| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Laboratory of Genetics, National Institute of Mental Health (T.B.U., J.H.), Basic Neurosciences Program, National Institute of Neurological Diseases and Stroke (T.V., G.H., E.M.), National Institutes of Health, Bethesda, Maryland 20892; and Endocrine Unit (G.S.), Massachusetts General Hospital, Boston, Massachusetts 02114
Address all correspondence and requests for reprints to: Ted Usdin, Laboratory of Genetics, NIMH, Building 36/Room 3D06, 36 Convent Drive MSC4094, Bethesda, Maryland 20892-4094. E-mail: usdin{at}codon.nih.gov
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Northern blots show that PTH2 receptor messenger RNA (mRNA) is most abundant in the brain and that it is also present in lung, pancreas, placenta, and testis (1). In situ hybridization histochemistry reveals that many more tissues express PTH2 receptor mRNA, and that it is expressed by distinct and often quantitatively minor cell populations within those tissues (17). In the cardiovascular system, it is expressed by vascular endothelium and smooth muscle, endocardium, and myocardium. In the gastrointestinal tract scattered cells that, based on morphology and distribution, appear to be mucus-producing cells and endocrine cells express PTH2 receptor mRNA. In the testis it is expressed by sperm, especially within the head of the epididymis, and it is also present within some ovarian follicles. Within the kidney, its mRNA was detected within a very small number of cells near the vascular pole of glomeruli.
We have now developed an antibody specifically recognizing the PTH2 receptor. We thought that it was important to confirm that the PTH2 receptor protein was expressed by cells where its mRNA was previously detected because protein expression does not necessarily parallel that of the mRNA encoding it. Labeling with the antibody also led to detection of cells not previously known to express the PTH2 receptor, and double-labeling contributed to their identification.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Cell culture and Western blots
HEK293 cells stably expressing the human PTH1 or PTH2 receptor
have previously been briefly described (19). Incubation with 1
µM PTH produces an approximately 50-fold stimulation of
cAMP accumulation in either cell line, 1 µM PTHrP
produces similar stimulation in only the PTH1 receptor expressing
cells, and there is no significant stimulation by either peptide in
nontransfected HEK293 cells. Saturation analysis using binding of
125I-rPTH (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34) to membranes prepared from these cells
indicates a receptor density of approximately 1 pmol/mg protein in each
cell line (Hoare, S., and T. Usdin, unpublished observations). For
Western blots P2 membranes were prepared from confluent plates of
HEK293 derived cell lines. Pellets were suspended directly in gel
loading buffer or first digested with PNGase F (New England Biolabs, Inc., Beverly MA) according to the suppliers protocol.
Electrophoresis and transfer to nitrocellulose membranes were performed
according to the protocols supplied with the 10% Nu-PAGE gels and
transfer buffer (Novex, San Diego, CA). Membranes were
stained with Ponceau-S to verify even transfer, the positions of
molecular weight standards marked, and then blocked by incubation in
Blotto (18) for 60 min followed by incubation with primary antisera,
and then horseradish peroxidase coupled secondary antibody for 1 h
diluted in Blotto. Antibody binding was detected using enhanced
chemiluminescence (SuperSignal Ultra; Pierce Chemical Co.).
Immunohistochemical methods
Immunostaining protocols and reagents are described in detail on
the world wide web
(http://intramural.nimh.nih.gov/lcmr/snge/Protocols/IHH/immuno.html).
Standard indirect immunofluorescence or avidin-biotin horseradish
peroxidase histochemistry (ABC) was performed on 4% paraformaldehyde
perfused 12-µm thick cryostat sectioned tissue. A few sections are
from tissue frozen and sectioned before fixation. This material was
postfixed in 4% paraformaldehyde and is noted in the relevant figure
legends. Tissue was from 150–200 g Sprague Dawley rats
(Taconic, Germantown, NY) or rat embryos of the noted ages
except for the mouse bone described (see Fig. 10
). Fixed,
decalcified, paraffin sectioned mouse femur was obtained from Molecular
Histology Laboratories (Rockville, MD). It was deparaffinized in xylene
and then rehydrated through decreasing concentrations of ethanol and
incubated in PBS before labeling, as performed for other tissues.
|
In situ hybridization histochemistry. The in situ hybridization data presented in this manuscript are from detailed reexamination of material generated in a previous study (17).
Immunological reagents
Fluorescent secondary antibodies were indocarbocyanine (Cy3),
fluorescein isothiocyanate (FITC), or aminomethylcoumarin acetate
conjugates of donkey immunoglobulin prepared for multiple labeling
(Jackson ImmunoResearch Laboratories, Inc., West Grove,
PA). Labeling with horseradish peroxidase used Vectastain
ABC elite reagents (Vector Laboratories, Inc., Burlingame,
CA). Antibody 10A8 recognizing MG160 (a Golgi selective marker (20, 21)
used at 1:100) was a gift of Nicholas K. Gonatas (University of
Pennsylvania). Mouse monoclonal antibody to caveolin-3 (used at 1
µg/ml) was from Transduction Laboratories, Inc.
(Lexington, KY). Rat monoclonal antibody to somatostatin was from
PharMingen (San Diego, CA). Rabbit antibody to
somatostatin (used at 1:400) was from INCSTAR Corp.
(Stillwater, ME). Guinea pig antibody to insulin (used at 1:2000) was
from INCSTAR Corp. Rabbit antibody to histidine
decarboxylase (used at 1:2000) was from Euro-Diagnostica (Malmö,
Sweden).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
) or a horseradish
peroxidase coupled secondary antibody (not shown). Preimmune serum does
not label the PTH2 receptor-expressing cells, and no labeling of either
the parent HEK293 cells (not shown) or HEK293 cells stably expressing
the human PTH1 receptor (which should express the same endogenous
epitopes as the parent cells) is detected. Similarly, there is intense
labeling of 20–30% of COS-7 cells transfected with PTH2 receptor cDNA
but no labeling of cells in mock transfected cultures (not shown).
Several bands are labeled in Western blots of PTH2 receptor-enriched
membranes (Fig. 2), probably representing
a combination of multiple glycosylation states and aggregation or
oligomerization of the receptor. The highest mobility major band
migrates with an apparent molecular weight of 84K, consistent
with the size seen following western blotting of a C-terminal epitope
labeled PTH2 receptor (Clark, J., and T. Usdin, unpublished
observations), and labeling of the receptor with a radioactive
photoaffinity ligand (2). Following digestion with PNGase F, the
mobility of the high mobility major band increases to an apparent
molecular weight of 63K, consistent with the predicted size of the
protein based on its cDNA sequence. No signal is seen in membranes
prepared from the parent HEK293 cells or ones expressing the PTH1
receptor. The limited sequence identity between the antigen and the rat
PTH1 receptor is identical to that with the human PTH1 receptor, and no
significant labeling is detected in rat kidney tubules (see Fig. 12).
Absorption of the antibody with the peptide used to generate it
eliminates tissue labeling (see examples shown in Figs. 5 and 10
), and
specific staining is absent when preimmune serum is used to label
tissue (see examples shown in Figs. 9 and 11).
|
|
|
|
|
|
, a–c). This position is characteristic
of several noninsulin-producing islet cell types. Double labeling shows
that the PTH2 receptor positive cells are somatostatin producing cells.
There is precise coincidence of labeling by PTH2 receptor and
somatostatin directed antibodies, but no overlap between the PTH2
receptor positive cells and antibody staining for insulin (Fig. 3c),
pancreatic polypeptide, or glucagon (not shown), which are present in
distinct cells (22, 23, 24).
|
).
|
), providing additional support for the
suggestion of low level persistence of exocrine expression of the PTH2
receptor in adult rats.
Thyroid gland
Parafollicular cells within the thyroid gland are labeled by the
PTH2 receptor antibody (Fig. 6). They
comprise a numerically minor population of cells within the thyroid
gland and, like the D-cells in pancreatic islets, were not obvious
following in situ hybridization. When slides from the
previous in situ hybridization study were reexamined,
increased grain density over cells with the distribution of
parafollicular cells was apparent (Fig. 7c). Double labeling with antibodies to
somatostatin and the PTH2 receptor demonstrates that many of the PTH2
receptor antibody-labeled cells contain somatostatin (Fig. 7, a and
b).
|
|
, d–g). Mucin-producing cells
(not shown), identified by their characteristic morphology and
distribution, are labeled in the gastric epithelium. In situ
hybridization previously demonstrated labeling of additional cells,
which seemed likely to be endocrine cells, based on their frequency and
location within the epithelium. This was confirmed, and the identity of
two of the cell types established by double labeling. There is partial
overlap here between labeling by the PTH2 receptor antibody and a
somatostatin directed antibody. There is also colocalization of
labeling by an antibody to histidine decarboxylase, which has recently
been established as a marker for gastrin secreting cells (25), and the
PTH2 receptor antibody. In addition, parasympathetic ganglion cells in
the submucosal and myenteric plexuses are distinctly labeled by the
antibody throughout the gastrointestinal tract (Fig. 8).
|
, h–j). Much of the labeling seems to be associated
with cell junctions, and could be within intercalated disks, although
we cannot be certain of this without additional markers or
immunoelectronmicro-scopy.
Vasculature in most tissues is labeled by the PTH2 receptor antibody
consistent with previous observations made using in situ
hybridization. Labeling of embryonic aorta is shown in Fig. 9.
Bone and cartilage
Chondrocytes in tracheal cartilage are clearly and intensely
labeled by the PTH2 receptor antibody (Fig. 6). Bone has relatively
high autofluorescence and endogenous peroxidase activity, but using
affinity purified PTH2 receptor antibody we are able to detect specific
labeling of some cells within bone (Fig. 10). The labeling has a punctate
pattern, like that of labeling by this antibody in other tissues,
whereas tissue autofluorescence has a more homogeneous appearance. The
punctate labeling is eliminated by absorption of the antibody with the
peptide antigen (Fig. 10c). Based on their distribution, the labeled
cells seem to be primarily chondrocytes in the growth plate and
subarticular cartilage. Expression is particularly strong in developing
bone (Fig. 11).
Kidney
Using in situ hybridization histochemistry, we
previously observed one or two cells expressing PTH2 receptor mRNA near
the vascular pole of glomeruli. The same pattern of staining is seen
with the PTH2 receptor recognizing antibody, and in this case the
signal to noise ratio is much better (Fig. 12).
Other tissues (not shown)
PTH2 receptor labeling in other tissues generally confirms the
distribution we previously determined from in situ
hybridization histochemistry (17). The most intense labeling is of
neurons within a limited number of nuclei in the brain, as demonstrated
by in situ hybridization. There is no labeling in the
pituitary gland by the PTH2 receptor antibody. There appears to be weak
labeling of a small population of cells within the adrenal medulla.
There is also weak labeling throughout the zona glomerulosa of the
adrenal cortex. Within the parathyroid gland, a very minor population
of cells, which may be oxyphils, appeared to be weakly labeled by the
PTH2 receptor antibody, and a similar labeling pattern is seen
following in situ hybridization. PTH2 receptor antibody
labeling is present in pulmonary bronchioles, some cells within both
the white and red pulp in the spleen, and supporting cells (not
neurons) in sympathetic ganglia. PTH2 receptor protein is detected in
the testis and epididymis with the same general pattern as previously
observed for expression of PTH2 receptor mRNA. However, the intensity
of the antibody staining, relative to the intensity of the in
situ hybridization signal in the testis, is lower than that in
other organs.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The antibody used in this study was raised against a peptide sequence within the C-terminus of the PTH2 receptor. Search of GenBank reveals no other proteins, or open reading frames, which contain this or a similar peptide sequence. Evidence for the specificity of the antibody labeling includes the following: 1) Using immunohistochemistry, the antibody labels tissue culture cells expressing the PTH2 receptor but not the parent cell lines or cells expressing the PTH1 receptor. 2) The same specificity is seen on Western blots of membranes prepared from these cells. 3) The size of the bands detected on Western blots is consistent with that expected for this receptor. 4) The tissue distribution of cells labeled by the antibody is consistent with that determined for PTH2 receptor mRNA by in situ hybridization histochemistry. 5) The antibody labeling is eliminated by preincubation of the antibody with the peptide antigen. Formally, the staining seen with this antibody must be considered PTH2 receptor-like immunoreactivity. However, we are confident that the immunoreactivity represents the PTH2 receptor in the majority of cells and tissues, where in situ hybridization demonstrates the corresponding mRNA. In several areas, including pancreatic islets and the thyroid gland, we saw labeling by the PTH2 receptor antibody where we had not previously observed an in situ hybridization signal. Guided by the immunocytochemistry we did see a clear hybridization signal on careful reexamination of the original material. The sensitivity of detection with this antibody, especially after affinity purification, is significantly greater than in situ hybridization. Thus, in chondrocytes we do not have corroboration of PTH2 receptor expression by an independent technique. The PTH2 receptor antibody labeling was blocked by the peptide used to raise the antibody, but we cannot completely exclude the possibility of a cross reacting antigen.
One of the most striking, and completely new, observations made in this study is precise overlap of expression of the PTH2 receptor and somatostatin within pancreatic islets. This labeling of somatostatin cells was the strongest signal observed, and it lead us to look for colocalization of the PTH2 receptor and somatostatin in other tissues. We observed partial overlap within the gastrointestinal tract and within thyroid parafollicular cells. Because activation of the PTH2 receptor in vitro causes cAMP accumulation (1), and stimulation of cAMP formation or treatment with cAMP analogs cause somatostatin release from pancreatic D cells (27), it seems logical that PTH2 receptor activation should stimulate or enhance somatostatin release. While somatostatin is generally an inhibitory factor, and influences release of both endocrine and exocrine pancreatic products (28, 29, 30), the precise physiological role of pancreatic somatostatin remains an area of active investigation. A biphasic effect of PTH on insulin secretion has been described (9). It is possible that the PTH2 receptor is responsible for the higher dose inhibition of insulin release via stimulation of somatostatin release. Pancreatic islet function is disturbed in both uremia and primary hyperparathyroidism (9, 31, 32, 33) and now a role of the PTH2 receptor in these conditions, through modulation of somatostatin synthesis or release, must be considered.
PTHrP, which acts on the PTH1 but not the PTH2 receptor, is found in all islet cell types and appears to regulate the formation and development of pancreatic islets (34, 35). There has not yet been cellular localization of the PTH1 receptor in islets. Regardless of whether PTHrP acts on the PTH1 receptor or an as yet unidentified receptor, it seems likely that its effects are quite different than those mediated by the PTH2 receptor, because the PTH2 receptor has such a distinct expression by one cell type.
We previously observed PTH2 receptor mRNA in the pancreas on Northern blots (1). It is not clear whether the small number of islet cells labeled by the PTH2 receptor directed antibody are responsible for that signal. In situ hybridization generated a weak signal throughout the exocrine pancreas suggesting that low level expression of PTH2 receptor mRNA by a large number of cells could be responsible for the band detected on Northern blots. Consistent with this, there also appears to be very weak labeling by the PTH2 receptor antibody in the adult exocrine pancreas and stronger labeling in the developing pancreas.
Thyroid parafollicular, or C, cells are labeled by the PTH2 receptor-directed antibody. This quantitatively minor cell population was not identified during our initial in situ hybridization survey. When that material was reexamined in light of the immunohistochemical observations, specific labeling of cells with the characteristic distribution of parafollicular cells was apparent. A subset of the PTH2 receptor antibody labeled cells contain somatostatin, as previously reported for parafollicular cells (36). Because not all somatostatin containing cells are labeled with the PTH2 receptor antibody, it seems likely that not all parafollicular cells express significant levels of the PTH2 receptor. Heterogeneity of parafollicular cells is well known (36). Calcitonin has counter regulatory effects to PTH. C cell release of calcitonin and serotonin is regulated by a calcium sensing receptor similar to the one that regulates PTH release from parathyroid cells (37). It is an intriguing possibility that the PTH2 receptor provides another regulatory input to these cells, perhaps stimulating calcitonin release when PTH levels are high. The PTH2 receptor could also be involved in regulation of other endocrine functions of parafollicular cells.
A question left unanswered by our in situ hybridization study (17) was whether the PTH2 receptor is present in the skeletal system. PTH2 receptor antibody labeling is clearly present in tracheal cartilage. We also see labeling of cells in bone growth plates and subarticular cartilage. Based on location, most of these cells appear to be chondrocytes. Because of the high autofluorescence and endogenous peroxidase activity of bone, we were not able to perform useful double labeling with the current reagents and techniques; thus, the potential expression of PTH2 receptors by other bone cells and the precise identity of the PTH2 receptor expressing cells remains to be addressed. The labeling pattern seen overlaps with the distribution of the PTH1 receptor demonstrated by in situ hybridization (38, 39). Mutation of either the PTH1 receptor or the peptide PTHrP has dramatic effects on bone and cartilage maturation (7, 40, 41). As in other areas of apparent PTH2 receptor expression, selective ligands and genetic studies will be required to determine its physiological function.
PTH2 receptor expression in parafollicular cells and chondrocytes suggests that it may be involved in development or maintenance of the skeletal system, or in calcium metabolism. Parafollicular cells are derivatives of the ultimobranchial body, a calcium regulating organ of phylogenetically older animals. Jüppner and colleagues (42) used the PTH1 receptor as a probe and identified a receptor more homologous to the PTH2 receptor in zebra fish. They did not detect a PTH1 receptor-like gene. These observations suggest that the PTH2 receptor may be part of phylogenetically older regulatory systems.
The PTH2 receptor has a punctate distribution in many cells. In the heart, where we have examined this most closely using confocal microscopy and double labeling with an antibody directed at caveolin, these clusters appear to be on the plasma membrane. Use of additional markers and/or electron microscopy may be required to further define the receptors subcellular localization. This aggregation could reflect the effects of tonic activation by a circulating ligand. A similar distribution is seen in tissue culture cells expressing the PTH2 receptor, and in this case incubation with PTH causes redistribution to a population of larger, clearly perinuclear, spots (unpublished observations).
Many of the cells that express the PTH2 receptor are hormone secreting, and the colocalization with somatostatin is striking. The PTH2 receptor could be involved in the specific regulatory functions of each of the cells that express it. These possibilities need to be experimentally addressed. There are reports of PTH action in many of these tissues including the heart, vasculature, and exocrine and endocrine pancreas but PTHrP effects have also been described. PTH activates the PTH2 receptor in tissue culture cells expressing it, but relatively high concentrations are required. Depending on the experimental conditions (buffer, incubation time, temperature, phosphodiesterase inhibition, etc.) we have observed EC50’s for PTH stimulation of cAMP accumulation ranging from 0.5–10 nM. Because the receptor density, efficiency of coupling, and second messenger systems may differ in the cells that endogenously express the PTH2 receptor, it is difficult to predict what concentration of PTH is required to cause a physiological effect. Circulating levels of PTH are reported to be in the picomolar range. We recently obtained evidence for another peptide that activates the PTH2 receptor (19). Either PTH or a novel ligand, or both, could be the endogenous effector(s) of the PTH2 receptor. Based on the broad, yet cell-specific, distribution demonstrated in this study the PTH2 receptor is likely to have a significant physiological, and possibly clinical, role.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received December 16, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  T. B. Usdin, M. Paciga, T. Riordan, J. Kuo, A. Parmelee, G. Petukova, R. D. Camerini-Otero, and E. Mezey Tuberoinfundibular Peptide of 39 Residues Is Required for Germ Cell Development Endocrinology, September 1, 2008; 149(9): 4292 - 4300. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Verhoeven and K. De Gendt Tuberoinfundibular Peptide of 39 Residues: A Neuromodulator Starting a Second Career in the Control of Meiosis Endocrinology, September 1, 2008; 149(9): 4289 - 4291. [Full Text] [PDF]
|
|
|
|
 |
|
 T. M. Murray, L. G. Rao, P. Divieti, and F. R. Bringhurst Parathyroid Hormone Secretion and Action: Evidence for Discrete Receptors for the Carboxyl-Terminal Region and Related Biological Actions of Carboxyl- Terminal Ligands Endocr. Rev., February 1, 2005; 26(1): 78 - 113. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. R. Papasani, R. C. Gensure, Y.-L. Yan, Y. Gunes, J. H. Postlethwait, B. Ponugoti, M. R. John, H. Juppner, and D. A. Rubin Identification and Characterization of the Zebrafish and Fugu Genes Encoding Tuberoinfundibular Peptide 39 Endocrinology, November 1, 2004; 145(11): 5294 - 5304. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Miao, B. He, B. Lanske, X.-Y. Bai, X.-K. Tong, G. N. Hendy, D. Goltzman, and A. C. Karaplis Skeletal Abnormalities in Pth-Null Mice Are Influenced by Dietary Calcium Endocrinology, April 1, 2004; 145(4): 2046 - 2053. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Sugimura, T. Murase, S. Ishizaki, K. Tachikawa, H. Arima, Y. Miura, T. B. Usdin, and Y. Oiso Centrally Administered Tuberoinfundibular Peptide of 39 Residues Inhibits Arginine Vasopressin Release in Conscious Rats Endocrinology, July 1, 2003; 144(7): 2791 - 2796. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Eichinger, N. Fiaschi-Taesch, T. Massfelder, S. Fritsch, M. Barthelmebs, and J.-J. Helwig Transcript Expression of the Tuberoinfundibular Peptide (TIP)39/PTH2 Receptor System and Non-PTH1 Receptor-Mediated Tonic Effects of TIP39 and Other PTH2 Receptor Ligands in Renal Vessels Endocrinology, August 1, 2002; 143(8): 3036 - 3043. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Dobolyi, H. Ueda, H. Uchida, M. Palkovits, and T. B. Usdin Anatomical and physiological evidence for involvement of tuberoinfundibular peptide of 39 residues in nociception PNAS, January 24, 2002; (2002) 42416199. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. L. Ward, C. J. Small, K. G. Murphy, A. R. Kennedy, M. A. Ghatei, and S. R. Bloom The Actions of Tuberoinfundibular Peptide on the Hypothalamo-Pituitary Axes Endocrinology, August 1, 2001; 142(8): 3451 - 3456. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Piserchio, T. Usdin, and D. F. Mierke Structure of Tuberoinfundibular Peptide of 39 Residues J. Biol. Chem., August 25, 2000; 275(35): 27284 - 27290. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. R. J. Hoare, J. A. Clark, and T. B. Usdin Molecular Determinants of Tuberoinfundibular Peptide of 39 Residues (TIP39) Selectivity for the Parathyroid Hormone-2 (PTH2) Receptor. N-TERMINAL TRUNCATION OF TIP39 REVERSES PTH2 RECEPTOR/PTH1 RECEPTOR BINDING SELECTIVITY J. Biol. Chem., August 25, 2000; 275(35): 27274 - 27283. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Dobolyi, H. Ueda, H. Uchida, M. Palkovits, and T. B. Usdin Anatomical and physiological evidence for involvement of tuberoinfundibular peptide of 39 residues in nociception PNAS, February 5, 2002; 99(3): 1651 - 1656. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||
