2. The Role of Dkk-3 in the Regulation of Cancer Fibroblasts and Stellate Cells
The fundamental role of the stroma in regulating cancer progression has been reported in several settings
[11][12]. Tumor microenvironment cells may be fibroblasts, endothelial cells, pericytes, or bone marrow-derived cells such as mesenchymal stem cells (MSC), neutrophils, macrophages, and mast cells
[13][14]. Signaling crosstalk among cancer cells, stromal cells, and epithelial cells is mediated by chemokines, cytokines
[15], and growth factors and proteases that remodel the extracellular matrix
[14][15]. In general, fibroblasts are the predominant cell type in the stromal compartment of most solid tumors
[15]. Cancer-associated fibroblasts (CAFs), also called cancer-reactive stromal cells, become activated to proliferate and increase the production of ECM, acquiring features of smooth muscle (SM) cells, similar to myofibroblasts
[15]. Recently, Dkk-3 was reported to modulate the response to TGF-β, which promotes the differentiation of fibroblasts into CAFs in prostate cancer
[16]. This link may reflect the interaction between cancer/stromal cell compartments and epithelial–mesenchymal transition (EMT) progression, since TGF-β can be shifted from being anti-tumorigenic to potently pro-tumorigenic
[17][18].
An association between Dkk-3 and tumor stroma has been reported in ER-negative breast cancer, ovarian cancer, and colon cancer cells
[19], where
DKK3 gene expression is upregulated in the tumor stroma, correlating with a more aggressive tumor type
[19]. Heat shock factor-1 (HSF-1) was found to regulate
DKK3 gene expression in stromal cells as part of the response to stress including oxidative stress, nutrient deprivation, and protein misfolding.
In benign prostatic hyperplasia, Dkk-3 is more highly expressed in the prostate stroma, particularly in myofibroblasts, compared to the epithelium
[16][20]. This alteration may be a response to the loss of Dkk-3 in the epithelium, which normally opposes the potential tumorigenic impact of TGF-β by reducing Smad3 phosphorylation in myofibroblasts; thus, stromal Dkk-3 may contribute to supporting a normal acinar architecture, thereby reducing proliferation and limiting prostate cancer cell invasion
[16]. Notably, these changes in the expression of Dkk-3 in the two compartments, epithelium, and stroma, are inversely correlated with changes in the expression of TGFBI
[16]. As noted earlier, it has also been proposed that Dkk-3 positively regulates Wnt/β-catenin signaling through an interaction with Krm
[21]. Another study reported the therapeutic effect of Ad-REIC, a DKK3-expressing adenovirus being developed for gene therapy, through the inhibition of CD147 (cluster of differentiation 147)
[22]. CD147 is a surface glycoprotein upregulated in many solid tumors including prostate cancer. Cell surface expression of CD147 in cancer promotes adjacent fibroblasts and cancer cells to secrete matrix metalloproteinases (MMPs), which are inducers of cancer cell invasion and metastasis
[22]. At the same time, CD147 induces VEGF and hyaluronan, thereby stimulating angiogenesis, drug resistance, and anchorage-independent growth. Ad-REIC treatment significantly reduces CD147 levels in prostate cancer cell lines
[22]. In adult human osteoarthritis, an inflammatory disease, increased expression of DKK3 may protect against cartilage degradation and aberrant signaling associated with this disease
[23]. Dkk-3 reduces metalloproteinases by attenuating NFκB signaling, thus preventing the breakdown of cartilage, which leads to irreversible degradation of ECM, especially type 2 collagen
[24].
Dkk-3 has been also described as a tumor-promoting factor, especially in cancers associated with the upregulation of
DKK3 expression such as in head and neck and pancreatic cancers
[25][26]. For example, a study conducted on human pancreatic duct adenocarcinoma (PDAC) and normal pancreas reported that Dkk-3 is mainly expressed in the stroma, where it is secreted by pancreatic stellate cells (PSC)
[27]. Dkk-3 was found to enhance PDAC metastasis and interfere with chemotherapy resistance, while a combination of a Dkk-3 function blocking antibody, JM6-6-1, and immune checkpoint inhibitors was found to limit tumor growth and survival
[27]. It is a common concept that PSCs play a role in promoting PDAC proliferation and survival as well as reducing PDAC cell responses to therapy
[27]. Thus, silencing of DKK3 in human PSCs was shown to inhibit PDAC cell proliferation and migration and blocking Dkk-3 might contribute to better responses to chemotherapy
[27].
3. The Role of Dkk-3 in the Regulation of Cancer Angiogenesis
Dkk-3 may play a role in the remodeling of stroma through its involvement in angiogenesis
[28][29]. Angiogenesis is fundamental for cancer invasion and metastasis
[30]. Dkk-3 has been reported to affect the expression levels of angiopoietin 1 (ANGPT1), angiopoietin 2 (ANGPT2), and vascular endothelial growth factors (VEGF)
[29][31][32]. During tumor progression, the ANGPT switch is considered to be rate-limiting in favor of a higher ANGPT2/ANGPT1 expression ratio
[33]. In benign prostatic hyperplasia (BPH) and prostate cancer, Dkk-3 increases ANGPT2 expression, a destabilizing factor that leads to the formation of vessel sprouting in the tumor microenvironment
[31]. Dkk-3 also supports human umbilical vein endothelial cell (HUVEC)-related tube formation through increasing VEGF stimulation by activating both ALK1 and TGF-β/Smad signaling
[34]. Busceti et al. (2018) used cultured endothelial cells to show that Dkk-3 induces VEGF expression to exert a protective effect against ischemic neuronal injury
[35]. Moreover, Dkk-3 has been reported to enhance new vessel growth in non-MYC-driven medulloblastoma
[36]. Interestingly, Dkk-3 is also an inducer of cavernous vascular endothelial cells, pericytes, and endothelial junctional proteins such as claudin-5 and ZO-1 in diabetic mice with erectile dysfunction (ED), increasing cavernous endothelial cell expression of angiogenesis factors including ANGPT1, VEGF, and bFGF. Thus, Dkk-3 might be useful as a therapy for diabetic patients with ED
[37]. Dkk-3 also promotes the migration and recruitment of Sca1
+ vascular stem/progenitor cells from the murine aortic adventitia. As already noted, Dkk-3 mediates vascular progenitor cell migration by binding to the chemokine receptor CXCR7
[32]. Dkk-3/CXCR7 engagement in vascular progenitor cells appears to lead to the activation of ERK1/2, phosphatidylinositol 3-kinase/AKT, and the small GTPases Rac1 and RhoA
[32].
Dkk-3 not only influences angiogenesis but may also be useful in regenerative medicine, where it has been found to rejuvenate ischemic tissue. In addition, Dkk-3 can promote endothelial cell regeneration, which contributes to protection against atherosclerosis
[28]. In contrast, in renal fibrosis, secreted Dkk-3 is described to contribute to trans-differentiation of endothelial cells to myofibroblasts, termed “endothelial to mesenchymal differentiation” in vivo
[38].
4. The Role of DKK3 in the Regulation of Cancer Immune Responses
Dkk-3 is not only involved in the activation of fibroblasts to myofibroblasts and angiogenesis but also modulates the inflammatory and immune responses of inflammatory cells including macrophages, mast cells, eosinophils, and neutrophils, which are part of the stromal compartment
[39]. During inflammation, inflammatory cells are stimulated to release cytokines and chemokines that stimulate the initiation of the immune response. Persistent or chronic inflammation may incite carcinogenesis by causing DNA damage, cell proliferation, and angiogenesis
[39]. Non-alcoholic fatty liver is a chronic inflammatory liver disease that is exacerbated in liver-specific Dkk3 knockout mice and ameliorated in liver-specific Dkk3-overexpressing transgenic mice, in a mechanism that requires the MAP kinase Ask1
[40]. Studies in mouse models of chronic renal inflammatory disease, in contrast, found that Dkk3 induces cytokines that lead to fibrosis
[41]: Dkk3 gene knockout or treatment with a Dkk3 function-blocking antibody was accompanied by increased accumulation of IFNγ-producing Th1 and Tregs at the expense of Th2 cells, which are profibrotic
[42][43]. A third example from studies in mice is in the pancreas, where Dkk3 levels are increased in caerulein-induced acute and chronic pancreatitis. In this context, Dkk3 knockout increased recovery rates by limiting canonical Wnt signaling and increasing Hedgehog (Hh) signaling during pancreatic regeneration. These effects have been proposed to involve a reduction in the expression of the transcriptional repressor Gli3
[44]. The same study reported a similar scenario in the context of liver regeneration, leading the authors to propose that Dkk3 functions as a roadblock in liver and pancreas repair/regeneration upon injury
[44]. Dkk-3 activation of non-canonical Wnt/β-catenin-independent signaling involving c-JUN N-terminal kinase (JNK) increases the secretion of inflammatory factors, such as TNF-α and IL-1β, in the brain, suggesting a neuroprotective mechanism
[45] Furthermore, Dkk-3 is a factor required for smooth muscle cell differentiation
[45]; the absence of Dkk-3 may therefore contribute to accelerating blood vessel inflammation and subsequent atherosclerotic plaque formation. At the same time, overexpressed Dkk-3 ameliorates myocardial infarction and inhibits the associated inflammatory process through the regulation of ASK1–JNK/p38 signaling
[45].
The expression of Dkk-3 in mesenchymal stem cells also limits CD8
+ and CD4
+ T cell-mediated responses to suppress the process of inflammation
[46]. Mesenchymal stem cells are known to suppress the growth of tumors by inhibiting T cell proliferation, the induction of T regulatory cells, and facilitating the generation of immunosuppressive M2-type macrophages
[46]. Treating mice with intracerebral hemorrhage with recombinant Dkk-3 results in a significant reduction in the release of TNF-α, cleaved caspase-1, and IL-1β. Dkk-3 thus appears to reduce JNK/AP-1-mediated inflammation, which may improve the neurological outcomes
[47].
Human fibroblasts infected with Ad-REIC showed increased IL-7 production, which indirectly inhibits cancer tumorigenesis
[48][49]. Furthermore, the expression of Dkk-3 ameliorates graft transplantation through the regulation of T cell-mediated rejection
[50]. In addition to the induction of IL-7 secretion, Ad-REIC facilitates an anti-tumorigenic microenvironment by promoting tumor-associated antigen-specific CD8
+ cytotoxic T lymphocytes (CTLs)
[9]. Injection of Ad-REIC leads to the presentation of proteins derived from apoptotic cancer cells to dendritic cells (DCs), thereby inducing the conversion to cytotoxic T cells and CD8
+ CTLs
[9]. On the other hand, deletion and/or neutralization of Dkk3 in TCR/MHC class-I double transgenic mice increases local CD8
+ T cell infiltration and enhances MHC class-I mismatched anti-tumor and skin graft rejection
[50]. Thus, organs that are classified as tissues with limited regenerative capacity such as the nervous system, eye, uterus, and placenta show the highest Dkk3 expression, possibly creating an immunosuppressive microenvironment and protecting against autoimmunity
[51]. In addition, a Dkk3
−/− experimental autoimmune encephalitis (EAE) CNS mouse model showed an increase in the total number of infiltrating T cells, particularly the IFNγ-producing CD4
+ and CD8
+ T cells, leading to disease exacerbation. In this EAE and Dkk3 deficiency model, the hypothesis is that the action of Dkk3 is local and does not include secondary lymphoid organs
[50]. In human glioblastoma, a machine-learning approach found an inverse correlation between DKK3 gene expression and anti-tumoral immunity, particularly in CD8
+ and CD4
+ T cells
[52]. These observations merit further studies of mouse and human tumor models to clarify the role of Dkk-3 as a tissue-derived immune modulator and potential immunotherapy. Of potential relevance, serum autoantibodies to cancer/testis antigens (CTAs), which are used as biomarkers for the anti-tumor immune response, were recently measured in Ad-REIC-treated CRPC patients and it was concluded that monitoring CTAs in this setting can contribute to clinical management
[53].
While most mouse studies have highlighted potential effects on T cells, Dkk3 is also a player in B cell-mediated autoimmune disease. B cells are divided into B1 and B2 subsets, with distinct functions, time of maturation, and anatomical location
[54][55]. B1 cells secrete the majority of natural immunoglobulins (IGs) in the innate immune response, while B2 cells are responsible for adaptive immune responses
[55]. Dkk3-deficient mice show suppression of B2 cell maturation and reduced proliferation and self-maintenance of peripheral B1 cells
[54].
5. The Role of DKK3 in Stem Cell Differentiation
The role of Dkk-3 in orchestrating the quiescence and/or the differentiation of stem/progenitor cells has been investigated in prostate, breast, pancreatic, liver, and oral submucosal diseases
[44][56][57][58]. Over the last decade, it has been of interest to identify the genes of somatic cells involved in re-programming to improve organogenesis for tissue engineering
[59]. In this context, Dkk-3 has been highlighted as a molecule with overlapping regulatory effects on the Wnt and Hh signaling pathways during tissue regeneration and somatic reprogramming
[44]. As noted above, studies in mouse models found that loss of Dkk3 ameliorates liver and pancreatic injury in acute and chronic disease via increased expansion and differentiation of the liver progenitor cell pool
[44]. Additionally, Dkk3 induces the differentiation of mouse embryonic stem cells to smooth muscle cells
[60]. Consistent with a role for Dkk-3 in human progenitor cell differentiation, silencing of DKK3 in benign prostate epithelial (RWPE-1) and stromal (WPMY-1) cell lines revealed reductions in the expression of the stem cell marker SOX2
[16]. In prostate stromal cells, DKK3 silencing also reduced the expression of the stem/progenitor cell marker
s-SHIP [56] as well as other putative stem/progenitor cell markers
[16]. Importantly, Dkk3 was identified as one of the top hits in a shRNA screen to identify modulators of mouse embryonic fibroblast reprogramming to generate induced pluripotent stem cells (iPSCs)
[44], providing further evidence for the contribution of fibroblast DKK3 in the regulation of cell fate
[19][27]. Mesenchymal stem cells (MSCs), which are thought to be progenitors of myofibroblasts in the cancer-associated stroma
[61], secrete soluble factors that reduce undesirable immune reactions such as hypersensitivity, autoimmune disease, and graft rejection
[62]. Importantly, MSCs secrete Dkk3
[46], and tumors containing Dkk3
−/− MSCs show increased CD8
+ T cell invasion and reduced M2-type macrophage infiltration, suggesting MSC-derived Dkk3 maintains the immune-suppressive capacity of the tumor microenvironment
[46]. Moreover, MSCs exposed to radiotherapy showed increased expression and secretion of Dkk-3, and MSCs were shown to enhance the effects of radiotherapy on tumor growth in a mouse melanoma xenograft model
[63].