October 15, 2016 (Vol. 36, No. 18)

Richard A. A. Stein M.D., Ph.D.

Drug Developers Are Finding Ways to Bend the Transcriptome to Their Will

Among the key obstacles in drug discovery and development are effective data integration, target identification, increasing costs, high failure rates, and the difficulty to predict adverse effects.

In addition to challenges that are shared by virtually all compounds being developed, specific classes of therapeutic agents face additional bottlenecks.

Recent advances that incorporate gene expression analyses into drug discovery efforts promise to catalyze the development of novel, safer, and more effective therapeutics, as well as decrease costs and shorten the time for translating benefits from the bench to the bedside.

“There are many animal studies that knocked out single genes and characterized the resulting perturbations through changes in those specific genes, but transcriptome analyses showed that the resulting phenotypes resulted from changes in the landscape of the entire system, as a result of the dysregulation of many other genes,” says Krzysztof Palczewski, Ph.D., distinguished university professor and chair of the department of pharmacology at Case Western Reserve University School of Medicine.

The up- or downregulation of many genes as a result of perturbations in a single gene is illustrated by biochemistry experiments, where mutagenesis of a single amino acid may establish a new thermodynamic equilibrium resulting from dysregulation of several pathways in which the protein is involved. “Many of the drugs that fail on long-term tests or show adverse effects do so because they cause other transcripts to become up- or downregulated and make the system unstable,” says Dr. Palczewski.

In a recent study, Dr. Palczewski and colleagues used a systems pharmacology approach and showed that a combination of FDA-approved drugs that act on G-protein-coupled receptors (GPCRs) protected mice with bright light-induced retinal degeneration. Two combinations of three therapeutics—bromocriptine, metoprolol, and tamsulosin, or bromocriptine, metoprolol, and doxazosin—each of them individually at therapeutically subeffective doses, achieved morphological and functional protection of the photoreceptor, bipolar, and horizontal cells. The compounds acted by modulating GPCR signaling through adrenergic and dopaminergic ligands.

Using transcriptome analyses, Dr. Palczewski and colleagues revealed that these combination treatments preserved a genome-wide retinal gene expression pattern that was comparable to the one seen in undamaged retinas. The possibility to target GPCRs pharmacologically promises the use of this therapeutic strategy for other complex diseases. One of the technical challenges in systems pharmacology is the vast datasets that are being generated.

“Approaches that we now have at our disposal, such as RNA sequencing (RNA-Seq), have become more precise, and the throughput and coverage are high. All these things are in place, but the bioinformatics will be the talking point, because so much data is generated that extracting the noise from the signal will be one of the major bottlenecks,” says Dr. Palczewski.


Case Western Reserve University scientists in the laboratory of Krzysztof Palczewski, Ph.D., found that a combination of FDA-approved drugs could protect photoreceptor cells against light-induced damage. The combination therapy, which reflects a systems pharmacology approach, was shown to be effective with four different imaging techniques: optical coherent tomography (OCT), scanning laser ophthalmoscopy (SLO), immunocytochemistry for calbindin, and two-photon microscopy (TPM).

Gene Expression in Profiling iPSC-Derived Cells

“We are performing RNA-Seq and gene expression analyses to examine gene expression profiles before and after drug treatment,” says Joseph Wu, M.D., Ph.D., Simon H. Stertzer, M.D. professor of medicine and radiology at Stanford University School of Medicine and director of the Stanford Cardiovascular Institute.

For most drug discovery studies performed as recently as a decade ago, after drugs were administered to human patients or experimental animals, RNA-Seq or microarray studies were performed on mononuclear blood cells to visualize their effects on gene expression. “The problem is that when studying medications that target specific organs, such as the brain or the heart, one cannot use blood cells as a surrogate for the drug response in other organs,” says Dr. Wu.

A major effort in Dr. Wu’s lab involves differentiating induced pluripotent stem cells (iPSCs) cells into other cell types and treating the differentiated cells with medications. “This allows us to interrogate the effect of specific drugs, such as beta-blockers, angiotensin converting enzyme (ACE) inhibitors, and chemotherapy drugs on the patient’s organs, not just on the patient’s blood,” says Dr. Wu.

In a study to examine the effect of the cellular origin on differentiation, phenotype, and gene expression signature of iPSC-derived endothelial cells, Dr. Wu and colleagues used fibroblasts, endothelial cells, and cardiac progenitor cells from the same individuals to generate human iPSCs and subsequently differentiated them into endothelial cells.

Endothelial cell-derived iPSCs at under 20 passages showed a higher differentiation propensity toward endothelial cells and expressed more endothelial cell-specific gene expression markers as compared to fibroblast-derived and cardiac progenitor cell-derived iPSCs, showing that their in vitro and in vivo identity was different from the one of endothelial cells derived from fibroblast-derived and cardiac progenitor cell-derived iPSCs.

This indicated that the somatic memory carried by early-passage iPSCs influences lineage differentiation propensity, a therapeutically relevant and actionable finding. “Thanks to the discovery of iPSCs, the next phase in drug discovery will be a lot of gene expression profiling on these specific cell types rather than on blood cells or lymphoblastoid cell lines,” explains Dr. Wu.

These developments also promise a conceptual shift from the existing strategy, in which virtually every drug discovery effort relies on animal, frequently rodent, studies. The divergence between humans and rodents approximately 75 million years ago resulted in marked differences between human and rodent physiology, presenting challenges in understanding the human relevance of rodent experiments.

The self-renewal of iPSCs, the possibility to differentiate them into any cell type, and the use of iPSCs derived from healthy individuals and from those with specific conditions make iPSCs an important alternative to animal studies that model human diseases. Drug discovery research can be performed directly on patient-specific and disease-specific iPSC-derived cells.

“We are currently developing a large biobank of patient- and disease-specific iPSCs, and we are also investigating the effect of various classes of medications on gene expression in broad patient groups,” says Dr. Wu. This biobank offers opportunities to capture interindividual variability concomitantly from thousands of patients—“drug discovery in a dish,” adds Dr. Wu.

Performing clinical trials in a dish promises benefits in terms of time, efforts, and costs, which are among the major bottlenecks in identifying the best compounds during drug discovery. “With an RNA microarray or a PCR reaction as the readout, one may be able to quickly distinguish nonresponders from responders and also figure out why a specific drug does not work on nonresponders,” Dr. Wu adds.


At the Stanford Cardiovascular Institute, Ming-Tao Zhao, Ph.D., and colleagues obtained these images of endothelial cells (ECs) derived from human induced pluripotent stem cells (iPSCs). Notice the expression, on the cell membranes, of EC-specific markers (CD31, green; CD144, red). The ECs were originally derived from endothelial cells (EC-iPSC-ECs, left) and fibroblasts (FB-iPSC-ECs, right). The nuclei were counterstained by Hoechst 33342.

Controlling Gene Expression with Synthetic Nucleic Acids

“For many years, the overall direction in my laboratory has been to investigate the use of synthetic nucleic acids to control gene expression,” says David R. Corey, Ph.D., Rusty Kelley professor of medical sciences in the department of pharmacology and biochemistry at the University of Texas Southwestern Medical Center. Advances in nucleic acid chemistry have helped identify therapeutic targets and develop approaches to modulate their expression. “We recognize our limitations, that our job is not to make a drug, but to bring an area of science far enough so that companies with more resources can capitalize on a starting point for further investigations in a particular area,” explains Dr. Corey.

The RNA transcribed from the vast majority of the human genome does not encode proteins and is called noncoding RNA (ncRNA). ncRNAs have become valuable targets for drug discovery due to their ability to modulate gene expression in multiple ways. Dr. Corey and colleagues have shown that synthetic single-stranded microRNAs (ss-miRNAs) can mimic the function of miRNAs and silence the expression of target genes through the RNA interference (RNAi) pathway.

“Single-stranded RNAs do not behave like double-stranded RNAs (dsRNAs), but they are pretty close,” explains Dr. Corey. He and his co-workers showed that a ss-miRNA mimic of miR-34a reduced the expression of SIRT1 and other targets, and this inhibition requires expression of the Argonaute 2 protein. “The question is how well the ss-miRNAs work in vivo, and whether they can be competitive with dsRNAs and antisense nucleotides,” says Dr. Corey.

An important challenge is to gain a deep understanding of not only whether an agent activates or inhibits a specific gene, but also of how it is doing so. Particularly for miRNAs, which are regulated by several messenger RNAs (mRNAs) and function through many different mRNAs, this may be challenging.

Modulating Bacterial Expression with Synthetic DNA

“We are trying to develop therapeutics to modulate the expression of bacterial genes,” says Bruce L. Geller, Ph.D., professor of microbiology at Oregon State University. The strategy developed in Dr. Geller’s lab involves generating a synthetic DNA mimic that binds to specific bacterial mRNAs based on sequence complementary.

“The synthetic DNA molecule binds to an essential bacterial gene and reduces or silences the expression of those genes and, as a result, the bacterial cell will die,” Dr. Geller explains. This approach, which is novel and different from the mechanisms used by currently approved antibiotics, also works on some antimicrobial-resistant pathogens, which have become particularly challenging to treat as antibiotic resistance has been emerging increasingly as a worldwide concern. 

In a recent study, Dr. Geller and colleagues showed that peptide-conjugated phosphorodiamidate morpholino oligomers (PPMOs) can target essential genes in clinically relevant Acinetobacter species in vitro and in vivo. Acinetobacter infections are particularly challenging due to antibiotic resistance. PPMOs designed to bind complementary bases in the mRNAs of essential Acinetobacter genes conjugated to membrane-penetrating peptides and exhibited bactericidal effects at clinically relevant minimum inhibitory concentrations, even against multiresistant strains, thus promising a viable therapeutic approach.

Researchers in Dr. Geller’ lab showed that PPMOs were able to restore susceptibility to carbapenems in Escherichia coli, Pseudomonas aeruginosa, and Acinetobacter baumanii, three therapeutically challenging Gram-negative pathogens. “We were able to use combinations of the antibiotic with gene-specific PPMOs to kill the carbapenem-resistant pathogens,” says Dr. Geller. Carbapenems, the most potent type of beta-lactam antibiotics, are often the last drug of choice for many bacterial pathogens, and targeting PPMOs to genes that cause carbapenem resistance can restore the sensitivity or susceptibility of the pathogens.

Targeting bacterial genes with synthetic DNA eliminates undesired effects on eukaryotic cells, which do not harbor the same resistance genes and are unlikely to be affected by oligomers directed against bacterial genes. “These oligomers have a lot of great promise, and this is a direction to which the future of medicine is headed,” concludes Dr. Geller. 

Achieving Tissue- or Cell-Specific Gene Expression In Vivo

Tissue- or cell-specific gene expression is essential for gene function study in both gene knockout and gene overexpression experiments. Two technologies enable accurate tissue-specific gene expression in vivo.

One involves AAVs (adeno-associated virus). Thanks to the capsid proteins, different serotypes of AAV have distinct tissue/organ specificity of transduction and gene delivery. The other relies on tissue-specific promoters. A large collection of tissue specific promoters are available.

When combined, the AAV and tissue-specific promoter can achieve remarkable targeting of transgene expression, making possible tissue-specific knock-in or knockout and tissue-specific gene expression.

According to Jeffrey Hung, Ph.D.,  chief commercial officer at Vigene Bioscience, all the major AAV serotypes and tissue-specific promoters are available from his company.

“Recently the advancement of Cre-based conditional gene knockout or knock-in further underlined the need for tissue-targeted gene expression,”  he said.  “AAV and tissue-specific promoter are critical for controlling the tissue-specific cre/lop based gene knockout or gene expression.” 

By injecting AAV vector carrying tissue specific Cre into animals, one can potentially save a lot of time, bypassing the need to cross and breed transgenic animals, added Dr. Hung. 

The tissues can be targeted by different AAVs are muscles, liver, lung, neurons, glial cells, retina, pancreas, and kidney. The tissue-specific promoters available include universal promoters (CAG, EF1A, EFFS, UBC, PGK), liver promoters (ALB, TBG), neuron promoters (CamKIIa, Synapsin, Mecp2, NSE), astrocyte (GFAP), and cardiomyocytes (cTNT), retinal pigment (Rpe65), Beta-cells (Insulin1), to name a few.

Previous articleDolphin Proteins May Hold the Secret to Protection against Hypoxia
Next articleTapping the Human Gut Microbiome, Part 1