Probes for GATA6

Relevant protein expression of GATA6, CK5, vimentin, and mucins using immunohistochemistry was assessed for predicting the prognosis and chemotherapy efficacy in pancreatic cancer (PC). The protein expression was examined in 159 PCs resected after neoadjuvant chemotherapy (NAC-PCs) with 120 matched biopsy specimens taken before NAC. KRAS mutations were assessed by digital PCR. NAC-PCs were classified by GATA6 expression initially and CK5 expression subsequently into four types, i.e., classical type (n = 22) showing GATA6-high (≧ 50%)/CK5-low (< 10%), hybrid type (n = 45) showing GATA6-high/CK5-high (≧ 10%), basal-like type (n = 53) showing GATA6-low (< 50%)/CK5-high (≧ 30%), and null type (n = 39) showing GATA6-low/CK5-low (< 30%), which resulted in a well-stratification of the patients’ prognosis. The classical type showed the most favorable prognosis, while the null type showed the worst prognosis (multivariate hazard ratio 3.56, 95% confidence interval (CI) 1.63−7.77, p = 0.0015). The hybrid and basal-like types were in between. The risk for hepatic recurrence was lower in the classical type than null (multivariate odds ratio (mOR) 0.18, CI 0.04−0.96, p = 0.0449) and basal-like (mOR 0.24, CI 0.05−1.16, p = 0.0750) types. In contrast, the risk for loco-regional recurrence was higher in classical type than the basal-like type (mOR 5.03, CI 1.20−21.1, p = 0.0272). The hybrid type was subclassified into transition and co-expression patterns with different gastric mucin expression. Vimentin-high (≧ 10%, n = 30) in pre-NAC-PC tissues was associated with poor prognosis (p = 0.0256). Phenotypic transitions between pre- and post-NAC were common (73/120; 61%). PCs with NAC regression grades 2 and 3 showed a transition to poorer prognostic phenotypes (p = 0.0497). KRAS mutations were not associated with these phenotypes. In conclusion, GATA6 and CK5 immunohistochemical expression phenotypes may stratify the survival of NAC-PCs and reflect post-NAC phenotypic transitions associated with poor prognosis. Prompt evaluation of immunohistochemical phenotypes may contribute to designing a precision therapeutic strategy for PC patients.

Dickkopf-2 (DKK2) has been described as Wnt/beta-catenin pathway antagonist and its expression is mediated by micro RNA-221 (miRNA-221). So far, there is only limited data characterizing the role of DKK2 expression in esophageal cancer. A tissue micro array of 192 patients with esophageal adenocarcinoma was analyzed immunohistochemically for DKK2, miRNA-221 expression by RNA scope, and GATA6 amplification by fluorescence in-situ hybridization. The data was correlated with clinical, pathological and molecular data (TP53, HER2, c-myc, GATA6, PIK3CA, and KRAS amplifications). DKK2 expression was detectable in 21.7% and miRNA-221 expression in 33.5% of the patients. We observed no correlation between DKK2 or miRNA-221 expression and clinico-pathological data DKK2 expression was correlated with TP53 mutations and amplification of GATA6. We did not detect a survival difference in dependence of DKK2 for the total cohort, however, in patients without neoadjuvant treatment DKK2 expression correlated with a prolonged survival (median overall-survival 202 vs. 55 months, p = 0.012) which turned opposite in patients that underwent neoadjuvant treatment. High amounts of miRNA-221 were in trend associated with a prolonged overall-survival (p = 0.070). DKK2 as a Wnt antagonist is associated with prolonged survival in patients without neoadjuvant treatment and changes its prognostic value to the contrary in patients after neoadjuvant therapy. The modulatory effects of neoadjuvant treatment in connection with DKK2 expression are not fully understood, but when considering DKK2 as a tumor marker, it is necessary to see it in the context of neoadjuvant therapy

A,B, Expression correlation analysis in 78 PDAC patient tumors12 [/articles/s43018-021-00258-w#ref-CR12] (E-MTAB-6134 [http://www.ebi.ac.uk/arrayexpress/experiments/E-MTAB-6134/]) with high tumor cellularity between cJUN and VIM (A) as well as GATA6 and VIM (B). RMA-normalized probe intensity values were plotted. A linear regression with 95% CI is shown in orange. Pearson’s correlation (_R_) and corresponding two-tailed _P_ value are indicated. C, Representative bright-field images of GCDX62 cells transduced with empty vector (EV) or cJUN overexpression (cJUN-OE) constructs. Morphology was monitored over several passages. D-F, RNA-seq analysis was performed on GCDX62 cells transduced with EV or cJUN-OE. n = 3 independent cultures. D, PCA plot. E,F, Enrichment plots for gene set enrichment analysis between cJUN-OE and EV samples for ‘classical’ and ‘quasi-mesenchymal’ PDAC13 [/articles/s43018-021-00258-w#ref-CR13] (E), as well as the top genes up- and downregulated following TNFα treatment in CLA (CAPAN1) cells (F). G, WB for indicated targets in CAPAN1 cells transduced with EV or cJUN-OE. Representative of n = 3 independent experiments. H, Representative bright-field images of CAPAN1 cells transduced with EV or cJUN-OE. Morphology was monitored over several passages. C,H, Scale bar: 200 µm. I,J, Trans-well invasion assay of CAPAN1 cells transduced with EV or cJUN-OE, showing representative DAPI staining of invaded cells (I) as well as quantification thereof (J). I, Scale bar: 100 µm. J, Data given as average counts per F.o.V., with means ± s.d. Unpaired, two-tailed Student’s t-test. n = 3 independent experiments. K-M, Mean cell viability ± s.d. at different concentrations of gemcitabine (K), oxaliplatin (L) and SN38 (M) in CAPAN1 cells transduced with EV or cJUN-OE. IC50 values for each drug are indicated. n = 3 independent experiments.