Species-Specific Detection of Growth Factor Gene Expression in Developing Murine Prostatic Tissue¹

^a Departments of Urology ^b Anatomy, University of California, San Francisco, California 94143-0540

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
The aim of the present study was to develop a method by which the expression of paracrine signaling molecules could be localized to either epithelial or stromal cells of developing prostatic tissue. Heterospecific tissue recombinants composed of mouse urogenital epithelium (mouse UGE) plus rat urogenital mesenchyme (rat UGM) and the reciprocal tissue recombinants, rat urogenital epithelium (rat UGE) plus mouse urogenital mesenchyme (mouse UGM), were grafted under the renal capsule in intact, athymic male mouse and rat hosts. After 2 wk of growth, RNA from the grafts was analyzed by species-specific reverse transcription-polymerase chain reaction for the expression of the mRNA for the following molecules: transforming growth factors ß1, ß3, and ; epidermal growth factor; epidermal growth factor receptor; and keratinocyte growth factor. The species of expression of these growth factor and receptor gene products within the heterospecific tissue recombinants was identified, allowing determination of the cell layer in which the genes were expressed. Identification of the tissue-specific expression of the growth factor and growth factor receptor profiles of the epithelium and mesenchyme of this in vivo model provides a basis for understanding the autocrine and paracrine mediators of cell-cell interactions in prostatic development.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
The prostate gland develops from the urogenital sinus, which is composed of epithelial (UGE) and mesenchymal (UGM) components. Development of the prostate occurs through androgen-driven mesenchymal-epithelial interactions [1]. During prostatic development, UGM induces epithelial growth and differentiation, and in a reciprocal fashion UGM induces the differentiation of UGE [2, 3]. Tissue recombinants composed of wild-type UGM and androgen receptor-deficient epithelium from the testicular feminization mouse have established the concept that androgens elicit prostatic epithelial growth and development via paracrine mechanisms [1]. Soluble growth factors have been proposed to be the paracrine mediators of these cell-cell interactions during prostatic development [4–6].

Current methods to assess gene expression include Northern blot hybridization, in situ hybridization, RNase protection (RP), and reverse transcription-polymerase chain reaction (RT-PCR). Sensitivity is relatively low for Northern detection and higher for detection by RP. RT-PCR and in situ hybridization give the highest sensitivity. These methods also differ in that RT-PCR gives qualitative gene expression; in situ hybridization provides tissue localization of the signal; and RP and Northern blot analyses provide quantitative data, but usually without tissue localization. The technique presented herein utilizes heterospecific tissue recombinations and species-specific RT-PCR (SS-RT-PCR) to provide maximum sensitivity in assessing gene expression and tissue-specific expression of growth factor and growth factor receptor genes in chimeric organs.

In this study, a tissue recombination model was used to assess the tissue-specific expression of growth factor and growth factor receptor genes of epithelial-mesenchymal interactions during in vivo prostatic development. Methods were developed to detect and identify individual tissue-type growth factor and growth factor receptor expression by species origin. This study describes a SS-RT-PCR technique that discriminates between rat and mouse growth factor and growth factor receptor gene expression. The transcripts examined in this study were transforming growth factor (TGF)ß1, TGFß3, TGF, keratinocyte growth factor (KGF), epidermal growth factor (EGF), and EGF receptor (EGF-R).

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Tissue Recombinant Preparation and RNA Isolation

Rat and mouse urogenital sinuses were dissected from 18-day embryonic Sprague-Dawley rats (Simonsen Laboratories, Inc., Gilroy, CA) and 16-day embryonic Balb/c mice (Simonsen; plug day is designated Day 0 for both mice and rats). Urogenital sinuses were digested with 0.1% bacto-trypsin 1:250 (Difco Laboratories, Detroit, MI) in calcium-/magnesium-free Hanks' solution (UCSF Cell Culture Facility, San Francisco, CA) for 90 min at 0°C. The epithelium was manually separated from the mesenchyme as previously described [7]. The tissue recombinants constructed were rat urogenital mesenchyme plus mouse urogenital epithelium (rat UGM + mouse UGE) and mouse urogenital mesenchyme plus rat urogenital epithelium (mouse UGM + rat UGE). The tissue recombinants were placed on agar and cultured overnight at 37°C in 5% CO₂ as previously described [7]. The tissue recombinants were grafted under the renal capsule of both athymic Balb/c mouse hosts (Simonsen) and athymic rat hosts (Hsd:RH-rnu, Harlan-Sprague Dawley, Indianapolis, IN), as previously described [7]. The rat UGM + mouse UGE and rat UGE + mouse UGM tissue recombinants were grafted into both mouse and rat hosts on three separate occasions. Hosts were killed by barbiturate overdose and cervical dislocation 2 wk after grafting. This is a time point at which prostatic growth and development is still occurring within the tissue recombinants. The grafts were recovered and homogenized in 0.5 ml RNA STAT-60 (Tel-Test "B" Inc., Friendswood, TX) in a 1.5-ml microfuge tube. Total RNA was extracted after chloroform, isopropanol, and 75% ethanol treatment (according to Tel-Test "B" Bulletin No. 1). RNA was dissolved in diethyl pyrocarbonate (DEPC; Sigma Chemical Co., St. Louis, MO)-treated H₂O and quantitated by spectrophotometry. RNA samples had a 260/280 absorbance ratio greater than 1.6.

Sequence Analysis

Messenger RNA and/or cDNA sequences were retrieved from the Nentrez database (National Center for Biotechnology Information, National Library of Medicine, National Institutes of Health, Bethesda, MD). Primer sequences were designed and analyzed using Oligo 4.03 software (National Biosciences Inc., Plymouth, MN) on an Apple Macintosh IIcx (Apple Computer, Cupertino, CA), and the nucleic acid sequence of the amplified RT-PCR product was determined. Restriction sites of the RT-PCR products were identified using DNA Strider 1.0 analysis software (Commissariat a l'Energie Atomique-France, Gif-Sur-Yvette, Cedex, France). For some genes, one set of primers amplified both mouse and rat sequences and yielded an RT-PCR product of the same size for both the rat and mouse genes. Restriction analysis of the amplified products using DNA Strider 1.0 usually identified distinctive restriction enzyme sites that allowed the two RT-PCR products to be distinguished. For example, the mouse and rat RT-PCR products of a single primer set could be distinguished either by one enzyme that gave fragments of different sizes upon digestion, or by enzymes that digest one product but not the other. The enzymes utilized were tested in each single species to confirm that the expected digestion had occurred. If the nucleotide sequence amplified by a primer set yielded mouse and rat RT-PCR products that were too homologous, then the two PCR products might have identical restriction enzyme digestion patterns. In this case, primers were selected that amplified a less conserved region of the cDNA sequences. For gene sequences with low homology of the mouse and rat sequences, species-specific primers were utilized, as in the case of TGF, EGF, and EGF-R shown in Table 1.

Reverse Transcriptase (RT) Reaction

For a 50-µl reaction, 1 µg of total RNA was dissolved in DEPC-treated H₂O, heated to 75°C for 3 min, and immediately placed on ice. The following were then added to the RNA: 160 ng random hexamer primers (Life Technologies, Gaithersburg, MD); 0.4 mM dNTP mixture, consisting of equal concentrations of dATP, dCTP, dGTP, dTTP (Perkin Elmer, Foster City, CA); 40 units RNasin ribonuclease inhibitor (Promega Corp. Madison, WI); and 200 units SuperScript II reverse transcriptase (Life Technologies) in single-strength PCR reaction buffer of 10 mM Tris-HCl, 1.5 mM MgCl₂, 50 mM KCl, pH 8.3 (Perkin Elmer). Before addition of the RT enzyme, the reaction tubes containing all other components were heated to 37°C for 2 min, and the RT enzyme was then added. The RT reaction was carried out at 37°C for 1 h; this was followed by a 5-min incubation at 95°C. At least three RT reactions were performed for each RNA sample, and the validity of the RT reaction was determined by PCR amplifications using rat or mouse ß-actin primers (data not shown).

PCR Amplification

Five microliters of the RT reaction was amplified in either a 25-µl or a 50-µl reaction volume in thin-walled PCR reaction tubes in single-strength PCR reaction buffer of 10 mM Tris-HCl, 1.5 mM MgCl₂, 50 mM KCl, pH 8.3 (Perkin Elmer), using 2.5 units AmpliTaq polymerase (Perkin Elmer), 0.55 µg TaqStart antibody (Clontech, Palo Alto, CA), 0.2 mM sense and antisense primers, and 0.4 mM dNTP mixture consisting of equal concentrations of dATP, dCTP, dGTP, and dTTP (Perkin Elmer). The reaction was performed in either a PTC-150 Programmable Thermal Controller (MJ Research, Inc., Watertown, MA) or a GeneAmp PCR System 2400 (Perkin Elmer). Cycle parameters for reactions performed in the PTC-150 were generally a 45-sec melting step at 94°C, a 45-sec annealing step, and a 90-sec extension step at 72°C. Thirty-five cycles were used per amplification except as noted below. Cycles were preceded by a 3-min 95°C melting step and followed by an 8-min 72°C extension step after the final cycle. Cycle parameters for reactions performed in the GeneAmp 2400 were similar to those outlined above except that the melting, annealing, and extension steps were all 30 sec in length. Primer sequences, product sizes, and annealing temperatures are given in Table 1. The optimal annealing temperatures of the 35 amplification cycles are shown in Table 1. For most of the primer sets, a touchdown amplification method was used. The first four cycles of the touchdown method use annealing temperatures beginning 4° above the final temperature as listed in Table 1 and decreasing by 1° each cycle until the optimal annealing temperature is reached; this method preferentially amplifies the expected RT-PCR product. RT-PCR products were electrophoresed on 2% SeaKem LE agarose (FMC BioProducts, Rockland, ME) gels in single-strength Tris-acetate-EDTA (TAE) or single-strength Tris-borate-EDTA (TBE) running buffer, stained with 0.5 µg/ml ethidium bromide (Fisher Scientific, Pittsburgh, PA), and photographed under UV transillumination. For the rat KGF RT-PCRs, after 35 amplification cycles, very little or no RT-PCR product was seen. Secondary amplifications of 30 cycles were performed after addition of more primers and AmpliTaq polymerase; the results shown for rat KGF expression are derived from RT-PCR products seen after these secondary amplifications. RT-PCR primers for TGFß1, TGFß3, mouse and rat TGF, mouse and rat EGF, mouse and rat KGF, rat EGF-R, and mouse p25 (mp-25) were designed or obtained at our institution; the RT-PCR primers for mouse EGF-R have been previously described [8]; the M-40 RT-PCR primers were kindly provided by Dr. David Danielpour (NCI, Bethesda, MD).

PCR Product Verification

The PCR product for a given reaction was digested with a designated restriction enzyme to confirm its identity: the specific enzymes used for digestions are shown in Table 2. Typically, to 20 µl of the PCR product, 5 µl of reaction buffer (see below) was added, which optimized the buffer conditions for the given restriction enzyme. The final PCR buffer conditions were adjusted to the necessary conditions of the restriction enzyme utilized. This generally involved increasing the concentrations of MgCl₂, Tris, and NaCl, lowering the pH, and adding DTT (1,4-dithiothreitol; Boehringer-Mannheim, Indianapolis, IN). As an example of buffer adjustments, some of the digestions used restriction enzymes Msp I and HinfI, which work optimally in Promega buffer B (single-strength concentrations are 6 mM Tris-HCl, 6 mM MgCl₂, 50 mM NaCl, 1 mM DTT, at pH 7.5). As the standard single-strength RT-PCR reaction buffer is 10 mM Tris-HCl, 1.5 mM MgCl₂, 50 mM KCl, pH 8.3, the changes needed to optimize the PCR reaction buffer for the enzymes Msp I and HinfI include increasing the MgCl₂ concentration to 6 mM, decreasing the pH to 7.5, maintaining the salt concentration at 50 mM, and lowering the Tris-HCl concentration to 6 mM. A stock of 5-strength buffer B was prepared (30 mM Tris-HCl, 30 mM MgCl₂, 250 mM NaCl, 5 mM DTT), and 5 µl of this 5-strength stock was added to 20 µl of the PCR reaction to be digested; the final digestion conditions were approximately 14 mM Tris-HCl, 7.2 mM MgCl₂, 90 mM NaCl, and 0.7 mM DTT. Lambda DNA was digested under the adjusted buffer conditions to determine enzyme functionality in these buffer adjustments; most restriction enzymes are flexible in their ability to digest under suboptimal conditions, and long digestion times and higher amounts of enzyme (> 1 unit/µg DNA) help ensure proper digestion in these conditions. Digestions were carried out for at least 2 h (often overnight) at 37°C, using between 1 and 10 units of enzyme. The digestion reactions were electrophoresed with undigested RT-PCR product controls and enzyme controls on 4% NuSieve 3:1 agarose (FMC BioProducts) gels in single-strength TBE running buffer. Gels were stained with 0.5 µg/ml ethidium bromide and photographed using UV transillumination. Restriction enzymes were purchased from Life Technologies, Promega, Boehringer-Mannheim, and New England Biolabs (Beverly, MA). The rat UGM + mouse UGE and rat UGE + mouse UGM tissue recombinants were grafted into both mouse and rat hosts on three separate occasions. Because of the size of the grafts, the RNA yields were not great enough to allow analysis of all growth factors and receptors for every graft harvested. The gene expression data represent RT-PCRs and digestions performed in duplicate using RNA from separately grafted specimens. In total this study was based upon analysis of 36 tissue recombinants.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
The heterospecific tissue recombinants utilized in this experiment grew and formed prostatic glandular structures as previously described [9]. For each transcript examined, it was possible to distinguish signal for the mouse gene from that of the rat gene using the primers indicated in Table 1 and the restriction enzymes indicated in Table 2.

In order to determine the possibility of contamination of the tissue recombinants, RT-PCRs were performed to detect gene expression of prostatic epithelial secretory proteins. The rationale for this method of monitoring contamination is that the nature of tissue separations means that epithelium is almost always completely devoid of mesenchymal contamination, and furthermore that any contamination is obvious with visual inspection. By contrast, traces of epithelial contamination of mesenchymal preparations can be extremely difficult to detect. The inductive nature of mesenchyme in these tissue recombinants means that a few contaminating epithelial cells can proliferate into a major contaminant of the harvested graft. Thus, monitoring of possible epithelial contamination of mesenchyme on a graft-by-graft basis was considered extremely important. Probasin is a well-characterized abundant secretory protein expressed exclusively by rat prostatic epithelium and representing a high percentage of total mRNA output of these cells. Therefore, if probasin gene expression was detected in the rat UGM + mouse UGE tissue recombinants, the only source of that signal would be contaminating rat epithelial cells. Likewise, mp-25 is a mouse-specific epithelial prostatic secretory protein. Its presence in the mouse UGM + rat UGE tissue recombinants indicates the presence of mouse epithelial cells contaminating the mouse UGM. Transcripts of the secretory protein genes (both probasin and mp-25) should be considerably more abundant than those of any of the growth factor or receptor genes. Thus, these transcripts should be very reliable as markers of epithelial contamination of UGM. The RT-PCR primers for these secretory proteins are included in Table 2. Expression of probasin gene mRNA was examined in the rat UGM + mouse UGE tissue recombinants. The gene expression of mp-25 was examined in the mouse UGM + rat UGE tissue recombinants. The results of these experiments indicated that cellular contamination occurred during the preparation of a minor subset of the tissue recombinants. This screening method thus allows screening of samples for cross-species contamination and subsequent elimination of contaminated samples.

The growth factor expression seen in the tissue recombinants grafted into mouse hosts is summarized in Table 3, and that found in rat hosts is summarized in Table 4. Tables 3 and 4 represent pooled data from multiple repeats of these experiments in which epithelial contamination of the UGM used to prepare the graft had been excluded. Figures 1–4 illustrate examples of specific RT-PCR amplifications and species-specific digestions of the amplified products to confirm identity and species of origin. Figures 1 and 2 show expression of TGFß1 and -ß3, respectively, representing a situation in which one primer set was used to amplify sequences from both species, with restriction enzyme analysis being used to differentiate the signal by species. In contrast, Figures 3 and 4 show the amplification of TGF signal where separate species-specific primers were used for the two species; digestion was used to confirm the nature of the amplified fragment. The use of both rat and mouse hosts was necessary to distinguish between graft and host gene expression. For example, when rat UGM + mouse UGE tissue recombinants were grafted into mouse hosts, transcripts from host tissues (e.g., vasculature and adjacent kidney) would not be distinguished from transcripts of the mouse UGE. Fortunately, signal from the mouse host cannot interfere with signal from the rat UGM. Likewise, in mouse UGM + rat UGE tissue recombinants grafted into rat hosts, the signal from the rat UGE would be affected by contaminating signal from host tissues, but the signal from mouse UGM would not be affected by host tissue signals: potential confounding effects of host cells can only affect signal associated with one of the tissues. Thus, by growing each tissue recombinant in both rat and mouse hosts, it was possible to accurately assess the gene expression of both epithelial and stromal cells in the graft. RT-PCRs were also performed for all growth factors and receptors using RNA from the mouse and rat host kidneys, and expression for all factors analyzed was seen in the kidney tissue of both hosts (data not shown). The rat expression columns in Table 3 and the mouse expression columns in Table 4 comprise expression signals that are not affected by contaminating host signal, thus giving the tissue source of gene expression.

View larger version (59K): [in this window] [in a new window]   FIG. 1. Rsa I (Rs) restriction enzyme digestion of the 655-base pair (bp) TGFß1 RT-PCR product. a) The undigested (Un) TGFß1 RT-PCR product and its digestion from the tissue recombinant specimens grown in the mouse host: the 350-bp band is indicative of mouse TGFß1 expression, and the 284-bp and 269-bp bands are indicative of rat TGFß1 expression. b) The TGFß1 RT-PCR product and its digestion from the tissue recombinant specimens grown in the rat host.

View larger version (32K): [in this window] [in a new window]   FIG. 2. TGFß3 RT-PCR product amplification and digestion. Restriction enzyme Msp I (Ms) digests the undigested (Un) 288-bp RT-PCR product of the mouse TGFß3 gene, giving digestion bands of 231 bp and 59 bp. Restriction enzyme Bsp1286I (Bs) digests the rat TGFß3 RT-PCR product, showing digestion bands of 252 bp and 38 bp. The 59-bp and 38-bp bands are not shown. a) Amplification and digestion of the specimens grown in the mouse host. b) Amplification and digestion of the specimens grown in the rat host. In b, the Bsp1286I digestion of the mouse UGM + rat UGE sample gives bands of 252 bp and 175 bp; the 252-bp band is indicative of rat gene expression, and the 175-bp band is indicative of mouse gene expression (the 175-bp band is also present in a in the Bsp1286I digestion of the rat UGM + mouse UGE specimen).

View larger version (30K): [in this window] [in a new window]   FIG. 3. Rat TGF gene expression as determined by HinfI (Hi) digestion of the undigested (Un) 428-bp RT-PCR product giving a 334-bp fragment. a) Products and digestions of recombinants grown in a mouse host; b) products and digestions of recombinants grown in a rat host.

View larger version (33K): [in this window] [in a new window]   FIG. 4. Mouse TGF undigested (Un) 262-bp RT-PCR product and digestion by Msp I (Ms), giving fragments of 142 bp and 120 bp. a) Expression in recombinants grown in a mouse host; b) expression in recombinants grown in a rat host.

Focusing on epithelial gene expression, the following specimens are informative: the rat epithelium of mouse UGM + rat UGE tissue recombinants grown in mouse hosts (Table 3) and the mouse epithelium of rat UGM + mouse UGE tissue recombinants grown in rat hosts (Table 4). Note that in both tissue recombinants the prostatic epithelium expresses TGFß3, TGF, EGF, and EGF-R, but not TGFß1. For stromal gene expression, instructive specimens are the rat mesenchyme of the rat UGM + mouse UGE recombinants grown in mouse hosts (Table 3) and the mouse mesenchyme of the mouse UGM + rat UGE grown in rat hosts (Table 4). In both of these tissue recombinants, stromal expression was consistently seen for KGF, TGFß3, TGF, EGF, and EGF-R. Transcripts for TGFß1 were detected in the rat mesenchyme of the rat UGM + mouse UGE tissue recombinants grown in mouse hosts but not in the mouse mesenchyme of the mouse UGM + rat UGE grown in rat hosts. The results are representative of gene expression in tissue recombinants of at least three different experiments, and all RT-PCR amplifications and digestions were performed in duplicate for a given tissue recombinant in a specific host using separate RNA samples.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Information on the expression of growth factors and receptors can be obtained by many methods. Immunocytochemistry and in situ hybridization are excellent methods for determining tissue localization of a particular gene product: in situ hybridization provides definitive information on the site of RNA synthesis while immunocytochemistry provides information on the localization of the protein itself. For both immunocytochemistry and in situ hybridization, sensitivity is variable, and the inherent background of a particular probe or can antibody affect the results. Another method of assessing the growth factor and growth factor receptor profiles of epithelial cells and mesenchymal/stromal cells is to isolate these cell types and perform Northern blot, RP, or RT-PCR analysis on purified cell fractions. Unfortunately, for many organs or organ rudiments, the complexity of epithelial architecture is such that it is not possible to directly make pure epithelial, mesenchymal, or stromal cell preparations. Invariably, a persistent problem is the cross-contamination of one cell type with the other. Clearly, the results from such epithelial and stromal fractions are only as good as the purity of the preparations.

SS-RT-PCR is an alternate method that overcomes some of the obstacles described. The technique is applicable for analysis of heterospecific tissue recombinants, whereas in situ hybridization, immunohistochemistry, and regular RT-PCR have broader applications. A principal advantage of utilizing SS-RT-PCR is that the heterospecific tissue recombinants are processed directly from the in vivo setting into RNA so that gene expression is truly representative of the in vivo condition. RT-PCR is presently the most sensitive widely available method with which to detect gene expression. One consideration of the SS-RT-PCR method is the high degree of technical precision required in the preparation of the tissue recombinants, because cross-contamination of the mesenchyme or epithelium during the construction of the tissue recombinants will give erroneous results. In this regard, it is noteworthy that contamination of trypsin-isolated epithelium by mesenchyme is an extremely rare occurrence because any residual mesenchymal cells attached to the epithelium are readily observed using a dissecting microscope and can be removed; conversely, epithelial cells attached to mesenchyme are more difficult to observe. The fact that the growth factor and growth factor receptor profiles of both the mouse and the rat epithelium in mouse UGM + rat UGE and rat UGM + mouse UGE tissue recombinants were identical is indicative of the purity of the epithelial preparations and the consistency of the results. Performing the separation of epithelial and mesenchymal tissues on embryonic organ rudiments, when epithelial architecture is simple (before ductal branching morphogenesis), increases the likelihood of obtaining pure epithelial and mesenchymal preparations. Finally, it should be noted that through use of RT-PCR reactions to abundant species-specific secretory proteins such as the rat C3 gene, DP-1, or probasin [10] in rat UGM + mouse UGE tissue recombinants, it is possible to monitor the possible carryover of contaminating residual rat UGE cells in the rat UGM and thus exclude from analysis those tissue recombinants that are contaminated. Likewise, the mp-25 RT-PCR primers can be used to detect mouse epithelial cells that may contaminate mouse UGM + rat UGE tissue recombinants. As discussed above, epithelial contamination of mesenchyme is far more likely to occur than mesenchymal contamination of epithelium. This is borne out by the detection of KGF expression (which is known to be restricted to the mesenchymal cells [11]). KGF was not detected in epithelium but, as expected, was found exclusively in the mesenchymal compartments of grafts. This observation confirms both the expression of KGF by mesenchymal cells during prostatic morphogenesis and the absence of significant mesenchymal contamination of the epithelium used to make the tissue recombinants.

The data presented here show expression of TGF, EGF, and EGF-R in both epithelium and mesenchyme of prostatic tissue recombinants regardless of their species. From their immunohistochemical study of TGF and EGF-R in human prostatic tissue, Cohen et al. [12] suggest a paracrine mechanism during normal prostatic development involving TGF production in the stroma that signals through the EGF-R in the epithelium. Through use of immunohistochemistry in rat prostatic cells, an autocrine loop has been described, as TGF is detected in epithelial cells that also express the EGF-R [13]. The existence of such autocrine loops has been suggested to be a feature characteristic of unregulated tumorigenic growth. Using SS-RT-PCR, EGF ligands and the EGF-R are present simultaneously in both the epithelium and mesenchyme of developing chimeric prostatic tissue recombinants. Epithelial-produced EGF has been described previously [14], and evidence suggests that it is secreted apically into ductal lumina as a component of prostatic secretion. EGF (and any other growth factor) secreted apically may be biologically irrelevant to prostatic growth because of an inability to gain access to EGF-R located on the epithelial basal lateral surface and on stromal cells. Thus, in the normal prostate, both autocrine and paracrine pathways may exist both within and between these tissues.

KGF is a classical example of a paracrine-acting growth factor. The ligand is normally exclusively expressed in stromal cells, while its receptor, the -IIIb splice variant of the FGF-R2 gene, is found exclusively in epithelial cells of the growing prostate [11]. The inclusion of KGF in this study acted to confirm a lack of stromal contamination of epithelial samples used to make tissue recombinants.

The TGFß family are commonly associated with inhibiting cell proliferation and have been linked to apoptotic cell death in the prostate after castration [15]. TGFß mRNA has been detected in the growing prostate of the mouse. Timme et al. [16] describe high levels of expression for TGFß1, -ß2, and -ß3 mRNA in the UGM as compared to the UGE, which is consistent with the findings reported here.

In conclusion, the present study presents a novel and informative method of detecting gene expression in heterospecific tissue recombinants. Determining the tissue-specific gene expression of epithelial and mesenchymal organ components is essential in developing an understanding of the autocrine and paracrine mediators of cell-cell interactions in prostatic development, in the growth-quiescence adult prostate, and potentially in prostatic carcinogenesis. The SS-RT-PCR techniques could be used to investigate stromal-epithelial interactions in any of the developing and adult organ systems that are amenable to investigation by tissue recombination techniques; historically these have included the urogenital and gastrointestinal tracts, as well as branching and secretory structures such as the lung, salivary gland, and mammary gland.

	FOOTNOTES

 
¹ Supported by NIH grants: CA 59831, DK 45861, CA 64872, DK 47517, and DK52721.

² Correspondence: Simon W. Hayward, UCSF Mt Zion Cancer Center, 2340 Sutter Street, Box 0540, San Francisco, CA 94115. FAX: (415) 502-2270; simonh{at}itsa.ucsf.edu

Accepted: February 25, 1998.

Received: December 29, 1997.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

1. Cunha GR, Donjacour AA, Cooke PS, Mee S, Bigsby RM, Higgins SJ, Sugimura Y. The endocrinology and developmental biology of the prostate. Endocr Rev 1987; 8:338–363.[Medline]
2. Cunha GR, Fujii H, Neubauer BL, Shannon JM, Reese BA. Stromal influence on expression of morphological and functional characteristics or urogenital epithelia. In: Kogan S, Hafez ESE (eds.), Clinics in Andrology, Vol. 7. Boston: Martinus Nijhoff Pub.; 1981: 21–36.
3. Cunha GR, Hayward SW, Dahiya R, Foster BA. Smooth muscle-epithelial interactions in normal and neoplastic prostatic development. Acta Anat 1996; 155:63–72.[Medline]
4. Cunha GR, Battle E, Young P, Brody J, Donjacour A, Hayashi N, Kinbara H. Role of epithelial-mesenchymal interactions in the differentiation and spatial organization of visceral smooth muscle. Epithelial Cell Biol 1992; 1:76–83.[Medline]
5. Cunha GR, Foster B, Thomson A, Sugimura Y, Tanji N, Tsuji M, Terada N, Finch PW, Donjacour AA. Growth factors as mediators of androgen action during the development of the male urogenital tract. World J Urol 1995; 13:264–276.[Medline]
6. Culig Z, Hobisch A, Cronauer MV, Radmayr C, Hittmair A, Zhang J, Thurnher M, Bartsch G, Klocker H. Regulation of prostatic growth and function by peptide growth factors. Prostate 1996; 28:392–405.[CrossRef][Medline]
7. Cunha GR, Donjacour AA. Mesenchymal-epithelial interactions: technical considerations. In: Coffey DS, Bruchovsky N, Gardner WA, Resnick MI, Karr JP (eds.), Assessment of Current Concepts and Approaches to the Study of Prostate Cancer. New York: A.R. Liss; 1987: 273–282.
8. Paria BC, Das SK, Andrews GK, Dey SK. Expression of the epidermal growth factor receptor gene is regulated in mouse blastocysts during delayed implantation. Proc Natl Acad Sci USA 1993; 90:55–59.[Abstract/Free Full Text]
9. Cunha GR, Sekkingstad M, Meloy BA. Heterospecific induction of prostatic development in tissue recombinants prepared with mouse, rat, rabbit, and human tissues. Differentiation 1983; 24:174–180.[CrossRef][Medline]
10. Hayashi N, Cunha GR, Parker M. Permissive and instructive induction of adult rodent prostatic epithelium by heterotypic urogenital sinus mesenchyme. Epithelial Cell Biol 1993; 2:66–78.[Medline]
11. Finch PW, Cunha GR, Rubin JS, Wong J, Ron D. Pattern of KGF and KGFR expression during mouse fetal development suggests a role in mediating morphogenetic mesenchymal-epithelial interactions. Dev Dynam 1995; 203:223–240.[Medline]
12. Cohen D, Simak R, Fair W, Melamed J, Scher H, Cordon-Cardo C. Expression of transforming growth factor-alpha and the epidermal growth factor receptor in human prostate tissues. J Urol 1994; 152:2120–2124.[Medline]
13. Taylor TB, Ramsdell JS. Transforming growth factor-alpha and its receptor are expressed in the epithelium of the rat prostate gland. Endocrinology 1993; 133:1306–1311.[Abstract]
14. Wu HH, Kawamata H, Kawai K, Lee C, Oyasu R. Immunohistochemical localization of epidermal growth factor and transforming growth factor alpha in the male rat accessory sex organs. J Urol 1993; 150:990–993.[Medline]
15. Kyprianou N, Isaacs JT. Expression of transforming growth factor-beta in the rat ventral prostate during castration-induced programmed cell death. Mol Endocrinol 1989; 3:1515–1522.[Abstract]
16. Timme TL, Truong LD, Merz VW, Krebs T, Kadmon D, Flanders KC, Park SH, Thompson TC. Mesenchymal-epithelial interactions and transforming growth factor-beta expression during mouse prostate morphogenesis. Endocrinology 1994; 134:1039–1045.[Abstract]
17. Mullhaupt B, Feren A, Fodor E, Jones A. Liver expression of epidermal growth factor RNA. Rapid increases in immediate-early phase of liver regeneration. J Biol Chem 1994; 269:19667-19670.[Abstract/Free Full Text]