- Email: [email protected]
Excess deposition of collagen in mammary glands of tamoxifen-treated Holstein heifers is associated with impaired mammary growth
Domestic Animal Endocrinology 65 (2018) 49–55
Contents lists available at ScienceDirect
Domestic Animal Endocrinology journal homepage: www.journals.elsevier.com/ domestic-animal-endocrinology
Excess deposition of collagen in mammary glands of tamoxifen-treated Holstein heifers is associated with impaired mammary growth H.L.M. Tucker, J. Holdridge, C.L.M. Parsons, R.M. Akers* Department of Dairy Science, Virginia Tech, Blacksburg, VA 24061, USA
a r t i c l e i n f o
a b s t r a c t
Article history: Received 7 February 2018 Received in revised form 26 April 2018 Accepted 11 May 2018
It is established that the ovary and estrogen are essential to bovine mammary development with the onset of puberty. Recent studies have shown that ovariectomy in the very early prepubertal period, well before onset of puberty, also dramatically impairs mammary growth. Similarly, prepubertal heifers treated with the antiestrogen tamoxifen (TAM) also exhibit markedly impaired mammary growth in correspondence with reduced estrogen receptor a (ESR1) expression. Our objective was to evaluate the effect of TAM on the mammary stroma and specifically to determine if the reported decrease in mammary development was related to changes in TAM-induced alterations in the stroma surrounding the mammary parenchyma. Briefly, 16 Holstein heifers calves were randomly assigned to one of 2 treatment groups: TAM-injected or control. Calves were administered TAM (0.3 mg kg1 d1) or placebo from 28 to 120 d of age. At day 120, calves were euthanized and udders removed. Mammary tissue from near the boundary between the parenchyma and surrounding mammary fat pad was collected for histology and morphometric analysis, expression of selected extracellular matrix–related genes, and quantitation of stromal collagen deposition by study of Sirius Red-stained tissue sections imaged with polarized light. Compared with tissue from control heifers, TAM heifers frequently exhibited areas with abundant fibroblasts and mesenchymal cells especially within the intralobular stroma, as well as less complex ductal structures. Among the array of extracellular matrix–related genes tested, only a small difference (P < 0.05) in expression of laminin was found between treatments. The relative tissue area occupied by stromal tissue was not impacted by treatment. However, the deposition of collagen within the stromal tissue was more than doubled (P < 0.0001) in TAM-treated heifers. These data suggest that blocking ESR1 expression with TAM allows for excessive collagen deposition in the stroma surrounding the developing epithelial structures and that this interferes with both the degree of overall mammary parenchymal development, as well as the pattern of normal ductal morphogenesis. Ó 2018 Elsevier Inc. All rights reserved.
Keywords: Tamoxifen Mammary ECM Stroma Estrogen
1. Introduction Progressive growth and development of mammary ducts in the peripubertal mammary gland depends on hormonal and growth factor mediation of interactions * Corresponding author. Tel.: þ1 540 231 8895; fax: þ1 540 231 5014. E-mail address: [email protected] (R.M. Akers). 0739-7240/$ – see front matter Ó 2018 Elsevier Inc. All rights reserved. https://doi.org/10.1016/j.domaniend.2018.05.006
between the epithelial ducts and the surrounding stromal tissues to produce the tubulo-alveolar structures present in the mature functional mammary gland at the onset of lactation [1]. Estrogen and growth hormone are especially critical in regulation of ductal elongation and morphogenesis [2,3]. In an earlier study [4], we showed that prepubertal heifers treated with tamoxifen (TAM) exhibited a 50% reduction in overall mammary development (reduced
50
H.L.M. Tucker et al. / Domestic Animal Endocrinology 65 (2018) 49–55
mass and DNA content). These reductions were similar to those observed in heifers ovariectomized at similar ages [5,6]. Together these results strongly support the relevance of estrogen and/or estrogen signaling in early prepubertal mammary development in the bovine. While the impairment in mammary growth after TAM treatment was not related to either location or proportion of epithelial cells expressing estrogen receptor a (ESR1) or progesterone receptor, there was 623% more ESR1 expression in control (CON) vs TAM calves at the protein level (quantitative immunocytochemistry) and a significant reduction in ESR1 mRNA. Interestingly, there was no significant difference in circulating concentrations of estradiol in TAM- and placebo-treated heifers [4]. Moreover, Geiger et al [7] noted that increased mammary growth in enhanced-fed preweaning heifers was positively correlated with increased expression of ESR1. These data suggest that loss of ESR1 expression at least partially explains the negative effects of ovariectomy (OVX) and TAM treatment in prepubertal heifers and alternatively that increases in ESR1 expression is associated with improved mammary growth in heifers. The impaired mammary development we reported in TAM-treated heifers led us to evaluate stromal development and extracellular matrix (ECM) deposition and expression of a select group of ECM-related genes. These genes were selected based on a comparison between ovariectomized and ovary-intact heifers [6]. Our general observations of reduced complexity of ductal morphology [4] and ongoing recognition of the importance of the ECM in regulation of ductal development [8] also supported our evaluation of ECM-related proteins. We were also influenced by observations of altered expression of collagen and fibronectin and ECM-associated genes in mammary tissue of ovariectomized heifers [9,10] and reports of impacts of TAM on tissue fibrosis in breast cancer models [11]. Our primary objective was to determine if the marked impairment in overall mammary development we previously reported in TAM-treated heifers [4] was associated with alterations in the mammary stroma surrounding the developing ductal structures. Such results would suggest that estrogen signaling in the prepubertal heifer mammary gland regulates not just epithelial cells (as noted previously) but that estrogen signaling might also impact the mammary stroma to influence overall mammary development. This might well provide opportunities to modulate parenchymal development and possibly future milk production by targeting development of the stromal tissue within the developing mammary gland. 2. Materials and methods 2.1. Animals For this evaluation, archived tissues from Tucker et al [4] were utilized. Briefly, all experimental procedures were conducted under the review and approval of the Virginia Polytechnic Institute and State University Institutional Animal Care and Use Committee (11–208 DASC). Sixteen female Holstein heifers were housed individually in calf hutches and randomly assigned to 1 of 2 groups: CON (n ¼ 7) or TAM (n ¼ 8). All calves were fed milk replacer
twice daily and were weaned at 8 wk of age. One CON calf died during the trial. The data from this animal were omitted before statistical analysis. Calves were managed and reared following standard practices at the Virginia Tech Dairy Center. 2.2. Treatments Because no information exists in the literature to suggest an optimal dose of TAM to use to suppress ESR1 action in the bovine mammary gland, we chose an amount equivalent to that administered to human breast cancer patients on TAM therapy [12]. Details describing solution preparation appear in Tucker et al [4]. Heifers were given daily subcutaneous injections of TAM (0.3 mg/kg) or the equivalent volume of excipient in placebo-treated heifers. Heifers were euthanized using Euthasol (Virbac Animal Health, Fort Worth, TX) and exsanguination at 121 1 d of age. 2.3. Mammary tissue sampling At slaughter, udders were removed and bisected medially into left and right hemispheres. The left fore quarter was used to collect samples of parenchyma for real-time quantitative polymerase chain reaction (RT-qPCR) and stored at 80 C until RNA was isolated. The left rear quarter was trimmed of excess mammary fat pad tissue, butterflied, and formalin-fixed for histological analysis and immunohistochemistry. After 24 h in fixative, tissue blocks for embedding and sectioning were collected from each of 3 zones. Zone 1 was just above the teat and gland cistern, zone 2 was approximately midway between the gland cistern and the outer edge of the parenchyma, and zone 3 was near the outer edge at the interface between the mammary parenchymal and mammary fat pad. For Sirius red staining (described in the following), only zone 3 sections were used. Following fixation tissues were stored in 70% ethanol. 2.4. Real-time quantitative polymerase chain reaction Protocols for isolation and purification of RNA followed manufacture instructions based on use of the RNAse Mini Kit (catalog number 74,104; Qiagen, Valencia, CA) and DNAse 1 digestion (catalog number 79,254; Qiagen Inc). Purity of resulting RNA was evaluated with a Nanodrop ND1000 Spectrophotometer (Nanodrop Technologies Inc, Wilmington, DE). Only samples with a ratio of optical density measurements (260 and 280 nm) greater than 1.8 were accepted. The integrity of 18S and 28S ribosomal RNA was evaluated with gel electrophoresis using a 1% agarose gel and visualized by ethidium bromide staining under UV light. Reverse transcription involved synthesizing single stranded cDNA via the High Capacity cDNA Archive Kit (Life Technologies Corporation). Briefly, 4 mg of RNA was reverse transcribed to single-stranded cDNA in a final reaction volume of 40 mL using random primers. The cDNA and no reverse transcriptase controls products were diluted 1:100 in sterile nuclease-free water. A total of 2 mL of cDNA was
H.L.M. Tucker et al. / Domestic Animal Endocrinology 65 (2018) 49–55
used in the RT-qPCR in conjunction with 12.5 mL of SYBR Green dye (Applied Biosystems, Foster City, CA), 9.5 mL of sterile distilled water, 0.5 mL of 10 mM forward primer, and 0.5 mL of 10 mM reverse primer (Table 1). The PCR conditions were 95 C for 10 min, 95 C for 15 s, and 60 C for 1 min. This was repeated for 40 cycles. Reactions were performed in a 7,300 Series Real-Time System, and data were collected and analyzed using SDS software (Applied Biosystems). Triplicates of each sample were assayed, and average coefficient of variance was 2.3%. Each PCR plate contained a no reverse transcriptase control for each sample and a no template control (nuclease-free water instead of cDNA template). Cycle threshold values for replicate samples were collected using the SDS software for each target gene and endogenous reference genes and were exported to Microsoft Office Excel. All primers used have been previously reported [10] and a listing of evaluated genes, associated primers, predicted product sizes, and references appear in Table 1. Three endogenous reference genes were used to normalize the data: protein 13 phosphatase 1 regulatory (inhibitor) subunit 11 (PPP1R11), ribosomal protein S15 A (RPS15 A), and mitochondrial GTPase 1 homolog (MTG1) [15]. Expression of the genes of interest was normalized using the geometric mean of these 3 reference genes. We [15] previously reported that use of these 3 genes were appropriate for normalization of gene expression in mammary parenchymal or mammary fat pad from prepubertal heifers. 2.5. Sirius red staining Formalin-fixed tissues were embedded in paraffin via use of an automated tissue processor (Leica TP 1020; Leica Microsystems Inc, Buffalo Grove, IL). One random tissue block from zone 3 per animal was used to produce microscope slides by slicing 5-mm sections from each block with a microtome (Model HM 340 E; Microm International GmbH, Germany). Sections were mounted on SuperFrost Plus glass slides (VWR International; Radnor, PA) with 3 or 4 serial tissue sections mounted on each slide. Slides were deparaffinized in xylene (3 3 min) and hydrated through a descending graded series of ethanol washes (100%, 2 3 min; 95%, 2 3 min; 70%, 1 3 min) into distilled water (3 min). Hydrated sections were
51
immersed in fresh Wiegert’s Iron hematoxylin solution (ENG Scientific, Inc., Clifton, NJ) for 15 min and then quickly rinsed with tap water. Slides were immersed in Picrosirius red solution (0.5 g Direct Red 80 þ 500 mL of saturated aqueous solution of picric acid) for 1 h. Then slides were washed with acidified water (5 mL glacial acetic acid þ1 L of distilled water; 2 min). Slides were counter-stained with fast green (1% in acidified water) then dehydrated to 100% ethanol followed by xylene (3 3 min each) for mounting of coverslips using Permount (Thermo Fisher Scientific, Waltham, MA). Slides were viewed on a Nikon Eclipse E600 microscope (Nikon Instruments, Inc., Melville, NY) using crosspolarized light and a 4 objective lens. Sirius red-stained collagen exhibited bright birefringence of fibers perpendicular to the transmission axis of the polarized light. Four random images of parenchymal tissue taken from zone 3 tissue from each animal were collected digitally with a QColor 3 color camera (QImaging, Surrey, British Columbia, CA) at a common exposure time. These images were then evaluated using image analysis software (Image Pro Plus version 7.0, Media Cybernetics, Silver Spring, MD) to first determine the relative area occupied by stromal tissue and secondly within stromal tissue regions, the proportion of stromal tissue area occupied by collagen fibers detected under polarized light conditions. Epithelial areas, including lumen, and total image area were outlined using either the irregular area of interest (AOI) or rectangular AOI tools within Image Pro and the resulting object areas were measured via the count/size tool. Stromal tissue area was calculated by the difference between total tissue area and total epithelial area. Bright birefringence (red, orange, and yellow) areas were identified using the count/size tool combined with manual selection of color based on the histogram ranges for red, green, and blue. The sum of each area of birefringence provided the total collagen area within the stromal tissue region for each image. Data compiled for each image were saved and submitted for statistical analysis. 2.6. Statistical analysis Statistics were performed using the Glimmix procedure from SAS statistical package, version 9.2 (SAS Inc.,
Table 1 Primer sequence data for analyzed genes. Gene name
Sequences (50 -30 ) Forward/Reverse
Beta catenin Collagen I E cadherin Epimorphin Fibronectin HSP 90 Laminin MTG1a PPP1R11a RPS15 Aa
TGG CTA CCC AAG CTG ATT TGA/ATG GAT TCC AGA GTC CAG GTA AGA BT030683.1 ATA CCT CCG CCG GTG ACC/AGT CCG CGT ATC CAC AAA GC NM_174520 ATG GGA GGC TGT TTA CAC AGT ATT AA/TAG CTG TTT TCA GAG TGC CTT CAT AY508164 CAG CGT CAG CTA GAA ATA ACT GGA A/AGA GCC TGT CTA GTA ATTT GTG AAT CTG XP597361.3 GGA GAA CAG TGG CAG AAG GAA/AGG TCT GCG CAG TTG TCA K00800.1 TGA AAA GGT GGT TGT GTC AAA CC/TCC AGC CGT ATG TGC TTG TG AB072368.1 GCT GAG GGC ATG GTT CAT G/GAT AGA AAT CCT GAC ACT GCT CAC A BC105436.1 CTT GGA ATC CGA GGA GCC A/CCT GGG ATC ACC AGA GCT GT XM_010819872.1 CCA TCA AAC TTC GGA AAC GG/ACA GCA GCA TTT TGA TGA GCG XP_005223660.1 GAA TGG TGC GCA TGA ATG TC/GAC TTT GGA GCA CGG CCT AA NM_001024541.2
GenBank accession number
Predicted Reference size (bp) 107 90 95 128 95 71 82 101 101 101
Huderson et al. [10] Nicodemus et al. [13] Huderson et al. [10] Huderson et al. [10] Musters et al. [14] Huderson et al. [10] Huderson et al. [10] Piantoni et al. [15] Piantoni et al. [15] Piantoni et al. [15]
a Mitochondrial GTPase 1 homolog (MTG1), protein 13 phosphatase 1 regulatory (inhibitor) subunit 11 (PPP1R11), and ribosomal protein S15 A (RPS15 A) served as endogenous reference genes and were used to normalize gene expression data.
52
H.L.M. Tucker et al. / Domestic Animal Endocrinology 65 (2018) 49–55
gland. The main effects were treatment (TAM vs CON) with random error. Data are given as mean SEM, and significance was declared at P < 0.05. 3. Results 3.1. Effect of tamoxifen on expression of selected ECM-related genes Figure 1 provides a summary of the gene expression data. For the array of tested genes, there was only a small but significant difference (P < 0.05) for expression of tissue laminin (LN). The relative expression of transcripts for these genes varied between genes, for example, collagen I vs fibronectin, but average expression for individual genes was nearly identical in TAM and CON heifers.
Fig. 1. Gene expression for selected ECM-related proteins in mammary parenchyma from TAM and CON heifers is shown. There were substantial differences in overall expression between the various genes but with the exception of a small but significant treatment difference for laminin (*P < 0.05); there were no other treatment differences in gene expression. CON, control; ECM, extracellular matrix; TAM, tamoxifen.
Cary, NC). Essentially this was as described previously [4] with the exception being that there were no zone effects, that is, all samples were from zone 3 within the mammary
3.2. Effect of tamoxifen on stromal tissue area and collagen deposition Figure 2 illustrates some of the detail of ductal structures (20 objective lens) in mammary tissue from zone 3 in TAM and CON heifers. Ductal profiles were generally similar between treatments but structures in TAM heifers were frequently less complex, that is, less evident branching. Furthermore, regions of interlobular stroma in TAM heifers often exhibited large numbers of closely packed fibroblasts (see double arrows, panel A). Panel C illustrates
Fig. 2. Mammary ductal structures in TAM (Panel A) and CON (Panel B) heifers are illustrated. While ductal structures were generally similar between treatments, stromal regions in TAM-treated heifers frequently contained large concentrations of fibroblasts and other mesenchymal cells (double arrows, Panel A). Panel C provides quantitation of the stromal area for TAM and CON heifers within the parenchymal tissue compartment. The proportion averaged 79 2% and did not differ between treatments. The scale bar equals 50 mm. CON, control; TAM, tamoxifen.
H.L.M. Tucker et al. / Domestic Animal Endocrinology 65 (2018) 49–55
the average tissue area occupied by stromal tissue in CONand TAM-treated heifers. Stromal tissue area averaged 79.6 2% and was not different between treatments (CON 80.9 1.2 vs 78.3 1.2%). Figure 3 provides examples of collagen fibrils identified by polarized light imaging of Sirius red-stained tissue sections in TAM (panel A) and CON (panel B) heifers. These low power images (4 objective lens) approximate the average deposition of collagen within stroma by treatment as quantified in panel C. These images also illustrate the automated capture of AOI-containing birefringent collagen fibrils via the Image Pro software. In addition, the epithelial structures were outlined manually. These areas were excluded from total tissue area to provide a measure of total stromal tissue area. In contrast with overall stromal tissue area, deposition of collagen fibrils within the stroma area was more than doubled in TAM-treated heifers (6.9 0.5 vs 15.4 1.4%; P < 0.0001). 4. Discussion It is not surprising that mammary biologists have focused on the growth, development, and functionality of the mammary epithelium. The epithelial structures of the mammary gland are responsible for the synthesis, transport, and secretion of milk. However, it has become
53
increasingly clear that interactions between the stroma and epithelium, in conjunction with hormones and growth factors, ultimately regulate mammary epithelial growth, morphogenesis, lactogenesis, and galactopoiesis [1,8,16]. In this report, we evaluated the impact of the antiestrogen TAM on development of the mammary stroma in the prepubertal bovine mammary gland. To our knowledge, our prior report [4] was the first use of an antiestrogen to determine impacts of estrogen signaling on mammary development in the bovine. It is also relevant to note that TAM treatment also had a dramatic negative effect on the development of the reproductive tract in these calves and notably ESR1 and growth factor signaling [17]. These results support our contention that the negative effects of TAM treatment on mammary development were due to the antiestrogen effects of the drug. Many of the molecular details identifying key ECM components involved in mammary growth and morphogenesis have derived from in vitro experiments, cancer models, and lab animal studies [1,18]. Studies focused on dairy animals are rare. However, several researchers have evaluated the impacts of various ECM components on expression of milk proteins in cultured bovine mammary epithelial cells [19,20], as well as some study of effects of ECM and growth factors on proliferation and morphogenesis [20,21].
Fig. 3. Examples of images showing collagen deposition corresponding to mean treatment values for TAM (Panel A) and CON (Panel B) heifers are provided. Areas of birefringence corresponding with deposition and arrangement of collagen fibrils exposed to polarized light and captured by the Image Pro Software are shown in the multiple area of interest (AOI) outlined in the images (arrows). Also illustrated are the outlines (red) of epithelial structures manually traced so that these regions could be excluded in the measurements of collagen deposition per unit of stromal tissue area. Quantitation of collagen deposition within stromal tissue regions is provided in Panel C. In contrast with overall stromal tissue area, deposition of collagen within the stromal tissue was more than doubled in TAM-treated heifers (*P ¼ 0.0002). The scale bar equals 500 mm. CON, control; TAM, tamoxifen. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)
54
H.L.M. Tucker et al. / Domestic Animal Endocrinology 65 (2018) 49–55
Except for mRNA expression of LN, we did not find significant differences at the gene level for the ECM components tested. It is surprising that heat shock protein 90, collagen1, and epimorphin gene expression was unchanged by TAM. Heat shock protein 90 is involved in regulation of steroid hormone receptors and expression was decreased by OVX, as were expression of ESR1 and PGR [10]. Tamoxifen-treated heifers displayed significant differences in protein level expression of steroid hormone receptors ESR1 (decrease) and PGR (increase), whereas gene expression data displayed more modest impacts [4]. Epimorphin, as a regulator of morphogenesis, would be predicted to be reduced in correspondence with reduced complexity of ductal morphology in TAM-treated heifers. Perhaps epimorphin gene expression was altered earlier in the treatment period. On the other hand, differences in the abundance of collagen were evident at the protein level. Collagen 1 is the predominant type of collagen in stromal tissue. It is likely that increased deposition of collagen by TAM was linked to increased gene expression of collagen 1 at some point during the treatment period. Perhaps changes in mRNA expression that presumably produced increased collagen deposition evident at the time of tissue collection occurred earlier during the treatment period. It is also relevant that Picrosirius red binds to collagen types 1, 2, and 3 [22]. It was not unexpected that expression of ecadherin and b-catenin, which are involved in cellular adhesion cadherin containing cell-cell junctions, were unaffected by TAM given the expression of neither of these genes was impacted by OVX [10]. A limitation of tissue sampling for gene expression is that collection of samples of parenchyma includes both epithelial and stromal tissue components. It is likely that much of the collected RNA was derived from the epithelium rather than the stroma tissue. This makes separation of gene expression of stromal cells vs epithelial cells within the parenchymal tissue difficult. However, in in situ hybridization, use of emerging RNA scope technology, or laser capture dissection followed by gene expression analysis could provide approaches to evaluate a more focused and localized evaluation of ECM-related genes. While the small difference in LN expression between TAM- and placebo treated-heifers is not likely to be biologically relevant, the importance of LNs in mammary ductal development is not in doubt. As reviewed by Maller et al 2010 [18], LNs are large complex glycoproteins that occur in at least 15 combinations of 5 different a, 3 b , and 3 g subunits. Several of these LN complexes are localized in the basal lamina of specific cells or tissue types. The complexes LN111 (a 1, b 1, g1) and LN332 ( a 3, b 3, g2) also referenced as LN1 and LN5, respectively, are common to the mammary epithelium. In the developing bovine mammary gland, overall deposition of LN based on Western blotting was lower in ovariectomized compared with ovary-intact heifers [9] but the antibody used did not distinguish LN complexes associated with the basal lamina of epithelial cells versus other structures, that is, blood vessels. Regardless, LN is important in regulation of mammary morphogenesis in multiple species [18]. Whether or not
changes in synthesis or deposition of LN complexes explain impaired mammary growth after TAM treatment remains to be determined. In contrast, imaging of specific ECM components, at the protein level, allows clear definition of expressed ECM molecules in various cells or within tissue compartments. Certainly, there are major differences in types and abundance of ECM molecules depending on populations of adjacent cells [23]. Despite recognition of the importance of mammary stroma in the development of mammary tissue in dairy animals, direct evaluation is very limited. Some older histological studies have noted apparent changes in composition of the mammary stroma across stages of development or in response to hormone treatment [23,24]. In addition, there has been some emphasis on the role of TGFb [25,26] and differences in stromal and epithelial gene expression in late prepartum cows [27], but stromal tissue development is largely unstudied. 5. Conclusions To our knowledge, our results are the first to quantify deposition of collagen in the stromal tissue of the prepubertal bovine mammary gland. Certainly, use of Sirius red staining to evaluate connective tissues is not new [28] but our quantitative approach and demonstration of a likely role for estrogen signaling in stromal tissue development in the developing bovine mammary gland is novel. Use of quantitative imaging to evaluate changes in stromal tissue development in the mammary gland may open new approaches to better understand important interactions between ductal morphogenesis and the ECM. Acknowledgments The authors acknowledge support to R. M. A. provided by the endowment for the Horace E. and Elizabeth F. Alphin Professorship and support from the Virginia Tech Agricultural Experiment Station. We also acknowledge support from USDA-NIFA-AFRI grant no. 2006–35206-16699,‘Ovarian Regulation of Stem Cells and IGF-I Axis Molecules in prepubertal Heifer Mammary Gland’ to R.M. A. and S.E. Ellis which provided some reagents and supplies. As well as similar support from USDA-NIFA-AFRI, 2016–67015-24575, ‘Impact of Pre-Weaning Nutrition on Endocrine Induction of Mammary Development in Dairy Heifers’ awarded to R.M. Akers. References [1] Nelson CM, Bissell MJ. Of extracellular matrix, scaffolds, and signaling: tissue architecture regulates development, homeostasis, and cancer. Annu Rev Cell Dev Biol 2006;22:287–309. [2] Akers RM. A 100-Year review: mammary development and lactation. J Dairy Sci 2017;100:10332–52. [3] Capuco AV, Ellis S, Wood DL, Akers RM, Garrett W. Postnatal mammary ductal growth: three-dimensional imaging of cell proliferation, effects of estrogen treatment, and expression of steroid receptors in prepubertal calves. Tissue Cell 2002;34:143–54. [4] Tucker HLM, Parsons CLM, Ellis S, Rhoads ML, Akers RM. Tamoxifen impairs prepubertal mammary development and alters expression of estrogen receptor a (ESR1) and progesterone receptors (PGR). Domest Anim Endocrinol 2016;54:95–105.
H.L.M. Tucker et al. / Domestic Animal Endocrinology 65 (2018) 49–55 [5] Purup S, Sejrsen K, Foldager J, Akers RM. Effect of exogenous bovine growth hormone and ovariectomy on prepubertal mammary growth, serum hormones and acute in-vitro proliferative response of mammary explants from Holstein heifers. J Endocrinol 1993;139:19–26. [6] Velayudhan BT, Huderson BP, McGilliard ML, Jiang H, Ellis SE, Akers RM. Effect of staged ovariectomy on measures of mammary growth and development in prepubertal dairy heifers. Animal 2012;6:941–51. [7] Geiger AJ, Parsons CLM, Akers RM. Feeding an enhanced diet to Holstein heifers during the preweaning period alters steroid receptor expression and increases cellular proliferation. J Dairy Sci 2017;100:8534–43. [8] Bonnans C, Chou J, Werb J. Remodeling the extracellular matrix in development and disease. Nat Rev Mol Cell Biol 2014;15:786–801. [9] Berry SD, Howard RD, Akers RM. Mammary localization and abundance of laminin, fibronectin, and collagen IV proteins in prepubertal heifers. J Dairy Sci 2003;86:2864–74. [10] Huderson BP, Velayudhan BT, Pearson RE, Ellis SE, Akers RM. Effect of exogenous somatotropin and staged ovariectomy on mRNA expression of select ECM-related proteins in mammary tissue of prepubertal calves. Open J Anim Sci 2013;3:160–8. [11] Hattar R, Maller O, McDaniel S, Hansen KC, Hedman KJ, Lyons TR, Lucia S, Wilson Jr RS, Schedin P. Tamoxifen induces pleiotrophic changes in mammary stroma resuling in extracellular matrix that suppresses transformed phenotypes. Breast Cancer Res 2009;11:R5. [12] Jordan VC. Tamoxifen: catalyst for the change to targeted therapy. Eur J Cancer 2008;44:30–8. [13] Niocodemus GD, Villanueva I, Bryant SJ. Mechanical stimulation of TMJ condylar chondrocytes encapsulated in PEG hydrogels. J Biomed Mater Res A 2007;83:323–31. [14] Swanson EW, Poffenbarger JI. Mammary gland development of dairy heifers during their first gestation. J Dairy Sci 1979;62:702–14. [15] Piantoni P, Bionaz M, Graugnard DE, Daniels KM, Akers RM, Loor JJ. Gene expression ratio stability evaluation in prepubertal bovine mammary tissue from calves fed different milk replacers reveals novel internal controls for quantitative polymerase chain reaction. J Nutr 2008;138:1158–64. [16] Schedin P, Hovey RC. Editorial: the mammary stroma in normal development and function. J Mammary Gland Biol Neoplasia 2010; 15:275–7.
55
[17] Al Naib A, Tucker HLM, Xie G, Keisler DH, Bartol FF, Rhoads RP, Akers RM, Rhoads ML. Prepubertal tamoxifen treatment affects development of heifer reproductive tissues and related signaling pathways. J Dairy Sci 2016;99:5780–92. [18] Maller O, Martinson H, Schedin P. Extracellular matrix composition reveals complex and dynamic stromal-epithelial interactions in the mammary gland. J Mammary Gland Biol Neoplasia 2010;15:301–18. [19] Riley LG, Gardiner-Garden M, Thomson PC, Wynn PC, Williamson P, Raadsma HW, Sheehy PA. The influence of extracellular matrix and prolactin on global gene expression profiles of primary bovine mammary epithelial cells in vitro. Anim Genet 2010;41:55–63. [20] Stiening CM, Hoying JB, Abdallah MB, Hoying AM, Pandey R, Greer K, Collier R. The effects of endocrine and mechanical stimulation on stage I lactogenesis in bovine mammary epithelial cells. J Dairy Sci 2008;91:1053–66. [21] Ellis S, Purup S, Sejrsen K, Akers RM. Growth and morphogenesis of epithelial cell organoids from peripheral and medial mammary parenchyma of prepubertal heifers. J Dairy Sci 2000;83:952–61. [22] Rittie L. Method for picosirius red-polarization detection of collage fibers in tissue sections. Methods Mol Biol 2017;1627:395–407. Chapter 26, Laure Rittie (ed.), Springer ScienceþBusiness Media LLC. [23] Vracko R. Basal lamina scaffold-anatomy and significance for maintenance of orderly tissue structure. Am J Pathol 1974;77:314–46. [24] Sud SC, Tucker HA, Meites J. Estrogen-progesterone requirements for udder development in ovariectomized heifers. J Dairy Sci 1968; 51:210–4. [25] Musters S, Coughlan K, McFadden T, Maple R, Mulvey T, Plaut K. Exogenous TGF-b1 promotes stromal development in heifer mammary gland. J Dairy Sci 2004;87:896–904. [26] De Vries LD, Casey T, Dover H, VandeHaar MJ, Plaut K. Effects of transforming growth factor-b on mammary remodeling during the dry period of dairy cows. J Dairy Sci 2011;94:6036–46. [27] Casey T, Dover H, Lieshman J, DeVries L, Kiupel M, VandeHaar M, Plaut K. Transcriptome analysis of epithelial and stromal contributions to mammogenesis in three week prepartum cows. PLoS One 2011;6:e22541. [28] Junqueira LCU, Bignolas G, Brentani RR. Picosirius staining plus polarization microscopy, a specific method for collage detection in tissue sections. Histochem J 1979;11:447–55.
Contents lists available at ScienceDirect
Domestic Animal Endocrinology journal homepage: www.journals.elsevier.com/ domestic-animal-endocrinology
Excess deposition of collagen in mammary glands of tamoxifen-treated Holstein heifers is associated with impaired mammary growth H.L.M. Tucker, J. Holdridge, C.L.M. Parsons, R.M. Akers* Department of Dairy Science, Virginia Tech, Blacksburg, VA 24061, USA
a r t i c l e i n f o
a b s t r a c t
Article history: Received 7 February 2018 Received in revised form 26 April 2018 Accepted 11 May 2018
It is established that the ovary and estrogen are essential to bovine mammary development with the onset of puberty. Recent studies have shown that ovariectomy in the very early prepubertal period, well before onset of puberty, also dramatically impairs mammary growth. Similarly, prepubertal heifers treated with the antiestrogen tamoxifen (TAM) also exhibit markedly impaired mammary growth in correspondence with reduced estrogen receptor a (ESR1) expression. Our objective was to evaluate the effect of TAM on the mammary stroma and specifically to determine if the reported decrease in mammary development was related to changes in TAM-induced alterations in the stroma surrounding the mammary parenchyma. Briefly, 16 Holstein heifers calves were randomly assigned to one of 2 treatment groups: TAM-injected or control. Calves were administered TAM (0.3 mg kg1 d1) or placebo from 28 to 120 d of age. At day 120, calves were euthanized and udders removed. Mammary tissue from near the boundary between the parenchyma and surrounding mammary fat pad was collected for histology and morphometric analysis, expression of selected extracellular matrix–related genes, and quantitation of stromal collagen deposition by study of Sirius Red-stained tissue sections imaged with polarized light. Compared with tissue from control heifers, TAM heifers frequently exhibited areas with abundant fibroblasts and mesenchymal cells especially within the intralobular stroma, as well as less complex ductal structures. Among the array of extracellular matrix–related genes tested, only a small difference (P < 0.05) in expression of laminin was found between treatments. The relative tissue area occupied by stromal tissue was not impacted by treatment. However, the deposition of collagen within the stromal tissue was more than doubled (P < 0.0001) in TAM-treated heifers. These data suggest that blocking ESR1 expression with TAM allows for excessive collagen deposition in the stroma surrounding the developing epithelial structures and that this interferes with both the degree of overall mammary parenchymal development, as well as the pattern of normal ductal morphogenesis. Ó 2018 Elsevier Inc. All rights reserved.
Keywords: Tamoxifen Mammary ECM Stroma Estrogen
1. Introduction Progressive growth and development of mammary ducts in the peripubertal mammary gland depends on hormonal and growth factor mediation of interactions * Corresponding author. Tel.: þ1 540 231 8895; fax: þ1 540 231 5014. E-mail address: [email protected] (R.M. Akers). 0739-7240/$ – see front matter Ó 2018 Elsevier Inc. All rights reserved. https://doi.org/10.1016/j.domaniend.2018.05.006
between the epithelial ducts and the surrounding stromal tissues to produce the tubulo-alveolar structures present in the mature functional mammary gland at the onset of lactation [1]. Estrogen and growth hormone are especially critical in regulation of ductal elongation and morphogenesis [2,3]. In an earlier study [4], we showed that prepubertal heifers treated with tamoxifen (TAM) exhibited a 50% reduction in overall mammary development (reduced
50
H.L.M. Tucker et al. / Domestic Animal Endocrinology 65 (2018) 49–55
mass and DNA content). These reductions were similar to those observed in heifers ovariectomized at similar ages [5,6]. Together these results strongly support the relevance of estrogen and/or estrogen signaling in early prepubertal mammary development in the bovine. While the impairment in mammary growth after TAM treatment was not related to either location or proportion of epithelial cells expressing estrogen receptor a (ESR1) or progesterone receptor, there was 623% more ESR1 expression in control (CON) vs TAM calves at the protein level (quantitative immunocytochemistry) and a significant reduction in ESR1 mRNA. Interestingly, there was no significant difference in circulating concentrations of estradiol in TAM- and placebo-treated heifers [4]. Moreover, Geiger et al [7] noted that increased mammary growth in enhanced-fed preweaning heifers was positively correlated with increased expression of ESR1. These data suggest that loss of ESR1 expression at least partially explains the negative effects of ovariectomy (OVX) and TAM treatment in prepubertal heifers and alternatively that increases in ESR1 expression is associated with improved mammary growth in heifers. The impaired mammary development we reported in TAM-treated heifers led us to evaluate stromal development and extracellular matrix (ECM) deposition and expression of a select group of ECM-related genes. These genes were selected based on a comparison between ovariectomized and ovary-intact heifers [6]. Our general observations of reduced complexity of ductal morphology [4] and ongoing recognition of the importance of the ECM in regulation of ductal development [8] also supported our evaluation of ECM-related proteins. We were also influenced by observations of altered expression of collagen and fibronectin and ECM-associated genes in mammary tissue of ovariectomized heifers [9,10] and reports of impacts of TAM on tissue fibrosis in breast cancer models [11]. Our primary objective was to determine if the marked impairment in overall mammary development we previously reported in TAM-treated heifers [4] was associated with alterations in the mammary stroma surrounding the developing ductal structures. Such results would suggest that estrogen signaling in the prepubertal heifer mammary gland regulates not just epithelial cells (as noted previously) but that estrogen signaling might also impact the mammary stroma to influence overall mammary development. This might well provide opportunities to modulate parenchymal development and possibly future milk production by targeting development of the stromal tissue within the developing mammary gland. 2. Materials and methods 2.1. Animals For this evaluation, archived tissues from Tucker et al [4] were utilized. Briefly, all experimental procedures were conducted under the review and approval of the Virginia Polytechnic Institute and State University Institutional Animal Care and Use Committee (11–208 DASC). Sixteen female Holstein heifers were housed individually in calf hutches and randomly assigned to 1 of 2 groups: CON (n ¼ 7) or TAM (n ¼ 8). All calves were fed milk replacer
twice daily and were weaned at 8 wk of age. One CON calf died during the trial. The data from this animal were omitted before statistical analysis. Calves were managed and reared following standard practices at the Virginia Tech Dairy Center. 2.2. Treatments Because no information exists in the literature to suggest an optimal dose of TAM to use to suppress ESR1 action in the bovine mammary gland, we chose an amount equivalent to that administered to human breast cancer patients on TAM therapy [12]. Details describing solution preparation appear in Tucker et al [4]. Heifers were given daily subcutaneous injections of TAM (0.3 mg/kg) or the equivalent volume of excipient in placebo-treated heifers. Heifers were euthanized using Euthasol (Virbac Animal Health, Fort Worth, TX) and exsanguination at 121 1 d of age. 2.3. Mammary tissue sampling At slaughter, udders were removed and bisected medially into left and right hemispheres. The left fore quarter was used to collect samples of parenchyma for real-time quantitative polymerase chain reaction (RT-qPCR) and stored at 80 C until RNA was isolated. The left rear quarter was trimmed of excess mammary fat pad tissue, butterflied, and formalin-fixed for histological analysis and immunohistochemistry. After 24 h in fixative, tissue blocks for embedding and sectioning were collected from each of 3 zones. Zone 1 was just above the teat and gland cistern, zone 2 was approximately midway between the gland cistern and the outer edge of the parenchyma, and zone 3 was near the outer edge at the interface between the mammary parenchymal and mammary fat pad. For Sirius red staining (described in the following), only zone 3 sections were used. Following fixation tissues were stored in 70% ethanol. 2.4. Real-time quantitative polymerase chain reaction Protocols for isolation and purification of RNA followed manufacture instructions based on use of the RNAse Mini Kit (catalog number 74,104; Qiagen, Valencia, CA) and DNAse 1 digestion (catalog number 79,254; Qiagen Inc). Purity of resulting RNA was evaluated with a Nanodrop ND1000 Spectrophotometer (Nanodrop Technologies Inc, Wilmington, DE). Only samples with a ratio of optical density measurements (260 and 280 nm) greater than 1.8 were accepted. The integrity of 18S and 28S ribosomal RNA was evaluated with gel electrophoresis using a 1% agarose gel and visualized by ethidium bromide staining under UV light. Reverse transcription involved synthesizing single stranded cDNA via the High Capacity cDNA Archive Kit (Life Technologies Corporation). Briefly, 4 mg of RNA was reverse transcribed to single-stranded cDNA in a final reaction volume of 40 mL using random primers. The cDNA and no reverse transcriptase controls products were diluted 1:100 in sterile nuclease-free water. A total of 2 mL of cDNA was
H.L.M. Tucker et al. / Domestic Animal Endocrinology 65 (2018) 49–55
used in the RT-qPCR in conjunction with 12.5 mL of SYBR Green dye (Applied Biosystems, Foster City, CA), 9.5 mL of sterile distilled water, 0.5 mL of 10 mM forward primer, and 0.5 mL of 10 mM reverse primer (Table 1). The PCR conditions were 95 C for 10 min, 95 C for 15 s, and 60 C for 1 min. This was repeated for 40 cycles. Reactions were performed in a 7,300 Series Real-Time System, and data were collected and analyzed using SDS software (Applied Biosystems). Triplicates of each sample were assayed, and average coefficient of variance was 2.3%. Each PCR plate contained a no reverse transcriptase control for each sample and a no template control (nuclease-free water instead of cDNA template). Cycle threshold values for replicate samples were collected using the SDS software for each target gene and endogenous reference genes and were exported to Microsoft Office Excel. All primers used have been previously reported [10] and a listing of evaluated genes, associated primers, predicted product sizes, and references appear in Table 1. Three endogenous reference genes were used to normalize the data: protein 13 phosphatase 1 regulatory (inhibitor) subunit 11 (PPP1R11), ribosomal protein S15 A (RPS15 A), and mitochondrial GTPase 1 homolog (MTG1) [15]. Expression of the genes of interest was normalized using the geometric mean of these 3 reference genes. We [15] previously reported that use of these 3 genes were appropriate for normalization of gene expression in mammary parenchymal or mammary fat pad from prepubertal heifers. 2.5. Sirius red staining Formalin-fixed tissues were embedded in paraffin via use of an automated tissue processor (Leica TP 1020; Leica Microsystems Inc, Buffalo Grove, IL). One random tissue block from zone 3 per animal was used to produce microscope slides by slicing 5-mm sections from each block with a microtome (Model HM 340 E; Microm International GmbH, Germany). Sections were mounted on SuperFrost Plus glass slides (VWR International; Radnor, PA) with 3 or 4 serial tissue sections mounted on each slide. Slides were deparaffinized in xylene (3 3 min) and hydrated through a descending graded series of ethanol washes (100%, 2 3 min; 95%, 2 3 min; 70%, 1 3 min) into distilled water (3 min). Hydrated sections were
51
immersed in fresh Wiegert’s Iron hematoxylin solution (ENG Scientific, Inc., Clifton, NJ) for 15 min and then quickly rinsed with tap water. Slides were immersed in Picrosirius red solution (0.5 g Direct Red 80 þ 500 mL of saturated aqueous solution of picric acid) for 1 h. Then slides were washed with acidified water (5 mL glacial acetic acid þ1 L of distilled water; 2 min). Slides were counter-stained with fast green (1% in acidified water) then dehydrated to 100% ethanol followed by xylene (3 3 min each) for mounting of coverslips using Permount (Thermo Fisher Scientific, Waltham, MA). Slides were viewed on a Nikon Eclipse E600 microscope (Nikon Instruments, Inc., Melville, NY) using crosspolarized light and a 4 objective lens. Sirius red-stained collagen exhibited bright birefringence of fibers perpendicular to the transmission axis of the polarized light. Four random images of parenchymal tissue taken from zone 3 tissue from each animal were collected digitally with a QColor 3 color camera (QImaging, Surrey, British Columbia, CA) at a common exposure time. These images were then evaluated using image analysis software (Image Pro Plus version 7.0, Media Cybernetics, Silver Spring, MD) to first determine the relative area occupied by stromal tissue and secondly within stromal tissue regions, the proportion of stromal tissue area occupied by collagen fibers detected under polarized light conditions. Epithelial areas, including lumen, and total image area were outlined using either the irregular area of interest (AOI) or rectangular AOI tools within Image Pro and the resulting object areas were measured via the count/size tool. Stromal tissue area was calculated by the difference between total tissue area and total epithelial area. Bright birefringence (red, orange, and yellow) areas were identified using the count/size tool combined with manual selection of color based on the histogram ranges for red, green, and blue. The sum of each area of birefringence provided the total collagen area within the stromal tissue region for each image. Data compiled for each image were saved and submitted for statistical analysis. 2.6. Statistical analysis Statistics were performed using the Glimmix procedure from SAS statistical package, version 9.2 (SAS Inc.,
Table 1 Primer sequence data for analyzed genes. Gene name
Sequences (50 -30 ) Forward/Reverse
Beta catenin Collagen I E cadherin Epimorphin Fibronectin HSP 90 Laminin MTG1a PPP1R11a RPS15 Aa
TGG CTA CCC AAG CTG ATT TGA/ATG GAT TCC AGA GTC CAG GTA AGA BT030683.1 ATA CCT CCG CCG GTG ACC/AGT CCG CGT ATC CAC AAA GC NM_174520 ATG GGA GGC TGT TTA CAC AGT ATT AA/TAG CTG TTT TCA GAG TGC CTT CAT AY508164 CAG CGT CAG CTA GAA ATA ACT GGA A/AGA GCC TGT CTA GTA ATTT GTG AAT CTG XP597361.3 GGA GAA CAG TGG CAG AAG GAA/AGG TCT GCG CAG TTG TCA K00800.1 TGA AAA GGT GGT TGT GTC AAA CC/TCC AGC CGT ATG TGC TTG TG AB072368.1 GCT GAG GGC ATG GTT CAT G/GAT AGA AAT CCT GAC ACT GCT CAC A BC105436.1 CTT GGA ATC CGA GGA GCC A/CCT GGG ATC ACC AGA GCT GT XM_010819872.1 CCA TCA AAC TTC GGA AAC GG/ACA GCA GCA TTT TGA TGA GCG XP_005223660.1 GAA TGG TGC GCA TGA ATG TC/GAC TTT GGA GCA CGG CCT AA NM_001024541.2
GenBank accession number
Predicted Reference size (bp) 107 90 95 128 95 71 82 101 101 101
Huderson et al. [10] Nicodemus et al. [13] Huderson et al. [10] Huderson et al. [10] Musters et al. [14] Huderson et al. [10] Huderson et al. [10] Piantoni et al. [15] Piantoni et al. [15] Piantoni et al. [15]
a Mitochondrial GTPase 1 homolog (MTG1), protein 13 phosphatase 1 regulatory (inhibitor) subunit 11 (PPP1R11), and ribosomal protein S15 A (RPS15 A) served as endogenous reference genes and were used to normalize gene expression data.
52
H.L.M. Tucker et al. / Domestic Animal Endocrinology 65 (2018) 49–55
gland. The main effects were treatment (TAM vs CON) with random error. Data are given as mean SEM, and significance was declared at P < 0.05. 3. Results 3.1. Effect of tamoxifen on expression of selected ECM-related genes Figure 1 provides a summary of the gene expression data. For the array of tested genes, there was only a small but significant difference (P < 0.05) for expression of tissue laminin (LN). The relative expression of transcripts for these genes varied between genes, for example, collagen I vs fibronectin, but average expression for individual genes was nearly identical in TAM and CON heifers.
Fig. 1. Gene expression for selected ECM-related proteins in mammary parenchyma from TAM and CON heifers is shown. There were substantial differences in overall expression between the various genes but with the exception of a small but significant treatment difference for laminin (*P < 0.05); there were no other treatment differences in gene expression. CON, control; ECM, extracellular matrix; TAM, tamoxifen.
Cary, NC). Essentially this was as described previously [4] with the exception being that there were no zone effects, that is, all samples were from zone 3 within the mammary
3.2. Effect of tamoxifen on stromal tissue area and collagen deposition Figure 2 illustrates some of the detail of ductal structures (20 objective lens) in mammary tissue from zone 3 in TAM and CON heifers. Ductal profiles were generally similar between treatments but structures in TAM heifers were frequently less complex, that is, less evident branching. Furthermore, regions of interlobular stroma in TAM heifers often exhibited large numbers of closely packed fibroblasts (see double arrows, panel A). Panel C illustrates
Fig. 2. Mammary ductal structures in TAM (Panel A) and CON (Panel B) heifers are illustrated. While ductal structures were generally similar between treatments, stromal regions in TAM-treated heifers frequently contained large concentrations of fibroblasts and other mesenchymal cells (double arrows, Panel A). Panel C provides quantitation of the stromal area for TAM and CON heifers within the parenchymal tissue compartment. The proportion averaged 79 2% and did not differ between treatments. The scale bar equals 50 mm. CON, control; TAM, tamoxifen.
H.L.M. Tucker et al. / Domestic Animal Endocrinology 65 (2018) 49–55
the average tissue area occupied by stromal tissue in CONand TAM-treated heifers. Stromal tissue area averaged 79.6 2% and was not different between treatments (CON 80.9 1.2 vs 78.3 1.2%). Figure 3 provides examples of collagen fibrils identified by polarized light imaging of Sirius red-stained tissue sections in TAM (panel A) and CON (panel B) heifers. These low power images (4 objective lens) approximate the average deposition of collagen within stroma by treatment as quantified in panel C. These images also illustrate the automated capture of AOI-containing birefringent collagen fibrils via the Image Pro software. In addition, the epithelial structures were outlined manually. These areas were excluded from total tissue area to provide a measure of total stromal tissue area. In contrast with overall stromal tissue area, deposition of collagen fibrils within the stroma area was more than doubled in TAM-treated heifers (6.9 0.5 vs 15.4 1.4%; P < 0.0001). 4. Discussion It is not surprising that mammary biologists have focused on the growth, development, and functionality of the mammary epithelium. The epithelial structures of the mammary gland are responsible for the synthesis, transport, and secretion of milk. However, it has become
53
increasingly clear that interactions between the stroma and epithelium, in conjunction with hormones and growth factors, ultimately regulate mammary epithelial growth, morphogenesis, lactogenesis, and galactopoiesis [1,8,16]. In this report, we evaluated the impact of the antiestrogen TAM on development of the mammary stroma in the prepubertal bovine mammary gland. To our knowledge, our prior report [4] was the first use of an antiestrogen to determine impacts of estrogen signaling on mammary development in the bovine. It is also relevant to note that TAM treatment also had a dramatic negative effect on the development of the reproductive tract in these calves and notably ESR1 and growth factor signaling [17]. These results support our contention that the negative effects of TAM treatment on mammary development were due to the antiestrogen effects of the drug. Many of the molecular details identifying key ECM components involved in mammary growth and morphogenesis have derived from in vitro experiments, cancer models, and lab animal studies [1,18]. Studies focused on dairy animals are rare. However, several researchers have evaluated the impacts of various ECM components on expression of milk proteins in cultured bovine mammary epithelial cells [19,20], as well as some study of effects of ECM and growth factors on proliferation and morphogenesis [20,21].
Fig. 3. Examples of images showing collagen deposition corresponding to mean treatment values for TAM (Panel A) and CON (Panel B) heifers are provided. Areas of birefringence corresponding with deposition and arrangement of collagen fibrils exposed to polarized light and captured by the Image Pro Software are shown in the multiple area of interest (AOI) outlined in the images (arrows). Also illustrated are the outlines (red) of epithelial structures manually traced so that these regions could be excluded in the measurements of collagen deposition per unit of stromal tissue area. Quantitation of collagen deposition within stromal tissue regions is provided in Panel C. In contrast with overall stromal tissue area, deposition of collagen within the stromal tissue was more than doubled in TAM-treated heifers (*P ¼ 0.0002). The scale bar equals 500 mm. CON, control; TAM, tamoxifen. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)
54
H.L.M. Tucker et al. / Domestic Animal Endocrinology 65 (2018) 49–55
Except for mRNA expression of LN, we did not find significant differences at the gene level for the ECM components tested. It is surprising that heat shock protein 90, collagen1, and epimorphin gene expression was unchanged by TAM. Heat shock protein 90 is involved in regulation of steroid hormone receptors and expression was decreased by OVX, as were expression of ESR1 and PGR [10]. Tamoxifen-treated heifers displayed significant differences in protein level expression of steroid hormone receptors ESR1 (decrease) and PGR (increase), whereas gene expression data displayed more modest impacts [4]. Epimorphin, as a regulator of morphogenesis, would be predicted to be reduced in correspondence with reduced complexity of ductal morphology in TAM-treated heifers. Perhaps epimorphin gene expression was altered earlier in the treatment period. On the other hand, differences in the abundance of collagen were evident at the protein level. Collagen 1 is the predominant type of collagen in stromal tissue. It is likely that increased deposition of collagen by TAM was linked to increased gene expression of collagen 1 at some point during the treatment period. Perhaps changes in mRNA expression that presumably produced increased collagen deposition evident at the time of tissue collection occurred earlier during the treatment period. It is also relevant that Picrosirius red binds to collagen types 1, 2, and 3 [22]. It was not unexpected that expression of ecadherin and b-catenin, which are involved in cellular adhesion cadherin containing cell-cell junctions, were unaffected by TAM given the expression of neither of these genes was impacted by OVX [10]. A limitation of tissue sampling for gene expression is that collection of samples of parenchyma includes both epithelial and stromal tissue components. It is likely that much of the collected RNA was derived from the epithelium rather than the stroma tissue. This makes separation of gene expression of stromal cells vs epithelial cells within the parenchymal tissue difficult. However, in in situ hybridization, use of emerging RNA scope technology, or laser capture dissection followed by gene expression analysis could provide approaches to evaluate a more focused and localized evaluation of ECM-related genes. While the small difference in LN expression between TAM- and placebo treated-heifers is not likely to be biologically relevant, the importance of LNs in mammary ductal development is not in doubt. As reviewed by Maller et al 2010 [18], LNs are large complex glycoproteins that occur in at least 15 combinations of 5 different a, 3 b , and 3 g subunits. Several of these LN complexes are localized in the basal lamina of specific cells or tissue types. The complexes LN111 (a 1, b 1, g1) and LN332 ( a 3, b 3, g2) also referenced as LN1 and LN5, respectively, are common to the mammary epithelium. In the developing bovine mammary gland, overall deposition of LN based on Western blotting was lower in ovariectomized compared with ovary-intact heifers [9] but the antibody used did not distinguish LN complexes associated with the basal lamina of epithelial cells versus other structures, that is, blood vessels. Regardless, LN is important in regulation of mammary morphogenesis in multiple species [18]. Whether or not
changes in synthesis or deposition of LN complexes explain impaired mammary growth after TAM treatment remains to be determined. In contrast, imaging of specific ECM components, at the protein level, allows clear definition of expressed ECM molecules in various cells or within tissue compartments. Certainly, there are major differences in types and abundance of ECM molecules depending on populations of adjacent cells [23]. Despite recognition of the importance of mammary stroma in the development of mammary tissue in dairy animals, direct evaluation is very limited. Some older histological studies have noted apparent changes in composition of the mammary stroma across stages of development or in response to hormone treatment [23,24]. In addition, there has been some emphasis on the role of TGFb [25,26] and differences in stromal and epithelial gene expression in late prepartum cows [27], but stromal tissue development is largely unstudied. 5. Conclusions To our knowledge, our results are the first to quantify deposition of collagen in the stromal tissue of the prepubertal bovine mammary gland. Certainly, use of Sirius red staining to evaluate connective tissues is not new [28] but our quantitative approach and demonstration of a likely role for estrogen signaling in stromal tissue development in the developing bovine mammary gland is novel. Use of quantitative imaging to evaluate changes in stromal tissue development in the mammary gland may open new approaches to better understand important interactions between ductal morphogenesis and the ECM. Acknowledgments The authors acknowledge support to R. M. A. provided by the endowment for the Horace E. and Elizabeth F. Alphin Professorship and support from the Virginia Tech Agricultural Experiment Station. We also acknowledge support from USDA-NIFA-AFRI grant no. 2006–35206-16699,‘Ovarian Regulation of Stem Cells and IGF-I Axis Molecules in prepubertal Heifer Mammary Gland’ to R.M. A. and S.E. Ellis which provided some reagents and supplies. As well as similar support from USDA-NIFA-AFRI, 2016–67015-24575, ‘Impact of Pre-Weaning Nutrition on Endocrine Induction of Mammary Development in Dairy Heifers’ awarded to R.M. Akers. References [1] Nelson CM, Bissell MJ. Of extracellular matrix, scaffolds, and signaling: tissue architecture regulates development, homeostasis, and cancer. Annu Rev Cell Dev Biol 2006;22:287–309. [2] Akers RM. A 100-Year review: mammary development and lactation. J Dairy Sci 2017;100:10332–52. [3] Capuco AV, Ellis S, Wood DL, Akers RM, Garrett W. Postnatal mammary ductal growth: three-dimensional imaging of cell proliferation, effects of estrogen treatment, and expression of steroid receptors in prepubertal calves. Tissue Cell 2002;34:143–54. [4] Tucker HLM, Parsons CLM, Ellis S, Rhoads ML, Akers RM. Tamoxifen impairs prepubertal mammary development and alters expression of estrogen receptor a (ESR1) and progesterone receptors (PGR). Domest Anim Endocrinol 2016;54:95–105.
H.L.M. Tucker et al. / Domestic Animal Endocrinology 65 (2018) 49–55 [5] Purup S, Sejrsen K, Foldager J, Akers RM. Effect of exogenous bovine growth hormone and ovariectomy on prepubertal mammary growth, serum hormones and acute in-vitro proliferative response of mammary explants from Holstein heifers. J Endocrinol 1993;139:19–26. [6] Velayudhan BT, Huderson BP, McGilliard ML, Jiang H, Ellis SE, Akers RM. Effect of staged ovariectomy on measures of mammary growth and development in prepubertal dairy heifers. Animal 2012;6:941–51. [7] Geiger AJ, Parsons CLM, Akers RM. Feeding an enhanced diet to Holstein heifers during the preweaning period alters steroid receptor expression and increases cellular proliferation. J Dairy Sci 2017;100:8534–43. [8] Bonnans C, Chou J, Werb J. Remodeling the extracellular matrix in development and disease. Nat Rev Mol Cell Biol 2014;15:786–801. [9] Berry SD, Howard RD, Akers RM. Mammary localization and abundance of laminin, fibronectin, and collagen IV proteins in prepubertal heifers. J Dairy Sci 2003;86:2864–74. [10] Huderson BP, Velayudhan BT, Pearson RE, Ellis SE, Akers RM. Effect of exogenous somatotropin and staged ovariectomy on mRNA expression of select ECM-related proteins in mammary tissue of prepubertal calves. Open J Anim Sci 2013;3:160–8. [11] Hattar R, Maller O, McDaniel S, Hansen KC, Hedman KJ, Lyons TR, Lucia S, Wilson Jr RS, Schedin P. Tamoxifen induces pleiotrophic changes in mammary stroma resuling in extracellular matrix that suppresses transformed phenotypes. Breast Cancer Res 2009;11:R5. [12] Jordan VC. Tamoxifen: catalyst for the change to targeted therapy. Eur J Cancer 2008;44:30–8. [13] Niocodemus GD, Villanueva I, Bryant SJ. Mechanical stimulation of TMJ condylar chondrocytes encapsulated in PEG hydrogels. J Biomed Mater Res A 2007;83:323–31. [14] Swanson EW, Poffenbarger JI. Mammary gland development of dairy heifers during their first gestation. J Dairy Sci 1979;62:702–14. [15] Piantoni P, Bionaz M, Graugnard DE, Daniels KM, Akers RM, Loor JJ. Gene expression ratio stability evaluation in prepubertal bovine mammary tissue from calves fed different milk replacers reveals novel internal controls for quantitative polymerase chain reaction. J Nutr 2008;138:1158–64. [16] Schedin P, Hovey RC. Editorial: the mammary stroma in normal development and function. J Mammary Gland Biol Neoplasia 2010; 15:275–7.
55
[17] Al Naib A, Tucker HLM, Xie G, Keisler DH, Bartol FF, Rhoads RP, Akers RM, Rhoads ML. Prepubertal tamoxifen treatment affects development of heifer reproductive tissues and related signaling pathways. J Dairy Sci 2016;99:5780–92. [18] Maller O, Martinson H, Schedin P. Extracellular matrix composition reveals complex and dynamic stromal-epithelial interactions in the mammary gland. J Mammary Gland Biol Neoplasia 2010;15:301–18. [19] Riley LG, Gardiner-Garden M, Thomson PC, Wynn PC, Williamson P, Raadsma HW, Sheehy PA. The influence of extracellular matrix and prolactin on global gene expression profiles of primary bovine mammary epithelial cells in vitro. Anim Genet 2010;41:55–63. [20] Stiening CM, Hoying JB, Abdallah MB, Hoying AM, Pandey R, Greer K, Collier R. The effects of endocrine and mechanical stimulation on stage I lactogenesis in bovine mammary epithelial cells. J Dairy Sci 2008;91:1053–66. [21] Ellis S, Purup S, Sejrsen K, Akers RM. Growth and morphogenesis of epithelial cell organoids from peripheral and medial mammary parenchyma of prepubertal heifers. J Dairy Sci 2000;83:952–61. [22] Rittie L. Method for picosirius red-polarization detection of collage fibers in tissue sections. Methods Mol Biol 2017;1627:395–407. Chapter 26, Laure Rittie (ed.), Springer ScienceþBusiness Media LLC. [23] Vracko R. Basal lamina scaffold-anatomy and significance for maintenance of orderly tissue structure. Am J Pathol 1974;77:314–46. [24] Sud SC, Tucker HA, Meites J. Estrogen-progesterone requirements for udder development in ovariectomized heifers. J Dairy Sci 1968; 51:210–4. [25] Musters S, Coughlan K, McFadden T, Maple R, Mulvey T, Plaut K. Exogenous TGF-b1 promotes stromal development in heifer mammary gland. J Dairy Sci 2004;87:896–904. [26] De Vries LD, Casey T, Dover H, VandeHaar MJ, Plaut K. Effects of transforming growth factor-b on mammary remodeling during the dry period of dairy cows. J Dairy Sci 2011;94:6036–46. [27] Casey T, Dover H, Lieshman J, DeVries L, Kiupel M, VandeHaar M, Plaut K. Transcriptome analysis of epithelial and stromal contributions to mammogenesis in three week prepartum cows. PLoS One 2011;6:e22541. [28] Junqueira LCU, Bignolas G, Brentani RR. Picosirius staining plus polarization microscopy, a specific method for collage detection in tissue sections. Histochem J 1979;11:447–55.