| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Department of Pathology (E.v.G., H.J.A.d.P., J.F.S., E.H.J.W., J.A.S.), and Department of Clinical Sciences of Companion Animals (G.R.R., J.A.M., J.A.S.), Faculty of Veterinary Medicine, Utrecht University, 3508 TD Utrecht, The Netherlands; and Department of Anatomy and Histology (E.H.), Faculty of Veterinary Medicine, SLU, S75007 Uppsala, Sweden
Address all correspondence and requests for reprints to: E. van Garderen, Department of Pathology, Faculty of Veterinary Medicine, Utrecht University, P.O. Box 80.158, 3508 TD Utrecht, The Netherlands. E-mail: e.vangarderen{at}vet.uu.nl
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Second, the canine GHR was characterized at the molecular level. Northern blot analysis revealed a major GHR transcript of approximately 4.2 kb. The coding sequence of the canine GHR shows extensive homology with the GHR of several species. Seminested RT-PCR (using primers annealing in exons 4–5, exon 6, and exon 9) generated, next to the primary product, four different products in mammary tissues and the canine mammary tumor cell line CMT-U335, which seemed to be alternative GHR transcripts. These alternative GHR transcripts were characterized by exon 8 skipping, exon 7 skipping, and use of alternative splice donor and acceptor sites. Especially, the transcript that is missing exon 8 may encode a GH binding protein. In most malignant mammary samples, only the primary transcript was present; and alternative transcripts could not be detected. The absence of alternative GHR transcripts in mammary carcinomas, and thus putative inhibitors of GH-induced signal transduction, may contribute to enhanced sensitivity of malignant tumors to GH.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Whereas in various species (rabbit, human, mouse, rat, sheep, cow, chicken, pig, and monkey) (4), the GHR has been studied, there are no reports addressing the canine GHR. Recently, research from our group has identified the progesterone or progestin-exposed canine mammary gland as a major site of extrapituitary GH production (5, 6, 7). In addition, GH expression was found in benign and malignant mammary tumors, which may indicate that locally produced GH can have a pathobiological role both in organogenesis and tumorigenesis of the mammary gland. Because GH does not bind to the PRL receptor (PRLR) in the dog (8), locally produced GH can only be effective in the presence of the GHR. Therefore, to further substantiate the importance of the GH/GHR axis in mammary tissue, we began investigations to evaluate the presence and the characteristics of the GHR in canine mammary tissue, canine mammary tumors, and canine mammary tumor cell lines. In addition, we investigated the tissues and cell lines for the presence of alternative GHR transcripts, which, if present, may indicate that there is a molecular basis for the generation of GHBPs in the dog.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
From the selected tissues, 5-µm-thick sections were cut and collected onto poly-L-lysin-coated slides (Sigma Diagnostics, St. Louis, MO). After this, slides were deparaffinized and rehydrated. Immunostaining was performed using the avidin-biotin peroxidase method. For antigen retrieval, sections were immersed in trypsin (0.1% trypsin in 0.1% CaCl2 in distilled water, pH 7.4) for 20 min at room temperature. Endogenous peroxidase activity was blocked by incubation with 1% H2O2 in methanol 100% for 30 min. Nonspecific antibody binding was blocked by preincubation with 10% normal horse serum for 15 min at room temperature. The primary antibodies used were a commercial monoclonal antibody against the rat/rabbit GHR (Mab 263, Agen Biomedical Ltd, Queensland, Australia), which recognizes a GH-binding site with high affinity (9, 10). In addition, a GHR-unrelated monoclonal antibody, raised against human melanomas (HMB 45, DAKO Corp., Carpinteria, CA) and of the same isotype (IgGlk) as Mab 263, was used as a control for aspecific binding of IgG1k to the tissue. After blocking endogenous peroxidase activity, sections were sequentially treated with Mab 263, used at a concentration of 25 µg/ml in PBS, overnight, with biotinylated horse antimouse IgG (Vector Laboratories, Inc., Burlingame, CA) for 30 min (diluted 1:125 in PBS) and with ABC solution for 30 min, each in a moist chamber at room temperature. Immunoreactivity was visualized using 0.3% H2O2 and 3,3-diaminobenzidine tetrachloride (Sigma) diluted in 0.05 M Tris-HCI, pH 7.6, for 10 min at room temperature. In between steps, sections were washed three times in PBS/Tween and once in PBS. Sections were counterstained with Mayer’s hematoxylin for 1 min, dehydrated with ascending ethanol series and xylene, and mounted with glass coverslips using Eukitt. The specificity of immunoreactivity was verified by absence of immunostaining when primary and secondary antibodies were omitted and when Mab 263 was substituted with HMB45 in the same dilution as Mab 263.
RNA isolation
Total RNA was isolated from frozen mammary tissues and mammary
tumor cell lines using TRIzol (Life Technologies, Inc.,
Grand Island, NY). RNA concentrations were determined by UV
spectrophotometry. Two micrograms of isolated RNA was separated on a
1% agarose gel to assess the quality by the presence of 28S and 18S
ribosomal RNA. The quality of RNA isolated from all samples was
satisfactory.
Mammary tissue was obtained from canine patients that were referred to the Department of Clinical Sciences of Companion Animals for the evaluation of mammary nodules. Postsurgically, a section of the nodular tissue was fresh-frozen in liquid nitrogen and stored at -80 C until use. The remaining sections of the nodules were used for histopathological diagnosis. Total RNA was isolated from 23 tissues, including 12 nontumorous mammary samples in different stages of glandular development, 1 benign mammary tumor, 7 malignant mammary tumors, and 3 lymph node metastases.
Canine mammary tumor cell lines that were examined for the presence of GHR messenger RNA (mRNA) were the CMT-U335 (isolated in the laboratory of Dr. Hellmén), the SH3 (11), the SH15, the SH27, and the P95/168 (all three isolated in our laboratory).
Complementary DNA (cDNA) cloning
First, a canine mammary cDNA library was made. Total RNA from
normal canine mammary tissue was used to construct this library with
the Smart PCR cDNA synthesis kit (CLONTECH Laboratories, Inc., Palo Alto, CA). This kit preferentially enriches the cDNA
library for full-length cDNAs and overcomes the general problem of
previous methods in which the 5'-ends of the genes tended to be
underrepresented in the library. One microliter of the library was
subjected to PCR to amplify the cDNA encoding the full-length canine
GHR, using the primer combination GHRcds5'/GHRcds3' (see Table 1
). These primers are deduced from the
coding sequence of the full-length human GHR. The specificity of the
resulting PCR product was confirmed by Southern blot analysis and was
cloned into pBluescript (Stratagene GmbH, Heidelberg,
Germany). To prepare for sequencing, the product was digested by
EcoRI, HindIII, and BamHI into smaller
fragments, which were then subcloned in pGEM-T easy (Promega Corp., Madison, WI).
|
In the first reaction, the thermal cycle consisted of a single RT step of 45 min at 50 C, followed by 5 min at 95 C. Further amplification of the cDNA products was obtained in 30 cycles of denaturation for 30 sec at 94 C, annealing for 30 sec at 50 C, and extension for 30 sec at 72 C. Samples were kept at 72 C for 10 min after the last cycle.
Southern blot analysis
Five percent of the (seminested) PCR product was separated on a
1% agarose gel and was transferred to Hybond-N+
(Amersham Pharmacia Biotech, Buckinghamshire, UK)
by capillary blotting. DNA was cross-linked to the membrane using UV
light (UV Stratalinker 1800, Stratagene, La Jolla, CA),
followed by washing the blot twice in 2x SSC. Then, the blot was
prehybridized for 2 h at 65 C in 100 ml PEG Hybmix (0.25
M NaPi, 0.25 M NaCl, 7% SDS, 10% PEG 6000, 1
mM EDTA), containing 10 mg herring sperm DNA. During
prehybridization, the probe was radioactively labeled
(
-[32P]deoxy-ATP) by random prime labeling. This probe
was a GHR-specific cDNA fragment of 471 bp, which was produced in our
laboratory using canine liver tissue. This fragment was validated by
sequencing and had 86% homology, compared with the human GHR sequence
in exons 6–9. Hybridization was performed overnight at 65 C. After
this, the blot was washed in 3x SSC, 1x SSC, and 0.2x SSC.
Hybridization was visualized on film, during 4 h of exposure at
room temperature.
Northern blot analysis
From each sample, poly A+ mRNA was isolated using the Oligotex
mRNA Midi kit (QIAGEN, Valencia, CA), according to the
manufacturer’s protocol. Five micrograms of poly A+ mRNA was
glyoxalated and separated on a 1% agarose gel in 20 mM
phosphate buffer and transferred to a Hybond-N+ membrane by
capillary blotting. 
-DNA, cut with HindIII, was
glyoxylated and used as a marker. The GHR probe, also used in the
Southern blot analysis, was radioactively labeled with
-[32P]deoxy-ATP by random prime labeling.
Hybridizations were performed as described before (12), and
autoradiographs were produced by exposure of the membrane to film
(Hyperfilm MP, Amersham Pharmacia Biotech) for 72 h
using intensifying screens. Then, the blot was stripped and
rehybridized with a ß-actin probe as a control for the amounts of RNA
loaded.
Cycle sequencing
First, the PCR product was purified using the QIAquick PCR
Purification Kit (QIAGEN Inc.). Sixty nanograms of
purified PCR product was used as template for the cycle sequencing
reaction with the BigDye Terminator Cycle Sequencing Kit
(Perkin-Elmer Corp., Foster, CA) using the GHR6a or
the GHR9b primer (Table 1
). The cycle sequence conditions were: 5 min
at 94 C, followed by 30 cycles of denaturation for 30 sec at 94 C,
annealing for 30 sec at 50 C, and extension for 4 min at 60 C. The
sequence was determined using the ABI PRISM system (Perkin-Elmer Corp.).
Alignment of sequences. Alignment of sequences was performed with the computer software Lasergene (DNASTAR Inc., Madison, WI).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, see also Materials and
Methods).
|
In atrophic glandular tissue (n = 2), typical for anestrus,
predominantly ductal structures were present. The epithelium of these
ducts displayed diffuse immunoreactivity. Also, the myoepithelial cells
that underly the ductal epithelium were diffusely immunopositive (Fig. 1A).
In the proliferation phase of the glandular tissue, reflecting early-
and middiestrus (n = 7), the tissue is characterized primarily by
the presence of ductal budding structures and by further elongation and
branching of the ductal system. Ductal epithelial cells, including the
epithelial cells in ductal buds, were diffusely immunopositive (Fig. 1B). Immunoreactivity was also extensively found in the myoepithelial
cell population in this tissue. Columnar epithelial cells of
nonsecreting alveoli were generally immunopositive, but occasionally
single cells in alveoli were immunonegative.
In the differentiation phase (n = 9), occurring in late
diestrus, the tissue is characterized by the presence of lobuloalveolar
structures, in which secretion may occur. Both ductal epithelium and
myoepithelial cells of inter- and intralobular ducts displayed diffuse
immunostaining. Alveolar epithelial cells, however, demonstrated
heterogenous immunoreactivity. Here, columnar nonsecretory epithelium
was, in general, immunopositive; whereas in the flattened epithelial
cells of alveoli filled with secretum, the cells were either weakly
immunoreactive or were immunonegative (Fig. 1C).
Tumorous mammary tissue. Of tumorous mammary tissue, 12 benign
tumors and 7 malignant tumors were examined. Tumors were classified
according to the World Health Organization classification (17). The
benign tumors were comprised of complex tumors and 1 fibroadenoma. In
the dog, complex adenomas are the most frequently occurring benign
mammary tumors and consist of both epithelial and spindle cell
proliferations. The spindle cell component probably has a myoepithelial
origin. In these complex tumors, immunostaining was observed in both
the epithelial and the spindle cell components (Fig. 1D). The
epithelial component showed, in general, a more prominent and
homogeneous immunoreactivity. In chondroid metaplasia that occurred in
one complex/mixed adenoma, chondroblast-like cells displayed intense
immunoreactivity.
The malignant mammary tumors comprised only simple-type tumors,
including four solid carcinomas, two tubular adenocarcinomas, and one
anaplastic carcinoma. Obviously, carcinoma cells were heterogeneously
immunostained (Fig. 1F). In the adenocarcinomas particularly and in one
case of solid carcinoma, immunonegative areas alternated with clusters
of immunopositive cells. Metastatic tumor cells in lymphatic vessels
were also immunopositive. Interestingly, in carcinomas, a nuclear
immunostaining was sometimes observed in tumor cells (Fig. 1G). Also,
in the anaplastic carcinoma, tumor cells were intensely immunoreactive.
Desmoplastic changes in tumor stroma present in the adenocarcinomas
showed that the activated fibroblasts were also immunopositive.
Molecular characterization of the GHR
Full-length GHR. The PCR product, using as template the canine
mammary cDNA library and the primers GHRcds5'/GHRcds3', was 1946 bp and
hybridized with the canine GHR probe. After sequencing, the canine GHR
coding sequence could be aligned with the sequence of the human,
porcine, and rabbit GHR without any insertions or deletions. To
analyze the canine sequence, we numbered the exons corresponding
to the homologous exons in the human GHR. In the human GHR gene, there
are 9 coding exons (exons 2–10). Exon 1 is noncoding. Exon 2 encodes
the final 11 bp of the 5'-untranslated regio (UTR), the 18-amino
acid signal sequence, and the first 5 aminoacids of the extracellular
hormone-binding domain. Exons 3–7 together encode the majority of the
extracellular hormone-binding domain. Exon 8 encodes the final 3 amino
acids of the hormone-binding domain, the 24-amino-acid hydrophobic
transmembrane domain, and the first 4 amino acids of the intracellular
domain. Exons 9 and 10 together encode the remaining 346 cytoplasmic
amino acids. The remaining part of exon 10 is noncoding (approximately
2-kb) (4).
The ATG startcodon in the canine sequence is located at position 12, and the stopcodon is found at position 1926 in the PCR product, resulting in a coding sequence of 1914 bp. This encodes a protein of 638 amino acids. This is consistent with the characteristics of the human GHR described above, including the GHR signal peptide of 18 amino acid residues and 620 residues in the human plasma membrane GHR. Also compared with the rabbit and porcine GHR sequence, the canine GHR has extensive homology, both on nucleotide and on amino acid levels. The coding sequence of the canine full-length GHR has been submitted to GenBank (accession number AF133835).
The translated canine GHR features six extracellular cysteines, which are linked by disulfide bonds in the human GHR. The first of these disulfide bonds is required for GH binding and also contains the epitope for Mab 263 (18). The YGEFS motif, another important site for the binding of GH (19), is present in both the human and the translated canine GHR. This motif is located just proximal to the membranous part of the GHR.
Primary and alternative GHR transcripts in exons 6–9. Because
exon 8 encodes the membranous part of the protein, alternative
transcription in exons 6–9 may result in the generation of proteins
that are unable to anchor in the membrane, yet retain the ability to
bind GH, giving rise to GHBPs. To investigate whether these alternative
GHR transcripts were present, a seminested PCR was performed. In all
normal and tumorous mammary tissues and in the mammary tumor cell lines
CMT-U335, SH3, SH27, and P95/168, a predominant PCR product of 471 bp
was amplified, which corresponds to the primary, normal-spliced GHR
transcript. The SH15 cell line, however, was negative for any of the
GHR transcripts. In addition to this primary product, in nontumorous
mammary tissue and in the benign tumor, additional PCR products were
present of 380 bp and 305 bp, which showed a positive hybridization in
the Southern blot analysis, indicating that these were alternative GHR
transcripts. Also, in CMT-U335 cells, there was a primary 471-bp
product and three different transcripts of 316 bp, 305 bp, and 277 bp
(see Fig. 2). Remarkably, in nearly all
malignant mammary samples, including the lymph node metastases, only
the primary 471-bp product was present. Only in one carcinoma, there
was an additional PCR product that comigrated with the 380-bp product
(not shown).
|
.
|
, in
normal canine mammary tissue and the mammary tumor cell lines CMT-U335,
P95/168, SH27, and SH3, a GHR transcript that comigrated with the
marker at approximately 4.2 kb was detected. The SH15 cell line,
however, was negative. In the normal mammary tissue sample and the
CMT-U335 cell line, an additional transcript seemed to be present at
approximately 4.1 kb. In the normal mammary sample, this band may
represent the alternative transcript characterized by exon 8
skipping; especially this transcript, which was identified in
the RT-PCR experiments, will be approximately 4.1 kb. No additional
transcripts of the size that have been found in the rat and mouse (1.2
kb), specifically coding GHBPs in these rodents, were found.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The immunohistochemical results in the present study, obtained with Mab 263 in canine mammary tissues, are in agreement with findings that have been reported for rat mammary tissues (28) and human mammary tissues (30). In canine nontumorous mammary tissue, myoepithelial and epithelial cells of inter- and intralobular ducts, including ductal budding structures, were diffusely immunoreactive in all stages of glandular development. Alveolar epithelium present in the later stage of the luteal phase showed a heterogenous immunoreactivity; nonsecreting columnar alveolar epithelium was generally immunopositive, whereas low cuboidal epithelium of alveoli, filled with secretum, displayed a strongly diminished or absent GHR expression. These findings are consistent with data reported in rat mammary tissue (28), including a markedly reduced level of immunoreactivity in alveolar cells during lactation, particularly in the later stages. Therefore, we conclude that whereas GH expression in canine mammary tissue is found in epithelial cells in the early- and midluteal phases of the ovarian cycle (7), GHR expression seems only to be down-regulated in completely differentiated alveolar epithelial cells at the end of the diestrus phase.
The findings in canine mammary tumors are in agreement with GHR expression found in benign and malignant human breast tissues (30). Also, in canine mammary tumors, we found extensive GHR immunoreactivity, both in benign and malignant samples, that was irrespective of the histopathological subtype. However, GHR expression in carcinomas was heterogenous, characterized by the presence of both immunopositive and immunonegative tumor cells. As in canine normal mammary tissue, the presence of GHR in canine mammary tumors is an important finding, in view of the frequently observed GH production by such tumors. Though GH expression in canine mammary tumors of the complex type was found to be restricted to the epithelial component, with virtual absence of reactivity in the spindle cell component (7), we now find that, in complex mammary tumors, GHR is expressed both by the epithelial and the spindle cell part. Consequently, locally produced GH might function both as an autocrine and paracrine growth factor.
In addition to the membranous and cytoplasmic GHR immunoreactivity, we sometimes observed a nuclear immunostaining of cells, which has also been frequently reported by others (21, 22, 24, 29). This nuclear translocation of the GHR indicates that GH might have a direct nuclear effect, in addition to ligand-dependent activation of cytoplasmic signal transduction pathways.
The molecular characterization of the canine full-length GHR in mammary tissue revealed extensive homology with the GHR sequence of several species, including the human GHR, the rabbit GHR, and the porcine GHR. In the present study, the 5'-UTR of the canine GHR, including a putative exon 1, has not been analyzed. It is now well documented that the 5'-UTRs of the GHR gene in several species is characterized by a marked heterogeneity. For example, eight different 5'-UTRs have been detected in the human gene, and five 5'-UTRs were found in the rat gene (for a review on 5'-UTR heterogeneity, see Ref. 4). The significance of multiple 5'-UTRs is not clear, but it is suggested that 5'-UTR exons are associated with unique promoters.
In addition to the primary product, we also found evidence for the presence of alternative GHR transcripts, which had resulted from alternative splicing between exons 6–9. In one of these, exon 8 was skipped. This alternative transcript particularly might encode a GHBP, because the transmembrane domain is missing, whereas the YGEFS motif is conserved. All other alternative transcripts, however, lack this motif. At the moment, it is not clear whether these alternative transcripts are translated into GHBPs. Although we found CMT-U335 cells positive for GHR/GHBP in immunocytochemistry (picture not shown), we were not able to demonstrate either GHR protein or GHBPs in cell lysate and supernatants of CMT-U335 in Western blot analysis. This, however, is probably caused by technical failure, because Mab 263 is not suited for Western blot analysis.
Northern blot analysis of polyadenylated mRNA of normal mammary tissue and most of the investigated canine mammary tumor cell lines revealed the presence of a GHR transcript of approximately 4.2 kb, which is in the range of sizes of the GHR transcripts described in other species (4). The additional transcripts of approximately 4.1 kb in normal mammary tissue and in CMT-U335 cells may be related to the alternative processings of the GHR gene that were identified in the RT-PCR experiments. We also conclude from our Northern blot analysis that, during routine cell culture, significant amounts of mRNA coding the GHR can be present in canine mammary tumor cell lines. This correlates well with recent results in our laboratory in which we found that a GH-induced signal transduction path, indicating the presence of a functional GHR, could be detected in CMT-U335 cells (manuscript in preparation).
The significance of the alternative GHR transcripts is not known at the moment. Although it is concluded from the Northern blot analysis that the amount of alternative transcripts in normal mammary tissue seems to be low, it is an interesting observation that we could not detect these alternative GHR transcripts in most malignant mammary tumors by seminested RT-PCR. Recently, it was reported that similar alternative GHR transcripts found in nonmammary human IM9 cells, encoded proteins that negatively influenced the full-length GHR when transfected and expressed in 293 cells (31). Consequently, absence of alternative GHR transcripts in canine mammary carcinomas might result in increased signaling of the full-length GHR and transcriptional activities in mammary carcinomas, compared with nontumorous mammary tissue. In view of the reported GH production in normal and malignant canine mammary tissue, the difference in alternative GHR processing in these tissues may be of relevance.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received March 24, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  M. J. LeBaron, T. J. Ahonen, M. T. Nevalainen, and H. Rui In Vivo Response-Based Identification of Direct Hormone Target Cell Populations Using High-Density Tissue Arrays Endocrinology, March 1, 2007; 148(3): 989 - 1008. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
