Machine learning approach to predict body weight in adults.

Background: Obesity is an established risk factor for non-communicable diseases such as type 2 diabetes mellitus, hypertension and cardiovascular disease. Thus, weight control is a key factor in the prevention of non-communicable diseases. A simple and quick method to predict weight change over a few years could be helpful for weight management in clinical settings. Methods: We examined the ability of a machine learning model that we constructed to predict changes in future body weight over 3 years using big data. Input in the machine learning model were three-year data on 50,000 Japanese persons (32,977 men) aged 19-91 years who underwent annual health examinations. The predictive formulas that used heterogeneous mixture learning technology (HMLT) to predict body weight in the subsequent 3 years were validated for 5,000 persons. The root mean square error (RMSE) was used to evaluate accuracy compared with multiple regression. Results: The machine learning model utilizing HMLT automatically generated five predictive formulas. The influence of lifestyle on body weight was found to be large in people with a high body mass index (BMI) at baseline (BMI ≥29.93 kg/m2) and in young people (<24 years) with a low BMI (BMI <23.44 kg/m2). The RMSE was 1.914 in the validation set which reflects ability comparable to that of the multiple regression model of 1.890 (p = 0.323). Conclusion: The HMLT-based machine learning model could successfully predict weight change over 3 years. Our model could automatically identify groups whose lifestyle profoundly impacted weight loss and factors the influenced body weight change in individuals. Although this model must be validated in other populations, including other ethnic groups, before being widely implemented in global clinical settings, results suggested that this machine learning model could contribute to individualized weight management.

ASCL1-coexpression profiling but not single gene expression profiling defines lung adenocarcinomas of neuroendocrine nature with poor prognosis.

BACKGROUND: Lung adenocarcinoma is heterogeneous regarding histology, etiology and prognosis. Although there have been several attempts to find a subgroup with poor prognosis, it is unclear whether or not adenocarcinoma with neuroendocrine (NE) nature has unfavorable prognosis. MATERIALS AND METHODS: To elucidate whether a subtype of adenocarcinoma with NE nature has poor prognosis, we performed gene expression profiling by cDNA microarray for 262 Japanese lung cancer and 30 normal lung samples, including 171 adenocarcinomas, 56 squamous cell carcinomas and 35 NE tumors. A co-expression gene set with ASCL1, an NE master gene, was utilized to classify tumors by non-negative matrix factorization, followed by validation using an ASCL1 knock-down gene set in DMS79 cells as well as an independent cohort (n=139) derived from public microarray databases as a test set. RESULTS: The co-expression gene set classified the adenocarcinomas into alveolar cell (AL), squamoid, and NE subtypes. The NE subtype, which clustered together almost all the NE tumors, had significantly poorer prognosis than the AL subtype that clustered with normal lung samples (p=0.0075). The knock-down gene set also classified the 171 adenocarcinomas into three subtypes and this NE subtype also had the poorest prognosis. The co-expression gene set classified the independent database-derived American cohort into two subtypes, with the NE subtype having poorer prognosis. None of the single NE gene expression was found to be linked to survival difference. CONCLUSION: Co-expression gene set with ASCL1, rather than single NE gene expression, successfully identifies an NE subtype of lung adenocarcinoma with poor prognosis.

Adhesion-induced drug resistance in leukemia stem cells.

The co-culture of TF-1 leukemia cells and MS-5 stromal cells produces a cobblestone area which partially mimics the leukemia stem cell niche. The adhering leukemia cells are shown to become less sensitive to cytarabine, etoposide and daunorubicin. These changes are associated with an increased proportion of the G0/G1 phase, increased upregulation of cyclin-dependent kinase inhibitors, and increased levels of Bcl-2, but not with any change in the expression of BAX or drug transporters such as ABCG2 and MDR1, compared to monocultured leukemic cells. In addition, we demonstrate using a bioimaging technique that daunorubicin accumulates in the lysosomes of the adherent leukemic cells and that V-ATPase is activated. These findings suggest that adhesion alone can lead to drug resistance in leukemic stem cells by various mechanisms.

Adhesion-mediated self-renewal abilities of Ph+ blastoma cells.

The Philadelphia chromosome-positive blastoma, maintained by serial subcutaneous transplantation in nude mice, is a highly proliferating biological mass consisting of homogenous CD34(+)CD38(-) myeloblastoid cells. These cells newly evolved from pluripotent leukemia stem cells of chronic myeloid leukemia in the chronic phase. Therefore, this mass may provide a unique tool for better understanding cellular and molecular mechanisms of self-renewal of leukemia stem cells. In this paper, we demonstrated that intravenously injected blastoma cells can cause Ph+ blastic leukemia with multiple invasive foci in NOD/SCID mice but not in nude mice. In addition, using an in vitro culture system, we clearly showed that blastoma cell adhesion to OP9 stromal cells accelerates blastoma cell proliferation that is associated with up-regulation of BMI1 gene expression; increased levels of beta-catenin and the Notch1 intra-cellular domain; and changed the expression pattern of variant CD44 forms, which are constitutively expressed in these blastoma cells. These findings strongly suggest that adhesion of leukemic stem cells to stromal cells via CD44 might be indispensable for their cellular defense against attack by immune cells and for maintenance of their self-renewal ability.

An evidence for adhesion-mediated acquisition of acute myeloid leukemic stem cell-like immaturities.

For long-term survival in vitro and in vivo of acute myeloid leukemia cells, their adhesion to bone marrow stromal cells is indispensable. However, it is still unknown if these events are uniquely induced by the leukemic stem cells. Here we show that TF-1 human leukemia cells, once they have formed a cobblestone area by adhering to mouse bone marrow-derived MS-5 cells, can acquire some leukemic stem cell like properties in association with a change in the CD44 isoform-expression pattern and with an increase in a set of related microRNAs. These findings strongly suggest that at least some leukemia cells can acquire leukemic stem cell like properties in an adhesion-mediated stochastic fashion.

Centenário da fundação da Comissão Rondon (1907-2007): personagens, descobertas e produção bibliográfica / Centenary of the foundation of the Comissão Rondon (1907-2007): characters, discoveries and bibliographic production

O objetivo deste artigo é dar informações biográficas sobre Cândido Mariano da Silva Rondon (1865-1958), que foi indicado pelo Presidente da República do Brasil, Affonso Augusto Moreira Penna, em 1907, para organizar e conduzir a Comissão de Linhas Telegraphicas Estratégicas de Mato-Grosso ao Amazonas, mais tarde conhecida como Comissão Rondon. Essa Comissão, levantou muitos fatos novos sobre Etnografia, Geografia, Astronomia, Saneamento, Geologia, Mineralogia, Saúde Pública, Zoologia e Botânica dos Estados de Mato Grosso e Amazonas, devido ao próprio Rondon e seus colaboradores - Alípio de Miranda Ribeiro (zoólogo), Frederico Carlos Hoehne (botânico) e alguns outros pesquisadores, cujas biografias são também apontadas. (AU)

Gene expression profiling of the rat endometriosis model.

PROBLEM: To investigate the molecular mechanism of endometriosis, gene expression profiling was analyzed in a rat endometriosis model. METHOD OF STUDY: An endometriosis model was induced by uterine autotransplantation in the peritoneal cavity on a female-SD rat (8 weeks old). As control samples, the normal uterine tissues were used. The gene expression was compared between endometriotic lesions and normal uterine tissues by cDNA microarray analysis, quantitative real time RT-PCR and immunohistochemistry. RESULTS: The expression of 71 genes was upregulated and that of 45 genes was downregulated in the endometriotic lesions compared to normal uterine tissues. The upregulated genes included genes encoding cytokines, chemokines, growth factors and cell adhesion molecules. The levels of transcripts of osteopontin, Lyn, Vav1, Runx1, and l-selectin in the endometriotic lesions were 130, 10, 10, 12 and 46-fold higher than the respective levels in the eutopic endometrial samples. CONCLUSION: The results suggest that osteopontin, Lyn, Vav1, Runx1, and l-selectin play important roles in the pathogenesis of endometriosis.

Two subclasses of lung squamous cell carcinoma with different gene expression profiles and prognosis identified by hierarchical clustering and non-negative matrix factorization.

Current clinical and histopathological criteria used to define lung squamous cell carcinomas (SCCs) are insufficient to predict clinical outcome. To make a clinically useful classification by gene expression profiling, we used a 40 386 element cDNA microarray to analyse 48 SCC, nine adenocarcinoma, and 30 normal lung samples. Initial analysis by hierarchical clustering (HC) allowed division of SCCs into two distinct subclasses. An additional independent round of HC induced a similar partition and consensus clustering with the non-negative matrix factorization approach indicated the robustness of this classification. Kaplan-Meier analysis with the log-rank test pointed to a nonsignificant difference in survival (P = 0.071), but the likelihood of survival to 6 years was significantly different between the two groups (40.5 vs 81.8%, P = 0.014, Z-test). Biological process categories characteristic for each subclass were identified statistically and upregulation of cell-proliferation-related genes was evident in the subclass with poor prognosis. In the subclass with better survival, genes involved in differentiated intracellular functions, such as the MAPKKK cascade, ceramide metabolism, or regulation of transcription, were upregulated. This work represents an important step toward the identification of clinically useful classification for lung SCC.

Notas para a história da medicina no Brasil / Notes for the history of the medicine in Brasil

Cita algumas obras publicadas sobre História da Medicina no Brasil para auxiliar o tratamento das doenças.

Two prognostically significant subtypes of high-grade lung neuroendocrine tumours independent of small-cell and large-cell neuroendocrine carcinomas identified by gene expression profiles.

BACKGROUND: Classification of high-grade neuroendocrine tumours (HGNT) of the lung currently recognises large-cell neuroendocrine carcinoma (LCNEC) and small-cell lung carcinoma (SCLC) as distinct groups. However, a similarity in histology for these two carcinomas and uncertain clinical course have led to suggestions that a single HGNT classification would be more appropriate. Gene expression profiling, which can reproduce histopathological classification, and often defines new subclasses with prognostic significance, can be used to resolve HGNT classification. METHODS: We used cDNA microarrays with 40?386 elements to analyse the gene expression profiles of 38 surgically resected samples of lung neuroendocrine tumours and 11 SCLC cell lines. Samples of large-cell carcinoma, adenocarcinoma, and normal lung were also included to give a total of 105 samples analysed. The data were subjected to filtering to yield informative genes before unsupervised hierarchical clustering that identified relatedness of tumour samples. FINDINGS: Distinct groups for carcinoids, large-cell carcinoma, adenocarcinoma, and normal lung were readily identified. However, we were unable to distinguish LCNEC from SCLC by gene expression profiling. Three independent rounds of unsupervised hierarchical clustering consistently divided SCLC samples into two main groups with LCNEC samples largely integrated with these groups. Furthermore, patients in one of the groups identified by clustering had a significantly better clinical outcome than the other (83% vs 12% survived for 5 years; p=0.0094. None of the highly proliferative SCLC cell lines subsequently analysed clustered with this good-prognosis group. INTERPRETATION: Our findings show that HGNT of the lung can be classified into two groups independent of SCLC and LCNEC. To this end, we have identified many genes, some of which encode well-characterised markers of cancer that distinguish the HGNT groups. These results have implications for the diagnosis, classification, and treatment of lung neuroendocrine tumours, and provide important insights into their underlying biology.

Integrated classification of lung tumors and cell lines by expression profiling.

The utility of cancer cell lines depends largely on their accurate classification, commonly based on histopathological diagnosis of the cancers from which they were derived. However, because cancer is often heterogeneous, the cell line, which also has the opportunity to alter in vitro, may not be representative. Yet without the overall architecture used in histopathological diagnosis of fresh samples, reclassification of cell lines has been difficult. Gene-expression profiling accurately reproduces histopathological classification and is readily applicable to cell lines. Here, we compare the gene-expression profiles of 41 cell lines with 44 tumors from lung cancer. These profiles were generated after hybridization of samples to four replicate 7,685-element cDNA microarrays. After removal of genes that were uniformly up- or down-regulated in fresh compared with cell-line samples, cluster analysis produced four major branch groups. Within these major branches, fresh tumor samples essentially clustered according to pathological type, and further subclusters were seen for both adenocarcinoma (AC) and small cell lung carcinoma (SCLC). Four of eight squamous cell carcinoma (SCC) cell lines clustered with fresh SCC, and 11 of 13 SCLC cell lines grouped with fresh SCLC. In contrast, although none of the 11 AC cell lines clustered with AC tumors, three clustered with SCC tumors and six with SCLC tumors. Although it is possible that preexisting SCC or SCLC cells are being selected from AC tumors after establishment of cell lines, we propose that, even in situ, AC will ultimately progress toward one of two poorly differentiated phenotypes with expression profiles resembling SCC or SCLC.

Inhibition of human endothelial cell proliferation by ShIF, a vacuolar H(+)-ATPase-like protein.

ShIF is a bone marrow stroma cell-derived factor originally identified to support proliferation of bone marrow cells in vitro. This protein shares high sequence homology to the yeast vacuolar H(+)-ATPase subunit, Vph1p, and the 116 kDa proton pump of the rat and bovine synaptic vesicle, Vpp1. We examined the function of ShIF in the proliferation of human umbilical vein endothelial cells (HUVEC). ShIF inhibited HUVEC proliferation in a dose-dependent manner. Recombinant ShIF added at 10 and 20 ng/ml inhibited HUVEC proliferation by 21.6 and 44.3%, respectively and increasing the concentration of ShIF to 100 ng/ml inhibited proliferation by as much as 55.5%. When HUVEC cells were cultured at various concentrations of ShIF in the presence of anti-ShIF antibody, the inhibitory effects of ShIF to HUVEC proliferation were abrogated by 89-91% indicating that the activity of ShIF to HUVEC was specific. HUVEC cultured in the presence of ShIF and bafilomycin, a specific inhibitor of ATPase, resulted to a 90% growth inhibition. Thus, ShIF may act as an antagonist to the ATPase complex by disrupting the production of cellular ATP thereby decreasing the ability of HUVEC to proliferate.

Genetic approach and phenotype-based complementation screening for identification of stroma cell-derived proteins involved in cell proliferation.

The functional capacities of stromal cell lines to support stem cell activity are heterogeneous and the mechanism of how they support bone marrow cultures remains unclear. Recently, we reported a strategy of functional analysis in which a genetic approach is combined with phenotype-based complementation screening to search for a novel secreted growth factor from mouse bone marrow stroma called ShIF that supported proliferation of bone marrow cells. To investigate the role of stromal cells in hemopoiesis, we extended this strategy to search for stroma-derived proteins that induce cell proliferation by establishing stroma-dependent Ba/F3 mutants of three stroma cell lines from two mouse tissues. Seven stroma-dependent Ba/F3 mutants were used as responder cells to identify cDNAs from stroma cell lines whose products supported proliferation not only to the mutant cells but also to hemopoietic progenitor cells in vitro.