| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Emory University School of Medicine (S.D.B., Y.L., G.A.H., M.R.), Department of Orthopaedic Surgery, and Atlanta Veterans Affairs Medical Center, Decatur, Georgia 30033; Division of Endocrinology and Metabolism (F.L.T., M.S.N.), Emory University School of Medicine, Atlanta Veterans Affairs Medical Center, Decatur, Georgia 30033; and University of California San Francisco School of Medicine (J.A.H., D.H.), Department of Orthopaedic Surgery, San Francisco, California 94143
Address all correspondence and requests for reprints to: Scott D. Boden, M.D., The Emory Spine Center, 2165 North Decatur Road, Decatur, Georgia 30033.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Glucocorticoid, and subsequently BMP-6, was found to induce a novel rat intracellular protein, LIM mineralization protein-1 (LMP-1), that in turn resulted in synthesis of one or more soluble factors that could induce de novo bone formation. Blocking expression of LMP-1 using antisense oligonucleotide prevented osteoblast differentiation in vitro. Overexpression of LMP-1 using a mammalian expression vector was sufficient to initiate de novo bone nodule formation in vitro and in sc implants in vivo. These data demonstrate that LMP-1 is an essential positive regulator of the osteoblast differentiation program as well as an important intermediate step in the BMP-6 signaling pathway.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To better discern the unique physiological role of different BMP signaling proteins, we recently compared the potency of BMP-6 with that of BMP-2 and BMP-4, for inducing rat calvarial osteoblast differentiation (3). We studied this process in first passage (secondary) cultures of fetal rat calvarial osteoblasts that require BMP or glucocorticoid for initiation of differentiation. In this model of membranous bone formation, glucocorticoid (GC) or a BMP will initiate differentiation to mineralized bone nodules capable of secreting osteocalcin, an osteoblast-specific protein. This secondary culture system is distinct from primary rat osteoblast cultures that undergo spontaneous differentiation. In this secondary system, glucocorticoid treatment resulted in a 10-fold induction of BMP-6 messenger RNA (mRNA) and protein expression that was responsible for the enhancement of osteoblast differentiation (4).
Here we report on a novel positive regulator of rat osteoblast differentiation. Owing to the presence of two LIM finger structures in its sequence, its pattern of expression, and its role in formation of mineralized bone, we have named this protein LMP (for LIM mineralization protein). LIM domain proteins were originally named for the three homeodomain proteins in which they were first described: Lin-11, Isl-1, and Mec-3 (5, 6, 7).
LMP was identified in RNA from osteoblasts stimulated by glucocorticoid and isolated from an osteosarcoma complementary DNA (cDNA) library. Based on its association with bone development in vivo and on the results of suppression and overexpression experiments in vitro and in vivo, our findings indicate that LMP is an essential intracellular positive regulator of the osteoblast differentiation program. Furthermore, its temporal and spatial association with bone morphogenetic protein-6 (BMP-6) suggests that it may be involved in the signaling pathway of BMPs, a family of secreted proteins important in bone formation.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Calvarial cell preparation
Following approval by the Institutional Animal Care and Use
Committee, fetal Sprague Dawley rats were removed at 20 days gestation,
decapitated, and the heads submerged in sterile PBS with 1%
penicillin/streptomycin-5000U (Gibco BRL). The crania were
dissected using sterile technique in the laminar flow hood. Parietal
and frontal bones were dissected free from the sutures and subjected to
four collagenase digestions (type 1:type 2 = 6:1). The specific
activity of collagenase (Worthington Enzymes, Freehold, NJ) was 43
IU/ml in the first digestion and 172 IU/ml for the remaining three
digestions. All digestions were carried out at 32 C for 20 min each.
Cells from the latter two digestions were pooled to provide a rat
osteoblastic (rOB)-enriched cell suspension (3). The pooled cells were
washed, pelleted, resuspended in MEM/10% FBS, counted by
hemocytometer, and seeded in T-75 vented flasks (Corning, Inc., Corning, NY) at 1 x 106 cells/flask.
Cells were grown at 37 C in 5% CO2 with humidification.
The cells were fed at 48 h and again at 96 h with MEM/10%
FBS. Seven days after seeding, the primary culture was trypsinized and
passed into 6-well plates at 1 x 105 cells/35 mm
well) as first subculture or secondary cells. Secondary cultures were
grown for an additional 7 days during which they reached second
confluence (day 0). To initiate osteoblast differentiation in the
secondary cultures 1 nM triamcinolone acetonide, a
glucocorticoid (GC), was applied for 7 days; alternatively 50 ng/ml of
rhBMP was applied in certain experiments to initiate differentiation.
Beginning on day 0, media were changed and treatments (GC and/or
cytokines) were applied under a laminar flow hood every 3–4 days. The
standard culture protocol was as follows: days 0–7 = MEM, 10%
FBS, 50 µg/ml ascorbic acid; days 8–14 = BGJb
medium, 10% FBS, 5 mM ß-GlyP (as a source of inorganic
phosphate to permit mineralization). Endpoint analysis of bone nodule
formation and osteocalcin secretion was performed at day 14.
Quantitation of bone nodule formation
Cultures were fixed overnight in 70% ethanol and stained with
von Kossa silver stain. A semiautomated computerized video image
analysis system (Optomax 5, Optomax, Hollis, NH) was used to quantitate
nodules in each well. This automated technique was previously validated
against a manual counting technique and demonstrated a correlation
coefficient of 0.92 (P < 0.000001) (3). All data are
expressed as the mean ± SEM (SEM)
calculated from 5–6 wells at each condition. Each experiment was
reconfirmed at least two times using cells from different calvarial
preparations.
Quantitation of osteocalcin secretion
Osteocalcin levels in the medium were measured using a
competitive RIA with a monospecific polyclonal antibody (PAb) raised in
our laboratory against the C-terminal octapeptide of rat osteocalcin as
described previously except for use of an acetylated peptide analog as
radioligand and standard (8). Osteocalcin values were reported as
pmol/ml (nM) medium (3-day production). Values were
expressed as the mean ± SEM of triplicate
determinations for 5–6 wells for each condition. Each experiment was
reconfirmed at least two times using cells from different calvarial
preparations.
Differential display PCR
Secondary osteoblast cultures were prepared as previously
described and treated in the presence or absence of 1 nM
GC. Total RNA was isolated by our standard methods (4) and differential
display RT-PCR was performed using four RNAimage kits (GenHunter Corp.)
according to the manufacturer’s protocol. Radiolabeled PCR products
were fractionated by electrophoresis on a DNA sequencing gel; dried
gels were exposed overnight to Kodak X-OMAT film (Eastman Kodak Co., Rochester, NY). Bands of differentially expressed cDNAs
were excised from the gel, reamplified by PCR and the products were
cloned into the PCR-II vector (TA cloning kit,
Invitrogen). Sequence analysis of the PCR product revealed
a novel 260-bp cDNA fragment
(BLASTN;http://www.ncbi.nlm.nih.gov/).
Isolation and sequencing of clones
The 260-bp DNA fragment was random primer labeled
(Amersham Pharmacia Biotech, Piscataway, NJ) with
-(32P)-dCTP (New England Nuclear, Boston, MA) and used
to probe an osteoblast cDNA library (custom UMR 106 library,
Stratagene, La Jolla, CA, generously provided by
Dr. Laura Mauro). The library was plated (5 x 104
pfu/ml) onto agar plates, grown 8 h at 37 C, and filter membranes
(Duralon-UV, Stratagene) overlaid for 2 or 4 min onto the
plates. Filters were denatured, rinsed, UV cross-linked, prehybridized
for 2 h at 42 C, hybridized with the probe overnight at 42 C,
washed under moderately stringent conditions, and exposed to
Kodak X-OMAT film overnight. Four positive clones were
identified that hybridized strongly to the 260 bp probe. Positive
plaques were rescued as Bluescript SK(-) phagemids
(Stratagene). Nucleotide sequence of the clones were
obtained using the Amplicycle Sequencing Kit (Perkin-Elmer Applied Biosciences, Foster, CA) and sequence-specific
oligonucleotides as sequencing primers. Autoradiographic bands were
analyzed manually. A total of 8 primers were used to assemble a cDNA
consisting of 1696 bp.
RT-PCR
Secondary rat osteoblast cultures were untreated or treated with
1 nM GC or 50 ng/ml BMP-6 for the indicated times. Total
RNA was pooled and isolated from two 35-mm wells as previously
described (4), and triplicate samples from each treatment were analyzed
by RT-PCR. Briefly, cDNAs were generated from total RNA using MMLV
reverse transcriptase (Promega Corp., Madison, WI) and
oligo-dT17 primer. PCR was performed on one twentieth of
the total RT reaction using -(32P)-dCTP, Amplitaq DNA
polymerase (Perkin-Elmer Corp.), and specific primers for
LMP-1 (forward = 5'CCACGTATGAGCACCTCCTC3'; reverse =
5'CACAGCTACATACAGGTTTATTG3'. PCR was performed for 22 cycles (94 C,
30"; 58 C, 30"; 72 C, 20"). Products were separated by PAGE,
analyzed by Phosphorimager (Molecular Dynamics, Inc., Sunnyvale, CA), and their intensities were normalized to those of
glyceraldehyde phosphate dehydrogenase (GAPDH, a constitutive marker,
not regulated by treatments; forward = 5'CTGGTCATCAATGGGAAAC3';
reverse = 5'AAAGTTGTCATGGATGACC3'). BMP-2 primers: forward =
5'TATGCTCGACCTGTACCGC3'; reverse = 5'CACTTCCACCACAAACCC3'.
Cbfa-1 primers: forward = 5'CCAGATGGGA- CTGTGGTTACC3';
reverse = 5'ACTTGGTGCAGAGTTCAGGG3'. To evaluate the
expression of LMP-1 mRNA in other tissues, total RNA from various rat
tissues was purchased (Stratagene), treated with DNase I
(5 U, Gibco BRL) at 37 C for 20 min and RT-PCR was
performed as described above and repeated using another set of primers
specific for a unique region of LMP-1 (forward =
5'ATCCTTGCTCACCTCAC-GGG3'; reverse =
5'GCACTGTGCTGGTTTTGTCTGG3'.
Antisense oliogonucleotide blocking experiments
Specific antisense oligonucleotides were designed by computer
analysis, synthesized and HPLC purified by the Emory University
Microchemical Facility as follows: BMP-6 antisense:
5'CCTGTAGTGTCGTTGATCGT3' against sequence beginning 7 bp downstream
from the translation start site. LMP-1 antisense:
5'GCACTACCTTGAAGGAATCCATGGT3', spanned the putative translation start
site. Cbfa1 antisense: 5'TTGTGAGGCGAATGAAGCAT3'. NONSENSE–5'AGC
TTGTTGCTGAGTTGTCC3' had no significant homologies to any known rat
sequence. Osteoblast cultures were stimulated to differentiate using 1
nM GC. In addition, cultures were treated with antisense
oligonucleotides or random oligonucleotides (0.4 µM for
the first 7 days). Oligonucleotides were preincubated in MEM without
serum at least 10 min or for 2 h at 4 C and further diluted in MEM
containing 10% FBS (Hyclone Laboratories, Inc.) and 50
µg/ml ascorbic acid to achieve 0.4 µM. RNA was
harvested at the indicated time points and analyzed by RT-PCR.
Northern analysis and RNase protection assay
Total RNA (30 µg/lane) was electrophoresed using a
formaldehyde 1% agarose gel and osmotically transblotted to
Gene-Screen (DuPont NEN). Membranes were hybridized
overnight with a 625 bp LMP-1 probe radiolabeled by random primer
method (Boehringer Mannheim, Mannheim, Germany).
Membranes were washed under moderately stringent conditions and exposed
to Kodak X-OMAT film overnight.
A nearly full-length 1498 bp riboprobe of LMP-1 was prepared (Boehringer Mannheim Transcription Kit) by T7 initiation of complementary sequence polymerization in linearized LMP-1 PCR-II (Invitrogen) clones. Total RNA (30 µg/lane) was hybridized (Boehringer Mannheim RNase Protection Kit) to the riboprobe (2.5 x 104 cpm/µl), the mixture was digested and precipitated according to the manufacturer’s protocol. Digested products were separated by electrophoresis on a 6% urea gel; the gel was dried and exposed to Kodak X-OMAT film overnight.
Protein analysis
A polyclonal antibody to rat LMP-1 was generated. A hydrophillic
hexapeptide (Gln-Asp-Pro-Asp-Glu-Glu, amino acids 389–394), determined
using the OMIGA 1.0 (Intelligenetics, Inc.) subprogram
ANTIGEN, was synthesized and conjugated to keyhole limpet hemocyanin by
the Emory University Microchemical Facility; 100 µg of the conjugated
peptide was mixed with Freund’s complete adjuvant and injected into 10
sites sc along the back of an 8-week-old rabbit in accordance with
Emory University guidelines. Rabbits were reinjected 15 days later and
every 30 days for 5 months to generate acceptable LMP-1 antibody
titers. In selected experiments, BMP-2 protein levels were determined
using a specific antibody for BMP-2 (provided by Genetics Institute).
Western blot analysis (9) was performed on 50 µg of cell culture homogenate (cells and matrix for LMP-1, medium for BMP-2), separated by SDS-PAGE, and electrophoretically transblotted to PVDF filter paper (Bio-Rad Laboratories, Inc.). LMP-1 antiserum (diluted 1:250 in TBS/1% goat serum) was incubated with membrane (blocked with TBS/5% goat serum) for 36 h at 4 C and, after washing, was detected using biotinylated goat antimouse secondary IgG (1:500) followed by Streptavidin (1:500) and TMB developer (Kirkegaard & Perry Laboratories) under standard conditions. For in vitro translation studies, an 35S-labeled LMP-1 protein was synthesized in vitro from a pcDNA3.1His plasmid (Invitrogen) containing an LMP-1 cDNA insert using the Promega Corp. TNT T7 Quick Coupled Transcription/Translation System (no. L1170).
In situ hybridization
Tissue postfixation was performed with 4% paraformaldehyde-PBS
with 0.1% sodium borohydride followed by acetylation to reduce
nonspecific binding. Riboprobes were prepared by standard methods
(Promega Corp. Kit) using 35S initially with
the full length LMP-1 cDNA and repeated with a 650 bp cDNA (nucleotides
-70 to +580) that did not contain the LIM domain sequences of LMP-1 to
confer greater specificity. In situ hybridization was
performed overnight at 55 C with high stringency washes (60 C) and
RNase A. Slides were exposed to emulsion (Kodak) for
appropriate times, developed, fixed and counterstained with Hoechst
33258. Images were captured using a 3-chip color CCD camera (Optronics)
and stored as Adobe Photoshop files. Images were displayed as
superimpositions of the in situ hybridization signal
obtained with transmitted light overlaid on the blue nuclear stain
revealed by the fluorescence of Hoechst 33258 dye. Results were
repeated on several specimens and with riboprobes to different regions
of LMP-1 to ensure specificity.
Construction of LMP-1 expression vector and transient transfection
of cells
The 1696 bp cDNA for LMP-1 was excised from the PCR-II cloning
vector and ligated into the mammalian expression vector, pCMV2 (5.5 kb,
Invitrogen), by standard methods (10). A control vector
was constructed in which the LMP-1 cDNA was inserted in the reverse
orientation, so as to not be translated. These vectors were applied to
osteoblast cultures for 2 h on day 0, before any other treatment,
using 7.5 µl/well Superfect transfection reagent (3 mg/ml,
Quiagen) and a modification of the manufacturer’s
protocol. DNA was resuspended in MEM to 450 µl and vortexed 10 sec;
Superfect was added, the solution vortexed and incubated at room
temperature for 10 min. MEM/10% FBS (1 ml/well) was added, mixed and
applied to osteoblast cultures immediately. Following 2 h
incubation at 37 C, the Superfect/DNA mixture was removed by
aspiration, the cultures were washed and MEM/10%FBS/50 µg/ml
ascorbic acid was added to begin the usual differentiation protocol.
Nodule number and osteocalcin were determined 14 days post
transfection.
Preparation of LMP-1 conditioned medium
Transfections were performed using 6 µg pCMV2-LMP-1-forward or
pCMV2-LMP-1-reverse DNA per culture. The medium was removed 4 days
posttransfection, frozen overnight, concentrated (10-fold) and desalted
using a Centriprep 3 centrifugal concentrator (3000 mol wt cut off),
and frozen. The concentrate was resuspended in fresh complete medium to
its original concentration and applied to secondary osteoblast cultures
with or without 1 nM GC (day 0).
Transfection of bone marrow cells for in vivo bone induction
Marrow was extracted from the hindlimbs of 4- to 5-week-old
normal rats (rnu-/+) in MEM, centrifuged, and the red blood cells
lysed by resuspending the pellet in 0.83% NH4Cl in 10
mM Tris, pH 7.4. Remaining marrow cells were washed x
3 with medium and transfected for 2 h with 9 µg pCMV2-LMP-1 in
the forward or reverse (control) orientation/1.5–3 million cells.
Cells were washed x 2 with medium, resuspended to 30 million
cells/ml, and 100 µl of the suspension applied to a sterile sheet (4
x4 x 3 mm) of human devitalized bone matrix or a bovine type I
collagen disc (2 x 5 mm). Implants were surgically placed sc on
each side of the chest wall, abdominal wall, or on the skull of 4- to
5-week-old athymic rats (rnu/rnu). Animal care was in accordance with
institutional guidelines. The animals were euthanized after 4 weeks and
the explants were removed, fixed in 70% ethanol, and analyzed by
radiography and undecalcified histologic examination using Goldner
trichrome. These experiments were repeated in 14 animals spread over
three groups of animals with consistent results.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
).
|
The LIM protein induced by glucocorticoid, hereafter named LIM
Mineralization Protein-1 (LMP-1), showed 54.9% homology (51.4%
nucleotide homology) to the most similar rat LIM family member, ENH
(17). In addition, LMP-1 protein was 78.5% homologous (83.9%
nucleotide) with the most similar human LIM family member, enigma (16).
Although the amino acid homology of the two LIM domains in LMP-1 and
the corresponding domains in human enigma is 96.8%, human enigma (and
rat ENH) contain a third LIM domain that is not present in LMP-1. In
addition, the remainder of the protein is considerably different;
therefore, we do not know if LMP-1 is the rat homologue of human enigma
(Fig. 1B
). The remainder of LMP-1 (nonLIM domain regions) had less than
25% overall homology with other rat LIM proteins.
To examine the time course of LMP-1 expression in our in
vitro model of membranous bone formation, we studied mRNA
expression using RT-PCR. Experiments with glucocorticoid (GC) as the
stimulus for osteoblast differentiation revealed a peak increase in
LMP-1 message at 48 h (Fig. 2A). In
light of earlier data showing that glucocorticoid treatment increased
BMP-6 mRNA levels by 6 h, we hypothesized that the increase in
LMP-1 message by GC could be mediated by BMP-6. Using RT-PCR, we
demonstrated that BMP-6 treatment resulted in a 4-fold increase in
LMP-1 message by 2 h (Fig. 2B), which increased to 30-fold on day
3 and returned to baseline on day 7 (Fig. 2C). We further established
the downstream relationship of LMP-1 to BMP-6 by showing that the
GC-induced increase of LMP-1 mRNA was blocked by the addition of
antisense oligonucleotide to BMP-6.
|
). GC induction of LMP-1 mRNA was
confirmed by the RNase protection assay (Fig. 3B). Western blot
analysis using an LMP-1 polyclonal antibody demonstrated that GC
stimulated an approximately 50 kDa band (Fig. 3C). Thus, the increase
in steady-state LMP-1 mRNA levels resulted in increased synthesis of
LMP-1 protein. In vitro translation of LMP-1 cDNA in a
rabbit reticulocyte lysate system also yielded a 50-kDa protein,
confirming the size of LMP-1 (data not shown).
|
). Rat embryonic day
14 (e14) and mouse e15 sections showed LMP-1 transcripts in mesenchymal
tissue surrounding the cartilaginous anlage of immature bones, and in
the future joint spaces (Fig. 4, A, B, and D). At the same time during
embryonic development, BMP-6 was expressed in the hypertrophic
cartilage region in the center of the long bone anlage and in the joint
spaces (Fig. 4C) as has been previously reported (18). Later, as
endochondral ossification progressed (Fig. 4L) and the hypertrophic
cartilage zone began to be replaced by mineralized bone, LMP-1
expression appeared in the mineralizing portion of the bone (Fig. 4, F
and G) adjacent to the BMP-6 expression domain in the hypertrophic
chondrocytes (Fig. 4E).
|
, H,
I, and J). In addition, LMP-1 transcripts were found in
mesoderm-derived bones of the skull base and neural crest-derived
endochondral bones such as the proximal mandible (Fig. 4, I and K).
LMP-1 is an essential positive regulator of osteoblast
differentiation
To explore the potential functional role of LMP-1 during
membranous bone formation, we synthesized an antisense oligonucleotide
to block LMP-1 mRNA translation and treated secondary osteoblast
cultures that were undergoing differentiation initiated by
glucocorticoid. LMP-1 antisense oligonucleotide inhibited mineralized
nodule formation and osteocalcin secretion (measured 14 days later) in
a dose-dependent manner (Fig. 5) similar
to the effect of BMP-6 oligonucleotide (4). The LMP-1 antisense block
in osteoblast differentiation could not be rescued by addition of
exogenous BMP-6 (50 ng/ml for 6 days), whereas the BMP-6 antisense
oligonucleotide inhibition was overcome, by addition of BMP-6, to 84%
of GC-stimulated levels (data not shown). This experiment further
confirmed the downstream position of LMP-1 relative to BMP-6 in the
osteoblast differentiation pathway.
|
LMP-1 Overexpression induces osteoblast differentiation in vitro
via secretion of a soluble factor
From the experiments described above, we concluded that LMP-1 was
necessary for osteoblast differentiation. We then performed experiments
to determine if forced expression of LMP-1 could enhance GC-stimulated
bone formation. We therefore cloned the LMP-1 cDNA into a pCMV
expression vector in the forward and reverse (control) orientations.
Overexpression of pCMV-LMP-1 in secondary fetal calvarial cells
stimulated with GC resulted in a 3- to 5-fold enhancement of
mineralized nodule formation and osteocalcin secretion (Fig. 6, A–D).
|
, E–H)
comparable with that induced by glucocorticoid. Control cultures
treated with overexpression of LMP-1 in the reverse orientation or the
empty vector did not undergo osteoblast differentiation. Western blot
analysis of cellular protein confirmed a 50-fold induction of LMP-1
protein 72 h after application of LMP-1 cDNA. The induction of
osteoblast differentiation by LMP-1 overexpression was blocked by
treatment with LMP-1 antisense oligonucleotide confirming the
functional importance of LMP-1 expression. Osteoblast cultures were
exposed to 3 µg pCMV2-LMP-1 and treated for 7 days with 1
µM antisense oligonucleotide to LMP-1 or nonsense
oligonucleotide for the next 7 days. The LMP-1 antisense
oligonucleotide completely inhibited osteocalcin secretion and
mineralized nodule formation measured on day 14, whereas the nonsense
oligonucleotide had no apparent effect on the GC-stimulated
differentiation (Fig. 7). RT-PCR
demonstrated consistent high levels of LMP-1 PCR product even in the
presence of LMP-1 antisense oligonucleotide, suggesting that the
oligonucleotide inhibition was not due to nonspecific cell toxicity. In
addition, Western blot analysis confirmed that the LMP-1 antisense
oligonucleotide inhibited LMP-1 protein synthesis (Fig. 7, inset).
|
). A time course
analysis suggested that at least 72 h were required for the LMP-1
transfected cells to condition the medium. The enhancement of
differentiation by conditioned medium mimicked the effect caused by
overexpression of LMP-1.
|
). This conditioned
medium experiment designed to rescue osteoblast differentiation
confirmed the lack of toxicity of the LMP-1 antisense oligonucleotide
treatment and further strengthened the evidence for the existence of a
secreted factor.
Bone marrow cells transfected with LMP-1 cDNA induce fone formation
in vivo
To determine if expression of LMP-1 could induce bone in
vivo, we transfected bone marrow cells with pCMV-LMP-1 and
implanted them on devitalized bone matrix placed on the chest of
athymic rats. Radiographs of explants at 4 weeks revealed extensive
bone formation in the implants which contained cells transfected with
the LMP-1 gene in the forward orientation, whereas cells transfected
with the reverse-oriented LMP-1 cDNA failed to form bone (Fig. 9, A and B). In 16 animals, 16/16
implants containing cells transfected with the LMP-1 (forward) cDNA
induced bone formation, whereas none (0/16) of the implants containing
cells with the control cDNA induced bone. Histology revealed new bone
trabeculae lined with osteoblasts in the LMP-1-transfected implants and
absence of new bone with partial resorption of the carrier in the
controls (Fig. 9, C–F). These studies demonstrate the feasibility of
local gene therapy using the LMP-1 gene to induce new bone formation
in vivo.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Transient expression of an antisense LMP-1 construct in rat osteoblast cells suppresses LMP-1 expression and prevents osteoblast differentiation in these cells. In addition, forced overexpression of LMP-1 in osteoblast cultures enhances ongoing differentiation and induces de novo differentiation of resting (unstimulated) osteoblast cultures. The in vitro effects are easily replicated in vivo, as demonstrated by ectopic bone formation in rats, even with relatively poor gene transfection efficiency. These profound physiologic effects are most likely facilitated by the secretion of a soluble factor induced by LMP-1 expression.
Role of LMP-1 in bone formation
LMP-1 appears to be involved with both of the two fundamental
mechanisms of bone formation. During membranous bone formation (seen in
embryonic flat bones such as the mandible and cranium), osteoblasts
differentiate, mature, and synthesize bone matrix which mineralizes.
This is the type of direct bone formation most closely simulated by our
osteoblast culture systems. In endochondral bone formation (seen in
embryonic long bone development and recapitulated postnatally during
fracture repair), chondrocytes proliferate, hypertrophy, and by some
unknown mechanism initiate the ingrowth of blood vessels and osteoblast
precursors resulting in a conversion of the cartilage matrix to
trabecular bone. This later bone formation phase of endochondral
ossification is similar to membranous bone formation.
BMP-6 is known to be expressed in hypertrophic chondrocytes, but its precise role has been unclear. The expression patterns of BMP-6 and LMP-1 suggest that the BMP-6 secreted by the long bone hypertrophic chondrocytes may induce the expression of LMP-1 in perichondral cells. The expression of LMP-1 in perichondral cells at e15 with subsequent expression in cells within the primary ossification center at e17 implies that LMP-1 in perichondral cells may facilitate the migration of osteoblastic cells into the primary ossification center. LMP-1 may represent a critical signal involved in coordinating the linkage between cartilage maturation and initiation of bone formation. Thus, BMP-6, acting through LMP-1, seems to provide an important signal to initiate membranous bone formation, the final phase of endochondral ossification. The expression of LMP-1 in bones of neural crest and mesoderm origin, as well as in both intramembranous and endochondral bone suggests that LMP-1 may be involved in a final common pathway of bone formation.
Possible mechanism of action of LMP-1
The data presented here strongly support a critical role of LMP-1
in the regulation of the complex program of osteoblast differentiation.
The association of LMP-1 with other proteins known to be important to
bone formation enables us to begin to build a more complete temporal
sequence of events comprising the pathway of osteoblast
differentiation. The observation that LMP-1 is regulated by BMP-6 and
not by BMP-2 or BMP-4 is consistent with its unique role early in
osteoblast differentiation because BMP-6 is the earliest of these three
BMPs to be expressed during differentiation in our culture system and
in vivo (4, 24).
Given the small number of cells actually transfected, the striking physiologic effects suggest that overexpression of LMP-1 results in the synthesis of an unidentified soluble factor or factors which act on cells in the osteoblast lineage causing them to differentiate and secrete BMP-2, a growth factor whose effects are well documented in the osteoinductive process (25, 26). The secretion of a soluble factor was confirmed by the conditioned medium experiments. In addition, the finding that LMP-1 up-regulates the recently described osteoblast transcription factor cbfa1 is further evidence for the vital role of this LIM protein.
The precise intracellular mechanism of action of LMP-1 is unknown. This protein appears to activate an intracellular switch that induces secretion of soluble factors that initiate and promote osteoblast differentiation during embryogenesis and adult bone formation. LMP-1, stimulated by BMP-6, may initiate the transition from hypertrophic cartilage to primary bone formation during endochondral ossification. Because this protein is also involved in membranous bone formation, it may be part of a final common pathway linking the processes of endochondral and membranous ossification. Studies are underway to further elucidate the relationship of LMP-1 with bone formation, BMP signaling, and the function of LMP-1 in tissues other than bone.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received May 15, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-Interferon inhibits 1,25-Dihydroxyvitamin D3
stimulated synthesis of bone gla protein in rat osteosarcoma cells by a
pretranslational mechanism. Endocrinology 127:588–594[Abstract]
This article has been cited by other articles:
 |
|
 |
|
  T. Camarata, B. Bimber, A. Kulisz, T.-L. Chew, J. Yeung, and H.-G. Simon LMP4 regulates Tbx5 protein subcellular localization and activity J. Cell Biol., July 31, 2006; 174(3): 339 - 348. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Sangadala, S. D. Boden, M. Viggeswarapu, Y. Liu, and L. Titus LIM Mineralization Protein-1 Potentiates Bone Morphogenetic Protein Responsiveness via a Novel Interaction with Smurf1 Resulting in Decreased Ubiquitination of Smads J. Biol. Chem., June 23, 2006; 281(25): 17212 - 17219. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Loughran, N. C. Healy, P. A. Kiely, M. Huigsloot, N. L. Kedersha, and R. O'Connor Mystique Is a New Insulin-like Growth Factor-I-regulated PDZ-LIM Domain Protein That Promotes Cell Attachment and Migration and Suppresses Anchorage-independent Growth Mol. Biol. Cell, April 1, 2005; 16(4): 1811 - 1822. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Minamide, S. D. Boden, M. Viggeswarapu, G. A. Hair, C. Oliver, and L. Titus Mechanism of Bone Formation with Gene Transfer of the cDNA Encoding for the Intracellular Protein LMP-1 J. Bone Joint Surg. Am., May 28, 2003; 85(6): 1030 - 1039. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Hannallah, B. Peterson, J. R. Lieberman, F. H. Fu, and J. Huard Gene Therapy in Orthopaedic Surgery J. Bone Joint Surg. Am., June 1, 2002; 84(6): 1046 - 1061. [Full Text]
|
|
|
|
|
|
Y. G. Amaar, G. R. Thompson, T. A. Linkhart, S.-T. Chen, D. J. Baylink, and S. Mohan Insulin-like Growth Factor-binding Protein 5 (IGFBP-5) Interacts with a Four and a Half LIM Protein 2 (FHL2) J. Biol. Chem., March 29, 2002; 277(14): 12053 - 12060. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Viggeswarapu, S. D. Boden, Y. Liu, G. A. Hair, J. Louis-Ugbo, H. Murakami, H. S. Kim, M. T. Mayr, W. C. Hutton, and L. Titus Adenoviral Delivery of LIM Mineralization Protein-1 Induces New-Bone Formation in Vitro and in Vivo J. Bone Joint Surg. Am., March 1, 2001; 83(3): 364 - 364. [Abstract] [Full Text]
|
|
|
|
 |
|
 L. Gilbert, X. He, P. Farmer, S. Boden, M. Kozlowski, J. Rubin, and M. S. Nanes Inhibition of Osteoblast Differentiation by Tumor Necrosis Factor-{alpha} Endocrinology, November 1, 2000; 141(11): 3956 - 3964. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Vallenius, K. Luukko, and T. P. Makela CLP-36 PDZ-LIM Protein Associates with Nonmuscle alpha -Actinin-1 and alpha -Actinin-4 J. Biol. Chem., April 6, 2000; 275(15): 11100 - 11105. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||
