Fibroblast Growth Factor 8 Isoform b Overexpression in Prostate Epithelium: A New Mouse Model for Prostatic Intraepithelial Neoplasia¹ Free

PIN⁵ is generally considered as a preneoplastic lesion in humans (1, 2, 3). To understand mechanisms involved in the genesis and progression of prostatic preneoplastic lesions, the availability of suitable animal models for the disease process is critical. Although several mouse models for the study of prostate tumorigenesis have been described recently (4, 5, 6, 7, 8), there are, as yet, no good mouse models of prostate adenocarcinoma. As such, attempts to generate other animal models targeting specific genes for growth factors, oncoproteins, or tumor suppressors known to be naturally involved in this malignancy should have a high priority. Perhaps, through a logical combination of these animals, the resultant hybrid animals may more closely resemble either early or late, if not the complete histopathological and clinicophysiologic characteristics of the human prostate cancer.

Dysregulation of several growth factors has been implicated in prostate tumorigenesis (9, 10). One such factor, FGF8, was reported to be associated with the development or progression of this malignancy (11, 12, 13, 14, 15, 16). The conventional knockout of the fgf8 gene in the mouse led to early embryonic lethality (17) because FGF8 is a crucial signaling molecule for outgrowth and patterning, such as the elongating body axis, midbrain/hindbrain junction, limb, and face (18, 19, 20, 21, 22, 23, 24). Alternative splicing of the first exon of the fgf8 gene in mouse gives rise to eight potential protein isoforms that vary in their amino termini (20, 25, 26, 27). In humans, however, only four protein isoforms (FGF8a, 8b, 8e, and 8f) are predicted because of a blocked reading frame in the exon 1B of the human gene (28, 29). Of the four possible isoforms, FGF8b has been demonstrated to possess the most transforming and tumorigenic potential (11, 26, 30, 31), it appears to be the primary species in prostate epithelial cell lines or malignant epithelium (11, 13, 15), and its expression is practically undetected in the stromal component of prostate cancer (13, 15, 32). The overexpression of FGF8b in LNCaP cells increases their growth rate, soft agar clonogenicity, and in vitro and in vivo invasion ability (16). Furthermore, it is demonstrated that the growth of stromal cells can be strongly up-regulated when cocultured with FGF8b-producing LNCaP cells (16). Down-regulation FGF8b mRNA by antisense RNA expression reduces the growth rate, inhibits the soft agar clonogenic activity, and decreases in vivo tumorigenicity of prostate tumor cells (14). Among the FGF receptors, the “c” splice isoform of FGFR2 or FGFR3, as well as FGFR4, is reported to be most efficiently activated by FGF8b (33, 34). Although these receptor isoforms are considered to be largely expressed in mesenchymal cells (35), there is evidence of aberrant expression of FGFR isoforms in prostate cancer cells (9, 36). It is also interesting that prostate appears to exhibit wide expression of the general classes of FGFRs. For example, a moderate level of expression of FGFR1 and FGFR2 is found in prostate epithelium and the microvasculature, whereas stromal smooth muscle cells exhibit a weak level of expression of FGFR3 (37). In another study (35) with primary cultures of human prostatic epithelial and stromal cells, FGFR3 was found to be the primary product in epithelial cells with a smaller amount of FGFR2, whereas stromal cells express primarily FGFR3 and a smaller amount of FGFR1 and FGFR2. More recently, increased expression of FGF8 isoforms and their receptors were reported in PIN lesions (32). Together, these observations indicate that FGF8-FGFR signaling plays an important role in prostate biology and cancer.

Thus, to study the in vivo effects of FGF8b, it was important to establish transgenic mouse lines by targeting FGF8b overexpression in the prostate epithelium. For this purpose, the small rat composite 468-bp probasin promoter ARR₂PB was chosen. This promoter has been demonstrated to confer a high level of reporter transgene expression specifically in the luminal prostatic epithelium and is strongly regulated by androgens (38, 39). Four independent transgenic lines, in which the FGF8b transgene expression is driven by this promoter, were generated. The specificity of expression of the transgene was confirmed by RT-PCR and in situ hybridization, and the transgenic animals heterozygous for the transgene were followed for study of the histopathology for up to 24 months of age.

The XhoI-XbaI fragment of ARR₂PB (38) was blunted and ligated into pSV plasmid vector containing the SV40 poly(A) sequence and splicing signal sequence. Subsequently the full-length human FGF8b cDNA fragment (11), which harbored 100% amino acid sequence identity to the mouse FGF8b sequence (11), was inserted into the EcoRI site after the ARR₂PB promoter (Fig. 1 A). The sequence of the FGF8b coding region and junction of each fragment were confirmed by automated DNA sequencing. The fragment containing the ARR₂PB promoter, FGF8b cDNA, and SV40 poly(A) sequence was released by digestion with NotI and KpnI, isolated by agarose gel, and purified by Qiagen spin column (Qiagen, Valencia, CA) and Elutip column according to the manufacturer’s protocol.

Two rounds of pronuclear injection of ARR₂PB promoter-FGF8b-SV40 poly(A) fragment were performed. The (C57BL/6 × DBA2)F₁ hybrid fertilized eggs containing the transgene construct were placed into pseudo-pregnant females. Potential founder animals were screened by PCR and confirmed by Southern blot analysis using clipped tissue DNA samples. Four productive transgenic lines were established by mating the founder animals with nontransgenic (C57BL/6 × DBA2)F₁ mice. Offspring were genotyped by PCR from tail DNA at 3–4 weeks of age.

An aliquot of 150–200 μl of 10 mg/ml BrdUrd (Sigma Chemical Co., St. Louis, MO) was injected i.p. 1 h before animals were sacrificed. The urogenital system was removed, and the individual prostate lobes were dissected under a dissecting microscope. Tissues for histopathological observation were fixed overnight in 10% neutral buffered formalin (Surgipath, Richmond, IL). Fixed tissues were processed and embedded in paraffin. Thin sections (5 μm) were cut and stained with H&E. Tissues for mRNA assays were frozen in liquid nitrogen at the time of dissection.

The tissue specimens were digested with 20 mg/ml proteinase K (Life Technologies, Inc., Buffalo, NY) in 500 μl of a buffer containing 50 mm Tris-HCl (pH 8.0), 100 mm EDTA (pH 8.0), 100 mm NaCl, and 1% SDS at 50°C for overnight. After centrifugation, the supernatant containing the genomic DNA was collected. After boiling for 5 min, 2 μl of the supernatant were used as the template in 30 μl of reaction mixture containing 0.2 mm dNTP, 1.0 mm MgCl₂, 0.02% (w/v) DMSO, 6 pmol of each primer, and 0.3 unit of Tag polymerase (Life Technologies, Inc.). The sequences of the primers F8b-3 and SV40-a used for PCR and RT-PCR were 5′-AACTACACAGCGCTGCAGAATG-3′, which is complementary to the FGF8b cDNA sequence (11), and 5′-GTTGAGAGTCAGCAGTAGCCTC-3′, which is complementary to the SV40 poly(A) signal sequence (Fig. 1 A). The PCR was started at 94°C for 4 min, followed by 35 cycles at 94°C for 1 min, 58°C for 90 s, and 72°C for 90 s, and ended with 72°C for 5 min. Founder animals were further confirmed by Southern blot analysis. Briefly, 10 μg of tail DNA were digested by BamHI, run on a 1.5% agarose gel, and transferred to Nylon membrane. A ³²P-labeled SV40 signal sequence was used to probe the Southern blot.

The tissue RNA was extracted using RNeasy Mini kit (Qiagen, Germany). The ThermoScript RT-PCR System (Life Technologies, Inc.) was used for RT-PCR assay. A solution of 1 μg of RNA was mixed with 1 μl of random hexamer primers provided in 10 μl volume and denatured at 65°C for 5 min. After cooling on the ice, 10 μl of cDNA synthesis mixture were added. The samples were incubated at 25°C for 10 min, followed by 60 min at 50°C, and terminated at 85°C for 5 min. Aliquots of 2–4 μl of cDNA synthesis reaction mixtures were used as templates for PCR as described above.

The transgene construct was amplified by PCR using primers of F8b-T7 and F8b-T3 or primers of SV40-T7 and SV40-T3 (Fig. 1 A). The sequence of F8b-T7 was 5′-GCGCTAATACGACTCACTATAGGGTAAGCTTGCTGCCATGGGCAGC-3′, which contained the T7 promoter sequence and a segment that was complementary to FGF8b sequence at the 5′ end (11). The sequence of F8b-T3 was 5′-GCGCAATTAACCCTCACTAAAGGGGCTTGATATCGAATTCAGGATG-3′, which contained the T3 promoter sequence and segment that was complementary to FGF8b sequence at the 3′ end (11). The sequence of SV40-T7 was 5′-GTAATACGACTCACTATAGGGCGCAGTGGTGGAATGCCTTTAATG-3′, which contained the T7 promoter sequence and a segment that was complementary to SV40 poly(A) signal at 5′ end. The sequence of SV40-T3 is 5′-GCGCAATTAACCCTCACTAAAGGGACCTCTACAAATGTGGTATGGCT-3′, which contained the T3 promoter sequence and segment that was complementary to SV40 poly(A) signal at 3′ end. PCR products were transcribed in vitro into probes that were labeled with digoxigenin using a digoxigenin labeling kit (Boehringer Mannheim, Mannheim, Germany). In Situ hybridization assay was performed as described by Nieto et al. (40) with some modifications (41). In short, frozen tissues (5 μm) were hybridized overnight at 60°C in hybridization mixture with a probe diluted to 1:100. Sections were incubated in the blocking solution containing alkaline phosphatase labeled sheep anti-digoxigenin Fab fragments (1:1500; Boehringer Mannheim) overnight at room temperature. Signals were detected with a detection solution containing 4-nitro blue tetrazolium chloride and 5-bromo-4-chloro-3-indolyl-1-phosphate (Boehringer Mannheim).

Consistent with previously described criteria (8, 42), a similar grading system for PIN-like lesions was used to evaluate PB-FGF8b animals. Generally, PIN lesions were categorized into LGPIN and HGPIN based on their degree of cytological atypia. LGPINs showed mild cytological atypia. HGPINs were distinguished from LGPINs by more epithelial cell proliferation, nuclear stratification, and cytological atypia. Other lesions, such as papillary hyperplasia, that cannot be characterized by this grading system are described separately, based on their architecture and cytology.

BrdUrd staining was performed using the Zymed BrdUrd staining kit (Zymed, South San Francisco, CA). Sections were incubated with a biotinylated monoclonal anti-BrdUrd antibody. Signals were generated by the Streptavidin-peroxidase with diaminobenzidine as a chromogen. The slides were counterstained with methylene green (KPL Labs, Gaithersburg, MD). The staining for androgen receptor was performed in a similar manner. After antigen retrieval by microwave heating in 1 m urea, specimens were incubated with a rabbit anti-androgen receptor polyclonal antibody (Santa Cruz Technology, Santa Cruz, CA) at 0.8 μg/ml concentration. The bound antibody was detected with the biotinylated goat antirabbit immunoglobulin. Sections incubated without primary antibody served as negative controls. For SMA staining, a mouse monoclonal antibody against α-SMA (Sigma Chemical Co.) was used at 1.1 μg/ml concentration with the DAKO ARK kit (DAKO, Carpinteria, CA) to eliminate mouse background staining. A rabbit polyclonal antihuman CD3 antibody (DAKO) was used at 0.8 μg/ml concentration for T-cell staining. For B-cell staining, a rat antimouse CD45R/B220 monoclonal antibody (PharMingen, San Diego, CA) was used at 0.6 μg/ml concentration.

Although an association between FGF8 overexpression and human prostate cancer was described previously, little was known about this association in mice. Thus, it was important to investigate the FGF8 expression status in prostatic lesions from mice. For this purpose, LPB-Tag sublines 12T-7f and 12T-7s were used (5). PIN lesions of these mice were analyzed for FGF8 expression by in situ hybridization assay in which FGF8-specific riboprobes were used. Signals were readily detected by a F8b-T3 antisense probe in the epithelial lesions of these lines. This is illustrated in Fig. 2,A for the 12T-7s line. Weak signals were also detected in prostate tissues of wild-type control animals only after prolonged exposure (data not shown). Sections incubated with a F8b-T7 sense probe served as an internal negative control as shown in Fig. 2 B. The major FGF8 RNA species expressed in the tumor sections was identified to be FGF8b following the same RT-PCR protocol that we reported before (11). A similar RT-PCR analysis of a tumor cell line, TRAMP-C, derived from the TRAMP model (43), also showed a single FGF8b species as the primary product (data not shown).

The founder animals were identified by the presence of a 550-bp PCR product amplified from tail DNA by the primers that spanned the FGF8b cDNA gene and the SV40 poly(A) signal sequence of the transgene (Fig. 1,B). These founder animals were further confirmed by the presence of a 1.5-kb band on a Southern blot of BamHI-digested genomic DNA probed with the SV40 poly(A) signal sequence probe (Fig. 1 C). Four founder animals were identified from the first pronuclear injection and marked as 1, 3, 5, and A. Another series of five founder mice, named 6, 8, 18, 20, and 24, was obtained from the second injection. Founder A, a male animal, was determined to be infertile. Founders 5, 6, and 8 did not transmit the transgene to the offspring, as determined by PCR screening of tail DNA, after testing of four litters of total 25 offspring. Animal 24 contained the transgene but failed to express it in the prostate tissues as determined by RT-PCR analysis using three different age groups of the offspring. Eventually, four productive lines, 1, 3, 18, and 20, that contained the transgene and expressed the mRNA in prostate tissues were established.

Because the primers for the detection of transgene expression were designed to flank an intron in SV40 poly(A) signal sequence, an additional band of 485-bp long was expected to be amplified if the transgene was transcribed and spliced (Fig. 1,D). For all productive lines, the expression of the spliced version was detected in all prostate lobes, i.e., AP, DLP, and VP. A detectable level of expression was also seen in the DD, SV, and epididymis. To determine whether there was spurious expression of the transgene, various other tissues (testes, thymus, liver, kidney, lung, spleen, and heart) from animals of different lines were examined for transgene expression by RT-PCR. There was no detectable FGF8b mRNA in these tissues. When prostates of animals of different ages (4–24 months) were assayed by RT-PCR, it was found that the transgene continued to be expressed throughout the time period of this investigation (data not shown). The transgene expression was localized by in situ hybridization assay using a SV40 poly(A)-specific riboprobe. In prostate tissue sections, transgene mRNA was readily detected by the SV40-T3 antisense probe, whereas the SV40-T7 sense probe served as an internal negative control. As illustrated in Fig. 2,C, the signals for the transgene transcripts were strikingly confined to the epithelial cells of the prostate. No positive signals were found on the prostate tissue sections of the nontransgenic littermate control animals under identical experimental conditions (Fig. 2 D).

Throughout the experimental period, no significant difference in gross body weight between the transgenic animals and wild-type siblings was found. Among the four productive lines, line 3 was studied most extensively (Table 1). In this line, after 2 to 3 months, the LP and VP of the transgenic animals were consistently larger in size, varying between 2 and 15 times those of the age-matched control animals (Fig. 3). Histopathologically, hyperplasia was detected in the LP and VP of some animals at time points as early as 2–3 months of age, which was consistent with the results of the gross examination. The lesions were usually multifocal in that epithelial hyperplasia was identified in many ducts of one lobe, as illustrated for VP of one transgenic in Fig. 4,A. Fig. 4,B represents the VP of a littermate nontransgenic animal. Generally, multifocal hyperplasia developed later in the DP and AP, with the earliest starting at 6 and 16 months, respectively. Stromal proliferation surrounding the epithelial hyperplasia (Fig. 4,A) became prominent in older transgenic animals, generally beginning at around 17 months of age, although in two cases, stromal hypercellularity in VP was noted as early as 6–7 months of age (Table 1).

Between 5 and 7 months, LGPIN was readily detected in the DP, LP, and VP. These lesions were organized in flattened, papillary, cribriform, or tufting patterns. The atypical cells were generally larger than adjacent hyperplastic cells. Although the nuclei were larger and hyperchromatic, they were mildly pleomorphic (Fig. 4, C and D). Abundant eosinophilic cytoplasm was also readily seen. Occasionally, mitotic figures could be identified (Fig. 4,C). Beginning at 15–17 months, HGPINs appeared in the DP, LP, and VP. In these relatively advanced lesions, atypical cells filled or almost filled the lumina (Fig. 5, A–C). Cellular atypia in HGPIN was much more pronounced as compared with LGPIN. This was characterized by an increased nuclear:cytoplasmic ratio, marked nuclear atypia, hyperchromasia, and prominent nucleoli. Mitotic figures were more common in HGPIN than LGPIN (Fig. 5, A and D). As shown in Fig. 5,C, a bulging duct filled with atypical cells arranged in cribriform pattern stood out from the adjacent abnormal glands. Apoptotic bodies and an inflammatory reaction were noted (Fig. 5,D). Although the epithelium was still surrounded by a thin layer of laminin focally, a part of fibromuscular sheath appeared disrupted as revealed by anti-SMA immunostaining (Fig. 5,E). Androgen receptor expression appeared to be intact in epithelial cells of HGPIN (Fig. 5 F).

Papillary hyperplasia, a frequent phenotype in PB-FGF8b animals, was found around 16 months of age. It was manifested as multiple nodules bulging into and filling the lumina with a stalk connected to the basement membrane (Fig. 6,A). Although VP, LP, and DP of some but not all animals exhibited papillary hyperplasia, these lesions were not observed in AP. Microscopically, the individual lesion presented as an exophytic papillary proliferation of epithelial cells overlying a solid fibromuscular core in which some blood vessels were readily visible (Fig. 6, A and B). Atypical epithelial cells with different degrees of cytological abnormalities, as described above, were evident. These nodules, shown in Fig. 6,C, when examined for BrdUrd immunostaining, displayed a much higher proliferation index as compared with that of the tissue from age-matched control animals (Fig. 6 D).

In general, as summarized in Table 1 and Fig. 7, between 2 and 14 months, 100% (17/17) of line 3 transgenic animals developed multifocal hyperplasia in at least one lobe of the prostate, and 35% (6 of 17) developed LGPIN. No papillary hyperplasia or HGPIN was detected in animals up to the age of 14 months. After 15 months, 100% (39 of 39) transgenic animals continued to display hyperplasia in increased number of lobes; 23% (9 of 39) developed papillary hyperplasia with atypia, the incidence of LGPIN increased to 66% (26 of 39), and 51% (20 of 39) developed HGPIN. Similar lesions were also identified in the transgenic animals of other lines (Table 2). Throughout the investigation period, none of the nontransgenic control animals developed PIN lesions, whereas a mild hyperplasia was noted in ∼20% of aging controls.

In line 3, chronic inflammation was frequently noticed starting at 5 months and became more common in older animals (Table 1). Generally, the inflammation was found in the LP and VP, although AP and DP were also involved but to a lesser extent. Similar changes were also present in animals of line 1, occasionally with a higher intensity when compared with animals of other lines (Table 2). Although the earliest time points of their appearances were similar, a direct correlation between the inflammation and PIN lesions could not be made, because there were specimens with PINs that lacked overt evidence of inflammation and vice versa. Microscopically, clusters of lymphocytes were sparsely located in the stroma (Figs. 4, 5, and 6,E). While in close continuity with the ducts, these lymphocytes did not seem to invade the epithelial cells in most animals. Immunohistochemically, most of these lymphocytes were CD3 positive (Fig. 6,F), whereas a small percentage was of B-cell lineage as determined by immunoreactivity to an anti-CD45R/B220 antibody (Fig. 6,G). In rare cases (2 of 56 animals in line 3), a mixed acute and chronic inflammation was found where the glandular profile was severely disrupted. In one animal of line 3 and one of line 1 (Tables 1 and 2), the CD3-positive inflammatory cells appeared in a pattern suggestive of lymphoma within the prostate tissues, a matter that remains to be further investigated. In these animals, the inflammatory infiltration was also found in some extraprostatic tissues, such as small intestine and stomach. The stromal proliferation, which was relatively more prominent in the AP than other lobes, was another remarkable change in PB-FGF8b transgenic animals. It was characterized by significantly thick stroma with hypercellularity (Figs. 4,A and 6, A and B). Smooth muscle cells were identified by immunohistochemistry as the major component of the stroma (Fig. 6,H). These stromal cells also exhibited a high proliferation index (Fig. 6,C). In some animals, proliferation of stromal cells apparently led to stromal papilloma (Fig. 6, A and B), which manifested the phyllodes-like pattern. These cells, having increased nuclear:cytoplasmic ratio, displayed pleomorphic, hyperchromatic nuclei with prominent nucleoli. A few scattered mitotic figures could also be noted (Fig. 6 B).

The mild chronic inflammation was also found in the stroma of the DD, SV, and epididymis in some of the transgenic animals. Interestingly, these were also the nonprostatic tissues that manifested a low level of transgene expression as detected by RT-PCR assay. However, these abnormalities did not affect the fertility of the transgenic animals because all established lines could be successfully maintained. Chronic inflammation, much less severe than that identified in transgenics, was also noted in ∼10% of old control animals. However, none of those control animals exhibited significant stromal hypercellularity.

Although there is ample evidence for the up-regulation of FGF8 gene expression in human prostatic premalignant and malignant lesions (12, 13, 15, 32), until now it was not known whether this observation is unique to the human disease or more general in nature in terms of prostate carcinogenesis. Here, we demonstrate that mouse fgf8 transcription is also elevated in PIN lesions of SV40 Tag-driven mouse models (5). Similar to the observations in human prostate cancers, we find that fgf8, isoform b mRNA, whose amino acid coding sequence is 100% identical between the human and mouse species (11), is specifically up-regulated in mouse PINs. Thus, it was considered logical to develop mouse models in which the fgf8b is targeted for overexpression in the prostatic epithelium.

We describe four productive lines of transgenic mice, in each of which the expression of FGF8b transgene is under the control of the androgen-regulated, prostate epithelium-specific ARR₂PB promoter (38). These independent lines display a similar sequence of development of phenotypic changes in prostatic tissues. In general, the results point to an increased expression of FGF8b that is sufficient to drive proliferation in the prostatic epithelium preceding the development of histopathologically identifiable lesions, many of which resemble human preneoplastic prostatic lesions. A stochastic pattern of disease progression in these transgenic mice is noteworthy (Fig. 7). The prostatic abnormalities, beginning with multifocal epithelial hyperplasia, are followed by appearance of LGPIN and, subsequently, HGPIN lesions. If all of the prostatic pathology is combined, irrespective of whether one of more lobes are involved, 100% of the transgenic animals manifest prostatic hyperplasia. Although this high incidence of hyperplasia is followed by the development of LGPIN at a rate of 35% within the first 14 months, no HGPINs could be detected up to this time point. However, during further aging (15–24 months), as the incidence of LGPIN increases from 35 to 66%, there is also the first appearance of HGPIN at a remarkable high frequency of 51%. Thus, the overexpression of FGF8b appears to be a distinct initiating event in the development of hyperplasia which, in turn, is perhaps conducive to the manifestation of other genetic lesions, which may represent the rate-limiting factors responsible for a potentially temporal progression from hyperplastic changes to HGPIN lesions. These properties of the transgenic mice, which are markedly prostate restricted, also closely reflect the usual slow progression of prostatic disease in humans.

Although the specific mechanisms by which FGF8b drives tumorigenesis are not known, some of the observations made in vivo with this transgenic model are concordant with those reported previously in other systems. For example, in vitro mitogenic and transforming activity of FGF8b (11, 26, 30, 31) or tumorigenicity in mouse mammary tumor virus promoter-driven mammary or ovarian epithelium (31) are consistent with the current findings. Moreover, a delayed but fairly common development of stromal hypercellularity in the prostate of the FGF8b transgenic mice mimics the in vitro coculture experiments described before (16), implicating an indirect effect of FGF8b signaling in epithelial cells on the stromal cells. Thus, FGF8b is likely to act not only as an initiation factor but also, possibly, as a progression factor. Because transgene expression in our model is found to be continuous throughout the 2-year period of life investigated and considering that FGF8b overexpression in weakly tumorigenic human prostatic tumor LNCaP cells could significantly enhance their tumorigenicity and invasiveness (16), this potential should not be overlooked. However, because the model has yet to yield invasive cancer, and as described above, there are likely to be other rate-limiting factors in the progression of the lesions, FGF8b is only proven to be an etiological factor in prostate tumorigenesis in this model. Its role as a progression factor, along with other hitherto unidentified secondary factors, remains to be investigated.

In summary, transgenic mice overexpressing FGF8b in prostate epithelial cells are found to develop progressively epithelial hyperplasia, LGPIN, and ultimately HGPIN. Interestingly, abnormalities of stroma including hyperplasia and chronic inflammation are also observed in these animals. These findings are biologically significant because FGF8b expression is associated with progression of human prostate cancer, beginning with PIN lesions. The model is likely to be valuable in examining how FGF8b may be involved in influencing the autocrine and paracrine pathways in the prostate tissue. It is also important to note that the potential of the model could be greatly enhanced by seeking genetic synergy between FGF8b transgenics with mutated mice known to reflect other “natural” changes in prostate cancer, such as Nkx 3.1 (6, 44), p27 (45), and PTEN (46). With the development of an efficient prostate epithelium-specific Cre-loxP model (39), it is also now possible to generate animals with conditional alleles of genes whose disruption is otherwise embryonically lethal. One such allele, RXRα retinoid receptor (47, 48), has already been combined in the background of FGF8b overproduction, and there is indication of acceleration of the phenotypic changes in the prostate from this intercrossing.⁶

We thank Dr. Robert Maxon for providing us the pSV vector containing SV40 poly(A) and splicing signal sequence and for use of the transgenic facility under his direction. We are grateful to Dr. Scott Shappell of Vanderbilt University for assessment of some of the pathology slides and all members of the Roy-Burman laboratory for assistance in various aspects of the work.