CD4+ lymphocytes modulate prostate cancer progression in mice

Abstract

Chronic inflammation contributes to the development of prostate cancer in humans. Here, we show that male Apc^Min/+ mice also develop prostate carcinoma with increasing age, mimicking that seen in humans in their 5th or 6th decade of life. Proinflammatory cytokines were significantly linked with cancer and increasing age in our mouse model; however, prostate and bowel tissues lacked evidence of inflammatory cell infiltrates other than mast cells. Lymphocytes protected against cancer, and protection from prostate cancer resided in antiinflammatory CD4⁺CD25⁺ regulatory (T_REG) cells that downregulated inflammatory cytokines. Supplementation with syngeneic T_REG cells collected from wild-type mice reduced the levels of interleukin (IL)-6 (p < 0.05) and IL-9 (p < 0.001) and lowered prostate cancer risk (p < 0.05). Depletion of CD25⁺ cells in 2-month-old animals increased the expression of IL-6 (p < 0.005) within prostate and increased the frequency of high-grade prostatic intraepithelial neoplasia (p < 0.05) and microinvasive prostatic carcinoma (p < 0.05) in dorsolateral prostate. Depletion of CD25⁺ cells in young animals also increased the frequency of intestinal cancer in Min mice. Taken together, chronically elevated proinflammatory cytokines promoted carcinoma in Apc^Min/+ mice. T_REG lymphocytes downregulated inflammation-associated carcinogenic processes and contributed to immune and epithelial homeostasis.

Prostate cancer is the 3rd leading cause of cancer deaths in American men.¹ Histological features consistent with prostate neoplasia increase in frequency with increasing age, such that prevalence of high-grade prostatic intraepithelial hyperplasia (HGPIN) rises to levels of 60% by the 6th decade of life.² Clearly, it is of great interest to further our understanding of exogenous and endogenous risk factors that promote prostate cancer. One such potentially important risk factor is chronic inflammation. On the basis of epidemiology, molecular pathology and sequellae resulting from chronic inflammation, it has been proposed that recurrent or chronic inflammation may be responsible for the increased risk of prostate cancer with increasing age in men.³ Here, we have directly addressed this possibility, and we indeed find that conditions that promote inflammation increase the frequency and severity of prostate cancer.

A link between chronic inflammation and cancer was originally described more than a century ago.⁴ Although acute inflammation can help destroy tumor cells, extensive epidemiologic data from human patients show that chronic inflammation is associated with disruptive changes to tissues that predispose susceptible individuals to cancer in many tissues throughout the body.^5–9 For example, a link between inflammation and cancer has been demonstrated in lower bowel with colitis¹⁰ and prostate in individuals with prostatitis^11,12 who are at increased risk for cancer. Furthermore, individuals on chronic NSAID therapy including cyclooxygenase (Cox)-2 inhibitors have reduced colonic and prostate cancer risk.^13–15 Along these lines, several studies have observed increased expression of Cox-2 within prostate cancer.⁵ With the growing appreciation of the importance of inflammation and Cox-2 in cancer development, there is a need for improved animal models that allow dissection of the molecular underpinnings that connect chronic inflammation and cancer.¹⁶

The adenomatosis polyposis coli (APC) gene is a critical tumor suppressor gene that prevents malignancies of the colon^17,18 and prostate^19,20 in humans. Indeed, families that carry a mutation in one copy of this gene suffer multiple intestinal neoplasia (Min) because of the spontaneous loss of the remaining wild-type copy.^21,22 Loss of function of both copies of the Apc gene leads to disruptions in the critical β-catenin and the wnt signaling pathway in the prostate.²³ Nearly one-quarter of advanced prostate cancers in humans have alterations in β-catenin expression, which is consistent with an important role of Apc in preventing prostate cancer.^24,25 Indeed, inactivation of the Apc gene in mouse prostate leads to the development of prostate carcinoma.²⁵ In mice that are heterozygous for the Apc gene (Apc^Min/+), cells completely lacking Apc frequently arise spontaneously because of the loss of heterozygosity. Consequently, by 3 months of age, Apc^Min/+ mice suffer a large number of intestinal polyps that are associated with alterations in β-catenin expression and upregulation of Cox-2.²⁶ Interestingly, Apc^Min/+ mice that lack functional lymphocytes (Rag2^−/−) are at increased risk for intestinal polyps and submucosal invasion, possibly because of a deficiency or malfunction in cells that normally suppress inflammation.²⁷

T lymphocytes are thought to play an important role in suppressing inflammation, and we hypothesized that they may therefore play an important role in suppressing prostatitis and subsequently prostate cancer. Roles for T lymphocytes in immune homeostasis have been evident for several decades, as first observed in mice undergoing neonatal thymectomy.²⁸ In the mid-1990s, Sakaguchi et al.²⁹ showed that T cells expressing CD25 prevent immune disorders by preventing proliferation of autoreactive T lymphocytes. Likewise, Powrie isolated CD4⁺ T regulatory (T_REG) cells with potent antiinflammatory properties that prevent inflammatory bowel disease³⁰ and promote systemic immune homeostasis.³⁰ Since that time, adoptive transfer or selective lymphocyte depletion techniques have been widely utilized in mice to elucidate roles for CD4⁺CD25⁺ cells in immune homeostasis. Adoptive transfer studies reveal that CD4⁺CD45RB^loCD25⁻T_REG³¹ and CD4⁺ CD45loCD25⁺ T_REG^32–36 cells are potent inhibitors of intestinal inflammatory responses. Relationships between T_REG cells and cancer are less clear. CD4⁺ CD25⁺ T_REG have been shown to inhibit intestinal polyposis in Apc^Min/+ mice^37,38; however, CD4⁺CD25⁺ T_REG cells may also undermine beneficial inflammatory responses such as for vaccination purposes.^39–42

Here, we combined our ability to manipulate the levels of CD25⁺ cells in vivo with the prostate-cancer prone Apc^Min/+ mice to test the possibility that T_REG cells help suppress endogenously arising prostate cancer. In the Apc^Min/+ mice, we observed age-associated inflammatory cytokine-dependent development of prostate carcinoma, replete with mast cells but otherwise lacking overt inflammation. Interestingly, we not only found that CD4⁺ T_REG cells downregulate inflammatory cytokines but also observed suppression of malignancy, thus opening doors to new possibilities for the prevention and possibly the treatment of prostate cancer.

Material and methods

Experimental animals

All animals were housed in AAALAC approved facilities and maintained according to protocols approved by the IACUC at Massachusetts Institute of Technology. Apc^Min/+ mice on a C57BL/6J background were originally obtained from the Jackson labs and bred in house as (heterozygous X wildtype) crosses to provide Apc^Min/+ mice and wildtype littermates for experimental recipients and donors.

Experimental design

A total of 160 mice were used for these experiments. Eighty-four Apc^Min/+ mice were included in various treatment regimens or as experimental controls. Sixty C57BL/6 wild-type littermates were used as experimental animals or as cell donors for adoptive transfer experiments. Sixteen Rag2-deficient Apc^Min/+ mice housed under identical conditions were used for comparisons of prostate pathology. Experiments were conducted using separate trials with 5–8 mice each. Trials were then run in duplicate or triplicate, as noted.

Adoptive transfer of T_REG cells

A total of 10 (5 mice per trial, 2 trials) Apc^Min/+ mice aged 3.5–4 months were dosed with 3 × 10⁵ T_REG cells. CD4⁺CD45RB^loCD25⁺ (T_REG) lymphocytes were isolated from spleen and mesenteric lymph nodes and adoptively transferred as previously described.³⁷ Upon necropsy, tissues of these mice were compared to 10 (2 trials of 5 mice per trial) untreated 6-month-old Apc^Min/+ mice. In a 2nd experiment, a total of 8 Rag2-deficient Apc^Min/+ mice aged 3.5–4 months were dosed with 3 × 10⁵ T_REG cells. These mice were compared to 8 untreated Rag2-deficient Min mice at 6 months of age. The donor mice for T_REG cells included male and female C57BL/6 wild-type littermate mice. The recipients used in this study were all male, in order to study prostate lesions.

Depletion of CD25⁺ cells

A total of 16 (2 trials of 8 mice each) male Min mice aged 4–6 weeks were treated with anti-CD25 antibody (clone PC-61; Bio-Express, West Lebanon, NH) at 150 µg intraperitoneally per mouse 2 times weekly for 4–6 weeks. Treated mice were compared to age-matched Min mice that received sham isotype antibody alone (N = 16). Depletion of CD25⁺ cells was confirmed by undetectably low fractions of CD25+ cells in spleens of 4 mice treated with anti-CD25 antibody compared to 4 sham-treated controls using flow cytometry upon completion of the study. In addition, depletion was confirmed by absence of Foxp3+ cells in target tissues, including prostatic and mesenteric lymph nodes at 42 days (N = 8) after onset of the depletion protocol. Depletion results were then confirmed in an additional 8 anti-CD25-treated and 8 sham-treated mice at 10 days (N = 4 mice per time point per group) and 24 days (N = 4 mice per time point per group) after onset of treatment.

Quantitation of intestinal tumors

Location of tumors was recorded using a stereomicroscope at 10× magnification. Location of tumors in the small intestine was recorded as distance from the pylorus with duodenum, jejunum and ileum comprising one-third of small intestine each and in the colon as distance from cecocolic junction.³⁷

Detection of systemic cytokine protein expression

Serum cytokine levels of 5 animals in each experimental group were analyzed using the Bioplex assay system (BioRad, Hercules, CA) according to the manufacturers protocol. Briefly, serum samples were diluted using the sample diluent kit, incubated with antibody coated beads followed by a secondary antibody incubation and analyzed in duplicates on a Bio-Plex 200 system (BioRad, Hercules, CA). The levels of TNF-alpha, IL-1b, IL-6, IL-9, IL-13, IL-17, IFN-gamma and G-CSF were assessed. Statistical analysis was performed using 2-tailed student’s t-test; a p-value of <0.05 was considered as statistically significant.

Histologic evaluation and Immunohistochemistry

As described previously,³⁸ the formalin-fixed tissues were processed, and tissue-sections were stained with H&E, toluidine-blue or IHC-evaluated by a veterinary pathologist blinded to sample identity. The frequency of invasive adenocarcinoma lesions in prostate⁴³ or bowel⁴⁴ was recorded for the various experimental groups. The preneoplastic and early neoplastic prostate lesions, which were detected based on recent consensus report criteria,⁴³ were quantitatively assessed as follows: From each group of mice, 30 high-power fields (20×) containing the glandular profiles that show the most advanced preneoplastic and early neoplastic lesions were captured using a Nikon eclipse 50i microscope and a Nikon DS-5 M-L1 digital camera. Ten images were randomly selected per treatment group. Lesions were then quantified and recorded as the number of low-grade PIN, high-grade PIN and microinvasive carcinoma foci counted per image.

Primary antibodies used for immunohistochemistry included rabbit anti-Apc, rabbit anti-β-catenin (ThermoFisher Scientific/Lab Vision, Fremont, CA), rabbit anti-E-cadherin (Santa Cruz Biotechnology, Santa Cruz, CA) and mouse anti-SMA (Dako, Glostrup, Denmark). Antigens were retrieved with Trypsin (ThermoFisher Scientific/Lab Vision) for Apc detection. Heat-induced antigen retrieval was performed with citrate buffer, pH 6, for β-catenin and E-cadherin or with EDTA buffer, pH 8, for SMA detection. Primary antibody binding was detected with goat antirabbit IgG Poly-HRP (Chemicon International, Temecula, CA) or the mouse-to-mouse detection system and goat anti-mouse IgG Poly-HRP (Chemicon). Signal was detected with diaminobenzidine, and tissues were counterstained with haematoxylin.

Detection of cytokine mRNA expression

Total RNA from ileum tumors of Apc^Min/+ mice was prepared using Trizol reagent according to the supplier’s instruction (Invitrogen, Carlsbad, CA). Five micrograms of the respective RNA was used to generate cDNA using the High Capacity Archive Kit from A/B Applied Biosystems according to the supplier’s procedure. Levels of Cox-2, TNF-α and IL-6 transcripts in the cDNA samples were quantified with the corresponding commercial primers and probes in the ABI Prism Sequence Detection system 7700 (A/B Applied Biosystems). The levels of these cytokines among the samples were normalized by the level of the GAPDH transcripts and compared between the CD4⁺CD25⁺ T_REG cells-treated and -untreated tumors by the ΔΔCt method described by A/B Applied Biosystems (User Bulletin #2).

Statistical analyses

Total prostate lesion counts and intestinal tumor counts were analyzed by unpaired t-test with Welch’s correction. For all statistical analyses, Graphpad Prism version 4.0 for windows, Graph-Pad software, San Diego, CA, USA was used.

Results

Prostate carcinoma arises spontaneously in Apc^Min/+ (Min) mice

Mutations of the Apc gene have been associated with prostate cancer development in both humans^24,45 and mice.²⁵ Here, we examined male mice that undergo spontaneous loss of heterozygosity of Apc (Apc^Min/+) and found that unmanipulated Apc^Min/+ mice between 5 and 6 months of age develop significantly (p < 0.01) more frequent prostate cancer when compared to age-matched wildtype (wt) littermate controls housed under the same conditions (∼40% in Apc^Min/+ vs. 0% for the wt controls; see Table I). Furthermore, histological features of both preneoplastic and neoplastic lesions were evident among the Apc^Min/+ mice (Fig. 1), while being completely absent among the wild-type controls (Table I). Lesions seen in Apc^Min/+ mice were consistent with prostate carcinoma according to criteria of Shappell et al. in a consensus report entitled “Prostate Pathology of Genetically Engineered Mice: Definitions and Classification” published in Cancer Research,⁴³ and also matched prostate cancer previously described in mice with inducible inactivation of the Apc gene.²⁵

TABLE I.

Pre–Neoplastic Findings (In Dorsolateral Prostate) of Male Apc^Min/+ (Min) Mice (Mean ± SE Per 10 20 × Images)

Treatment group	Low grade PIN		High grade PIN			Microinvasive carcinoma
wt (control; age 6 months)	0.4 ± 0.1	**	0.0 ± 0.0	**	**	0.0 ± 0.0	** *
Min (age 6 months)	1.5 ± 0.31		1.2 ± 0.20			1.0 ± 0.21		*	*
Min (age 6 months) + anti-TNF	1.6 ± 0.30		1.1 ± 0.31			0.6 ± 0.16
Min (age 6 months) + TREG	1.7 ± 0.30		1.1 ± 0.31			0.4 ± 0.16
Min (age 3 months) untreated	1.6 ± 0.30		0.9 ± 0.34			0.2 ± 0.13
Min (age 3 months) + sham IgG	1.4 ± 0.22		0.5 ± 0.16	*		0.3 ± 0.15	*
Min (age 3 months) + anti-CD25 IgG	1.5 ± 0.30		1.6 ± 0.16			1.0 ± 0.21

Frequency and features of prostate lesions are given in this table. CD25⁺ cells are important in impeding age-related progression of neoplasia in the prostate of Min mice. Min mice depleted of CD25⁺ cells (N = 8 mice per trial) from age 1 to 3 months had significantly more HGPIN and microinvasive carcinoma prostate lesions when compared to agematched sham antibody treated (N = 8 mice per trial) control Min mice. Min mice that received supplemental CD25⁺ T_REGS (N = 5 mice) showed a statistically significant decrease in the frequency of microinvasive carcinoma lesions when compared to Min mice of matched age = 6 months (N = 5 mice). Grading of neoplastic progression in the DLP of mice and analysis of results was performed as described in Material and methods.

^*p < 0.05;

^**p < 0.01;

^***p < 0.001.

Figure 1.

Depletion of CD25⁺ cells increases the risk of prostate cancer in young mice, but complete lack of lymphocytes and increasing age are the major factors contributing to the progression of prostate neoplasia in Apc^Min/+ mice. (a, b) HGPIN lesions in the anterior prostate of CD25-depleted Min mice at 3 months of age. Glands in a are distorted, surrounded by irregular fibromuscular sheaths and filled with atypical epithelial cells. The stroma adjacent to gland profile with PIN lesions in b contains neutrophils (open arrow-head) and mast cells (arrow-head). (c, d) Early invasive neoplastic lesions in the dorsolateral prostate of a RagMin mouse at 6 months of age. Highly irregular glands profiles with tufting epithelia and edematous stromal reaction (c). Higher magnification of the boxed area is shown in D. Early invasion of neoplastic epithelium is found in association with HGPIN. Note the hyperchromasia and the nuclear pleomorphism of the epithelial cells in PIN lesions (open arrow) and the large, euchromatinic nuclei bearing prominent nucleoli of invasive epithelial structures (arrow). (e, f) Prostate adenocarcinoma in the dorsolateral prostate of Min mice at 6 months of age. Moderately differentiated adenocarcinoma with desmoplastic reaction. Abnormal mitotic figure is indicated by arrow-head (e). Nuclear stabilization of β-catenin characterized advanced prostate adenocarcinoma lesions (f). (g, h) Prostate adenocarcinoma in the dorsolateral prostate of Min mice at 6 months of age. Well-differentiated adenocarcinoma with prominent desmoplasia (g). Increased numbers of mast cells (arrow-head) in stroma associate with abnormal glands (h). (a, b, c, d, e, g, h) Hematoxylin and Eosin. (f) β-catenin-specific immunohistochemistry; Hematoxylin counterstain, DAB chromogen. Bars a, c and g: 100 µm; b, d, e, f and h: 25 µm.

Prostate cancer is more frequent with increasing age

In humans, prostate cancer is more common with increasing age.³ To test whether prostate cancer frequency is associated with aging in this model, we examined male Apc^Min/+ mice at 3 or 6 months of age. We discovered that Apc^Min/+ male mice of age 5–6 months developed significantly more frequent prostatic microadenocarcinoma lesions than the younger 3-month-old Apc^Min/+ males (compare groups in Table I, Figs. 2 and 3), showing that prostate disease is dependent on increasing age of the animals. We analyzed a spectrum of nonneoplastic and neoplastic proliferations, including multifocal hyperplasia with or without atypia and low-grade (LG) to high-grade (HG) prostatic intraepithelial neoplasia (PIN). The HGPIN was occasionally found associating with microinvasive carcinoma. Importantly, only aged Min mice had moderately to well-differentiated adenocarcinoma accompanied by desmoplastic reaction. We also assessed signs of inflammation, and although aggregates of inflammatory cells were not observed, stroma adjacent to preneoplastic and neoplastic epithelia consistently contained mast cells and occasional neutrophils. The seminal vesicles and the anterior prostate were most frequently affected sites, followed by the dorsolateral and the ventral prostate lobes. These data indicated that Apc^Min/+ mice have an age-associated predilection to spontaneously arising carcinoma in the prostate. These findings are consistent with prostate carcinoma described in mice with inducible inactivation of the Apc gene.²⁵

Figure 2.

Frequency of prostate carcinoma lesions rises with increasing age. Prostate carcinoma arose spontaneously in male Min mice but not wt mice. Aged (6 months old) Min mice (N = 5 mice per trial) had greater frequency of early neoplastic prostate lesions, when compared to younger (3 months old) Min mice (N = 5 per trial). Pre-neoplastic and early neoplastic prostate lesions, which were detected based on recent consensus report criteria,⁴² were quantitatively assessed for each group of mice based upon review of 30 high-power fields ×20 (data presented as mean value) containing the glandular profiles low-grade PIN, high-grade PIN and microinvasive carcinoma as shown earlier. The most advanced preneoplastic and early neoplastic lesions were captured using a Nikon eclipse 50i microscope and a Nikon DS-5 M-L1 digital camera. Ten images were randomly selected per treatment group. Lesions were then quantified and recorded as the number of low-grade PIN, high-grade PIN and microinvasive carcinoma foci counted per image.

Figure 3.

(a) Frequency of prostate cancer lesions in Min and Rag-deficient Min mice. Depletion of CD25⁺ cells (Min⁺anti-CD25) (N = 8 mice per trial) accelerated prostate carcinogenesis at age of 3 months, and the frequency of HGPIN and microinvasive adenocarcinoma was significantly higher than those of age-matched (N = 5 mice per trial) or sham-treated (N = 8) mice (p < 0.01 and <0.05, respectively). Compared to the age-matched Min mice (N = 5), RagMin mice (N = 8) had significantly higher frequency of prostate neoplasia at 6 months old (p < 0.05). Assays used tissues from 5 to 8 mice per treatment group (as shown) with review of microscopic fields as described in Material and methods. (b) Histology of dorsolateral prostate illustrating selected key intermediate steps in progression of carcinogenesis in Apc^Min/+ mice. (a) Normal-appearing prostate gland. (b) Low-grade prostatic intraepithelial neoplasia (PIN) with epithelial tufting and nuclear stratification. Cellular atypia is evidenced by the presence of nuclear enlargement, increased but not severe nuclear pleomorphism, hyperchromasia and occasional prominent nucleoli. (c) High-grade PIN. Note the irregular contour of the prostate gland. Highly atypical cells with severe nuclear pleomorphism and hyperchromasia fill the lumen. (d) Microinvasive carcinoma. Epithelial cells with notably large, euchromatinic nuclei bearing prominently enlarged nucleoli with penetration through the basement membrane into surrounding stroma. (e) Invasive adenocarcinoma. Moderately differentiated small, irregular malignant glands are bounded by desmoplastic stroma. Hematoxylin and Eosin. Bars: 25 µm. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

Interleukin-6 is elevated in mice with prostate cancer

The pleiotropic cytokine IL-6 has been linked with the development of prostate cancer in humans. To determine whether prostate carcinoma is associated with changes in the levels of IL-6 and other cytokines in this murine model, we examined sera of Apc^Min/+ mice and wt mice. We found that levels of certain cytokines in sera were indeed significantly higher in Apc^Min/+ mice, a group that exhibited higher risk of carcinoma. Specifically, we observed significant elevations of IL-6 (p < 0.01) and IL-9 (p < 0.001) in aged Apc^Min/+ mice (Table II; compare wt and Min mice). Additional analysis of gene expression within the prostate tissue revealed a clear increase in IL-6 (p < 0.005) in groups of Apc^Min/+ mice also prone to carcinoma, but not in wt animals (Fig. 4a). Elevations in TNF-α and Cox-2 expression in prostate tissue did not reach significance (data not shown).

TABLE II.

Serum Cytokine Levels in Pg/Ml

	TNF-a	IL-1b	IL-6	IL-17	IFN-g	G-CSF	IL-9	IL-13
wt 6 months vs. APCMin/+ 6 months
WT	638 (135)	225 (22)	18 (6)	45 (7)	142 (12)	81 (101)	138 (40)	108 (23)
APCmin/+	153 (359)	314 (38)	90 (16)	52 (11)	136 (45)	229 (93)	591 (77)	12 (35)
p-value	ns	ns	0.027	ns	ns	ns	0.00002	ns
wt 6.months vs.RagAPCmin/+ 6 months
WT	638 (135)	228 (22)	18 (6)	45 (7)	142 (12)	81 (101)	138 (40)	108 (23)
RagAPCmin/+	699 (56)	32 (153)	113 (17)	66 (16)	190 (58)	238 (114)	538 (74)	128 (14)
p-value	ns	ns	0.039	0.034	ns	ns	0.00002	ns
Apcmin/+ 6 months vs.Rag APCmin/+ 6 months
APCmin/+	1537 (359)	314 (38)	90 (16)	52 (11)	136 (45)	229 (93)	591 (77)	120 (35)
RagAPCmin/+	699 (56)	327 (153)	113 (17)	66 (16)	190 (58)	238 (114)	538 (74)	128 (14)
p-value	ns	ns	ns	ns	ns	ns	ns	ns
Rag Apcmin/+ 6 months untreated vs.Treg
Untreated	699 (56)	327 (153)	113 (17)	66 (16)	190 (58)	238 (114)	538 (74)	128 (14)
Treg	500 (114)	225 (28)	20 (3)	46 (13)	115 (7)	60 (20)	151 (74)	89 (11)
p-value	0.016	ns	0.043	ns	0.021	0.010	0.00032	0.002
Apcmin/+ 3 months vs.APCmin/+ 6 months
3 months	619 (55)	228 (22)	16 (3)	44 (7)	143 (11)	57 (29)	112 (17)	112 (6)
6 months	1537 (359)	314 (38)	90 (16)	52 (11)	136 (45)	229 (93)	591 (77)	120 (35)
p-value	ns	ns	0.023	ns	ns	0.013	0.00006	ns

Regulatory T cells restore immune homeostasis. Serum levels of cytokines IL-6 and IL-9 were significantly increased in aged Min mice at high risk of intestinal polyposis and prostate cancer. Supplementation with wt T_REG cells restored immune homeostasis, prevented cancer development and decreased serum cytokine levels of aged Rag2−/−Min mice to those of wt mice or young adult Min mice. Serum cytokine levels in pg/ml; assay used sera of n = 5 mice per group. Standard error is presented in parentheses.

Statistics calculated using 2-tailed Student’s t-test ns, not significant (p = 0.05).

Figure 4.

mRNA expression levels of IL-6 and Cox-2 in the murine prostate and bowel. There were significant increases in expression of IL-6 in (a) prostate and (b) ileum of Min mice treated with anti-CD25 antibody, and also in Rag-deficient Min mice, when compared to sham-treated Min counterparts. Supplementation with T_REG cells from wt donor mice decreased expression of IL-6 in prostate tissue, whereas depletion of CD25⁺ cells increased IL-6 gene expression in prostate tissue. Assay used tissues from 5 to 8 mice per treatment group. For comparison of mRNA levels, the target mRNA was normalized to that of the housekeeping gene GAPDH. Numbers on the y-axis represent mean fold change of target mRNA levels in reference to the control levels (B6 wt, defined as 0, standard deviation represented by solid bars). mos, age in months upon necropsy.

Although both the serum cytokine levels and the tissue cytokine mRNA levels point to an association between prostate cancer and cytokines, we wished to learn more about the possibility that cytokines may actually promote cancer development. To test whether prostate cancer depends on inflammatory cytokines in Apc^Min/+ mice, we neutralized a proinflammatory cytokine, tumor necrosis factor (TNF)-α, which is involved in initiation of immune activation and also amplification of other cytokines, using intraperitoneal injection of anti-TNF-α-antibodies for 10 days starting at age 20 weeks, as previously described.^27,38 We observed, in particular, a reduction in features of neoplastic invasion (Table I, rows 3 and 6) in treated mice when compared to age-matched controls. This requirement of sustained inflammation for progression of neoplastic invasion has been previously shown in other murine models.^38,46

Lymphocytes suppress the risk of developing prostate carcinoma

We next sought to determine whether or not lymphocytes, which have roles in both promoting and suppressing cancer, may be needed for prostatic neoplasia in Apc^Min/+ mice. To address this question, we exploited Rag2-deficient C57BL/6 mice to create Rag2-defcient Apc^Min/+ mice that completely lack functional lymphocytes. The lymphocyte compartment harbors cells with both pro- and anticancer functions, so it was unknown at the outset whether lack of lymphocytes would promote or suppress prostate cancer development. Interestingly, we found that Rag2-deficient Apc^Min/+ mice aged 5–6 months developed more frequent neoplastic features in the prostate gland (Fig. 1) when compared to age-matched Apc^Min/+ controls that were housed under the same conditions (Table I; Fig. 3).

Qualitatively, however, the spectrum of prostatic lesions tabulated in aged Rag2-deficient Apc^Min/+ mice was similar to that of Apc^Min/+ mice (shown in Fig. 1). One exception was that Rag2-deficient Apc^Min/+ prostate stroma had a more prominent edematous alteration and a greater frequency of mast cells. Mast cells (Figs. 1e and 1f) were a frequent feature of malignancy in this model. We have previously shown that TNF-α-dependent activities of mast cells are essential for intestinal polyp development in Apc^Min/+ mice.⁴⁷ Significant elevations of IL-6 (p < 0.05) and IL-9 (p < 0.001) were observed in Apc^Min/+ mice (Table II). Use of the Rag2-deficient host highlights the innate immune contributions toward IL-6 in prostate tissue in Apc^Min/+ mice and role for these cells in subsequent pathogenic events in humans and mice.⁴⁸

Taken together, it is clear that lymphocytes are not absolutely required for prostate tumorigenesis in the Apc^Min/+ mouse model, an observation that is consistent with earlier studies.²⁷ The observation of increased invasive prostate lesions in Rag2-deficient Apc^Min/+ mice when compared to their Apc^Min/+ counterparts (Table I; Fig. 3) indicates that lymphocytes are among cells that protect against malignancy. On the basis of prior data derived from Apc^Min/+ mice,^27,37,46,47 we postulated that CD4⁺ T_REG cells may function to maintain immune homeostasis and inhibit prostate cancer in Apc^Min/+ mice.

CD4⁺ T_REG cells inhibit prostate carcinogenesis

Having observed that prostatic neoplasia was exacerbated by a deficiency in lymphocytes, we wondered whether supplementation with T_REG cells might inhibit developing prostate cancer in Rag2^−/− Apc^Min/+ mice. The Rag2^−/− Apc^Min/+ mice permit direct assessment of the impact of adoptively transferred CD4⁺CD45RB^loCD25⁺ T_REG in the complete absence of endogenous T_REG cells or other lymphocytes. Highly purified T_REG cells derived from wild-type cell donors were adoptively transferred into 14-week-old Rag2^−/−Apc^Min/+ mice. We found that recipients of T_REG cells examined 8 weeks later, at 5–6 months of age, had a significant (p < 0.01) reduction in prostate pathology (N = 8; Fig. 3), in cytokine levels of TNF-α (p < 0.05), IL-6 (p < 0.05) and IL-9 (p < 0.005) in the serum (Table II), and in cytokine expression levels in the prostate tissue (p < 0.01) (Fig. 4a) when compared to untreated Rag2^−/−Apc^Min/+ mice. Taken together, these data provide evidence that T_REG cells alone, when collected from immune-competent donors, are sufficient for inhibition of prostate cancer.

We next examined whether supplementation with T_REG cells might inhibit progression of developing prostate cancer in Apc^Min/+ mice that have endogenous lymphocytes. To test this possibility, 10 4- to 5-month-old male Apc^Min/+ mice underwent adoptive transfer of T_REG cells from syngeneic wildtype (wt) donors. Despite the fact that the recipient mice already had T_REG cells present, we nevertheless observed a significant (p < 0.05) reduction in prostate pathology in T_REG-treated mice (N = 10, with 2 trials of 5 mice per trial; see Table I, Fig. 3). Examination of bowel from the same treated and untreated Apc^Min/+ mice showed a significant (p < 0.01) reduction in intestinal adenoma multiplicity (N = 10; µ = 18 ± 3.68) in Apc^Min/+ recipients of T_REG cells when compared to untreated Apc^Min/+ mice (N = 10; µ = 36 ± 5.40). These data, taken together with prior studies,^27,46,49 support an essential protective role for T_REG cells of immune-competent animals in maintenance of epithelial homeostasis and prevention of cancer.

Depletion of CD25⁺ cells accelerates progression of prostate carcinogenesis

Finally, we tested whether depletion of T_REG cells may alter development of prostate cancer. We reasoned that if T_REG are critical in preventing development of prostate cancer, then removing them from young mice should increase cancer risk later in life.^37,38 Our lab³⁸ and others³¹ have previously shown that both CD4⁺CD25⁺ and CD4⁺CD25⁺ subsets of T_REG cells exhibit potent antiinflammatory properties in mice. For these experiments, we performed depletion of the CD25⁺ subset of T_REG cells using antibodies against CD25. We depleted CD25⁺ cells in Apc^Min/+ mice (2 trials; total N = 16) using anti-CD25 antibody injection starting at 6–8 weeks of age, before immune dysregulations are known to arise.⁵⁰ Upon completion, we confirmed depletion by determining the fraction of CD25⁺ cells among splenocytes using flow cytometry, as well as inability to detect Foxp3+ cells in target tissues (prostatic lymph and mesenteric nodes) in situ using IHC (data not shown). Afterward, comparisons of CD25⁺ depleted and sham IgG-treated animals showed that there was a significant (p < 0.01) increase in prostate lesion frequency after depletion of CD25 when compared to either the sham antibody-treated or the unmanipulated age-matched Apc^Min/+ mice (Table I). Severity and progression of PIN and microinvasive carcinoma in male Min mice that had undergone CD25 depletion matched those of Apc^Min/+ mice of much more advanced age (compare in Table I: 3-month-old Apc^Min/+ undergoing CD25 depletion versus 6-month-old Apc^Min/+ mice).

The frequency of preneoplastic foci within prostate tissue arose rapidly after depletion of CD25⁺ cells (Table I) and were clearly correlated with disruption of β-catenin and E-cadherin (Fig. 5). In general, loss of the Apc gene predisposed to decline in the normal lateral epithelial cell membrane staining pattern of E-cadherin (as shown in Fig. 5e) and β-catenin (as shown in Fig. 5g). The immunohistochemical signal of E-cadherin progressively attenuated (Fig. 5f) as the lesions progressed from LGPIN to prostate adeno-carcinoma. The abnormal and intense cytoplasmic (as with PIN, Fig. 5h), or both cytoplasmic and nuclear (as with adenocarcinoma, Fig. 1f) stabilization of β-catenin typified prostate neoplasia in Apc^Min/+ mice. Gene expression analyses of prostate tissue after depletion of CD25⁺ cells also revealed increased expression of IL-6 when compared to age-matched sham-treated mice (p < 0.0001) (Fig. 4a).

Figure 5.

Immunohistochemical staining illustrates critical steps of carcinogenesis in the Min mouse model of prostate carcinoma. The aberrant immunostaining pattern observed in PIN lesions of Min mice that were depleted of CD25⁺ cells (right row) can be appreciated by comparison with normal prostate epithelia of wildtype mice (left row). (a–d) Dorsolateral prostate. (e–h) Anterior prostate. (a, b) The normal cytoplasmic expression of Apc (A) is absent in Min mouse prostate shown in b. (c, d) SMA immunohistochemistry highlights the fibromuscular sheath of prostate glands. Compare the smooth and continuous contour of normal prostate gland in c with the irregular and discontinuous (arrow-heads) sheath of prostate gland in d. (e, f) Normal lateral epithelial cell membrane staining pattern of E-cadherin in normal prostate (e) is by large lost in HGPIN shown in f. (g, h) Normal immunolabeling of β-catenin is largely restricted to adherence junctions in the prostate of wildtype mice (g). HGPIN lesions (h) instead of normal staining pattern show intense cytoplasmic staining, which indicates accumulation of β-catenin in the cytoplasm of abnormal prostate epithelial cells. Hematoxylin counterstain, DAB chromogen. Bars: 25 µm.

On the basis of these findings, we constructed a continuum of inflammation-associated prostate pathology in Apc^Min/+ mice (see Fig. 3b–3e), progressing from LGPIN to prostatic adenocarcinoma. Upon comparison between all groups, it was Rag2-deficient Apc^Min/+ mice that developed the highest frequency of advanced adenocarcinoma and invasive lesions (Table I and Fig. 3). The most advanced types of adenocarcinoma lesions were infrequent in 3-month-old Apc^Min/+ mice, even after they had undergone CD25 depletion for 30–45 days. In general, however, carcinogenesis, including high-grade PIN (p < 0.01) and microinvasive adenocarcinoma (p < 0.05), was accelerated by depletion of antiinflammatory CD25⁺ cells in Apc^Min/+ mice (Table I and Fig. 3).

Depletion of CD25⁺ cells increased frequency of intestinal ampullary tumors in the duodenal papilla of Vater

Apc^Min/+ mice are a widely applied and well-established model for intestinal carcinogenesis because of their inherent susceptibility to intestinal tumorigenesis. The Apc mutation is an early event in 80% of sporadic intestinal cancers in humans,^51,52 and the Apc^Min/+ mouse has been frequently used for study of familial intestinal adenomatous polyposis (FAP) in humans.^21,22 To examine whether the carcinogenic effect of CD25⁺ cell depletion might also affect other types of cancers, we inspected intestinal tissues of the same Apc^Min/+ mice used for prostate cancer analyses: 8 males (2 trials; total N = 16, as described earlier) with the depletion protocol starting at 6–8 weeks of age. As previously described for the study of prostate, mice had been treated with anti-CD25 antibodies for 4 weeks starting at 4–6 weeks of age, alongside sham IgG-treated age-matched controls.

As we found for prostate cancer, Apc^Min/+ mice treated with anti-CD25 antibody had an increased susceptibility to development of intestinal carcinoma. Specifically, we observed an increased frequency of ampullary cancer of the duodenal papillae of Vater in 25% (2/8) of mice treated with anti-CD25 (Fig. 6). This type of ampullary cancer was not observed even after extensive examination of untreated Apc^Min/+ mice of all ages. Lesions were comprised of variably sized duct profiles with abnormal contours that infiltrated duodenal mucosa and the muscular coat of the pancreatic duct. Ductal epithelia lacking goblet cells were pseudostratified and highly atypical. Also, we found that mice that had been treated with anti-CD25 antibody showed signs of increased inflammation; we observed a desmoplastic and inflammatory reaction, characterized mainly by neutrophils and macrophages prominent within ampullary cancer lesions. Histologic features of the tumor matched those of human patients⁵³ and also those previously described in compound mutant Apc^Min/+ and SMAD4-deficient mice.⁵⁴ Mast cells were a frequent feature in invasive intestinal lesions (Fig. 6b). We hypothesized that accelerated carcinogenesis was due to unmitigated inflammatory response and upregulation of inflammatory cytokines in intestinal tissues after depletion of CD25⁺ cells.

Figure 6.

Depletion of lymphocytes correlates with an invasive neoplastic phenotype in the small intestine of Apc^min/+mice. (a) Ampullary cancer arising from pancreatic duct epithelium in 25% (2/8) of 3-month-old Apc^Min/+ mice after undergoing depletion of CD25⁺ cells (N = 8 mice per trial). High magnification (inset in a) reveals features of the abnormal pancreatic duct epithelium including pseudostratification and nuclear pleomorphism with associated inflammation. (b) Ileum of Rag2^−/− Apc^Min/+ mouse at 6 months of age revealing submucosal and incipient tunica muscularis invasion of neoplastic glands in an adenomatous polyp. Large numbers of mast cells infiltrate the submucosa and muscle layers at the base of the polyp. The higher magnification (inset) reveals the close topographic association of mast cells with the invasive front of the tumor. (a) Hematoxylin and Eosin. (b) Toluidine Blue. Bars a and b: 250 µm; insets: 25 µm. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

We next examined whether proinflammatory cytokines were upregulated in bowel tissue after depletion of CD25⁺ cells. Upon analysis of snap frozen ileum tissue, we found that anti-CD25-treated Apc^Min/+ mice had significantly (p < 0.01) increased levels of IL-6 gene expression in tissue after depletion of CD25+ cells when compared to tissue of sham-treated age-matched controls (Fig. 4b). Gene expression of Cox-2 (Fig. 4b) was also significantly increased when compared to sham IgG-treated counterparts. These findings match data from humans showing that cancer risk in the prostate and bowel is reduced after Cox-2 inhibitor therapy.^13–15 Taken together, the results of the CD25-depletion studies support that antiinflammatory CD25⁺ T_REG cells significantly inhibit carcinogenesis in Apc^Min/+ mice.

Discussion

These studies demonstrate for the first time to our knowledge that male Apc^Min/+ mice exhibit features of neoplasia in the prostate. Untreated Apc^Min/+ mice developed age-associated prostate lesions with a high frequency of HGPIN and microadenocarcinoma resembling prostate cancer in humans. These findings were consistent with prostate carcinoma described in mice with inducible inactivation of the Apc gene.²⁵ Overt inflammatory disease was not a feature of cancer; however, factors previously linked with prostate cancer in men, i.e., IL-6, were significantly associated with development of carcinoma in this mouse model. In this model, CD4⁺ T_REG cells inhibited development of prostate carcinoma, at least in part because of downregulation of IL-6 and IL-9 cytokine expression. Not only did supplementation of purified CD4⁺ Treg cells from immune-competent donors serve to inhibit cancer development, but also the inverse was demonstrated when carcinoma frequency was increased after depletion of CD25⁺ cells in young mice. Data shown in Table I and Figure 5 after depletion of CD25⁺ cells were consistent with observations of Bruxvoort et al. showing that loss of the Apc gene leads to disruption of β-catenin and E-cadherin during development of prostatic adenocarcinoma in mice.²⁵

Among genetic factors contributing to cancer, nearly 85% of sporadic colon cancers and 15% of prostate cancers in humans have been linked with inactivation of the Apc gene. Alterations in β-catenin and the wnt signaling pathway have also been correlated with advanced prostate cancer in humans.^24,55 Development of cancer was coincident with abnormalities in β-catenin and E-cadherin in prostatic epithelia (see Results Figs. 1 and 5). These findings support data from other mouse models implicating β-catenin^25,56 and E-cadherin⁵⁷ in prostate carcinogenesis, and match earlier findings in humans.^58,59 Disruption of cytoskeletal complexes because of reduction in E-cadherin is postulated to contribute to the loss of tissue architecture⁶⁰ and has relevancy to human prostate health and disease.²⁵

Although recent data involving protective versus destructive roles for T_REG cells in cancer appear contradictory on the surface,^39–42 we postulate that function of T_REG depends on the age of the animal and context within the microenvironment of the animal. In young and immune-competent animals, T_REG cells actually serve an important protective role against cancer by maintaining immune homeostasis and constructively regulating levels of TNF-α, IL-6, IL-17 and IL-9 that when dysregulated may contribute to cancer growth. In support of this, data in Table II show that levels of TNF-α, IL-6 and IL-9 are significantly lowered when exogenous T_REG cells are added in young mice. Likewise, targeted depletion of CD25 leads to upregulation of cytokines such as IL-6 (p < 0.001) in prostate (see Fig. 4; groups undergoing CD25 cell depletion) and contribute to onset of cancer. Inability to properly regulate IL-6 was shown in other models to divert homeostatic functions of CD4⁺ cells toward a destructive Th17 pathway that may contribute to cancer growth.^48,61 We propose that this process can be interrupted to benefit the host by downregulation of inflammation (as with anti-TNF) or by supplementation with exogenous T_REG cells collected from immune competent wt hosts.^27,38,46,62 Reenforcement of the antiinflammatory arm of immunity with T_REG cells significantly (p < 0.05) reduces levels of IL-6 in mice (see Table II).

Depletion of CD25⁺ cells in 2- to 3-month-old mice was shown here to increase the rate of prostatic neoplasia that arose spontaneously at a later age of 5–6 months in untreated Apc^Min/+ mice (see Table I). One interesting question raised by these data is whether Apc^Min/+ mice have an intrinsic defect that diminishes ability of resident T_REG cells to prevent prostatic and bowel cancer. One possible explanation is that the growing intestinal tumor burden increases levels of TNF-α and IL-6 that subsequently accelerate a Th-17-driven spiral of T_REG-mediated events that subsequently contribute to cancer growth.^49,50,63 An alternative interpretation in this “chicken-and-egg” scenario is that thymic depletion, evident at 90 days of age in Apc^Min/+ mice,⁵⁰ may lead to widespread immune dysfunction predisposing to elevated IL-6 that increases cancer risk. By this 2nd way of thinking, immune impairment arises first and then allows for increased levels of TNF-α and IL-6 that impede T_REG homeostatic capabilities.⁴⁸ In either event, maintenance of immune homeostasis is disrupted in the Apc^Min/+ mice, providing an opportunity to dissect roles for cytokines that may contribute to cancer growth. In this setting, supplemental CD4⁺ T_REG cells collected from immune competent donors would, at least initially, be untainted by high levels of IL-6 within the tumor microenvironment. Our experimental depletion of CD25⁺ cells started at 30–45 days of age, preceding the natural onset of thymic involution in Apc^Min/+ mice.⁵⁰ Importantly, at 30– 45 days of age, prostatic neoplasia was otherwise nearly nonexistent in untreated male Apc^Min/+ mice (Table I). Thus, removal of CD25⁺ cells greatly hastened the onset of cancer in an animal otherwise predisposed to preneoplastic pathology.

Our depletion of CD25⁺ cells matched data of mice undergoing neonatal thymectomy that were found to subsequently develop prostatitis.⁶⁴ Likewise, cytokine expression was increased in prostate and bowel tissue after CD25 cell depletion (Fig. 4a), and conversely decreased after T_REG cell supplementation (Fig. 4a, Table II). It is important to note that the anti-CD25 treatment approach²⁹ may also deplete other nonregulatory cell CD25⁺ populations, such as activated effector T cells,⁶⁵ that may also contribute to experimental outcome; although, large numbers of activated cells bearing CD25 are unlikely in young mice lacking tumors and without targeted infections. Rapid repopulation of CD25⁺ cells after depletion has been previously described.⁶⁵ In addition, CD4⁺CD25-T_REG cells are previously shown by our lab³⁸ and others³¹ to downregulate inflammation in mice. These data do not invalidate a key role for CD25-T_REG cells in prevention of cancer. Indeed, cotransfer of CD4⁺CD45RB^loCD25⁺ cells together with CD4⁺CD45RB^hiCD25⁻cells³⁸ was previously shown to provide the most robust enhancement of longevity and protection from cancer.

The pivotal relationship between T_REG cells and IL-6 was previously highlighted in wound repair after epithelial injury in the bowel of mice.⁶² It was postulated in that model that antiinflammatory efficiency of T_REG cells dictates outcome of wound healing. The normal aging process contributes to increased susceptibility to IL-6-mediated disruptions in T_REG efficacy manifesting as delayed wound healing and dysregulated wound repair.⁶² This progression of IL-6-mediated events was found to be most evident in male animals.⁶² It is interesting that aged male Apc^Min/+ mice supplemented with CD4⁺CD25⁺ cells in our study exhibited not only decreased levels of TNF-α and IL-6 but also reduction in cancer cachexia, along with improved longevity and fecundity, when compared to age-matched untreated Apc^Min/+ mice (data not shown). The Apc^Min/+ mouse has been proposed elsewhere⁶⁶ as a model for IL-6-associated wasting disease for cancer and aging in men, highlighting broader relevancy for our findings.

Another noteworthy observation is the absence of overt inflammatory disease associated with prostate carcinoma development in this model. In humans, there is controversy over a role for inflammation in prostate cancer, although molecular pathogenesis of prostate cancer is characterized by somatic alterations of genes involved in inflammatory defenses and tissue recovery.³ Interestingly, lymphocytes and macrophages were infrequent within prostate tissue in this model. Only mast cells were clearly linked with neoplastic lesions and invasive features (see Figs. 1 and 6). The absence of inflammatory cell infiltrate in Apc^Min/+ mice matches prior data of TNF-α-dependent bowel neoplasia in mice lacking overt IBD.^27,46 These findings raise the possibility that the interplay between TNF-α, IL-6, IL-17 and T_REG cells may have broader relevancy beyond cancers that are clearly linked with overt inflammatory disease.

Inflammatory cytokine IL-9 was significantly elevated in mouse sera (Table II) and correlated with the development of prostatic neoplasia in our study. IL-9 is a growth factor linked with recruitment and activation of mast cells.^67,68 It is possible that IL-9 promoted prostate carcinogenesis in Apc^Min/+ mice through activation of mast cells; indeed, mast cells may be important for creating a prostatic environment conducive to cancer development.^12,69 Mast cells and IL-9 are previously linked with activated T_REG cells in maintenance of immune tolerance⁶⁸; thus, it is interesting that supplementation with exogenous T_REG cells serves to lower levels of IL-9 (see Table II) and also effectively suppresses carcinogenesis in Apc^Min/+ mice. TNF-α-dependent mast cell accumulations were previously shown to be an essential feature of intestinal polyposis in mice,⁴⁷ and mast cells were a frequent feature in invasive intestinal lesions in our study (see Figs. 6c and 6d). Alternatively, carcinoma may be due to a direct carcinogenic effect of IL-9 in prostate epithelia, although thus far IL-9 has proven oncogenic activities only in lymphoid cell malignancies of mice and humans.⁷⁰

Although similar prostate pathology was reported in male mice with inducible inactivation of the Apc gene,²⁵ it was unclear from the present studies whether intestinal polyposis contributed in any way to prostate cancer growth. In many murine models, bowel pathology depends on pathogenic bacterial infections that trigger a local and systemic proinflammatory carcinogenic host response. We have previously shown that infections with pathogenic intestinal bacteria may upregulate proinflammatory cytokines, such as TNF-α^34,37 and IL-6,⁶² that subsequently contribute to carcinoma development and malignancy in distant sites. Infections with pathogenic bacteria have previously been linked with prostatitis in humans⁷¹ and in rats.¹² Interestingly, in an immune-competent host, pathogenic bacteria or their toxins may also induce potent IL-10-dependent CD4⁺ T_REG cells that are highly protective against inflammation-associated cancer.^27,46,72 Preliminary data using experimental bacterial challenges reveal similar dual harmful and beneficial roles for microbiota in prostate cancer in male Apc^Min/+ mice (data not shown).

The appearance of ampullary carcinomas in the duodenal papilla of Vater after depletion of CD25⁺ cells in 25% (2/8) of young Apc^Min/+ mice is noteworthy. No such carcinomas were found after extensive examination of bowel tissues from age-matched or older Apc^Min/+ mice. In humans with FAP, ampullary tumors are a serious cause of comorbidity.⁷³ Duodenal tumors in FAP patients are often clustered in the ampullary region and frequently become malignant. Compound mutant mice with disruption of Apc and SMAD4 suffer from a similar tumor phenotype. This invasive phenotype mirrors mice lacking Tgf-β1 and SMAD3,^74–76 as well as murine recipients of dysfunctional T_REG cell,⁶² suggesting that dysregulation of Tgf-β signaling may be a shared feature in all of these pathological syndromes.

In conclusion, spontaneous prostate cancer arose in untreated male Apc^Min/+ mice with a high frequency of HGPIN and microadenocarcinoma resembling age-associated prostate cancer in humans. Factors previously found with human prostate cancer, i.e., IL-6, were significantly associated with development of carcinoma in this model. Although accumulated epidemiological and medical data support a link between chronic inflammation and prostate cancer, this is the first report providing evidence that T_REG cells reduce development of preneoplastic and neoplastic findings in the prostate. Inflammatory cytokines were significantly elevated and linked with prostate cancer development, but without the presence of overt inflammatory disease in the prostate. In particular, serum IL-6 levels were correlated with cancer in these mice and may serve as a useful biomarker in humans. Future prophylactic and therapeutic strategies may focus upon reinforcing relevant immune fortification for protection against cancer development.