| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
Department of Pathology, Helsinki University Hospital and
* Haartman Institute, University of Helsinki, Finland; and
Karolinska Institute and Hospital, Stockholm, Sweden
1Correspondence: University of Helsinki, Haartman Institute, Department of Pathology, P.O. Box 21 (Haartmaninkatu 3), FIN-00014 Helsinki, Finland. E-mail: Leif.Andersson{at}helsinki.fi
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: PMA • neural sprouting • familial amyloidosis of the Finnish type (FAF) • brain
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. The activity of
gelsolin is stimulated by Ca2+ and/or low pH,
which promotes its association to actin filaments (2)
. The
gelsolin molecule consists of six structurally related 15 kDa domains.
These form a compact globular structure in a
Ca2+-free environment (3)
, in which
the putative actin binding sequences are not sufficiently exposed to
bind actin. A rise in Ca2+ or lowered pH is
proposed to release connections joining the amino- and
carboxyl-terminal halves of the molecule, thereby allowing the molecule
to adopt its active configuration. Polyphosphoinositides,
particularly phosphatidylinositol 4,5-bisphosphate
(PIP2)2
(2)
, down-regulate the activity of gelsolin by inducing
its dissociation from actin (4
, 5)
. There is also evidence
indicating that the in vivo levels of
Ca2+ and PIP2 may regulate
the compartmental distribution of gelsolin between the
membrane-associated cytoskeleton and the cytosol (reviewed in ref
1
).
The gene for human gelsolin is located on chromosome 9q32-q34
(6)
. It codes for cytoplasmic gelsolin and a secreted
plasma form of gelsolin that carries an additional amino-terminal
extension of 23 amino acids. Both the cytoplasmic and secreted forms of
gelsolin are expressed in most adult tissues (7)
. The
expression pattern in the developing rat brain (8)
, where
initial low levels of gelsolin are followed by increased expression
levels around day 10 and a subsequent decrease around day 30, suggests
that gelsolin activity is involved in early brain development.
A single nucleotide substitution G654 to A654 in the genomic DNA
sequence of gelsolin (9
, 10)
replaces aspartic acid at
residue 187 with asparagine, giving rise to Finnish type familial
amyloidosis (FAF), an autosomal dominant disease. The mutation exposes
a novel protease cleavage site in the mature gelsolin protein, which is
the proposed mechanism responsible for the generation of the
amyloidogenic fragment. An accumulation of the amyloid material gives
rise to a systemic amyloidosis characterized by corneal lattice
dystrophy, skin changes, renal complications, and a neuropathy that
affects the cranial nerves in particular (11)
.
Furthermore, plasma gelsolin isolated from FAF patients has been shown
to have a defective actin-severing activity in vitro
(12)
.
The neuropathy occurring in FAF includes signs of decreased
conductibility and demyelination (13)
. The amount of
amyloid found in the vicinity of nerves does not, however,
satisfactorily explain the severity of the neuropathy. This raises the
question of whether the presence of mutated gelsolin directly impairs
neural functions. To address this question, we created stable
transfectants of wild type and G654 to A654 point-mutated gelsolin in
the human, neural crest-derived Paju cell line.
We demonstrate that gelsolin activity is a regulator of neural differentiation. Overexpression of wild type gelsolin inhibited neural sprouting induced by treatment with phorbol 12-myristate 13-acetate (PMA). The induced sprouting, however, which requires actin polymerization, was not affected by overexpression of mutated gelsolin.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
was subcloned into the expression vector
pcDNA3 (Invitrogen, The Netherlands). The G654 to A654 point mutation
was introduced by polymerase chain reaction (PCR) mutagenesis, using
5'-TCAACAATGGCAACTGCTTCATCCTG-3' and 5'-CAGGATGAAGCAGTTGCCATTGTTGA-3'
as inside primers containing the point mutation. As outside primers,
5'-CGAGTTCCTCAAGGCAGGGA-3' and 5'-CACGGGCACCTTGTTGGAAC-3' were used. A
1137 bp BglII fragment was excised from the native sequence
and substituted with the corresponding 1137 bp mutated PCR fragment.
Sequence integrity was verified by sequencing.
Cells
The Paju tumor cell line was established in our laboratory from
the pleural fluid of a 16-year-old girl who had a widespread,
metastatic neural crest-derived tumor. The cells grow surface adherent
and were cultivated in RPMI 1640 medium, supplemented with 5% fetal
calf serum (FCS), penicillin G (10 U/ml), streptomycin sulfate (50
mg/ml), and 1 mM glutamine. For subculturing, the cells were detached
by treatment with Versene/EDTA (Gibco BRL, Grand Island, N.Y.). PMA was
obtained from Sigma (St. Louis, Mo.), dissolved in ethanol, and used at
an optimal concentration of 10 nM. Paju cells were transfected with the
vector construct using lipofectamine reagent according to the
instructions of the manufacturer (Gibco BRL). Transfected cells were
selected for resistance to G418 (700 µg/ml) and single-cell cloned.
All experimental results were confirmed with separate single-cell
cloned cell lines of the wild type and G654A gelsolin transfected Paju
cells.
Western blotting
The cells were detached and washed with phosphate-buffered
saline (PBS), lysed in ice-cold Pawson lysis buffer containing 50 mM
HEPES (pH 7.0), 150 mM NaCl, 10% glycerol, 1% Triton X-100, 1.5 mM
MgCl2, 1 mM EGTA, 100 mM NaF, 10 mM NaPPi, 1 mM
orthovavadate, 1 mM PMSF, and 10 mg/ml each of aprotinin and leupeptin.
The samples were centrifuged at 14,000 x g for 5 min
and the supernatants were collected. An aliquot was removed for total
protein estimation using Bio-Rad protein assay (Bio-Rad Laboratories,
Hercules, Calif.). An aliquot corresponding to 20 µg of total protein
of each sample was separated by sodium dodecyl sulfate-polyacrylamide
gel electrophoresis under reducing conditions and transferred
electrophoretically to nitrocellulose filters. Nonspecific binding of
antibody was blocked with 3% bovine serum albumin in PBS/0.1% Triton
X-100 for 1 h. Immunoblotting was carried out with the respective
antibody, followed by peroxidase-conjugated secondary
anti-immunoglobulin (anti-Ig) antibodies, and the blots were developed
by the enhanced chemiluminescence method (Amersham, U.K.). Monoclonal
antibodies to gelsolin and ß-actin were purchased from Sigma.
Immunoprecipitation was carried out by incubating the gelsolin antibody
for 2 h at +4°C with the cell lysates, followed by an incubation
for 1 h with agarose-conjugated goat anti-mouse IgG antibody
(Sigma). Monoclonal anti-CD20 (Oy Medix AB, Finland) was used as a
control antibody. Precipitated proteins were eluted with Laemmli sample
buffer and used for Western blotting.
Cell staining
Cells were grown on coverslips in the presence of 10 nM PMA for
2 days. The coverslips were collected, rinsed in PBS, fixed in 3.5%
paraformaldehyde, and carefully rinsed three times in PBS. The cells
were permeabilized with PBS/0.05% Triton X-100 for 10 min and rinsed
in PBS. For phalloidin staining, the coverslips were incubated for 45
min with a 13 U/ml dilution of rhodamine-conjugated phalloidin
(Molecular Probes, Eugene, Oreg.) in PBS. The coverslips were rinsed
three times in PBS and mounted in glycerol:PBS (1:1) for microscopic
analysis. The cells stained for gelsolin were incubated for 15 min in
PBS/2% FCS, followed by incubation for 1 h with the gelsolin
antibody diluted 1:150 in PBS/2% FCS. The coverslips were rinsed three
times in PBS/2% FCS and incubated for 45 min with a fluorescein
isothiocyanate-conjugated goat anti-mouse antibody (DAKO A/S,
Glostrup, Denmark) diluted 1:20 in PBS/2% FCS. The coverslips were
rinsed three times in PBS/2% FCS and mounted in glycerol:PBS (1:1) for
microscopic analysis.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
A). The gelsolin encoded by the transfected cDNA had a
carboxyl-terminal truncation and was therefore easily distinguishable
from the endogenous gelsolin. Carboxyl-terminal truncation of gelsolin
is known to result in a loss of calcium requirement for filament
severing and monomer binding (15)
.
This ensured that the transfected gelsolin was constitutively active
and allowed us to study its effect regardless of internal or external
calcium stimuli. Clones of transfected Paju cells expressing equal
amounts of WT gelsolin and G654A gelsolin were selected. The level of
overexpression in the transfected cells studied was densitometrically
measured to be 2.4-fold the endogenous level for the
wild type and 3.8-fold the endogenous level for the G654A clone. Cells
transfected with the empty vector served as controls. Since gelsolin is
also secreted as plasma gelsolin, we assayed the amounts of secreted
gelsolin in conditioned media of the various transfected clones (Fig. 1B
). Relatively large amounts were secreted as compared to
control clones containing the empty vector (pcDNA3). A differential
electrophoretic pattern was observed between the secreted WT and G654A
gelsolins. It is known that the proteolytic cleavage of mutant gelsolin
differs from normal gelsolin (16)
and leads to the
production of the amyloidogenic fragment.
|
Overexpression of WT gelsolin and G654A gelsolin had strikingly
different effects on the shape of Paju cells. By 10 h after
plating, the G654A gelsolin overexpressing cells already displayed a
flat, polygonic morphology with tiny cytoplasmic protrusions, whereas
the WT overexpressing cells retained their round shape (Fig. 2
A). The Paju cells transfected with the empty vector and
untransfected cells had acquired an intermediate morphology.
This prompted us to investigate the impact of gelsolin
overexpression on PMA-induced neural sprouting.
|
Effect of gelsolin on PMA-induced neural differentiation
When exposed to PMA, Paju cells undergo neural differentiation by
producing extensive neurite outgrowths. Cells overexpressing WT or
G654A gelsolin were treated with 10 nM PMA for 2 days (Fig. 2A
). Control cells and cells overexpressing G654A gelsolin
responded by characteristic neurite outgrowths, whereas cells
overexpressing WT gelsolin displayed a completely different behavior.
Treatment with PMA did not induce typical neural sprouting, and the
cells acquired only some short cytoplasmic protrusions (Fig. 2A
). The endogenous levels of gelsolin were not affected by
PMA treatment (Fig. 2B
) in any of the cell clones.
Since one of the major roles of gelsolin is to modify actin filaments,
and the organization of the actin cytoskeleton is largely decisive for
the cellular shape, we stained F-actin in PMA-treated cells with
rhodamine-phalloidin (Fig. 3
). The slender neurite-like extensions of PMA-treated control cells or
cells overexpressing G654A gelsolin contained long polymerized actin
cables. These were absent from the cells overexpressing WT gelsolin.
Taken together with the previously reported loss of actin-severing
ability in vitro of plasma gelsolin isolated from FAF
patients (12)
, our findings suggest that the different
responses to PMA treatment of cells overexpressing WT or G654A gelsolin
could be due to the reported defective severing ability of mutated
cytoplasmic gelsolin in vivo.
|
Actin binding of wild type and mutant gelsolin
To investigate whether the morphological differences between the
transfected Paju cells were due to the inability of the mutant gelsolin
to bind to actin, we immunoprecipitated gelsolin and measured the
amount of coprecipitated actin (Fig. 4
). The amount of coprecipitated actin was identical for both wild type
and mutant gelsolin, showing that no differences exist in the monomeric
actin binding capacity of WT and mutated gelsolin.
|
Distribution of wild type and mutated gelsolin in Paju cells
Knowing the impact of WT gelsolin vs. G654A gelsolin on the
induced neural differentiation, we investigated by indirect
immunofluorescence the distribution of gelsolin in control Paju cells
and in cells transfected with the gelsolin cDNA constructs. A
membrane-associated immunoreactivity was seen in the control cells,
with more intense staining in the emerging growth cones and neurite
extensions. In the Paju cells transfected with the gelsolin constructs,
the overexpressed protein seemed to accumulate, at least partially, in
surface-associated aggregates. This resembles previous reports of
endogenous gelsolin immunostaining of rat fibroblasts
(17)
. In cells overexpressing G654A gelsolin, these
intensely staining aggregates were also seen along the neurite-like
extensions (Fig. 5
). Immunostaining with the secondary antibody alone did not display any
reactivity (data not shown). Analysis of gelsolin amounts in
fractionated cell lysates, however, did not reveal any differences in
subcellular localization of WT and G654A gelsolin (data not shown).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 19)
]. The functional role of gelsolin
became particularly evident from studies of gelsolin (-/-) knockout
mice (20)
. Although gelsolin was not required for normal
embryonal development or growth of the mouse, defective platelet shape
changes caused prolonged bleeding times, neutrophile migration was
impaired, and skin fibroblasts displayed excessive actin stress fibers
and lowered migratory capacity. Chen et al. (21)
also
reported that gelsolin plays a central role in epidermal growth factor
receptor-induced cell motility, and Azuma et al. (22)
recently provided evidence for gelsolin acting as a downstream effector
of rac (a member of the rho family of GTPases) in fibroblast membrane
ruffling. Together, these observations indicate that gelsolin activity
is required at sites where rapid modification of the membrane
actin-cytoskeleton takes place. Our findings demonstrate that gelsolin activity is involved in the regulation of morphological differentiation of a neural cell line. In fact, the extensive neural differentiation displayed by the Paju cell model offers a sensitive cellular readout for gelsolin activity. It is conceivable that strictly regulated gelsolin activity is required not only to create and maintain the extended actin cables forming a cytoskeletal core in neuronal processes, but also to maintain the growth cones representing the lamellipodia of sprouting neurons.
Overexpression of wild type and mutant gelsolin had a dramatic effect
on the behavior of the neural cell line Paju. This was already clearly
seen shortly after plating of the cells. Paju cells overexpressing WT
gelsolin remained rounded for a prolonged period of time, whereas cells
containing overexpressed mutant gelsolin rapidly acquired a flat
polygonic shape with developed actin stress fibers. This difference was
even more pronounced after induction of differentiation by treatment
with PMA. Whereas overexpressed WT gelsolin prevented neural sprouting,
the G654A mutated gelsolin did not inhibit the polymerization of long
actin filaments necessary for dendritic extensions to form. The
fundamental importance of gelsolin activity for neural differentiation
is also evident from the findings of Lu et al. (23)
.
Cultured hippocampal neurons from gelsolin (-/-) null mice were shown
to have delayed or absent retraction of filopodia on the extending
neurites as compared to hippocampal neurons from normal mice. An excess
of gelsolin would conceivably prevent the formation of filopodial
structures, as seen in Paju cells overexpressing WT gelsolin.
The different action of WT vs. G654A mutant gelsolin on neural
sprouting is not due to altered actin binding capacity. Coprecipitation
with antibodies to gelsolin brought down equal amounts of actin. Also,
no notable difference in intracellular distribution of overexpressed WT
and G654A gelsolin could be detected by indirect immunofluorescence or
cellular fractionation studies. Therefore, together with the previously
reported defective actin-severing activity, in vitro, of
plasma gelsolin isolated from FAF patients (12)
, it
appears that our findings may be attributed to defective in
vivo actin-severing activity of the G654A mutated gelsolin.
The cells overexpressing G654A gelsolin acquired a polygonic shape more rapidly than control Paju cells. This implies stabilization of the actin filaments, which may lead to reduced plasticity during development of the neural system in FAF patients. The slower cell cycle progress of Paju cells overexpressing G654A gelsolin is also in line with this notion (J. A. Westberg et al., unpublished observations). These observations suggest that mutant gelsolin may be exerting dominant negative activity. Moreover, our in vitro findings provide a pathogenetic mechanism for the dominant autosomal inheritance of FAF.
The effects seen by overexpression of mutant gelsolin may explain some
of the symptoms of FAF patients. It has been suggested that gelsolin is
involved in the lamellipodial movement of myelin-forming cells to wrap
axons by modulation of their actin polymerization (8)
. The
aberrant actin modulation of mutated gelsolin in FAF patients would
therefore impair the efficiency of the myelination of axons, which
could explain the electrophysiological findings of Kiuru and
Seppäläinen (13)
, i.e., signs of slow nerve
conduction and prolonged distal motor latencies suggestive of
demyelination. The deregulation of neural process formation in itself
would also contribute to the neuropathy of FAF patients. In conclusion,
it is evident that the neuropathy of FAF patients is not only caused by
amyloid deposits, but appears to arise as a combination of amyloid
deposits and a loss of function of the cellular gelsolin protein. These
findings, in turn, reveal an important function for gelsolin in normal
neural development.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Received for publication January 25, 1999. Revised for publication April 24, 1999.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. H. Song, C. X. Wang, D. K. Song, P. Wang, A. Shuaib, and C. Hao Interferon {gamma} Induces Neurite Outgrowth by Up-regulation of p35 Neuron-specific Cyclin-dependent Kinase 5 Activator via Activation of ERK1/2 Pathway J. Biol. Chem., April 1, 2005; 280(13): 12896 - 12901. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. J. Hernandez-Deviez, M. G. Roth, J. E. Casanova, and J. M. Wilson ARNO and ARF6 Regulate Axonal Elongation and Branching through Downstream Activation of Phosphatidylinositol 4-Phosphate 5-Kinase {alpha} Mol. Biol. Cell, January 1, 2004; 15(1): 111 - 120. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Kanungo, Z. Kozmik, S. K. Swamynathan, and J. Piatigorsky Gelsolin is a dorsalizing factor in zebrafish PNAS, March 18, 2003; 100(6): 3287 - 3292. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
