Deciphering the Role of Protein Kinase D1 (PKD1) in Cellular Proliferation

PKD1, also called PKCμ, is a serine/threonine kinase that belongs to the PKD family, a subgroup of the calcium/calmodulin-dependent kinase (CAMK) family (1). PKD1 is a 912 amino acid residue protein with an apparent molecular weight of 115 kDa that contains a carboxy-terminus catalytic domain and a regulatory domain at the amino-terminus. The latter regulates the catalytic activity of PKD1 by maintaining the protein in an inactive state through an autoinhibitory mechanism exerted toward the catalytic domain (2). PKD1 can be activated by a wide variety of extracellular stimuli including growth factors, vasoactive peptides, chemokines, neuropeptides, phorbol esters, and others. To date, the best characterized signaling pathway responsible for the activation of PKD1 involves the activation of phospholipases Cβ or γ (PLC_β or PLCγ; ref. 3). These proteins synthesize inositol-triphosphate (IP₃) and diaglycerol (DAG), which allows the activation of several protein kinase C (PKC) isoforms and their recruitment close to PKD1. Once nearby, PKCs phosphorylate PKD1 onto two serine residues (738 and 742, or 744 and 748, human or murine numbering, respectively) localized in its activation loop leading to the stimulation of the catalytic domain and its autophosphorylation onto its serine 910 (or 916 for murine PKD1) residue (4). Activated PKD1 thus translocates into different cellular compartments modulating its targets. The wide diversity of its substrates makes PKD1 a main actor in several biological processes such as cell proliferation, migration, invasion, apoptosis, angiogenesis, cardiac contraction, and immune regulation (5). In this context, its dysregulation (over- or underexpression) was shown to be associated to diverse pathologies such as inflammation, cardiac hypertrophy, and cancer (6). However, it remains largely unknown what regulates PKD1 gene (prkd1) expression in tumors. PKD1 gene promoter was shown to be either activated by the oncogenic KRas–NFκB pathway increasing the expression of PKD1 in pancreatic cancer cells (7) or inhibited by β-catenin in prostate cancer (8). It was also shown to be the target of epigenetic methylation decreasing PKD1 expression in some breast tumor cells (9–11). These different molecular mechanisms lead to tumor tissue–specific PKD1 mRNA expression profiles. According to TCGA data, PKD1 mRNAs are mostly expressed in prostate cancer and melanoma (Fig. 1). Also, the data relative to 11 studies [Breast Invasive Carcinoma, Colorectal Adenocarcinoma, Head and Neck Squamous Cell Carcinoma, Kidney Renal Clear Cell Carcinoma, Kidney Renal Papillary Cell Carcinoma, Lung Adenocarcinoma, Lung Squamous Cell Carcinoma, Pancreatic Adenocarcinoma, Prostate Adenocarcinoma, Skin Cutaneous Melanoma, Stomach Adenocarcinoma (TCGA, PanCancer Atlas)], has shown PKD1 mRNA levels to be upregulated by 1.8% (pancreas) to 18% (lung adenocarcinoma) with a mean value of 7.8% (additional files 1 to 8: Supplementary Figs. S1–S8). Moreover, analysis of the prkd1 gene reveals that only 4% of the tumors analyzed (206 patients over 5,615) carry a mutation or a copy number alteration in the 11 abovementioned TCGA studies (Fig. 2). Taken together, these results suggest, at least for tumors with the lowest upregulated values, that a dysregulated PKD1 activity may certainly play a more significant role in tumor progression than its gene overexpression or amplification.

Although PKD1 seems to play an essential role in oncogenesis and is activated by a large variety of stimuli, especially by many growth factors, it has a complex relationship with cell proliferation both in normal and cancer cells. In fact, the specificity of the role of PKD1 with regards to cell proliferation depends not only on the tissue type but also on the phenotype (normal vs. tumor) because PKD1 has been described to be either proproliferative or antiproliferative (Table 1). Moreover, this complexity drastically increases when, for a given cell type, some controversial data exist. However, increasing data described PKD1 activity as affecting tumor behavior both in vitro and in vivo through its ability to regulate cell proliferation making PKD1 a putative pertinent pharmacologic target in oncology. In this context, it becomes obviously and urgently crucial to have a clear knowledge of its role in cell proliferation. Therefore, this review aims to list the major data existing to date concerning the pro- or antiproliferative effects of PKD1 and tries to bring elements of discussion to explain, when necessary, potentially contradictory results.

Angiogenesis is the process by which new blood vessels are formed and is of pivotal importance in processes such as wound healing and embryonic vascular development (12). It also plays a fundamental role in tumor growth and metastasis (13). It provides tumors with oxygen and nutrients, crucial for their development and also helps in discarding tumor metabolites (13). Inhibition of angiogenesis has thus been regarded as a valuable new approach to cancer therapy (14). A number of stimulators and inhibitors regulate angiogenesis (15). In the case of PKD1, a consensus exists with regard to its proangiogenic and thus proproliferative role whatever the cell model [endothelial progenitor cells (EPC) or cell lines] or the species (human or even zebrafish). VEGF, a major component of angiogenesis under both physiologic and pathologic conditions, induces the phosphorylation of PKD1 in human umbilical vein endothelial cells (HUVEC), bovine aortic endothelial cells (BAEC) (16), and EPCs (17). This occurs within minutes upon binding of VEGF to its receptor, VEGFR2, through a PLCγ/PKCα-dependent signaling pathway (16). VEGF-stimulated ERK1/2 phosphorylation and DNA synthesis in HUVEC (16), and VEGF-induced microvessels sprouting from mouse aortic rings (18) were markedly inhibited by PKD1 knockdown and PKD1 kinase–negative mutant expression, respectively, making PKD1 a proproliferative protein in endothelial cells. As previously mentioned, histone deacetylases (HDAC) help control gene expression by regulating the acetylation state of the chromatin. VEGF stimulates HDAC7 and HDAC5 phosphorylation and their nucleocytoplasmic shuttling through a PKD1-dependent signaling pathway (17–19). Once phosphorylated and in the cytosolic compartment, HDAC7 is localized away from its substrates promoting gene expression leading to cell proliferation. Moreover, PKD1 was shown to regulate the VEGF-induced expression of metalloproteinases such as MT1-MMP (13), whose gene expression is implicated in angiogenesis in vivo (20). Thus, it is very interesting to note that PKD1 has a particular role with regard to VEGF because, on one hand, PKD1 is an actor of the VEGF signaling pathway and, on the other hand, PKD1 regulates VEGF secretion as shown in pancreatic cancer cells (21). Thus, in a tumor context, PKD1 could allow the activation of a self-sustained loop allowing the formation of new vessels promoting the progression of the tumor (22).

The tumor microenvironment (TME) of cancer cells has been found to be a key determinant in tumor progression and metastasis and has thus been gaining increasing interest in cancer research. Fibroblasts, a major component of the TME, are responsible for the synthesis, deposition, and remodeling of the extracellular matrix and are a source of paracrine growth factors that regulate the growth of cancer cells (23). PKD1 is portrayed as a proproliferative protein in fibroblasts. Overexpression of PKD1 in murine Swiss-3T3 cells enhances the proliferative response to G-protein–coupled receptor agonists and to phorbol esters. In fact, treatment with angiotensin, bombesin or phorbol 12,13-dibutyrate (PDBu) led to PKD1 phosphorylation onto its 744, 748, and 916 serine residues (murine numbering). Furthermore, PKD1 overexpression potentiated neuropeptide-induced mitogenesis (24) probably through an increased duration of the ERK signaling pathway characterized by a significant increase in the phosphorylation of FAK and RSK, and by the accumulation of the early gene c-Fos (25).

Skin cancer is characterized by an abnormal growth of skin cells. Depending of the skin cell type involved, two major categories of cancers were defined: basal and squamous cell skin cancer and melanoma (26). Basal and squamous cell cancers are the most common types and are mainly caused by UV exposure, thus usually developing on body parts exposed to sunlight (27). On the other hand, melanoma develops from melanocytes and has less well-defined origins. It can also be caused by UV light but, unlike basal and squamous cell cancers, can also develop on body parts unexposed to sunlight (28). Melanoma is more likely to form metastasis in other tissues making it usually more aggressive than basal and squamous cell carcinomas (29).

PKD1 expression was first described in 1999 in mouse epidermis and positively correlated with cell proliferation (30). Keratinocytes proliferation was decreased after treatment with PKD1 pharmacologic inhibitor, Goedecke 6976 (Gö6976), and enhanced in PKD1-overexpressing cells (30). Moreover, carcinoma (but not papilloma) from a two-stage carcinogenesis induced mouse model, using first 7,12-dimethylbenz[a]anthracene (DMBA) as initiator, and then tetradecanoylphorbol-13-acetate (TPA) as promoter, expressed high levels of PKD1 (30) and were strongly impaired for their development after peracetylated EGCG (AcEGCG)-induced PKD1 inhibition (31). According to these data, murine PKD1 was shown to be mostly expressed in the basal, proliferative, layer of the epidermis and, although present, less expressed in suprabasal layers (32). Consistently, overexpressing PKD1 in primary keratinocytes stimulates keratin 5 (proproliferative marker), but inhibits involucrin (prodifferentiative marker) promoter activities, respectively (32). Moreover, genetic PKD1 depletion not only inhibited cell proliferation but also strongly potentiated the calcium-induced expression of late, intermediate, and early differentiation markers of mouse keratinocytes such as loricrin, involucrin, and keratin 10 (33). Finally, further evidence of a proproliferative role of PKD1 came from adenovirus-transfected primary mouse epidermal keratinocytes showing that a constitutively active PKD1 mutant significantly increased DNA synthesis. In contrast, a dominant-negative PKD1 mutant inhibited it (34). Altogether, these results clearly established PKD1 as a proproliferative and antidifferentiating protein in mouse keratinocytes. Nevertheless, one study reveals that these proproliferative and antidifferentiative PKD1 functions would only be revealed in particular situations. In fact, mice carrying conditional and specific disruption of PKD1 in keratinocytes (K14-Cre-PKD1-cKO) displayed no alteration in epidermal proliferation and differentiation suggesting that PKD1 would be dispensable for skin development and homeostasis under normal conditions (35). However, PKD1-cKO–deficient mice displayed strongly impaired wound healing and reepithelialization and became mostly totally refractory to DMBA/TPA–induced tumor formation (35). These results are of crucial importance because they underline the role of PKD1 in adaptive responses such as skin carcinogenesis and are finally in total accordance with data showing that PKD1 was activated by UVB and that its overexpression protected keratinocytes from UVB-induced apoptosis (36). In fact, through the stimulation of prosurvival signaling pathways, PKD1 could thus allow the proliferation of mutated cells leading to cancer formation.

When considering normal human skin cells, the status of PKD1 expression remains more elusive. In fact, PKD1 was initially described to be expressed throughout the superbasal layers with a predominant expression in the stratum basalis, in accordance with its proproliferative role as previously described in mice (37). However, further results indicated that PKD1 was not detected in human keratinocytes, these cells being more dependent on the two other isoforms of the PKD family, PKD2 (prodifferentiative), and PKD3 (proproliferative), for their proliferation (38). As hypothesized by the authors, such discrepancies could be the consequence of a certain lack of specificity of the antibodies used by Ristich and colleagues (37) because sc-935 antibody has been reported by others to cross-react with PKD2 in Western blot analyses (39). This hypothesis does not however provide an explanation as to why the prodifferentiative PKD2 would be the most abundantly expressed in actively dividing cells. In hyperplastic human skin disorders, such as melanoma (40), basal cell carcinoma, and psoriasis (37), PKD1 was found to be upregulated. Taken together, these data suggest that, despite the fact that PKD1 seems not to be of primary importance in normal skin homeostasis (35), it has, whatever the species models, an evident proproliferative role in the context of skin carcinogenesis, wound healing, and other skin hyperproliferative diseases like psoriasis. Consequently, these data also suggest why targeting PKD1, by relatively selective inhibitors, has to be considered as an option for the treatment and prevention of epidermal tumorigenesis, and for other hyperproliferative diseases such as psoriasis. To this goal, we showed that inhibition of PKD1 in melanoma cells (i) decreased their colony-forming capacities probably through the regulation of the ERK, JNK, and NFκB signaling pathways, and (ii) induced the relocation of β-catenin from nucleus to plasma membrane, and the subsequent expression decrease of some proproliferative target genes such as cyclin D1 (40).

Although PKD1 was shown to have a main role in insulin secretion of pancreatic islets (41), our review will only focus on the exocrine function of the organ because tumors arise mainly from these structures. Among pancreatic cancers, the most common is the pancreatic ductal adenocarcinoma (PDAC) representing 90% of all cancers and considered among the most lethal cancers with a very low 5-year survival rate of about 3%–5%. It is characterized by an early metastatic state associated with a rapidly succeeding chemoresistance (42).

The expression status of PKD1 is often incorrectly formulated due to sentences that, using shortcuts, become inaccurate in some publications. Indeed, stating that PKD1 is not expressed in the normal exocrine pancreatic cells is incorrect because PKD1 was detected in untreated rat pancreatic acini and very rapidly phosphorylated (detectable effect after 30 seconds) and activated by cholecystokinin (CCK) through a PKC-δ–dependent signaling pathway (43). PKD1 plays a major role in rat pancreatic acini modulating experimental pancreatitis. Pharmacologic inhibitors of PKD1 attenuated early pancreatitis events (44), but also significantly attenuated pancreatic injury when used as posttreatment (45). These results are of crucial importance because, by promoting pancreatitis-associated necrosis (46), PKD1 would promote a proinflammatory state, especially characterized by the secretion of IL6 and MCP-1 proteins (45), which could give rise to pancreatic lesions, known as risk factors for cancer development. This is consistent with data showing that PKD1 is upregulated in mice pancreatic acinar cells that undergo acinar-to-ductal metaplasia (ADM; ref. 47). ADM occurs after inflammation or injury and is reversible unless in a persistent proproliferative context where it can progress to neoplasia and cancer (48, 49). During ADM, PKD1 was shown to act downstream of TGFα and K-Ras, and upstream of the Notch pathway, to promote the formation of ductal structures (47).

On the other hand, PKD1 is very moderately expressed in normal mouse and human pancreatic tissue contrary to PKD3 that represents the major, if not the single, isoform (50). However, transformed acinar cells and human PDAC show a strong increase in PKD1 expression compared with normal tissue (7, 51, 52). PKD1 shortened the doubling time of PKD1-transfected Colo357 cells by 20% probably through an enhanced expression and activity of hTERT (52) and dose-dependently increased DNA synthesis and cell proliferation of inducible PKD1-expressing Panc-1 cells (53). The latter result is of main importance because it makes a clear proportional link between PKD1 expression levels and cell proliferation rates. By strengthening the duration of ERK signaling and inhibiting G protein–coupled receptors (GPCR)-induced c-Jun phosphorylation, increased PKD1 levels stimulate cell-cycle progression allowing the accumulation of immediate gene products such as c-Fos, whereas inhibition of c-Jun phosphorylation leads to the attenuation of the JNK signaling switching its proapoptotic action to a proproliferative one (53). Another hallmark of PDAC is a highly increased NFκB signaling, linked to an increased proliferation of tumor cells. Oncogenic K-Ras induces canonical NFκB signaling and upregulates PKD1 expression and activity (7). Moreover, overexpression of PKD1 increased anchorage-independent growth of PDAC cells (21), whereas its pharmacologic inhibition by CRT0066101 (21), or the expression of a PKD1 kinase–dead mutant (53), or its molecular silencing (21) decreased the number of colonies formed in a semi-solid medium. The prooncogenic role of PKD1 was further demonstrated in vivo because orally given PKD1 inhibitor CRT0066101 significantly reduced the volume of established tumors in subcutaneous Panc-1 xenograft models, or inhibited the final tumor volume of orthotopic implanted Panc-1 cells (51). These results therefore highlighted the role of PKD1 not only in the genesis, but also in the maintenance of pancreatic tumors. These effects could be the consequence of the regulation of angiogenesis. In fact, PKD1 expression stimulates the secretion of proangiogenic factors such as VEGF and CXCL-8 and enhances the association between pancreatic cancer cells and endothelial cells on Matrigel, whereas the PKD1 inhibitor CRT0066101 reduces angiogenesis in orthotopic PDAC tumor explant in vivo (21).

Taken all together, these data define PKD1 as a clear and prominent proproliferative factor in PDAC making this protein a putative target for the development of new therapeutic strategies against pancreatic cancer.

Despite the scarcity of studies, it seems that PKD1 plays different roles depending on which part of the tract is considered. Compared with normal tissues, PKD1 expression was shown to be markedly downregulated in gastric (11, 48, 54) and colorectal human cancer cells (55, 56), with a more pronounced decrease in higher grade tumors (54, 55) suggesting an antioncogenic role of this protein in these tissues. In fact, overexpression of PKD1 in human gastric adenocarcinoma cells (AGS cells; ref. 54) or in human SW480 colorectal cancer cells (55) inhibited cell proliferation, clonogenicity, and motility, and delayed tumor growth in a xenograft mouse model. Such effect could be dependent on PKD1-induced nuclear exclusion of β-catenin and the subsequent decrease of its transcriptional activity (55) toward several proto-oncogenic genes like cyclin D1 or c-Myc (57). Although these results were both provided upon PKD1 overexpression, potentially generating false, very low-level affinity interactions, and in a cell model, SW480 cells, chosen upon its particular PKD1 and β-catenin expression and localization levels, they suggest that PKD1 may negatively regulate cell proliferation through a β-catenin–dependent mechanism in normal colorectal cells. Therefore, loss of PKD1 expression during steps of tumorigenesis would release this break, promoting cell proliferation. Thus, downregulation of PKD1 expression levels appears as a key determinant for gastric and colorectal tumorigenesis process and could be the consequence of an epigenetic inactivation occurring on the PKD1 promoter as demonstrated in gastric cancer cells (11). However, this mechanism does not seem to be generalizable to other parts of the gastrointestinal tract because authors mentioned that DNA methyltransferase inhibitors were unsuccessful to reexpress PKD1 in colorectal cancer cells (56).

In contrast, PKD1 was shown to have a proproliferative role in intestinal cells both in vitro and in vivo suggesting a potential prooncogenic role in the intestine. Selective knockdown of endogenous PKD1 inhibited DNA synthesis and cell proliferation induced by angiotensin II and vasopressin in IEC-18 rat intestinal cell line (58). Moreover, overexpression of PKD1 in small intestine of transgenic mice increased the proliferation rate and the number of intestinal cells per crypt in vivo (58). PKD1 overexpression would enhance the proliferation induced by GPCR agonist–dependent signaling pathways through the phosphorylation and the subsequent nuclear export of class IIa histone deacetylase (such as HDAC4, HDAC5, and HDAC7; ref. 59). In fact, HDACs regulate gene expression by interacting with and repressing various transcription factors (60). More recently, PKD1 overexpression was also shown to promote angiotensin II–stimulated cell proliferation by inducing β-catenin translocation to the nucleus (61). These intriguing results were in total contradiction with previous ones found in human colon cancer cells (55), highlighting the apparent complexity of the molecular mechanisms regulated by PKD1. However, it is essential to notice that these contradictory results were (i) not conducted in the same species (murine vs. human) suggesting that PKD1 “species-specific roles” cannot be excluded; (ii) performed either in tumor models, for colon and stomach, or in normal cells, for the intestine, each expressing specific cellular contexts that could be determinant to define the role of PKD1. Among the suspected proteins, the two other members of the PKD family are of interest and most particularly PKD2. Indeed, PKD2 is often described to display opposite functions to PKD1 and their respective expression is often inversely regulated as demonstrated in colon (56) and gastric (54) cancers. Therefore, one may suppose that depending on the relative expression level of PKD2, the apparent role of PKD1 could be consequently modulated. Therefore, whenever possible, this point should be taken into consideration and further studies are still needed for a better understanding of the role of PKD1 in the whole gastrointestinal tract.

Kidney cancer is among the tumors with the fastest growth rate and is the deadliest type of urinary tract cancer (62). Aldosterone is a mineralocorticoid hormone that regulates ion fluxes among nephron epithelium. In addition to its well-characterized role as an ion transport modulator (63), aldosterone was also shown to stimulate the proliferation of human RCC (renal cell carcinoma) cell lines (64) as well as the murine M1 cortical collecting duct cell line (M1-CCD; ref. 65). In M1-CCD cells, aldosterone was shown to stimulate the phosphorylation of PKD1 through a mineralocorticoid receptor- (MR) and EGFR-dependent mechanism (66). PKD1 knockdown inhibited aldosterone-stimulated proliferation demonstrating a proproliferative role of PKD1 in this cell line (65). PKD1 may promote aldosterone-induced cell proliferation by maintaining a sustained activation of ERK1/2 and inducing its translocation to the nucleus (65). Taken together, these findings highlight the proproliferative role of PKD1 in renal collecting duct cells. However, it is important to notice that among 17 tissues analyzed from TCGA data, high PKD1 mRNA expression is a good prognostic factor in kidney tumors (Table 2). Despite their apparent contradiction, these results mainly highlight, as previously mentioned in the introduction, that a direct correlation between tumor mRNA expression levels and PKD1 activity cannot be assumed and that the relevant and interesting prkd1 gene expression analysis cannot be freed from that of PKD1 activity.

Lung cancers are the leading cause of cancer mortality worldwide (67). Among the two subtypes, non–small cell lung cancers (NSCLC) are the most common (85%), whereas small-cell lung cancers (SCLC) are usually more likely to spread and become life threatening (68). Because of a few studies about PKD1 in the lung, it remains difficult to have a clear idea about the pro- or antiproliferative role of this protein in this tissue.

PKD1 was shown to be highly expressed and phosphorylated in bronchiolar and regenerative alveolar epithelia from patients with idiopathic pulmonary fibrosis (69). This pathology is characterized by lung fibroblast activation and proliferation suggesting a proproliferative role of PKD1 in this tissue (70). In contrast, PKD1 mRNA expression was shown to be downregulated in NSCLC compared with normal tissues especially when patients with NSCLC displayed venous invasion or lymph node metastasis, suggesting that PKD1 would negatively regulate NSCLC tumor development (71). Consistent with this, although PKD1 was shown to induce a prolonged activation of the ERK1/2 signaling pathway in Swiss-3T3 cells (see Chapter Fibroblast), PKD1 mediates the inhibition of PMA-induced ERK phosphorylation in A549 cells. Thus, pharmacologic PKD1 inhibition or its downregulation resulted in enhanced PMA-induced S6K1 and ERK phosphorylation and A549 cell proliferation, whereas constitutively active PKD1 results in S6K1 and ERK inhibition (71). Interestingly, the antiproliferative role of PKD1 may also be dependent upon its ability to maintain a low-proliferative epithelial phenotype of lung cells. In fact, PKD1 was also described to directly bind to E-cadherin leading to its membrane redistribution and activation independently of DAG or PKC in A549 cells (72). Because PKD1 positively regulates E-cadherin transcription, this interaction/activation generates a positive feedback loop favoring the maintenance of a low proliferative epithelial phenotype. Conversely, and in accordance with these results, knocking down PKD1 induces the loss of expression of E-cadherin promoting the epithelial-to-mesenchymal transition and the acquisition of migratory capacities (72).

However, it is important to note that in the study by Ni and colleagues (71), PKD1 protein expression levels were not determined in NSCLC tissue specimens and even if mRNA levels decreased, no one can conclude that protein levels will automatically follow the same profile. However, if these results are also later confirmed at the protein level, future studies would have to carefully consider these data to study the role of PKD1 in lung. Indeed, the use of the adenocarcinoma-derived human alveolar basal epithelial cell model, A549 cells, cannot then represent a good model enough for NSCLC investigations insofar as this tumor cell line expresses large amounts of PKD1 (73) unlike what has been found in human tissues.

Prostate cancer is the second leading cause of cancer-related death and the most commonly diagnosed cancer in males in the United States (74). Despite advances in the screening methods, effective treatments of advanced androgen-independent tumors are still to be found.

Scores of studies analyzed the roles of PKD1 in prostate cancer cells. PKD1 was shown to be downregulated in metastatic androgen-independent prostate cancers compared with their respective primary tumor (75, 76). Moreover, PKD1 is highly expressed in low proliferative and low metastatic androgen-sensitive LNCaP cells and downregulated in the castration-resistant LNCaP-derivative cell line, C4-2 cells (androgen-hypersensitive), or in the highly metastatic androgen-insensitive DU145 and PC3 cells (75–77). These results suggested an association between the downregulation of PKD1 and the progression and aggressiveness of prostate cancer. Knockdown of PKD1 using shRNA enhanced cell growth (77), whereas its overexpression inhibited cell proliferation (77, 78). Moreover, curcumin was also thought to inhibit prostate cancer cell proliferation through a PKD1-dependent mechanism (79). All these results indicated that PKD1 can be considered as an antiproliferative protein in prostate cancer cells. Such an effect could be mediated through the binding of PKD1 with β-catenin and the subsequent inhibition of β-catenin–mediated proliferation function (78), or through the secretion of matrix metalloproteinase-2 and -9 (75). However, despite many studies, the precise function of PKD1 with regards to cell proliferation and the mechanisms it controls remain somewhat unclear. For instance, inhibition of PKD1 expression drastically decreased ERK phosphorylation in DU145 cells (75) although this protein is described as mediating proproliferative signaling pathways (80). Moreover, the molecular mechanisms by which PKD1 would inhibit cell proliferation have been mostly demonstrated in PKD1-overexpressing prostate cancer cells (75, 77, 78, 81). This technical approach is not necessarily the better way to proceed because a recent study showed that according to TCGA data [Prostate Adenocarcinoma (TCGA, PanCancer Atlas)], mRNA PKD1 expression levels in prostate cancer are upregulated in about 5% tumors suggesting that PKD1 hyperactivity may play a more important role in tumor progression than overexpression (82). Moreover, LNCaP cells already express very high amounts of PKD1 making the relevance of such a model questionable insofar as overexpression can only lead to nonspecific and nonrelevant interactions. Furthermore, the comparison of results obtained in different cell lines is also complicated and very hazardous. Indeed, the most commonly used cell lines (i.e., LNCaP, C4-2, and PC3) display different sensitivities to androgens. However, PKD1 has a particular relationship toward androgen receptor, AR. Indeed, PKD1 would inhibit AR-mediated transcriptional activity (observed in PKD1 and AR-overexpressing cells; ref. 81) while androgens would inhibit PKD1 expression through the expression of a repressor, FRS2 (83). Consistent with the latter results, incubation of cells in an androgen-depleted medium increased PKD1 expression (83) indicating that AR expression and androgen sensitivity status of the cell lines must be considered in a serious way and that the extrapolation of results between different cell models cannot be done as easily as this. Therefore, even if these results should not be questioned, one must be aware of their limits. In fact, it is interesting to note that contradictory results exist even in studies conducted by the same team in which inhibition of PKD1 expression was shown to have either an effect (77) or not (78) on the growth of LNCaP cells.

Although PKD1 was described as an antiproliferative protein in prostate cancer cells, many PKD1-targeting pharmacologic inhibitory compounds were assayed in these cell lines (84–88). Despite the fact that these molecules are not totally specific toward PKD1, they all induced cell growth arrest that can be reversed through infection with adenovirus carrying PKD1 gene (87, 88) suggesting a proproliferative role of PKD1 that appears discrepant with data presented in the previous paragraph. However, due to the drastic effects induced by these molecules on cell growth, PKD1 appears more likely as a prosurvival factor as also suggested in LNCaP where PKD1 was demonstrated to protect cells from PMA-induced apoptosis by promoting ERK and NFκB activities (84). Although an antiproliferative effect is compatible with a prosurvival role, these results indicate that more data are needed to better understand the role of PKD1 in prostate cancer cells taking into account each particular cellular environment.

Breast cancer is the most common cancer, and the second leading cause of cancer-related death in women (89). Despite improvement in early detection and treatment of breast tumors, advanced metastatic breast cancer remains life threatening. Accumulating evidence show a potential role of PKD1 in breast tumor progression. However, the link between PKD1 expression levels, PKD1 activation/activity, and tumor aggressiveness remains unclear. In fact, PKD1 expression was shown to be high in normal breast tissues and reduced in more than 95% of invasive breast cancer tissue and triple-negative tumors (9, 90). Conversely, its expression is markedly increased in breast cancer cell lines compared with normal cells where it is undetectable (91). Consistently, a large-scale analysis performed in 152 malignant breast tissues showed that patients with poor prognosis overexpressed PKD1, while those with good prognosis had significantly lower PKD1 expression levels (91) suggesting that PKD1 expression was positively linked to disease progression. Although PKD1 expression was shown to be regulated through epigenetic modifications such as DNA methylation of its promoter sequence (9), its expression status should be more carefully studied in tissues and cell lines taking also into account the expression levels of other determinant markers of breast cancers' progression such as ERα, Her2, and progesterone receptor (PR).

Overall, PKD1 was shown to positively regulate cell proliferation. In fact, PKD1 overexpression strongly and specifically increased MCF-7 cell growth by promoting G₀–G₁ to S-phase transition of the cell cycle through a MEK/ERK-dependent signaling pathway (92). Moreover, PKD1 overexpression improved anchorage- and growth factor–independent proliferation in vitro and promoted tumor growth in vivo (92). It also increased ERα expression further demonstrating the link that exists between these two proteins (91). Furthermore, PKD1 overexpression increased MCF-7 cells' sensitivity to estradiol, their independence toward estrogen for proliferation, and their partial resistance to the antiestrogen, ICI 182,780 (91). Interestingly, this new cell behavior looks like the one of prostatic C4-2 cells which, contrary to their parental cells, LNCaP, display an androgen-independent and -hypersensitive phenotype for proliferation associated, in this case, with a loss of PKD1 expression (see Chapter Prostate). More recently, our data were further confirmed in the drug-resistant model of MCF-7-ADR cells expressing high levels of PKD1 as well as cancer stemness markers compared with parental MCF-7 cells (93). In fact, knockdown of PKD1 by siRNA- or miRNA-targeting PKD1 in MCF-7-ADR cells was shown to reduce the number of tumorspheres, to increase doxorubicin-induced apoptosis in vitro, as well as to suppress tumor formation in xenograft models in vivo (93). Altogether, these results strongly define PKD1 as a proproliferative and protumorigenic factor in breast cancer cells.

As proactive members of the tumor microenvironment, immune cells are main actors in tumor progression because they can either positively or negatively regulate tumor growth depending on their nature, activity, and reciprocal interactions. In the perspective of developing antitumor strategies, it is therefore relevant to know whether PKD1 regulates immune cell proliferation to anticipate, as much as possible, the potential consequences of targeting PKD1 in tumors. Studies have shown that PKD1 is not expressed in murine T- and B-lymphocytes, nor in malignant B cells, nor in thymus and spleen (94–96), PKD2 being the major PKD isoform expressed. However, ectopic expression of a constitutively active form of PKD1 induced pre-T-cell proliferation (97) illustrating again the proproliferative role of PKD1 when expressed and raising questions about its putative function in hematopoietic malignancies. Nevertheless, PKD1 was shown not to be expressed in non-Hodgkin and Hodgkin lymphoma (98) and, although it regulates migration, PKD1 has no proproliferative role in multiple myeloma (99). Nonetheless, Epstein–Barr virus (EBV) latent membrane protein-1 (LMP1) induces PKD1 expression in B-cell lymphoma and protects them from apoptosis. This contributes to the LMP-1–induced drug resistance and progression of the pathology and makes PKD1 a potential molecular target in EBV-associated B-cell lymphoma (100).

Furthermore, cancer development is associated with a local inflammatory response that generally surrounds the tumor. PKD1 has a dual role considering inflammation because it was shown either to promote or inhibit inflammation through, among others, the secretion of chemokines by mast cells (101–105). However, the use of distinct inflammation-inducing agents and methods to analyze the inflammatory response makes it impossible to conclude, until now, about the role of PKD1 in this physiopathologic response.

Because PKD1 was mainly described as a proproliferative protein in several cancer cells both in vitro and in vivo, it emerged as an interesting putative therapeutic target to fight against tumors and different strategies were developed in the attempt to inhibit its activity. Among them, different pharmacologic inhibitors were developed and characterized as to their effectiveness and specificity toward PKD1 both in vitro and in vivo (Table 3). Their inhibitory characteristics (i.e., IC₅₀) vary from one study to another because the experimental approaches used by authors to determine them were not normalized. Three main techniques were commonly used such as the measurement of the phosphorylation of a PKD1 substrate, mainly syntide 2 in an in vitro kinase assay, the quantification of the cellular inhibition of the autophosphorylation of the PKD1 serine 910 (human numbering) residue analyzed by Western immunoblotting (cellular inhibition of PKD1 autophosphorylation) and the analysis of their effects on cell viability. Despite very variable characteristics, these compounds were found to be effective in blocking proliferation and other cellular functions such as invasion and migration of different cell models making them promising inhibitors for cancer treatment (85). Unfortunately, they were also described to be too rapidly metabolized, which limited their efficacy in vivo. However, among them, CRT0066101 was shown to inhibit growth of pancreatic, colorectal, bladder and triple-negative breast cancer cells xenografts in vivo [review in ref. (106); refs. 107, 108]. It thus appeared as a relatively good therapeutic candidate because it blocked cell-cycle progression at the G₁ phase and increased apoptosis by inhibiting the phosphorylation of “classical” proproliferative proteins such as Myc, MAPK1/3, Akt, Yap, and Cdc2 (107). Studies have also suggested that the overexpression and activation of PKD1 observed in CD34⁺ skin stem cells and skin tumors are potential targets for the treatment of skin carcinogenesis (31). However, to our knowledge, no PKD1 inhibitors have been used in clinical trials and further studies are absolutely necessary to notably increase the specificity of such compounds toward not only the two other members of the PKD family, PKD2 and PKD3, but also toward other protein kinases to only interfere with PKD1-regulated (or -dysregulated) signaling pathways. In fact, many of these compounds cannot be considered as specific inhibitors of PKD1 and should be used with great caution in experiments concluding about the specific role of PKD1 in several cell functions. To this end, molecular extinction of PKD1 protein expression, when possible, remains a reliable technical approach to strengthen and comfort results obtained with pharmacologic compounds.

Whatever the cell type, the tissue, and its normal versus cancer status considered, it remains clear that PKD1 plays a crucial role in growth-dependent signaling pathways. However, due to its potential opposite functions, pro- or antiproliferative, illustrated in Figs. 3 and 4, respectively, the development and the putative use of PKD1-targeting inhibitors as therapeutic tools may be considered with major caution. Some contradictory data exist but are sometimes the consequences of studies in which the analysis of the PKD1 phosphorylation level was too rapidly correlated to the activity of the protein. But many results clearly indicated that such a transposition cannot be made directly. Indeed, although the phosphorylation of PKD1 onto its serine S738/742 and S910 residues seems to be important for its activation process, a direct relationship between PKD1 activation and its catalytic activity is not always observed. Consistently, the development of robust techniques allowing the direct and precise measurement of PKD1 activity would be a major breakthrough in the field of PKD1 studies. Moreover, a very recent study showed that a direct correlation cannot be automatically made between the relative PKD1 expression levels in normal and tumor tissues and its role in tumorigenesis (109). In fact, PKD1 is significantly downregulated in head and neck localized tumors and metastases compared with normal tissues due to epigenetic modifications, suggesting an antiproliferative role of this protein. However, its expression has been positively correlated with both the subcutaneous head and neck squamous cell carcinoma xenografts growth and a sustained bombesin-induced ERK1/2 activation demonstrating a proproliferative role of this protein (109). In addition, despite numerous studies concerning PKD1, most of the data come from experimental studies and very few information come from cohort ones. Thus, obtaining in vivo data on large scales is also an important point to be considered to better understand the role of this protein in the different tissues.

No potential conflicts of interest were disclosed.

The authors thank Dr. Sylvie Babajko for critical reading of the manuscript. This work was supported by the Centre National de la Recherche Scientifique (CNRS) and the Ecole Normale Supérieure Paris-Saclay (ENS Paris-Saclay).