Discriminant analysis and machine learning approach for evaluating and improving the performance of immunohistochemical algorithms for COO classification of DLBCL.

BACKGROUND: Diffuse large B-cell lymphoma (DLBCL) is classified into germinal center-like (GCB) and non-germinal center-like (non-GCB) cell-of-origin groups, entities driven by different oncogenic pathways with different clinical outcomes. DLBCL classification by immunohistochemistry (IHC)-based decision tree algorithms is a simpler reported technique than gene expression profiling (GEP). There is a significant discrepancy between IHC-decision tree algorithms when they are compared to GEP. METHODS: To address these inconsistencies, we applied the machine learning approach considering the same combinations of antibodies as in IHC-decision tree algorithms. Immunohistochemistry data from a public DLBCL database was used to perform comparisons among IHC-decision tree algorithms, and the machine learning structures based on Bayesian, Bayesian simple, Naïve Bayesian, artificial neural networks, and support vector machine to show the best diagnostic model. We implemented the linear discriminant analysis over the complete database, detecting a higher influence of BCL6 antibody for GCB classification and MUM1 for non-GCB classification. RESULTS: The classifier with the highest metrics was the four antibody-based Perfecto-Villela (PV) algorithm with 0.94 accuracy, 0.93 specificity, and 0.95 sensitivity, with a perfect agreement with GEP (κ = 0.88, P < 0.001). After training, a sample of 49 Mexican-mestizo DLBCL patient data was classified by COO for the first time in a testing trial. CONCLUSIONS: Harnessing all the available immunohistochemical data without reliance on the order of examination or cut-off value, we conclude that our PV machine learning algorithm outperforms Hans and other IHC-decision tree algorithms currently in use and represents an affordable and time-saving alternative for DLBCL cell-of-origin identification.

Cytotoxic Activity of a Black Bean (Phaseolus vulgaris L.) Extract and its Flavonoid Fraction in Both In Vitro and In Vivo Models of Lymphoma.

BACKGROUND: Black bean (Phaseolus vulgaris L.) is a very common legume seed in Mexican diet. Flavonoids and crude extracts from different plants have been reported as effective agents for chemoprevention and cytotoxicity in several cancer cell lines. We investigated the effects of black bean hulls extract (BBE) and its flavonoid fraction (FF) on lymphoma cells. METHODS: BBE and FF were characterized by high-performance liquid chromatography. Viability and flow cytometry assays were carried out. Finally, a mouse model was generated to test the in vivo effect of both fractions. RESULTS: Both BBE and FF inhibited cell proliferation in a dose-dependent way. In addition, cells underwent apoptosis, and the cellular population at S-phase increased after exposure to these fractions. Furthermore, mice treated with BBE or FF increased the overall survival by 5 or 6 days, respectively, in comparison with a placebo group (p = 0.056). DISCUSSION: BBE and FF had cytotoxic action by driving OCI-Ly7 cells into apoptosis as well as blocking progression to G2/M phase. In addition, BBE and FF treatments were effective in xenograft models.

Entrainment of Breast Cell Lines Results in Rhythmic Fluctuations of MicroRNAs.

Circadian rhythms are essential for temporal (~24 h) regulation of molecular processes in diverse species. Dysregulation of circadian gene expression has been implicated in the pathogenesis of various disorders, including hypertension, diabetes, depression, and cancer. Recently, microRNAs (miRNAs) have been identified as critical modulators of gene expression post-transcriptionally, and perhaps involved in circadian clock architecture or their output functions. The aim of the present study is to explore the temporal expression of miRNAs among entrained breast cell lines. For this purpose, we evaluated the temporal (28 h) expression of 2006 miRNAs in MCF-10A, MCF-7, and MDA-MB-231 cells using microarrays after serum shock entrainment. We noted hundreds of miRNAs that exhibit rhythmic fluctuations in each breast cell line, and some of them across two or three cell lines. Afterwards, we validated the rhythmic profiles exhibited by miR-141-5p, miR-1225-5p, miR-17-5p, miR-222-5p, miR-769-3p, and miR-548ay-3p in the above cell lines, as well as in ZR-7530 and HCC-1954 using RT-qPCR. Our results show that serum shock entrainment in breast cells lines induces rhythmic fluctuations of distinct sets of miRNAs, which have the potential to be related to endogenous circadian clock, but extensive investigation is required to elucidate that connection.

Identification of circadian-related gene expression profiles in entrained breast cancer cell lines.

Cancer cells have broken circadian clocks when compared to their normal tissue counterparts. Moreover, it has been shown in breast cancer that disruption of common circadian oscillations is associated with a more negative prognosis. Numerous studies, focused on canonical circadian genes in breast cancer cell lines, have suggested that there are no mRNA circadian-like oscillations. Nevertheless, cancer cell lines have not been extensively characterized and it is unknown to what extent the circadian oscillations are disrupted. We have chosen representative non-cancerous and cancerous breast cell lines (MCF-10A, MCF-7, ZR-75-30, MDA-MB-231 and HCC-1954) in order to determine the degree to which the circadian clock is damaged. We used serum shock to synchronize the circadian clocks in culture. Our aim was to initially observe the time course of gene expression using cDNA microarrays in the non-cancerous MCF-10A and the cancerous MCF-7 cells for screening and then to characterize specific genes in other cell lines. We used a cosine function to select highly correlated profiles. Some of the identified genes were validated by quantitative polymerase chain reaction (qPCR) and further evaluated in the other breast cancer cell lines. Interestingly, we observed that breast cancer and non-cancerous cultured cells are able to generate specific circadian expression profiles in response to the serum shock. The rhythmic genes, suggested via microarray and measured in each particular subtype, suggest that each breast cancer cell type responds differently to the circadian synchronization. Future results could identify circadian-like genes that are altered in breast cancer and non-cancerous cells, which can be used to propose novel treatments. Breast cell lines are potential models for in vitro studies of circadian clocks and clock-controlled pathways.

A PER3 polymorphism is associated with better overall survival in diffuse large B-cell lymphoma in Mexican population.

BACKGROUND: Diffuse large B-cell lymphoma (DLBCL) is the most common subtype of malignant lymphoma. Presently, one of the most important clinical predictors of survival in DLBCL patients is the International Prognostic Index (IPI). Circadian rhythms are the approximate 24 hour biological rhythms with more than 10 genes making up the molecular clock. OBJECTIVE: Determine if functional single nucleotide polymorphism in circadian genes may contribute to survival status in patients diagnosed with diffuse large B-cell lymphoma. METHODS: Sixteen high-risk non-synonymous polymorphisms in circadian genes (CLOCK, CRY2, CSNK1E, CSNK2A1, NPAS2, PER1, PER2, PER3, PPP2CA, and TIM) were genotyped by screening PCR. Results were visualized by agarose gel electrophoresis and confirmed by two-direction sequencing. Clinical variables were compared between mutated and non-mutated groups. LogRank survival analysis and Kaplan-Meier method were used to calculate the overall survival. RESULTS: PER3 rs10462020 variant showed significant difference in overall survival between patients containing mutated genotypes and those with non-mutated genotypes (p = 0.047). LDH levels (p = 0.021) and IPI score (p < 0.001) also showed differences in overall survival. No clinical differences were observed in mutated vs. non-mutated patients. CONCLUSIONS: This work suggests a role of PER3 rs10462020 in predicting a prognosis in DLBCL overall survival of patients.

A role for 1α,25-dihydroxyvitamin d3 in the expression of circadian genes.

The active form of vitamin D, 1α,25-(OH)2D3, has been associated with metabolism control, cell growth, differentiation, antiproliferation, apoptosis, and adaptive/innate immune responses, besides its functions in the integrity of bone and calcium homeostasis. The circadian rhythm regulates a variety of biological processes, many of them related to the functions associated with 1α,25-(OH)2D3. In the present study, we determine whether 1α,25-(OH)2D3 alters the expression of circadian genes in adipose-derived stem cells (ADSCs). The effect of 1α,25-(OH)2D3 on the expression of circadian genes BMAL1 and PER2 was measured by qPCR, over a 60-h period every 4 h, in serum shocked ADSCs, serum shocked ADSCs supplemented with 1α,25-(OH)2D3, and ADSCs under the presence of only 1α,25-(OH)2D3. The results showed that 1α,25-(OH)2D3 was able to synchronize circadian clock gene expression in ADSCs. The expression of circadian genes BMAL1 and PER2 in ADSCs that contained only 1α,25-(OH)2D3 has a profile similar to that found in the ADSCs synchronized by a serum shock. The results suggest an important role of 1α,25-(OH)2D3 in the regulation of the molecular clock.

Mapping ligand interactions with the hyperpolarization activated cyclic nucleotide modulated (HCN) ion channel binding domain using a soluble construct.

Hyperpolarization activated cyclic nucleotide modulated (HCN) ion channel currents are activated by hyperpolarization and modulated in response to changes in cytosolic adenosine 3',5'-cyclic monophosphate (cAMP) concentrations. A cDNA chimera combining the rat HCN2 cyclic nucleotide binding domain and the DNA binding domain of the cAMP receptor protein (CRP) from E. coli and the histidine tag (HCN2/CRP) was expressed and purified. The construct is capable of forming only non-ligand dependent dimers because the C-linker region of the channel is not present in this construct. The construct binds 8-[[2-[(fluoresceinylthioureido) amino] ethyl] thio] adenosine-3',5'-cyclic monophosphate (8-fluo cAMP) with a Kd of 0.299 microM as determined with a monomer binding model. The Ki values of 20 ligands related to cAMP were measured in order to determine the properties necessary for a ligand to bind to the HCN2 binding domain. This is the first report of cAMP and gunaosine 3',5'-cyclic monophosphate (cGMP) affinities to the HCN2 binding domain being equivalent, even though they modulate the channel with a 10-fold difference in K0.5. Furthermore, the array of ligands measured allows the preference rank order for each purine ring position to be determined: position 1, H > NH2 > O; position 2, NH2 > Cl > H > O; position 6, NH2 > Cl > H > O; and position 8, NH2 > Cl > H > O. Finally, the ability of HCN2/CRP to bind cyclic nucleotide pyrimidine rings at concentrations approximately 1.33 times greater than cAMP suggests that ribofuranose is key for binding.

Proposed structural mechanism of Escherichia coli cAMP receptor protein cAMP-dependent proteolytic cleavage protection and selective and nonselective DNA binding.

The Escherichia coli cAMP receptor protein (CRP) displays biphasic characteristics in protease and beta-galactosidase induction assays at increasing cAMP concentrations in response to ligand binding at the secondary binding site located between the primary binding site and the DNA binding domain. Two mutants were created to determine the mechanistic reason for the CRP biphasic response by inhibiting binding of cAMP to the secondary site via interference with the Arg 181 interaction with the ligand's phosphate. The S179A/R180D/E181H mutant binds two cAMP molecules per dimer, does not exhibit a biphasic response, lacks selective DNA binding, and has inhibited nonselective DNA binding. The R180K mutant binds four cAMP molecules per dimer, exhibits a biphasic response, nonselective DNA binding similar to CRP, but has inhibited selective DNA binding characteristics. The results are consistent with a 2 x 2-binding site scheme were both primary binding sites must be occupied before the secondary binding sites are occupied. A structural mechanism suggesting the secondary sites are formed by binding of cAMP to the primary sites is proposed. AMMP-generated molecular models suggest that R180 orients E181 to produce selective DNA binding, Arg 169 interactions are necessary for nonselective DNA binding, and the position of Leu 57 inhibits chymotrypsin cleavage of Phe 136. DNA binding results suggest that CRP may be the unknown transcription factor which binds to the temperature sensitive dsrA promoter.