| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Regular Articles |
Institute for Molecular Biology and Biotechnology, Department of Biology, McMaster University, Hamilton, Canada [E. R. A., M. A. L., A. A. G-G.]; Cell Biology Program, Memorial Sloan-Kettering Cancer Center, New York, New York 10021 [P. M. S.]; Department of Pathology, School of Medicine, University of California, Davis, California 95616 [R. D. C.]; and Molecular Oncology Labs, McGill University, Royal Victoria Hospital, Montreal, Quebec, H3A 1A1 Canada [W. J. M.]
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Direct evidence for the importance of Neu in mammary tumor development stems from observations made with transgenic mice expressing the activated form of neu under the control of the MMTV promoter enhancer. Mammary epithelial specific expression of activated neu in transgenic mice resulted in the rapid induction of multifocal metastatic mammary tumors (12, 13, 14) . Additionally, several other mouse models with various neu alleles have been created and revealed insights into the requirement for activating mutations or splice variants and expression of other family members (13 , 15 , 16) . Although these transgenic studies suggest that activation of neu is a critical step in tumor progression, one limitation of the transgenic mouse models is that expression of neu is driven by a strong viral promoter. In addition, because this promoter is hormonally regulated, it is difficult to assess the role of hormones in tumor induction.
In an attempt to more closely mimic the events involved in human ErbB2-induced mammary tumor progression, we have recently derived transgenic mice that carry a Cre inducible activated neu allele under the transcriptional control of the endogenous promoter (17) . In contrast to the rapid tumor progression observed in the MMTV-mediated Neu strains, focal mammary tumors arose only after an extended latency period. Tumor progression in this strain was also associated with a dramatic elevation of both Neu protein and transcript levels. Remarkably, the elevated expression of Neu was further correlated with selective genomic amplification of the activated neu allele (17) . Thus, like human breast cancers, amplification of neu appears to be a critical event in mammary tumor progression in this unique transgenic mouse model. Consistent with these initial studies, comparative genome hybridization and spectral karyotyping analysis have revealed that these tumors consistently carry amplification of the regions of mouse chromosomes 11, 10, and 15 (18) . Significantly, this amplified region of chromosome 11 corresponds to the mapped location of the recombinant neu allele that we have demonstrated previously to be amplified in these Neu-induced tumors. In addition to specific amplicons and deletions, a region of mouse chromosome 4 that is analogous to human chromosome 1p32–36, a region demonstrated previously to be site of allelic loss in human cancers (19) , was also noted (18) .
Tumor progression in this Neu knock-in model involves a dramatic transcriptional up-regulation of Neu protein and transcript that cannot simply be accounted for by 2–20-fold amplification of the activated neu allele. One possible explanation for the dramatic up-regulation of activated Neu expression is that the levels of key transcription factors involved in the regulation of the promoter have been altered. To further explore this possibility, we have compared the gene expression profile between tumors derived from either transgenic mice expressing activated neu under the transcriptional control of the endogenous promoter (17) or expressing activated neu from the MMTV promoter (16) .
Although both sets of tumors histologically resembled each other, one notable distinction is that in contrast to the high rates of metastasis observed in the MMTV/Neu model, the mammary tumors induced by Neu expression under the endogenous promoter were poorly metastatic. Consistent with these analyses, expression profiling of these nonmetastatic tumors revealed elevated expression of a number of differentiated markers. In addition, these analyses showed that Grb7 and Cab1, which are coamplified and coexpressed with HER2 in human breast cancer (20) , are also coexpressed and coamplified in transgenic mice expressing activated Neu under the endogenous promoter. Taken together, these observations suggest that mammary tumors induced by this mouse model resemble noninvasive ductal carcinoma in situ lesions observed in human breast cancer.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
24 h. Thin sections of the paraffin-blocked samples were cut and stained with H&E using standard methods. To determine whether the adenocarcinomas had metastasized to the lungs, several sections through various lobes of the lung were examined at high magnification after H&E staining.
Animal Care and Tumor Analysis.
Animal care was conducted in accordance with the policies of McMaster University’s Central Animal Facility Guidelines. The detection of tumors was through routine palpation of the #2, 3, 4, and 5 mammary glands at weekly intervals. Furthermore, the #1 set of mammary glands was assessed by eye.
RNA Analysis.
RNA from mammary glands and tumor samples was prepared by a CsCl sedimentation gradient after homogenization in GIT as described previously for quantitative PCR and Northern analysis (21)
. RNA was resuspended in diethyl pyrocarbonate-treated water and quantitated by UV absorbance. For quantitative RT-PCR, the reverse transcription, denaturation, PCR, and quantification were carried out in a single capillary using the LightCycler RNA Amplification Kit SYBR Green 1 (Roche). The manufacturer’s protocols were followed to determine optimal MgCl2 concentration for each primer pair. The suggested protocol was also followed in the RT-PCR reactions using 200 ng of total RNA. The primers used to amplify WDNM1 were as follows: (a) 5' TCT TTG TTC TGG TAG CTT TGA TTT 3' and (b) 5' GTT TGC AGG CAT GAC CAC AG 3'. The PCR reaction was stopped, whereas all samples were in the log-linear phase of amplification, and when necessary, samples were diluted so that all reactions would be in this phase of amplification at the same cycle. The PCR reaction was then terminated, and the product was subject to a melting curve analysis. To determine that the product defined by the melting temperature was the actual amplified product, the samples were electrophoresed on a 1.5% agarose gel during optimization of the primers. To standardize the level of RNA in these samples, RT-PCR was also completed for glyceraldehyde-3-phosphate dehydrogenase using the following primers: (a) 5' TCA TGA CCA CAG TGG ATG CC 3' and (b) 5' GGA GTT GCT GTT GAA GTC GC 3'. Northern analysis was conducted as described previously (22)
. The primers used to amplify MDGI were 5' GAA TTC ACC ATC ATC GAG AAG AAC GG 3' and 5' GGA TCC TTA TTG ACC TTG GAG CAC CC 3'; Serglycin, 5' GAA TTC CTG TTT TGA TGG AAG GAC CC 3' and 5' GGA TCC CGA GTT AAT GAG GAA AGG GG 3'; and Glycam1, 5' GAA TTC ACC TCT CTT GCT CTC CTG CC 3' and 5' GGA TCC TCA GAT GAG CAG AAT GTG GC 3'. The italicized sequence corresponds to restriction sites added for cloning purposes.
RNA was also isolated for Affymetrix GeneChip analysis using the RNeasy Midi Kit (Qiagen) following the manufacturer’s directions. RNA was resuspended, and concentration was determined as above. Affymetrix GeneChips were conducted on a pay per service basis by the OHRI gene expression facility.
Amplification and Overexpression.
DNA was obtained from tumors and examined for amplification of neu as described previously (17)
. Probes for Grb7 and Cab1 were also generated and used on the same Southern blot after stripping using standard methods. Overexpression of Grb7 was monitored through a Western analysis as described previously using an antibody to Grb7 (Santa Cruz Biotechnology).
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, both the conditionally activated and MMTV-regulated models exhibit altered branching (Fig. 1, C–F)
. Careful examination of mammary wholemounts revealed several notable differences between the two activated Neu models. Firstly, the mice expressing activated neu under the control of the endogenous promoter exhibit mammary epithelium that escapes the normal confines of the fat pad. This ductal outgrowth typically follows the small deposit of stroma that surrounds the blood vessels that enter and exit the mammary gland (Fig. 1C
, arrows). This phenotype was observed in 22 of the 28 mice that were examined. Secondly, although both activated Neu models exhibit extensive side branching that is not commonly observed in the wild-type control (Fig. 1, D and F)
, the hyperplastic progression is distinct. In the mice expressing the activated allele under the endogenous promoter, the lobuloalveolar endbuds are frequently hyperplastic, but in the MMTV-based system, these hyperplastic regions quickly develop into adenocarcinomas. On tumor formation, the two Neu mouse models were compared histologically (Fig. 2, A
versus B). The tumors in the two different model systems appear quite similar and are both readily classified as comedo-adenocarcinomas typical of Neu-induced tumors.
|
|
. From previous studies, it was noted that this increased latency is attributable to the requirement for neu amplification and overexpression. However, in addition to the increased latency, a striking difference was noted in the percentage of observed lung metastasis. Although 
60% of the MMTV Neu-mediated tumors metastasize to the lungs, only a single metastasis has been noted in >40 tumor-bearing mice where neu is regulated by the endogenous promoter (Fig. 2D)
.
One explanation for the preoneoplastic outgrowths observed in the mammary glands of the activated Neu models is that these lesions represent expansion of cells targeted by Neu. Previous studies have suggested Neu- and Ras-initiated tumors share a number of common genetic markers that reflect a common cellular lineage (23)
. One such gene is WDNM1, characterized previously as a neoplastic molecular marker. To explore whether preneoplastic growths observed in transgenic mice expressing activated neu under the control of the endogenous promoter express this tumor marker, we compared the levels of WDNM1 transcript in Neu-induced tumors and pretumorous mammary glands with wild-type glands. The results of these analyses revealed that like the Neu-induced mammary tumors, mammary glands carrying preneoplastic growths expressed elevated levels of WDNM1. In contrast, virgin wild-type mammary tissue expressed relatively low levels of WDNM1 (Fig. 3)
. These observations suggest that preneoplastic growths observed in this model reflect expansion of a cell lineage that is targeted by activated Neu.
|
. The fold elevation is expressed as a positive value when the transcript level from the endogenously regulated neu tumor was higher than the transcript level from the MMTV-regulated tumor. In general, most of the genes detected were expressed at higher levels in the endogenous promoter model than there were in the MMTV-based system. One category of genes that arose from these analyses are mammary gland differentiation markers, including genes such as whey acidic protein (128-fold elevation), several members of the casein family (5–26-fold elevation), and members of the keratin family (23–99-fold elevation). To further validate this analysis, we performed Northern blot or real-time PCR analyses with probes specific to each of these markers (Table 1)
. Representative Northern blot analyses for Glycam1, Serglycin, and MDGI are shown in Fig. 4
. Although the average fold up-regulation of these markers was in close agreement with the Microarray analyses, the Northerns revealed that some individual tumors were highly variable in the expression of these differentiation markers, e.g., although three of the five tumors expressed high levels of Glycam1 and MDGI relative to primary tumors derived from the MMTV/activated ErbB2 strains (Fig. 4
, compare Lanes 2–4 with 6–10 in A and C), two of the tumors derived the endogenous promoter-driven model expressed relatively low amounts of these differentiation markers (Fig. 4
, Lanes 1 and 5 in A and C). These analyses suggest that the state of differentiation of these ErbB2-induced tumors varies.
|
|
Tumors Derived from the Endogenous Promoter-driven Model Coamplify neu, grb-7, and cab-1.
Another category of genes that was expressed in the endogenously regulated model included grb7 and cab1, which have been reported previously to be coamplified in human breast cancer (20)
. To assess the extent of grb7 and cab1 gene amplification, we performed Southern blot analyses on these tumor samples. Using a probe that will detect both the wild-type allele (7.5 kb) and knock-in activated neu allele (4.5 kb), we showed that in these samples, the degree of amplification of neu correlated with the extent of gene amplification grb7 and cab1 (Fig. 5A)
. The Southern analysis for neu and grb7 was quantitated to show that the fold elevation correlated between the two Southerns (Fig. 5B)
. Taken together, these data show that neu is coamplified with grb7 and cab1. Additionally, the gene expression data suggested that unlike the MMTV/Neu model, the expression of Grb7 and Cab1 is closely linked to the extent of neu amplification. To confirm that Grb7 was overexpressed in addition to being amplified, we performed immunoblot analyses on primary tumors from both transgenic models with Grb7-specific antisera. As shown in Fig. 5C
, tumors derived from the endogenously regulated model expressed elevated levels of Grb7 protein (Fig. 5C
, Lanes 1 and 2). By contrast, tumors derived from the MMTV/Neu model did not express detectable levels of Grb 7 (Fig. 5C
compare Lanes 3–5 with 1 and 2). Taken together, these observations indicate that like HER2-expressing human breast cancers, coamplification of neu, grb7, and cab1 is correlated with elevated expression of the individual components of this amplicon.
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Although the tumors that developed in the two model systems were histologically similar, they were quite distinct in both latency and metastatic potential. To address the differences between the two tumor types, a gene expression experiment was conducted. This comparison of two Neu-induced tumor models allowed the identification of differentially expressed genes while eliminating common genes with high or low expression in both sample sets. The results of this analysis identified numerous targets that were subdivided into various categories, including genes that were coamplified with neu. The coamplification of grb7 and cab1 with HER2 has been well documented in the human condition (32 , 33) , and the identification of these genes in the mouse model illustrates the similarity of the amplification process. Additionally, we have illustrated that these genes are not only amplified but also overexpressed, leading to questions about their role in the progression of Neu-mediated breast cancer. Although we have shown previously that overexpression of Neu is limited to the tumors (1) , these data have illustrated that there was not a direct correlation between the extent of amplification and level of overexpression. Furthermore, because many of the targets identified in the gene expression analysis reflected that the poorly metastatic, Neu-induced tumors were more differentiated, it suggests that the amplification or chromosomal deletions may be responsible for these changes. Amplification of cab1 is seen both in this model system and the human condition, yet the effect of the enhanced steroidogenesis mediated by this protein (34) is not well understood with respect to its effects on breast cancer. Clearly, given the hormonal nature of breast cancer, enhanced steroidogenesis may play a role in the progression of the disease, and the differential expression of cab1 in the two tumor models may account for a portion of the difference in metastatic potential and differentiation status.
In addition to analyzing how similar the conditional mouse model is to the human condition, the experimental approach has also identified transcription factors that are essential in mediating tumorigenesis. Although both tumor sets overexpress Neu and have histologically similar tumors, the tumors differentially express transcription factors dependent on the promoter-driving expression of neu. In the model where the endogenous promoter drives neu expression, several interesting transcription factors are up-regulated. In addition to several others, a notable transcription factor involved in Neu-mediated tumorigenesis under the endogenous promoter is Krox-20. Krox20 (EGR2) has been linked previously to Neu-induced tumorigenesis (35) , and studies of other family members suggest a role in signaling mediated by Neu. Specifically, both the muscle-specific deletion of neu and knockout of EGR3 both resulted in a lack of muscle spindles (36 , 37) . Furthermore, the general neurological defects seen with neu knockouts are remarkably similar to the defects in the Krox20 knockouts (38, 39, 40) . Clearly, a detailed examination of this signaling pathway is warranted and currently underway.
Although this analysis revealed overexpression of several transcription factors, it primarily highlighted the differences in the differentiation status of the tumors. Histologically, these tumors are quite similar and both easily categorized as Neu-mediated adenocarcinomas. Moreover, although there are differences in the gene profile, these differences are not appreciated at the histological level. However, the metastatic frequency of the two tumor types is quite different with the expression of neu under the endogenous promoter, resulting in very few metastatic lesions. Not surprisingly, these tumors appear to be far more differentiated when the differentially expressed genes are examined. This type of gene profiling may dictate patient treatment in the future.
The identification of a number of glycoproteins is also intriguing, given that MUC4 may serve in the regulation of Neu signaling through the facilitation of heterodimerization with ErbB3 (41) . An examination of the differentially expressed glycoproteins reveals several interesting findings. AEG-1 (cysteine-rich secretory protein-1) is overexpressed 153-fold and normally expressed in the male genital tract and salivary gland. However, the same analysis of the AEG-1 promoter revealed a consensus binding domain for PEA3 (42) , which was also overexpressed in the endogenously regulated, Neu-mediated tumors. Interestingly, muc1 is also overexpressed in these tumors and has been observed to be overexpressed in many human tumors. Furthermore, it has been illustrated that EGFR and MUC1 interact at the cell membrane, resulting in the recruitment of c-Src (43) . Clearly, these various glycoproteins may serve to recruit and activate specific pathways that will alter the properties of the carcinoma.
In summary, these results have further characterized two Neu-mediated mouse models of mammary tumorigenesis. The endogenously regulated expression of an activated neu allele was shown to induce mammary tumors that, while histologically similar to the MMTV system, were quite distinct at the molecular level. Indeed, these results correlated with the differential metastatic potential of the tumors and suggest that human therapies may need to be tailored to the specific genetic profile of the tumor for effective treatment. Additionally, this study has identified exciting new areas for study with respect to the control of neu expression through the identification of numerous promoters that may be required for Neu-mediated tumorigenesis.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 Supported by CBCRI (to W. J. M.). E. R. Andrechek was supported by the United States Army Scholarship DAMD17-99-1-9285. 
2 To whom requests for reprints should be addressed, at Molecular Oncology Labs, McGill University, Royal Victoria Hospital, Montreal, Quebec, H3A 1A1 Canada. Phone: (514) 842-1231, extension 36383; Fax: (514) 843-1478; E-mail: muller{at}molonc.mcgill.ca
3 The abbreviations used are: EGFR, epidermal growth factor receptor; AEG-1, acidic epididymal glycoprotein 1; MMTV, mouse mammary tumor virus; RT-PCR, reverse transcription-PCR.
Received 2/25/03. Revised 5/ 1/03. Accepted 6/ 2/03.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. L. Dillon, R. Marcotte, B. T. Hennessy, J. R. Woodgett, G. B. Mills, and W. J. Muller Akt1 and Akt2 Play Distinct Roles in the Initiation and Metastatic Phases of Mammary Tumor Progression Cancer Res., June 15, 2009; 69(12): 5057 - 5064. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Dourdin, B. Schade, R. Lesurf, M. Hallett, R. J. Munn, R. D. Cardiff, and W. J. Muller Phosphatase and Tensin Homologue Deleted on Chromosome 10 Deficiency Accelerates Tumor Induction in a Mouse Model of ErbB-2 Mammary Tumorigenesis Cancer Res., April 1, 2008; 68(7): 2122 - 2131. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. L. Dillon, S. T. Brown, C. Ling, T. Shioda, and W. J. Muller An EGR2/CITED1 Transcription Factor Complex and the 14-3-3{sigma} Tumor Suppressor Are Involved in Regulating ErbB2 Expression in a Transgenic-Mouse Model of Human Breast Cancer Mol. Cell. Biol., December 15, 2007; 27(24): 8648 - 8657. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Lu, K. L. Knutson, E. Gad, and M. L. Disis The Tumor Antigen Repertoire Identified in Tumor-Bearing Neu Transgenic Mice Predicts Human Tumor Antigens Cancer Res., October 1, 2006; 66(19): 9754 - 9761. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. G. Hodgson, T. Malek, S. Bornstein, S. Hariono, D. G. Ginzinger, W. J. Muller, and J. W. Gray Copy Number Aberrations in Mouse Breast Tumors Reveal Loci and Genes Important in Tumorigenic Receptor Tyrosine Kinase Signaling Cancer Res., November 1, 2005; 65(21): 9695 - 9704. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Astolfi, L. Landuzzi, G. Nicoletti, C. De Giovanni, S. Croci, A. Palladini, S. Ferrini, M. Iezzi, P. Musiani, F. Cavallo, et al. Gene Expression Analysis of Immune-Mediated Arrest of Tumorigenesis in a Transgenic Mouse Model of HER-2/neu-Positive Basal-Like Mammary Carcinoma Am. J. Pathol., April 1, 2005; 166(4): 1205 - 1216. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. R. Andrechek, W. R. Hardy, M. A. Laing, and W. J. Muller Germ-line expression of an oncogenic erbB2 allele confers resistance to erbB2-induced mammary tumorigenesis PNAS, April 6, 2004; 101(14): 4984 - 4989. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. K. Galang, W. J. Muller, G. Foos, R. G. Oshima, and C. A. Hauser Changes in the Expression of Many Ets Family Transcription Factors and of Potential Target Genes in Normal Mammary Tissue and Tumors J. Biol. Chem., March 19, 2004; 279(12): 11281 - 11292. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
) and MMTV/Neu transgenics (
) reveals remarkable differences in tumor latency. Average onset in 50% of female mice is 15.8 and 4.7 months, respectively. The difference in metastatic frequency in these two strains is also shown (D). Greater than 60% of MMTV/Neu-driven tumors result in lung metastasis, whereas only 1 of the 41 tumors detected in the endogenous regulation of activated Neu resulted in metastasis to the lung.
