Secreted frizzled-related protein 1

There are 3 types of tumor suppressor genes:[11]

• Genes that affect cell growth
• Genes that limit the cell cycle and induce apoptosis
• Genes that repair damaged DNA

SFRP1 appears to fall in the first category of genes, those that affect cell growth.

The role of SFRP1 as a tumor suppressor has been proposed in many cancers, based on its loss in patient tumors. Its frequent inactivation by methylation-induced silencing is consistent with it behaving as a tumor suppressor.[12] Also, the SFRP1 gene is located in a region on chromosome 8 that is frequently lost in many cancer types.[13] Expression levels of several targets of the Wnt signaling pathways are increased in tumor tissue compared with normal, and the expression of SFRP1 is lost in patient tumor samples. The role for the Wnt/β-catenin signaling in cancer has been well defined: β-catenin drives transcription of genes that contribute to the tumor phenotype by regulating processes such as proliferation, survival and invasion.[12]

Gumz et al. showed that SFRP1 expression in UMRC3 cells (clear cell renal cell carcinoma cell line) resulted in a growth-inhibited phenotype. SFRP1 expression not only reduced the expression of Wnt target genes, but also markedly inhibited tumor cell growth in culture, soft agar and xenografts in athymic nude mice. Growth in culture and anchorage-independent growth were inhibited in SFRP1-expressing UMRC3 cells. The growth-inhibitory effects of SFRP1 were due primarily to decreased cell proliferation rather than an increase in apoptosis.[12] This was consistent with the effect of SFRP1 on cellular proliferation as seen in prostate cancer, where retroviral-mediated expression of SFRP1 resulted in inhibited cellular proliferation but had no effect on apoptosis.[14] Also, restoration of SFRP1 expression attenuated the malignant phenotype of cRCC; moreover, other studies showed reexpression of SFRP1 resulted in decreased colony formation in colon and lung cancer models.[15][16]

The Wnt signaling pathways are initiated by the binding of the Wnt ligand to the Fz receptor. There are three different molecular pathways downstream of the Wnt/Fz interaction. The majority of research has focused on the Wnt/β-catenin pathway (also known as the "canonical" Wnt pathway), which manages cell fate determination by regulating gene expression. The Wnt/Ca²⁺ and Wnt/polarity pathways are known as the "non-canonical pathways". The decision of which pathway is activated most likely depends on which Wnt ligand and Fz receptor are present, as well as the cellular context. Nineteen Wnt ligands and ten different members of the Fz seven-transmembrane receptor family have been described in the human genome. As a result, a large variety of responses could be initiated from the Wnt/Fz interactions.[17]

The Wnt/β-catenin pathway starts with the binding of Wnt to a receptor complex encompassing a Fz receptor and LRP co-receptor. After Wnt binds, an intracellular protein named Dishevelled (Dvl) is activated via phosphorylation. β-catenin degradation complexes in the cytoplasm are composed of adenomatous polyposis coli (APC), glycogen synthase kinase 3β (GSK3β) and Axin. APC promotes the degradation of β-catenin by increasing the affinity of the degradation complex to β-catenin. Axin is a scaffolding protein which holds the degradation complex together. The activated Dvl associates with Axin and prevents GSK3β and casein kinase 1α (CK1α) from phosphorylating critical substrates, such as β-catenin. Phosphorylation of β-catenin marks the protein for ubiquitylation and rapid degradation by proteasomes. Thus, the binding of Wnt to the receptor results in a non-phosphorylated form of β-catenin which localizes to the nucleus and, after displacing the Groucho corepressor protein, forms a complex with Tcf/Lef transcription factors and co-activators (such as CREB binding protein) and induces the expression of downstream target genes.[17]

β-catenin is actively stabilized in over 50% of breast cancers and its nuclear localization correlates with poor patient prognosis. Several target genes of the Wnt signaling pathway, such as cyclin D1, are activated in a significant proportion of breast tumours.[6] It has been shown that SFRP1 transcription can be driven by B-catenin in normal intestinal epithelial cells. Neoplastic epithelial cells were treated with lithium chloride, which inhibits GSK3B and thus stabilizes B-catenin. Lithium chloride is widely used to mimic Wnt signaling. Rather than suppressing SFRP1 expression, B-catenin/TCF activity was associated with the induction of SFRP1. This is consistent with a negative feedback response restricting the exposure of a normal cell to a prolonged Wnt growth factor signal.[18]

Hedgehog signaling in the intestinal epithelium represses the canonical Wnt signaling to restrict expression of Wnt target genes to stem or progenitor cells. It was thought that the Hedgehog signaling pathway does this via the induction of the secreted-type Wnt inhibitor. Katoh et al. searched for the GLI-binding site within the promoter region of Wnt inhibitor genes. GLI are transcription factors that activate the transcription of Hedgehog target genes. The GLI-binding site was identified within the 5’-flanking promoter region of the human SFRP1 gene. The GLI-binding site was conserved among promoter regions of mammalian SFRP1 orthologs. These facts indicate that the SFRP1 gene was identified as the evolutionarily conserved target of the Hedgehog-GLI signaling pathway. SFRP1 was found to be expressed in mesenchymal cells. Hedgehog is secreted from differentiated epithelial cells to induce SFRP1 expression in mesenchymal cells, which keeps differentiated epithelial cells away from the effect of canonical Wnt signaling. Thus, SFRP1 is most likely the Hedgehog target to confine canonical Wnt signaling within stem or progenitor cells. Epigenetic CpG hypermethylation of the SFRP1 promoter during chronic persistent inflammation and aging leads to the occurrence of gastrointestinal cancers, such as colorectal cancer and gastric cancer, through the breakdown of Hedgehog-dependent Wnt signal inhibition.[19]

Three out of 10 advanced colorectal tumors had mutations leading to premature termination of the SFRP1 translation product. The mutations were two single-base deletions (26delG and 67delG) and a single-base change (G450A), which generates an in-frame stop codon. These three mutations were found within the first exon, which was shown previously to be sufficient for Wnt antagonist activity by itself [26, 32]. Of the 10 tumors analyzed, no truncating mutations were found in the second or third exons of SFRP1.[13]

An additional 51 tumors were analyzed via direct sequence analysis, yielding 49 clearly interpretable results. Only the first exon was sequenced for stop codon mutations, but none were found. This indicates that point mutation is not a frequent method of inactivation of the SFRP1 gene in colorectal cancer.[13]

The primary translation product of SFRP1 contains an atypical signaling sequence, where a chain of 15 hydrophilic amino acids precede the hydrophobic domain. Looking at 7 tumors without the truncating mutation, the retained SFRP1 allele contained an in-frame three-base insertion after nucleotide 37. This is thought to lead to an extra alanine in the protein after codon 13. However, no significant association was found between the development of colorectal cancer and the presence of the 3-bp insertion.[13]

An unspliced form of SFRP1 is the dominant form in the lung and liver, leading to an extended protein. This extended sequence contains a hydrophobic region that may act as a transmembrane anchor, modifying the localization of the protein. This may then influence the function of SFRP1 in different tissues because an untethered protein may be more effective in antagonizing Wnt signaling to tumor cells than would a membrane-bound form.[13]

• NSCLC (48% of investigated cell lines)[15]
• SCLC (38% of investigated cell lines)[15]
• Bladder cancers (38%)[20]
• Breast cancers (46%)[21]
• Malignant mesotheliomas (48%)[22]
• Colorectal cancers (76%)[13]

DNA methylation involves the addition of a methyl group to the carbon-5 position of the cytosine ring in the CpG dinucleotide and converting it to methylcytosine. This process is catalyzed by DNA methyltransferase. In numerous cancers, the CpG islands of selected genes are aberrantly methylated (hypermethylated) which results in transcriptional repression. This may be an alternate mechanism of gene inactivation.[7]

Multiple genes have been discovered to be frequently methylated in cancers and leukemias.[23][24] More specifically, the deregulation of the Wnt signaling pathway has been implicated in a wide array of cancers[25][26] that is mainly seen as a result of loss-of-function mutations of APC and axin or as a gain-of-function mutation of CTNNB1 (B-catenin).[25][27] The GC content of the SFRP1 promoter in humans is 56.3%.[19]

It has been found that the overexpression of B-catenin may lead to enhanced proliferation in myeloma plasma cells; thus, soluble Wnt inhibitors are potential tumor suppressor genes and, if inactivated, may contribute to myeloma pathogenesis. This led Chim et al. to investigate the role of aberrant gene methylation of a panel of soluble Wnt antagonists, including SFRP1. Complete methylation led to silencing of respective genes (no transcripts), whereas absence of gene methylation was associated with constitutive gene expression. Methylation of soluble Wnt antagonists would be important in the pathogenesis of multiple myeloma if Wnt signaling was regulated by an autocrine loop by Wnt and Fz. If an autocrine loops exists, then both the ligand (Wz) and receptor (Fzd) should be simultaneously expressed in myeloma cells and growth of tumour cells should be inhibited upon addition of SFRP1. Chim et al. demonstrated simultaneous expression of Wz and Fzd in myeloma plasma cells. Moreover, treatment with recombinant SFRP1 inhibited the growth of myeloma cells in a dose-dependent manner. These findings implicate soluble Wnt inhibitors as tumor suppressors that could be inactivated by methylation.[7]

Veeck and colleagues found all of their eight breast cancer cell lines had complete methylation in the SFRP1 promoter region, while no methylation was detectable in non-malignant cell lines. After treatment with 5-Aza-2’-deoxycytidine (DAC), an inhibitor of DNA methyltransferase, SFRP1 expression was restored in all four treated breast cancer cell lines, supporting the hypothesis of methylation-mediated SFRP1 gene silencing in breast cancer.[6]

Furthermore, the transcriptional silencing mechanism underlying DNA methylation which is brought about through the hypermethylation of CpG-rich islands present in the promoter region of genes, can cooperate with histone deacetylation to change chromatin structure to a repressed form. Lo and colleagues looked at the effects of DAC and trichostatin A (TSA, selectively inhibits the mammalian histone deacetylase family of enzymes) on cancer cells. In 4 breast cancer cell lines, SFRP1 expression was significantly restored after treatment with DAC alone. TSA, only in combination with DAC, had a slightly enhanced effect on SFRP1 expression in these cell lines. A different breast cancer cell line (SKBR3, showed loss of SFRP1 expression without significant methylation of the SFRP1 promoter. Lo et al. hypothesized that this may be due to silencing via histone deacetylation. After SKBR3 cells were treated with TSA, SFRP1 expression was restored in a dose- and time-dependent manner. Yet another breast cancer cell line (T47D) required both DAC and TSA to upregulate SFRP1 expression. This indicates that T47D cells are tightly regulated by two layers of epigenetic control (DNA methylation and histion deacetylation) and relieving inhibition by both mechanisms is necessary for reactivation of SFRP1. This study shows that both the epigenetic mechanisms, DNA methylation and histone deacetylation, are involved in silencing of SFRP1.[28]

Twenty-three out of 25 patients showed high expression of SFRP1 mRNA in leiomyoma than the matched normal myometrium. During the menstrual cycle, the level of SFRP1 mRNA in leiomyoma was highest in the follicular phase. Gonadotropin releasing hormone analogue (GnRHa) decreases estrogen secretion from the ovary. Patients treated with (GnRHa) presurgically showed the lowest expression of SFRP1 in both myometrial and leiomyoma tissues. These findings suggest that SFRP1 could be under the control of estrogen. Gene expression of estrogen receptors in leiomyomas is stronger than that in the myometrium. This suggests that leiomyoma possess increased sensitivity to E2 (estradiol, a form of estrogen) and the estrogen-dependent expression of SFRP1 in leiomyoma could be associated with the growth and pathogenesis of leiomyoma.[29]

Smooth muscle cells cultured from the myometrium showed no significant induction of SFRP1 mRNA in response to treatment with E2 and/or progesterone. Conversely, cells cultured from leiomyomas showed significant dose-dependent induction of SFRP1 mRNA in response to treatment with E2; however, progesterone had no effect on SFRP1 even when coapplied with E2.[29]

Both hypoxic conditions and serum deprivation induced increased expression of SFRP1 in leiomyoma cells. However, the smooth muscle cells cultured from the myometrium showed no significant correlation between SFRP1 expression and oxygen concentration. This suggests that SFRP1 may protect the cells from the damage caused by these stresses.[29]

Aberrant promoter hypermethylation of SFRP1 occurs frequently during the pathogenesis of human cancers and has been found to be one of the primary mechanisms in SFRP1 down-regulation. Methylation-specific PCR (MSP) is able to detect this epigenetic change and could be used for cancer detection.[35] Detection and quantification of promoter CpG methylation in body fluid is both feasible and noninvasive. Combined MSP analyses of multiple genes in voided urine could provide a reliable way to improve cancer diagnosis.[36]

Urakami et al. were able to detect cancers cells using conventional MSP analysis of Wnt-antagonist genes (including SFRP1) in voided urine of patients with bladder tumor. Their results showed a high percentage of identical methylation with tumor-tissue DNA. Conversely, no aberrant methylation was detected in >90% of urine DNA from normal controls. This demonstrates that methylation detection of SFRP1 is both feasible and reliable and that the urine methylation score (M score) of Wnt antagonist genes could be used as an excellent noninvasive diagnostic biomarker for bladder tumor. Furthermore, the M score of Wnt-antagonist genes may reflect the presence of bladder tumor that progresses to invasive disease that would signal for future aggressive treatment. An optimal hypermethylation panel of Wnt-antagonist genes could contribute significantly to early detection of bladder tumor and predict bladder tumor aggressiveness. In fact, the methylation of SFRP1 genes in fecal DNA isolated from stool samples has been used to screen for colorectal cancer.[37]

Immunotherapy is a treatment used to produce immunity to a disease or enhance the resistance of the immune system to an active disease process, such as cancer

Wnt and Fz genes are frequently overexpressed in head and neck squamous cell carcinoma (HNSCC). Treatment of a HNSCC cell line (SNU 1076) with anti-Wnt1 antibodies reduced the activity of the Wnt/Fz dependent transcription factor LEF/TCF and diminished the expression of cyclin D1 and B-catenin proteins. Similar to anti-Wnt antibodies, treatment with recombinant SFRP1 inhibited growth of SNU 1076 cells as well. This suggests that Wnt and Fz receptors may be attractive targets for immunotherapy and drug therapy of HNSCC.[38]

Epigenetic Therapy is the use of drugs or other epigenome-influencing techniques to treat medical conditions.

It is recently regarded as promising therapy to NSCLC.[39] As can be seen in the above, SFRP1 was downregulated epigenetically in NSCLC and was recently proposed as one of epigenetic therapy target.[40]