Abstract P5-05-04: MALDI mass spectrometry imaging profile of low molecular metabolites in breast carcinoma tissues embedded in frozen tissue microarray

[Background]

Metabolomics is now widely utilized for searching disease markers or identification of new drug targets. In common method, tissue samples originated in human resected specimens are stored by formalin-fixed, paraffin-embedded (FFPE) blocks. However, these samples are inadequate to measure low molecular metabolites or lipids. Furthermore, extraction process that is required for conventional mass-spectrometry causes the loss of information on the spatial localization of the metabolites. In this study, we directly analyzed breast carcinoma tissues embedded in frozen tissue microarray (fTMA) using MALDI mass spectrometry imaging (MALDI-MSI). With our original method, we could obtain profiles of low molecular metabolite and mapping images of several tissues at one time.

[Method]

Six fTMA blocks were constructed by 119 breast tissues (carcinoma 84, normal 35) from 99 patients and sectioned at 10 um thickness. MADLI-MSI were performed by AXIMA Confidence (Shimadzu, Japan) with 9-aminoacridine as a matrix (m/z range:50∼1000, Negative Ion mode). Carcinoma and normal area in individual tissues were confirmed by H&E staining of slide grasses after MADLI-MSI analysis. Acquired MSI data were processed with the freely available software BioMap (Novartis, Switzerland).

[Result]

We could detected 1,915 peaks derive from endogenous metabolite by direct tissue MALDI-MSI analysis of breast carcinoma fTMA. Among them, 185 peaks that could be commonly detected were subjected to further analysis. Among these peaks, we could identify 18 metabolites related to energy metabolism such as ATP. By comparison of metabolite profiles obtained from carcinoma with normal tissues, we found that the energy charge (EC; which is related to ATP, ADP and AMP concentrations) and the sum of adenosine phosphate compound intensities (AXP) were significantly higher than that of normal tissue (EC; T : N = 0.56 : 0.35, AXP; 17453 : 2066, p<0.0001), but there were no significant difference with lymph node metastasis, tumor histological type and tumor size. In comparison with tumor subtype, higher EC was observed in ER(+)/ Her2(-) tumor than others but AXP showed no significant among all subtypes including Ki-67 labeling index.

[Conclusion]

A combination of fTMA and MALDI-MSI is promising approach for biomarker discovery because it can achieve high throughput metabolic mapping without obvious artifact or other problem. In this study, even though high EC value were indicated in carcinoma tissue than normal but newly biomarker candidate was indeterminate at this moment. Identification of the candidates of novel carcinoma tissue biomarker is now underway.

Citation Format: Torata N, Kubo M, Miura D, Ohuchida K, Miyazaki T, Fujimura Y, Hayakawa E, Kai M, Oda Y, Mizumoto K, Hashizume M, Nakamura M. MALDI mass spectrometry imaging profile of low molecular metabolites in breast carcinoma tissues embedded in frozen tissue microarray. [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr P5-05-04.

Over-expression of S100A2 in pancreatic cancer correlates with progression and poor prognosis

Controversy exists regarding the clinical significance of S100A2 in the progression of tumours. In pancreatic cancer, little is known about the role of S100A2. The aim of this study was to clarify the clinical significance of S100A2 expression in pancreatic carcinogenesis. We microdissected invasive ductal carcinoma (IDC) cells from 22 lesions, pancreatic intraepithelial neoplasia (PanIN) cells from five lesions, intraductal papillary mucinous neoplasm (IPMN) cells from 38 lesions, pancreatitis-affected epithelial (PAE) cells from 16 lesions, and normal ductal cells from 18 normal pancreatic tissues. S100A2 expression in 14 pancreatic cancer cell lines, microdissected cells and formalin-fixed paraffin-embedded (FFPE) samples was examined by quantitative reverse transcription-polymerase chain reaction (qRT-PCR). Microdissection analyses revealed that IDC cells expressed higher levels of S100A2 than did IPMN, PAE or normal cells (all comparisons, p < 0.007). Cell lines from metastatic sites expressed higher levels of S100A2 than those from primary sites. PanIN cells expressed higher levels of S100A2 than normal cells (p = 0.018). IDC cells associated with poorly differentiated adenocarcinoma expressed higher levels of S100A2 than did IDC cells without poorly differentiated adenocarcinoma (p = 0.006). Analyses of FFPE samples revealed that levels of S100A2 were higher in samples from patients who survived < 1000 days after surgery than in those from patients who survived > 1000 days (p = 0.043). Immunohistochemical analysis was consistent with qRT-PCR. S100A2 may be a marker of tumour progression or prognosis in pancreatic carcinogenesis and pancreatic cancer.

Sonic hedgehog is an early developmental marker of intraductal papillary mucinous neoplasms: clinical implications of mRNA levels in pancreatic juice

Intraductal papillary mucinous neoplasms (IPMNs) are common cystic tumours of the pancreas. Sonic hedgehog (SHH) is involved in gastric epithelial differentiation and pancreatic carcinogenesis. However, a comprehensive analysis of SHH expression in IPMN has not yet been performed. In the present study, one-step quantitative real-time reverse transcription-polymerase chain reaction with gene-specific priming was used to examine mRNA levels in various types of clinical samples. SHH expression in IPMN was measured and the possible association of gastric epithelial differentiation with development of IPMN was evaluated. In bulk tissue analyses (IPMNs, 11 pancreatic cancer, and 20 normal pancreatic tissues), IPMN expressed significantly higher levels of SHH than did normal pancreas (IPMN versus normal pancreas, p = 0.0025; pancreatic cancer versus normal pancreas, p = 0.0132), but SHH expression did not differ between IPMN and pancreatic cancer (p = 0.3409). In microdissection analyses (infiltrating ductal carcinoma cells from 20 sections, IPMN cells from 20 sections, pancreatitis-affected epithelial cells from 11 sections, and normal epithelial cells from 12 sections), IPMN cells expressed significantly higher levels of SHH than did cancer cells, normal cells, or pancreatitis-affected ductal cells (all comparisons, p < 0.008). Pancreatic juice analyses (20 samples from pancreatic cancers, 31 samples from IPMNs, and 27 samples from chronic pancreatitis) revealed that SHH expression differed significantly between IPMN juice and pancreatitis juice (p < 0.0001), and between cancer juice and pancreatitis juice (p = 0.0125). Receiver operating characteristic curve analyses revealed that SHH measurement in pancreatic juice was useful for discriminating IPMN from chronic pancreatitis (area under the curve = 0.915; 95% confidence interval: 0.796-0.976). The data suggest that overexpression of SHH is an early event in the development of IPMN and that SHH measurement in pancreatic juice may provide some advantages for the treatment or follow-up of a subset of patients with IPMN or chronic pancreatitis.