| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,1
Departments of
* Neuroscience and Cell Biology,
Pathology, and
Internal Medicine, University of Texas Medical Branch, Galveston, Texas, USA
1Correspondence: Department of Neuroscience and Cell Biology, University of Texas Medical Branch, 301 University Blvd., Galveston, TX 77555-0620, USA. E-mail: sweinman{at}utmb.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2-fold. Entry was entirely inhibited by the mitochondrial Ca2+ uniporter inhibitor, Ru-360, but not influenced by an Na+/Ca2+ exchanger inhibitor or ROS scavengers. These results indicate that core directly increases mitochondrial Ca2+ uptake via a primary effect on the uniporter. This enhanced the ability of mitochondria to sequester Ca2+ in response to ER Ca2+ release, and increased mitochondrial ROS production and MPT. Thus, the mitochondrial Ca2+ uniporter is a newly identified target for viral modification of cell function.—Li, Y., Boehning, D. F., Qian, T., Popov, V. L., Weinman. S. A. Hepatitis C virus core protein increases mitochondrial ROS production by stimulation of Ca2+ uniporter activity.
Key Words: HCV core • mitochondria • cytosolic Ca2+ • MAVS
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, but the pathogenesis of this viral infection is still not well characterized. Mitochondrial dysfunction has been noted in infected patients and may be an important factor in disease progression (2
3
4)
. Recent studies have demonstrated that several HCV proteins interact with mitochondria. The viral NS3–4A protease cleaves a mitochondrial antiviral signaling protein (MAVS) and results in its dissociation from the mitochondrial membrane. This effect contributes to viral escape from the innate immune response (5
6
7)
. Two other HCV proteins, core and p7, have also been found to partially localize to mitochondria, but the significance of these interactions is still unknown (8
, 9)
. The HCV core protein represents the first 191 amino acids of the viral precursor polyprotein, and a 179 amino acid mature form is released from the endoplasmic reticulum (ER) after cleavage of a signal peptide (10)
. Suzuki et al. reported that amino acids 112–152 can mediate association of the core protein to both ER and mitochondria (11)
, and Schwer et al. identified a mitochondrial outer membrane localization signal at position 149–158 (12)
. Our previous studies have demonstrated that core protein reduced electron transport complex I activity, increased mitochondrial reactive oxygen species (ROS) production, and increased mitochondrial Ca2+ entry with a subsequent increase in ROS in isolated mitochondria (9)
.
Mitochondria are recognized as an important intracellular Ca2+ store and actively participate in Ca2+ signaling. Increased mitochondrial Ca2+ has been noted to have a number of effects including increasing electron transport, increasing ROS production, and, under some conditions, causing the opening of the permeability transition pore (13)
. The mitochondrial Ca2+ content is regulated by both Ca2+ uptake and efflux. Mitochondria have a strong negative matrix electrochemical potential, and under normal conditions they take up Ca2+ specifically via a Ca2+-selective channel in the inner membrane called the Ca2+ uniporter. Efflux of Ca2+ from mitochondria occurs as well, but needs to be coupled to another ion gradient through Na+/Ca2+ exchange or H+/Ca2+ exchange (14)
.
In many cell types, including hepatocytes, ER and mitochondria are in close contact, and ER-mitochondria Ca2+ transfer has been recognized as a mechanism by which mitochondrial function is regulated (15
, 16)
. In addition to its effects on mitochondria, HCV core protein also produces ER stress with activation of an ER unfolded protein response, increased expression of ER chaperone proteins, and release of ER Ca2+ stores (17
, 18)
. However, it is not known whether core protein has direct effects on mitochondrial Ca2+ uptake. In this study we examined the effect of HCV core on mitochondrial Ca2+ and ER to mitochondrial Ca2+ transfer. The results show that core protein directly affects mitochondria by increasing Ca2+ uptake via the uniporter, sensitizing mitochondria to MPT, and increasing mitochondrial ROS production secondarily as a result of the Ca2+ increase. In addition, it enhances the fraction of agonist-induced ER Ca2+ release that enters the mitochondria. In combination with its ability to produce ER stress and ER Ca2+ release, the net effect is to sensitize mitochondria to changes in cellular Ca2+ homeostasis. Stimulation of the mitochondrial Ca2+ uniporter thus is an additional mechanism by which viruses can alter cell function.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. They were maintained in DMEM/F12 with glutamine and supplemented with 10% fetal bovine serum (FBS), 1% penicillin-streptomycin, 200 µg/ml G418, 4 µg/ml blasticidin, and 2 µg/ml tetracycline. To induce HCV core protein, cells were cultured in medium free of tetracycline for 48–72 h. HCV transgenic mice expressing core, E1, and E2 (strain SL-139) were described previously (9)
. C57BL/6 mice were purchased from Jackson Lab (Bar Harbor, ME, USA). Female mice 6 months old were used in the experiments. Recombinant core protein (HCV NG1b amino acids 1–179 derived from Escherichia coli) was kindly provided by Dr. S. Watowich. As an independent source, a second preparation of recombinant core protein (HCV NG1b amino acids 1–186) was purchased from GenScript Corp. (Piscataway, NJ, USA). Human synaptotagmin I C2A was provided by Dr. R. B. Sutton. Calf thymus histone was purchased from AXXORA, LLC (San Diego, CA, USA). MitoSOXTM Red, Rhod-2AM, Fura-2AM, rhodamine123, and 5,5',6,6'-tetrachloro-1,1',3,3'-tetraethylbenzimidazolcarbocyanine iodide (JC-1) were purchased from Molecular Probes (Carlsbad, CA, USA). Reagents other than those specifically mentioned were purchased from Sigma-Aldrich (St. Louis, MO, USA).
Isolation of liver mitochondria
Mouse liver mitochondria were isolated by a modification of the method of Johnson and Lardy as reported previously (9
, 20)
. Briefly, liver samples were gently homogenized in isolation buffer (250 mM sucrose, 10 mM HEPES, 0.5 mM EGTA, 0.1% BSA, pH 7.4) by three or four strokes with a Dounce homogenizer. The homogenate was centrifuged at 600 g for 5 min at 4°C. The supernatant was centrifuged at 7800 g for 10 min at 4°C to obtain a crude mitochondrial pellet.
Measurement of Ca2+ in isolated mitochondria
Freshly prepared mitochondria were pretreated with or without recombinant HCV core protein on ice for 30 min. For mitochondrial Ca2+ determination, mitochondria were first loaded with the Ca2+ indicator Rhod-2 AM, 4 µM in isolation buffer for 1 h at 4°C, washed twice with 250 mM sucrose, 2 mM K-HEPES buffer, then diluted to a final concentration of 0.33 mg/ml with "respiration buffer" (100 mM KCl, 20 mM Tris, 20 mM HEPES, 10 mM NaCl, 5 mM sodium succinate, 1 mM KH2PO4, 20 µM K-EGTA, 2 µM rotenone, and 1 µg/ml oligomycin, pH 7.2). The mitochondrial suspension was aliquoted in 96-well plates and incubated with Ru360 (Calbiochem, San Diego, CA, USA), glutathione ethyl ester (Calbiochem) or CGP37157 (Tocris Cookson Inc., Ellisville, MI, USA) at 4°C for 30 min. The mitochondria were then exposed to Ca2+ containing respiration buffer (250 µM Ca2+) and Rhod-2 fluorescence was measured every 30 s for 30 min with a FLUOstar OPTIMA fluorescence plate reader (Ex/Em 544/590 nm) (BMG LABTECH, Durham, NC, USA).
Measurement of cytosolic Ca2+
Cytosolic Ca2+ measurements were performed as described previously by Venkatachalam et al. (21)
. Cells grown on coverslips were loaded with 5 µM Fura-2-AM for 30 min at room temperature in buffer (107 mM NaCl, 7.2 mM KCl, 1.2 mM MgCl2, 1 mM CaCl2, 11.5 mM glucose, 20 mM HEPES-NaOH, pH 7.2). Cells were washed briefly and incubated for another 20 min in Fura-2 AM-free solution. Coverslips were then mounted in chamber with nominally Ca2+-free buffer (107 mM NaCl, 7.2 mM KCl, 1.2 mM MgCl2, 11.5 mM glucose, 20 mM HEPES-NaOH, pH 7.2) to minimize Ca2+ entry and visualized with a TE200-IUC Quantitative Fluorescence Live-Cell Imaging System (Nikon, Melville, NY, USA). Fluorescence emission images at 505 nm were acquired every 2 s after sequential excitation at 340 and 380 nm. Fluorescence intensity at each wavelength was determined in the cytosolic region and corrected by subtraction of background fluorescence. Data analysis was performed using MetaMorph software and cytosolic Ca2+ was represented as F340/380 nm ratio. Data are presented as mean ± SE and are the average of 5 independent experiments, each consisting of at least 10 separate cells measured.
Measurement of mitochondrial Ca2+ in live cells
Cells were seeded on coverslips and transfected with the mitochondrial targeted calcium indicator protein (ratiometric pericam) plasmid kindly provided by Dr. Atsushi Miyawaki (Wako, Saitama, Japan) (22)
. After 48 h, coverslips with cells were mounted in chamber with Ca2+-free buffer (107 mM NaCl, 7.2 mM KCl, 1.2 mM MgCl2, 11.5 mM glucose, 20 mM HEPES-NaOH, pH 7.2). Images were acquired every 2 s (emission at 510 nm and excitation at 495 and 405 nm) using the TE200 system. Mitochondrial Ca2+ was determined as the F 495/405 nm ratio. Data are presented as mean ± SE.
Mitochondrial membrane potential determination
Cells were plated on glass coverslips and incubated with rhodamine 123 (500 nM) at 37°C for 20 min. Cells were then bathed in nominally Ca2+-free buffer and observed in a Nikon TE200 fluorescence microscope (emission at 510 nm and excitation at 488 nm). At time zero, thapsigargin (200 µM) was added and images were recorded at 15 s intervals. Loss of 
was indicated by loss of the mitochondrial punctate distribution of rhodamine123 and the appearance of diffuse cytoplasmic fluorescence.
As a second method to measure , cells were incubated with the mitochondrial membrane potential-sensitive fluorophore, JC-1 (5 µg/ml), at 37°C for 15 min. They were subsequently analyzed by flow cytometry as described (19)
. In some cases cells were pretreated with cyclosporin A (1 µM) before thapsigargin treatment. Data are presented as mean ± SE, n = 4 samples for each treatment. JC-1 fluorescence of 10000 cells was collected in each sample.
Determination of mitochondrial superoxide
Cells were incubated in DMEM/F12 medium (phenol red free) containing 2.5 µM mitoSOXTM Red at 37°C for 30 min (23)
. After a brief wash, cells were mounted in a chamber with Ca2+-free buffer, and images (Ex/Em 510 nm/570 nm) were recorded at various time points after drug treatment using an LSM510 META advanced laser scanning confocal microscope (Carl Zeiss MicroImaging, Inc., Thornwood, NY, USA). MitoSOX fluorescence was quantitated by analysis of cell fluorescence intensity on image data using Metafluor software. Alternatively, L14 cells ("on " and "off") were plated in 96-well plates and incubated with mitoSOX. Cells were then washed briefly with Ca2+-free buffer and treated with thapsigargin (10 µM) for 2 h. MitoSOX fluorescence was subsequently measured in a fluorescence plate reader. Data are presented as mean ± SE.
Determination of mitochondrial permeability transition (MPT)
MPT was monitored as a decrease in light scattering from mitochondrial suspensions. It was determined from the optical density (O.D.) at 540 nm in a medium containing 150 mM KCl, 25 mM NaHCO3 1 mM MgCl2, 3 mM KH2PO4, 20 mM HEPES, and 5 mM sodium succinate pH 7.4 (24
, 25)
. Mitochondrial protein (1 mg/ml) was present in each sample and O.D. was monitored in a spectrophotometer at 24°C for 20 min at 10 s intervals.
Electron microscopy
Six-month-old female SL139 and wild-type mice were anesthetized with pentobarbital (0.1 mg/10 g weight) and the portal vein was canulated for perfusion. The portal vein was perfused with 10 µM thapsigargin in Hank’s buffer (NaCl 120 mM, KCl 5 mM, NaH2PO4 0.6 mM, NaHCO3 19 mM, pH 7.4) at a rate of 8 ml/min for 2 min. Livers were then immediately perfused with fixative (2.5% formaldehyde, 0.1% glutaraldehyde in 50 mM cacodylate buffer containing 0.03% trinitrophenol and 0.03% CaCl2 pH 7.2). Livers were then removed and cut into cubes of 1 mm (3)
for further fixation. Samples were washed in 100 mM cacodylate buffer, postfixed in 1% OsO4 in the same buffer, en bloc stained with 2% aqueous uranyl acetate, dehydrated in ethanol, and embedded in Poly/Bed 812 (Polysciences, Warrington, PA, USA). Ultrathin sections were cut on a Reichert-Leica Ultracut S ultramicrotome, stained with lead citrate, and examined in a Philips 201 or Philips CM-100 electron microscope at 60 kV. Assessment of mitochondrial morphology was performed by categorizing mitochondria into two types. Normal mitochondria were round, 0.8–1.5 µm in diameter, with either dense or light matrix, and had preserved cristae. Abnormal mitochondria had elongated and/or oval shapes, 0.8–1 x 3 µm with very light matrix and a small amount of cristae.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 19
, 26
, 27)
. To test whether core protein alters the mitochondrial response to ER Ca2+ release, we induced ER Ca2+ release in L14 cells with thapasigargin (TG), an ER Ca2+ pump inhibitor, and measured mitochondrial superoxide production with MitoSOXTM Red, a mitochondrially localized, superoxide-sensitive fluorescent ROS indicator. As demonstrated in Fig. 1
A, B, treatment of core-expressing ("on") cells with TG resulted in greater mitochondrial superoxide production compared with cells not expressing core ("off"). The failure of the "off" cells to produce ROS was related specifically to a different response to TG, since elevated mitochondrial superoxide production was readily observed in "off" cells after treatment with the redox recycling agent menadione (MND) (Fig. 1C, D
). TG-induced ROS production was Ca2+ dependent since pretreatment with the Ca2+ chelator 1,2-bis(2-aminophenoxy)ethane-N,N,N<<,N<<-tetraacetic acid prevented the short-term increase of mitochondrial superoxide measured in the imaging studies (Fig. 1B
). MitoSOX fluorescence was also measured from a larger number of cells in a plate reader assay over a longer period (2 h). These results (Fig. 1E
) show that TG increased superoxide production in both core-expressing cells and control cells, but there was an 1.5-fold increase in superoxide production in core-expressing cells compared with control cells. The increase of mitoSOX fluorescence was abolished by pretreating cells with the mitochondrial Ca2+ uptake inhibitor Ru360. Core protein thus stimulates mitochondrial ROS production in response to ER Ca2+ release. This stimulation requires mitochondrial Ca2+ uptake.
|
Effect of HCV core protein on cytosolic and mitochondrial Ca2+ uptake
To determine whether increased ROS production resulted from greater mitochondrial Ca2+ uptake, we measured cytosolic and mitochondrial Ca2+ after ER Ca2+ release. Figure 2
A demonstrates that cytosolic Ca2+ increases similarly in control and core-expressing cells in response to ER Ca2+ store release with TG. This indicates that both cell types have identical total ER Ca2+ stores. Despite its lack of effect on the cytosolic Ca2+ response, core protein expression dramatically increased mitochondrial Ca2+ uptake as measured with the mitochondrially localized calcium indicator protein ratiometric pericam (22)
. Core-expressing cells had a more rapid increase in mitochondrial Ca2+, followed by a quick decrease to levels lower than baseline. In cells not expressing core, there was a smaller increase, followed by a slow reduction of mitochondrial Ca2+ back to baseline (Fig. 2B
).
|
To determine whether core protein alters ER to mitochondrial Ca2+ transfer in response to physiological agonist-induced ER Ca2+ release, we measured both cytosolic and mitochondrial Ca2+ after exposure of cells to extracellular ATP, an agonist for purinergic receptors that induces ER Ca2+ release via IP3 production. ATP induced an increase in cytosolic Ca2+ concentration in both "off" and "on" cells. While both cell types had identical baseline cytosolic Ca2+, cytosolic Ca2+ responses to ATP peaked at greater levels in "off" cells than in "on" cells (P=0.018) (Fig. 2C
). This could be explained if core increased the "buffering" of released Ca2+. To test this possibility, we measured mitochondrial Ca2+ content with mitochondrially targeted pericam. As shown in Fig. 2D
, mitochondria in core-expressing cells had a higher baseline Ca2+ and took up more Ca2+ in response to ATP than did mitochondria in cells not expressing core protein (Fig. 2D
). Mitochondrial Ca2+ increases in response to ATP were considerably smaller than after TG-induced Ca2+ release, possibly because the cytosolic Ca2+ increase after ATP was smaller and of shorter duration than that induced by TG (Fig. 2A, C
).
HCV core protein promotes Ca2+-induced mitochondrial permeability transition
Since we have previously shown that core protein increases mitochondrial depolarization in response to oxidative stress (19)
, we hypothesized that the secondary TG-induced decrease in mitochondrial Ca2+ below baseline (Fig. 2B
) might be a result of the onset MPT resulting in sudden loss of mitochondrial membrane potential () and the opening of a large conductance permeable to Ca2+. To test this hypothesis, we measured the effect of TG on mitochondrial membrane potential. Cells were loaded with rhodamine 123 (Rh123), a positively charged fluorescent molecule actively sequestered in mitochondria due to but that rapidly exits mitochondria in response to MPT (28)
. As shown in Fig. 3
A, all mitochondria initially concentrate Rh123 and display a punctate fluorescence pattern. In response to TG treatment, core-expressing cells quickly lost the punctate distribution of Rh123, whereas mitochondria in cells not expressing core maintained the normal mitochondrial punctuate distribution. As an alternative way to measure changes in mitochondrial membrane potential, cells were also loaded with the mitochondrial potential sensing fluorophore, JC-1, and examined by flow cytometry for evidence of TG-induced depolarization. Figure 3B
shows that core-expressing cells were more significantly depolarized by TG than non-core-expressing cells and that this depolarization could be blocked by the MPT inhibitor, cyclosporine A (CSA). Taken together, these results indicate that core protein first increases mitochondrial Ca2+ uptake in response to TG, then subsequently sensitizes mitochondria to TG-induced permeability transition and depolarization.
|
Core protein sensitizes mitochondria MPT in HCV core transgenic mice
We next tested whether HCV proteins alter mitochondrial responses to Ca2+ and ROS by using HCV transgenic mice that express the structural proteins core, E1, and E2. These mice have an HCV core in the liver comparable to those in human disease and have been shown to have an oxidative stress phenotype (9)
. In the absence of Ca2+, mitochondria isolated from wild-type and transgenic mouse livers were stable and did not undergo MPT over 15 min (Fig. 4
A). In response to 25 µM Ca2+, transgenic mitochondria had a more rapid decrease in light scattering compared with wild-type mice (Fig. 4A
), demonstrating they are more sensitive to Ca2+-induced MPT. Previous work from our lab has shown that core protein expression in cells increases ROS-induced cell death (19)
. We therefore tested whether this was also true for MPT in isolated mitochondria. When liver mitochondria were exposed to the lipophillic peroxide tBOOH (100 µM), transgenic mitochondria underwent MPT more rapidly than did mitochondria from wild-type livers (Fig. 4B
). In both experiments, CSA prevented the loss of light scattering, demonstrating that the measurement correctly reflects opening of the permeability transition pore.
|
TG-induced mitochondrial damage in HCV core transgenic mice
To test whether similar effects on mitochondria occur in liver in vivo, we perfused livers from control and SL-139 transgenic mice with TG. After 2 min perfusion with TG, 80% of liver mitochondria in transgenic mice exhibited swelling and loss of cristae (Fig. 5
A, B). Neither transgenic mice without treatment nor wild-type mice after TG treatment showed similar mitochondrial morphological changes. These studies thus indicate that HCV core protein sensitizes mitochondria to the effects of ER Ca2+ release in intact liver.
|
Core protein effects on mitochondrial Ca2+ uniporter activity
To determine the direct effect of core protein on mitochondria, we isolated mitochondria from C57BL/6 mouse liver and measured mitochondrial Ca2+ uptake with Rhod-2. Although this mitochondrial prep does contain ER membranes, the absence of ATP and prolonged incubation with a Ca2+ chelator eliminates ER as a source of Ca2+ or a site of active Ca2+ uptake. The mitochondria, however, are energized by incubation with succinate and retain the ability for potential-driven Ca2+ uptake. Figure 6
shows the characteristics of this assay. In the absence of Ca2+, either mitochondria alone or mitochondria loaded with Rhod-2 displayed minimal fluorescence. Upon addition of 250 µM Ca2+, there was a large increase of Rhod-2 fluorescence. Collapse of the mitochondrial membrane potential with the uncoupler FCCP prevented the increase in fluorescence, indicating an active transport process. This demonstrates that the fluorescence signal represents Ca2+ uptake and not Ca2+ interaction with extramitochondrial Rhod-2. As expected, prevention of Ca2+ uptake by chelation with EGTA similarly prevented the increase in fluorescence (Fig. 6A
).
|
We next determined the kinetics of Ca2+ uptake in this system (Fig. 6B
). Addition of Ca2+ resulted in a monotonic uptake process that reached a plateau after 10 min. Since the Ca2+ uniporter is known to be the only significant source of Ca2+ uptake (29)
, we expected that uptake would be prevented by the uniporter inhibitor, Ru-360. Figure 6B
shows the unexpected finding that whereas Ru-360 dramatically inhibited the early time uptake of Ca2+, eventually Ca2+ entry occurred and increased toward the control plateau value. We hypothesized that this second Ca2+ entry pathway might be the permeability transition pore. This would occur because the Ca2+ gradients present in this in vitro system are different from those present in live cells and would result in net Ca2+ uptake after permeability transition pore opening (see Discussion). We tested this hypothesis by observing Ca2+ uptake in mitochondria treated with the MPT inhibitor CSA in addition to Ru-360. As shown in Fig. 6B
, the combination of these two inhibitors prevented Ca2+ entry. This indicates that in the presence of CSA, this system is able to measure Ca2+ uptake due exclusively to the uniporter.
We next determined the effect of core protein on mitochondrial Ca2+ uptake. CSA was present in all these experiments. Figure 7
A demonstrates that the initial rate of mitochondrial Ca2+ uptake was increased by a factor of 1.8 ± 0.1 when mitochondria were preincubated with HCV core protein (100 ng core/mg mitochondrial protein). The plateau value was also increased. Ru360 completely abolished the increase of Ca2+ entry induced by core. Since the total mitochondrial Ca2+ content reflected by the Rhod-2 fluorescence signal is a net result of Ca2+ influx and efflux, we also studied whether the core effect resulted from a change in the activity of the Ca2+ efflux transporter, the Na+/Ca2+ exchanger. Figure 7B
shows that an inhibitor of Na+/Ca2+ exchange, CGP37157, did not affect the mitochondrial Ca2+ increase. Since core protein increases mitochondrial ROS production in addition to Ca2+ uptake, we tested whether the ROS increase was responsible for increased Ca2+ entry. We pretreated mitochondria with a permeable ROS scavenger, glutathione ethyl ester (GEE), and found that GEE did not affect the core-induced mitochondrial Ca2+ increase (Fig. 7C
).
|
To determine whether the effect of core in mitochondria is a specific phenomenon, we tested several control proteins as well as a second independent source of core protein. Neither calf thymus histone, another small positively charged protein with nucleic acid binding ability, nor synaptotagmin I C2A, a small protein produced by expression in E. coli, had any effect on mitochondrial Ca2+ uptake (Fig. 7D
). A second preparation of core protein obtained from a commercial source, core NG1b (1–186), had an effect similar to the previously used core 1–179 prep. Both core protein preparations increased mitochondrial Ca2+ uptake (Fig. 7E
).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. In the current study we further investigated the mechanism of these core protein effects. The results demonstrate that in isolated mitochondria, HCV core protein increases mitochondrial Ca2+ uptake via the mitochondrial Ca2+ uniporter (MCU); in intact cells this results in greater mitochondrial Ca2+ uptake in response to ER Ca2+ release, with increased mitochondrial ROS production and onset of MPT. Stimulation of the MCU is thus a previously unrecognized site at which viral proteins can alter cell function.
Effects of core on mitochondrial Ca2+ uptake
Mitochondrial Ca2+ influx is dependent on the activity of a voltage-driven permeability pathway called the mitochondrial Ca2+ uniporter (MCU) (30
, 31)
. This is a Ca2+-selective ion channel that utilizes the strongly negative matrix potential to drive concentrative Ca2+ uptake. Although it has yet to be molecularly identified, it can be inhibited selectively by ruthenium red or Ru-360. Another pathway for Ca2+ movement is through the MPT pore. This is a large protein complex localized at the contact site of inner and outer membranes. Opening of this relatively nonselective pore leads to an immediate depolarization of the mitochondria and a large conductive pathway for many small molecules and ions, including Ca2+. In living cells, cytosolic-free Ca2+ concentration is in the range of 100 nM while intramitochondrial Ca2+ is hundreds of times greater. Opening of MPT would therefore be a mechanism for Ca2+ exit from the mitochondria. In our isolated mitochondria experiments, however, the Ca2+ gradients are reversed and MPT would initially function as a Ca2+ influx pathway. In fact, we observed that this was indeed the case, but were able to eliminate this competing "nonphysiological" pathway for Ca2+ entry by using the MPT inhibitor CSA in the uptake solution.
In addition to the uptake step, steady-state mitochondrial Ca2+ might also depend on an export transporter such as Na+/Ca2+ exchanger (32)
. Our data show that core protein increases the initial rate of mitochondrial Ca2+ uptake as well as the steady-state level of intramitochondrial Ca2+ (Fig. 7)
. These effects were both due to changes in activity of the MCU since they were entirely inhibited by Ru360. This suggests that efflux is not required for the core effect. This is further substantiated by the observation that an inhibitor of the Na+/Ca2+ exchanger, CGP37157, had no effect. Furthermore, previous studies have demonstrated that core protein does not hyperpolarize mitochondria (9)
, and thus the increased Ca2+ uptake can only be explained by an increase in either the conductance or open probability for the MCU and not by a change in the driving force for Ca2+ entry. Recent work shows that the MCU is itself regulated by Ca2+ and appears to close at a particular Ca2+ set point, thus limiting mitochondrial Ca2+ uptake to a maximum level (31)
. The ability of core protein to increase both rate and steady-state mitochondrial Ca2+ suggests that it alters this regulatory function of the MCU.
The effect of core protein on the MCU was readily observed at a concentration (100 ng/mg mitochondrial protein) similar to that at which core protein caused oxidation of mitochondrial glutathione (9)
. While this was higher than the concentrations estimated in infected human liver (equivalent to 5 ng/mg mitochondrial protein), only a small percentage of hepatocytes express viral proteins, and thus the effective core concentration in a single cell would be much higher. Stimulation of MCU is specific for core protein and did not occur for two different control proteins: a positively charged histone of similar size and a small human protein synaptotagmin I C2A also derived from expression in E. coli.
Relationship between ER and Ca2+ effects
The mechanism by which core protein alters MCU function is unclear. The MCU mediates Ca2+ movement across the inner mitochondrial membrane, while core binds to the outer membrane and is not present in the intermembrane space or inner membrane (9
, 33)
. However, there is growing evidence of spatial proximity of the complexes responsible for ER Ca2+ release, mitochondrial Ca2+ uptake, and the mitochondrial permeability transition pore. The voltage-dependent anion channel (VDAC) is the major outer membrane component of the MPT pore; it binds to the adenine nucleotide translocator on the inner membrane and plays an important role in transporting Ca2+ across the outer membrane. The MCU inhibitors ruthenium red and Ru360 are known to specifically interact with the Ca2+ binding site in VDAC (34
, 35)
. It was recently demonstrated that a VDAC isoform on the mitochondrial outer membrane is able to form a complex with the IP3 receptor Ca2+ release channel of the ER (36)
, and the core protein has been localized to sites of contact between ER and mitochondria (12)
. This raises the possibility that core protein interacts with this complex, affecting both ER Ca2+ release and mitochondrial Ca2+ uptake.
The idea that core protein may alter events in the ER-mitochondrial axis is important since earlier work reveals ER stress in HCV infection. The HCV proteins NS5A, NS4B, and core have each been shown to produce an ER stress response characterized by activation of signaling pathways associated with the unfolded protein response and release of ER Ca2+ (37
38
39)
. Core protein’s direct effects on mitochondria would thus serve to increase mitochondrial Ca2+ uptake in response to ER Ca2+ release, precisely as we observed in this study (Fig. 2)
.
Although core protein consistently increased Ca2+ transfer from ER to mitochondria, the effects were greater for TG-induced release than for agonist-induced release. In the case of ATP-induced Ca2+ release, mitochondrial uptake was readily observed in the core-expressing cells but was quite small in the cells not expressing core protein (Fig. 2)
. This was a consistent observation and occurred for histamine and vasopressin as well (data not shown). The relatively small increase in mitochondrial Ca2+ after ATP may simply be due to inadequate sensitivity of the pericam measurement. The ATP-induced cytosolic Ca2+ spike was of smaller magnitude and shorter duration than that induced by TG, and thus less mitochondrial accumulation is expected.
This transfer of Ca2+ from ER to mitochondria is likely to explain some apparent differences between our work and that published by others. For example, Bergqvist et al. (18)
observed that core protein expression reduced the peak of cytosolic Ca2+ induced by TG in Jurkat cells. This observation was interpreted as a depletion of ER Ca2+ stores but could also have been due to greater uptake of released Ca2+ by mitochondria. Benali-Furet et al. (17)
directly measured ER Ca2+ with ER-targeted aequorins and showed that core protein reduced the rate of ER refilling on extracellular Ca2+ addition to Ca2+-depleted cells. While this might reflect an intrinsic decrease in ER Ca2+ transport, it could also have resulted from greater mitochondrial Ca2+ uptake, which would thus reduce the accumulation of Ca2+ in other compartments. In fact, we observed this phenomenon directly for ATP-induced Ca2+ release, where decreased cytosolic Ca2+ accumulation was accompanied by increased mitochondrial Ca2+ accumulation (Fig. 3)
. Thus it is apparent that changes in either ER or mitochondrial Ca2+ homeostasis can influence events in the other organelle.
Viral advantages of the mitochondrial effects
Viral modulation of mitochondrial function may serve several purposes. First, since mitochondria are sites of integration of pro- and antiapoptotic signals, viral effects may either prevent apoptosis or induce apoptosis, and this may serve the functions of preventing viral clearance or promoting dissemination at different stages of the viral life cycle. The recent understanding that a mitochondrial protein, MAVS, is a critical signaling molecule in the dsRNA-induced innate immune response is an example of how viral-induced cleavage of a mitochondrial protein promotes viral replication (7
, 40)
. The Ca2+ uniporter has not been recognized as a mitochondrial viral target, but its activation increases ROS production, changes the redox state of the cell, and thus may influence redox-dependent signaling processes that can promote the viral life cycle.
By increasing mitochondrial Ca2+ uptake, HCV may increase ATP production, ROS production, and alter apoptotic signaling (41
, 42)
. Mitochondria thus serve as effectors where increased ROS production affects other processes, as well as final targets where Ca2+ and ROS induce MPT (43
44
45)
. MPT can result in apoptosis of hepatocytes. Increasing frequency of hepatocellular apoptosis has been observed in patients with chronic hepatitis C, where it contributes both to necroinflammation and fibrosis development (46)
.
In addition to Ca2+ effects, core has other mechanisms by which it influences apoptosis. Some studies have observed both pro- and antiapoptotic effects of core. Inhibition of core protein expression increased proapoptotic Bax and decreased Bcl-xL proteins in immortalized human hepatocytes (47)
, while other studies demonstrated core-induced apoptosis through CHOP/GADD153 or TRAIL via a mitochondria signaling pathway (48)
. Understanding the overall apoptotic effects of the Ca2+ transport changes observed in this study therefore will require additional investigation. Nonetheless, the triangle of Ca2+, ROS, and MPT plays an important role in mitochondrial physiology, and our results suggest that specific disruption of this loop may play a key role in HCV-related liver diseases.
|
ACKNOWLEDGMENTS |
|---|
Received for publication October 10, 2006. Accepted for publication February 22, 2007.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH |
