sm-20: a novel mitochondrial protein that causes caspase ... · pdf filekeywords: apoptosis,...

SM-20: A Novel Mitochondrial Protein That Causes Caspase-Dependent

Cell Death in Nerve Growth Factor-Dependent Neurons

Running title: SM-20: a novel death-inducing mitochondrial protein

Elizabeth A. Lipscomb1, Patrick D. Sarmiere2, and Robert S. Freeman2

1Department of Environmental Medicine and 2Department of Pharmacology and PhysiologyUniversity of Rochester School of Medicine and Dentistry

Rochester, NY 14642

Corresponding author: Robert S. Freeman, Ph.D.Department of Pharmacology and PhysiologyUniversity of Rochester, School of Medicine and Dentistry601 Elmwood AvenueRochester, NY 14642Phone: (716) 273-4893Fax: (716) 273-2652Email: [email protected]

M0:08407 SM-20: a novel death-inducing mitochondrial protein

1

Copyright 2000 by The American Society for Biochemistry and Molecular Biology, Inc.

JBC Papers in Press. Published on November 1, 2000 as Manuscript M008407200 by guest on M

ay 17, 2018http://w

ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

Keywords: apoptosis, nerve growth factor, egl-9, mitochondria, sympathetic neuron, caspase, cytochrome c

SM-20: a novel death-inducing mitochondrial protein

2

by guest on May 17, 2018

http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

SUMMARY

Sympathetic neurons undergo protein synthesis-dependent apoptosis when deprived of

nerve growth factor (NGF). Expression of SM-20 is upregulated in NGF-deprived sympathetic

neurons and ectopic SM-20 is sufficient to promote neuronal death in the presence of NGF. We

now report that SM-20 is a mitochondrial protein that promotes cell death through a caspase-

dependent mechanism. SM-20 immunofluorescence was present in the cytoplasm in a punctate

pattern that co-localized with cytochrome oxidase I and with mitochondria-selective dyes.

Analysis of SM-20/dihydrofolate reductase fusion proteins revealed that the first 25 amino acids

of SM-20 contain a functional mitochondrial targeting sequence. An amino-terminal truncated

form of SM-20 was not restricted to mitochondria but instead localized throughout the cytosol

and nucleus. Nevertheless, the truncated SM-20 retained the ability to induce neuronal death,

similar to the wild type protein. SM-20 induced death was accompanied by caspase-3

activation and was blocked by a general caspase inhibitor. Additionally, overexpression of SM-

20, under conditions where cell death is blocked by a general caspase inhibitor, did not result in

widespread release of cytochrome c from mitochondria. These results indicate that SM-20 is a

novel mitochondrial protein that may be an important mediator of neurotrophin-withdrawal

mediated cell death.


3


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

INTRODUCTION

Sympathetic neurons undergo apoptosis when deprived of their survival factor, nerve

growth factor (NGF)1. Inhibitors of RNA or protein synthesis block this death, suggesting that

gene expression is important for apoptosis in these cells (1, 2). Consistent with this idea, a small

number of genes have been shown to increase in expression in neurons after NGF withdrawal (3-5).

Among these, c-jun and cyclin D1 have been implicated as upstream regulators of neuronal

death (4-9). Recently, we found that expression of the SM-20 gene is also upregulated in NGF-

deprived sympathetic neurons (10). SM-20 expression is also induced in sympathetic neurons

treated with the anti-tumor agent cytosine arabinoside or with inhibitors of phosphatidylinositol

3-kinase, agents that cause apoptosis in the presence of NGF (11, 12). Although a requirement for SM-

20 in cell death remains to be established, overexpression of SM-20 is sufficient to induce cell

death in neurons maintained in the continual presence of NGF (10). These findings suggest that SM-

20 may function as a regulator of neuronal death.

Regulation of SM-20 expression also occurs in muscle cells and in response to activation

of the p53 protein in fibroblasts. SM-20 mRNA levels increase in vascular smooth muscle cells

stimulated with growth factors and are downregulated when muscle cells are induced to

differentiate (13, 14). In vivo, SM-20 mRNA levels are elevated in muscle cells in response to injury to

the blood vessel wall (15). SM-20 expression also increases following activation of a temperature-

sensitive p53 protein in rat embryo fibroblasts (16). Activation of p53 in these cells results in both

growth arrest and apoptosis. Interestingly, stable expression of SM-20 in tumor cells lacking

functional p53 resulted in greatly reduced numbers of colonies in colony formation assays,

suggesting that SM-20 might act downstream of p53 in either growth-arrest or apoptosis (16).


4


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Although the biochemical function of the SM-20 protein is unknown, it contains a 218-

amino acid region that is closely related to sequences in the Caenorhabditis elegans Egl-9

protein (17). Mutations in the egl-9 gene had been originally shown to disrupt normal egg laying in

nematodes (18). More recently, egl-9 was implicated as a mediator of toxin-induced neuromuscular

paralysis in C. elegans infected with the pathogenic bacterium Pseudomonas aeruginosa (17).

To help elucidate the function of SM-20, its subcellular localization was investigated.

We report that the amino terminus of SM-20 contains a mitochondrial targeting sequence that

localizes SM-20 to mitochondria. Because mitochondria have a prominent role in the regulation

of cell death (19), we tested the significance of the mitochondrial targeting of SM-20 on its ability to

promote neuronal death. Our results indicate that targeting of SM-20 to mitochondria is not

required for SM-20 induced death. In addition, we find that caspase-3 activation occurs during

SM-20 induced cell death. In contrast to Bax, expression of SM-20 in neurons maintained in

the presence of a general caspase inhibitor did not lead to detectable cytochrome c release from

mitochondria. These results indicate that SM-20 is a novel mitochondrial protein that may be an

important mediator of neurotrophin-withdrawal mediated cell death.


5


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

EXPERIMENTAL PROCEDURES

Cell Culture. CV-1 (ATCC; Rockville, MD) and NIH3T3 cells were maintained in DMEM (Gibco-BRL;

Gaithersburg, MD) supplemented with 10% fetal bovine serum (Hyclone; Indianapolis, IN), 2 mM L-glutamine,

100 U/ml penicillin, and 100 µg/ml streptomycin (all from Gibco-BRL). Primary cultures of

sympathetic neurons were prepared from the superior cervical ganglia of embryonic day 21 rats

as described previously (12). Cultures were maintained in vitro for 5-6 d in media consisting of 90%

MEM, 10% fetal bovine serum (Harlan Bioproducts; Madison, WI), 2 mM L-glutamine, 20 µM

uridine, 20 µM fluorodeoxyuridine, 100 U/ml penicillin, 100 µg/ml streptomycin, and 50 ng/ml

NGF (Harlan Bioproducts).

Expression Vectors and Transient Transfections. The expression vectors containing the

SM-20 open reading frame in Rc-CMV (Invitrogen; Carlsbad, CA) and the SM-20/green

fluorescent protein (GFP) cDNA in pcDNA3 (Invitrogen) were described previously (10). For

expressing T7-epitope tagged forms of SM-20, dihydrofolate reductase (DHFR), SM-

20(60–355), and SM-20/DHFR fusions, complementary oligonucleotides encoding the T7 epitope tag

(Novagen; Madison, WI) flanked by restriction sites for EcoRI–NdeI (5’) and EcoRV (3’) were

annealed and ligated into Bluescript KS+ (Stratagene; La Jolla, CA) between the EcoRI and

EcoRV sites. The various cDNAs were inserted into the T7-modified Bluescript vector upstream of

the T7 epitope using EcoRI and NdeI. The SM-20 sequences used to make SM-20(60-355),

SM-20(1-25)/DHFR, and SM-20(1-50)/DHFR encoded an amino-terminal methionine and

were generated by polymerase chain reaction using Pfu polymerase (Stratagene). The EcoRI to

EcoRV fragments containing each cDNA fused at its 3’ end to the T7 epitope were then subcloned


6


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

into pcDNA3. For SM-20(60-355)/GFP, a DNA fragment consisting of an amino-terminal

methionine followed by amino acids 60-355 of SM-20 was amplified using Pfu polymerase,

fused to the 5’ end of GFP, and inserted into pcDNA3 between the EcoRI and EcoRV sites. All

amplified DNAs were sequenced to confirm that they were correct.

CV-1 cells (2 x 105 cells) were plated onto glass coverslips in 35-mm plastic tissue

culture dishes 24 h prior to transfection with Lipofectamine (Gibco-BRL) as described by the

manufacturer. NIH3T3 cells (1 x 107 cells) were plated onto 100-mm tissue culture dishes and

transfected with Lipofectamine. Cells were used for experiments on the second day after

transfection.

Mitochondrial Labeling. CV-1 cells were incubated for 5 h in media containing 1 µM

MitoTracker Green FM (Molecular Probes; Eugene, OR), a mitochondria-selective dye that is

retained in fixed cells. To establish the specificity of this dye for labeling mitochondria, cells

were plated on gridded cover slips and then stained with Mitotracker Green FM as described

above. After staining, the cells were visualized using a fluorescein isothiocyanate (FITC) filter

set (λex=450–490 nm, λem=520–560 nm) and images were captured. The same cells were then

stained with the non-fixable mitochondria-specific dye Rhodamine 123 (20) and images were

obtained using a tetramethylrhodamine isothiocyanate (TRITC) filter set (λex=546±10 nm,

λem>590 nm). In every cell, the fluorescence of Mitotracker Green FM and Rhodamine 123 coincided

completely. For labeling of mitochondria in sympathetic neurons, Mitotracker Orange (25 nM,

Molecular Probes) was added to the culture medium for 15 min after which images of live cells


7


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

were captured using a Leica confocal microscope and TCS-NT software.

Immunofluorescence. CV-1 cells were fixed and permeabilized with 4%

paraformaldehyde containing 0.2% Triton X-100 in phosphate-buffered saline (PBS) for 10

min, and then incubated in block buffer consisting of 3% goat serum and 0.2% bovine serum

albumin in PBS. To detect SM-20, DHFR, or SM-20/DHFR proteins, cells were incubated with

either a 1:200 dilution of rabbit polyclonal anti-SM-20 antibody (10) or 0.2 µg/ml anti-T7

monoclonal antibody (Novagen) in block buffer. Coverslips were then rinsed with PBS and

incubated with TRITC-conjugated goat anti-rabbit or goat anti-mouse secondary antibodies

(Jackson ImmunoResearch; West Grove, PA) at 15 µg/ml in block buffer. The cells were

counterstained with Hoechst 33,258 (Molecular Probes) and then mounted onto slides for

viewing by epi-fluorescence with a Nikon Diaphot 300 microscope. Digital images were

captured using a Dage-MTI CCD camera and Scion Image software.

For dual immunofluorescence, CV-1 cells were first incubated with IgG-purified anti-

SM-20 primary antibody and TRITC-conjugated goat anti-rabbit secondary antibody (Jackson

ImmunoResearch) as described above. After extensive washing of the coverslips, antibodies to

cytochrome oxidase I (#A-6403; Molecular Probes) or cytochrome oxidase VI (#A-6401;

Molecular Probes) were added at 1 µg/ml or 4 µg/ml in block buffer, respectively, followed by

FITC-conjugated goat anti-mouse IgG at 15 µg/ml in block buffer. Secondary antibodies used

in the double labeling experiments were cross absorbed against mouse or rabbit IgG.

For visualization of active caspase-3 and cytochrome c, sympathetic neurons were fixed

with 4% paraformaldehyde in PBS for 30 min at 4oC and then incubated in block buffer

consisting of 5% goat serum and 0.3% Triton X-100. Primary antibodies to active caspase-3


8


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

(Pharmingen; San Diego, CA) and cytochrome c (Pharmingen) were added at 0.5 µg/ml in PBS

containing 1% goat serum and 0.3% Triton X-100 and the neurons were labeled overnight at

4oC. The appropriate FITC-conjugated secondary antibodies were diluted to 15 µg/ml in PBS

containing 1% goat serum and 0.3% Triton X-100 and the cultures were incubated for 2 h.

Cultures were then rinsed in PBS and counterstained with Hoechst 33,342 (Molecular Probes).

Subcellular Fractionation, Immunoprecipitation, and Immunoblotting. CV-1 cells stably

expressing SM-20 were preincubated for 30 min in media lacking cysteine and methionine and

then incubated for 4 h in the same media containing 0.5 mCi/ml 35S-TransLabel (ICN; Costa

Mesa, CA) and 10% dialyzed calf serum. For preparing whole cell lysates, cells were rinsed

twice with cold PBS and then lysed in RIPA buffer (50 mM Tris (pH 8.0), 150 mM NaCl, 1%

Nonidet-P40, 1% deoxycholate, 0.1% SDS, 0.1 mM phenylmethylsulfonyl fluoride (PMSF), 5

µg/ml leupeptin, 5 µg/ml aprotinin) for 15 min at 4ºC. Cell lysates were clarified by centrifugation

at 10,000 x g for 10 min, preabsorbed with protein A-Sepharose (Pharmacia; Piscataway, NJ),

and then incubated with anti-SM-20 antibody or, in some cases, antibody that had been

preincubated with excess blocking peptide. Immune complexes were precipitated with protein

A-Sepharose, washed four times with RIPA buffer, separated by 12.5% SDS-PAGE, and

analyzed using a Phosphorimager and ImageQuant software (Molecular Dynamics; Sunnyvale,

CA).

Mitochondrial enriched fractions were prepared essentially as described by others (21, 22). For

these experiments NIH3T3 cells were used rather than CV-1 cells because higher levels of SM-

20 expression were obtained. (Note that both CV-1 and NIH3T3 cells showed the same pattern


9


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

of SM-20 immunofluorescence.) Cells were transiently transfected with SM-20/Rc-CMV or an

empty control vector and metabolically labeled as described above. After labeling, each dish

was rinsed twice with PBS before adding ice-cold homogenization buffer (17 mM

morpholinopropane sulfonic acid, 2.5 mM EDTA, 250 mM sucrose, 0.1 mM PMSF, 5 µg/ml

leupeptin, and 5 µg/ml aprotinin). The cells were then scraped from the dishes, homogenized in

a Dounce homogenizer (Kontes; Vineland, NJ) for 20-25 strokes, and then centrifuged at 500 x

g for 10 min at 4ºC to pellet nuclei. The supernatants were then centrifuged at 10,000 x g for 15

min. The resultant pellets from this centrifugation were resuspended in homogenization buffer

and centrifuged again at 10,000 x g for 15 min, yielding a mitochondria-enriched heavy

membrane pellet that also contains lysosomes, golgi, and rough endoplasmic reticulum. The

mitochondria-enriched fraction was resuspended in RIPA buffer and subjected to

immunoprecipitation with anti-SM-20 antibody as described above. A soluble cytosolic

fraction was prepared from the supernatants obtained after the 10,000 x g spin by further

centrifuging the supernatants at 150,000 x g for 60 min. To verify that the heavy membrane

fraction was enriched for mitochondria, 25 µg of protein from the heavy membrane and cytosolic

fractions were separated by SDS-PAGE and transferred to nitrocellulose membranes. The

membranes were blocked in 5% non-fat dry milk in Tris-buffered saline (TBS) prior to labeling

overnight with primary antibody against cytochrome c (Pharmingen) at 5 µg/ml in 1X TBS

containing 1% non-fat dry milk and 0.05% Tween-20. The blots were then incubated with

biotinylated goat anti-mouse IgG (Jackson ImmunoResearch) at 0.3 µg/ml and detected using a

streptavidin-conjugated alkaline phosphatase assay kit (BioRad; Hercules, CA).

Intracellular Microinjections and Quantitation of Cell Death. Sympathetic neurons were


10


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

microinjected with expression plasmids and scored for viability as described previously (12).

Neurons were injected directly into the nucleus with each plasmid at 50 µg/ml in a buffer

containing non-fixable rhodamine-dextran (survival experiments) or lysine-fixable rhodamine-

dextran (immunofluorescence experiments) to permit visualization of injected cells. For some

experiments the neurons were treated immediately following microinjection by including the

general caspase inhibitor BOC-Asp-CH2F (BAF) (Enzyme Systems Products; Dublin, CA) at

50 µM in the culture medium. The number of successfully injected (rhodamine-positive)

neurons was determined 1215 h after injection. Three days after injection, cells were stained

with the fluorescent DNA-binding dye Hoechst 33,342 (Molecular Probes) and evaluated for

survival by counting the number of rhodamine-positive cells that were phase-bright with

smooth and intact neurites, a discernible nucleus and diffusely stained chromatin. In contrast,

dying cells are characterized by condensed or undetectable chromatin, fragmented neurites, and

atrophic cell bodies. Percent survival is reported as the number of healthy cells divided by the

number of rhodamine-positive cells counted 12–15 h after injection.


11


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

RESULTS

Localization of SM-20 to mitochondria.

To determine where the SM-20 protein is localized in cells, we first performed

immunofluorescence on transfected cells using a polyclonal anti-peptide antibody directed

against the carboxy terminus of SM-20. This antibody specifically recognizes recombinant

SM-20 protein and it immunoprecipitates SM-20 from cells transfected with an SM-20

expression plasmid, but not from cells transfected with an empty vector (10). CV-1 cells were

chosen for these experiments because of their flattened morphology and large cytoplasmic

volume and because transient expression of SM-20 does not lead to death in these cells (use of

NIH3T3 cells yielded identical results, data not shown). SM-20 immunofluorescence occurred

in a punctate pattern throughout the cytoplasm of expressing cells (Fig. 1, panels B and E). In

contrast, neighboring untransfected cells or cells transfected with an empty expression vector did

not show detectable immunofluorescence.

The distribution of SM-20 immunofluorescence appeared similar to the distribution of

mitochondria described previously in CV-1 cells (23). Therefore, experiments were performed to

determine if SM-20 expression co-localizes with known mitochondrial markers. In dual

labeling experiments, SM-20 immunofluorescence completely coincided with the labeling

generated by a monoclonal antibody that recognizes cytochrome oxidase subunit I, an inner

mitochondrial membrane protein (Fig. 1, panels A–C). SM-20 immunofluorescence also co-

localized with a second mitochondrial protein, cytochrome oxidase subunit VI (data not shown).

The SM-20 localization pattern was also compared with the staining pattern of the

mitochondria-selective dye Mitotracker Green FM. SM-20 immunofluorescence co-localized


12


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

perfectly with mitochondria stained by Mitotracker Green FM (Fig. 1, panels D–F). The co-

localization of SM-20 with two distinct mitochondrial proteins and a mitochondria-selective

dye strongly suggests that SM-20 is localized to mitochondria in these cells.

Identification of a 33-kD processed form of SM-20 in mitochondria.

The SM-20 open reading frame predicts a 355 amino acid protein with a calculated molecular mass of 39.8

kilodaltons (kD) (13). When whole cell lysates were prepared from CV-1 cells stably expressing SM-20 and

immunoprecipitated with anti-SM-20 antibody, proteins of approximately 33 and 40 kD were specifically

recovered (Fig. 2A). Pre-incubating the antibody with the cognate peptide blocked immunoprecipitation of both

proteins. Moreover, proteins of the same sizes were detected in cells expressing a T7-epitope tagged form of SM-

20 (tagged at its carboxy-terminus) when an antibody specific for the T7 epitope was used for the

immunoprecipitation (data not shown). These results indicate that both the 33 and 40 kD proteins are products of

the SM-20 cDNA. Because antibodies specific for the carboxy terminus of SM-20 recognize the 33 kD protein, it

most likely corresponds to a processed form of SM-20 lacking amino-terminal sequences.

Proteins that are synthesized in the cytosol and then targeted to mitochondria often contain mitochondrial

targeting sequences at their amino terminus. Mitochondrial targeting sequences are typically 15–35 residues long,

rich in hydroxylated and positively-charged amino acids, and devoid of acidic residues (24). Inspection of the SM-20

protein sequence revealed that the first 59 amino acids contain 13 hydroxylated residues, 13 basic residues, and no

acidic residues. Because mitochondrial targeting sequences are often proteolytically removed during import into

mitochondria (25), we suspected that the 33 kD protein represented a processed form of SM-20 present in mitochondria.

NIH3T3 cells transiently transfected with SM-20 cDNA or a control plasmid were used to prepare cytosolic and

mitochondria-enriched protein fractions. Immunoprecipitation of these fractions with anti-SM-20 antibody

revealed that the 33 kD SM-20 protein was the predominant form present in the mitochondria-enriched fraction

(Fig. 2B). In contrast, the soluble cytosolic fraction contained almost exclusively the full-length SM-20 protein.

Because the antibody used to detect SM-20 recognizes its carboxy terminus, these results provide evidence that an

approximately 7 kD polypeptide is removed from the amino terminus of SM-20 when it is targeted to mitochondria.


13


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

SM-20 contains an amino-terminal mitochondrial targeting sequence.

To determine whether the amino-terminus of SM-20 is required for its mitochondrial localization, we

constructed a truncated form of SM-20 (SM-20(60–355)) expected to encode a protein similar in size to the 33 kD

form of SM-20 detected by immunoprecipitation. As predicted, SM-20(60–355) was not targeted to mitochondria

in CV-1 cells but instead was diffusely dispersed throughout the cytosol and nucleus, similar to DHFR (Fig. 3).

Thus the loss of 59 amino terminal residues apparently disrupted a mitochondrial targeting sequence within SM-20.

To ascertain whether sequences within the amino-terminal end of SM-20 are capable of targeting a

normally cytosolic protein to mitochondria, the first fifty amino acids of SM-20 were fused to the amino-terminus

of DHFR. When expressed in CV-1 cells, the fusion protein (SM-20(1–50)/DHFR) co-localized with Mitotracker

Green FM indicating that SM-20(1–50)/DHFR was efficiently targeted to mitochondria (Fig. 4, panels A–C). A

second fusion protein consisting of the first 25 amino acids of SM-20 fused to DHFR (SM-20(1–25)/DHFR) was

also targeted to mitochondria (Fig. 4, panels D–F). These data indicate that the amino-terminal 25 residues of SM-

20 can function as a mitochondrial targeting sequence.

Mitochondrial targeting of SM-20 is not required for its ability to induce cell death in

sympathetic neurons.

SM-20 mRNA levels and SM-20 protein synthesis increase in sympathetic neurons undergoing apoptosis (10).

Despite this, SM-20 protein levels are still relatively low in these cells and this has impeded our attempts at

confirming the localization of the endogenous SM-20 in neurons. To determine if SM-20 can be targeted to

mitochondria in neurons, we microinjected sympathetic neurons with a cDNA encoding a SM-20/GFP fusion

protein. The distribution of SM-20/GFP in sympathetic neurons visualized with confocal microscopy coincided

with the labeling pattern of Mitotacker Orange, a mitochondria-selective dye (Fig. 5, panels A–C). The

mitochondrial labeling was specific to the full length SM-20 protein; SM-20(60-355)/GFP was diffusely localized


14


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

throughout the cytosol and nucleus and did not co-localize with Mitotracker Orange (Fig. 5, panels D–F).

Expression of SM-20 in sympathetic neurons maintained in the presence of NGF induces cell death in

approximately half of the microinjected neurons (10). To determine whether the mitochondrial localization of SM-20 is

necessary for its ability to promote neuronal death, we microinjected sympathetic neurons with expression plasmids

for SM-20(60-355)/GFP or, as a control, GFP. Three days after microinjection, most GFP-injected neurons

appeared healthy with phase-bright cell bodies and readily discernible nuclei and nucleoli (Fig. 6A). In contrast,

most of the SM-20(60-355)/GFP-injected neurons displayed morphological features of apoptosis, including

atrophic cell bodies and condensed or degraded chromatin. Similar results were obtained when the SM-20(60-355)

cDNA fused to the T7-epitope tag was expressed in the neurons. Quantifying these results showed that greater than

60% of the SM-20(60-355)-injected neurons contained condensed or degraded chromatin compared to only 25%

of the control cells injected with β-galactosidase (LacZ) (Fig. 6B). The extent of cell death induced

by SM-20(60-355) in these experiments is comparable to the level of cell death produced by

microinjection of full-length SM-20 (10). Since SM-20(60-355) is not targeted to mitochondria in

sympathetic neurons, the mitochondrial localization of SM-20 does not appear to be critical for

its death promoting activity.

SM-20 promoted death in sympathetic neurons is caspase-dependent and leads to activation of

caspase-3.

Sympathetic neurons deprived of NGF undergo caspase-dependent apoptosis that can be

blocked by the general caspase inhibitor BAF (26). To investigate whether SM-20 induced death is

also caspase-dependent, we microinjected sympathetic neurons with expression vectors

containing SM-20 or LacZ and then incubated the cells in NGF-containing media with or


15


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

without BAF (50 µM). After three days, the SM-20 injected neurons maintained in the absence

of BAF displayed morphological features of apoptosis, including condensed, crescent-shaped or

degraded chromatin (Fig. 7A, panels ac). In contrast, SM-20 injected neurons maintained in the

presence of BAF were healthy with intact cell bodies and diffuse chromatin (Fig. 7A, panels

d–f). Greater than 70% of SM-20 injected neurons maintained in the presence of BAF survived

after 72 h compared to 38% in the absence of BAF (Fig. 7B). LacZ injected cells, maintained

under the same conditions, showed no significant difference in percent survival. Thus SM-20

induced neuronal death is caspase-dependent and can be blocked by a general caspase inhibitor.

To test whether SM-20 induced death leads to activation of effector caspases such as

caspase-3, sympathetic neurons expressing SM-20, LacZ or Bax were analyzed by

immunofluorescence using an antibody that recognizes activated caspase-3. As shown in Fig. 8

(panels a–d), LacZ injected neurons appeared healthy and lacked active caspase-3

immunofluorescence. In contrast, both SM-20 and Bax injected neurons showed significant

staining for active caspase-3 (Fig. 8, panels e–l). The percent of injected cells expressing

activated caspase-3 at 72 h was significantly greater for both SM-20 (18.5 ± 1.5%) and Bax

(47.2 ± 9.8%) when compared with LacZ injected cells (6.3 ± 1.7%). For SM-20 injected

neurons (see Fig. 7B) and neurons injected with Bax or deprived of NGF (data not shown), the

percent of cells with activated caspase-3 is less than the fraction that undergoes cell death, as

assayed by visualizing Hoechst-stained nuclei. Thus, activated caspase-3 may only accumulate

to detectable levels in cells with advanced chromatin condensation and degradation. Taken

together, these data suggest that SM-20 induced cell death is caspase-dependent and that it


16


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

involves activation of caspase-3.

SM-20 expression does not cause cytochrome c release from mitochondria.

The release of cytochrome c from mitochondria in sympathetic neurons following NGF

withdrawal is thought to contribute to caspase activation (27). Once in the cytosol, cytochrome c can

form a complex with Apaf-1 and caspase-9 resulting in the activation of caspase-9 and

eventually downstream caspases such as caspase-3 (28). To determine whether ectopic expression

of SM-20 can cause release of cytochrome c from mitochondria, we performed cytochrome c

immunofluorescence on NGF-maintained neurons microinjected with expression plasmids for

SM-20, LacZ or Bax. For these experiments, the injected neurons were incubated in the

presence of BAF (50 µM), which prevents the final stages of cell death caused by NGF

withdrawal without blocking cytochrome c release (27). Greater than 85% of SM-20 or LacZ

injected neurons showed a punctate mitochondrial pattern following labeling with an anti-

cytochrome c antibody indicating that cytochrome c was not released from mitochondria in these

cells (Fig. 9A and 9B). In contrast, approximately 75% of Bax injected neurons displayed a

diffuse cytoplasmic staining pattern that closely resembled the cytochrome c

immunofluorescence observed after NGF-deprivation (27), consistent with a role for Bax as an

upstream regulator of cytochrome c release (29). These data indicate that SM-20 expression, by

itself, is not sufficient to cause cytochrome c release.


17


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

DISCUSSION

In nonneuronal and neuronal cells, SM-20 immunofluorescence co-localized with

mitochondria-selective dyes as well as the immunofluorescence for two distinct mitochondrial

proteins. Deletion of 59 amino acids from the amino terminus of SM-20 resulted in the

truncated protein being distributed uniformly throughout the cell. Consistent with its

mitochondrial localization, the amino-terminal 25 amino acids of SM-20 were found to be

sufficient to target a normally cytosolic protein to mitochondria. In addition to the expected 40

kD SM-20 protein, both CV-1 cells and NIH3T3 cells expressed a 33 kD form of SM-20. In

cell fractionation experiments, the smaller SM-20 protein was present predominantly in the

mitochondria-enriched fraction. This result, together with the presence of an amino-terminal

mitochondrial targeting sequence in SM-20, suggests that the mitochondrial targeting sequence

is proteolytically removed when SM-20 is imported into mitochondria. Together, these results

identify SM-20 as a novel mitochondrial protein.

SM-20 immunocytochemistry was previously detected in a punctate pattern in the

cytoplasm of vascular smooth muscle cells (15). The SM-20 immunostaining was shown to be

distinct from that of α-actin but efforts to further localize SM-20 were not undertaken. The

appearance of SM-20 immunostaining in smooth muscle cells was very similar to the

distribution of SM-20 in neurons and CV-1 cells that we observed. The identification of a

mitochondrial targeting sequence in SM-20 suggests that the punctate distribution of SM-20

detected previously in smooth muscle cells reflects targeting of the endogenous SM-20 protein

to mitochondria.

Although SM-20 is predominantly localized to mitochondria, a truncated form of SM-20


18


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

lacking its mitochondrial targeting sequence unexpectedly retained the ability to induce death in

sympathetic neurons, suggesting that endogenous SM-20 may not need to be localized to

mitochondria to induce cell death. One explanation for this apparent discrepancy is that SM-20

might be released from mitochondria into the cytosol after NGF withdrawal, analogous to the

release of cytochrome c (27, 30). Expression of truncated SM-20(60–355) that is present in the

cytoplasm could mimic this event. An alternative explanation is that endogenous SM-20 does

not get released from mitochondria during cell death and that it is the mitochondrial form of

SM-20 that contributes to apoptosis. Expression of SM-20(60–355) may result in a small

amount of SM-20(60–355) being imported into mitochondria or binding to undefined proteins

present on the surface of mitochondria, which in this case may be sufficient to promote death.

To further examine the mechanism of SM-20 induced cell death, we investigated the

role of SM-20 in caspase activation and in cytochrome c release from mitochondria. As with

NGF-withdrawal induced death, SM-20 induced death was efficiently blocked by the general

caspase inhibitor BAF. In addition, we detected activated caspase-3 in a significant fraction of

SM-20 injected neurons suggesting that SM-20 may directly or indirectly participate in the

activation of this caspase in dying neurons. Release of cytochrome c from mitochondria occurs

after NGF withdrawal and has been suggested to contribute to cell death in this paradigm (27, 30). In

contrast to Bax, ectopic expression of SM-20 did not cause widespread cytochrome c release.

Since SM-20 does not act by directly causing cytochrome c release, our results suggest that SM-

20 functions during cell death in a pathway that is either downstream or independent of

cytochrome c release.


19


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Although SM-20 does not resemble known mammalian proteins, it contains a 218-

amino acid region that is 43% identical (61% similar) to a portion of the protein encoded by the

C. elegans egl-9 gene (17). Deletions and mutations that disrupt the egl-9 gene confer resistance to a

Pseudomonas aeruginosa-derived toxin that causes a lethal neuromuscular paralysis in C. elegans.

The genetics in this model indicate that the toxin results in aberrant activation of Egl-9 rather

than in its inactivation. Consistent with a role for Egl-9 as a regulator of neuronal signaling or

muscle contraction, the egl-9 promoter is most active in muscle cells and in neurons (17).

Expression of the rat SM-20 gene is also greatest in muscle-containing tissues and in brain

suggesting that the function of SM-20/Egl-9 in these tissues may be conserved (13). In humans,

several SM-20-related genes or pseudogenes appear to be present on distinct chromosomes (D.

M. Hasbani and R. S. Freeman, unpublished observation). Future studies will examine the

ability of these SM-20 related proteins including Egl-9 to induce cell death, as well as their

subcellular localization.


20


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

ACKNOWLEDGMENTS

We are grateful to Dr. Daniel Donoghue for providing the DHFR plasmid, Dr. Patricia

Hinkle for providing NIH3T3 cells, and Dr. Jane Chisholm for expert assistance with confocal

microscopy. We also thank Daphne Hasbani and Leah Larocque for excellent technical

assistance. R. S. F. acknowledges the generous support of the Paul Stark Endowment at the

University of Rochester. E. A. L. and P. D. S. were supported by an NIEHS institutional training

grant (ES07026). This work was supported by NIH grant NS34400.


21


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

FOOTNOTES

1The abbreviations used are: BAF, BOC-Asp-CH2F; DHFR, dihydrofolate reductase; FITC,

fluorescein isothiocyanate; GFP, green fluorescent protein; kD, kilodaltons; LacZ, β-

galactosidase; NGF, nerve growth factor; PBS, phosphate-buffered saline; PMSF,

phenylmethylsulfonyl fluoride; TRITC, tetramethylrhodamine isothiocyanate.


22


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

REFERENCES

1. Martin, D. P., Schmidt, R. E., DiStefano, P. S., Lowry, O. H., Carter, J. G., and Johnson, E. M. (1988) J. Cell Biol. 106, 829-844

2. Edwards, S. N., Buckmaster, A. E., and Tolkovsky, A. M. (1991) J. Neurochem. 57, 2140-2143

3. Freeman, R. S., Estus, S., and Johnson, E. M. (1994) Neuron 12, 343-355

4. Estus, S., Zaks, W. J., Freeman, R. S., Gruda, M., Bravo, R., and Johnson, E. M. (1994) J. Cell Biol. 127, 1717-1727

5. Ham, J., Babij, C., Whitfield, J., Pfarr, C. M., Lallemand, D., Yaniv, M., and Rubin, L. L. (1995) Neuron 14, 927-939

6. Watson, A., Eilers, A., Lallemand, D., Kyriakis, J., Rubin, L. L., and Ham, J. (1998) J. Neurosci. 18, 751-762

7. Kranenburg, O., van der Eb, A. J., and Zantema, A. (1996) EMBO J. 15, 46-54

8. Park, D. S., Levine, B., Ferrari, G., and Greene, L. A. (1997) J. Neurosci. 17, 8975-8983

9. Park, D. S., Morris, E. J., Padmanabhan, J., Shelanski, M. L., Geller, H. M., and Greene, L. A. (1998) J. Cell Biol. 143, 457-467

10. Lipscomb, E. A., Sarmiere, P. D., Crowder, R. J., and Freeman, R. S. (1999) J. Neurochem. 73, 429-432

11. Martin, D. P., Wallace, T. L., and Johnson, E. M. (1990) J. Neurosci. 10, 184-193

12. Crowder, R. J. and Freeman, R. S. (1998) J. Neurosci. 18, 2933-2943

13. Wax, S. D., Rosenfield, C. L., and Taubman, M. B. (1994) J. Biol. Chem. 269, 13041-13047

14. Moschella, M. C., Menzies, K., Tsao, L., Lieb, M. A., Kohtz, J. D., Kohtz, D. S., and Taubman, M. B. (1999) Gene Exp. 8, 59-66

15. Wax, S. D., Tsao, L., Lieb, M. E., Fallon, J. T., and Taubman, M. B. (1996) Lab. Invest. 74, 797-808

16. Madden, S. L., Galella, E. A., Riley, D., Bertelsen, A. H., and Beaudry, G. A. (1996) Cancer Res. 56, 5384-5390

17. Darby, C., Cosma, C. L., Thomas, J. H., and Manoil, C. (1999) Proc. Natl. Acad. Sci. U. S. A. 96, 15202-15207


23


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

18. Trent, C., Tsung, N., and Horvitz, R. (1983) Genetics 104, 619-647

19. Green, D. R. and Reed, J. C. (1998) Science 281, 1309-1312

20. Johnson, L. V., Walsh, M. L., and Chen, L. B. (1980) Proc. Natl. Acad. Sci. U. S. A. 77, 990-994

21. Krajewski, S., Tanaka, S., Takayama, S., Schibler, M. J., Fenton, W., and Reed, J. C. (1993) Cancer Res. 53, 4701-4714

22. Reed, J. C., Meister, L., Tanaka, S., Cuddy, M., Yum, S., Geyer, C., and Pleasure, D. (1991) Cancer Res. 51, 6529-6538

23. Modica-Napolitano, J. S., Steele, G. D. J., and Chen, L. B. (1989) Cancer Res. 49, 3369-3373

24. Hurt, E. C. and van Loon, A. P. G. M. (1986) Trends Bioch. Sci. 11, 204-207

25. Neupert, W. (1997) Annu. Rev. Biochem. 66, 863-917

26. Deshmukh, M., Vasilakos, J., Deckwerth, T. L., Lampe, P. A., Shivers, B. D., and Johnson, E. M. (1996) J. Cell Biol. 135, 1341-1354

27. Deshmukh, M. and Johnson, E. M. (1998) Neuron 21, 695-705

28. Kluck, R. M., Bossy-Wetzel, E., Green, D. R., and Newmeyer, D. D. (1997) Science 275, 1132-1136

29. Putcha, G. V., Deshmukh, M., and Johnson, E. M. (1999) J. Neurosci. 19, 7476-7485

30. Neame, S. J., Rubin, L. L., and Philpott, K. L. (1998) J. Cell Biol. 142, 1583-1593


24


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

FIGURE LEGENDS

Fig. 1. Localization of SM-20 to mitochondria. CV-1 cells were transiently transfected with

SM-20-Rc/CMV plasmid DNA. Dual immunofluorescence was performed for SM-20 and

cytochrome oxidase I (panels A-C) 48 h after transfection. Primary antibodies for cytochrome

oxidase I (panel A) and SM-20 (panel B) were detected using FITC- and TRITC-conjugated

secondary antibodies, respectively. Other cells were labeled with Mitotracker Green FM and

then subjected to immunofluorescence with anti-SM-20 antibody (panels D-F). Yellow in the

overlay images (panels C and F) appears where the red fluorescence of SM-20 co-localizes with

the green fluorescence of either cytochrome oxidase I (panel A) or Mitotracker Green FM (panel

D). Bar = 10 µm.

Fig. 2. A 33-kD processed form of SM-20 is enriched in mitochondria. (A) Whole cell extracts

prepared from metabolically labeled CV-1 cells (lane 1) or from CV-1 cells stably expressing

SM-20 (lanes 2 and 3) were immunoprecipitated with a SM-20 anti-peptide antibody.

Immunoprecipitations were done in the presence (+) or absence (–) of excess blocking peptide

corresponding to the last 14 amino acids of SM-20. The positions of molecular weight markers

are indicated. Arrows indicate the presence of approximately 40 and 33 kD proteins in SM-20

expressing cells that were specifically immunoprecipitated in the absence of the blocking peptide

(lane 2). (B) Cytosolic and mitochondria-enriched fractions were prepared from NIH3T3 cells

transiently transfected with a SM-20 expression vector (lanes 2 and 4) or with pcDNA3 (lanes 1

and 3) and metabolically labeled for 4 h with 35S-TransLabel. Fractions were

immunoprecipitated with anti-SM-20 antibody and then the immunoprecipitated proteins were


25


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

subjected to SDS-PAGE and Phosphorimager analysis. Note that the 33 kD protein (lower

arrow) is present only in the mitochondria-enriched fraction. The cytosolic and mitochondrial-

enriched fractions were immunoblotted with anti-cytochrome c antibody to confirm that a

fraction enriched in mitochondrial proteins was obtained.

Fig. 3. An amino-terminal deletion of SM-20 prevents its mitochondrial localization.

Expression vectors encoding DHFR tagged with a carboxy-terminal T7 epitope (A), SM-20(60-

355)/T7 (SM-20 minus its first 59 amino acids) (B), or full-length SM-20 (C) were transfected

into CV-1 cells. Indirect immunofluorescence was performed 48 h later using anti-T7 antibody

or anti-SM-20 antibody. Both DHFR and SM-20(60–355)/T7 show diffuse cytoplasmic and

nuclear fluorescence in marked contrast to the punctate distribution of the full-length SM-20.

Bar = 10 µm.

Fig. 4. SM-20 contains an amino terminal mitochondrial targeting sequence. CV-1 cells

transfected with expression vectors for SM-20(150)/DHFR (panels AC) and SM-

20(125)/DHFR fusion proteins (panels DF) were incubated with Mitotracker Green FM (panels A and D).

Indirect immunofluorescence using anti-T7 antibody and TRITC-conjugated secondary

antibody was performed to detect the T7-tagged fusion proteins (panels B and E). Yellow in the

overlay images (panels C and F) appears where the red fluorescence of the fusion proteins co-

localizes with the green fluorescence of Mitotracker Green FM. Bar = 10 µm.


26


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Fig. 5. SM-20, but not SM-20(60-355), is localized to mitochondria in sympathetic neurons.

Sympathetic neurons were microinjected with expression vectors encoding SM-20/GFP (panels

A-C) or SM-20(60-355)/GFP fusion proteins diluted to 50 µg/ml in injection buffer without

rhodamine-dextran (panels D-F). Cells were labeled with the mitochondria-selective dye

Mitotracker Orange 24 h after microinjection (panels A and D). Images of live cells were

acquired by confocal microscopy. The green fluorescence in cells injected with SM-20/GFP

fusion constructs is shown in panels B and E. These images are superimposed with the red

fluorescence of Mitotracker Orange in panels C and F. The letter “n” represents the position of

the nucleus in the injected neurons. Bar = 10 µm.

Fig. 6. SM-20(60-355) induces cell death in NGF-maintained sympathetic neurons. (A)

Sympathetic neurons were microinjected with GFP (panels a-c) or SM-20(60-355)/GFP (panels

d-i) expression plasmids and after 72 h stained with the DNA-binding dye Hoechst 33,342.

Injected cells were identified by the green fluorescence of GFP (panel b) or SM-20(60-

355)/GFP fusion protein (panels e and h). The Hoechst-stained nuclei of the same cells are shown in

panels c, f, and i and phase-contrast images are shown in panels a, d, and g. The white arrows

indicate the injected cells and their corresponding phase-contrast and Hoechst-stained nuclei.

Bar = 15 µm. (B) NGF-maintained neurons were microinjected with expression vectors

encoding LacZ or SM-20(60-355). After 72 h, the injected cells were stained with Hoechst

33,342 and scored for viability as described in Experimental Procedures. Data are mean + SEM

from six independent experiments with 200–300 neurons scored per injected DNA in each


27


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

experiment. The mean survival of SM-20-injected cells was significantly less than that of

LacZ-injected cells (*, two-tailed t-test, p = 0.004).

Fig. 7. The cell death induced by SM-20 in sympathetic neurons is blocked by a general caspase

inhibitor. (A) Sympathetic neurons were microinjected with a SM-20 expression plasmid and

treated with (panels d-f) or without (panels a-c) 50 µM BAF for 72 h. Injected cells were

identified by inclusion of a fixable rhodamine-dextran dye in the injection buffer (panels b and

e). The white arrows indicate the injected cells and their corresponding phase-contrast (panels a

and d) and Hoechst-stained nuclei (panels c and f). Bar = 15 µm. (B) NGF-maintained neurons

were microinjected with expression vectors encoding LacZ or SM-20 and incubated in the

presence or absence of 50 µM BAF. After 72 h, the injected cells were stained with Hoechst

33,342 and scored for viability as described in Experimental Procedures. Data are mean + SEM

from four independent experiments. The mean survival of SM-20-injected cells in the absence

of BAF was significantly less than that of SM-20-injected cells in the presence of BAF (two-

tailed t-test, p = 0.0013).

Fig. 8. Expression of SM-20 in sympathetic neurons leads to caspase-3 activation. NGF-

maintained sympathetic neurons were microinjected with expression plasmids encoding LacZ

(panels a–d), SM-20 (panels e–h) or Bax (panels i–l). Injected cells were identified by inclusion of

a fixable rhodamine-dextran dye in the injection buffer (panels b, f, and j). After three days, the

cells were fixed and immunofluorescence was performed using anti-activated caspase-3

antibody (panels c, g, and k). The white arrows indicate the injected cells and their


28


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

corresponding phase-contrast (panels a, e, and i) and Hoechst-stained nuclei (panels d, h, and l).

Bar = 15 µm.

Fig. 9. Absence of detectable cytochrome c release from mitochondria in SM-20 injected

sympathetic neurons. (A) Sympathetic neurons were microinjected with LacZ (panels a-c),

SM-20 (panels d-f) or Bax (panels g-i) expression vectors and maintained in BAF-containing

media. After 24-48 h, the cells were fixed and immunofluorescence was performed using anti-

cytochrome c antibody (panels c, f, and i). Injected cells were identified by inclusion of a fixable

rhodamine-dextran dye in the injection buffer (panels b, e, and h). The white arrows indicate the

injected cells and their corresponding phase-contrast images (panels a, d, and g). Bar = 15 µm.

(B) NGF and BAF-maintained neurons were microinjected with expression vectors encoding

LacZ, SM-20 or Bax. After 48 h, the percent of injected cells (rhodamine-positive) showing a

loss of punctate mitochondrial cytochrome c labeling was determined. Data are mean + SEM

from 3–7 independent experiments. The mean number of SM-20 injected cells showing release

of cytochrome c was not significantly different than that for LacZ (two-tailed t-test, p > 0.2).


29


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Elizabeth A. Lipscomb, Patrick D. Sarmiere and Robert S. Freemannerve growth factor-dependent neurons

SM-20: a novel mitochondrial protein that causes caspase-dependent cell death in

published online November 1, 2000J. Biol. Chem.

10.1074/jbc.M008407200Access the most updated version of this article at doi:

Alerts:

When a correction for this article is posted•

When this article is cited•

to choose from all of JBC's e-mail alertsClick here


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/lookup/doi/10.1074/jbc.M008407200

http://www.jbc.org/cgi/alerts?alertType=citedby&addAlert=cited_by&cited_by_criteria_resid=jbc;M008407200v1&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/early/2000/11/01/jbc.M008407200.citation

http://www.jbc.org/cgi/alerts?alertType=correction&addAlert=correction&correction_criteria_value=early/2000/11/01/jbc&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/early/2000/11/01/jbc.M008407200.citation

http://www.jbc.org/cgi/alerts/etoc

http://www.jbc.org/

sm-20: a novel mitochondrial protein that causes caspase ... · pdf filekeywords: apoptosis,...

Documents