The mammary gland: a unique organ for the study of development and tumorigenesis

The microanatomy and development of the mammary gland are unique and a reflection of its function to synthesize and deliver milk to the newborn offspring. The uniqueness of the mammary gland resides in several factors. First, the mammary parenchyma undergoes the vast majority of its growth postpubertally, thus enabling experiments on development to be performed in the juvenile or adult and presenting opportunities for experimental manipulation of the gland not available with other organs. On the basis of this characteristic, the fat pad transplantation method was developed, which resulted in the elaboration of important concepts in senescence, immortalization, and preneoplasia. Second, the accessibility of the gland and the ductal organization allows delivery and localization of specific molecules to mammary parenchyma cells, the cells which are the site of origin of neoplastic development. Third, the organ is the target of viral, chemical, and physical carcinogens, allowing development of unique and complex models for neoplastic development. Finally, the complexity of hormone and growth factor regulation of mammary gland function allows a sophisticated approach to the study of hormone action. The purpose of this review is to illustrate some unique properties of the gland which provide the basis for specialized approaches to developmental, neoplastic, and functional problems.

Incidence and risk factors for early small bowel obstruction after celiotomy for penetrating abdominal trauma

The reported overall incidence of postoperative Small Bowel Obstruction (SBO) is 0.69 per cent. This study examined the incidence and risk factors for early postoperative SBO for penetrating abdominal trauma (PAT), with early SBO defined as SBO within 6 months of operation. This was a prospective cohort study of patients admitted to a Level 1 Trauma Center from 5/91 to 12/93 who required celiotomy for PAT. Patients were followed at least 6 months for readmission for SBO to be considered evaluable. Of 341 patients undergoing celiotomy for PAT and surviving to discharge, 298 (87.4%) were evaluable. The overall incidence of early SBO was 7.4 per cent and varied between 2.3 per cent (nontherapeutic celiotomy) and 10.8 per cent (small/large bowel injury). All patients with celiotomy for PAT are at increased risk for early SBO compared with elective surgery patients. Those with small/large bowel penetration or gunshot wounds are at the highest risk. Previous abdominal surgery is not a risk factor for early SBO in PAT patients. Surgeons and Managed Care case managers should devote special attention to close follow-up in PAT patients, particularly those with the risk factors identified in this study.

Delay of dimethylbenz[a]anthracene-induced mammary tumorigenesis in transgenic mice by apoptosis induced by an unusual mutant p53 protein

Murine p53 containing an Arg-->Leu substitution at amino acid 172 possesses many properties characteristic of wild-type p53, including the ability to induce p21/WAF/Cip1 and apoptosis. To determine if p53-dependent apoptosis plays a critical role in mammary tumorigenesis, transgenic mice were generated in which the expression of this mutant p53 protein was targeted to the mammary gland by using the rat whey acidic protein gene promoter. Mice bearing pituitary isografts were treated with 7,12-dimethylbenz[a]anthracene (DMBA) and examined for mammary tumor development. Mice overexpressing the p53 transgene exhibited a statistically significant increase in apoptosis in the mammary gland and a statistically significant decrease in the incidence of DMBA-induced mammary tumors. No difference in tumor incidence was observed in mice without pituitary isografts who were treated with DMBA, because the transgene is not overexpressed in the absence of hormone stimulation provided by the pituitary isograft. The unexpected wild-type properties of the 172Arg-->Leu mutant p53, including its ability to stimulate apoptosis, make it a possible candidate for use in gene therapy protocols.

Mouse mammary hyperplasias and neoplasias exhibit different patterns of cyclins D1 and D2 binding to cdk4

Deregulated expression of G1 cyclins D1 and D2 is a feature of some neoplasias. This study examined the altered expression of D1 and D2 cyclins, both the total pool and as associated with cdk4 and cdk2, at different stages of mouse mammary tumorigenesis. Three different mammary hyperplastic outgrowth lines, TM2, TM10 and TM12, and their respective tumors were examined. Increasing levels of the cyclin D1 protein pool, D1 binding to cdk4 and cdk2 and cdk4 kinase activity were closely correlated with tumorigenesis. In constrast, cyclin D2 binding to cdk4 was predominant in hyperplasias and much less in tumors, where cyclin D1 became predominant. However, the cyclin D2 pool showed increases of 15-65 times in hyperplasias compared with normal gland and further increases of 11-15 times in two of three different tumors. The message level for cyclin D1 increased only 2-3 times in tumors compared with normal gland. Cyclin D2 mRNA was highest in normal tissue and decreased only marginally in tumors. These results suggest that cyclin D2 functions uniquely from cyclin D1 in the early stages of mouse mammary tumor development. Cyclin D2 bound to cdk4 may act to guarantee a low level of kinase activity in hyperplasias and may be an attempt to direct the mammary epithelial cells through differentiation rather than proliferation. This interaction may be one of the negative regulatory mechanisms in the early stages in mouse mammary tumor development, until cyclin D1 totally replaces cyclin D2 binding to cdk4, which would activate the high levels of cdk4 kinase activity observed in neoplasias.

Expression of the mouse mammary tumor virus long terminal repeat open reading frame promotes tumorigenic potential of hyperplastic mouse mammary epithelial cells

The retrovirus mouse mammary tumor virus (MMTV) 3' long terminal repeat (LTR) contains an open reading frame (ORF) for a 36-kDa protein and encodes a superantigen activity [pORF(sag)]. We have tested the potential oncogenic activity of pORF(sag) in two immortalized mouse mammary epithelial cells. We subcloned MMTV LTR ORF DNA into the pRc/CMV mammalian expression vector in order to place LTR ORF transcription under the control of the constitutive CMV promoter. Mouse mammary epithelial cell lines TM3 and FSK7e4 were transfected and G418-resistant cell clones were isolated. Reverse transcription-polymerase chain reaction and Northern blot analyses revealed modest overexpression of LTR RNA in several transfected cell clones of each line. Individual cell clones were transplanted into cleared mammary gland fat pads of syngeneic BALB/c mice. The parental cell lines and FSK7e4-derived clones did not form tumors, whereas ORF-transfected clones derived from the TM3 cells formed tumors within 8 weeks in 100% of transplanted fat pads in multiple experiments. The tumor cells expressed exogenous LTR ORF RNA and were proven to be derivatives of TM3 cells based on a marker p53 mutation. Immunohistochemistry using a polyclonal antiserum raised against pORF(sag) expressed in insect cells revealed a cytoplasmic reaction in TM3-CMV-LTR tumor cells; a much weaker cytoplasmic reaction was detected in the transfected tissue culture cells. These observations suggest that MMTV pORF(sag) may act as an oncogene in certain mouse mammary epithelial cells and raise the possibility that pORF(sag) may have a role in mammary tumorigenesis. As the parental FSK7 cell line has produced only ductal outgrowths upon transplantation in vivo and the TM3 cell line produces a nontumorigenic hyperplasia, the results suggest further that pORF(sag) may influence the latter stages of mammary tumorigenesis, namely, the preneoplastic to neoplastic transformation.

Prostate cancer in a transgenic mouse

Progress toward understanding the biology of prostate cancer has been slow due to the few animal research models available to study the spectrum of this uniquely human disease. To develop an animal model for prostate cancer, several lines of transgenic mice were generated by using the prostate-specific rat probasin promoter to derive expression of the simian virus 40 large tumor antigen-coding region. Mice expressing high levels of the transgene display progressive forms of prostatic disease that histologically resemble human prostate cancer, ranging from mild intraepithelial hyperplasia to large multinodular malignant neoplasia. Prostate tumors have been detected specifically in the prostate as early as 10 weeks of age. Immunohistochemical analysis of tumor tissue has demonstrated that dorsolateral prostate-specific secretory proteins were confined to well-differentiated ductal epithelial cells adjacent to, or within, the poorly differentiated tumor mass. Prostate tumors in the mice also display elevated levels of nuclear p53 and a decreased heterogeneous pattern of androgen-receptor expression, as observed in advanced human prostate cancer. The establishment of breeding lines of transgenic mice that reproducibly develop prostate cancer provides an animal model system to study the molecular basis of transformation of normal prostatic cells and the factors influencing the progression to metastatic prostate cancer.

Deficiency of p53 accelerates mammary tumorigenesis in Wnt-1 transgenic mice and promotes chromosomal instability

By crossing mice that carry a null allele of p53 with transgenic mice that develop mammary adenocarcinomas under the influence of a Wnt-1 transgene, we have studied the consequences of p53 deficiency in mammary gland neoplasia. In Wnt-1 transgenic mice homozygous for the p53 null allele, tumors appear at an earlier age than in animals heterozygous or wild-type at the p53 locus. About half of the tumors arising in p53 heterozygotes exhibit loss of the normal p53 allele, implying selection for p53-deficient cells. Mammary tumors lacking p53 display less fibrotic histopathology and increased genomic instability with aneuploidy, amplifications, and deletions, as detected by karyotype analysis and comparative genomic hybridization. In one tumor, the amplified region of chromosome 7 had an ectopically expressed int-2/FGF3 proto-oncogene, a gene known to cooperate with Wnt-1 in the production of mammary tumors. These findings favor a model in which p53 deficiency relaxes normal restraints on chromosomal number and organization during tumorigenesis.

Cell cyclins and cyclin-dependent kinase activities in mouse mammary tumor development

Breast cancer in humans, as in mice and rats, is thought to be the result of sequential changes in the epithelial cells of the mammalian glands. This study examines the altered expression or activation of cell cycle related proteins in an in situ system composed of hyperplasia, preneoplasia and neoplasia of mouse mammary glands. The results showed a high level of cdc2/cdk2 kinase activities in tumors compared to hyperplasias which was independent of cdc2/cdk2 protein levels. Some of the cdk-associated proteins which are thought to regulate cdk kinase activity were examined in these tissues. Cyclin A was overexpressed in all hyperplasias irrespective of their tumorigenic potentials. However, a number of alterations in cyclin E protein were associated with cdk2 and its associated kinase activity during mammary tumorigenesis. First, the level of normal cyclin E (p50) expression was positively correlated with the tumorigenic potentials of different hyperplasia lines. Second, several cyclin E isoforms (p48, p43, p35, p34, p32) were detected only in tumor tissues. Third, a 2.3- and 8.3-fold increase in cyclin E-associated cdk2 kinase activity was present in highly tumorigenic hyperplasias and neoplasias respectively compared to the low tumorigenic hyperplasias. Polymorphic cell nuclear antigen (PCNA) protein bound to cdk2 was a better indicator for cell proliferation and cdk2 kinase activity than the PCNA labeling index. These results suggest a sequential pattern of multiple derangements in factors regulating cdk2 protein function during mammary tumorigenesis. High levels of cdk2 kinase activity are observed only in tumors and appear to be closely related to alterations in cyclin E protein expression.

Tyrosine phosphorylation in mouse mammary hyperplasias

Tyrosine phosphorylation status was investigated during mouse mammary tumor development using non-tumorigenic and tumorigenic hyperplastic outgrowth lines. These outgrowth lines were compared with normal mammary glands from pregnant mice and with their corresponding tumors. The levels of total tyrosine phosphorylation in proteins of hyperplastic and neoplastic tissues were 4.7- and 3.4-fold higher than in the normal gland respectively. These results indicate that increases in tyrosine phosphorylation occur in the earliest stages of neoplastic development and are not restricted to neoplastic cells per se. These results led to the identification of the specific proteins showing high levels of tyrosine phosphorylation. Of the eight molecular weight bands of proteins exhibiting detectable levels of tyrosine phosphorylation, the only proteins exhibiting consistently different degrees of phosphorylation between hyperplasias and tumors were of approximately 34 kDa. In a series of six different hyperplasias with tumorigenic potentials ranging from 0 to 93%, the extent of tyrosine phosphorylation of 34 kDa proteins correlated inversely with tumorigenic potential. The levels of p34cdc2 and p33cdk2 proteins were examined, using antibodies specific for the cdc2 and cdk2 proteins. The amounts of p34cdc2 and p33cdk2 proteins were low in non-tumorigenic (TM3 and TM2L) compared to tumorigenic hyperplasias and correlated inversely with tyrosine phosphorylation of 34 kDa proteins during tumor development. Thus in the non-tumorigenic hyperplasias (TM2L and TM3) the majority of p34cdc2 was phosphorylated on tyrosine, in contrast to the p34cdc2 in tumorigenic (TM2H) hyperplasias and tumors. Two-dimensional PAGE analysis of mammary tumor samples with antibodies specific to cdc2, cdk2 and phosphorylated tyrosine revealed one p34cdc2 form, two p33cdk2 isoforms and two phosphotyrosine isoforms of about 33-34 kDa. The results suggest that the high levels of tyrosine phosphorylation in cdc2 and cdk2 reflect the low tumorigenic potential of a subset of mammary preneoplastic hyperplasias. This interpretation is in accord with current concepts on the role of tyrosine phosphorylation in the regulation of the cyclin-dependent kinases.

Preferential overexpression of a 172Arg-->Leu mutant p53 in the mammary gland of transgenic mice results in altered lobuloalveolar development

Regulatory sequences derived from the rat whey acidic protein gene have been used to preferentially overexpress a murine 172Arg-->Leu mutant p53 in the mammary gland of transgenic mice. Several different lines of mice expressing the 172Arg-->Leu mutant p53 displayed an impaired ability to lactate, and the mice expressing the highest levels of mutant p53 were unable to nurse their young. This failure was related to the inhibition of normal lobuloalveolar development that occurred during late pregnancy and a marked decrease in milk protein gene expression at early lactation. Interestingly, immunohistochemical analysis revealed that the mutant p53 was localized predominantly in the cytoplasm of alveolar cells. Ductal development was not overtly impaired in these mice. Expression of the 172Arg-->Leu mutant p53 resulted in radiation-induced apoptosis, and transactivation or repression of the expression of a number of genes, including mdm-2 and proliferating cell nuclear antigen, known properties of wild-type p53. The availability of lines of mice preferentially expressing specific p53 mutants in the mammary gland should facilitate evaluation of the roles of other factors, such as hormones, oncogenes and chemical carcinogens, in the etiology of breast cancer.

Infrequent p53 mutations in 7,12-dimethylbenz[a]anthracene-induced mammary tumors in BALB/c and p53 hemizygous mice

We conducted experiments to determine if p53 alterations, which are frequent in human breast cancers, were also common in murine mammary tumors. In 13 mammary tumors from 7,12-dimethylbenz[a]anthracene (DMBA)-treated BALB/c mice were immunohistochemically analyzed for overexpression of p53; p53 protein was not detectable. Three of the tumors were established as cell lines in vitro. p53 protein was rarely detected at passage 4 in these lines but was overexpressed by passage 8 in two of them. The p53 nucleotide sequence was shown to be wild type in one primary mammary tumor and in the two p53-overexpressing cell lines. One cell line that overexpressed p53 in vitro was implanted into BALB/c mice. The resulting tumors retained the wild-type p53 nucleotide sequence but no longer expressed detectable levels of p53 protein, suggesting that the overexpression of wild-type p53 was related to in vitro culture conditions. The effect of DMBA on mammary-tumor development was also tested in mice rendered hemizygous for p53. These mice and wild-type littermate controls had no differences in susceptibility to induction of mammary tumors by oral administration of DMBA. Furthermore, Southern blot hybridization detected no gross alterations in the wild-type p53 allele in mammary tumors from the p53-deficient mice. Point mutation of the wild-type p53 allele was also infrequent in the DMBA-induced mammary tumors from hemizygous p53 mice; it occurred in only one of seven tumors. Thus, the p53 gene is apparently not a primary target for genetic alterations in DMBA-induced mammary tumors. Next, we examined mammary tumors derived from D1 and D2 transplantable hyperplastic alveolar nodule (HAN) outgrowths, which rapidly form tumors containing Ha-ras mutations after DMBA treatment. As ras and p53 mutants can cooperate in transformation, we examined whether D1 and D2 HAN outgrowths have p53 mutations. Unlike in the DMBA-induced primary mammary tumors, nuclear p53 accumulation was observed frequently (10 of 14) in tumors that arose from D1 and D2 HAN outgrowths. Direct sequencing of the entire coding region of the p53 cDNA from six D1 and D2 tumors confirmed that the sequence was wild type. Although wild-type p53 was retained in both DMBA-induced mammary tumors and mammary tumors derived from D1 and D2 preneoplastic outgrowths, wild-type p53 overexpression was detected only in D1 and D2 tumors. Therefore, D1 and D2 tumors appear to arise by a pathway in which p53 expression is altered, whereas DMBA induction affects a different pathway that does not require such alteration.

Immunophenotypic properties and estrogen dependency of budding cell structures in the developing mouse mammary gland

The initial phase of growth of the parenchymal component of the mouse mammary gland is ductal clongation, which is mainly accomplished by proliferating cells in a specialized structure termed end bud. End buds are composed of multiple layers of epithelial cells (so called body cells) which are capped by a single layer of morphologically unique cells termed cap cells. We sought to examine the interrelationship between cap cells and other epithelial cell subclasses using a variety of antibodies to different keratin proteins and also antibodies to vimentin, actin and collagen IV. An extensive immunohistochemical characterization of the epithelial components of the developing and differentiating mammary gland demonstrated that cap cells were devoid of any immunohistochemically-detectable keratins but were positive for collagen IV. In contrast, the majority of cells in the end bud along with the luminal epithelial and myoepithelial cells were keratin positive. The body cells of the end bud were the only cells which were positive for antibody to keratin 6, a keratin which previously has been reported to be expressed in proliferating mammary epithelial cells. In addition, estrogen receptor was localized only to epithelial cells of ducts, alvcoli and body cells of end buds, but not to cap cells or myoepithelial cells. We interpret these results to suggest that cap cells are not totpotent stem cells but rather cells specialized in paving the way for ductal elongation as well as serving as precursors to myoepithelial cells.

p53 mutations in mouse mammary epithelial cells: instability in culture and discordant selection of mutations in vitro versus in vivo

The phenotypes of p53 mutations found in human and murine tumors often are analyzed using a variety of transformation assays in vitro, but data have not been available to correlate in vitro effects with in vivo activities. We have assessed the effects of p53 mutations using mouse mammary epithelial cell lines which can be analyzed both in vitro and in vivo. Parental mammary epithelial cell lines (FSK series) injected into cleared mammary fat pads of syngeneic mice frequently give rise to preneoplastic lesions (HAN) which can be reestablished in culture (TM lines) to permit analysis of genetic changes important in the development of preneoplasia. Characterization of the FSK3 cell line revealed a cell population mixed with respect to p53 genotypes. One subpopulation of mutant (Ser233-234) p53-expressing cells was selected in a preneoplastic mammary outgrowth in vivo (TM3), whereas another minor population of mutant (Pro135) p53-expressing cells was selected during culturing of FSK3 cells in vitro. When FSK3 cells were subdivided and passaged in parallel in vitro, p53 genotypes continued to evolve and diverge. These findings reveal that selective pressures exerted on mammary epithelial cell populations in vivo are different from pressures present in vitro. Selective forces may vary from one cell culture environment to another. The growth advantage conferred by a specific p53 mutation appears to differ in vivo versus in vitro. We propose that caution should be exercised when attempting to correlate the effects of p53 mutations assayed in cell culture systems with the in situ phenotypes of mutant p53 in cancer.

Significance of selenium-labeled proteins for selenium's chemopreventive functions

A 58 kDa selenium-labeled protein purified from mouse mammary epithelial cells (MMEC) was used to examine whether selenium modulates protein synthesis or is just a marker for cellular selenium status. The protein was isolated using Sephadex G150 gel filtration and DEAE-Sephadex A50 ion-exchange chromatography. It was further analysed using 2-dimensional polyacrylamide gel electrophoresis (2D-PAGE) and was found as a single spot with a pI of 4.6. The immunoreactivity with anti-58 kDa antiserum and the 75Se signal co-localized on a single 58 kDa protein band on both 1D- and 2D-PAGE. Partial amino acid analysis of the peptide showed homology with the thiol protein disulfide oxidoreductase (TPDO). Varying the selenium concentration in culture medium did not affect the protein content or the immunoreactivity of the 58 kDa protein. Additionally, selenium did not seem to regulate the activity of TPDO in TM6 cells. The glutathione peroxidase activity of TM6 cells, taken as the internal positive control, was enhanced with the increase in selenium concentration in the medium. The results suggest that selenium is attached to the 58 kDa protein, but does not regulate either its protein synthesis or its functional activity. We conclude that selenium labeling of the 58 kDa protein reflects the cellular selenium status but probably is not involved in its chemopreventive ability.

Mutations in p53 are frequent in the preneoplastic stage of mouse mammary tumor development

Preneoplastic lesions in the mammary gland represent a population of cells at increased risk of progression to tumors. Because p53 is the most commonly mutated gene in human breast cancer, we sought to determine whether mutations in p53 were present in preneoplastic lesions or were acquired during progression to overt tumors. In the mouse mammary gland, hyperplastic alveolar nodules (HAN) are the most common preneoplastic lesion. Analysis of the TM series of transplantable murine HAN outgrowths and tumors allowed the status of p53 to be determined at distinct stages of mammary tumorigenesis. Alterations in the p53 gene or in the pattern of p53 protein expression were observed in all five HAN outgrowth lines examined. Altered expression of p53 protein was detected in 3 of 5 TM HAN outgrowth lines as determined by immunohistochemistry. Overexpression of nuclear p53 was detected in only a fraction of the cells (10-50%) in TM3 and TM4 HAN outgrowths, whereas in tumors that arose from TM4 HAN outgrowths, the proportion of cells overexpressing p53 protein approached approximately 100%. Despite overexpression of p53 in TM3 HAN outgrowths, no tumors have developed in this line. The TM9 outgrowth line exhibited a different pattern of p53 expression by immunohistochemistry: p53 protein was overexpressed in the cytoplasm of virtually all cells in the HAN outgrowths but was localized to the nuclei of TM9 tumor cells. Direct sequencing of p53 transcripts from tumors and cell lines revealed various genetic changes: point mutations in exons 4 and 5 (TM2H, nonsense; TM4, missense); a deletion in exon 5 (TM4); and an insertion in exon 7 (TM3). Although p53 protein was overexpressed in TM9 tumors, it was shown to be wild-type both by immunoprecipitation and direct sequencing of the entire coding region of the cDNA. TM4 cells were homogeneous with respect to mutant p53 genotype and uniformly expressed p53 by immunohistochemical staining in vitro, but transplantation of TM4 cells to fat pads of BALB/c hosts resulted in HAN outgrowths in situ in which < 50% of the cells expressed the mutant p53 at detectable levels. In summary, mutation of the p53 gene and overexpression of p53 protein can occur in preneoplastic mammary epithelial cells, and those mutations are maintained in tumors that arise from the HAN. Conversely, expression of mutant p53 was decreased when cells were grown in situ, implicating the presence of cellular factors that can suppress p53 expression in vivo. These observations demonstrate that the p53 pathway may be a common target for mutation in murine mammary tumorigenesis.(ABSTRACT TRUNCATED AT 400 WORDS)

p53 mutations selected in vivo when mouse mammary epithelial cells form hyperplastic outgrowths are not necessary for establishment of mammary cell lines in vitro

Breast cancer is a consequence of multiple alterations occurring over a long period of time. Genetic changes in early stages of tumorigenesis have not been defined. A recently developed murine system permits the study of mammary preneoplastic cells in vivo and in vitro (F. S. Kittrell et al., Cancer Res., 52: 1924-1932, 1992). To assess the potential role of p53 mutations in early stages of breast cancer, the status of p53 was determined in a series of mouse mammary epithelial cell lines which give rise to preneoplastic outgrowths (hyperplastic alveolar nodules) when transplanted into cleared mammary fat pads of syngeneic mice. Protein stability and conformation were analyzed using immunoprecipitations and immunochemical assays; p53 transcripts were sequenced using a polymerase chain reaction approach. The parental cell lines (FSK lines) showed no evidence of p53 alterations at either the protein or the nucleic acid level, indicating that p53 mutations are not essential for the establishment of mammary epithelial cell lines in vitro. In contrast, cell lines (TM lines) derived from hyperplastic alveolar nodule outgrowths induced by FSK cells expressed only mutant p53 genes. The mutation in one outgrowth cell line (TM-2H) resulted in the loss of p53 protein synthesis, whereas two other outgrowth lines (TM-3, TM-4) overexpressed mutant p53 protein. Mutation of p53 appears to correlate with preneoplastic growth in vivo. Although it is not clear if the mutations occur before or after transplantation of cells in vivo, there appears to be a pronounced growth advantage in the mammary gland for cells expressing mutant p53.

Combination therapy with a nuclear type-ii site agonist and 5-Fluorouracil - inhibition of mammary-tumor growth

Methyl p-hydroxyphenyllactate (MeHPLA) is an endogenous ligand for nuclear type II sites which apparently regulates cellular growth and proliferation through this binding interaction. Occupancy of type II sites by MeHPLA agonists such as dihydroxybenzylidene acetophenone (DHBA), 2,6-bis ([3,4-dihydroxyphenyl]-methylene) cyclohexanone (BDHPC) and 2,6-bis ([3-methoxy-4-hydroxyphenyl]-methylene)cyclohexanone (BMHPC) is directly correlated with the inhibition of malignant cell proliferation. Most importantly, these compounds inhibit mammary tumor growth in vivo with minimal non-specific cytotoxicity. Therefore, combination therapy with MeHPLA agonists plus standard anti-neoplastic agents such as 5-Fluorouracil (FU) may result in tumor growth inhibition with minimal non-specific cytotoxicity. The results of these studies demonstrated that low doses of BMHPC (2 mug/mL) and FU (0.2 mug/mL) failed to significantly affect MCF-7 human breast cancer cell proliferation. However, combination therapy with these sub-inhibitory doses of BMHPC plus FU resulted in significant inhibition of cell proliferation in vitro, suggesting that BMHPC acts in an additive or synergistic fashion with FU to inhibit MCF-7 cell proliferation. Similarly, when orally administered to mice at a dose level of 50 mug/mL drinking water, neither BMHPC or FU alone substantially inhibited the growth of estrogen-independent transplantable mammary tumors. However, combination therapy with BMHPC plus FU antagonized tumor growth and no significant treatment effects were observed on fluid consumption or the body weights of these animals. These results demonstrate that MeHPLA agonists such as BMHPC are capable of acting additively or synergistically with FU to maintain therapeutic response with reduced non-specific systemic toxicity.

Expression of fatty acid-binding proteins in the developing mouse mammary gland

Proteins of 14,000 daltons (SLP-14) were isolated and partially characterized from mouse mammary glands of different developmental stages. The purified proteins were partially sequenced at the amino acid level. The SLP-14 belong to the family of fatty acid-binding (FABP) proteins. The major SLP-14 expressed was a reflection of the differentiation stage of the mammary gland. Within the regions sequenced for each protein, virgin mouse mammary gland (primarily adipocytes) expressed a protein 100% homologous to adipocyte lipid binding protein whereas pregnant and lactating mouse mammary glands expressed a protein highly homologous to mouse heart fatty acid binding protein. In the lactating gland, the isolated protein was 97% identical to the heart FABP over a 98 amino acid stretch. We could not detect the 14Kd protein identified as mammary gland growth inhibitor (MDGI) which is also a member of the FABP family and highly homologous to heart FABP. A rabbit antiserum to the rat mammary SLP-14 recognized the SLP-14 proteins in mouse and rat mammary gland, skeletal muscle and heart, whereas it failed to recognize the SLP-14 in liver, intestine and other organs. These data indicate that the SLP-14 detected in rodent mammary gland is of the fatty acid-binding protein family. On the basis of amino acid sequence, the major form in the differentiated mouse mammary gland is apparently FABP and not MDGI. MDGI may be a protein in low abundance and/or localized to a specific group of mammary epithelial cells.

Growth factor dependency and gene expression in preneoplastic mouse mammary epithelial cells

The TM preneoplastic mammary outgrowth lines were established in vivo from mammary epithelial cell lines and have been characterized with respect to their tumorigenic, morphological, and functional properties. The TM outgrowth lines were then established as in vitro cell lines. A comparison of the growth factor dependencies of the TM preneoplastic lines and their progenitor cell lines grown in monolayer cell culture indicated that the TM preneoplastic cell lines exhibited a decreased dependence on epidermal growth factor for growth in vitro. The exception to this result was the TM3 cell line which still exhibited a marked dependence on epidermal growth factor for growth. An examination of several genes for mRNA levels indicated that the expression of c-neu, c-H-ras, c-myc, and retinoblastoma was not elevated in those TM preneoplasias which exhibited a decreased dependence on epidermal growth factor. Additionally, there was no evidence that c-H-ras was activated in the preneoplastic outgrowths or tumors. In contrast, mouse mammary tumor virus long terminal repeat mRNA was increased in preneoplasias and tumors, whereas gelsolin mRNA was decreased in tumors but not in preneoplasias. The down-regulation of gelsolin mRNA is consistent with recent reports in human breast cancers. These results together with those reported in another paper (D. Medina et al., Cancer Res., 53: 663-667, 1993) indicate that the TM3 outgrowth line is a minimally deviated preneoplasia which does not share many of the molecular alterations exhibited in tumorigenic TM preneoplastic outgrowth lines. These data, along with other recent data, are interpreted in a hypothesis which views the three essential characteristics of the mammary preneoplastic state as independent and dissociable genetic alterations. Importantly, each of the three characteristics is independently isolated in one or more of the in vivo outgrowth populations. These outgrowth lines should allow identification of the nature and function of the molecular alterations associated with each stage of mammary preneoplasia.