong Province (2016A020215152), by the Medical Scientific Research Foundation of Guangdong Province, China (No. A2014484), by the Foundation for Distinguished Young Talents in Higher Education of Guangdong, China (2013LYM_0035), by the Sail Plan of Talents Recruitment Grant, Guangdong, China (Yue Ren Cai Ban [2014] 1) to Dr. Du Feng.
AUTOPHAGIC PUNCTUM
FUNDC1 is a novel mitochondrial-associated-membrane (MAM) protein required for hypoxia-induced mitochondrial fission and mitophagy Wenxian Wua, Wen Lia, Hao Chena, Lei Jiangb, Runzhi Zhua, and Du Fenga,c a
Guangdong Key Laboratory of Age-related Cardiac-cerebral Vascular Disease, Institute of Neurology, Department of Neurology, Affiliated Hospital of Guangdong Medical University, Guangdong Medical University, Zhanjiang, Guangdong China; bDepartment of Cardiology, Guangdong Cardiovascular Institute, Guangdong Provincial Key Laboratory of Coronary Disease, Guangdong General Hospital, Guangdong Academy of Medical Sciences, Southern Medical University, Guangzhou China; cDepartment of Developmental Biology, Harvard School of Dental Medicine, Harvard Medical School, Boston, MA USA
ABSTRACT
ARTICLE HISTORY
Mitochondria need to be fragmented prior to engulfment by phagophores, the precursors to autophagosomes. However, how these 2 processes are finely regulated and integrated is poorly understood. We have shown that the outer mitochondrial membrane protein FUNDC1 is a novel mitochondrial-associated membrane (MAM) protein, enriched at the MAM by interacting with the ER resident protein CANX (calnexin) under hypoxia. As mitophagy proceeds, it dissociates from CANX and preferably recruits DNM1L/DRP1 to drive mitochondrial fission in response to hypoxic stress. In addition, knocking down of FUNDC1, DNM1L or CANX in hypoxic cells increases the number of elongated mitochondria and also reduces the colocalization of autophagosome and mitochondria, thus preventing mitophagy. These findings identify FUNDC1 as a molecular hub integrating mitochondrial fission and mitophagy at the MAM in response to hypoxia.
Received 13 May 2016 Revised 16 May 2016 Accepted 17 May 2016
Stresses such as hypoxia or membrane potential uncouplers, etc. will send at least 2 signals to cells; one is to trigger mitochondrial fission, and the other is to initiate autophagosome formation to finely control the quality and quantity of a healthy mitochondrial pool. The MAM has been recently shown to be not only the site for selective autophagosome (mitophagosome) formation but also the spot for mitochondrial fission. In light of this, we speculated that there should exist some key molecules at the MAM controlling both processes. In search of key proteins at the MAM, we first needed to get highly purified MAM fractions from hypoxic cells. Twenty (20 cm) culture dishes were used to homogenize cells and the lysates were subjected to Percoll density-gradient centrifugation for enriching highly purified MAM; the purity was assessed by different organelle and MAM markers. Proteomics and mass spectrometry were performed to identify the proteins in the MAM fraction. The experiments were repeated 3 times and only the overlapping proteins were counted. Finally, we obtained 45 highly abundant proteins and then eliminated nonmitochondrially localized proteins, leaving »10 proteins for further analysis. Similar to the previous reports, ACSL4/ FACL4, MFN2, and VDAC, etc. are enriched in this fraction but we also found some new MAM proteins (manuscript in preparation). Interestingly, the newly identified mitophagy receptor FUNDC1 is in this fraction. We then tested the fission and mitophagy induction roles of the remaining 10 proteins. Among them, only FUNDC1 shows the ability to both promote
calnexin; DRP1; ER; FUNDC1; mitochondrial-associated membranes; mitophagy
mitochondrial fragmentation and LC3 puncta formation. So, we focused our study on FUNDC1. To prove that FUNDC1 is indeed a MAM protein, we compared its subcellular localization either under normoxia or hypoxia conditions. FUNDC1 substantially accumulates at the MAM in response to hypoxia, although a small amount of FUNDC1 can be also found in the ER-mitochondria contact sites under normoxic conditions. Interestingly, the protein level of DNM1L at the MAM is also prominently increased under hypoxia and colocalizes with FUNDC1 at these contact interfaces. Immunogold labeling confirmed that FUNDC1 indeed accumulates at the MAM together with the MAM marker protein ACSL4. The latter is enriched at this site in response to hypoxia. Notably, the number of MAMs increases under hypoxia conditions, implying that MAM may have a functional role during mitochondrial fission. The reason why FUNDC1 accumulates at MAM is still not very clear. We hypothesized that some protein might pull FUNDC1 to this site. Thus, we tested whether FUNDC1 could be immunoprecipitated by different reported MAM proteins, including STX17, ACSL4, ITPR/IP3Rs and CANX, etc., and found that the N terminus of CANX associates with the hydrophilic domain of FUNDC1 (96–139) at the MAM. The binding is indirect because the N terminus of CANX is in the lumen of the ER. It is unlikely that the hydrophilic domain of FUNDC1 (96–139) can bind to the lumenal part of CANX. Accordingly, there should be intermediate protein(s) between FUNDC1 and CANX, which need to be further clarified. Depletion of CANX
57#, People’s road, Zhanjiang, China, 524001; Runzhi Zhu CONTACT Du Feng [email protected] No. 96 Dongchuan Road, Guangzhou, China, 510000 jiang, China, 524001; Lei Jiang [email protected] Color versions of one or more of the figures in the article can be found online at www.tandfonline.com/kaup. © 2016 Taylor & Francis
KEYWORDS
[email protected]
57#, People’s road, Zhan-
1676
W. WU ET AL.
Figure 1. The mitophagy receptor FUNDC1 accumulates at the MAM in the early stage of hypoxia, which might be driven by association with the ER-resident protein CANX through binding to an unknown intermediate protein by an unclear mechanism. As mitophagy proceeds, the binding between CANX and FUNDC1 weakens and much of the cytosolic DNM1L is recruited by FUNDC1 for the final scission of mitochondria.
in hypoxic cells prevents the accumulation of FUNDC1 at the MAM, suggesting that this resident ER protein might drive the accumulation of FUNDC1. Several pieces of data suggest that hypoxia induces fission of mitochondria but the mechanism of action remains unclear. MFF, FIS1, or MIEF1/MID51-MIEF2/MID49 is involved in recruiting DNM1L to mediate mitochondrial scission in many physiological or stress conditions. But whether these proteins are involved in mitochondrial fission and DNM1L recruitment in response to hypoxia is less studied. Silencing of MFF, FIS1, or MIEF1-MIEF2 does not prevent mitochondrial translocation of DNM1L and subsequent mitochondrial fragmentation in hypoxic cells, although the effect of MFF depletion is less severe. However, in FUNDC1-depleted cells, the number of translocated DNM1L is substantially reduced and the mitochondria are dramatically elongated under hypoxia treatment. Interestingly, when FUNDC1 was overexpressed, most DNM1L colocalized with the fragmented mitochondria. In addition, a double-immunogold labeling experiment indicates they both accumulate at the MAM. Based on the above strong phenotype, we tested if they have a physical interaction. Immunoprecipitation and GST affinity isolation assays both demonstrate that they have direct binding. Similar to CANX, DNM1L also interacts with the cytosolic loop of FUNDC1 (96–139). Furthermore, the truncated FUNDC1 that lacks binding to DNM1L cannot induce mitochondrial fission. Notably, FUNDC1 binds to CANX at a very early stage of hypoxia and then dissociates from CANX and preferentially interacts with DNM1L instead when the length of time under hypoxic conditions is prolonged. This is possible because both proteins bind to the same region of FUNDC1 and they may compete in
different situations by mechanisms yet to be determined. An enhanced binding between LC3 and FUNDC1 at later stages of hypoxia is observed, and FUNDC1 (AA 1–50) is essential for its binding with LC3. In addition to the effects of FUNDC1, DNM1L, CANX, MFF or FIS1 in mitochondrial fission, we also compared their role in mitophagy. Silencing of FUNDC1, CANX or DNM1L causes mitochondrial elongation and reduces mitophagy in response to hypoxia. In contrast, mitochondrial fragmentation and mitophagy frequently occur in hypoxic MFF or FIS1 knockdown cells, or control cells treated with scramble siRNA. Together, we concluded that the mitophagy receptor FUNDC1 is not only a novel MAM protein, but also a new mitochondrial receptor for DNM1L to facilitate mitochondrial fragmentation and mitophagy in response to hypoxia (Fig. 1).
Disclosure of potential conflicts of interest No potential conflicts of interest were disclosed.
Funding This work is supported by the NSFC (No. 31401182), by the National Basic Research Program of China (2013CB910100), by the Natural Science Foundation of Guangdong Province, China (2014A030313533), by the Science and Technology Planning Project of Guangdong Province (2016A020215152), by the Medical Scientific Research Foundation of Guangdong Province, China (No. A2014484), by the Foundation for Distinguished Young Talents in Higher Education of Guangdong, China (2013LYM_0035), by the Sail Plan of Talents Recruitment Grant, Guangdong, China (Yue Ren Cai Ban [2014] 1) to Dr. Du Feng.
