| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Short Communications |
-Secretase Site Mutation Develop a Myopathy Similar to Human Inclusion Body Myositis
From the Department of Pathology,*
Laboratory of
Neuropathology,
and Department of Comparative
Medicine,
University of Washington,
Seattle, Washington
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-secretase recognition and to preserve the Aß domain of
C99. The majority (87%) of the 24-month-old transgenic mice showed
myopathic changes, and approximately one-third of them had
degenerating fibers with sarcoplasmic vacuoles and
thioflavin-S-positive deposits. Ultrastructurally, the
inclusions were aggregates of short thin amyloid-like fibrils,
6 to 8 nm in diameter. These features are similar to those of human
IBM. Immunocytochemistry using an antibody against Aß showed
membranous staining in most muscle fibers of transgenic mice,
as well as granular or vacuolar cytoplasmic staining in the atrophic
fibers. Western blots showed a high level of accumulation of
carboxyl-terminal fragments of ßPP in the muscles of the transgenic
mice with the most severe IBM-like lesions. The expression of IBM-like
lesions was age dependent. These transgenic mice provide a model for
the study of IBM and for the peripheral expression of a key element in
the pathogenesis of Alzheimer disease.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1-antichymotrypsin,9
apolipoprotein
E,10,11
ubiquitin,12,13
cellular prion
protein,14
and presenilin-1.15
All of
them are proteins that are likely to play key roles in the degeneration
of neurons in AD and other neurodegenerative disorders. Therefore, AD
and IBM appear to share some common pathogenetic mechanisms, or at
least similar mechanisms of amyloidogenesis. There is strong evidence that intracellular events mediated by ßPP metabolites may cause neuronal death. Several laboratories have reported that, by overexpression, either full-length16,17 or the amyloidogenic carboxyl-terminal fragments (CTFs) of ßPP18,19 are neurotoxic in vitro and cause AD-like lesions in vivo. Our laboratory has studied the effects of a fusion protein termed SßC composed of the carboxyl-99 residues of ßPP fused to the 17-amino acid signal peptide of ßPP. SßC overexpression in neuron-like cells resulted in altered proteolysis and was associated with neurotoxicity.20-22 The toxic effects were related to CTFs of approximately 14 and 15 kd that contain intact Aß domains and are potentially amyloidogenic. These CTFs are normal metabolic products of ßPP in the human brain.23 Several ßPP mutations linked to familial AD cause intraneuronal accumulation of potentially amyloidogenic CTFs.24 So far, however, no intraneuronal amyloid has been found in these model systems.
Here we report a line of transgenic mice showing an abnormal accumulation of CTFs in the skeletal muscles, with associated lesions similar to IBM. This is the first animal model for this untreatable disease.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The K612V substitution in pCA-SßCK612V was made by modifying the
pCA-SßC construct21
using the Alter Sites Mutagenesis
System (Promega, Madison, WI). The pCA-SßCK612V vector encodes a
17-amino-acid signal peptide fused to C99 of ßPP containing a valine
in place of the lysine at the site of -secretase cleavage (K612V of
ßPP695). Expression is under the control of the cytomegalovirus
enhancer/chicken ß-actin promoter. The sequence of the entire
mutagenized cDNA was confirmed by automated DNA sequencing (Department
of Pharmacology, University of Washington). A 3.1-kb
SalI/PstI fragment containing the enhancer,
promoter, and mutated cDNA sequences was isolated and injected into the
fertilized eggs produced by mating between F1s of C57BL/6 and C3H mice.
Founders were backcrossed to C57BL/6 mice (three to four backcrossings
at the time of this analysis).
Western Blot Analysis
Quadriceps muscle from one side was homogenized in SDS-polyacrylamide gel electroporesis (SDS-PAGE) sample buffer. Quadriceps muscle from the other side was saved for histochemistry and immunocytochemistry. Fifty micrograms of each sample were run on a 16.5% SDS-PAGE gel and transferred to Immobilon-P (Millipore, Bedford, MA). The membrane was blocked with 5% nonfat milk and sequentially incubated with B994, a polyclonal antibody to the carboxyl-terminal 39 amino acids of ßPP (generously provided by Thomas R. Hinds) and a polyclonal antibody to desmin (Sigma Chemical Co., St. Louis, MO). The signal was detected using biotinylated secondary antibodies, Vectastain peroxidase (Vector Laboratories, Burlingame, CA), and the enhanced chemiluminescence system (Amersham, Arlington Heights, IL).
Histochemistry and Immunocytochemistry
Quadriceps muscles from control and transgenic mice were divided transversely into two halves. One half was immediately fixed in formalin, and some small pieces from it were fixed in 2.5% glutaraldehyde for electron microscopic examination. The other half was immediately frozen in isopentane cooled in liquid nitrogen. Adjacent cryostat sections were stained with H&E and modified Gomori's trichrome stains as well as NADH dehydrogenase reactions. Adjacent formalin-fixed, paraffin-embedded tissue sections were stained with H&E, Gomori's trichrome, thioflavin-S, myosin, and Aß immunocytochemistry. Standard peroxidase-anti-peroxidase25 or avidin-biotin complex (ABC)26 techniques with hematoxylin counterstain were used for immunocytochemistry. Tissue pretreated with formic acid27 was used for the immunocytochemistry applying Aß1-28 antibody (Athena Neurosciences, South San Francisco, CA).
Electron Microscopy
One-millimeter pieces of quadriceps were fixed in 2.5% glutaraldehyde in 0.1 mmol/L sodium cacodylate buffer. Each tissue sample was post-fixed in buffered osmium tetroxide and embedded in PolyBed (Polysciences, Warrenton, PA). Plastic sections (1 µm) were prepared from each tissue block, and these sections were stained with Azure II methylene blue. Thin sections were picked up on copper grids and stained with uranyl acetate and lead citrate. Each thin section was examined with a Philips 410 electron microscope at 80 kV.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Most molecules of ßPP undergo proteolytic release from the cell
membrane by the action of an -secretase that cleaves in the middle
of the Aß domain28
and abolishes the amyloidogenecity of
CTFs. To make the product of SßC construct more amyloidogenic, we
replaced lysine at amino acid 612 of ßPP-69529
with
valine (K612V mutation) and produced transgenic mice harboring the
SßCK612V construct. As the major action of the -secretase is
directed at either amino or carboxyl to K612, it was expected that the
K612V substitution would abolish much of the -secretase recognition
and preserve an intact Aß domain. This effect has been demonstrated
in a COS-1 cell transfection system.30
Our SßCK612V
construct, when transfected into P19 cells, directed the
over-expression of 10- to 12-kd proteins specifically recognized by
B994, a polyclonal antibody to the carboxyl-terminal 39 amino acids of
ßPP (M. Hearn, in preparation). We studied line 5590, which has two
unlinked arrays of SßCK612V transgenes, as revealed by Southern blot
analysis using BglII-digested genomic DNA probed with DNA
from the SßC construct. One array contained approximately three
copies and the other one copy of the transgene (data not shown).
Transgenic mice and nontransgenic sibs were sacrificed for histological
and immunochemical analysis at ages 9 to 24 months. By Western blots,
the relative abundance of the transgene protein products among various
organs were found to be similar to those of SßC mice without the
K612V mutation;31
the brain showed comparatively less
expression than the heart, intestine, kidney, and skeletal muscle (data
not shown). As in the case of SßC mice,31
no brain
lesions were found in SßCK612V mice of all ages. Modified
Bielschowsky silver and Aß immunocytochemical stains failed to
demonstrate any senile plaque formation in the transgenic brains.
Muscles of 24-Month-Old Transgenic Mice Showed a Myopathy Similar to Human IBM
Although the quadriceps muscle of the SßCK612V mice showed at
least a fivefold greater expression of the transgene protein product
(C99K612V) as compared with the brain (data not shown), they did not
show any lesions or any evidence of amyloid formation at ages 9 to 13
months. However, at 24 months of age, the transgenic mice exhibited
myopathic changes of varying severity in the quadriceps muscle. In the
mild cases, variation of muscle fiber size, increased central nuclei,
and split fibers were found. In more advanced cases, the quadriceps
demonstrated scattered angulated atrophic fibers, occasional necrotic
fibers with macrophage infiltration, occasional regenerating fibers
with basophilic cytoplasm, and markedly increased central nuclei (many
fibers showed a caravan pattern of central nuclei). The modified
Gomori's trichrome stain showed a patchy increase in subsarcolemmal
red granular staining, indicating an increased number of mitochondria
in these fibers. The trichrome stain did not show any increase in
interstitial fibrous tissue. These are nonspecific myopathic changes
and were seen in 26 of the 30 transgenic mice examined. Ten
nontransgenic sibs examined showed none of these changes, except for
one with a slight increase of subsarcolemmal mitochondria and two with
a few muscle fibers with central nuclei. Although the presence of
angulated atrophic fibers may indicate neurogenic atrophy, it is not
likely due to the absence of fiber type grouping in NADH dehydrogenase
reactions and myosin immunocytochemically stained sections (data not
shown). In 9 of 30 transgenic mice examined, more striking myopathic
changes were noted. As shown in Figure 1
,
in addition to the severe myopathic changes indicated above, there were
patchy lymphocytic infiltrates around endomysial blood vessels and
connective tissue (Figure 1A)
seen on H&E-stained paraffin sections as
well as cryostat sections. On H&E-stained cryostat sections, but not on
paraffin sections, there were sarcoplasmic vacuoles, which were oval or
irregular in small degenerating fibers with small basophilic granules
clinging to their edges (Figure 1C)
. The basophilic granules were also
irregularly scattered within these vacuole-containing fibers. These
granules stained red with a modified Gomori's trichrome stain (Figure 1D)
. Muscle fibers with vacuoles were much less frequent relative to
those with other myopathic features and were not found in the
nontransgenic sibs.
|
, but fibers with these
thioflavin-S-positive deposits were rather rare. Immunocytochemistry
using an antibody against Aß1-28 showed no staining in the control
nontransgenic muscles (Figure 1E)
whereas the muscles from the
transgenic mice showed pronounced sarcolemmal staining in most fibers
(Figure 1F)
. Anti-Aß1-28 also highlighted the cytoplasm of almost all
angulated atrophic fibers (Figure 1F)
, many of which showed
Aß-positive, vacuole-like inclusions (Figure 1G)
or coarse granules
(Figure 1H)
in approximately one-third of SßCK612V mice, the features
of which included angulated atrophic fibers, lymphocytic infiltration,
sarcoplasmic vacuoles, and thioflavin-S- and Aß-positive sarcoplasmic
deposits. This combination of pathological changes is similar to that
seen in human IBM. Sarcoplasmic Inclusions of Amyloid-Like Fibrils Were Seen in Degenerating Muscle Fibers by Electron Microscopy
We carried out electron microscopic examination to find additional
evidence of IBM-like pathology. On semi-thin sections stained with
Azure II methylene blue, myopathic changes corresponding to those found
on H&E sections were noted in transgenic mice. In particular, round to
oval, rather pale sarcoplasmic inclusion bodies were detected in six of
the nine transgenic mice with IBM-like lesions (Figure 2A)
. Corresponding to the sarcoplasmic
vacuoles seen on the cryostat sections, these inclusion bodies were
relatively less frequent (involving less than 1% of the muscle fibers
on sections) compared with other pathological features and tended to
occur in a small cluster, often subsarcolemmal (Figure 2D)
, in one
degenerating fiber. These inclusion bodies had a smooth contour with
dark granules of various sizes within or around the inclusions. The
muscle fibers harboring these inclusion bodies invariably were small,
with disorganized myofibrils and dense debris. Focal zones of
degeneration within one muscle fiber were frequent (Figure 2, A and C)
.
There were small clusters of dark granules in the degenerating fibers
(Figure 2B)
, which are the characteristic feature of human IBM on resin
sections.2
Specimens from two such transgenic mice and one
nontransgenic sib were studied by EM. The above inclusions consisted of
aggregates of short thin fibrils, 6 to 8 nm in diameter (Figure 2, E and F)
, not unlike the amyloid-like fibrils that were decorated
with antibodies against Aß in human IBM.5
The fibrils
were haphazardly arranged. The individual fibrils appear straight and
beaded with a poorly defined periodicity. There were also granular
materials admixed with the fibrils in the inclusions. The filamentous
inclusions were surrounded by osmiophilic membranous whorls and
cellular debris (Figure 2, D and E)
. The inclusions resided solely
in the sarcoplasm. No such inclusions were found in the control
muscles. Thus, EM confirmed the presence of inclusions composed
of amyloid-like fibrils in the degenerating muscle fibers from the
transgenic mice with severe muscle pathology.
|
Western blots were performed to estimate the relative abundance of
the transgene protein products in the quadriceps muscles of both
transgenic and nontransgenic mice. The CTFs of ßPP, ranging in size
from 10 to 12 kd, were visualized by B994, an antibody to the
carboxyl-terminal 39 amino acids of ßPP (Figure 3)
. CTFs were not present in the muscles
from the nontransgenic sibs at any ages. Muscles obtained from the
transgenic mice at ages ranging from 9 to 13 months showed a baseline
expression level (approximately five times that of the brain) of CTFs
(data not shown), although there were no pathological changes found in
these muscles. In contrast, the muscles obtained from the 24-month-old
mice showed a large variation in the abundance of CTFs above the
baseline. As the occurrence of the inclusions in the muscle sections
was relatively rare, it was impossible to make a correlation between
the number of inclusions and the abundance of CTFs. A subjective
three-tier histological grading of the severity of myopathic changes
with a consideration of all myopathic features was performed in a blind
fashion by the neuropathologist (L-W. Jin). The levels of CTFs were
normalized with the levels of desmin detected on the same blot. The
histological grades and the levels of CTFs did not correlate well (data
not shown). However, it is noteworthy that among the nine transgenic
mice with inclusions, five showed very high levels of CTFs (at least
five times the average) and more severe myopathic changes as well.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Among the aged (24 months old) transgenic mice, the phenotypes varied. At the protein level, the expression of CTFs in the muscles also varied widely among the transgenic mice. At the histological level, the changes varied from no lesions in 4 of 30 transgenic mice to IBM-like lesions in 9 of 30 transgenic mice. The level of CTFs and the severity of myopathy did not correlate well. One possible explanation for the lack of correlation is that the quadriceps for the Western blot analysis and the quadriceps for the microscopic examination were from different sides. In human IBM, the muscle lesions may be rather focal. There would be sampling errors if there was asymmetric involvement of IBM-like pathology in the transgenic mice, which may have also occurred in a rather focal fashion. For example, in one case we found moderately severe myopathy in the semi-thin sections but no lesions in the proportionally much larger H&E-stained sections. This focal nature of the lesions has posed difficulties for both adequate sampling for pathological evaluation and Western blot analysis of transgene expression. Other uncontrolled variables may have contributed to this variability. The most interesting possibility would be differential contributions of background alleles, as the transgenic mice were only backcrossed (to C57BL/6 mice) three to four times.
A heterogeneous group of ßPP molecules is generated by alternative splicing and post-translational modification as ßPP molecules transit the secretory pathway. Although the majority of ßPP molecules remain associated with internal membranes, especially the Golgi, some full-length ßPP is transported to the cell surface.36 The extracellular domains of approximately 30% of the ßPP molecules at the cell surface are secreted.37 The remainder of the molecules are internalized and either recycled or processed in the endosomal/lysosomal pathway. Internalization of ßPP is mediated through regions of the cytoplasmic domain.38,39 The SßCK612V mice express the CTFs of ßPP, which contains Aß as well as the cytoplasmic domain. Immunocytochemistry demonstrated the Aß-positive sarcolemmal staining in almost all fibers from transgenic mice. This is consistent with the data of Sisodia,30 who showed that ßPP-770 molecules with the same K to V substitution are associated with the plasma membrane instead of being secreted. Interestingly, the positive cytoplasmic staining was present only in the degenerating fibers, indicating abnormal metabolism of CTFs and accumulation of Aß-containing metabolites in the cytoplasm. In view of the low abundance of inclusions or amyloid deposition relative to the high levels of Aß epitopes in the degenerating fibers, most Aß-containing fragments in the cytoplasm were not in the amyloid form. It cannot be ascertained at this point whether the high level of Aß-containing fragments has caused individual fiber degeneration or the fiber degeneration has caused an accumulation of Aß-containing fragments. The accumulation of osmiophilic membranous whorls and cellular debris in or around the filamentous inclusions on EM may argue for toxicity induced by the amyloid-like filaments. However, in most fibers, myopathic changes (central nuclei, fiber splitting, and increased number of mitochondria) occurred in the absence of inclusions or cytoplasmic Aß positivity.
The extent to which the K612V mutation enhances the myopathy remains to be determined by quantitative comparison of the muscle pathology of mutant and wild-type constructs having similar degree of expression. The laboratory of K. Fukuchi (see related paper in this issue) has shown similar myopathy in the absence of the K612V mutation. Their transgenic mice,31 however, had a genetic background different from ours and, therefore, direct comparisons cannot be readily established. Our discovery of these lesions was made completely independently.
Similar to human IBM, age appears to be a risk factor for the development of lesions in the transgenic mice. Transgenic mice at ages of 9 to 13 months, although expressing high levels of CTFs in the muscles, had not developed any muscle lesions. It is possible that the high levels of CTFs in the sarcoplasmic membranes in the transgenic mice render them more susceptible to the oxidative stress that accumulates with age. Environmental factors, such as cumulative minor traumas, may contribute to the CTF accumulations with age.
The study of muscle degeneration and amyloidosis in IBM may yield insights into the mechanism of neuronal degeneration and amyloidosis in AD. The idea that amyloid can form intracellularly is gaining acceptance40,41 and may have a broader pathological significance. For example, intracellular amyloid has been observed in scrapie-infected hamster brains42 and in a case of telencephalic variant of Gerstmann-Sträussler syndrome.43 In this regard, our transgenic mice should facilitate the study of the pathogenesis of IBM and the mechanism of amyloid formation in general. Its greatest significance, however, may be its potential utility as a surrogate for investigation of the cerebral ß-amyloid cascade hypothesis as a primary mechanism in the neuronal lesions of AD and its pharmacological modulation.44 The biochemical and morphological evolution of the peripheral muscle lesions in our transgenic mice should be amenable to longitudinal studies via biopsy.
|
Acknowledgements |
|---|
|
Footnotes |
|---|
Supported by grants from the National Institute on Aging (T32AG00057, R35AG10917, P50AG05136, and P30AG13280).
The first two authors contributed equally to this paper.
Accepted for publication September 3, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1-antichymotrypsin colocalizes with ß-amyloid protein and ubiquitin in vacuolated muscle fibers of inclusion-body myositis. Acta Neuropathol 1993, 85:378-382[Medline]
This article has been cited by other articles:
 |
|
 |
|
  J. Schmidt, K. Barthel, A. Wrede, M. Salajegheh, M. Bahr, and M. C. Dalakas Interrelation of inflammation and APP in sIBM: IL-1{beta} induces accumulation of {beta}-amyloid in skeletal muscle Brain, May 1, 2008; 131(5): 1228 - 1240. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. E-H. Moussa, Q. Fu, P. Kumar, A. Shtifman, J. R. Lopez, P. D. Allen, F. LaFerla, D. Weinberg, J. Magrane, T. Aprahamian, et al. Transgenic expression of {beta}-APP in fast-twitch skeletal muscle leads to calcium dyshomeostasis and IBM-like pathology FASEB J, October 1, 2006; 20(12): 2165 - 2167. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. E. Finch A perspective on sporadic inclusion-body myositis: The role of aging and inflammatory processes Neurology, January 24, 2006; 66(1_suppl_1): S1 - S6. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. P. Murphy and T. E. Golde Inclusion-body myositis and Alzheimer disease: Two sides of the same coin, or different currencies altogether? Neurology, January 24, 2006; 66(1_suppl_1): S65 - S68. [Full Text] [PDF]
|
|
|
|
|
|
V. Askanas and W. K. Engel Inclusion-body myositis: A myodegenerative conformational disorder associated with A{beta}, protein misfolding, and proteasome inhibition Neurology, January 24, 2006; 66(1_suppl_1): S39 - S48. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L.-W. Jin, I. Maezawa, I. Vincent, and T. Bird Intracellular Accumulation of Amyloidogenic Fragments of Amyloid-{beta} Precursor Protein in Neurons with Niemann-Pick Type C Defects Is Associated with Endosomal Abnormalities Am. J. Pathol., March 1, 2004; 164(3): 975 - 985. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. C. Sugarman, T. R. Yamasaki, S. Oddo, J. C. Echegoyen, M. P. Murphy, T. E. Golde, M. Jannatipour, M. A. Leissring, and F. M. LaFerla Inclusion body myositis-like phenotype induced by transgenic overexpression of beta APP in skeletal muscle PNAS, April 30, 2002; 99(9): 6334 - 6339. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. C. Sugarman, T. R. Yamasaki, S. Oddo, J. C. Echegoyen, M. P. Murphy, T. E. Golde, M. Jannatipour, M. A. Leissring, and F. M. LaFerla Inclusion body myositis-like phenotype induced by transgenic overexpression of beta APP in skeletal muscle PNAS, April 30, 2002; 99(9): 6334 - 6339. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
