Large scale microarray profiling and coexpression network analysis of CHO cells identifies transcriptional modules associated with growth and productivity

Molecular biomarkers identification and applications in CHO bioprocessing

Biomarkers are valuable tools in clinical research where they allow to predict susceptibility to diseases, or response to specific treatments. Likewise, biomarkers can be extremely useful in the biomanufacturing of therapeutic proteins. Indeed, constraints such as short timelines and the need to find hyper-productive cells could benefit from a data-driven approach during cell line and process development. Many companies still rely on large screening capacities to develop productive cell lines, but as they reach a limit of production, there is a need to go from empirical to rationale procedures. Similarly, during bioprocessing runs, substrate consumption and metabolism wastes are commonly monitored. None of them possess the ability to predict the culture behavior in the bioreactor. Big data driven approaches are being adapted to the study of industrial mammalian cell lines, enabled by the publication of Chinese hamster and CHO genome assemblies which allowed the use of next-generation sequencing with these cells, as well as continuous proteome and metabolome annotation. However, if these different -omics technologies contributed to the characterization of CHO cells, there is a significant effort remaining to apply this knowledge to biomanufacturing methods. The correlation of a complex phenotype such as high productivity or rapid growth to the presence or expression level of a specific biomarker could save time and effort in the screening of manufacturing cell lines or culture conditions. In this review we will first discuss the different biological molecules that can be identified and quantified in cells, their detection techniques, and associated challenges. We will then review how these markers are used during the different steps of cell line and bioprocess development, and the inherent limitations of this strategy.

LncRNA analysis of mAb producing CHO clones reveals marker and engineering potential

Long non-coding RNAs (lncRNAs) are a potential new cell line engineering tool for improvement of yield and stability of CHO cells. In this study, we performed RNA sequencing of mAb producer CHO clones to study the lncRNA and protein coding transcriptome in relation to productivity. First, a robust linear model was used to identify genes correlating to productivity. To unravel specific patterns in expression of these genes, we employed weighted gene coexpression analysis (WGCNA) to find coexpressed modules, looking both for lncRNAs and coding genes. There was little overlap in the genes associated with productivity between the two products studied, possibly due to the difference in absolute range of productivity between the two mAbs. Therefore, we focused on the product with higher productivity and stronger candidate lncRNAs. To evaluate their potential as engineering targets, these candidate lncRNAs were transiently overexpressed or deleted by stable CRISPR Cas9 knock out both in a high and a low productivity subclone. We found that the thus achieved expression level of the identified lncRNAs, as confirmed by qPCR, does correlate well to productivity, so that they represent good markers that may be used for early clone selection. Additionally, we found that the deletion of one tested lncRNA region decreased viable cell density (VCD), prolonged culture time and increased cell size, final titer and specific productivity per cell. These results demonstrate the feasibility and usefulness of engineering lncRNA expression in production cell lines.

Single-cell RNA sequencing reveals homogeneous transcriptome patterns and low variance in a suspension CHO-K1 and an adherent HEK293FT cell line in culture conditions

Recombinant mammalian host cell lines, in particular CHO and HEK293 cells, are used for the industrial production of therapeutic proteins. Despite their well-known genomic instability, the control mechanisms that enable cells to respond to changes in the environmental conditions are not yet fully understood, nor do we have a good understanding of the factors that lead to phenotypic shifts in long-term cultures. A contributing factor could be inherent diversity in transcriptomes within a population. In this study, we used a full-length coverage single-cell RNA sequencing (scRNA-seq) approach to investigate and compare cell-to-cell variability and the impact of standardized and homogenous culture conditions on the diversity of individual cell transcriptomes, comparing suspension CHO-K1 and adherent HEK293FT cells. Our data showed a critical batch effect from the sequencing of four 96-well plates of CHO-K1 single cells stored for different periods of time, which was and may be therefore identified as a technical variable to consider in experimental planning. Besides, in an artificial and controlled culture environment such as used in routine cell culture technology, the gene expression pattern of a given population does not reveal any marker gene capable to disclose relevant cell population substructures, both for CHO-K1 cells and for HEK293FT cells. The variation observed is primarily driven by the cell cycle.

Key gene network related to primary ciliary dyskinesia in hippocampus of patients with Alzheimer’s disease revealed by weighted gene co-expression network analysis

Background

Alzheimer’s disease (AD) is closely related to aging, showing an increasing incidence rate for years. As one of the main brain regions involved in AD, hippocampus has been extensively studied due to its association with many human diseases. However, little is known about its association with primary ciliary dyskinesia (PCD).

Material and Methods

The microarray data of hippocampus on AD were retrieved from the Gene Expression Omnibus (GEO) database to construct the co-expression network by weighted gene co-expression network analysis (WGCNA). The gene network modules associated with AD screened with the common genes were further annotated based on Gene Ontology (GO) database and enriched based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) database. The protein-protein interaction (PPI) network was constructed based on STRING database to identify the hub genes in the network.

Results

Genes involved in PCD were identified in the hippocampus of AD patients. Functional analysis revealed that these genes were mainly enriched in ciliary tissue, ciliary assembly, axoneme assembly, ciliary movement, microtubule based process, microtubule based movement, organelle assembly, axoneme dynamin complex, cell projection tissue, and microtubule cytoskeleton tissue. A total of 20 central genes, e.g., DYNLRB2, ZMYND10, DRC1, DNAH5, WDR16, TTC25, and ARMC4 were identified as hub genes related to PCD in hippocampus of AD patients.

Conclusion

Our study demonstrated that AD and PCD have common metabolic pathways. These common pathways provide novel evidence for further investigation of the pathophysiological mechanism and the hub genes suggest new therapeutic targets for the diagnosis and treatment of AD and PCD.

Subjects

Bioinformatics, Cell Biology, Molecular Biology, Neurology.

Systems biology approach in the formulation of chemically defined media for recombinant protein overproduction

The cell culture medium is an intricate mixture of components which has a tremendous effect on cell growth and recombinant protein production. Regular cell culture medium includes various components, and the decision about which component should be included in the formulation and its optimum amount is an underlying issue in biotechnology industries. Applying conventional techniques to design an optimal medium for the production of a recombinant protein requires meticulous and immense research. Moreover, since the medium formulation for the production of one protein could not be the best choice for another protein, hence, the most suitable media should be determined for each recombinant cell line. Accordingly, medium formulation becomes a laborious, time-consuming, and costly process in biomanufacturing of recombinant protein, and finding alternative strategies for medium development seems to be crucial. In silico modeling is an attractive concept to be adapted for medium formulation due to its high potential to supersede laboratory examinations. By emerging the high-throughput datasets, scientists can disclose the knowledge about the effect of medium components on cell growth and metabolism, and via applying this information through systems biology approach, medium formulation optimization could be accomplished in silico with no need of significant amount of experimentation. This review demonstrates some of the applications of systems biology as a powerful tool for medium development and illustrates the effect of medium optimization with system-level analysis on the production of recombinant proteins in different host cells.

Competing Endogenous RNA and Coexpression Network Analysis for Identification of Potential Biomarkers and Therapeutics in association with Metastasis Risk and Progression of Prostate Cancer

Prostate cancer (PCa) is the most frequently diagnosed malignant neoplasm in men. Despite the high incidence, the underlying pathogenic mechanisms of PCa are still largely unknown, which limits the therapeutic options and leads to poor prognosis. Herein, based on the expression profiles from The Cancer Genome Atlas (TCGA) database, we investigated the interactions between long noncoding RNA (lncRNA) and mRNA by constructing a competing endogenous RNA network. Several competing endogenous RNAs could participate in the tumorigenesis of PCa. Six lncRNA signatures were identified as potential candidates associated with stage progression by the Kolmogorov-Smirnov test. In addition, 32 signatures from the coexpression network had potential diagnostic value for PCa lymphatic metastasis using machine learning algorithms. By targeting the coexpression network, the antifungal compound econazole was screened out for PCa treatment. Econazole could induce growth restraint, arrest the cell cycle, lead to apoptosis, inhibit migration, invasion, and adhesion in PC3 and DU145 cell lines, and inhibit the growth of prostate xenografts in nude mice. This systematic characterization of lncRNAs, microRNAs, and mRNAs in the risk of metastasis and progression of PCa will aid in the identification of candidate prognostic biomarkers and potential therapeutic drugs.

Mammalian Systems Biotechnology Reveals Global Cellular Adaptations in a Recombinant CHO Cell Line

Effective development of host cells for therapeutic protein production is hampered by the poor characterization of cellular transfection. Here, we employed a multi-omics-based systems biotechnology approach to elucidate the genotypic and phenotypic differences between a wild-type and recombinant antibody-producing Chinese hamster ovary (CHO) cell line. At the genomic level, we observed extensive rearrangements in specific targeted loci linked to transgene integration sites. Transcriptional re-wiring of DNA damage repair and cellular metabolism in the antibody producer, via changes in gene copy numbers, was also detected. Subsequent integration of transcriptomic data with a genome-scale metabolic model showed a substantial increase in energy metabolism in the antibody producer. Metabolomics, lipidomics, and glycomics analyses revealed an elevation in long-chain lipid species, potentially associated with protein transport and secretion requirements, and a surprising stability of N-glycosylation profiles between both cell lines. Overall, the proposed knowledge-based systems biotechnology framework can further accelerate mammalian cell-line engineering in a targeted manner.

Transient manipulation of the expression level of selected growth rate correlating microRNAs does not increase growth rate in CHO-K1 cells

Engineering of Chinese Hamster Ovary cells by manipulating microRNA (miRNA) expression levels has been shown to induce advantageous, desired phenotypes. Most of these studies so far were concerned with increasing productivity or reducing growth rate (with the implied intention of thus freeing cellular resources to also increase productivity). Here we evaluated the ability of growth correlating miRNAs to increase the growth rate of CHO-K1 cells by transient overexpression or knock down, respectively. Candidates were selected based on the correlation between growth rate and miRNA expression levels as observed in previous studies. These candidates were then up- or downregulated initially by transfection of mimics or inhibitors and subsequently by transfection of plasmids bearing the corresponding miRNAs or sponges. None of the 40 selected candidates was able to induce a better growth phenotype under these conditions. Overlap between miRNAs identified to correlate to growth in published miRNA expression studies and those identified to actively increase growth rate in a functional screen is minimal, indicating that the here selected approach of traditional overexpression/knock down engineering of miRNAs may not be a suitable strategy for the purpose of increasing growth rate.

The Long Non-Coding RNA Transcriptome Landscape in CHO Cells Under Batch and Fed-Batch Conditions

The role of non-coding RNAs in determining growth, productivity, and recombinant product quality attributes in Chinese hamster ovary (CHO) cells has received much attention in recent years, exemplified by studies into microRNAs in particular. However, other classes of non-coding RNAs have received less attention. One such class are the non-coding RNAs known collectively as long non-coding RNAs (lncRNAs). The authors have undertaken the first landscape analysis of the lncRNA transcriptome in CHO using a mouse based microarray that also provided for the surveillance of the coding transcriptome. The authors report on those lncRNAs present in a model host CHO cell line under batch and fed-batch conditions on two different days and relate the expression of different lncRNAs to each other. The authors demonstrate that the mouse microarray is suitable for the detection and analysis of thousands of CHO lncRNAs and validated a number of these by qRT-PCR. The authors then further analyzed the data to identify those lncRNAs whose expression changed the most between growth and stationary phases of culture or between batch and fed-batch culture to identify potential lncRNA targets for further functional studies with regard to their role in controlling growth of CHO cells. The authors discuss the implications for the publication of this rich dataset and how this may be used by the community.

Proteomic analysis of micro-scale bioreactors as scale-down model for a mAb producing CHO industrial fed-batch platform

The pharmaceutical production of recombinant proteins, such as monoclonal antibodies, is rather complex and requires proper development work. Accordingly, it is essential to develop appropriate scale-down models, which can mimic the corresponding production scale. In this work, we investigated the impact of the bioreactor scale on intracellular micro-heterogeneities of a CHO cell line producing monoclonal antibodies in fed-batch mode, using a 10 mL micro-bioreactor (ambr™) scale-down model and the corresponding 300 L pilot-scale bioreactor. For each scale, we measured the time evolution of the proteome, which enabled us to compare the impact of the bioreactor scale on the intracellular processes. Nearly absolute accordance between the scales was verified by data mining methods, such as hierarchical clustering and in-detail analysis on a single protein base. The time response of principal enzymes related to N-glycosylation was discussed, emphasizing major dissimilarities between the glycan fractions adorning the heavy chain and the corresponding protein abundance. The enzyme expression displayed mainly a constant profile, whereas the resulting glycan pattern changed over time. It is concluded that the enzymatic activity is influenced by the changing environmental conditions present in the fed-batch processes leading to the observed time-dependent variation.

A Comparative Transcriptomics Workflow for Analyzing Microarray Data From CHO Cell Cultures

Microarray-based comparative transcriptomics analysis is a powerful tool to understand therapeutic protein producing mammalian cell lines at the gene expression level. However, an integrated analysis workflow specifically designed for end-to-end analysis of microarray data for CHO cells, the most prevalent host for commercial recombinant protein production, is lacking. To address this gap, an automated data analysis workflow in R that leverages public domain analysis modules is developed to analyze microarray based gene expression data. In addition to testing the global transcriptome differences of CHO cells at different conditions, the workflow identifies differentially expressed genes and pathways with intuitive visualizations as the outputs. The utility of this automated workflow is demonstrated by comparing the transcriptomic profiles of recombinant protein expressing CHO cells with and without a temperature shift. Statistically significant differential expression at the gene, pathway, and global transcriptome levels are identified and visualized. An automated workflow like the one developed in this study will enable rapid translation of CHO culture microarray data into biologically relevant information for mechanism-driven cell line optimization and bioprocess development.

The Bioinformatic Analysis of the Dysregulated Genes and MicroRNAs in Entorhinal Cortex, Hippocampus, and Blood for Alzheimer's Disease

Aim

The incidence of Alzheimer's disease (AD) has been increasing in recent years, but there exists no cure and the pathological mechanisms are not fully understood. This study aimed to find out the pathogenesis of learning and memory impairment, new biomarkers, potential therapeutic targets, and drugs for AD.

Methods

We downloaded the microarray data of entorhinal cortex (EC) and hippocampus (HIP) of AD and controls from Gene Expression Omnibus (GEO) database, and then the differentially expressed genes (DEGs) in EC and HIP regions were analyzed for functional and pathway enrichment. Furthermore, we utilized the DEGs to construct coexpression networks to identify hub genes and discover the small molecules which were capable of reversing the gene expression profile of AD. Finally, we also analyzed microarray and RNA-seq dataset of blood samples to find the biomarkers related to gene expression in brain.

Results

We found some functional hub genes, such as ErbB2, ErbB4, OCT3, MIF, CDK13, and GPI. According to GO and KEGG pathway enrichment, several pathways were significantly dysregulated in EC and HIP. CTSD and VCAM1 were dysregulated significantly in blood, EC, and HIP, which were potential biomarkers for AD. Target genes of four microRNAs had similar GO_terms distribution with DEGs in EC and HIP. In addtion, small molecules were screened out for AD treatment.

Conclusion

These biological pathways and DEGs or hub genes will be useful to elucidate AD pathogenesis and identify novel biomarkers or drug targets for developing improved diagnostics and therapeutics against AD.

CHO-Omics Review: The Impact of Current and Emerging Technologies on Chinese Hamster Ovary Based Bioproduction

CHO cells are the most prevalent platform for modern bio-therapeutic production. Currently, there are several CHO cell lines used in bioproduction with distinct characteristics and unique genotypes and phenotypes. These differences limit advances in productivity and quality that can be achieved by the most common approaches to bioprocess optimization and cell line engineering. Incorporating omics-based approaches into current bioproduction processes will complement traditional methodologies to maximize gains from CHO engineering and bioprocess improvements. In order to highlight the utility of omics technologies in CHO bioproduction, the authors discuss current applications as well as limitations of genomics, transcriptomics, proteomics, metabolomics, lipidomics, fluxomics, glycomics, and multi-omics approaches and the potential they hold for the future of bioproduction. Multiple omics approaches are currently being used to improve CHO bioprocesses; however, the application of these technologies is still limited. As more CHO-omic datasets become available and integrated into systems models, the authors expect significant gains in product yield and quality. While individual omics technologies provide incremental improvements in bioproduction, the authors will likely see the most significant gains by applying multi-omics and systems biology approaches to individual CHO cell lines.

Improvements in protein production in mammalian cells from targeted metabolic engineering

Bioprocess optimization has yielded powerful clones for biotherapeutic production. However, new genomic technologies allow more targeted approaches to cell line development. Here we review efforts to enhance protein production in mammalian cells through metabolic engineering. Most efforts aimed to reduce toxic byproducts accumulation to enhance protein productivity. However, recent work highlights the possibility of regulating other desirable traits (e.g., apoptosis and glycosylation) by targeting central metabolism since these processes are interconnected. Therefore, as we further detail the pathways underlying cell growth and protein production and deploy diverse algorithms for their analysis, opportunities will arise to move beyond simple cell line designs and facilitate cell engineering strategies with complex combinations of genes that together underlie a phenotype of interest.

The 'Omics Revolution in CHO Biology: Roadmap to Improved CHO Productivity

Increased understanding of Chinese hamster ovary (CHO) cell physiology has been ushered in upon availability of the parental CHO-K1 cell line genome. Free and openly accessible sequence information has complemented transcriptomic and proteomic studies. The previous decade has also seen an increase in sensitivity and accuracy of proteomic methods due to technology development. In this genomic era, high-throughput screening methods, sophisticated informatic tools, and models continually drive major innovations in cell line development and process engineering. This review describes the various achievements in 'omics techniques and their application to improve recombinant protein expression from CHO cell lines.

Hyperosmotic stimulus study discloses benefits in ATP supply and reveals miRNA/mRNA targets to improve recombinant protein production of CHO cells

Biopharmaceuticals are predominantly produced by Chinese hamster ovary (CHO) cells cultivated in fed-batch mode. Hyperosmotic culture conditions (≥ 350 mOsmol kg(∑1) ) resulting from feeding of nutrients may enhance specific product formation rates (qp ). As an improved ATP supply was anticipated to enhance qp this study focused on the identification of suitable miRNA/mRNA targets to increase ATP levels. Therefor next generation sequencing and a compartment specific metabolomics approach were applied to analyze the response of an antibody (mAB) producing CHO cell line upon osmotic shift (280 → 430 mOsmol kg(-1) ). Hyperosmotic culture conditions caused a ∼2.6-fold increase of specific ATP formation rates together with a ∼1.7-fold rise in cytosolic and mitochondrial ATP-pools, thus showing increased ATP supply. mRNA expression analysis identified several genes encoding glycosylated proteins with strictly tissue related function. In addition, hyperosmotic culture conditions induced an upregulation of miR-132-3p, miR-132-5p, miR-182, miR-183, miR-194, miR-215-3p, miR-215-5p which have all been related to cell cycle arrest/proliferation in cancer studies. In relation to a previous independent CHO study miR-183 may be the most promising target to enhance qp by stable overexpression. Furthermore, deletion of genes with presumably dispensable function in suspension growing CHO cells may enhance mAB formation by increased ATP levels.

Comprehensive genome and epigenome characterization of CHO cells in response to evolutionary pressures and over time

The most striking characteristic of CHO cells is their adaptability, which enables efficient production of proteins as well as growth under a variety of culture conditions, but also results in genomic and phenotypic instability. To investigate the relative contribution of genomic and epigenetic modifications towards phenotype evolution, comprehensive genome and epigenome data are presented for six related CHO cell lines, both in response to perturbations (different culture conditions and media as well as selection of a specific phenotype with increased transient productivity) and in steady state (prolonged time in culture under constant conditions). Clear transitions were observed in DNA‐methylation patterns upon each perturbation, while few changes occurred over time under constant conditions. Only minor DNA‐methylation changes were observed between exponential and stationary growth phase; however, throughout a batch culture the histone modification pattern underwent continuous adaptation. Variation in genome sequence between the six cell lines on the level of SNPs, InDels, and structural variants is high, both upon perturbation and under constant conditions over time. The here presented comprehensive resource may open the door to improved control and manipulation of gene expression during industrial bioprocesses based on epigenetic mechanisms. Biotechnol. Bioeng. 2016;113: 2241–2253. © 2016 The Authors. Biotechnology and Bioengineering Published by Wiley Periodicals, Inc.

The use of 'Omics technology to rationally improve industrial mammalian cell line performance

Biologics represent an increasingly important class of therapeutics, with 7 of the 10 top selling drugs from 2013 being in this class. Furthermore, health authority approval of biologics in the immuno-oncology space is expected to transform treatment of patients with debilitating and deadly diseases. The growing importance of biologics in the healthcare field has also resulted in the recent approvals of several biosimilars. These recent developments, combined with pressure to provide treatments at lower costs to payers, are resulting in increasing need for the industry to quickly and efficiently develop high yielding, robust processes for the manufacture of biologics with the ability to control quality attributes within narrow distributions. Achieving this level of manufacturing efficiency and the ability to design processes capable of regulating growth, death and other cellular pathways through manipulation of media, feeding strategies, and other process parameters will undoubtedly be facilitated through systems biology tools generated in academic and public research communities. Here we discuss the intersection of systems biology, 'Omics technologies, and mammalian bioprocess sciences. Specifically, we address how these methods in conjunction with traditional monitoring techniques represent a unique opportunity to better characterize and understand host cell culture state, shift from an empirical to rational approach to process development and optimization of bioreactor cultivation processes. We summarize the following six key areas: (i) research applied to parental, non-recombinant cell lines; (ii) systems level datasets generated with recombinant cell lines; (iii) datasets linking phenotypic traits to relevant biomarkers; (iv) data depositories and bioinformatics tools; (v) in silico model development, and (vi) examples where these approaches have been used to rationally improve cellular processes. We critically assess relevant and state of the art research being conducted in academic, government and industrial laboratories. Furthermore, we apply our expertise in bioprocess to define a potential model for integration of these systems biology approaches into biologics development.

Microarray profiling of preselected CHO host cell subclones identifies gene expression patterns associated with increased production capacity

Over the last three decades, product yields from CHO cells have increased dramatically, yet specific productivity (qP) remains a limiting factor. In a previous study, using repeated cell-sorting, we have established different host cell subclones that show superior transient qP over their respective parental cell lines (CHO-K1, CHO-S). The transcriptome of the resulting six cell lines in different biological states (untransfected, mock transfected, plasmid transfected) was first explored by hierarchical clustering and indicated that gene activity associated with increased qP did not stem from a certain cellular state but seemed to be inherent for a high qP host line. We then performed a novel gene regression analysis identifying drivers for an increase in qP. Genes significantly implicated were first systematically tested for enrichment of GO terms using a Bayesian approach incorporating the hierarchical structure of the GO term tree. Results indicated that specific cellular components such as nucleus, ER, and Golgi are relevant for cellular productivity. This was complemented by targeted GSA that tested functionally homogeneous, manually curated subsets of KEGG pathways known to be involved in transcription, translation, and protein processing. Significantly implicated pathways included mRNA surveillance, proteasome, protein processing in the ER and SNARE interactions in vesicular transport.

van der Vaart A, Web BT, Bacanu SA, Kalsi G, Kendler KS, Miles MF, Dick D, Riley BP, Dumur C, Vladimirov VI. Integrating mRNA and miRNA Weighted Gene Co-Expression Networks with eQTLs in the Nucleus Accumbens of Subjects with Alcohol Dependence

Alcohol consumption is known to lead to gene expression changes in the brain. After performing weighted gene co-expression network analyses (WGCNA) on genome-wide mRNA and microRNA (miRNA) expression in Nucleus Accumbens (NAc) of subjects with alcohol dependence (AD; N = 18) and of matched controls (N = 18), six mRNA and three miRNA modules significantly correlated with AD were identified (Bonferoni-adj. p≤ 0.05). Cell-type-specific transcriptome analyses revealed two of the mRNA modules to be enriched for neuronal specific marker genes and downregulated in AD, whereas the remaining four mRNA modules were enriched for astrocyte and microglial specific marker genes and upregulated in AD. Gene set enrichment analysis demonstrated that neuronal specific modules were enriched for genes involved in oxidative phosphorylation, mitochondrial dysfunction and MAPK signaling. Glial-specific modules were predominantly enriched for genes involved in processes related to immune functions, i.e. cytokine signaling (all adj. p≤ 0.05). In mRNA and miRNA modules, 461 and 25 candidate hub genes were identified, respectively. In contrast to the expected biological functions of miRNAs, correlation analyses between mRNA and miRNA hub genes revealed a higher number of positive than negative correlations (χ2 test p≤ 0.0001). Integration of hub gene expression with genome-wide genotypic data resulted in 591 mRNA cis-eQTLs and 62 miRNA cis-eQTLs. mRNA cis-eQTLs were significantly enriched for AD diagnosis and AD symptom counts (adj. p = 0.014 and p = 0.024, respectively) in AD GWAS signals in a large, independent genetic sample from the Collaborative Study on Genetics of Alcohol (COGA). In conclusion, our study identified putative gene network hubs coordinating mRNA and miRNA co-expression changes in the NAc of AD subjects, and our genetic (cis-eQTL) analysis provides novel insights into the etiological mechanisms of AD.

Transcriptome analysis of a CHO cell line expressing a recombinant therapeutic protein treated with inducers of protein expression

The search for specific productivity (qP) determinants in Chinese hamster ovary (CHO) cells has been the focus of the biopharmaceutical cell line engineering efforts aimed at creating "super-producer" cell lines. In this study, we evaluated the impact of small-molecule inducers and temperature shift on recombinant protein production, and used transcriptomic analysis to define gene-phenotype correlations for qP in our biological system. Next-generation RNA Sequencing (RNA-Seq) analysis revealed that each individual inducer (caffeine, hexamethylene bisacetamide (HMBA) and sodium butyrate (NaBu)) or a combination treatment had a distinct impact on the gene expression program of the RANK-Fc cell line. Temperature shift to 31 °C impacted inducer action with respect to transcriptional changes and phenotypic cell line parameters. We showed that inducer treatment was able to increase expression level of the Fc- fusion mRNA and the selectable marker mRNA from 16% up to 45% of total mRNA in the cell. We further demonstrated that qP exhibited a strong positive linear correlation to transcript levels of both the RANK-Fc fusion protein and the dihydrofolate reductase (DHFR) selectable marker. In fact, these were 2 out of 7 transcripts with significant positive correlation to qP at both temperatures. Many more transcripts were anti- correlated to qP, and gene set enrichment analysis (GSEA) revealed that those were involved in cell cycle progression, transcription, mRNA processing, translation and protein folding. Therefore, we postulate that the transcript level of the recombinant protein is a major qP determinant in our biological system, while downregulation of routine activity within the cell is necessary to divert cellular resources towards recombinant protein production.

Seq-ing improved gene expression estimates from microarrays using machine learning

BACKGROUND

Quantifying gene expression by RNA-Seq has several advantages over microarrays, including greater dynamic range and gene expression estimates on an absolute, rather than a relative scale. Nevertheless, microarrays remain in widespread use, demonstrated by the ever-growing numbers of samples deposited in public repositories.

RESULTS

We propose a novel approach to microarray analysis that attains many of the advantages of RNA-Seq. This method, called Machine Learning of Transcript Expression (MaLTE), leverages samples for which both microarray and RNA-Seq data are available, using a Random Forest to learn the relationship between the fluorescence intensity of sets of microarray probes and RNA-Seq transcript expression estimates. We trained MaLTE on data from the Genotype-Tissue Expression (GTEx) project, consisting of Affymetrix gene arrays and RNA-Seq from over 700 samples across a broad range of human tissues.

CONCLUSION

This approach can be used to accurately estimate absolute expression levels from microarray data, at both gene and transcript level, which has not previously been possible. This methodology will facilitate re-analysis of archived microarray data and broaden the utility of the vast quantities of data still being generated.

Deep sequencing reveals different compositions of mRNA transcribed from the F8 gene in a panel of FVIII-producing CHO cell lines

Coagulation factor VIII (FVIII) is one of the most complex biopharmaceuticals due to the large size, poor protein stability and extensive post-translational modifications. As a consequence, efficient production of FVIII in mammalian cells poses a major challenge, with typical yields two to three orders of magnitude lower than for antibodies. In the present study we investigated CHO DXB11 cells transfected with a plasmid encoding human coagulation factor VIII. Single cell clones were isolated from the pool of transfectants and a panel of 14 clones representing a dynamic range of FVIII productivities was selected for RNA sequencing analysis. The analysis showed distinct differences in F8 RNA composition between the clones. The exogenous F8-dhfr transcript was found to make up the most abundant transcript in the present clones. No correlation was seen between F8 mRNA levels and the measured FVIII productivity. It was found that three MTX resistant, nonproducing clones had different truncations of the F8 transcripts. We find that by using deep sequencing, in contrast to microarray technology, for determining the transcriptome from CHO transfectants, we are able to accurately deduce the mature mRNA composition of the transgene and identify significant truncations that would probably otherwise have remained undetected.

Gene expression measurements normalized to cell number reveal large scale differences due to cell size changes, transcriptional amplification and transcriptional repression in CHO cells

Conventional approaches to differential gene expression comparisons assume equal cellular RNA content among experimental conditions. We demonstrate that this assumption should not be universally applied because total RNA yield from a set number of cells varies among experimental treatments of the same Chinese Hamster Ovary (CHO) cell line and among different CHO cell lines expressing recombinant proteins. Conventional normalization strategies mask these differences in cellular RNA content and, consequently, skew biological interpretation of differential expression results. On the contrary, normalization to synthetic spike-in RNA standards added proportional to cell numbers reveals these differences and allows detection of global transcriptional amplification/repression. We apply this normalization method to assess differential gene expression in cell lines of different sizes, as well as cells treated with a cell cycle inhibitor (CCI), an mTOR inhibitor (mTORI), or subjected to high osmolarity conditions. CCI treatment of CHO cells results in a cellular volume increase and global transcriptional amplification, while mTORI treatment causes global transcriptional repression without affecting cellular volume. Similarly to CCI treatment, high osmolarity increases cell size, total RNA content and antibody expression. Furthermore, we show the importance of spike-in normalization for studies involving multiple CHO cell lines and advocate normalization to spike-in controls prior to correlating gene expression to specific productivity (qP). Overall, our data support the need for cell number specific spike-in controls for all gene expression studies where cellular RNA content differs among experimental conditions.

Beyond modules and hubs: the potential of gene coexpression networks for investigating molecular mechanisms of complex brain disorders

In a research environment dominated by reductionist approaches to brain disease mechanisms, gene network analysis provides a complementary framework in which to tackle the complex dysregulations that occur in neuropsychiatric and other neurological disorders. Gene-gene expression correlations are a common source of molecular networks because they can be extracted from high-dimensional disease data and encapsulate the activity of multiple regulatory systems. However, the analysis of gene coexpression patterns is often treated as a mechanistic black box, in which looming 'hub genes' direct cellular networks, and where other features are obscured. By examining the biophysical bases of coexpression and gene regulatory changes that occur in disease, recent studies suggest it is possible to use coexpression networks as a multi-omic screening procedure to generate novel hypotheses for disease mechanisms. Because technical processing steps can affect the outcome and interpretation of coexpression networks, we examine the assumptions and alternatives to common patterns of coexpression analysis and discuss additional topics such as acceptable datasets for coexpression analysis, the robust identification of modules, disease-related prioritization of genes and molecular systems and network meta-analysis. To accelerate coexpression research beyond modules and hubs, we highlight some emerging directions for coexpression network research that are especially relevant to complex brain disease, including the centrality-lethality relationship, integration with machine learning approaches and network pharmacology.

Alternative promoters regulate cold inducible RNA-binding (CIRP) gene expression and enhance transgene expression in mammalian cells

The use of a temperature shift cultivation to enhance recombinant protein yield is widely utilised in the bioprocessing industry. The responses of mammalian cells to heat stress are well characterized; however, the equivalent cold stress responses are not. In particular, the transcriptional mechanisms that lead to enhanced gene-specific expression upon cold stress have yet to be elucidated. We report here in silico and experimental identification and characterization of transcriptional control elements that regulate cold inducible RNA-binding (CIRP) gene expression and demonstrate these can be used for enhanced transgene expression. In silico analysis identified the core CIRP promoter and a number of conserved transcription factor-binding sites across mammalian species. The core promoter was confirmed by experimental studies that located the basal transcriptional regulatory elements of CIRP within 264 nucleotides upstream of the transcription start site. Deletion analysis of a fragment from -264 to -64 that contained two putative CAAT-binding sites abolished promoter activity. A second promoter was identified in the region -452 to -264 of the transcription start site which was able to drive transcription independent of the core promoter. As the two CIRP promoters were transcriptionally active and possibly cold responsive, we used electrophoretic mobility shift assays to show that both promoter regions are able to bind factors within a nuclear extract in a dose-dependent manner and that the formation of these complexes was specific to the promoter regions. Finally, we successfully demonstrate using a reporter gene approach that enhanced transgene expression can be achieved using the identified CIRP promoter.

Proteins improving recombinant antibody production in mammalian cells

Mammalian cells have been successfully used for the industrial manufacture of antibodies due to their ability to synthesize antibodies correctly. Nascent polypeptides must be subjected to protein folding and assembly in the ER and the Golgi to be secreted as mature proteins. If these reactions do not proceed appropriately, unfolded or misfolded proteins are degraded by the ER-associated degradation (ERAD) pathway. The accumulation of unfolded proteins or intracellular antibody crystals accompanied by this failure triggers the unfolded protein response (UPR), which can considerably attenuate the levels of translation, folding, assembly, and secretion, resulting in reduction of antibody productivity. Accumulating studies by omics-based analysis of recombinant mammalian cells suggest that not only protein secretion processes including protein folding and assembly but also translation are likely to be the rate-limiting factors for increasing antibody production. Here, this review describes the mechanism of antibody folding and assembly and recent advantages which could improve recombinant antibody production in mammalian cells by utilizing proteins such as ER chaperones or UPR-related proteins.

Whole brain and brain regional coexpression network interactions associated with predisposition to alcohol consumption

To identify brain transcriptional networks that may predispose an animal to consume alcohol, we used weighted gene coexpression network analysis (WGCNA). Candidate coexpression modules are those with an eigengene expression level that correlates significantly with the level of alcohol consumption across a panel of BXD recombinant inbred mouse strains, and that share a genomic region that regulates the module transcript expression levels (mQTL) with a genomic region that regulates alcohol consumption (bQTL). To address a controversy regarding utility of gene expression profiles from whole brain, vs specific brain regions, as indicators of the relationship of gene expression to phenotype, we compared candidate coexpression modules from whole brain gene expression data (gathered with Affymetrix 430 v2 arrays in the Colorado laboratories) and from gene expression data from 6 brain regions (nucleus accumbens (NA); prefrontal cortex (PFC); ventral tegmental area (VTA); striatum (ST); hippocampus (HP); cerebellum (CB)) available from GeneNetwork. The candidate modules were used to construct candidate eigengene networks across brain regions, resulting in three "meta-modules", composed of candidate modules from two or more brain regions (NA, PFC, ST, VTA) and whole brain. To mitigate the potential influence of chromosomal location of transcripts and cis-eQTLs in linkage disequilibrium, we calculated a semi-partial correlation of the transcripts in the meta-modules with alcohol consumption conditional on the transcripts' cis-eQTLs. The function of transcripts that retained the correlation with the phenotype after correction for the strong genetic influence, implicates processes of protein metabolism in the ER and Golgi as influencing susceptibility to variation in alcohol consumption. Integration of these data with human GWAS provides further information on the function of polymorphisms associated with alcohol-related traits.

Translatome analysis of CHO cells to identify key growth genes

We report the first investigation of translational efficiency on a global scale, also known as translatome, of a Chinese hamster ovary (CHO) DG44 cell line producing monoclonal antibodies (mAb). The translatome data was generated via combined use of high resolution and streamlined polysome profiling technology and proprietary Nimblegen microarrays probing for more than 13K annotated CHO-specific genes. The distribution of ribosome loading during the exponential growth phase revealed the translational activity corresponding to the maximal growth rate, thus allowing us to identify stably and highly translated genes encoding heterogeneous nuclear ribonucleoproteins (Hnrnpc and Hnrnpa2b1), protein regulator of cytokinesis 1 (Prc1), glucose-6-phosphate dehydrogenase (G6pdh), UTP6 small subunit processome (Utp6) and RuvB-like protein 1 (Ruvbl1) as potential key players for cellular growth. Moreover, correlation analysis between transcriptome and translatome data sets showed that transcript level and translation efficiency were uncoupled for 95% of investigated genes, suggesting the implication of translational control mechanisms such as the mTOR pathway. Thus, the current translatome analysis platform offers new insights into gene expression in CHO cell cultures by bridging the gap between transcriptome and proteome data, which will enable researchers of the bioprocessing field to prioritize in high-potential candidate genes and to devise optimal strategies for cell engineering toward improving culture performance.

Correlating transcriptional networks to breast cancer survival: a large-scale coexpression analysis

Weighted gene coexpression network analysis (WGCNA) is a powerful 'guilt-by-association'-based method to extract coexpressed groups of genes from large heterogeneous messenger RNA expression data sets. We have utilized WGCNA to identify 11 coregulated gene clusters across 2342 breast cancer samples from 13 microarray-based gene expression studies. A number of these transcriptional modules were found to be correlated to clinicopathological variables (e.g. tumor grade), survival endpoints for breast cancer as a whole (disease-free survival, distant disease-free survival and overall survival) and also its molecular subtypes (luminal A, luminal B, HER2+ and basal-like). Examples of findings arising from this work include the identification of a cluster of proliferation-related genes that when upregulated correlated to increased tumor grade and were associated with poor survival in general. The prognostic potential of novel genes, for example, ubiquitin-conjugating enzyme E2S (UBE2S) within this group was confirmed in an independent data set. In addition, gene clusters were also associated with survival for breast cancer molecular subtypes including a cluster of genes that was found to correlate with prognosis exclusively for basal-like breast cancer. The upregulation of several single genes within this coexpression cluster, for example, the potassium channel, subfamily K, member 5 (KCNK5) was associated with poor outcome for the basal-like molecular subtype. We have developed an online database to allow user-friendly access to the coexpression patterns and the survival analysis outputs uncovered in this study (available at http://glados.ucd.ie/Coexpression/).

Transcriptomic analysis of clonal growth rate variation during CHO cell line development

The selection of clones displaying a high rate of cell growth is an essential component of Chinese hamster ovary (CHO) cell line development. In recent years various "omics" technologies have been utilised to understand the mechanisms underlying bioprocess phenotypes. In this study, gene expression analysis using a CHO-specific microarray was conducted for a panel of CHO-K1 MAb-secreting cell lines spanning a range of growth rates that were derived from a single cell line development project. In-silico functional analysis of the resulting transcriptomic data revealed the overrepresentation of biological processes such as cell cycle and translation within those genes upregulated during fast growth, while genes associated with cellular homeostasis were downregulated. Using differential expression and correlation analysis we identified a high priority group of 416 transcripts (190 upregulated; 226 downregulated) associated with growth rate. Expression changes of eight of these genes were independently confirmed by qPCR. Finally, we demonstrate the enrichment of predicted mRNA targets of miR17-92, a microRNA (miRNA) cluster known to be upregulated during rapid proliferation, within downregulated transcripts.

Integrated miRNA, mRNA and protein expression analysis reveals the role of post-transcriptional regulation in controlling CHO cell growth rate

Background

To study the role of microRNA (miRNA) in the regulation of Chinese hamster ovary (CHO) cell growth, qPCR, microarray and quantitative LC-MS/MS analysis were utilised for simultaneous expression profiling of miRNA, mRNA and protein. The sample set under investigation consisted of clones with variable cellular growth rates derived from the same population. In addition to providing a systems level perspective on cell growth, the integration of multiple profiling datasets can facilitate the identification of non-seed miRNA targets, complement computational prediction tools and reduce false positive and false negative rates.

Results

51 miRNAs were associated with increased growth rate (35 miRNAs upregulated and 16 miRNAs downregulated). Gene ontology (GO) analysis of genes (n=432) and proteins (n=285) found to be differentially expressed (DE) identified biological processes driving proliferation including mRNA processing and translation. To investigate the influence of miRNA on these processes we combined the proteomic and transcriptomic data into two groups. The first set contained candidates where evidence of translational repression was observed (n=158). The second group was a mixture of proteins and mRNAs where evidence of translational repression was less clear (n=515). The TargetScan algorithm was utilised to predict potential targets within these two groups for anti-correlated DE miRNAs.

Conclusions

The evidence presented in this study indicates that biological processes such as mRNA processing and protein synthesis are correlated with growth rate in CHO cells. Through the integration of expression data from multiple levels of the biological system a number of proteins central to these processes including several hnRNPs and components of the ribosome were found to be post-transcriptionally regulated. We utilised the expression data in conjunction with in-silico tools to identify potential miRNA-mediated regulation of mRNA/proteins involved in CHO cell growth rate. These data have allowed us to prioritise candidates for cell engineering and/or biomarkers relevant to industrial cell culture. We also expect the knowledge gained from this study to be applicable to other fields investigating the role of miRNAs in mammalian cell growth.

Recent advances in technology supporting biopharmaceutical production from mammalian cells

The demand for production of glycoproteins from mammalian cell culture continues with an increased number of approvals as biopharmaceuticals for the treatment of unmet medical needs. This is particularly the case for humanized monoclonal antibodies which are the largest and fastest growing class of therapeutic pharmaceuticals. This demand has fostered efforts to improve the efficiency of production as well as to address the quality of the final product. Chinese hamster ovary cells are the predominant hosts for stable transfection and high efficiency production on a large scale. Specific productivity of recombinant glycoproteins from these cells can be expected to be above 50 pg/cell/day giving rise to culture systems with titers of around 5 g/L if appropriate fed-batch systems are employed. Cell engineering can delay the onset of programmed cell death to ensure prolonged maintenance of productive viable cells. The clinical efficacy and quality of the final product can be improved by strategic metabolic engineering. The best example of this is the targeted production of afucosylated antibodies with enhanced antibody-dependent cell cytotoxicity, an important function for use in cancer therapies. The development of culture media from non-animal sources continues and is important to ensure products of consistent quality and without the potential danger of contamination. Process efficiencies may also be improved by employing disposable bioreactors with the associated minimization of downtime. Finally, advances in downstream processing are needed to handle the increased supply of product from the bioreactor but maintaining the high purity demanded of these biopharmaceuticals.

Integrative gene regulatory network analysis reveals light-induced regional gene expression phase shift programs in the mouse suprachiasmatic nucleus

We use the multigenic pattern of gene expression across suprachiasmatic nuclei (SCN) regions and time to understand the dynamics within the SCN in response to a circadian phase-resetting light pulse. Global gene expression studies of the SCN indicate that circadian functions like phase resetting are complex multigenic processes. While the molecular dynamics of phase resetting are not well understood, it is clear they involve a “functional gene expression program”, e.g., the coordinated behavior of functionally related genes in space and time. In the present study we selected a set of 89 of these functionally related genes in order to further understand this multigenic program. By use of high-throughput qPCR we studied 52 small samples taken by anatomically precise laser capture from within the core and shell SCN regions, and taken at time points with and without phase resetting light exposure. The results show striking regional differences in light response to be present in the mouse SCN. By using network-based analyses, we are able to establish a highly specific multigenic correlation between genes expressed in response to light at night and genes normally activated during the day. The light pulse triggers a complex and highly coordinated network of gene regulation. The largest differences marking neuroanatomical location are in transmitter receptors, and the largest time-dependent differences occur in clock-related genes. Nighttime phase resetting appears to recruit transcriptional regulatory processes normally active in the day. This program, or mechanism, causes the pattern of core region gene expression to transiently shift to become more like that of the shell region.

Rapid temporal changes in the expression of a set of neuromodulatory genes during alcohol withdrawal in the dorsal vagal complex: molecular evidence of homeostatic disturbance

BACKGROUND

Chronic alcohol exposure produces neuroadaptation, which increases the risk of cellular excitotoxicity and autonomic dysfunction during withdrawal. The temporal progression and regulation of the gene expression that contributes to this physiologic and behavioral phenotype is poorly understood early in the withdrawal period. Further, it is unexplored in the dorsal vagal complex (DVC), a brainstem autonomic regulatory structure.

METHODS

We use a quantitative polymerase chain reaction platform to precisely and simultaneously measure the expression of 145 neuromodulatory genes in more than 100 rat DVC samples from control, chronically alcohol-exposed, and withdrawn rats. To gain insight into the dynamic progression and regulation of withdrawal, we focus on the expression of a subset of functionally relevant genes during the first 48 hours, when behavioral symptoms are most severe.

RESULTS

In the DVC, expression of this gene subset is essentially normal in chronically alcohol-exposed rats. However, withdrawal results in rapid, large-magnitude expression changes in this group. We observed differential regulation in 86 of the 145 genes measured (59%), some as early as 4 hours into withdrawal. Time series measurements (4, 8, 18, 32, and 48 hours after alcohol removal) revealed dynamic expression responses in immediate early genes, γ-aminobutyric acid type A, ionotropic glutamate, and G-protein coupled receptors and the Ras/Raf signaling pathway. Together, these changes elucidate a complex, temporally coordinated response that involves correlated expression of many functionally related groups. In particular, the expression patterns of Gabra1, Grin2a, Grin3a, and Grik3 were tightly correlated. These receptor subunits share overrepresented transcription factor binding sites for Pax-8 and other transcription factors, suggesting a common regulatory mechanism and a role for these transcription factors in the regulation of neurotransmission within the first 48 hours of alcohol withdrawal.

CONCLUSIONS

Expression in this gene set is essentially normal in the alcohol-adapted DVC, but withdrawal results in immediate, large-magnitude, and dynamic changes. These data support both increased research focus on the biological ramifications of alcohol withdrawal and enable novel insights into the dynamic withdrawal expression response in this understudied homeostatic control center.