Dopamine and cAMP regulated phosphoprotein MW 32 kDa is overexpressed in early stages of gastric tumorigenesis

Abstract

Background: Gastric adenocarcinoma is a leading cause of cancer mortality. The role of dopamine and cAMP regulated phosphoprotein MW 32 kDa (DARPP-32) overexpression in the gastric tumorigenesis cascade remains unclear.

Methods: The expression of DARPP-32 in the multistep carcinogenesis cascade was examined using immunohistochemistry analysis on 533 samples. The contribution of DARPP-32 in cellular transformation and molecular signaling was investigated using NIH3T3, AGS, and SNU16 cells.

Results: The composite expression score (CES), calculated from immunostaining patterns, increased significantly from normal or gastritis to metaplasia, dysplasia, and adenocarcinoma (P < .001). In patients with normal stomach or gastritis and tumor samples, a 76% and 77% chance, respectively, was found (P < .001) that CES was higher in the tumor. High median CES correlated with well- or moderately differentiated (P = .03) gastric adenocarcinomas. NIH3T3 cells transfected with DARPP-32 demonstrated increased levels of phospho-AKT and a 5-fold increase in the number of foci as compared with the control (P = .02). DARPP-32 expression in AGS cells led to increased protein levels of phospho-AKT and BCL-2. For validation, the knockdown of endogenous DARPP-32 expression in SNU16 cells using shRNA resulted in decreased levels of phospho-AKT phosphorylation and BCL-2.

Conclusion: Our results suggest that DARPP-32 overexpression may participate in the transition to intestinal metaplasia and in the progression to neoplasia. The ability of DARPP-32 to transform NIH3T3 cells and to regulate AKT and BCL-2 underscores its possible oncogenic potential.

Figure 1. Immunohistochemistry analysis of DARPP-32 on tissue microarray

DARPP-32 immunohistochemistry was performed on tissue microarrays that contained 533 paraffin-embedded tissue samples (117 normal, 57 gastritis, 88 metaplasia, 30 dysplasia, and 241 adenocarcinoma). A) Samples of normal stomach adjacent to adenocarcinoma had absent to low DARPP-32 immunostaining. B) A proportion of samples with gastritis had elevated DARPP-32 expression compared to the normal samples, although the difference in median DARPP-32 CES was not statistically significant. Samples with intestinal metaplasia (C), dysplasia (D), and intestinal- (E) and diffuse-type (F) adenocarcinoma had high levels of DARPP-32 expression.

Figure 2. The pattern of DARPP-32 expression in gastric tumorigenesis

The intensity of shading in these diagrams represents the proportion of samples with a particular intensity and frequency. Samples with histologically normal tissue adjacent to adenocarcinoma had low to absent levels of DARPP-32 intensity and frequency by immunohistochemistry. A larger proportion of gastritis samples have elevated DARPP-32 intensity and frequency. With metaplastic samples, both the intensity and frequency of DARPP-32 immunostaining is high. This statistically significant trend is maintained in dysplastic and tumor samples, although there is a subset of adenocarcinoma samples (54/241, 24%) with no significant DARPP-32 expression.

Figure 3. Composite Expression Score Analysis

After calculation of the DARPP-32 composite expression score (CES, Table 1) it is possible to graph DARPP-32 expression on a single axis correlated to histology. A) Samples with metaplasia, dysplasia, and adenocarcinoma have significantly higher levels of DARPP-32 expression than samples of normal tissue or tissue with gastritis (p < 0.001). B) Median DARPP-32 CES significantly increases between normal (5 [1– 6]) and gastritis (5 [1– 6]) samples and samples in the metaplasia (10 [10– 11]), dysplasia (10 [7– 11]), and adenocarcinoma (10 [6– 11]) categories (p<0.001) Metaplasia and dysplasia have similar levels of DARPP-32 expression, as do dysplasia and adenocarcinoma. C) Well- and moderately-differentiated adenocarcinomas have higher CES than poorly differentiated adenocarcinomas (p = 0.03). D) There is a trend toward increased levels of DARPP-32 expression in intestinal-type adenocarcinomas compared to diffuse-type adenocarcinomas (p = 0.06). E) Analysis of paired samples in patients with a tissue sample of adenocarcinoma and at least one non-tumor sample was performed. These formed three groups, patients with normal and adenocarcinoma (n = 80), gastritis and adenocarcinoma (n = 50), and metaplasia and adenocarcinoma (n = 77). The hypothesis that the probability of an increase or maintenance in the level of DARPP-32 protein expression was greater than 0.50 during tumor progression was tested. Normal and gastritis samples had a probability of 0.77 (95% CI 0.66–0.85, p < 0.001) and 0.76 (95% CI 0.62–0.87, p < 0.001) respectively that DARPP-32 CES would be maintained or increase during the course of tumor progression. Paired samples of metaplasia and adenocarcinoma indicated that there was no significant additional increase in DARPP-32 expression in the transition from metaplasia to adenocarcinoma, as the probability was 0.37 (95% CI 0.26–0.48, p = 0.99). In few patients, the DARPP-32 expression decreased from metaplasia to adenocarcinoma and these were more likely to have a poorly differentiated tumor (p = 0.002).

Figure 4. DARPP-32 overexpression mediates cellular transformation and phosphorylation of AKT

A–B) A focus formation assay was performed using low passage NIH 3T3 cells. Cells transfected with DARPP-32 had 76.7 ± 17.6 foci per 10 cm plate as compared to 16.7 ± 5.0 foci per plate for empty vector control (p = 0.02). C) Western blotting was performed to confirm the expression of DARPP-32 and normalized to Actin. DARPP-32 transfected cells had higher levels of phosphorylated AKT at the Ser473 residue than vector control. D) AGS cells stably expressing a tetracycline-inducible DARPP-32 were exposed to 24 hours of doxycycline (lane 2) and followed for subsequent time points. There was a specific induction of DARPPP-32 protein expression with sequential increases in the levels of phospho- AKT and BCL-2 (lanes 4–7). The DARPP-32 and phosphorylated AKT levels returned to base line at 72 hours following the withdrawal of doxycycline (lane 8). E) SNU-16 cells stably expressing doxycycline-inducible DARPP-32-shRNA were exposed to doxycycline for 48 hours prior to obtaining cell lysates for Western blotting. The induction of DARPP-32-shRNA led to a remarkable decrease in the levels of DARPP-32, phospho-AKT and BCL-2.