Search

Abstract Background Patients with small intestinal neuroendocrine tumors (SINETs) frequently present with lymph node and liver metastases at the time of diagnosis, but the molecular changes that lead to the progression of these tumors are largely unknown. Sequencing studies have only identified recurrent point mutations at low frequencies with CDKN1B being the most common harboring heterozygous mutations in less than 10% of all tumors. Although SINETs are genetically stable tumors with a low frequency of point mutations and indels, they often harbor recurrent hemizygous copy number alterations (CNAs) yet the functional implications of these CNA are unclear. Methods Utilizing comparative genomic hybridization (CGH) arrays we analyzed the CNA profile of 131 SINETs from 117 patients. Two tumor suppressor genes and corresponding proteins i.e. SMAD4, and CDKN1B, were further characterized using a tissue microarray (TMA) with 846 SINETs. Immunohistochemistry (IHC) was used to quantify protein expression in TMA samples and this was correlated with chromosome number evaluated with fluorescent in-situ hybridization (FISH). Intestinal tissue from a Smad4+/− mouse model was used to detect entero-endocrine cell hyperplasia with IHC. Results Analyzing the CGH arrays we found loss of chromosome 18q and SMAD4 in 71% of SINETs and that focal loss of chromosome 12 affecting the CDKN1B was present in 9.4% of SINETs. No homozygous loss of chromosome 18 was detected. Hemizygous loss of SMAD4, but not CDKN1B, significantly correlated with reduced protein levels but hemizygous loss of SMAD4 did not induce entero-endocrine cell hyperplasia in the Smad4+/− mouse model. In addition, patients with low SMAD4 protein expression in primary tumors more often presented with metastatic disease. Conclusions Hemizygous loss of chromosome 18q and the SMAD4 gene is the most common genetic event in SINETs and our results suggests that this could influence SMAD4 protein expression and spread of metastases. Although SMAD4 haploinsufficiency alone did not induce tumor initiation, loss of chromosome 18 could represent an evolutionary advantage in SINETs explaining the high prevalence of this aberration. Functional consequences of reduced SMAD4 protein levels could hypothetically be a potential mechanism as to why loss of chromosome 18 appears to be clonally selected in SINETs.

Abstract Background Caveolae are membrane invaginations measuring 50–100 nm. These organelles, composed of caveolin and cavin proteins, are important for cellular signaling and survival. Caveolae play incompletely defined roles in human kidneys. Induction of caveolin-1/CAV1 in diseased tubules has been described previously, but the responsible mechanism remains to be defined. Methods Healthy and atrophying human kidneys were stained for caveolar proteins, (caveolin 1–3 and cavin 1–4) and examined by electron microscopy. Induction of caveolar proteins was studied in isolated proximal tubules and primary renal epithelial cells. These cells were challenged with hypoxia or H2O2. Primary tubular cells were also subjected to viral overexpression of megakaryoblastic leukemia 1 (MKL1) and MKL1 inhibition by the MKL1 inhibitor CCG-1423. Putative coregulators of MKL1 activity were investigated by Western blotting for suppressor of cancer cell invasion (SCAI) and filamin A (FLNA). Finally, correlative bioinformatic studies of mRNA expression of caveolar proteins and MKL1 were performed. Results In healthy kidneys, caveolar proteins were expressed by the parietal epithelial cells (PECs) of Bowman’s capsule, endothelial cells and vascular smooth muscle. Electron microscopy confirmed caveolae in the PECs. No expression was seen in proximal tubules. In contrast, caveolar proteins were expressed in proximal tubules undergoing atrophy. Caveolar proteins were also induced in cultures of primary epithelial tubular cells. Expression was not enhanced by hypoxia or free radical stress (H2O2), but proved sensitive to inhibition of MKL1. Viral overexpression of MKL1 induced caveolin-1/CAV1, caveolin-2/CAV2 and SDPR/CAVIN2. In kidney tissue, the mRNA level of MKL1 correlated with the mRNA levels for caveolin-1/CAV1, caveolin-2/CAV2 and the archetypal MKL1 target tenascin C (TNC), as did the MKL1 coactivator FLNA. Costaining for TNC as readout for MKL1 activity demonstrated overlap with caveolin-1/CAV1 expression in PECs as well as in atrophic segments of proximal tubules. Conclusions Our findings support the view that MKL1 contributes to the expression of caveolar proteins in healthy kidneys and orchestrates the induction of tubular caveolar proteins in renal injury.