Wnt-11 Expression Promotes Invasiveness and Correlates with Survival in Human Pancreatic Ductal Adeno Carcinoma

Pancreatic ductal adenocarcinoma (PDAC) is one of the deadliest forms of cancer, proving difficult to manage clinically. Wnt-11, a developmentally regulated gene producing a secreted protein, has been associated with various carcinomas but has not previously been studied in PDAC. The present study aimed to elucidate these aspects first in vitro and then in a clinical setting in vivo. Molecular analyses of Wnt-11 expression as well as other biomarkers involved qRT-PCR, RNA-seq and siRNA. Proliferation was measured by MTT; invasiveness was quantified by Boyden chamber (Matrigel) assay. Wnt-11 mRNA was present in three different human PDAC cell lines. Wnt-11 loss affected epithelial-mesenchymal transition and expression of neuronal and stemness biomarkers associated with metastasis. Indeed, silencing Wnt-11 in Panc-1 cells significantly inhibited their Matrigel invasiveness without affecting their proliferative activity. Consistently with the in vitro data, human biopsies of PDAC showed significantly higher Wnt-11 mRNA levels compared with matched adjacent tissues. Expression was significantly upregulated during PDAC progression (TNM stage I to II) and maintained (TNM stages III and IV). Wnt-11 is expressed in PDAC in vitro and in vivo and plays a significant role in the pathophysiology of the disease; this evidence leads to the conclusion that Wnt-11 could serve as a novel, functional biomarker PDAC.


Introduction
Pancreatic ductal adenocarcinoma (PDAC) is the main form of pancreatic cancer and has one of the highest death rates among human cancers with a 5-year survival of less than 4% [1][2][3]. The high mortality rate is due mainly to difficulties in both diagnosis and treatment [2], as the disease is relatively symptomless until the late stages. Carbohydrate antigen (CA) 19-9 is currently the only available biomarker for PDAC; however, it is not reliable due to its poor sensitivity and specificity [4]. Moreover, Lewis antigen-negative individuals, who compose approximately 5%-10% of PDAC cases,

RNA Sequencing
Total RNA was isolated from cells by using Trizol (Sigma, Gillingham, Dorest. UK). The reverse transcription to cDNA was performed with IonAmpliseqTM Transcriptome Human Gene Expression kit (Life Technologies). We then used the Human Gene Expression Core Panel primer set (Life Technologies) to prepare small amplicon gene expression libraries targeting 20,000 genes (95% of the RefSeq gene database). The cDNA amplicon libraries (125-300 bp) were ligated to adapters and amplified using IonXpress RNA-seq barcoded primers (5 ). cDNA libraries were clonally amplified using an Ion PI template OT2 200 kit (Life technologies, USA) on an Ion OneTouch2 system (Life technologies) as per the manufacturer's instructions. Samples were processed using the Ion Proton 200 sequencing kit and loaded onto a P1 chip and sequenced on an Ion Proton (Life technologies) using default parameters (single-end, forward sequencing). Base calling, adaptor trimming, barcode deconvolution, alignment and Ampliseq gene expression analysis was performed on Torrent Suite version 3.6 (Life technologies).

Invasion and Proliferation Assays
Details of cell invasion assay were as reported previously [16]. Briefly, 5 × 105 cells (transfected or control) were plated on Matrigel-coated transwell filters (BD Biosciences) in a chemotactic gradient of 1:10% FBS. After 16 hours, the total number of invaded cells was determined by MTT assay and this was confirmed by crystal violet assay. In parallel, the same number of cells plated and incubated for 16 h to determine the effect of the cell proliferation by MTT assay. Invasion and proliferation were presented as "Invasion" and "Proliferation", i.e., percentage (%) of the readings for invaded cells/original cell number.

Survival Analysis
Survival data was obtained on 22nd of Oct 2019 from Human Protein Atlas website: https: //www.proteinatlas.org/ENSG00000085741-WNT11/pathology/pancreatic+cancer#Location Wnt-11 expression of total 176 patients was analysed, (low expression; n = 121 and high expression; n = 55) and Kaplan-Meier survival estimators examined the prognosis of each group of patients, and the survival outcomes of the two groups were compared by log-rank tests [34].

Data Analysis
All data were analysed as means ± standard errors. Statistical significance was determined using Student's t-test or ANOVA with Newman-Keuls post-hoc analysis, as appropriate. Results were considered significant for p < 0.05.

Results
Three sets of results are presented. First, we describe the Wnt-11 mRNA profile of three different PDAC cell lines and using siRNA, we demonstrate that Wnt-11 controls the expression of several biomarkers associated with metastasis. Secondly, we show that silencing Wnt-11 expression results in significant inhibition of cellular invasiveness. Thirdly, we demonstrate that Wnt-11 mRNA levels were significantly higher in clinical samples of PDAC compared with matched normal tissues. Overall, these data suggest that Wnt-11 is involves in the pathophysiology of PDAC.

Effects of Silencing Wnt-11 on Biomarkers Associated with Metastasis
Silencing Wnt-11 in Panc-1 cells led to a reduction in mRNA expression to 22 + 2% of the control transfection, (n = 6; p < 0.05; Figure 1A). This was confirmed qualitatively at the protein level by immunocytochemistry ( Figure 1B,C). We then determined whether Wnt-11 was involved in expression of biomarkers associated with metastasis ( Figure 2). Two sets of such biomarkers were studied, those relating to neuronal characteristics and EMT [35,36]. As regards NEMs, Wnt-11 silencing led to decreases of the following (percentage reduction according to the control transfection that was used as a 'baseline' and presented as 100% and statistical significance was determined by using Student's t-test): NSE (53 + 2), NeuroD (89 + 2) and Hes6 (52 + 4); in addition, the stem cell marker Nanog was reduced to 94 + 2 (n = 6; p < 0.01 for all; Figure 2A). Initial qRT-PCR results suggested that Wnt-11 and E-cadherin mRNA levels were inversely correlated in PDAC cells. Indeed, following the Wnt-11 siRNA treatment in Panc-1 cells, the level of E-cadherin increased to 529 + 3% (n = 6; p < 0.01). In contrast, the other three EMT markers studied all showed decreases to the following percentages: Snail (73 + 4), Vim (83 + 3) and Twist (84 + 3) (n = 6; p < 0.01 for all; Figure 2B). Taken together, all results are consistent with Wnt-11 controlling EMT [22,37] and being involved in the promotion of neuronal characteristics [16,37,38].

Effect of Silencing Wnt-11 on Cellular Invasiveness
The invasiveness of Panc-1 cells was studied in Boyden chambers with Matrigel ( Figure 3). Cells demonstrated a noticeable invasion over 16 h. Silencing Wnt-11 resulted in significant suppression of invasiveness by 23 + 2% (n = 3; p < 0.01; Figure 3A). As expected, there was no change in the cell number over the 16 hours ( Figure 3B). Thus, Wnt-11 could promote invasiveness independently of proliferation.

Expression in Biopsies
We also determined the Wnt-11 mRNA levels in clinical samples of PDAC and compared these with 'matched' control tissues ( Figure 4A). In direct agreement with the in vitro data, the average level in PDAC was significantly (four-fold) higher than the matching controls (p = 0.02; n = 111). We then questioned any association between the Wnt-11 mRNA expression and TNM stage ( Figure 4B). Patient samples classified as T1, expressed the least Wnt-11. This was significantly (1931 + 125%) higher for T2 (p = 0.04). This increase was maintained for T3 and T4 with no further change (p > 0.05 for T3 and T4 cf. T2 and T3 cf. T4; Figure 4B).
transwell filters and the extent of invasion determined after 16 hours. (A) Wnt-11 siRNA decreased invasion by 23% (n = 3; p = 0.02). The results are plotted as Invasion Index (InvI, %), which is the percentage of invaded cells compared to the total number of cells seeded. (B) The total cell number/proliferation did not change during the course of the experiment (n = 3; p > 0.05). All data were analysed as means ± standard errors.

Expression in Biopsies
We also determined the Wnt-11 mRNA levels in clinical samples of PDAC and compared these with 'matched' control tissues ( Figure 4A). In direct agreement with the in vitro data, the average level in PDAC was significantly (four-fold) higher than the matching controls (p = 0.02; n = 111). We then questioned any association between the Wnt-11 mRNA expression and TNM stage ( Figure 4B). Patient samples classified as T1, expressed the least Wnt-11. This was significantly (1931 + 125%) higher for T2 (p = 0.04). This increase was maintained for T3 and T4 with no further change (p > 0.05 for T3 and T4 cf. T2 and T3 cf. T4; Figure 4B). In total, 111 pairs of tumour and adjacent control tissues were studied and Wnt-11 expression (relative) appeared to be increased in the tumour tissue compared to adjacent normal tissues (p = 0.02). (B) A high level of Wnt-11 expression (relative) was found in patients classified according to T staging (p < 0.05 for T1 vs T2 or T3 or T4; T2 vs T3 or T3 vs T4 p > 0.05). All data were analysed as means ± standard errors.

Discussion
The main results are as follows: 1) Wnt-11 mRNA expression occurred commonly in PDAC cell lines and was significantly higher in PDAC compared with matched control tissues; 2) Wnt-11 promoted NEM expression and EMT as well as cellular invasiveness without affecting proliferative activity; 3) an increased level of Wnt-11 expression in PDAC patients was associated with TNM stage as a relatively early event.

Wnt-11 Expression in PDAC Cells and Tissues: Control of NEMs, EMT and Invasiveness
We showed that Wnt-11 mRNA is expressed in three different PDAC cell lines. In agreement with this, Wnt-11 mRNA was present in PDAC tissues and this was at a significantly higher level compared with corresponding normal tissues. The in vitro pathophysiological role of the Wnt-11 was investigated by a siRNA approach. Thus, silencing Wnt-11 suppressed the expression of a range of biomarkers associated with metastasis, especially NEMs and those involved in EMT. Such effects Figure 4. Clinical evidence that Wnt-11 was significantly upregulated in a cohort of pancreatic tumour samples. (A) In total, 111 pairs of tumour and adjacent control tissues were studied and Wnt-11 expression (relative) appeared to be increased in the tumour tissue compared to adjacent normal tissues (p = 0.02). (B) A high level of Wnt-11 expression (relative) was found in patients classified according to T staging (p < 0.05 for T1 vs T2 or T3 or T4; T2 vs T3 or T3 vs T4 p > 0.05). All data were analysed as means ± standard errors.

Discussion
The main results are as follows: 1) Wnt-11 mRNA expression occurred commonly in PDAC cell lines and was significantly higher in PDAC compared with matched control tissues; 2) Wnt-11 promoted NEM expression and EMT as well as cellular invasiveness without affecting proliferative activity; 3) an increased level of Wnt-11 expression in PDAC patients was associated with TNM stage as a relatively early event.

Wnt-11 Expression in PDAC Cells and Tissues: Control of NEMs, EMT and Invasiveness
We showed that Wnt-11 mRNA is expressed in three different PDAC cell lines. In agreement with this, Wnt-11 mRNA was present in PDAC tissues and this was at a significantly higher level compared with corresponding normal tissues. The in vitro pathophysiological role of the Wnt-11 was investigated by a siRNA approach. Thus, silencing Wnt-11 suppressed the expression of a range of biomarkers associated with metastasis, especially NEMs and those involved in EMT. Such effects suggest that Wnt-11 would promote PDAC aggressiveness and this was confirmed by Matrigel invasion assays. Interestingly, the inhibitory effect on invasion was seen without any change in proliferative activity, which is consistent with the notion that primary and secondary tumourigenesis (i.e., proliferation vs. invasion) can be controlled differently, even independently [16,39].
Overall, members of the Wnt family of secreted proteins are known to play a significant role in determining the "stemness", EMT, acquisition of neuronal characteristics and invasiveness of carcinomas [16,17,20]. The level of Wnt-11 mRNA expression correlates strongly with levels of neuroendocrine differentiation in cancers of the prostate and breast [16,39]. Wnt-11 also controls expression of a neurone-specific enolase, which drives neuronal differentiation and enhanced cell viability and migration in prostate cancer [16]. These findings constitute strong evidence that Wnt-11 promotes migratory behaviour of cells, including multipotent stem/progenitor cells, during development and cancer progression, including cancers of breast, prostate, colon and cervix [16,19,[39][40][41][42]. This has now been extended to PDAC in vitro and in vivo. Thus, Wnt-11 seems to play a consistent role in the pathophysiology of cancer. Indeed, Wnt-11 is evolutionarily a highly conserved protein, regulated by TGFβ and calcium in a wide variety of cell types [20,42,43].

Mechanistic Aspects
At present, the mechanism(s) underlying the role of Wnt-11 in promoting cancer (including PDAC) cell invasiveness is/are unknown. Overall, Wnt-11 is a member of the non-canonical Wnt family, which is associated with three different signalling pathways: Planar cell polarity (PCP), Ca 2+ and other β-catenin/TCF independent events [44]. Here, insights can be obtained by focusing on Ca 2+ signalling. It is known that the Frizzled family of Wnt receptors, including Wnt-11, can elicit the release of intracellular Ca 2+ [42]. In turn, the rise in Ca 2+ can promote invasiveness by enhancing both cellular motility and proteolysis, integral to the Boyden chamber (Matrigel) assay. For example, Ca 2+ -activated K + channels were shown to enhance motility of Mia-Paca cells [45]. Another pathway could involve activation of ROCK [46,47]. As regards proteolysis, matrix metalloproteinases (MMPs) are known to be upregulated in response to non-canonical Wnt stimulation during development and tissue remodelling and promote cell migration [48,49]. Increased MMP (e.g., MMP-2, MMP-7 and MMP-9) expression occurs frequently upon Wnt stimulation of tumour cell invasion [50][51][52][53]. There is also substantial evidence for multi-modal regulation of MMPs by Ca 2+ [53]. Consistently with the proposed non-canonical Wnt-Ca 2+ association, it has been shown that intracellular Ca 2+ signalling can trigger EMT induction in human cancers [54]. Moreover, TGF-β1 and Ca 2+ have synergic effects in promoting EMT and osteochondral differentiation via Wnt-11 and L-type calcium channels, respectively [22]. Further research is required to determine the precise role of Ca 2+ signalling in the pathophysiology of PDAC.

Wnt-11 Expression in PDAC and Survival
Our tissue analyses complement our in vitro data. Thus, in our cohort of paired human tissue samples, the levels of Wnt-11 mRNA were significantly higher in PDACs compared to their respective normal adjacent tissues. Furthermore, within this tissue set, expression was already high at TNM stage II and maintained at stages III and IV. Previous work on different cancer types also suggested that Wnt family members play a significant role in TNM staging [55,56]. These results are consistent with Wnt-11 expression being a relatively early event in PDAC development/progression. Some of the critical data obtained from the Human Protein Atlas database [34] is that it includes 176 samples from different grading stages at PDAC. All data indicate that higher-grade patients showed diminished survival analysis with up-regulated Wnt-11 expression. Five-year survival for the "higher" group was 26%, compared with 31% for the group with lower Wnt-11 expression (p = 0.02, Figure 5). In the first instance, these data give further credence to the results from the in vitro and tissue analyses. More broadly, as noted above, since functional Wnt-11 expression occurs in other cancers as well, its role in determining survival may be more widespread in oncology [57]. Genes 2019, 10, x FOR PEER REVIEW 9 of 13 5-year survival high 26%; 5-year survival low 31%. [34]. Kaplan-Meier survival estimators examined the prognosis of each group of patients, and the survival outcomes of the two groups were compared. Bylog ranktests. https://www.proteinatlas.org/ENSG00000085741WNT11/pathology/pancreatic + cancer # Location. Accessed on 22 October 2019.

Conclusions
In conclusion, our data are consistent with Wnt-11 occurring early in the development of PDAC and controlling cellular expression of NEMs, EMT and invasiveness. More studies are needed to compare Wnt-11 expression and its role in non-malignant pancreatic cell lines and tissues and its function. Therefore, unsurprisingly, upregulation of Wnt-11 expression ultimately worsens patient survival. Thus, Wnt-11 may be considered as a novel prognostic biomarker for PDAC. Since this is a potential functional biomarker, inhibiting Wnt-11 expression and/or activity may also be of therapeutic importance [58].

Conclusions
In conclusion, our data are consistent with Wnt-11 occurring early in the development of PDAC and controlling cellular expression of NEMs, EMT and invasiveness. More studies are needed to compare Wnt-11 expression and its role in non-malignant pancreatic cell lines and tissues and its function. Therefore, unsurprisingly, upregulation of Wnt-11 expression ultimately worsens patient survival. Thus, Wnt-11 may be considered as a novel prognostic biomarker for PDAC. Since this is a potential functional biomarker, inhibiting Wnt-11 expression and/or activity may also be of therapeutic importance [58].
Supplementary Materials: The following are available online at http://www.mdpi.com/2073-4425/10/11/921/s1, Table S1: Expression analysis of RNA sequencing data. Funding: This research did not receive any specific grant from funding agencies in the public, commercial or non-for-profit sectors.