| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Research Centre of Centre Hospitallier Université de Québec (CHUQ), Department of Anatomy and Physiology, Laval University, Laurier, Québec, Canada; and
Molecular Neuroscience Research Center, Shiga University of Medical Science, Shiga, Japan
1Correspondence: Research Centre of CHUQ, 2705 Boul. Laurier, Québec, QC, G1V4G2, Canada. E-mail: jean-pierre.julien{at}crchul.ulaval.ca
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: digitonin • microsome • live imaging • green fluorescent protein • electron immunomicroscopy
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2
3
4
5)
. Initial studies focused on aberrant copper-mediated catalysis as potential source of toxicity (6
, 7)
, but subsequent mouse studies provided compelling evidence for a gain of a new toxic function that is independent of the enzymatic activity involving the copper catalytic site. Many lines of evidence now suggest that the toxic properties of mutant SOD1 species might be related to propensity to aggregate (8
9
10
11
12
13)
. However, the exact mechanism of disease remains unknown.
Recent studies demonstrated that a fraction of SOD1 can be secreted via the ER-Golgi network (14
15
16)
and that chromogranins, which are abundant proteins in motor neurons and interneurons, may act as chaperone-like proteins to promote secretion of misfolded SOD1 mutants. Moreover, extracellular mutant SOD1 can induce microgliosis and motor neuron death (16)
. Such pathogenic mechanism in ALS based on toxicity of secreted SOD1 mutant are in line with findings that the disease is not strictly autonomous to motor neurons and that toxicity can propagate from one cell to another. Indeed, extracellular mutant SOD1 is being considered as a target for development of therapeutic immunization (17)
.
SOD1 is mainly a cytosolic protein, but it may also distribute to nucleus (18)
, mitochondria (18
19
20)
, and lysosome (18
, 20
21
22)
. The presence of SOD1 in microsomes remains problematic. A previous study did not detect SOD1 in ER or Golgi by electron microscopy (18)
, and evidence indicates that SOD1 can be eliminated from microsome fractions by high-salt wash (20)
. This raises the possibility of SOD1 contaminating the microsomal fractions through subfractionation procedures (20)
. Moreover, it is unclear how SOD1, a major cytosolic protein, can be targeted to the ER-Golgi network (14
15
16)
. Proper folding and dimerization of SOD1 requires the formation of intramolecular disulfide bonds between Cys-57 and Cys-146. The monomeric form of SOD1 does exist in a reducing environment or in apo state (23
, 24)
. Moreover, it has been reported that a significant amount of SOD1 exists in reduced form in tissue lysates from transgenic mice expressing human WT or mutant SOD1 (25)
. The level of reduced SOD1 would favor conversion of a soluble form to an aggregated form (12
, 26)
. Since SOD1 does not contain any translocation signal, the folding state based on such redox or apo/holo state might be involved in the subcellular translocation from cytosol.
Here, we have used live cell imaging of EGFP-fused SOD1 and membrane permeabilization with digitonin to demonstrate the translocation of SOD1 species into vesicles of the ER-Golgi system. Moreover, subcellular fractionation procedures combined with chemical cross-linking experiments revealed that microsomes are the predominant cellular site of aggregation of mutant SOD1. Finally, evidence is presented that the monomeric SOD1 is a molecular form that can be translocated into microsomes in the presence of ATP.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
was fused to pEGFP-N2 (BD Bioscience, Palo Alto, CA, USA) at BamHI/EcoRI sites (pEGFP-SOD1). pDsRed-Golgi, in which Golgi-translocation signal from human beta-1,4-galactosyltransferase was fused with DsRed was a generous gift from Y. Imai (RIKEN Brain Science Institute, Tokyo, Japan) (16)
. The construct was verified by sequencing and Western blotting using anti-human SOD1 antibody. Recombinant protein of human SOD1 [wild-type (WT) and Gly93Ala mutant (G93A)] and mouse chromogranin A were produced as described elsewhere (16
, 27)
. All chemicals were purchased from Sigma (St. Louis, MO, USA).
Antibodies
Rabbit polyclonal anti-human SOD1 and rabbit polyclonal anti-calnexin antibodies were purchased from StressGen (Victoria, BC, Canada). Mouse monoclonal anti-COX4 antibody was purchased from Invitrogen (Carlsbad, CA, USA). Rat monoclonal anti-Lamp2 was purchased from BD-Bioscience (San Diego, CA, USA). Rabbit polyclonal anti-chromogranin A antibody (H300) was purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Rabbit polyclonal anti-bovine serum albumin antibody was obtained from Calbiochem (San Diego, CA, USA). Rabbit polyclonal anti-TGN38 antibody was prepared as previously reported (16)
.
Live cell imaging with confocal laser microscope
The mouse neuroblastoma neuro2a cells (American Tissue Culture Collection, Manassas, VA, USA) plated onto glass-bottom chamber slides (Lab-Tek, Nalge Nunc, Rochester, NY, USA) were transfected with pEGFP-SOD1 (WT, G85R, or G93A) using Lipofectamine 2000 (Invitrogen). At 24 h after transfection, cells were treated with various fluorescent marker dyes for subcellular organelles (Mito-tracker, Lyso-tracker or ER-tracker; Invitrogen) for 30 min. To study the localization in Golgi network, neuro2a cells were cotransfected with pEGFP-SOD1 and pDsRed-Golgi (16)
. The medium was replaced by phenol red-free DMEM containing 10% FBS, and the cells were observed under confocal laser microscope (LSM510 META; Carl Zeiss, Oberkochen, Germany). Image was obtained using the image analysis software equipped with the microscope.
Plasma membrane perforation by digitonin
Transfected cells were incubated with digitonin to wash out cytosolic protein according to the protocol previously reported (28)
. At 24 h after transfection, cells were washed in the ice-cold PBS(–) twice and incubated in EDTA-ATP buffer (25 mM HEPES, pH 7.4; 4 mM EDTA; 1 mM ATP; 50 mM KCl; 0.25 M sucrose) with or without 40 µM digitonin for 10 min on ice. Immediately after washing in ice-cold EDTA-ATP buffer twice, cells were observed with the confocal laser microscope at room temperature.
Transgenic mice
Wild-type human SOD1 mice (C57BL/6-TgN[SOD1]3Cje, hSOD1WT) were purchased from the Jackson Laboratory (Bar Harbor, ME, USA). Transgenic mice carrying G37R SOD1 (line 29) were a kind gift from D. Cleveland (University of California, San Diego, La Jolla, CA, USA) and were housed and bred with C57Bl/6. Mice were anesthetized and perfused in saline before spinal cords were dissected. Experimental protocols were approved by the Comité de Protection des Animaux de l'Université Laval according to Canadian Council on Animal Care guidelines.
Electron immunomicroscopy
Subcellular localization of human SOD1 was investigated using ultrathin slices from the spinal cord of SOD1 transgenic mice (8 months old) under an electron microscope, as described previously (16)
. In brief, mice were perfused with 4% parafolmaldehyde/0.5% glutaraldehyde. LR White resin-embedded tissues were sliced with a diamond knife (Reichert-Jung Ultracut E ultramicrotome; Leica, Wetzlar, Germany) and were placed onto nickel grids. Samples were reacted with rabbit polyclonal anti-human SOD1 antibody and were labeled with immunogold-conjugated secondary antibody (10 nm particles). Grids were analyzed under an electron microscope (Tecnai 12; FEI Co., Hillsboro, OR, USA).
Subcellular fractionation of spinal cord lysates from human SOD1 transgenic mice
The experiments described below, including sucrose density-gradient ultracentrifugation, floating ultracentrifugation, and chemical cross-linking experiments, were performed using subcellularly fractionated lysates of the spinal cord from human SOD1 transgenic mice according to the protocol described elsewhere (16)
. Briefly, after eliminating the pellet containing nuclei and debris from the homogenates in the homogenization buffer (50 mM HEPES, pH 7.4; 0.25 M sucrose; 1 mM magnesium acetate; protease inhibitor cocktail) by centrifuging at 1000 g for 10 min, the resulting supernatant (postnuclear fraction) was sequentially centrifuged at 9000 g for 15 min to separate a pellet (mitochondrial fraction) and supernatant that was further centrifuged at 100,000 g for 1 h at 4°C to obtain microsome (pellet) and cytosolic (supernatant) fractions.
Sucrose density-gradient ultracentrifugation
Postnuclear membrane components from the spinal cord of 8-month-old transgenic mice (WT and G37R SOD1) were subjected to sucrose density-gradient ultracentrifugation according to the established protocol (29)
. Briefly, the pellet obtained after ultracentrifugation of postnuclear lysates at 100,000 g for 1 h was washed twice and resuspended in homogenization buffer. Discontinuous sucrose density gradient was prepared with 0.3 ml of 2 M, 1.3 ml of 1.3 M, 1.2 ml of 1 M, and 0.7 ml of 0.6 M sucrose, finally overlaid with the 0.5 ml of the sample in the homogenization buffer. The samples were centrifuged in a swing rotor (SW60.1, Beckman-Coulter, Fullerton, CA, USA) at 100,000 g for 4 h at 4°C. Every 300 µl of fraction was taken from top to bottom, and 30 µl from each fraction was analyzed by Western blotting.
Floating ultracentrifugation of spinal cord lysates
To avoid the contamination of cytosolic protein in microsomes, floating ultracentrifugation of the spinal cord lysates from human SOD1 transgenic mice (WT and G37R, 8 months old for both) was performed based on the previous method (30)
. First, a postmitochondrial fraction was obtained from the supernatant after 15 min centrifugation of postnuclear lysates at 9000 g. In several experiments, the lysates were incubated in 0.5 M NaCl for 30 min at 4°C for high-salt wash or with 0.3 mM glycyl-L-phenylalanine-beta-naphthylamide (GPN), a lytic agent of lysosomes (31)
for 10 min at 37°C to disrupt lysosome. Subsequently, the lysates were mixed with 60% iodoxanol (Sigma), resulting in 30% solution. The tissue lysates (500 µl) were overlaid on 3.5 ml of 35% iodoxanol, and 200 µl of 5% iodoxanol solution was placed on the top. After ultracentrifugation for 2 h at 200,000 g at 4°C using an SW60.1 rotor, vesicles and cytosolic proteins were collected from top and bottom, respectively.
Chemical cross-linking of spinal cord lysates
To further investigate protein complexes and the monomer/dimer state of SOD1 in the mouse spinal cord, lysates from subcellularly fractionated spinal cord tissues were incubated with 2 mM disuccinimidyl suberate (DSS, Sigma) for 1 h at 37 or 22°C. Reaction was terminated by excess amount of Tris (100 mM Tris-HCl, pH 7.4; at final concentration) and incubated in sodium dodecyl sulfate (SDS) sampling buffer containing 2-mercaptethanol for 5 min at 95°C prior to Western blot analysis.
Western blotting
Protein concentration of tissue lysates was determined by the Bradford method (Bio-Rad, Hercules, CA, USA). Protein denatured by SDS sampling buffer containing 6% 2-mercaptoethanol for 5 min at 95°C was separated by SDS-polyacrylamide gel electrophoresis, followed by transfer onto polyvinylidene difluoride (PVDF) membrane (Perkin-Elmer, Wellesley, MA, USA). Target proteins labeled by primary and peroxidase-conjugated secondary antibody were visualized by chemoilluminescent kit (Perkin-Elmer).
Cell-free microsome binding assay
Translocation of human SOD1 into microsomes was analyzed using well-characterized rat liver microsome fractions (Cellzdirect, Pittsboro, NC, USA) and recombinant SOD1 protein as previously reported (32)
, with minor modification. Briefly, 5 µg of bacterially purified apo-state or metallated human SOD1 protein was incubated with purified rat liver microsomes (15 µg) with or without 5 mM ATP for 20 min at 37°C in 50 µl of reaction buffer (25 mM HEPES, pH 7.4; 50 mM potassium acetate; 2.5 mM magnesium acetate; and 10 mM DTT) in a siliconized Eppendorf tube (Fisher Scientific, Ottawa, ON, Canada). Subsequently, the mixture was treated with proteinase K (50 µg/ml) with or without 1% Triton X-100 for 2 h at 37°C. The reaction was terminated by protease inhibitor cocktail (Roche, Basel, Switzerland) followed by SDS sampling buffer with 95°C incubation for 5 min. In certain experiments, 0.3 mM GPN was included in the reaction mixture to disrupt lysosome during translocation experiments.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
A–D), unlike the previous report about the detection of SOD1 in lysosomes (33)
. The WT SOD1-EGFP was rarely observed in mitochondria (Fig. 1B
) but was occasionally detected in ER or Golgi (Fig. 1C, D
). We noted that cells expressing G93A SOD1-EGFP exhibited more aggregates or vesicular patterns. The G93A SOD1-EGFP showed more frequent localization in ER and Golgi than WT SOD1 (Fig. 1G, H
). Lysosomal or mitochondrial localization was also rare in the G93A mutant (Fig. 1E, F
).
|
Detection of SOD1 into Golgi vesicles after plasma membrane permeabilization with digitonin
Live cell imaging revealed a distribution of EGFP-SOD1 species into ER and Golgi, but a masking effect was due to the abundance of SOD1 in cytosol. To overcome this problem, transfected neuro2A cells with pEGFP-SOD1 were treated with digitonin to permeabilize the plasma membrane and to release cytosolic SOD1. Confocal laser microscope analysis showed that digitonin effectively washed out cytosolic EGFP (Fig. 2
A) and cytosolic EGFP-fused SOD1 proteins (Fig. 2B-D
for WT, G85R, and G93A, respectively). After digitonin treatment, both EGFP and EGFP-WT SOD1 yielded a fluorescent pattern mainly along plasma membranes. In contrast, EGFP-mutant SOD1 (Fig. 2C, D
for G85R and G93A, respectively) remained abundant in vesicular or aggregate-like structures (Fig. 2
, arrowheads).
|
Next, we examined by confocal laser microscope the subcellular localization of these EGFP-SOD1 remnants after digitonin treatment using the various trackers noted previously. Preliminary experiments revealed that only Mito-tracker and DsRed-Golgi remained detectable after digitonin treatment. The signals for other organelle markers were lost during digitonin treatment (data not shown). After cotransfection of neuro2a cells, the signals for EGFP-WT and EGFP-mutant SOD1 after digitonin treatment colocalized predominantly with DsRed-Golgi (Fig. 3
A, B). The mutant SOD in Golgi appeared in larger vesicles or aggregates than WT SOD1 (Fig. 3
, arrowheads). Unexpectedly, we found poor localization of EGFP-SOD1 species with Mito-tracker after digitonin treatment (Fig. 3C, D
), implying that mitochondria are not a major target for EGFP-SOD1 species. We also confirmed the SOD1 localization in Golgi using fluorescent microscopy (Supplemental Fig. 1).
|
Clusters of immunogold-labeled mutant SOD1 in ER detected by electron microscopy
We performed electron immunomicroscopy using ultrathin sections prepared from spinal cords of hSOD1 transgenic mice (WT and G37R, 8 months old). Immunogold-labeled SOD1 particles were detected in cytosol, mitochondria, and rough ER. Clusters of immunogold particles in rough ER were detected in samples from the G37R mutant SOD1 mice, whereas most of the particles in samples from WT SOD1 transgenic mice were singlet or doublets (Fig. 4
A, B). This result is consistent with localization and aggregation of mutant SOD1 in ER.
|
Distribution of SOD1 in ER-Golgi vesicles after subfractionation of spinal cord lysates
The presence of SOD1 in microsomes has been a matter of debate. Subcellular fraction analysis showed that both WT and mutant SOD1 can distribute in the microsomal fraction (15
, 16)
. Kikuchi et al. (15)
demonstrated by proteinase K protection assay that the G85R mutant localizes in microsomes intraluminally in the spinal cord lysates from mutant SOD1 transgenic mice. However, it has been argued that this could reflect contaminated cytosolic SOD1 binding externally to the microsomal membrane, which was shown to be released by high-salt wash (20)
. Moreover, it is difficult to eliminate cytosolic components from the microsome pellet by conventional ultracentrifugation.
Therefore, we attempted to clarify the distribution of SOD1 in luminal structures by more compelling methods, using sucrose density gradient ultracentrifugation and iodoxanol floating ultracentrifugation techniques. The postnuclear fraction from spinal cord lysates of presymptomatic G37R SOD1 mice and of age-matched WT SOD1 transgenic mice was recovered and further separated through discontinuous sucrose gradient from 0.6 to 2 M. As shown in Fig. 5
A, samples from both WT and G37R SOD1 distributed in fractions 2 to 8 and covered a broader density gradient than the mitochondrial fractions (fractions 5–8, with the peak at 7). TGN38, a marker of trans-Golgi network, was detected in fractions 3 to 6, with a peak at 5. Calnexin, an ER marker, distributed more broadly in fractions 2 to 11 (most apparently from 3 to 9) and covered the SOD1 distribution. The lysosome marker lamp2 was detected in fractions 5–7 (while smearing band was detected in lane 5). Thus, both WT and G37R SOD1 distributed abundantly in fractions 3 and 4, which are enriched in ER-Golgi vesicles. Nonetheless, as expected, the majority of SOD1 was recovered in the cytosolic protein fraction after centrifugation of the postmitochondrial supernatant at 100,000 g for 1 h to obtain the microsomal (pellet) and cytosolic fractions (data not shown).
|
To confirm further the intraluminal existence of SOD1, we utilized a floating ultracentrifugation technique using an iodoxanol cushion to efficiently separate vesicles and membranes from cytosol. The postmitochondrial fractions recovered from supernatant after centrifugation at 9000 g for 15 min were placed on the iodoxanol cushion and ultracentrifuged at 200,000 g for 2 h. The resulting vesicles and cytosol were separated on top and at bottom, respectively. Moreover, we treated the lysates with 0.5 M NaCl before ultracentrifugation to release SOD1 that might stick to the outer membrane, or with 0.3 mM GPN, a lytic agent specific for lysosomes. As shown in Fig. 5B
(left), both WT and G37R mutant SOD1 were detected in vesicles and high-salt wash did not alter the localization of SOD1 in this fraction, strongly indicating that SOD1 exists inside microsome vesicles. Furthermore, pretreatment with GPN did not change the amount of SOD1 in the vesicular fractions, excluding the possibility that majority of intraluminal SOD1 in the vesicular fraction came from lysosomes (Fig. 5B
, right). From the combined results, we conclude that both WT and mutant SOD1 can be translocated into the ER-Golgi pathway.
Formation of SOD1 aggregates within microsomes
Previous studies showed that mutant SOD1 forms high molecular weight species in the spinal cord lysates from transgenic mice in an age-dependent manner (15
, 16)
. Evidence indicates that the misfolding of monomeric SOD1 is an intermediate in the aggregation pathway (25
, 27
, 34)
. To assess the aggregation of SOD1 species, we utilized the chemical cross-linker DSS, which covalently binds molecules facing at a distance shorter than 11.4 Å (35)
. After the DSS cross-linking reaction using spinal cord lysates, it was possible to assess the monomer/dimer states of SOD1 through denaturing SDS-PAGE and immunoblotting. When each subcellular fraction from spinal cord lysates of WT or G37R SOD1 mice was incubated with DSS for 1 h at 37°C, cross-linked SOD1 dimers were clearly detected (Fig. 5C
, blank arrowhead). Note the multiple bands at the expected dimer size corresponding to human/human, human/mouse, and mouse/mouse SOD1 dimers, or to a mixture of oxidized and reduced forms of SOD1 (25)
. Actually, the majority of the SOD1 species were recovered in the monomeric form after SDS-PAGE, even after treatment with DSS at different temperatures (Fig. 5C
, black arrowheads; top panel, 37°C; bottom panel, 22°C) and despite the fact that the dimer interface of SOD1 is 2.8 Å, which is far shorter than 11.4 Å, the limit for cross-linking by DSS (36)
. Of particular interest is the finding that the DSS treatment produced high molecular weight smears in the mitochondrial and microsome fractions derived specifically from the mutant SOD1 mice (vertical lines). Such high molecular weight complexes were not detected in the cytosolic fractions even at the end-stage of disease in G37R mice. The high molecular weight complexes were the most prominent in the microsomal fraction.
ATP-dependent translocation of apo/monomeric SOD1 in cell-free microsomes
It is well-known that SOD1 dimer and metallated SOD1 are resistant to proteinase K. However, monomeric or apo-state SOD1 can be digested by proteinase K completely (Fig. 6
A). To test the hypothesis that monomer/apo SOD1 can be translocated into microsomes, as was reported for the mitochondrial translocation of apo SOD1 (37)
, bacterially purified recombinant human SOD1 protein (WT, G93A) of apo state (5 µg) was incubated with rat liver microsome for 20 min at 37°C with or without ATP or cytosolic fraction from COS7 cells (S100). Thereafter, reaction mixtures were digested by proteinase K. Western blot analysis revealed the presence of recombinant SOD1 in microsomes even after proteinase K treatment (Fig. 6B
, lane 2). The amount of protected recombinant SOD1 increased in the presence of ATP (Fig. 6B
, lane 4). ATP itself does not affect proteinase K digestion (Fig. 6C
). However, recombinant SOD1 proteins were completely digested by proteinase K in presence of Triton-X100 (Fig. 6B
, lanes 3 and 4). Cytosolic proteins in S100 lysates had no effect on this translocation (Fig. 6B
, lane 6), indicating that cytosolic protein is not required for SOD1 translocation into microsomes. We further investigated whether SOD1 uptake was due to contaminated lysosomes, since a previous paper indicated that purified lysosomes might incorporate some proteins in an ATP-dependent manner (38)
. Thus, we included 0.3 mM GPN in the reaction mixture during translocation time. The result clearly indicated that chemical disruption of lysosomes from the microsome fraction did not significantly affect the uptake of either endogenous or recombinant SOD1 (Fig. 6E
). It is noteworthy that treatment of rat liver microsomes with proteinase K did not decrease the rat endogenous SOD1, which was completely digested in the presence of Triton X-100, and that ATP treatment rendered endogenous SOD1 resistant against proteinase K (Fig. 6B
, asterisk). These findings indicate the presence of endogenous SOD1 in ER-Golgi vesicles under physiological conditions. It is unclear why the amount of endogenous rat SOD1 in microsomes increased in the presence of exogenous recombinant SOD1 (Fig. 6B, D, E
). We also confirmed that protein translocation to microsomes is not a general finding, because recombinant mouse chromogranin A lacking a signal peptide or BSA (signal peptide is also cleaved) was not detected in microsomes after proteinase K digestion (Fig. 6D
).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
or the exosomes (40)
. An explanation for this phenomenon is provided here. From our in vitro translocation assay, recombinant apo/monomeric SOD1 can be incorporated into rat liver microsomes in presence of ATP (Fig. 6)
. No microsomal uptake of other proteins such as BSA and signal peptide-deleted chromogranin A was found. It is plausible that exposure of hydrophobic sequences from monomeric SOD1 contributes to translocation into microsomes. For example, it has been reported that a hydrophobic sequence plays a key role in uptake into secretory pathway of FGF16, a protein lacking signal peptide for ER translocation (41)
. A similar phenomenon may underlie the translocation of monomeric SOD1 into luminal structures. In addition, the apo/monomeric form of both WT and mutant SOD1 is recognized by Hsp/Hsc70 (27)
and is implicated in aggregate formation under oxidative stress (34)
.
Live cell imaging revealed a very poor distribution of mutant EGFP-SOD1 in lysosomes or mitochondria. It is possible that our experimental protocol underestimated the amount of lysosomal SOD1, because the Lyso-tracker used in this experiment is a pH indicator fragile to environmental change, and its labeling efficiency is not so high. The reason for poor localization of mutant EGFP-SOD1 in mitochondria remains unclear. To date, mitochondria have been regarded as a major target of mutant SOD1 toxicity (42
, 43)
. So, we cannot exclude the possibility that the fusion to EGFP conferred new properties to mutant SOD1. Consistent with our finding, a recent paper by Bergemalm et al. (44)
showed that only very low levels of unstable SOD1 mutants such as G85R and G127X were detectable in mitochondria prepared from the mouse models. It is has been argued that the loading of the stable SOD1 mutant forms (G93A, G37R, G90A) into mitochondria is a nonphysiological event resulting from overexpression of human SOD1 species in 20-fold excess of endogenous mouse SOD1.
Once incorporated into ER, both WT and mutant SOD1 species can form soluble dimers, as revealed by the DSS cross-linking experiment performed with microsomes (Fig. 5C
). However, unlike WT SOD1, it appears that mutant SOD1 has a propensity to aggregate into high molecular weight species after translocation into luminal structures (Fig. 5C
). It is noteworthy that our cross-linking studies revealed the microsomal fraction as being a major compartment for the formation of mutant SOD1 aggregates. By comparison, much lower levels of mutant SOD1 aggregates were detected in cytosolic or mitochondrial fractions. The microsomal milieu may favor the folding state of mutant SOD1 (45)
and protein aggregation. High molecular weight species of mutant SOD1 were even detected in microsomes at 12 months of age in the absence of DSS treatment (Fig. 4C
). These aggregates occurred even in presence of 2-mercaptoethanol, indicating involvement of covalent links other than disulfide bonds. The smears on SDS-PAGE suggest that mutant SOD1 may aggregate with various proteins in microsomes. Substantial amounts of monomeric SOD1 may also exist in human SOD1 transgenic mice. A reduced form of SOD1 has been detected in transgenic mice (25)
, and monomeric SOD1 species have been detected in motor neurons of ALS mice using antibodies against an epitope at the dimer interface (46)
.
Although a consensus exists for the involvement of misfolding and aggregation of mutant SOD1 species in pathogenesis, the identity of noxious mutant SOD1 species and the precise mechanism of disease remains unknown. The finding that the ER-Golgi pathway is a predominant site of uptake and of age-dependent aggregation of mutant SOD1 suggests that secretion may play a key role in disease. Further investigation is required to clarify the nature of the proteins linked to mutant SOD1 by DSS treatment and their potential role in the disease process. Deleterious effects could arise from ER stress (15)
as well as secretion of noxious mutant SOD1 species (16)
. In addition, a pathogenic model of ALS based on secretion defects or release of toxic components is appealing because it would be consistent with the notion that the disease is not strictly autonomous to motor neurons (47)
and with the therapeutic effects of immunization aiming to reduce extracellular mutant SOD1 (17)
.
|
ACKNOWLEDGMENTS |
|---|
Received for publication July 11, 2007. Accepted for publication February 21, 2008.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
