Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Introduction

Prostatic acid phosphatase (PAcP; E.C.3.1.3.2) is known classically as a prostate-­specific differentiation antigen in differentiated prostate epithelia [1, 2]. Human PAcP protein is synthesized in differentiated columnar epithelia of the prostate gland [2–6]; some of which is secreted into prostatic fluid as the secretory form (sPAcP) and the rest is retained intracellularly as the cellular form (cPAcP) [7]. Recent studies have revealed a transmembrane form of PAcP (TM-PAcP) [8]. Since PAcP biochemically hydrolyzes a broad variety of small organic phosphomonoesters under acidic conditions, this enzyme is known as an AcP [3, 9–11].

PAcP levels are very low in normal circulation. In 1936, Gutman and colleagues made the seminal observation that human PAcP activity in serum is significantly increased in PCa patients, especially those with bone metastases [12]. Subsequently, Huggins and colleagues reported that the circulating PAcP activity correlates with prostate tumor progression [13]. Since then, serum PAcP has been studied extensively as a surrogate marker for the diagnosis of PCa until the availability of PSA [14].

PAcP expression is positively associated with prostate epithelial cell ­differentiation. Prior to puberty, cPAcP is expressed at a low level. In normal, well-­differentiated human prostate epithelial cells, the high level of cPAcP protein is in agreement with slow cellular growth [1, 2, 15]. In prostate adenocarcinoma cells cPAcP expression is lower than in adjacent noncancerous cells, despite elevated sPAcP activity in circulation [6, 16–18]. Studies have suggested that cPAcP acts as a tumor suppressor in PCa cells [19–21]. Several lines of evidence collectively support the concept that cPAcP functions as a histidine-dependent PTP in prostate epithelia and regulates its tumorigenicity by dephosphorylating p-Tyr of human ErbB-2 (also known as HER-2 or neu protein) at physiological pH [19, 21–24]. cPAcP also plays a critical role in regulating the cross-talk between androgens and Tyr-P signaling. In parallel, numerous studies have shown a therapeutic potential of PAcP for the treatment of PCa [19, 21, 25–28]. Hence, in this chapter, we review the structure and regulation of PAcP isoforms in prostate epithelia and the function and therapeutic role of cPAcP in human PCa.

Human Prostatic Acid Phosphatase

Physiology of PAcP

AcPs are a group of enzymes that biochemically hydrolyze phosphomonoesters optimally at acidic pH. In human cells, AcPs can be divided into at least five isoforms, including erythrocytic, lysosomal, prostatic, macrophagic, and testicular AcPs [2, 29, 30]. The mature form of PAcP is a glycoprotein consisting of two subunits of approximately 50 kDa each and is synthesized in differentiated columnar prostate epithelia [3, 5, 17, 31]. There are two forms of PAcP protein in well-­differentiated human prostate epithelia: the cellular form (cPAcP) and the secretory form (sPAcP). Recent results reveal the possible existence of a transmembrane form (TM-PAcP) [8].

The physiological level of PAcP is negligible in fetal tissue and young males. After puberty, cPAcP level can reach approximately 0.5 mg/g of wet tissue in normal, well-differentiated prostate epithelia [2, 15]. sPAcP is predominantly secreted into seminal fluid at approximately 1 mg/ml and has been used as a marker in forensic medicine [7, 14]. In PCa patients, the circulating level of sPAcP is elevated and correlates with the stage of PCa. Hence, PAcP has received much attention and has served as a surrogate marker for PCa [12–14].

Expression and Distribution of PAcP in Human Tissues

While PAcP is considered as a prostate-specific differentiation antigen, studies of PAcP expression in non-prostate tissues have yielded inconsistent results. Solin et al. [32] showed by RNA blot analysis that there is no detectable hPAcP mRNA in human liver, lung, pancreatic cancer tissue, placenta, breast cancer cells, mononuclear blood cells, or acute promyelocytic leukemia cells. Similarly, Zelivianski et al. [33] could not detect hPAcP mRNA expression in spleen, thymus, testis, ovary, small intestine, colon, or peripheral blood leukocyte by northern blotting.

On the contrary, immunologic studies demonstrate the positive reactivity of hPAcP Ab in some non-prostate cells such as leukocytes, kidney, spleen, placenta, pancreas, liver, stomach, granulocytes, neutrophils [34–38], male anal gland and in urethral gland of both sex [39], crypt epithelium of the duodenum [40], pancreatic islet cell carcinomas [41], and breast tumor cells [34]. Nevertheless, it has been proposed that the reactivity of hPAcP Ab in non-prostatic cells does not recognize the authentic hPAcP, but an immunologically cross-reactive AcP [22, 32, 42, 43], e.g., lysosomal acid phosphatase, a transmembrane phosphatase expressed in almost all tissues and cell types [38]. Supportively, purified AcPs from human spleen and lung having a similar molecular weight as hPAcP share at least one common antigenic epitope with hPAcP [22].

Recently, Graddis et al. [44] using quantitative RT-PCR reported that hPAcP is expressed at moderate to high levels in both normal and malignant prostate tissues. In non-prostate normal tissues examined, bladder cells express the highest ratio relative to prostate, though the expression level is still 50-fold lower than prostate [44]. The ratio of normal prostate PAcP mRNA to normal kidney PAcP mRNA is 178, which is comparable to the ratio of 192 reported previously [45]. The hierarchical tissue distribution of hPAcP mRNA by PCR in normal tissues is prostate >>>bladder>kidney>pancreas>cervix = testis>lung = ovary [44], which is similar to previous analyses prostate>>>placenta>kidney>testis>pancreas>small intestine = leukocytes>lung>ovary [45]. Among the tumor samples analyzed, cervical tumors express PAcP mRNA at the level similar to that seen in normal bladder; i.e., the level is less than 2 % of that in normal prostate [44]. Due to the clinical importance of hPAcP in PCa therapy and other medical applications, further experiments should clarify the identities of these proteins by determining their sequence.

Expression of PAcP in Prostate Epithelia

Immunohistochemistry staining has demonstrated that hPAcP is primarily localized in the differentiated columnar epithelial cells of prostate [37, 43, 46–48]. In situ hybridization analyses confirmed that hPAcP mRNA is detected in the glandular, ductal epithelial cells of prostate, and that the stromal cells are devoid of this mRNA [6]. An electron microscopic study showed that hPAcP is in the microvilli lining and vesicular bodies of apical cells in normal prostate [46]. The existence of human cPAcP in the cytosolic fraction has been clearly demonstrated by various biochemical approaches including sub-fractionation [9, 21]. Due to the importance of cPAcP in regulating Tyr-P signaling in PCa cells, it is imperative to clarify the subcellular localization of cPAcP where it interacts with ErbB-2 for growth regulation.

Structure of Human Prostatic Acid Phosphatase

Biochemical Characterization of Human PAcP Gene and mRNA

The human PAcP gene is located at chromosome 3q21–q23 [49] and has a size of more than 40 kb and distributed over 10 exons [50, 51]. Exon 1 encodes for the signal peptide and the first eight amino acids of the protein. Exons 2–10 encode the rest of the coding regions and 3′-untranslated region (Fig. 12.1). Several PAcP cDNA clones have been obtained in which additions, deletions, and/or substitutions of nucleotides [49, 52–54] lead to the heterogeneity of amino acid residues. Interestingly, two different signal peptide sequences have been identified [52–54]. The biological significance of these heterogeneities in PAcP sequence requires further investigations.

Fig. 12.1
figure 00121

Schematic diagram of human prostatic acid phosphatase (PAcP) gene and protein. (a) Localization of the hPAcP gene in the q-arm of the chromosome 3 (Chr 3). (b) hPAcP gene encoded by 10 exons. The number of nucleotides in each exon was noted. (c) The full length of hPAcP protein consisting of 386 amino acids with 32 amino acid signal peptides (SP). The signal peptide and the first eight amino acids were encoded by exon 1 and the rest of amino acids were coded by exons 2–10

Full size image

In human LNCaP prostate carcinoma cells, the major transcription start site is located at 50 nucleotides upstream of the gene’s ATG codon [55]. In normal differentiated human prostate epithelia, two species of PAcP mRNA are detected by Northern blot analysis with molecular sizes of 2.4 kb and 3.3 kb, essentially due to the variation in the number of Alu repeats in the 3′-noncoding sequence [32]. In prostate carcinomas, only expression of the 3.3 kb species is detected, which is lower than in noncancerous cells [32, 33, 52, 56]. The biological significance and the molecular mechanism of the loss of 2.4 kb PAcP mRNA expression in PCa cells is not clear.

Structure of the PAcP Protein

The PAcP protein is initially translated as a precursor form with 386 amino acids containing a 32-amino acid signal peptide, and the mature PAcP protein of 354 amino acids has a calculated molecular mass of 41,126 Da (Fig. 12.1). The signal peptide directs the nascent PAcP polypeptide into rough endoplasmic reticulum, which provides an environment for posttranslational modifications of PAcP protein [52, 57]. These modifications provide stability for the PAcP protein, where deletion of the signal peptide sequence from cDNA results in an extreme low level or nondetectable PAcP protein [58, 59], (Lingappa, Vishwanath and Lin, Ming-Fong, Unpublished observations). It has been proposed that the signal peptide directs secretion of PAcP protein [59]; moreover, no cPAcP protein is found intracellularly [58, 59]. These data support the concept that the function of signal peptide is primarily responsible for directing the nascent peptide via rough endoplasmic reticulum for its various post-translational modifications, including glycosylation, which stabilizes PAcP protein [58].

Sequence analysis has revealed that hPAcP protein contains three asparagine-­linked glycosylation sites (62Asn–Glu–Ser64, 188Asn–Phe–Thr190, and 301Asn–Glu–Thr303) and 6 cysteine residues forming two disulfide bonds (Cys129–340 and Cys314–319) and two free residues (Cys183 and Cys281). The glycosylation and disulfide linkages support the structural conformation and the stability of PAcP protein. Molecular sieving under native and denaturing conditions indicates that hPAcP is a dimer consisting of two subunits of similar molecular size [3, 9, 11, 60]. Analyses of the crystal structure of hPAcP reveal that each subunit has two domains: the larger domain is α/β type composed of a central seven-stranded mixed β-sheet with helices on both sides; while, the second, smaller one contains six α-helices and is formed mostly by long-chain excursions (residues 125–227) from the first domain and α-loop between residues 16–38 with no secondary structural elements [61].

PAcP Isoforms: Cellular, Secretory, and Transmembrane Forms

Recent studies have revealed that the signal peptide of hPAcP protein can direct differential biosynthetic pathways, which results in different biological functions of PAcP protein, depending on the growth environments [56], (Lingappa, Vishwanath and Lin, Ming-Fong, Unpublished observations). cPAcP and sPAcP proteins exhibit unique antigenic epitope(s), yet they share partial cross-reactivities [9]. Biochemically, they exhibit different, while overlapping pIs [9, 62].

The level of PAcP in normal circulation is negligible, while it is elevated in PCa patients and correlates with clinical progression [12, 13, 63, 64]. Thus, circulating PAcP has served as a surrogate marker for PCa detection prior to the availability of PSA [14, 65, 66]. The tumor-associated PAcP exhibits different biochemical properties from normal species and is hyper-glycosylated, including sialylation [11]. This hyper-sialylation prolongs the half-life of sPAcP protein in circulation, which contributes to the elevated PAcP level in the circulation of PCa patients despite its decreased mRNA level [11, 67]. sPAcP and cPAcP exhibit different sensitivities to endoglycosidase [Garcia-Arenas, Renee and Lin, Ming-Fong, Unpublished observations]. The biological significance of glycosylation involved in PAcP function and subcellular localization deserves further studies.

Large amounts of PAcP protein are found in normal prostate tissues by immunohistochemistry staining and biochemical analysis of tissue homogenates [9, 62, 68, 69]. The presence of cytosolic cPAcP protein is further validated by sub-­fractionation approaches, including ultracentrifugation [21]. Additionally, immunocytochemistry staining of intact, nonpermeabilized LNCaP PCa cells shows no significant staining of hPAcP. On the contrary, intensive staining is seen in the cytosolic area of permeabilized cells with higher intensity of staining in higher density cells [21, 70, 71]. Therefore, cPAcP is localized intracellularly and has served as a useful marker for identifying the prostate origin of metastatic cancer [1, 2, 14]. In prostate carcinomas, intracellular PAcP protein level decreases, correlating with PCa progression [6, 16]. The decreased protein level is at least in part by the decrease of mRNA [32]. Because of the importance of cPAcP in tumorigenicity and androgen sensitivity of PCa cells, the molecular structure of cPAcP relating to sPAcP should be further investigated.

Quintero et al. [8] reported the existence of a PAcP-spliced variant, which is a type І transmembrane protein in many mouse tissues. However, its expression profile in human tissues other than prostate is not yet known. Immunostaining with an anti-PAcP Ab showed PAcP expression in human skeletal muscle cells. They further demonstrated that PAcP colocalizes with lysosomal associated membrane protein 2. None-the-less, expression of PAcP is extremely low in skeletal muscle cells, and importantly, exhibits no lysosomal localization [1, 2, 38]. These observations have raised a concern that the staining may be due to a partial cross-reactivity of polyclonal Ab [22]. It has been further proposed that the active site of TM-PAcP is localized and functioning extracellularly [20]. Biochemical analyses have demonstrated that no phosphatase activity is significantly detected when small organophosphate substrates are incubated with intact LNCaP cells. These results collectively suggest that in prostate epithelia only a very low amount, if any, of the active domain of hPAcP faces extracellularly. While the decreased expression of classical PAcP mRNA and protein correlates with PCa progression, the expression of TM-PAcP mRNA is not significantly changed, indicating that TM-PAcP is not involved in prostate carcinogenesis [8]. A functional characterization of TM-PAcP in prostate epithelia utilizing its specific monoclonal Ab is required for further investigation.

Biological Function of Prostatic Acid Phosphatase Isozymes

Cellular PAcP (cPAcP) as a Growth Regulator and Tumor Suppressor in Prostate Cancer Cells

Expression of PAcP is correlated with the differentiation of normal prostate cells. In PCa, cPAcP levels of both mRNA and protein are decreased, compared to nonmalignant cells [6, 16, 63, 69, 72–74]. It has been proposed that normal prostate epithelia, having a low level of cPAcP, are at high risk of carcinogenesis [72]. In human PCa cell lines, cPAcP levels are correlating inversely with cell growth rates [16, 19, 21, 56]. Results of increased PAcP expression by cDNA transfection and decreased cPAcP expression by antisense cDNA or shRNA knockdown in PCa cells validate the growth regulatory role of cPAcP in PCa cells [17, 21, 71].

Several studies of clinical archival specimens showed that decreased cPAcP expression correlates with increased tumorigenicity and cancer progression [2, 6, 16, 17, 75]. Conversely, expression of cPAcP correlates with decreased tumorigenicity of PCa cells in xenograft animals [16]. Igawa et al. [19] further explored the direct tumor suppression activity of cPAcP in a xenograft animal model. Importantly, in a PAcP-knockout mouse model, the prostate develops adenocarcinomas [20]. The data collectively demonstrate that cPAcP expression suppresses the growth and tumorigenicity of PCa cells. This provides an explanation for the clinical phenomenon that the expression levels of cPAcP inversely correlate with the stage of PCa as well as its advanced progression under androgen deprivation therapy (ADT) [6, 72].

Further studies have revealed that cPAcP is involved in regulating androgen-­stimulated proliferation of human PCa cells [76–78]. Expression of cPAcP by cDNA transfection in androgen receptor (AR)-positive, androgen-independent LNCaP C-81 cells results in restoring the androgen sensitivity, i.e., the cell growth is sensitive to androgen treatment [16, 78]. Conversely, an androgen-independent phenotype is obtained by knockdown of cPAcP expression by shRNA in androgen-­senstitive LNCaP C-33 cells [21]. Thus, androgen-induced proliferation of prostate epithelia is at least in part due to an androgen effect on decreasing cPAcP activity [16, 71, 78]. The data taken together support the concept that cPAcP plays a critical role in regulating the basal as well as the androgen-stimulated proliferation of human PCa cells.

Transmembrane PAcP (TM-PAcP) as an Analgesic in Mice

While TM-PAcP is detected in several tissues from mice, thus far it has been detected only in prostates of humans [8]. Interestingly, PAcP knock-out (PAcP−/−) mice display enhanced noxious thermal sensitivity and sensitization through inactivation of adenosine A1 receptor (A1R) and phospholipase C-mediated elevation of phosphatidylinositol 4,5-bisphosphate (PIP2) [20, 79, 80]. It has been proposed that TM-PAcP is the fluoride-resistant acid phosphatase and functions as phosphoadenosine phosphatase upon expression in human embryonic kidney HEK293 cells [20]. Considering the long-lasting antinocieptive effect of secretory PAcP in the naïve mouse model, it has been proposed that recombinant PAcP can be used to treat chronic pain [20, 81, 82]. However, it is not known if TM-PAcP is indeed expressed in the corresponding human tissues and cells. The expression profile of TM-PAcP in human tissues and cells should be analyzed.

Role of Secretory PAcP (sPAcP) in Sperm Motility, HIV Transmission, and Forensic Marker

Due to the large amount of sPAcP protein in seminal fluid [83], it has been suggested that PAcP plays a physiological role in fertility [84] and may affect the motility of sperm [20, 85]. Dave and Rindani [86] observed that phosphatase activity is maximal in azoospermic men, and that this activity is decreased as the sperm number (concentration) increases. However, there is no significant difference in PAcP activity in seminal plasma between normal and vasectomized patients [87].

Importantly, a proteolytically cleaved PAcP peptide, PAPf39, which forms amyloid fibrils called Semen-derived Enhancer of Viral Infection (SEVI), can enhance the HIV’s ability to infect human cells by five orders of magnitude [88]. This observation is in agreement with a previous report that HIV replication component is detected at a tenfold higher concentration in seminal fluid than in blood, even in the presence of an antiretroviral drug [89]. Alternatively, PAcP may increase the pH of the vagina [90]. Thus, this postintercourse neutralization of pH may allow a female-­to-male transmission of HIV [91]. The role of sPAcP in sexually transmitted diseases requires further investigation.

It should be noted that secretory PAcP can also serve as a forensic marker. Due to the large quantity of sPAcP in seminal fluid, and also due to its specificity of expression in males, secretory PAcP was investigated and served as a surrogate marker in forensic medicine for sexual assault [92]. This was supported by the observations that elevated levels of AcP activity persist in the vaginal pool after sexual intercourse and in semen stains [93–100]. Nevertheless, there are some concerns, for example, the potential cross reactivity with other acid phosphatases and the instability of its enzymatic activity. The subsequent advancements including the enhancement in enzymatic assays and the development of detection methods such as ELISA, counter-immunoelectrophoresis, and radioimmunoassays made PAcP as a useful surrogate marker in forensic medicine [101–104]. Nevertheless, the discovery of PSA (also called γ-seminoprotein; γ-SM) in seminal fluid, due to its long-­term stability (identifiable for as long as 1 year) and easy identification, replaced PAcP as the forensic marker [14, 105–108].

Biochemical Characterization of Human Prostatic Acid Phosphatase Protein

PAcP: A Histidine-Dependent Neutral Protein Tyrosine Phosphatase

PAcP, which belongs to the histidine phosphatase superfamily, uses an active-site histidine in catalyzing the transfer of a phosphoryl group from phosphomonoesters to water at acidic pH. The copurification of cPAcP protein with the majority of PTP activity from noncancerous prostate tissue and purified PAcP protein from seminal fluid and tissue exhibiting the endogenous PTP activity together indicate that PAcP is an authentic PTP [109, 110]. Biochemical characterizations show that PAcP dephosphorylates p-Tyr of EGFR with a neutral pH optimum, supporting the concept that cPAcP can indeed function as a neutral PTP in cells [111]. Further, in PAcP K/O mice, Tyr-P activity is increased in prostate cells, suggesting that intracellular cPAcP functions as PTP [20, 112]. Several lines of evidence together support the notion that cPAcP indeed functions as a neutral PTP in prostate epithelia [17, 110, 113].

Structural analyses of PAcP protein reveal that it contains neither the PTP signature motif, C(X)5R(S/T), nor the extended active site signature sequence for the dual-specificity phosphatases, VXVHCXXGXXRS(X)5AY(L/I)M [52, 57, 114, 115]. Chemical titration experiments revealed that PAcP has two reactive sulfhydryl groups [116]. It was hypothesized that Cys183 is essential for the PTP activity of PAcP. Nevertheless, studies by site-directed mutagenesis determined that neither Cys183 nor Cys281 plays a role in the phosphatase enzymatic activity [58].

Covalent modifications and phosphoenzyme trapping studies revealed that PAcP contains histidine and carboxylic acid residues in the active site [117–119]. The role of His12 in both AcP and PTP activities is clearly evidenced by site-directed mutagenesis [58]. The His12 imidazole ring provides a pair of electrons for nucleophilic attack to the phosphate group. Cooperatively, Asp258 donates a proton from its carboxyl group to the substrate resulting in the formation of the phosphoenzyme intermediate and the liberation of dephosphorylated substrate. Additionally, Asp258 might also stabilize the phospho-His12 intermediate. Subsequently, the nucleophilic attack of the phosphoenzyme intermediate occurs through a water molecule to release the phosphate group and to return a proton to the Asp258 carboxyl group [116, 120, 121]. The data collectively from chemical modification, site-directed mutagenesis, and X-ray crystallographic approaches suggest the importance of His12 and Asp258 in both AcP and PTP activity of PAcP protein [58, 61, 116, 120–122]. These results further support the concept that PAcP represents a novel histidine-dependent PTP, which uses the same active site as well as the catalytic mechanism of AcP to execute its PTP activity.

ErbB-2/HER-2/neu: An Intracellular Substrate of cPAcPin Prostate Epithelia

Several lines of evidence validate cPAcP as an authentic PTP [17, 21, 109–111, 113, 121, 123]. In PCa cells, cPAcP activity inversely correlates with the p-Tyr level of a 185 kDa protein [23, 111]. The incorporation of purified PAcP protein into PAcP-­null DU145 PCa cells results in decreased Tyr-P of 185 kDa protein [23]. The 185 kDa protein was identified to be the ErbB-2 [24]. The notion of cPAcP dephosphorylating ErbB-2 Tyr-P is further supported by ectopic expression of the wild-­type cPAcP but not its phosphatase-inactive mutant in PCa cells [16, 58]. Conversely, small interfering RNA or antisense-mediated PAcP knockdown in LNCaP cells results in increased ErbB-2 Tyr-P and subsequently cell proliferation [17, 21]. Additionally, an intratumoral injection of the wild type PAcP, but not phosphatase-­inactive mutant, cDNA expression vector in xenograft tumors results in decreased ErbB-2 Tyr-P as well as tumorigenicity [19].

Transient expression of PAcP in PAcP-null PCa cells is associated with decreased Tyr1221/2 and Tyr1248 phosphorylation at ErbB-2 and reduced Tyr-P of p52Shc and cell growth [21]. Knockdown of endogenous PAcP expression by shRNA is associated with elevated Tyr-P of ErbB-2 at Tyr1221/2 as well as Tyr1248 and activation of downstream signaling, including Akt, STAT-3, and STAT-5 [21]. Importantly, reciprocal co-immunoprecipitation analyses showed an interaction between PAcP and ErbB-2 in the same complex under a nonpermissive growth condition [21]. This interaction by co-immunoprecipitation was decreased upon growth stimulation [124]. Thus, the effect of cPAcP on downregulation of PCa cell growth is at least in part through dephosphorylating the p-Tyr of ErbB-2 protein in those cells [16, 21, 24, 58, 124]. ErbB-2 serves as an in vivo substrate of cPAcP in PCa cells [17, 19, 21, 120, 124] (Fig. 12.2).

Fig. 12.2
figure 00122

Schematic representation of cPAcP interaction with ErbB-2 in prostate cancer cells. Progression of androgen-sensitive prostate cancer cells towards androgen independence is accompanied by early decrease/loss of cPAcP expression in prostate cancer cells results in hyperphosphorylation of HER-2 on tyrosine residues including Y1221/2 and Y1248 leading to androgen-independent cellular proliferation. Activated HER-2 can transduce its signals via p52Shc (blocked by dominant-negative (DN) HER-2 cDNA transfection or HER-2 inhibitors, AG825, AG879) to activate the downstream pathway (blocked by p52Shc Y317F mutant cDNA transfection or MEK inhibitors)

Full size image

The cPAcP dephosphorylation model indicates that dimeric cPAcP dephosphorylates two autophosphorylated residues on an activated receptor simultaneously because the presence of a second phosphorylated tyrosyl residue at the C terminus of ErbB-2 can enhance the binding affinity considerably [120]. Phosphopeptide-­binding analyses showed that cPAcP has the most favorable binding energy toward the synthesized peptide DNLpYYWD, corresponding to Tyr1221/2 phosphorylation of ErbB-2, with the possibility of acting on Tyr1248 as the additional site [120]. This is supported by kinetic studies on ErbB-2 activation that phosphorylation of Tyr1221/2 is elevated prior to Tyr1248 activation in PAcP-knockdown PCa cells [21]. Alternatively, due to the close proximity of Tyr1221/2 and Tyr1248, elevated phosphorylation on Tyr1248 in PAcP-knockdown cells may be secondary to the removal of cPAcP from Tyr1221/2. Further experiments should clarify the molecular mechanism.

ErbB-2 Signaling

Results of several studies clearly support the notion that ErbB-2 plays a critical role in PCa progression despite the fact that ErbB-2 gene is not amplified nor ErbB-2 protein is elevated in most carcinomas. It should be noted that ErbB-2-specific activity is increased as shown by increased overall Tyr-P and downstream signaling in AI PCa cells, higher than that in corresponding AS cells, and this increased Tyr-P is inversely correlated with cPAcP activity [21, 23, 24, 77]. Evidently, knockdown of cPAcP expression by antisense cDNA and siRNA in AS PCa cells leads to increased Tyr-P of ErbB-2, activation of downstream signaling and increased cell growth both in regular medium and in steroid-reduced condition [17, 21]. Thus, increased ErbB-2 protein-specific activity contributes to advanced CRPCa progression primarily by phosphorylation regulation, including ERK/MAPK, Akt as well as STAT-3 and STAT-5 activation, which leads to advanced PCa cell survival, proliferation, and PSA production under androgen-deprived conditions. In summary, the interaction between cPAcP and ErbB-2 is involved in controlling the basal as well as the androgen-stimulated proliferation of human PCa cells [21, 78]. Aberrant regulation of this interaction can lead to CRPCa progression under ADT.

Regulation of PAcP Expression

The expression of PAcP protein is regulated by multi factors at different levels. Due to the importance of cPAcP protein as a growth regulator in PCa cells, it is vital to delineate its regulatory mechanism for potential clinical applications.

Androgen Regulation of PAcP Expression and Secretion in Prostatic Carcinoma Cells

Since PAcP expression correlates with the differentiation of prostate epithelia after puberty, it has been proposed that the expression and secretion of PAcP is regulated by androgens [2]. The stimulated secretion of PAcP has served as a hallmark of androgen action in prostate epithelial cells for over 6 decades, and the circulating PAcP in PCa patients has been used as a surrogate marker in ADT for about five decades [1, 2, 13].

Lin and Garcia-Arenas [125] made the seminal observation that depending on the cultured cell density, DHT can upregulate or downregulate PAcP mRNA levels in LNCaP cells. These results clarify the inconsistent reports of the opposite regulations of PAcP mRNA by DHT [54, 126]. Nuclear run-on experiments showed that DHT regulation of PAcP expression can occur at the transcriptional level [33]. Further investigation is needed to delineate the molecular mechanism by which cell density modulates androgen regulation of PAcP mRNA level and to examine whether androgens regulate the stability of PAcP mRNA in addition to the transcriptional rate [33].

Sequence analyses revealed that the human PAcP gene promoter DNA within 3 kb upstream of the coding region lacks the canonical TATA box and the GC box, where there are five putative AREs [55, 127]. In prostate carcinoma cells, although the PSA promoter is regulated by androgens [128–130], PAcP expression is not androgen dependent [54, 126, 131]. Utilizing AR-negative, androgen-independent PC-3 and DU145 PCa cell lines, in the absence of androgen receptor or the addition of androgen, the PAcP promoter is highly active as determined by reporter assay. These results demonstrate that in PCa cells, the PAcP gene is regulated in an androgen-­independent, responsive manner [70, 126, 132]. Supportively, secretion of PAcP protein is observed when LNCaP cells are cultured in media supplemented with charcoal-stripped FBS [133] or dialyzed FBS [77, 134] in which steroids and growth factors are significantly reduced. Further, PAcP protein expression is at a high level in cells with serum-free media in the absence of added DHT and even higher than with regular medium containing FBS [54, 77]. These data together support the notion that in LNCaP cells, androgen-stimulated PAcP secretion is via two regulatory pathways: increasing levels of secretory PAcP mRNA and at the same time, promoting the secretory pathway [54].

It has been shown that JFC1 [also designated as synaptotagmin-like protein (slp1)], a Rab27a and PtdIns(3,4,5)P3-binding protein, can modulate androgen-­stimulated secretion of sPAcP in LNCaP cells [135–137]. In parallel, phosphoinositide 3-kinase (PI3K) also plays a critical role in regulating the exocytosis of sPAcP [135]. Additionally, PKC activator and androgens both increase PAcP secretion, which are blocked by PKC inhibitors [138]. These results together suggest that sPAcP secretion is mediated by a regulatory process including Rab27a, PI3K, and PKC, differing from PSA secretion primarily in response to androgens [135, 138, 139].

Effects of Growth Factors on PAcP Expression

Factors other than androgens can regulate PAcP gene expression in LNCaP cells. Interestingly, the effect of EGF is more pronounced than DHT in determining PAcP mRNA expression and can potentiate the downregulation by androgens, but there is no added effect by androgens on EGF suppression [54, 70, 140]. EGF treatment also results in decreased cPAcP activity, which may be due to its phosphorylation inactivation by EGFR or oxidation inactivation [124]. While both EGF and TGF-α can bind to the EGF receptor [141, 142], TGF-α has a less inhibitory effect than EGF on reducing hPAcP mRNA [71]. On the other hand, TGF-β1, which is inhibitory to normal prostatic epithelial cells [143, 144], upregulates the expression of PAcP mRNA [140]. Due to the importance of PAcP in clinical applications, further experiments should clarify the regulation of PAcP in prostate epithelia.

In the presence of androgens, the expression of growth factors, e.g., EGF, TGF-­α, TGF-β1, and TGF-β3, are also modulated. Androgens and growth factors and their receptors, represent cross-talk at several levels. Additional experiments are required for elucidating the role of this cross-talk on the regulation of PAcP gene expression.

Epigenetic Regulation of PAcP Expression in Prostate Carcinoma Cells

DNA methylation and histone modification, two common epigenetic mechanisms, play vital roles in regulating PCa cell growth and metastasis [145]. Histone modification, primarily by acetylation and deacetylation, leads to altered gene expression by changing chromosome structure and the level of gene transcription. Histone deacetylase (HDAC) activity is enhanced and upregulated in PCa and other carcinomas [146–148]. Hence HDAC is recognized as a promising target for cancer therapy, although the exact role of specific HDACs in the pathophysiology of PCa is still not well understood.

HDAC inhibitors have been shown to induce PCa cell growth arrest, differentiation, and apoptosis. cPAcP functions as a tumor suppressor in prostate carcinomas, and its decreased expression correlates with PCa progression. However, the molecular mechanism of its reduced expression in PCa remains an enigma [16, 19, 21]. Importantly, HDAC inhibitors, including sodium butyrate, trichostatin A (TSA), and valproic acid (VPA), suppress the growth of PCa cells and concurrently, cPAcP mRNA and protein expression are increased and ErbB-2 Tyr-P is decreased [149]. Conversely, knockdown of cPAcP expression by shRNA reduces the efficacy of HDAC inhibitor-induced growth suppression. Therefore, PAcP is involved in HDAC inhibitor-induced growth suppression and functions as a tumor suppressor gene in regulating PCa progression and metastasis (Fig. 12.3). Importantly, HDAC inhibitor-­treated PCa cells increase their androgen responsiveness [149]. Understanding the regulation of cPAcP expression by HDACs may lead to improved CRPCa therapy by HDAC inhibitors.

Fig. 12.3
figure 00123

Epigenetic regulation of prostate cancer cells. We propose that in PCa cells, upregulated histone deacetylases (HDACs) downregulate PAcP expression. This PAcP suppression leads to aberrant activation of ErbB2/HER-2/neu by Tyr-P followed by ERK/MAPK and Akt activation, leading to cell survival, proliferation, adhesion, and migration. Conversely, HDACs inhibitors (HDACi) restore cPAcP expression. This restored cPAcP dephosphorylates ErbB-2, which leads to inhibit prostate cancer progression and metastasis by p38 and JNK activation

Full size image

Prostatic Acid Phosphatase as a Therapeutic Agent as Well as a Target for Prostate Cancer Treatment

While the majority of patients with metastatic prostate cancer have an initial response to ADT, most patients will eventually relapse with castration-resistant tumors. With the limited efficacy of conventional therapeutic approaches and also with significant morbidities of surgical and radiation treatments in advanced PCa, other avenues for treating advanced prostate carcinoma are actively under investigation.

PAcP Per Se as a Therapeutic Agent

Several lines of biochemical evidence have demonstrated that cPAcP functions as a tumor suppressor. A single intratumoral injection of an expression vector encoding the wild-type PAcP protein into xenograft tumors results in the suppression of tumor growth and progression [19]. In PAcP-knockout mice, the prostate develops carcinomas in situ, indicating that cPAcP functions as a tumor suppressor [20]. In parallel, cPAcP plays a critical role in HDAC inhibitor-induced PCa cell growth suppression [149]. Importantly, the HDAC inhibitor-treated PCa cells exhibit an increase in androgen responsiveness, suggesting that intermittent treatment with HDAC inhibitors may prolong the duration of ADT [149]. Thus, the restoration of cPAcP expression in PCa cells may provide a novel avenue for treating CRPCa.

PAcP as an Antigen for Immunotherapy of Prostate Cancer

Immunotherapeutic vaccines induce an antitumor response [150] by targeting tumor-associated antigens (TAAs) or by disrupting molecular pathways that promotes tumor growth [151, 152]. Therefore, the primary goal of immunotherapy is to activate the effector T cells that can migrate to the developing tumors and facilitate the damage of individual cancer cells.

Prostate cells express several specific biomarkers, including PSA, PAcP, and prostate-specific membrane antigen, which serve as TAAs and can serve as immunogens. Previous studies demonstrated naturally occurring PAcP-specific binding IgG in human serum [65], the induction of a destructive prostatitis by PAcP-specific CTLs in rodents [153], and T-helper cell responses in men with PCa [154]. Studies have also identified Abs and circulating T cells against TAAs in PCa patients [155, 156]. These findings suggest that T-cells can break the tolerance and induce an immune response against tumor cells [155, 156]. These phenomena collectively indicate that an immune environment capable of supporting antigen-specific CTL may exist in vivo [154]. Small et al. [157] observed that dendritic cells loaded with an engineered antigen–cytokine fusion protein consisting of PAcP and granulocyte macrophage colony-stimulating factor (GM-CSF) are capable of inducing a potent cellular immune response in vivo to rodent tissues and tumors that express PAcP [157]. Subsequently, a dendritic cell product consisting of autologous dendritic cells loaded with the human PAcP–GM-CSF fusion protein was developed. It is hypothesized that when the vaccine is infused into the patient, the activated antigen-­presenting cells (APCs) displaying the fusion protein will induce an immune response against the TAA. Phase I/II clinical trials with a dendritic cell-based PAcP vaccine in CRPCa patients led to a greater than 50 % decrease in PSA [157, 158]. The study showed that all patients developed specific immune responses to the recombinant fusion protein, and 38 % developed immune responses to PAcP. The time to disease progression correlated with the development of an immune response to PAcP and with the dose of dendritic cells received. There were minimal side effects of the therapy.

Conclusion and Perspectives

Until the availability of PSA, circulating PAcP activity has served as a surrogate marker for diagnosing PCa and also has been used to monitor the efficacy of androgen deprivation therapy in treating PCa [14, 66]. In contrast, expression of PAcP and its intracellular level (cPAcP) is diminished in prostate carcinomas. Recent advances emphasize that cPAcP is an authentic protein tyrosine phosphatase and functions as a negative growth regulator in PCa cells. Importantly, cPAcP represents a novel subfamily of PTP super family [121]. The expression of PAcP is regulated at different levels as well as by different factors including androgens and growth factors in prostate carcinoma cells. The androgenic regulation of PAcP expression and secretion has been known to be a hallmark of androgen action for over six decades [2]. Nevertheless, the results of molecular studies demonstrate that the promoter activity of the PAcP gene is regulated in an androgen-independent manner [54, 126, 132].

Several lines of evidence support the importance of cPAcP enzyme in regulating PCa cell proliferation, particularly during the castration-resistant progression, at least in part by dephosphorylating p-Tyr of ErbB-2 intracellularly. It has been speculated that PAcP can also function as a phospholipid phosphatase because in cPAcP-­knockdown cells, phospholipid activity is enhanced and also PAcP has an open active domain [80, 121]. Additionally, PAcP expression is in part regulated by epigenetic mechanism including histone acetylation and possibly methylation. These emerging data support PAcP as a potential therapeutic target for advanced PCa. The recent clinical immunotherapy trial with PAcP protein as a vaccine is promising. Further studies are needed to improve the clinical efficacy, for example, by effective intracellular delivery of antigenic peptides into dendritic cells.

Taken together, the data clearly show that cPAcP functions as an authentic tumor suppressor in PCa. Due to the importance of the PAcP gene in prostate carcinogenesis, investigation of the basic biochemistry and molecular biology of cPAcP including its interaction with other oncogenic proteins should provide valuable insights into its potential therapeutic applications.