Academic Editors: Guoyao Wu and Graham Pawelec
Introduction: Accumulating evidence suggests that mitochondrial structural and functional defects are present in human placentas affected by pregnancy related disorders, but mitophagy pathways in human trophoblast cells/placental tissues have not been investigated. Methods: In this study, we investigated three major mitophagy pathways mediated by PRKN, FUNDC1, and BNIP3/BNIP3L in response to AMPK activation by AICAR and knockdown of PRKAA1/2 (AKD) in human trophoblast cell line BeWo and the effect of AKD on mitochondrial membrane potential and ATP production. Results: Autophagy flux assay demonstrated that AMPK signaling activation stimulates autophagy, evidenced increased LC3II and SQSTM1 protein abundance in the whole cell lysates and mitochondrial fractions, and mitophagy flux assay demonstrated that the activation of AMPK signaling stimulates mitophagy via PRKN and FUNDC1 mediated but not BNIP3/BNIP3L mediated pathways. The stimulatory regulation of AMPK signaling on mitophagy was confirmed by AKD which reduced the abundance of LC3II, SQSTM1, PRKN, and FUNDC1 proteins, but increased the abundance of BNIP3/BNIP3L proteins. Coincidently, AKD resulted in elevated mitochondrial membrane potential and reduced mitochondrial ATP production, compared to control BeWo cells. Conclusions: In summary, AMPK signaling stimulates mitophagy in human trophoblast cells via PRKN and FUNDC1 mediated mitophagy pathways and AMPK regulated mitophagy contributes to the maintenance of mitochondrial membrane potential and mitochondrial ATP production.
The placenta plays a critical role during pregnancy by maintaining pregnancy, nurturing the fetus and mediating bidirectional communication between the mother and the fetus. More importantly, the placenta is a key mediator of fetal programming by which the long term health and disease risk of offspring is predisposed by the in utero environment such as nutrition, inflammation, endocrine status [1, 2, 3]. Placental functions require a high demand of energy. Thus, the placenta is an active metabolic tissue, accounting for 40 percent of oxygen consumption  and one third of glucose uptake by the placental-fetal unit during late pregnancy . Like other metabolic tissues, mitochondria in the placenta provide the majority of ATP production by metabolizing three major energy substrates: glucose, lipid, and amino acids [6, 7]. In addition to the central role in cellular energy metabolism, mitochondria regulate a variety of biochemical events such as calcium signaling, reactive oxygen species production, apoptosis and steroidogenesis [8, 9]. Therefore, mitochondrial quality and quantity control is indispensable for mitochondrial functions and consequent cellular functions.
Mitophagy, a highly conserved and specific process to remove the dysfunctional/destroyed mitochondria, in concert with mitochondrial biogenesis, is critical for maintaining mitochondrial homeostasis and functions . Mitophagy pathways have been studied primarily in cardiac and neuronal cells due to the high demand of mitochondrial ATP production in the heart and neuronal tissues [11, 12]. Currently, mitohpagy is thought to proceed as follows. Depolarization of mitochondrial membrane or other impairment interrupts normal proteolytic processing of PINK1 kinase. As a consequence, PINK1 accumulates on the outer mitochondrial membrane and phosphorylates MFN2, promoting the recruitment of PRKN and PRKN mediated ubiquitinization of proteins on the outer mitochindrial membrane. Poly-ubiquitinated proteins bind to SQSTM1 protein, which is recognized and bound to LC3 protein in the autophagosome, leading to engulfing mitochondrial proteins into autophagosomes. Following the fusion of autophagosome and lysosome, mitochondrial proteins are degraded by lysosomal proteases . To date, in mammals, three major mitophagy pathways have been defined primarily by the key mediator proteins in mitochondrial or mitochondrial fragmentation recognition and engulfing in each pathway, PINK1-PRKN (mitophagy receptor independent pathway), BNIP3-BNIP3L, and FUNDC1 (mitophagy receptor dependent pathways), respectively [13, 14]. These mediators demonstrate a preferential association with LC3 family members in recruiting autophagosomes that encapsulate mitochondria . These mitophagy pathways are responsive to different cellular stimuli or stress. Briefly, PINK1-PRKN mediated mitophagy responds to mitochondrial membrane potential depletion affected by stresses, such as oxidative stress; BNIP3- BNIP3L mediated mitophagy responds to nutritional stress and/or stimuli  and hypoxia ; FUNDC1 mediated mitophagy responds to mitochondrial uncoupling, and hypoxia [17, 18].
It has been widely accepted that mitophagy is a selective form of autophagy by which the cell recycles macromolecules and organelles to maintain cell survival and thus, sharing similar regulatory mechanisms to autophagy [19, 20]. AMPK signaling and mTOR signaling are major counterregulatory mechanisms of the formation of autophagosomes, a critical regulatory procedure in the initiation of autophagy and subsequent events [21, 22, 23, 24]. AMPK signaling stimulates the initiation of autophagy by AMPK-ULK1 axis in response to nutritional deprivation, while mTOR signaling inhibits the initiation of autophagy . Accumulating evidence suggests that AMPK signaling regulates autophagy in the placenta and/or trophoblast cells, and its regulatory role in autophagy may be disrupted by major pregnancy related disorders. AMPK activation reflects its phosphorylation at Thr172 and was reduced in the placentas from women who were obese prior to their pregnancy [25, 26], gestational diabetes mellitus (GDM) , preeclampsia  and preterm birth , compared their counterparts with normal pregnancy. In addition, mitochondrial structural and functional defects have been found in the placentas from women with maternal obesity and GDM [27, 30, 31, 32, 33]. These observations suggest that a causal relationship exists between AMPK signaling and autophagy/mitophagy in human placentas and/or trophoblast cells.
In this study, we hypothesized that AMPK signaling stimulates mitophagy in human trophoblast cells. To test this hypothesis, first, using human trophoblast cell line, BeWo, a widely applied human trophoblast cell model, we investigated the effect of AMPK activator AICAR (5-Aminoimidazole-4-carboxamide ribonucleotide) on the following parameters, the phosphorylation of AMPK and its target proteins ACC and ULK1, autophagy mediators LC3II, SQSTM1, and mitochondrial receptors PRKN, BNIP3/BNIP3L, and FUNDC1 in total cell lysates and/or mitochondria-enriched fractions, and mitochondrial membrane potential. Second, to confirm the role of AMPK signaling in mitochondrial mitophagy in human trophoblast cell, we investigated mitophagy pathways, mitochondrial membrane potential, and mitochondrial ATP production in response to AMPK knockdown. Our study showed that AICAR activated AMPK signaling stimulates mitophagy in BeWo cells via PRKN and FUNDC1 mediated processes, while AMPK knockdown inhibited mitophagy, increased mitochondrial membrane potential, and reduced mitochondrial ATP production.
Human cytotrophoblast cell line, BeWo (Cat. CCL-98; ATCC, Manassas, VA, USA) was
cultured in F-12K culture medium (Cat. 30-2004; ATCC) supplemented 10% FBS (Cat.
30-2020; ATCC) and penicillin/streptomycin (Cat. 30-2300; ATCC) in a 5% CO
Using AMPK lentivirus shRNA targeting both alpha 1 and 2 subunits PRKAA1/2 (Cat. No. sc-45312-V; Santa Cruz Biotechnology, Dallas, TX, USA), AMPK knockdown was conducted according to the manufacturer’ instructions. Briefly, BeWo cells seeded in 24-well cell culture plate was transfected by shRNA Lentiviral Particles for 36 hours, followed by cell recovery, proliferation, and puromycin selection for 6 days. The AMPK knockdown was confirmed by reduction in both mRNA and protein levels, measured by q-PCR and Western blotting analysis, respectively. The gene knockdown was stable in as least 9 passages as we have determined recently.
Total mRNAs extraction and on-column genomic DNA cleanup were conducted using
PureLink™ RNA Micro Scale Kit (Cat. 12183016; Invitrogen, Waltham,
MA, USA) and all procedures were followed the manufacturer’s instructions. cDNAs
were made from 1
Total cell lysates and mitochondria-enriched components were prepared, following
the procedures described previously [34, 35] with minor modifications. Briefly,
after culture media was removed, cells were washed with phosphate buffered saline
(PBS), lysed in mitochondrial homogenization buffer (10 mM Tris, 1 mM EDTA, and
250 mM sucrose, pH 7.4 and supplemented with protease and phosphatase inhibitor
Cocktail (PPC1010; Sigma) and frozen at –80
Protein concentration in whole cell lysates and mitochondria enriched components
was determined using Pierce™ BCA Protein Assay Kit (Cat. 23227;
Thermo Fisher Scientific, Waltham, MA, USA) and NanoDrop™ One/OneC
Microvolume UV-Vis Spectrophotometer (Cat. ND-ONEC-W; ThermoFisher). Twenty
microgram of whole cell lysates or mitochondrial proteins was mixed with
|total AMPK||Cell Signaling||2532||rabbit||1:1000|
|p-AMPK (Thr172)||Cell Signaling||2535||rabbit||1:1000|
|total ACC||Cell Signaling||3676||rabbit||1:1000|
|p-ACC (Ser79)||Cell Signaling||11818||rabbit||1:1000|
|total ULK1||Cell Signaling||8054||rabbit||1:1000|
|p-ULK1 (Ser555)||Cell Signaling||5869||rabbit||1:1000|
Mitochondrial membrane potential was measured using Tetramethylrhodamine, Methyl
Ester (TMRM) staining. Briefly, cells were seeded into 96-well plates at a
density of 40,000 cells per well, attached to the plate overnight and treated
with or without Chloroquine (40
Seahorse Cell mito stress test is a golden standard assay on mitochondrial ATP
production in live cells by measuring oxygen consumption rate in the presence of
electron transport chain complex inhibitors (oligomycin, rotenone/antimycin) or
ATPase uncoupler Carbonyl cyanide-p-trifluoromethoxyphenylhydrazone (FCCP). AKD
and control BeWo cells were seeded into Seahorse Miniplate at a density of 20,000
cells per well (n = 3) and cultured in cell culture incubator with 5% CO
Data on gene expression and protein abundance were analyzed for the effect of
AICAR, Chloroquine and/or their interactions, and data on gene expression and
protein abundance, mitochondrial membrane potential and ATP production were
analyzed for the effect of AMPK knockdown, using least-squares analysis of
variance (ANOVA) and the general linear model procedures of the Statistical
Analysis System (Version 9.4., SAS Institute, Cary, NC, USA). Log transformation
of variables was performed when the variance of data was not homogenous among
groups, as assessed by the Levene’s test. A p-value
To confirm that AICAR can stimulate AMPK signaling in human trophoblast cells,
we treated BeWo cells with AICAR for 12 or 24 hours and measured the
phosphorylation of amino acids residue specific to AMPK activity and immediate
downstream target proteins ACC and ULK1. The phosphorylated AMPK at Thr172
(p-AMPK) levels were increased (p
AICAR activated AMPK signaling in human trophoblast cell line
BeWo. (A) Western blotting analysis of phosphorylated and total AMPK, ACC, ULK1
after AICAR treatment for 12 or 24 hours. Relative abundance of phosphorylated
AMPK at Thr172 proteins (B), total AMPK (C), phosphorylated ACC at Ser79 (D),
total ACC (E), phosphorylated ULK1 at Ser555 (F), at Ser757 (G) and total ULK1
(H), all normalized to ACTB. Data are presented as the mean
Autophagy flux with specific blockage of lysosomal degradation of autophagy
target protein is used in autophagy analysis. Chloroquine is one of widely used
chemical which inhibits the fusion of autophagosome and lysosome . There is
no report on autophagy/mitophagy flux in human trophoblast cells, but the dose of
chloroquine is critical in the mitophagy flux analysis due to its cytotoxicity at
high doses, so we optimized the dose of chloroquine by investigating the main
autophagy/mitophagy mediator LC3II in mitochondria enriched fractions to in
response to different doses of CLQ. Our data indicated that CLQ at the
concentration of 40
In whole cell lysates, AICAR alone increased the protein abundance of LC3II for
AICAR stimulated autophagy flux in human trophoblast cell line
BeWo. (A) Western blotting analysis of LC3II and SQSTM1 in mitochondrial
fractions and whole cell lysates extracted from BeWo cells treated with AICAR,
CLQ, their combination (AICAR+CLQ) and controls (CT) and MemCode staining.
Quantification of the relative abundance of LC3II and SQSTM1 in mitochondrial
fractions (B,D) and whole cell lysates (C,E) proteins by normalizing density
of a band in Western blot to MemCode staining signal in the entire lane. Data
are presented as the mean
In mitochondrial fractions, AICAR alone increased the protein abundance of LC3II
for 1.92-fold (p
We next analyzed mitophagy flux by measuring the abundance of mediators of 3
major mitophagy pathways by Western blotting (Fig. 3A). In mitochondrial
fractions, AICAR alone increased the protein abundance of PRKN, and FUNDC1
proteins by 1.24-,1.58-fold (both p
AICAR regulated mitophagy pathways human trophoblast cell line
BeWo. (A) Western blotting analysis of on PRKN, FUNDC1, BNIP3, and
BNIP3L in mitochondrial fractions from BeWo cells treated with AICAR, CLQ, their
combination (AICAR+CLQ) and controls (CT). Quantification of
the relative abundance of PRKN (B), FUNDC1 (C), BNIP3 (D) and BNIP3L (E) proteins
in mitochondrial fractions by normalizing density of a band in Western blot to
MemCode staining signal in the entire lane (shown in Fig. 2). Data are presented
as the mean
To confirm the regulation of AMPK on mitophagy and mitochondrial ATP production,
expression of the two genes encoding catalytic subunits of AMPK (PRKAA1/2) was
knocked down in BeWo cells and a stable cell line was established. The mRNA
levels of PRKAA1/2 were reduced by 2.03- and 2.07-fold (both p
The abundance of LC3II proteins in mitochondrial fractions was reduced by
Reduced autophagy flux and mitophagy mediators by PRKAA1/2
knockdown in human trophoblast cell line BeWo. (A) Western blotting analysis of
LC3II, SQSTM1, PRKN, FUNDC1, BNIP3, and BNIP3L proteins in mitochondrial
fractions in AKD and BeWo cells treated with CLQ or without treatment and MemCode
staining of the blot. Quantification of the relative abundance of LC3II (B),
SQSTM1 (C), PRKN (D), FUNDC1 (E), BNIP3 (F), and BNIP3L (G) proteins by
normalizing density of a band in Western blot to MemCode staining signal in the
entire lane. Data are presented as the mean
Mitochondrial oxidative phosphorylation and ATP production are dependent on the
finely tuned mitochondrial membrane potential . To investigate the effect of
AMPK knockdown on mitochondrial membrane potential, TMRM staining was conducted
and compared between AKD and control BeWo cells. The average TMRM Mean intensity
in AKD cells was increased by 1.094-fold (p
Mitochondrial membrane potentials increased by PRKAA1/2
knockdown in human trophoblast cell line BeWo. (A) Representative staining of
TMRM (red) and Hoechst 33342 (blue) in PRKAA1/2 knockdown (AKD) and control BeWo
cells (Bar = 200
To investigate whether impaired mitophagy and elevated mitochondrial membrane
potential affect mitochondrial ATP production, Seahorse cell mito stress test was
conducted on AKD and control BeWo cells. ATP production via mitochondrial
oxidative phosphorylation was lower in basal respiration and in presence of
oligomycin, FCCP, and rotenone/antimycin in AKD compared to CT cells (Fig. 6A).
The basal respiration (Fig. 6B), maximal respiration (Fig. 6C), spare respiratory
capacity (Fig. 6D), ATP production-coupled respiration (Fig. 6E),
non-mitochondrial oxygen consumption (Fig. 6F) and coupling efficiency (Fig. 6G)
were reduced by 1.54- (p
Reduced mitochondrial ATP production by PRKAA1/2 knockdown in
human trophoblast cell line BeWo. (A) Dynamic oxygen consumption rate (OCR) in
PRKAA1/2 knockdown (KD) and control (CT) cells in the presence of Oligomycin A,
FCCP, Rotenone/Antimycin measured by Seahorse Mito Cell Stress Test. (B–H)
Comparison of OCR for the basal respiration (B), maximal respiration (C), spare
respiratory capacity (D), ATP production-coupled respiration (E), and
non-mitochondrial oxygen consumption (F), and proton leak (H) between KD and CT
cells. Data are presented as the mean
Mitophagy plays a critical role in maintaining mitochondrial homeostasis in response to nutritional stresses, primarily monitored by the cell energy sensor AMPK and associated signaling. During human pregnancy, placental mitochondria are challenged by potent oxidative and nitrative stresses and many other deleterious factors [38, 39]. Thus, effective mitophagy is critical for maintaining proper mitochondrial hemostasis and functions. To date, accumulating evidence supports that major pregnancy related disorders are associated with altered mitochondrial functions and/or autophagy [27, 40, 41, 42, 43, 44, 45, 46], but mitophagy pathways and the role of AMPK signaling in the regulation of mitophagy remain unclear in human trophoblast cells. Our study for the first time delineated major mitophagy pathways and confirmed that AMPK signaling stimulates mitophagy in BeWo human trophoblast cells via PRKN and FUNDC1 mediated pathways (Figs. 2,3). Lower AMPK protein abundance reduces mitophagy and mitochondrial ATP production (Figs. 4,6), coincident with elevated mitochondrial membrane potential (Fig. 5).
We for the first time elucidate the regulatory effect of AMPK signaling in human trophoblast cells by analyzing three major mitophagy pathways by AMPK overactivation and AMPK knockdown. The major mitophagy pathways mediated by PINK1/PRKN, BNIP3/3L and FUNDC1 have been characterized in other cell types including cardiomyocytes and neurons [12, 17], reflecting a rescue strategy to promote cell survival in response to a variety of stresses [47, 48]. Our study found these mitophagy pathways are present in human trophoblast cells (Figs. 3,4). More importantly, for all mitochondrial receptors or mitophagy mediators (LC3II, SQSTM1, PRKN, BNIP3/BNIP3L and FUNDC1) investigated in this study, the pattern of changes in response to AMPK activation is opposite to that in response to AMPK knockdown (Figs. 3,4). Thus, AMPK signaling, interweaved in complicated cellular communications, plays a critical role in the regulation of mitophagy. Conversely, the pattern of changes in BNIP3/BNIP3L is opposite to that of PRKN and FUNDC1 and negatively corelated with mitophagy. Enhanced mitophagy in response to AMPK activation is coincident with reduced BNIP3/3L protein abundance in mitochondrial fractions (Fig. 3) while impaired mitophagy in response to AMPK knockdown is coincident with increased BNIP3/3L protein abundance in mitochondrial fractions (Fig. 4). These observations suggest that AMPK may regulate mitophagy via several pathways (PRKN, FUNDC1) and BNIP3/3L mediated mitophagy may serve as a counterregulatory mechanism in the regulation of mitophagy. Trophoblast cells are metabolically active and imposed many stresses such as nutritional, oxidative, nitrative and hypoxia stresses [2, 49], thus requiring different mitophagy pathways to respond to multiple stress conditions. How these different pathways are activated in the placenta remains unclear; our study indicates that AMPK is a key regulator.
AMPK knockdown in human trophoblast cells decreases mitochondrial ATP production, possibly via hyperpolarization of mitochondrial inner membrane and impaired mitophagy. Mitochondrial membrane potential is critical for mitochondrial ATP production and the membrane potential must be kept in a narrow range to maintain sustainable ATP production . However, unlike in excitable cells including cardiomyocytes and neutrons, how mitochondrial membrane potential is controlled in human trophoblast cells has not been reported, to the best of our knowledge. Our study found that AMPK knockdown reduces mitochondrial ATP production (Fig. 6A), which was similar to that found in the mouse trophoblast stem cell line SM-10 . In addition to reduced basal ATP production, the underlying mechanisms responsible for reduced mitochondrial ATP production in AKD cells include reduced maximal respiration (Fig. 6C), spare respiratory capacity (Fig. 6D), and ATP production-coupled respiration (Fig. 6E), indicating that AMPK knockdown exerts broad effects on electron transport chain complexes and ATPase activities. Interestingly, our study demonstrated that mitochondrial membrane potential in cells with PRKAA1/2 knockdown was increased 10% (Fig. 5) while mitochondrial ATP production was reduced 32% in basal levels (Fig. 6). These observations indicate that mitochondrial membrane potential contributes largely to mitochondrial ATP production in human trophoblast cells, and therefore, mitochondrial membrane potential must be controlled in a narrower range to optimize mitochondrial ATP production compared to other cell types. Conversely, hyperpolarization of the mitochondrial inner membrane in response to AMPK knockdown may contribute to impaired mitophagy. It is known that depolarization of mitochondrial inner membrane, indicated by reduced mitochondrial membrane potential, is considered to be the trigger of mitophagy, especially that mediated by PINK1/PRKN pathway [51, 52]. Therefore, the reduced PRKN protein levels in mitochondrial fractions in AKD cells, together with reduced LC3II and SQSTM1 (Fig. 4), support that mitophagy and mitochondrial membrane potential are orchestrated in response to AMPK signaling.
There are several weaknesses in our study due to unavoidable limitations. First, we could not include an AMPK antagonist as a negative control for AICAR induced AMPK activation because there is no reliable AMPK antagonist available . Dorsomorphin, also called Compound C, has been widely used as an AMPK antagonist  but its inhibition of AMPK is nonspecific [55, 56] and more than 50 other protein kinases are inhibited simultaneously . However, our AMPK knockdown strategy confirms the regulation of AMPK on mitophagy and mitochondrial function in an inhibitory context (Figs. 4,6). Second, human primary trophoblast cells do not proliferate in culture, so we used BeWo cells that are widely applied in mechanistic studies on placentas and trophoblast cells . However, as a cancer cell line , BeWo cells might have acquired additional features that affect mitophagy and mitochondrial functions. Thus, some findings from BeWo cells should be validated in human primary cytotrophoblast cells. Third, in the mitophagy flux assay, the accumulation of FUNDC1 and BNIP3 in mitochondrial fractions was similar in response to the combination of AICAR and CLQ and CLQ alone (Fig. 3C,D), possibly because mitophagy receptors are inactive under resting conditions, and their activities are elicited differently upon signals triggering mitochondrial damage . While Western blotting analysis of mitophagy receptors is the most reliable to study autophagy/mitophagy , other complementary methods may be included in future validation studies of some findings in this study.
To date, the dynamic changes of placental mitophagy during pregnancy and the cause-effect relationship of placental mitophagy and pregnancy related disorders remain unclear, but a handful of studies have demonstrated the link between autophagy/mitophagy defects in placental tissue or trophoblast cells in major pregnancy related disorders such as gestational diabetes mellitus and preeclampsia [27, 40, 41, 42, 43, 44, 45, 46]. However, due to technical limitations with in vivo studies and the lack of reliable animal model for mechanistic studies, mitophagy in the placenta has remained poorly understood. While characterizing mitophagy in human trophoblast cells, we are currently investigating the cause of altered placental mitophagy using primary trophoblast cells or tissues from women with major pregnancy related disorders, aiming to dig out the underlying mechanisms present in the placental-fetal unit during pregnancy. Thus, this study provides a conceptual foundation to conduct future mechanistic studies on placental mitophagy in normal pregnancy and pathophysiological status.
This study indicates that major mitophagy pathways mediated by PRKN, FUNDC1, BNIP3/BNIP3L are present in human trophoblast cells and AMPK signaling regulates mitophagy via PRKN and FUNDC1 mediated mitophagy pathways, which may affect mitochondrial membrane potential and mitochondrial ATP production.
AICAR, 5-Aminoimidazole-4-carboxamide ribonucleotide; ACC, Acetyl-CoA carboxylase; AMPK, 5’ adenosine monophosphate-activated protein kinase; BNIP3, BCL2/adenovirus E1B 19 kDa protein-interacting protein 3; BNIP3L, BCL2 Interacting Protein 3 Like; FUNDC1, FUN14 domain containing 1; LC3II, Microtubule-associated proteins 1A/1B light chain 3B; MFN2, Mitofusin-2; PINK1, PTEN-induced kinase 1; PRKN, Parkin; SQSTM1, Sequestosome 1; ULK1, Unc-51 like autophagy activating kinase 1.
HG and YC designed the research study. BW and HG performed the research. YC, RC, EA, PY and BH provided help and advice on research protocols, data interpretation and discussion. BW, YC and HG analyzed the data. HG wrote the manuscript. All authors contributed to editorial changes in the manuscript. All authors read and approved the final manuscript.
The authors thanks Gene Expression Core, Howard University RCMI program for sharing instruments in Western blotting imaging.
This research was funded by National Institutes of Health grants R03HD095417 (NICHD), U54MD007597 (NIMHD, Howard University RCMI Program), and Bridge Fund/Pilot Study Award (Dean’s Office Howard University College of Medicine).
The authors declare no conflict of interest.