Genome-wide identification of expression quantitative trait loci (eQTLs) in human heart

In recent years genome-wide association studies (GWAS) have uncovered numerous chromosomal loci associated with various electrocardiographic traits and cardiac arrhythmia predisposition. A considerable fraction of these loci lie within inter-genic regions. The underlying trait-associated variants likely reside in regulatory regions and exert their effect by modulating gene expression. Hence, the key to unraveling the molecular mechanisms underlying these cardiac traits is to interrogate variants for association with differential transcript abundance by expression quantitative trait locus (eQTL) analysis. In this study we conducted an eQTL analysis of human heart. For a total of 129 left ventricular samples that were collected from non-diseased human donor hearts, genome-wide transcript abundance and genotyping was determined using microarrays. Each of the 18,402 transcripts and 897,683 SNP genotypes that remained after pre-processing and stringent quality control were tested for eQTL effects. We identified 771 eQTLs, regulating 429 unique transcripts. Overlaying these eQTLs with cardiac GWAS loci identified novel candidates for studies aimed at elucidating the functional and transcriptional impact of these loci. Thus, this work provides for the first time a comprehensive eQTL map of human heart: a powerful and unique resource that enables systems genetics approaches for the study of cardiac traits.

Reduced activity of double-strand break repair genes in prostate cancer patients with late normal tissue radiation toxicity

PURPOSE

To investigate clinical parameters and DNA damage response as possible risk factors for radiation toxicity in the setting of prostate cancer.

METHODS AND MATERIALS

Clinical parameters of 61 prostate cancer patients, 34 with (overresponding, OR) and 27 without (non-responding, NR) severe late radiation toxicity were assembled. In addition, for a matched subset the DNA damage repair kinetics (γ-H2AX assay) and expression profiles of DNA repair genes were determined in ex vivo irradiated lymphocytes.

RESULTS

Examination of clinical data indicated none of the considered clinical parameters to be correlated with the susceptibility of patients to develop late radiation toxicity. Although frequencies of γ-H2AX foci induced immediately after irradiation were similar (P=.32), significantly higher numbers of γ-H2AX foci were found 24 hours after irradiation in OR compared with NR patients (P=.03). Patient-specific γ-H2AX foci decay ratios were significantly higher in NR patients than in OR patients (P<.0001). Consequently, NR patients seem to repair DNA double-strand breaks (DSBs) more efficiently than OR patients. Moreover, gene expression analysis indicated several genes of the homologous recombination pathway to be stronger induced in NR compared with OR patients (P<.05). A similar trend was observed in genes of the nonhomologous end-joining repair pathway (P=.09). This is congruent with more proficient repair of DNA DSBs in patients without late radiation toxicity.

CONCLUSIONS

Both gene expression profiling and DNA DSB repair kinetics data imply that less-efficient repair of radiation-induced DSBs may contribute to the development of late normal tissue damage. Induction levels of DSB repair genes (eg, RAD51) may potentially be used to assess the risk for late radiation toxicity.

NetWeAvers: an R package for integrative biological network analysis with mass spectrometry data

Summary: The discovery of functionally related groups in a set of significantly abundant proteins from a mass spectrometry experiment is an important step in a proteomics analysis pipeline. Here we describe NetWeAvers (Network Weighted Averages) for analyzing groups of regulated proteins in a network context, e.g. as defined by clusters of protein–protein interactions. NetWeAvers is an R package that provides a novel method for analyzing proteomics data integrated with biological networks. The method includes an algorithm for finding dense clusters of proteins and a permutation algorithm to calculate cluster P-values. Optional steps include summarizing quantified peptide values to single protein values and testing for differential expression, such that the data input can simply be a list of identified and quantified peaks.

Availability and implementation: The NetWeAvers package is written in R, is open source and is freely available on CRAN and from netweavers.erasmusmc.nl under the GPL-v2 license.

Contact: e.mcclellan@erasmusmc.nl

Supplementary information:Supplementary data are available at Bioinformatics online.

Consensus and conflict cards for metabolic pathway databases

Background

The metabolic network of H. sapiens and many other organisms is described in multiple pathway databases. The level of agreement between these descriptions, however, has proven to be low. We can use these different descriptions to our advantage by identifying conflicting information and combining their knowledge into a single, more accurate, and more complete description. This task is, however, far from trivial.

Results

We introduce the concept of Consensus and Conflict Cards (C₂Cards) to provide concise overviews of what the databases do or do not agree on. Each card is centered at a single gene, EC number or reaction. These three complementary perspectives make it possible to distinguish disagreements on the underlying biology of a metabolic process from differences that can be explained by different decisions on how and in what detail to represent knowledge. As a proof-of-concept, we implemented C₂Cards^Human, as a web application http://www.molgenis.org/c2cards, covering five human pathway databases.

Conclusions

C₂Cards can contribute to ongoing reconciliation efforts by simplifying the identification of consensus and conflicts between pathway databases and lowering the threshold for experts to contribute. Several case studies illustrate the potential of the C₂Cards in identifying disagreements on the underlying biology of a metabolic process. The overviews may also point out controversial biological knowledge that should be subject of further research. Finally, the examples provided emphasize the importance of manual curation and the need for a broad community involvement.

Mechanisms of age-dependent response to winter temperature in perennial flowering of Arabis alpina

Perennial plants live for more than 1 year and flower only after an extended vegetative phase. We used Arabis alpina, a perennial relative of annual Arabidopsis thaliana, to study how increasing age and exposure to winter cold (vernalization) coordinate to establish competence to flower. We show that the APETALA2 transcription factor, a target of microRNA miR172, prevents flowering before vernalization. Additionally, miR156 levels decline as A. alpina ages, causing increased production of SPL (SQUAMOSA PROMOTER BINDING PROTEIN LIKE) transcription factors and ensuring that flowering occurs in response to cold. The age at which plants respond to vernalization can be altered by manipulating miR156 levels. Although miR156 and miR172 levels are uncoupled in A. alpina, miR156 abundance represents the timer controlling age-dependent flowering responses to cold.

Dissection of a quantitative trait locus for PR interval duration identifies Tnni3k as a novel modulator of cardiac conduction

Atrio-ventricular conduction disease is a common feature in Mendelian rhythm disorders associated with sudden cardiac death and is characterized by prolongation of the PR interval on the surface electrocardiogram (ECG). Prolongation of the PR interval is also a strong predictor of atrial fibrillation, the most prevalent sustained cardiac arrhythmia. Despite the significant genetic component in PR duration variability, the genes regulating PR interval duration remain largely elusive. We here aimed to dissect the quantitative trait locus (QTL) for PR interval duration that we previously mapped in murine F2 progeny of a sensitized 129P2 and FVBN/J cross. To determine the underlying gene responsible for this QTL, genome-wide transcriptional profiling was carried out on myocardial tissue from 109 F2 mice. Expression QTLs (eQTLs) were mapped and the PR interval QTL was inspected for the co-incidence of eQTLs. We further determined the correlation of each of these transcripts to the PR interval. Tnni3k was the only eQTL, mapping to the PR-QTL, with an established abundant cardiac-specific expression pattern and a significant correlation to PR interval duration. Genotype inspection in various inbred mouse strains revealed the presence of at least three independent haplotypes at the Tnni3k locus. Measurement of PR interval duration and Tnni3k mRNA expression levels in six inbred lines identified a positive correlation between the level of Tnni3k mRNA and PR interval duration. Furthermore, in DBA/2J mice overexpressing hTNNI3K, and in DBA.AKR.hrtfm2 congenic mice, which harbor the AKR/J “high-Tnni3k expression” haplotype in the DBA/2J genetic background, PR interval duration was prolonged as compared to DBA/2J wild-type mice (“low-Tnni3k expression” haplotype). Our data provide the first evidence for a role of Tnni3k in controlling the electrocardiographic PR interval indicating a function of Tnni3k in atrio-ventricular conduction.

Atrio-ventricular (AV) conduction disease (delay), characterized by prolongation of the PR interval on the surface electrocardiogram (ECG), is a common feature in Mendelian rhythm disorders and is associated with sudden cardiac death. Prolongation of the PR interval is also a strong predictor of atrial fibrillation (AF), the most common sustained cardiac arrhythmia. Although there is a substantial heritable component to the variability of the PR interval, the causative genes remain largely elusive. The identification of these genetic factors in the human population has been difficult owing to wide genetic heterogeneity and an uncontainable environment. We here exploited the homogeneous genetic background and controlled environment of inbred laboratory mouse strains to detect a genetic modifier of the PR interval. We identify Tnni3k as prime candidate for the modulation of the PR interval duration and suggest a new role for this gene, in the modulation of atrio-ventricular conduction.

Knowledge representation in metabolic pathway databases

The accurate representation of all aspects of a metabolic network in a structured format, such that it can be used for a wide variety of computational analyses, is a challenge faced by a growing number of researchers. Analysis of five major metabolic pathway databases reveals that each database has made widely different choices to address this challenge, including how to deal with knowledge that is uncertain or missing. In concise overviews, we show how concepts such as compartments, enzymatic complexes and the direction of reactions are represented in each database. Importantly, also concepts which a database does not represent are described. Which aspects of the metabolic network need to be available in a structured format and to what detail differs per application. For example, for in silico phenotype prediction, a detailed representation of gene-protein-reaction relations and the compartmentalization of the network is essential. Our analysis also shows that current databases are still limited in capturing all details of the biology of the metabolic network, further illustrated with a detailed analysis of three metabolic processes. Finally, we conclude that the conceptual differences between the databases, which make knowledge exchange and integration a challenge, have not been resolved, so far, by the exchange formats in which knowledge representation is standardized.

A critical assessment of SELDI-TOF-MS for biomarker discovery in serum and tissue of patients with an ovarian mass

Background

Less than 25% of patients with a pelvic mass who are presented to a gynecologist will eventually be diagnosed with epithelial ovarian cancer. Since there is no reliable test to differentiate between different ovarian tumors, accurate classification could facilitate adequate referral to a gynecological oncologist, improving survival. The goal of our study was to assess the potential value of a SELDI-TOF-MS based classifier for discriminating between patients with a pelvic mass.

Methods

Our study design included a well-defined patient population, stringent protocols and an independent validation cohort. We compared serum samples of 53 ovarian cancer patients, 18 patients with tumors of low malignant potential, and 57 patients with a benign ovarian tumor on different ProteinChip arrays. In addition, from a subset of 84 patients, tumor tissues were collected and microdissection was used to isolate a pure and homogenous cell population.

Results

Diagonal Linear Discriminant Analysis (DLDA) and Support Vector Machine (SVM) classification on serum samples comparing cancer versus benign tumors, yielded models with a classification accuracy of 71-81% (cross-validation), and 73-81% on the independent validation set. Cancer and benign tissues could be classified with 95-99% accuracy using cross-validation. Tumors of low malignant potential showed protein expression patterns different from both benign and cancer tissues. Remarkably, none of the peaks differentially expressed in serum samples were found to be differentially expressed in the tissue lysates of those same groups.

Conclusion

Although SELDI-TOF-MS can produce reliable classification results in serum samples of ovarian cancer patients, it will not be applicable in routine patient care. On the other hand, protein profiling of microdissected tumor tissue may lead to a better understanding of oncogenesis and could still be a source of new serum biomarkers leading to novel methods for differentiating between different histological subtypes.

Proteomic analysis of HIV-T cell interaction: an update

This mini-review summarizes techniques applied in, and results obtained with, proteomic studies of human immunodeficiency virus type 1 (HIV-1)–T cell interaction. Our group previously reported on the use of two-dimensional differential gel electrophoresis (2D-DIGE) coupled to matrix assisted laser-desorption time of flight peptide mass fingerprint analysis, to study T cell responses upon HIV-1 infection. Only one in three differentially expressed proteins could be identified using this experimental setup. Here we report on our latest efforts to test models generated by this data set and extend its analysis by using novel bioinformatic algorithms. The 2D-DIGE results are compared with other studies including a pilot study using one-dimensional peptide separation coupled to MS^E, a novel mass spectrometric approach. It can be concluded that although the latter method detects fewer proteins, it is much faster and less labor intensive. Last but not least, recent developments and remaining challenges in the field of proteomic studies of HIV-1 infection and proteomics in general are discussed.

Abstract 4800: Quantitative detection of discriminatory proteins in tissue and serum of ovarian cancer patients by LC-MSE

Purpose: Using label-free UPLC-MSE quantification methods to identifying proteins as well as determining their abundance in serum and microdissected ovarian cancer tissue. Experimental procedures: Serum of six patients with a serous adenocarcinoma of the ovary (stage IIIB or higher) was collected before treatment. Matched control patients with a serous cystadenoma of the ovary were chosen as a control group. In addition, homogeneous regions of cells that exhibited uniform histology were isolated from cancer tissue by laser capture microdissection (LCM), normal ovarian epithelial cell samples and stroma samples were taken from matched controls. We subsequently employed label-free ultra-high pressure liquid chromatography tandem mass spectrometry (UPLC-MSE) to identify proteins and determine their absolute concentrations in serum and tissue lysates. Moderated t-tests were used to identify differentially expressed proteins between patient- and control-group. p-values were adjusted for multiple-testing using Benjamini-Hochberg false discovery rates and considered differentially expressed when &lt;0.05. Summary of the data: In the serum specimens, 13 differentially expressed proteins were identified by LC-MSE profiling. Proteins with different concentrations in patients versus control sera include abundant serum proteins such as apolipoprotein AI and transferrin, both exhibited a lower concentration in serum of cancer patients. Differential expression was also observed for apolipoproteins (APOA IV, APOA II and APOC III) and other proteins that have not been associated with ovarian cancer previously such as C9 and Afamin. In the tissue-lysates we identified 535 Proteins of which 332 could be reliably quantified. Of these 56 were differentially expressed proteins of which 12 had non-overlapping standard deviations. Literature searches confirmed a link between ovarian cancer and 6 of these proteins, suggesting a possible role in disease progression. Of special interest was the identification of prohibitin (PHB) and cofilin-1 (CFL1). Immunohistochemical detection of CFL1 in formalin-fixed paraffin-embedded tissues of our patient group showed heterogeneous expression in 50% of the serous ovarian carcinomas, whereas all the benign lesions were negative. These results again demonstrate the validity of the protein quantitation results obtained by MSE. Interestingly, no overlap was found between the discriminatory proteins in serum specimens and corresponding tissue lysates, illustrating the challenge to identify a single biomarker in the circulation reflecting ovary cancer. Conclusions: Our study revealed several protein changes in serum and tissue of patients suffering from serous adenocarcinoma of the ovary. Further investigation of these proteins is warranted to establish whether they could provide new insights into disease etiology and act as potential new markers.

Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 4800. doi:1538-7445.AM2012-4800

Critical assessment of human metabolic pathway databases: a stepping stone for future integration

Background

Multiple pathway databases are available that describe the human metabolic network and have proven their usefulness in many applications, ranging from the analysis and interpretation of high-throughput data to their use as a reference repository. However, so far the various human metabolic networks described by these databases have not been systematically compared and contrasted, nor has the extent to which they differ been quantified. For a researcher using these databases for particular analyses of human metabolism, it is crucial to know the extent of the differences in content and their underlying causes. Moreover, the outcomes of such a comparison are important for ongoing integration efforts.

Results

We compared the genes, EC numbers and reactions of five frequently used human metabolic pathway databases. The overlap is surprisingly low, especially on reaction level, where the databases agree on 3% of the 6968 reactions they have combined. Even for the well-established tricarboxylic acid cycle the databases agree on only 5 out of the 30 reactions in total. We identified the main causes for the lack of overlap. Importantly, the databases are partly complementary. Other explanations include the number of steps a conversion is described in and the number of possible alternative substrates listed. Missing metabolite identifiers and ambiguous names for metabolites also affect the comparison.

Conclusions

Our results show that each of the five networks compared provides us with a valuable piece of the puzzle of the complete reconstruction of the human metabolic network. To enable integration of the networks, next to a need for standardizing the metabolite names and identifiers, the conceptual differences between the databases should be resolved. Considerable manual intervention is required to reach the ultimate goal of a unified and biologically accurate model for studying the systems biology of human metabolism. Our comparison provides a stepping stone for such an endeavor.

Polymorphism in HIV-1 dependency factor PDE8A affects mRNA level and HIV-1 replication in primary macrophages

Four genome-wide RNAi screens have recently identified hundreds of HIV-1 dependency factors (HDFs). Previously, we reported a large variation in the ability of HIV-1 to replicate in monocyte-derived macrophages (MDM) derived from >400 healthy seronegative blood donors. Here we determined whether SNPs in genes encoding newly identified HDFs were associated with this variation in HIV-1 replication. We found a significant association between the minor allele of SNP rs2304418 in phosphodiesterase 8A (PDE8A) and lower HIV-1 replication (p=2.4×10(-6)). The minor allele of SNP rs2304418 was also significantly associated with lower PDE8A mRNA levels in MDM (p=8.3×10(-5)). In accordance with this, overexpression of PDE8A in HEK293T cells resulted in increased HIV-1 replication, while subsequent knock-down of PDE8A decreased replication. This study links host genetic variation in a newly identified HDF to variation in HIV-1 replication in a relevant primary target cell for HIV-1 and may provide new leads for treatment of this infection.

Genome-wide association scan in HIV-1-infected individuals identifying variants influencing disease course

Background

AIDS develops typically after 7–11 years of untreated HIV-1 infection, with extremes of very rapid disease progression (<2 years) and long-term non-progression (>15 years). To reveal additional host genetic factors that may impact on the clinical course of HIV-1 infection, we designed a genome-wide association study (GWAS) in 404 participants of the Amsterdam Cohort Studies on HIV-1 infection and AIDS.

Methods

The association of SNP genotypes with the clinical course of HIV-1 infection was tested in Cox regression survival analyses using AIDS-diagnosis and AIDS-related death as endpoints.

Results

Multiple, not previously identified SNPs, were identified to be strongly associated with disease progression after HIV-1 infection, albeit not genome-wide significant. However, three independent SNPs in the top ten associations between SNP genotypes and time between seroconversion and AIDS-diagnosis, and one from the top ten associations between SNP genotypes and time between seroconversion and AIDS-related death, had P-values smaller than 0.05 in the French Genomics of Resistance to Immunodeficiency Virus cohort on disease progression.

Conclusions

Our study emphasizes that the use of different phenotypes in GWAS may be useful to unravel the full spectrum of host genetic factors that may be associated with the clinical course of HIV-1 infection.

An evaluation protocol for subtype-specific breast cancer event prediction

In recent years increasing evidence appeared that breast cancer may not constitute a single disease at the molecular level, but comprises a heterogeneous set of subtypes. This suggests that instead of building a single monolithic predictor, better predictors might be constructed that solely target samples of a designated subtype, which are believed to represent more homogeneous sets of samples. An unavoidable drawback of developing subtype-specific predictors, however, is that a stratification by subtype drastically reduces the number of samples available for their construction. As numerous studies have indicated sample size to be an important factor in predictor construction, it is therefore questionable whether the potential benefit of subtyping can outweigh the drawback of a severe loss in sample size. Factors like unequal class distributions and differences in the number of samples per subtype, further complicate comparisons. We present a novel experimental protocol that facilitates a comprehensive comparison between subtype-specific predictors and predictors that do not take subtype information into account. Emphasis lies on careful control of sample size as well as class and subtype distributions. The methodology is applied to a large breast cancer compendium involving over 1500 arrays, using a state-of-the-art subtyping scheme. We show that the resulting subtype-specific predictors outperform those that do not take subtype information into account, especially when taking sample size considerations into account.

Interorgan coordination of the murine adaptive response to fasting

Starvation elicits a complex adaptive response in an organism. No information on transcriptional regulation of metabolic adaptations is available. We, therefore, studied the gene expression profiles of brain, small intestine, kidney, liver, and skeletal muscle in mice that were subjected to 0-72 h of fasting. Functional-category enrichment, text mining, and network analyses were employed to scrutinize the overall adaptation, aiming to identify responsive pathways, processes, and networks, and their regulation. The observed transcriptomics response did not follow the accepted "carbohydrate-lipid-protein" succession of expenditure of energy substrates. Instead, these processes were activated simultaneously in different organs during the entire period. The most prominent changes occurred in lipid and steroid metabolism, especially in the liver and kidney. They were accompanied by suppression of the immune response and cell turnover, particularly in the small intestine, and by increased proteolysis in the muscle. The brain was extremely well protected from the sequels of starvation. 60% of the identified overconnected transcription factors were organ-specific, 6% were common for 4 organs, with nuclear receptors as protagonists, accounting for almost 40% of all transcriptional regulators during fasting. The common transcription factors were PPARα, HNF4α, GCRα, AR (androgen receptor), SREBP1 and -2, FOXOs, EGR1, c-JUN, c-MYC, SP1, YY1, and ETS1. Our data strongly suggest that the control of metabolism in four metabolically active organs is exerted by transcription factors that are activated by nutrient signals and serves, at least partly, to prevent irreversible brain damage.

Involvement of the TGF-beta and beta-catenin pathways in pelvic lymph node metastasis in early-stage cervical cancer

PURPOSE

Presence of pelvic lymph node metastases is the main prognostic factor in early-stage cervical cancer patients, primarily treated with surgery. Aim of this study was to identify cellular tumor pathways associated with pelvic lymph node metastasis in early-stage cervical cancer.

EXPERIMENTAL DESIGN

Gene expression profiles (Affymetrix U133 plus 2.0) of 20 patients with negative (N(0)) and 19 with positive lymph nodes (N(+)), were compared with gene sets that represent all 285 presently available pathway signatures. Validation immunostaining of tumors of 274 consecutive early-stage cervical cancer patients was performed for representatives of the identified pathways.

RESULTS

Analysis of 285 pathways resulted in identification of five pathways (TGF-β, NFAT, ALK, BAD, and PAR1) that were dysregulated in the N(0), and two pathways (β-catenin and Glycosphingolipid Biosynthesis Neo Lactoseries) in the N(+) group. Class comparison analysis revealed that five of 149 genes that were most significantly differentially expressed between N(0) and N(+) tumors (P < 0.001) were involved in β-catenin signaling (TCF4, CTNNAL1, CTNND1/p120, DKK3, and WNT5a). Immunohistochemical validation of two well-known cellular tumor pathways (TGF-β and β-catenin) confirmed that the TGF-β pathway (positivity of Smad4) was related to N(0) (OR: 0.20, 95% CI: 0.06-0.66) and the β-catenin pathway (p120 positivity) to N(+) (OR: 1.79, 95%CI: 1.05-3.05).

CONCLUSIONS

Our study provides new, validated insights in the molecular mechanism of lymph node metastasis in cervical cancer. Pathway analysis of the microarray expression profile suggested that the TGF-β and p120-associated noncanonical β-catenin pathways are important in pelvic lymph node metastasis in early-stage cervical cancer.

Genome-wide association study identifies single nucleotide polymorphism in DYRK1A associated with replication of HIV-1 in monocyte-derived macrophages

BACKGROUND

HIV-1 infected macrophages play an important role in rendering resting T cells permissive for infection, in spreading HIV-1 to T cells, and in the pathogenesis of AIDS dementia. During highly active anti-retroviral treatment (HAART), macrophages keep producing virus because tissue penetration of antiretrovirals is suboptimal and the efficacy of some is reduced. Thus, to cure HIV-1 infection with antiretrovirals we will also need to efficiently inhibit viral replication in macrophages. The majority of the current drugs block the action of viral enzymes, whereas there is an abundance of yet unidentified host factors that could be targeted. We here present results from a genome-wide association study identifying novel genetic polymorphisms that affect in vitro HIV-1 replication in macrophages.

METHODOLOGY/PRINCIPAL FINDINGS

Monocyte-derived macrophages from 393 blood donors were infected with HIV-1 and viral replication was determined using Gag p24 antigen levels. Genomic DNA from individuals with macrophages that had relatively low (n = 96) or high (n = 96) p24 production was used for SNP genotyping with the Illumina 610 Quad beadchip. A total of 494,656 SNPs that passed quality control were tested for association with HIV-1 replication in macrophages, using linear regression. We found a strong association between in vitro HIV-1 replication in monocyte-derived macrophages and SNP rs12483205 in DYRK1A (p = 2.16 × 10(-5)). While the association was not genome-wide significant (p<1 × 10(-7)), we could replicate this association using monocyte-derived macrophages from an independent group of 31 individuals (p = 0.0034). Combined analysis of the initial and replication cohort increased the strength of the association (p = 4.84 × 10(-6)). In addition, we found this SNP to be associated with HIV-1 disease progression in vivo in two independent cohort studies (p = 0.035 and p = 0.0048).

CONCLUSIONS/SIGNIFICANCE

These findings suggest that the kinase DYRK1A is involved in the replication of HIV-1, in vitro in macrophages as well as in vivo.

Molecular profiling of cytomegalovirus-induced human CD8+ T cell differentiation

CD8+ T cells play a critical role in the immune response to viral pathogens. Persistent human cytomegalovirus (HCMV) infection results in a strong increase in the number of virus-specific, quiescent effector-type CD8+ T cells with constitutive cytolytic activity, but the molecular pathways involved in the induction and maintenance of these cells are unknown. We show here that HCMV infection induced acute and lasting changes in the transcriptomes of virus-reactive T cells collected from HCMV-seropositive patients at distinct stages of infection. Enhanced cell cycle and metabolic activity was restricted to the acute phase of the response, but at all stages, HCMV-specific CD8+ T cells expressed the Th1-associated transcription factors T-bet (TBX21) and eomesodermin (EOMES), in parallel with continuous expression of IFNG mRNA and IFN-γ-regulated genes. The cytolytic proteins granzyme B and perforin as well as the fractalkine-binding chemokine receptor CX3CR1 were found in virus-reactive cells throughout the response. During HCMV latency, virus-specific CD8+ T cells lacked the typical features of exhausted cells found in other chronic infections. Persistent effector cell traits together with the permanent changes in chemokine receptor usage of virus-specific, nonexhausted, long-lived CD8+ T cells may be crucial to maintain lifelong protection from HCMV reactivation.

Expression of a retinoic acid signature in circulating CD34 cells from coronary artery disease patients

Background

Circulating CD34+ progenitor cells have the potential to differentiate into a variety of cells, including endothelial cells. Knowledge is still scarce about the transcriptional programs used by CD34+ cells from peripheral blood, and how these are affected in coronary artery disease (CAD) patients.

Results

We performed a whole genome transcriptome analysis of CD34+ cells, CD4+ T cells, CD14+ monocytes, and macrophages from 12 patients with CAD and 11 matched controls. CD34+ cells, compared to other mononuclear cells from the same individuals, showed high levels of KRAB box transcription factors, known to be involved in gene silencing. This correlated with high expression levels in CD34+ cells for the progenitor markers HOXA5 and HOXA9, which are known to control expression of KRAB factor genes. The comparison of expression profiles of CD34+ cells from CAD patients and controls revealed a less naïve phenotype in patients' CD34+ cells, with increased expression of genes from the Mitogen Activated Kinase network and a lowered expression of a panel of histone genes, reaching levels comparable to that in more differentiated circulating cells. Furthermore, we observed a reduced expression of several genes involved in CXCR4-signaling and migration to SDF1/CXCL12.

Conclusions

The altered gene expression profile of CD34+ cells in CAD patients was related to activation/differentiation by a retinoic acid-induced differentiation program. These results suggest that circulating CD34+ cells in CAD patients are programmed by retinoic acid, leading to a reduced capacity to migrate to ischemic tissues.

Abstract 3980: Quantitative detection of discriminatory proteins in tissue and serum of ovarian cancer patients by LC-MSE

Purpose

To search for potential protein markers or activated pathways which play a role in ovarian cancer.

Experimental procedures

Serum of six patients with a serous adenocarcinoma of the ovary (stage IIIB or higher) was collected before treatment. Matched control patients with a serous cystadenoma of the ovary were chosen as a control group. In addition, homogeneous regions of cells that exhibited uniform histology were isolated from cancer tissue in three of these patients by laser capture microdissection (LCM), normal ovarian epithelial cell samples were taken from their matched controls. Subsequently we employed label-free liquid chromatography tandem mass spectrometry (LC-MSE) to identify proteins and determine their absolute concentrations in serum specimens and tissue lysates. Moderated t-tests were used to identify differentially expressed proteins between patient- and control-group, p-values were adjusted for multiple-testing using Benjamini-Hochberg false discovery rates. Proteins were considered to be differentially expressed if the adjusted p-values were &lt;0.05.

Summary of the data

In the serum specimens, 13 differentially expressed proteins were identified by LC-MSE profiling. Proteins with different concentrations in patients versus control sera include abundant serum proteins such as apolipoprotein AI and transferrin, that both exhibited a lower concentration in serum of cancer patients. Differential expression was also observed for apolipoproteins (APOA IV, APOA II and APOC III) and other proteins that have not been associated with ovarian cancer previously such as C9 and Afamin.

In the tissue lysates, 17 differentially expressed proteins were identified. Amounts of collagen found were decreased in cancer tissue. Furthermore, proteins associated with the respiratory chain and cell migration were found in increased amounts in the cancer tissue lysates.

Interestingly, no overlap was found between the discriminatory proteins in serum specimens and corresponding tissue lysates, illustrating the challenge to identify a single biomarker in the circulation reflecting ovary cancer.

Conclusions

Our study revealed several protein changes in serum and the ovary of patients suffering from serous adenocarcinoma of the ovary (stage IIIB or higher). Further investigation of these proteins is warranted to establish whether they could provide new insights into the etiology of the disease and act as potential new disease markers.

Note: This abstract was not presented at the AACR 101st Annual Meeting 2010 because the presenter was unable to attend.

Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 101st Annual Meeting of the American Association for Cancer Research; 2010 Apr 17-21; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2010;70(8 Suppl):Abstract nr 3980.

MiR423-5p as a circulating biomarker for heart failure

RATIONALE

Aberrant expression profiles of circulating microRNAs (miRNAs) have been described in various diseases and provide high sensitivity and specificity. We explored circulating miRNAs as potential biomarkers in patients with heart failure (HF).

OBJECTIVE

The goal of this study was to determine whether miRNAs allow to distinguish clinical HF not only from healthy controls but also from non-HF forms of dyspnea.

METHODS AND RESULTS

A miRNA array was performed on plasma of 12 healthy controls and 12 HF patients. From this array, we selected 16 miRNAs for a second clinical study in 39 healthy controls and in 50 cases with reports of dyspnea, of whom 30 were diagnosed with HF and 20 were diagnosed with dyspnea attributable to non-HF-related causes. This revealed that miR423-5p was specifically enriched in blood of HF cases and receiver-operator-characteristics (ROC) curve analysis showed miR423-5p to be a diagnostic predictor of HF, with an area under the curve of 0.91 (P<0.001). Five other miRNAs were elevated in HF cases but also slightly increased in non-HF dyspnea cases.

CONCLUSION

We identify 6 miRNAs that are elevated in patients with HF, among which miR423-5p is most strongly related to the clinical diagnosis of HF. These 6 circulating miRNAs provide attractive candidates as putative biomarkers for HF.

Proteomic profiling of plasma and serum in elderly patients with delirium

The aim of this study was to compare plasma and serum protein profiles in elderly acute hip fracture patients with and without delirium. The spectra from surface-enhanced laser desorption ionization (SELDI) using time-of-flight (TOF) mass spectrometry of 16 patients without and 16 patients with delirium (two groups of eight and eight) scored using the Confusion Assessment Method were compared. The most discriminating peak of 15.9 kDa in plasma in a testing group of eight patients with delirium to eight patients without delirium was confirmed in an independent validation group. Taking both groups together, three discriminating peaks of 7.97, 15.9, and 16.0 kDa were found in delirious patients. These peaks presumably correspond to hemoglobin-beta, its doubly charged ion, and its glycosylated form.

A comprehensive sensitivity analysis of microarray breast cancer classification under feature variability

BACKGROUND

Large discrepancies in signature composition and outcome concordance have been observed between different microarray breast cancer expression profiling studies. This is often ascribed to differences in array platform as well as biological variability. We conjecture that other reasons for the observed discrepancies are the measurement error associated with each feature and the choice of preprocessing method. Microarray data are known to be subject to technical variation and the confidence intervals around individual point estimates of expression levels can be wide. Furthermore, the estimated expression values also vary depending on the selected preprocessing scheme. In microarray breast cancer classification studies, however, these two forms of feature variability are almost always ignored and hence their exact role is unclear.

RESULTS

We have performed a comprehensive sensitivity analysis of microarray breast cancer classification under the two types of feature variability mentioned above. We used data from six state of the art preprocessing methods, using a compendium consisting of eight different datasets, involving 1131 hybridizations, containing data from both one and two-color array technology. For a wide range of classifiers, we performed a joint study on performance, concordance and stability. In the stability analysis we explicitly tested classifiers for their noise tolerance by using perturbed expression profiles that are based on uncertainty information directly related to the preprocessing methods. Our results indicate that signature composition is strongly influenced by feature variability, even if the array platform and the stratification of patient samples are identical. In addition, we show that there is often a high level of discordance between individual class assignments for signatures constructed on data coming from different preprocessing schemes, even if the actual signature composition is identical.

CONCLUSION

Feature variability can have a strong impact on breast cancer signature composition, as well as the classification of individual patient samples. We therefore strongly recommend that feature variability is considered in analyzing data from microarray breast cancer expression profiling experiments.

Modulation of glycosphingolipid metabolism significantly improves hepatic insulin sensitivity and reverses hepatic steatosis in mice

Nonalcoholic fatty liver disease (NAFLD) is associated with obesity, insulin resistance, and type 2 diabetes. The hyperinsulinemia that occurs as a consequence of insulin resistance is thought to be an important contributor to the development of fatty liver. We have shown that the iminosugar N-(5'-adamantane-1'-yl-methoxy)-pentyl-1-deoxynojirimycin (AMP-DNM), an inhibitor of the enzyme glucosylceramide synthase, is a potent enhancer of insulin signaling in rodent models for insulin resistance and type 2 diabetes. The present study was designed to assess the impact of AMP-DNM on insulin levels, liver triglyceride synthesis, and gene expression profile. Treatment of ob/ob mice with AMP-DNM restored insulin signaling in the liver, corrected blood glucose values to levels found in lean mice, and decreased insulin concentration. The expression of sterol regulatory element-binding protein 1c target genes involved in fatty acid synthesis normalized. AMP-DNM treatment significantly reduced liver to body weight ratio and reversed hepatic steatosis, comprising fat as well as inflammatory markers. In addition, AMP-DNM treatment corrected to a large extent the gene expression profile of ob/ob mice livers toward the profile of lean mice.

CONCLUSION

Pharmacological lowering of glycosphingolipids with the iminosugar AMP-DNM is a promising approach to restore insulin signaling and improve glucose homeostasis as well as hepatic steatosis.

Classification-based comparison of pre-processing methods for interpretation of mass spectrometry generated clinical datasets

Background

Mass spectrometry is increasingly being used to discover proteins or protein profiles associated with disease. Experimental design of mass-spectrometry studies has come under close scrutiny and the importance of strict protocols for sample collection is now understood. However, the question of how best to process the large quantities of data generated is still unanswered. Main challenges for the analysis are the choice of proper pre-processing and classification methods. While these two issues have been investigated in isolation, we propose to use the classification of patient samples as a clinically relevant benchmark for the evaluation of pre-processing methods.

Results

Two in-house generated clinical SELDI-TOF MS datasets are used in this study as an example of high throughput mass-spectrometry data. We perform a systematic comparison of two commonly used pre-processing methods as implemented in Ciphergen ProteinChip Software and in the Cromwell package. With respect to reproducibility, Ciphergen and Cromwell pre-processing are largely comparable. We find that the overlap between peaks detected by either Ciphergen ProteinChip Software or Cromwell is large. This is especially the case for the more stringent peak detection settings. Moreover, similarity of the estimated intensities between matched peaks is high.

We evaluate the pre-processing methods using five different classification methods. Classification is done in a double cross-validation protocol using repeated random sampling to obtain an unbiased estimate of classification accuracy. No pre-processing method significantly outperforms the other for all peak detection settings evaluated.

Conclusion

We use classification of patient samples as a clinically relevant benchmark for the evaluation of pre-processing methods. Both pre-processing methods lead to similar classification results on an ovarian cancer and a Gaucher disease dataset. However, the settings for pre-processing parameters lead to large differences in classification accuracy and are therefore of crucial importance. We advocate the evaluation over a range of parameter settings when comparing pre-processing methods. Our analysis also demonstrates that reliable classification results can be obtained with a combination of strict sample handling and a well-defined classification protocol on clinical samples.

Suppression of inflammatory signaling in monocytes from patients with coronary artery disease

Monocytes and T-cells play an important role in the development of atherosclerotic coronary artery disease (CAD). Transcriptome analysis of circulating mononuclear cells from carefully matched atherosclerotic and control patients will potentially provide insights into the pathophysiology of atherosclerosis and supply biomarkers for diagnostic purposes. From patients undergoing coronary angiography because of anginal symptoms, we carefully matched 18 patients with severe triple-vessel CAD to 13 control patients without angiographic signs of CAD. All patients were on statin and aspirin treatment. Elevated soluble-ICAM levels demonstrated increased vascular inflammation in atherosclerotic patients. RNA from circulating CD4+ T-cells, CD14+ monocytes, lipopolysaccharide-stimulated monocytes, and macrophages was subjected to genome-wide expression analysis. In CD14+ monocytes, few inflammatory genes were overexpressed in control patients, while atherosclerotic patients showed overexpression of a group of Krüppel-associated box - containing transcription factors involved in negative regulation of gene expression. These differences disappeared upon LPS-stimulation or differentiation towards macrophages. No consistent changes in T cell transcriptomes were detected. Large inter-individual variability prevented the use of single differentially expressed genes as biomarkers, while monocyte gene expression signature predicted patient status with an accuracy of 84%. In this comprehensive analysis of circulating cell transcriptomes in atherosclerotic CAD, cautious patient matching revealed only small differences in transcriptional activity in different mononuclear cell types. Only an indication of a negative feedback to inflammatory gene expression was detected in atherosclerotic patients. Transcriptome differences of circulating cells possibly play less of a role than hitherto thought in the individual patient's susceptibility to atherosclerotic CAD, when appropriately matched for clinical symptoms and medication taken.

Analysis of gene expression identifies differentially expressed genes and pathways associated with lymphatic dissemination in patients with adenocarcinoma of the esophagus

INTRODUCTION

The presence of lymphatic dissemination is an important predictor of survival in esophageal adenocarcinoma (EA). The aim of this study was to discover a prognostic gene expression profile for lymphatic dissemination in EA and to identify genes and pathways that provide oncological insight in lymphatic dissemination.

METHODS

Patients who had lymphatic dissemination (N = 55) were compared with patients without lymphatic dissemination (N = 22). Whole-genome oligonucleotide microarrays were used to evaluate the genetic signature of 77 esophageal cancers. Multiple random validation was used to analyze the stability of the molecular signature and predictive power. Gene set enrichment analysis (GSEA) was applied to elucidate oncogenetic pathways.

RESULTS

Lymphatic dissemination was correctly predicted in 75 +/- 14% of lymph node positive patients. The absence of lymphatic dissemination was correctly predicted in 41 +/- 23% of lymph-node-negative patients. Argininosuccinate synthetase (ASS) was selected for validation on the protein level because it was present in most prognostic signatures as well as the list of differentially expressed genes. ASS expression was lower (P = 0.048) in patients with lymphatic dissemination than in patients without. GSEA identified that arginine metabolism pathways and lipid metabolism pathways are related to less chance of developing lymphatic dissemination.

DISCUSSION

The predictive profile does not outperform current clinical practice to predict the presence of lymphatic dissemination in patients with EA. Several genes, including ASS, and genetic pathways which are important in the development of lymphatic dissemination in EA, were identified.

Interferon-β Signaling Is Enhanced in Patients With Insufficient Coronary Collateral Artery Development and Inhibits Arteriogenesis in Mice

Stimulation of collateral artery growth in patients has been hitherto unsuccessful, despite promising experimental approaches. Circulating monocytes are involved in the growth of collateral arteries, a process also referred to as arteriogenesis. Patients show a large heterogeneity in their natural arteriogenic response on arterial obstruction. We hypothesized that circulating cell transcriptomes would provide mechanistic insights and new therapeutic strategies to stimulate arteriogenesis. Collateral flow index was measured in 45 patients with single-vessel coronary artery disease, separating collateral responders (collateral flow index, >0.21) and nonresponders (collateral flow index, ≤0.21). Isolated monocytes were stimulated with lipopolysaccharide or taken into macrophage culture for 20 hours to mimic their phenotype during arteriogenesis. Genome-wide mRNA expression analysis revealed 244 differentially expressed genes (adjusted P , <0.05) in stimulated monocytes. Interferon (IFN)-β and several IFN-related genes showed increased mRNA levels in 3 of 4 cellular phenotypes from nonresponders. Macrophage gene expression correlated with stimulated monocytes, whereas resting monocytes and progenitor cells did not display differential gene regulation. In vitro, IFN-β dose-dependently inhibited smooth muscle cell proliferation. In a murine hindlimb model, perfusion measured 7 days after femoral artery ligation showed attenuated arteriogenesis in IFN-β–treated mice compared with controls (treatment versus control: 31.5±1.2% versus 41.9±1.9% perfusion restoration, P <0.01). In conclusion, patients with differing arteriogenic response as measured with collateral flow index display differential transcriptomes of stimulated monocytes. Nonresponders show increased expression of IFN-β and its downstream targets, and IFN-β attenuates proliferation of smooth muscle cells in vitro and hampers arteriogenesis in mice. Inhibition of IFN-β signaling may serve as a novel approach for the stimulation of collateral artery growth.

Expression differences for genes involved in lignin, glutathione and sulphate metabolism in response to cadmium in Arabidopsis thaliana and the related Zn/Cd-hyperaccumulator Thlaspi caerulescens

Cadmium (Cd) is a widespread, naturally occurring element present in soil, rock, water, plants and animals. Cd is a non-essential element for plants and is toxic at higher concentrations. Transcript profiles of roots of Arabidopsis thaliana (Arabidopsis) and Thlaspi caerulescens plants exposed to Cd and zinc (Zn) are examined, with the main aim to determine the differences in gene expression between the Cd-tolerant Zn-hyperaccumulator T. caerulescens and the Cd-sensitive non-accumulator Arabidopsis. This comparative transcriptional analysis emphasized the role of genes involved in lignin, glutathione and sulphate metabolism. Furthermore the transcription factors MYB72 and bHLH100 were studied for their involvement in metal homeostasis, as they showed an altered expression after exposure to Cd. The Arabidopsis myb72 knockout mutant was more sensitive to excess Zn or iron (Fe) deficiency than wild type, while Arabidopsis transformants overexpressing bHLH100 showed increased tolerance to high Zn and nickel (Ni) compared to wild-type plants, confirming their role in metal homeostasis in Arabidopsis.

Muramyl dipeptide-induced differential gene expression in NOD2 mutant and wild-type Crohn's disease patient-derived dendritic cells

BACKGROUND

Mutations in the gene encoding the nucleotide-binding oligomerization domain 2 (NOD2) protein are associated with Crohn's disease (CD), but the mechanism underlying this is not completely understood. To study the mechanism of CD resulting from NOD2 mutations, we analyzed NOD2-dependent whole-genome expression profiles of patient-derived antigen-presenting cells.

PATIENTS AND METHODS

Monocyte-derived dendritic cells (DCs) from CD carriers of double-dose NOD2 mutations, wild-type CD patients, and wild-type healthy volunteers were stimulated with the NOD2 ligand muramyl dipeptide. Whole-genome microarrays were used to assess the differential gene expression. The clustering of significantly changed genes was analyzed by online gene ontology mapping software.

RESULTS

In the DCs from the wild-type CD patient group, 683 genes were significantly changed, with most of the genes clustering in the pathways of inflammatory response. In addition, a significant number of genes clustered in the apoptosis regulation-related pathway. In the DCs from the healthy volunteer group, only 50 genes were significantly changed, predominantly those belonging to the response to pathogen pathway. Analysis of differentially expressed gene ontology pathways in the DCs from the NOD2 mutant CD patient group showed that the transcription of pathogen response genes was absent. In this group, 298 genes were significantly changed, predominantly clustering in the negative apoptosis regulation and cell organization and biogenesis pathways.

CONCLUSIONS

Our results suggest that NOD2 mutations may result in perpetuation of mucosal inflammation through insufficient pathogen elimination. Further, these observations implicate a possible role of defective regulation of dendritic cell apoptosis in CD pathogenesis.

Gene expression in early stage cervical cancer

OBJECTIVE

Pelvic lymph node metastases are the main prognostic factor for survival in early stage cervical cancer, yet accurate detection methods before surgery are lacking. In this study, we examined whether gene expression profiling can predict the presence of lymph node metastasis in early stage squamous cell cervical cancer before treatment. In addition, we examined gene expression in cervical cancer compared to normal cervical tissue.

METHODS

Tumour samples of 35 patients with early stage cervical cancer who underwent radical hysterectomy and pelvic lymph node dissection, 16 with and 19 without lymph node metastasis, were analysed. Also five normal cervical tissues samples were analysed. We investigated differential expression and prediction of patient status for lymph node positive versus lymph node negative tumours and for healthy versus cancer tissue. Classifiers were built by using a multiple validation strategy, enabling the assessment of both classifier accuracy and variability.

RESULTS

Five genes (BANF1, LARP7, SCAMP1, CUEDC1 and PEBP1) showed differential expression between tumour samples from patients with and without lymph node metastasis. Mean accuracy of class prediction is 64.5% with a 95% confidence interval (CI) of 40-90%. For healthy cervical tissue versus early stage cervical cancer, the mean accuracy of class prediction is 99.5% (95% CI of 90-100%). A subset of genes involved in cervical cancer was identified.

CONCLUSION

No accurate class prediction for lymph node status in early stage cervical cancer was obtained. Replication studies are needed to determine the relevance of the differentially expressed genes according to lymph node status. Early stage cervical cancer can be perfectly differentiated from healthy cervical tissue by means of gene expression profiling.

Distinctive expression of chemokines and transforming growth factor-beta signaling in human arterial endothelium during atherosclerosis

Knowledge about the in vivo role of endothelium in chronic human atherosclerosis has mostly been derived by insights from mouse models. Therefore, we set out to establish by microarray analyses the gene expression profiles of endothelium from human large arteries, as isolated by laser microbeam microdissection, having focal atherosclerosis of the early or the advanced stage. Within individual arteries, the endothelial transcriptomes of the lesional and unaffected sides were compared pairwise, thus limiting genetic and environmental confounders. Specific endothelial signature gene sets were identified with changed expression levels in either early (n = 718) or advanced atherosclerosis (n = 403), relative to their paired plaque-free controls. Gene set enrichment analysis identified distinct sets of chemokines and differential enrichments of nuclear factor-kappaB-, p53-, and transforming growth factor-beta-related genes in advanced plaques. Immunohistochemistry validated the discriminative value of corresponding endothelial protein expression between early (fractalkine/CX3CL1, IP10/CCL10, TBX18) or advanced (BAX, NFKB2) stages of atherosclerosis and versus their plaque-free controls. The functional involvement of transforming growth factor-beta signaling in directing its downstream gene repertoire was substantiated by a consistent detection of activated SMAD2 in advanced lesions. Thus, we identified truly common, local molecular denominators of pathological changes to vascular endothelium, with a marked distinction of endothelial phenotype between early and advanced plaques.

Three-dimensional measurement and visualization of morphogenesis applied to cardiac embryology

Volume growth and proliferation are key processes in heart morphogenesis, yet their regionalization during development of the heart has been described only anecdotally. To study the contribution of cardiomyocyte proliferation to heart development, a quantitative reconstruction method was designed, allowing the local mapping of this morphogenetic process. First, a morphological surface reconstruction is made of the heart, using sections stained specifically for cardiomyocytes. Then, by a comprehensive series of image processing steps, local three-dimensional (3D) information of proliferation is obtained. These local quantitative data are then mapped onto the morphological surface reconstruction, resulting in a reconstruction that not only provides morphological information (qualitative), but also displays local information on proliferation rate (quantitative). The resulting 3D quantitative reconstructions revealed novel observations regarding the morphogenesis of the heart.

Large expression differences in genes for iron and zinc homeostasis, stress response, and lignin biosynthesis distinguish roots of Arabidopsis thaliana and the related metal hyperaccumulator Thlaspi caerulescens

The micronutrient zinc has an essential role in physiological and metabolic processes in plants as a cofactor or structural element in 300 catalytic and noncatalytic proteins, but it is very toxic when available in elevated amounts. Plants tightly regulate their internal zinc concentrations in a process called zinc homeostasis. The exceptional zinc hyperaccumulator species Thlaspi caerulescens can accumulate up to 3% of zinc, but also high amounts of nickel and cadmium, without any sign of toxicity. This should have drastic effects on the zinc homeostasis mechanism. We examined in detail the transcription profiles of roots of Arabidopsis thaliana and T. caerulescens plants grown under deficient, sufficient, and excess supply of zinc. A total of 608 zinc-responsive genes with at least a 3-fold difference in expression level were detected in A. thaliana and 352 in T. caerulescens in response to changes in zinc supply. Only 14% of these genes were also zinc responsive in A. thaliana. When comparing A. thaliana with T. caerulescens at each zinc exposure, more than 2,200 genes were significantly differentially expressed (>or=5-fold and false discovery rate < 0.05). While a large fraction of these genes are of yet unknown function, many genes with a different expression between A. thaliana and T. caerulescens appear to function in metal homeostasis, in abiotic stress response, and in lignin biosynthesis. The high expression of lignin biosynthesis genes corresponds to the deposition of lignin in the endodermis, of which there are two layers in T. caerulescens roots and only one in A. thaliana.

Incorporation of liquid-crystal light valve nonlinearities in optical multilayer neural networks

Sigmoidlike activation functions, as available in analog hardware, differ in various ways from the standard sigmoidal function because they are usually asymmetric, truncated, and have a nonstandard gain. We present an adaptation of the backpropagation learning rule to compensate for these nonstandard sigmoids. This method is applied to multilayer neural networks with all-optical forward propagation and liquid-crystal light valves (LCLV) as optical thresholding devices. The results of simulations of a backpropagation neural network with five different LCLV response curves as activation functions are presented. Although LCLV's perform poorly with the standard backpropagation algorithm, it is shown that our adapted learning rule performs well with these LCLV curves.