Predicting drugs’ side effects, new uses

Bryan Roth, MD, PhD

It should come as no surprise that lots of target-specific drugs are no so targeted. After all, many of the most successful drugs on the market today are being prescribed for uses that are quite different from the ones they were originally designed to treat.

Sometimes serendipity in the form of unforeseen side effects happens. That’s how today’s Viagra avoided the pharmaceutical dustbin.  Having flunked a Phase 1 clinical trial as a heart drug for angina, researchers noted it was associated with penile erections. The rest is history, or herstory… whatever. 

Now, scientists at the UNC School of Medicine and the University of  California, San Francisco, have developed and tested a computational technique to predict a new drug’s side effects and new target diseases for existing drugs.

Their article published online November 1st in Nature, describes a “chemical similarity” approach using “statistics-based chemoinformatics” that compares drug targets by the similarity of their mutual binding partners, or ligands.

Co-senior study author Bryan Roth, M.D., Ph.D., professor of pharmacology and director of the National Institute of Mental Health Psychoactive Drug Screening Program at UNC, says  “We may now have a way to predict what side effects are likely to occur from treatment before we even put a drug into clinical testing.” 

Read our news release here.

Les Lang

FLASH dances with death and mRNA

Whether it was Daniel DeFoe or Benjamin Franklin, whoever coined the phrase, “Nothing is certain but death and taxes,” clearly hadn’t a clue of another of life’s immutable pairings:  histone proteins and DNA.

Histones are the chief protein components of chromatin — the protein/DNA combo of which chromosomes are made — and they act as a scaffold to allow the packaging of DNA into a condensed form that fits inside the nucleus of a cell. As the DNA interacts with histones and with metabolic signals from within the cell, these proteins help regulate gene expression.

But how does the histone component of this dynamic duo get started in the first place?  The answer, according to new research from UNC,  appears in [a] FLASH.

This protein is already shown to play a role in initiating apoptosis, or the normal process of programmed cell death, and is now the latest key player in the molecular dance essential for DNA replication within cells.

According to senior study author Zbigniew Dominski, Ph.D., associate professor of biochemistry and biophysics, the new study demonstrates that FLASH is  required for the proper synthesis of histone messenger RNA, which gives rise to histone proteins.

“FLASH is crucial for the production of histone messenger RNA, without which the cell can’t make the histone proteins around which DNA is packaged,” says Dominski. “Our study suggests for the first time that a potential link exists between the processes of histone messenger RNA formation and apoptosis.”

The research is described in the October 23^rd issue of the journal Molecular Cell.

For the study, Dominski adapted a laboratory system that reproduces in the test tube what normally occurs in the cell when FLASH participates in the biochemical cleavage event that results in mature histone messenger RNA. This enabled his team to explore what might occur when FLASH was added or removed.

“We could then figure out exactly what portion of FLASH would restore the protein’s function in generating histone mRNAs and remarkably, only the first 100 or so amino acids are required.  The remaining 2,000 amino acids of this large protein likely control other processes in the cell, including apoptosis and DNA replication” he explained.

Co-author William F. Marzluff, Ph.D., is Distinguished Professor of biochemistry & biophysics. He notes that FLASH is the first component found in this protein complex that integrates or initiates many cellular functions — DNA replication, apoptosis, histone production.  “Having this small piece of the puzzle makes it a lot easier to identify others.”

Except for a certain somebody who originally paired the certainty of death and taxes.

Les Lang

From gene deletion to schizophrenia?

Scientists have long thought the faulty neural wiring that predisposes individuals to behavioral disorders like autism and psychiatric diseases like schizophrenia must occur during development. But no one has ever shown that a risk gene for the disease actually disrupts brain development.

Now, UNC scientists  have found that the 22q11 gene deletion on chromosome 22 – a mutation that confers the highest known genetic risk for schizophrenia – is associated with changes in the development of the brain that ultimately affect how its circuit elements are assembled.

In a genetic animal model, the researchers discovered that the gene deletion alters the number of a critical subset of neurons that end up in the brain’s cerebral cortex – the region critical to reasoning and memory. The defect also causes another type of nerve cell – called GABAergic neurons – to be misplaced within the brain’s cortical layers, resulting in a subtle miswiring of the organ.

“For practically ever other disease, we know what cells take a hit,” said senior study author Anthony LaMantia, Ph.D., professor of cell and molecular physiology and co-director of the Silvio M. Conte Center for Research in Mental Disorders at the UNC School of Medicine.

“For multiple sclerosis the myelinating oligodendrocytes in the brain falter, for Lou Gehrig’s disease the motor neurons in the brain stem degenerate. But we really had no idea what was happening in schizophrenia, or in any of the psychiatric diseases for that matter – until now.”

The study was presented October 17, 2009  at the Society for Neuroscience meeting in Chicago, by Daniel Meechan, Ph.D., post-doctoral fellow in the LaMantia laboratory and the first author of the recent paper in Proceedings of the National Academy of Sciences that details the findings.

Les Lang

GO accelerator for genome research

Ever since the first genome sequence was published in 2001, scientists have been working to figure out what the sequence means.  An analogy is walking across a desert and finding a large book in a language you don’t know, then trying to figure out what the book is saying.

“In the case of the human genome, the book is a blueprint to building cells—and ultimately—the whole human.  But we don’t yet understand its language,” said Morgan Giddings, Ph.D., associate professor of microbiology and immunology and of biomedical engineering at the University of North Carolina at Chapel Hill.

Morgan Giddings, Ph.D., left, and Xian Chen, Ph.D., right. Photo by Courtney Potter

Giddings and UNC colleague Xian Chen, Ph.D., associate professor of biochemistry and biophysics, have been developing methods for decoding the human blueprint by studying the things the blueprint produces: proteins.  They relate the proteins back to the blueprint itself, to further decode the language of the genome blueprint.

Giddings and Chen have been awarded a $1.6 million 2-year “Grand Opportunities” (GO) grant from the National Human Genome Research Institute to accelerate this research.  Their effort will be part of a consortium of investigators studying the human genome blueprint, titled the “ENCyclopedia Of DNA Elements” (ENCODE).  The consortium’s overall goal is to assemble a comprehensive catalog of functional elements in the human genome.

With their GO grant, Giddings and Chen will generate, analyze, and release to the public large-scale data sets that allow linking of the protein products in cells to their genomic blueprints.  According to Giddings, “this will significantly promote our understanding of the language of the human genome, enhancing efforts to solve pressing human health issues like heart disease and cancer by understanding how errors in the blueprint lead to disease, and how we might fix those errors.”

Giddings is a member of the Carolina Center for Genome Sciences, and Chen is technology development director for the UNC Proteomics Core.

As to impact  on the local North Carolina economy, the new grant is expected to  bring 4-6 new high-tech jobs to the Triangle.

NHGRI has awarded approximately $22 million of American Recovery and Reinvestment Act 2009 funds to support research aimed at identifying and understanding the genome’s functional elements.

Les Lang

Neural lesson in migration

In laying down the neural circuitry of the developing brain, billions of neurons must first migrate to their correct destinations and then form complex synaptic connections with their new neighbors.

When the process goes awry, neurodevelopmental disorders such as mental retardation, dyslexia or autism may result.

Scientists led by Franck Polleux, Ph.D., associate professor of pharmacology at the University of North Carolina at Chapel Hill School of Medicine have now discovered that establishing the neural wiring necessary to function normally depends on the ability of neurons to make finger-like projections of their membrane called filopodia.

The finding, published as the cover story of the Sept. 4 issue of the journal Cell, indicates that the current notion regarding how cells change shape, migrate or differentiate needs to be revisited.

According to Polleaux, scientists have thought that the only way for a cell to morph and move is through the action of the cytoskeleton or the scaffold inside the cell, pushing membrane forward or sucking it in.

But Polleux’s study shows that the brain protein srGAP2 can also impose cell shape by directly bending membranes, forming filopodia as a mean to control the migration and branching of neurons during brain development.

Les Lang   (Video produced by Chris Carmichael)

Discovery, discovery, discovery

“Location. Location. Location,” is a mantra in real estate circles, but when it comes to molecular science nowadays, the phrase appears to be “Discovery. Discovery. Discovery.”

This takes on particular importance when the view from the lab bench after the last puff of smoke clears reveals what could be potential applications for human health and disease, especially for future interventions in cancer.

A good example are the recent findings from the lab of  Victoria Bautch, Ph.D., professor of biology at the University of North Carolina at Chapel Hill.

In a paper published recently in the journal Developmental Cell (the cover story for that issue), Bautch and co-authors reported having identified a new biological process that spurs the growth of new blood vessels.

Up until now, scientists thought that the molecular signals to form new blood vessel sprouts came from outside the vessel. But new research from UNC has shown that signals can also come from within the blood vessel, pushing new blood vessel sprouts outward.

In experiments using mouse embryonic stem cells and mouse retinas, the researchers found that defects in a protein called Flt-1 lead to abnormal sprouts and poor vessel networks. Other research recently showed that levels of Flt-1 protein are particularly low in the dilated and leaky blood vessels that supply tumors with oxygen.

“The blood vessels themselves seem to participate in the process guiding the formation of the vascular network,” said  Bautch. “They do not just passively sit there getting acted upon by signals coming from the outside in. Rather, they produce internal cues that interact with external cues to grow.”

The findings could give important insights into the formation of the vasculature needed to feed new tumors.

Bautch, who is also a member of the Program in Molecular Biology and Biotechnology, the UNC McAllister Heart Institute and UNC Lineberger Comprehensive Cancer Center, notes that the more scientists understand about the sophistication and complexity of the mechanisms guiding the formation of blood vessel sprouts, the better equipped they will be to develop therapeutic interventions to produce or to halt new blood vessels.

Les Lang. (Video produced by Courtney Potter)

CF gene therapy via common cold virus?

Cystic fibrosis is a nasty inherited chronic illness.  It affects the lungs and digestive system of about 30,000 children and adults in the United States and an estimated 70,000 people worldwide.

CF is the most common lethal genetic disease in the Caucasian population, affecting one in 3300 births. Other ethnic populations are affected less frequently, ranging from one in 10,000 – 15,000 births in Hispanic and African-American populations, to one in 30,000 Asian births.

The underlying cause is a mutation in the ion channel gene CFTR (cystic fibrosis transmembrane conductance regulator).  This gene is responsible for controlling salt and water transport across the cells lining the lung, pancreas, and other organs.

When this gene is abnormal, secretions in these organs become dehydrated and sticky, and eventually clog airways and may block other organs (pancreas, intestines, male reproductive tract, bile ducts) as well.

Not too many decades ago, few children lived to attend elementary school. Today, with advances in research and treatment,  the median age of survival for children diagnosed with CF  is more than 37  years. Many with the disease can now expect to live beyond their 30s and 40s.

Scientists have worked for 20 years to perfect gene therapy for the treatment of cystic fibrosis.  In recent research, UNC scientists have found what may be the most efficient way to deliver a corrected gene to lung cells collected from cystic fibrosis patients. They also showed that it may take this high level of efficiency for cystic fibrosis (CF) patients to see any benefit from gene therapy.

Using parainfluenza virus, one of the viruses that causes common colds, the UNC scientists found that delivery of a corrected version of the CFTR  gene to 25 percent of cells grown in a tissue culture model that resembles the lining of the human airways was sufficient to restore normal function back to the tissue.

“This is the first demonstration in which we’ve been able to execute delivery in an efficient manner,” says Ray Pickles, Ph.D., associate professor of microbiology and immunology at the UNC Cystic Fibrosis Research and Treatment Center. “When you consider that in past gene therapy studies, the targeting efficiency has been somewhere around 0.1 percent of cells, you can see this is a giant leap forward.”

“We discovered that if you take a virus that has evolved to infect the human airways, and you engineer a normal CFTR gene into it, you can use this virus to correct all of the hallmark CF features in the model system that we used,” Pickles said. For instance, the experiment improved the cells’ ability to hydrate and transport mucus secretions.

The next step is to work to ensure the safety of the delivery system. “We haven’t generated a vector that we can go out and give to patients now,” Pickles says, “but these studies continue to convince us that a gene replacement therapy for CF patients will some day be available in the future.”

Les Lang

Brain blood vessels love exercise

Here’s some new evidence that aerobic activity may keep the brain young.

A UNC School of Medicine study slated to appear July 9 in the American Journal of Neuroradiology, suggests that physically active elderly people have healthier cerebral blood vessels.

This is the first study to compare brain images of elderly subjects who exercise with those who don’t.

Researchers led by Elizabeth Bullitt, M.D., Van L. Weatherspoon Distinguished Professor of neurosurgery, used non-invasive magnetic resonance (MR) angiography to examine the number and shape of blood vessels in the brains of physically active elderly people, 7 men and 7 women, ages 60 to 80.

The study subjects were equally divided into 2 groups. The high activity group reported participating in an aerobic activity for a minimum of 180 minutes per week for the past 10 consecutive years, and the low activity group told investigators they had no history of regular exercise and currently spent less than 90 minutes a week in any physical activity. (The researchers did not know into which group participants were placed.)

Aerobically active subjects exhibited more small-diameter vessels with less tortuosity, or twisting, than the less active group, exhibiting a vessel pattern similar to younger adults.

Cerebral blood vessels, active elderly group

The authors, who were sponsored in part by the National Institutes of Health’s National Institute of Biomedical Imaging and Bioengineering, identified significant differences in the left and right middle cerebral artery regions confirmed by more than one statistical analysis.

The brain’s blood vessels naturally narrow and become more tortuous with advancing age, but the study showed the cerebrovascular patterns of active patients appeared “younger” than those of relatively inactive subjects. The brains of these less active patients had increased tortuosity produced by vessel elongation and wider expansion curves.

Cerebral blood vessels, inactive group

Bullitt says the pilot study lays the foundation for future research to determine whether aerobic activity improves anatomy, if older patients with “younger” brains are more likely to engage in physical activity, and whether elderly adults who begin a program of aerobic activity can reverse the cerebrovascular, anatomic and functional changes associated with advancing age.

Portions of this paper were presented in December 2008 at the Radiological Society of North America annual meeting in Chicago.

Les Lang

Xist not essential for X-inactivation

Science writer Marla Vacek Broadfoot reports this story for us:

Because females carry two copies of the X chromosome to males’ one X and one Y.  they harbor a potentially toxic double dose of the over 1000 genes that reside on the X chromosome.

To compensate for this imbalance, mammals such as mice and humans shut down one entire X-chromosome through a phenomenon known as X-inactivation. For almost two decades, researchers have believed that one particular gene, called Xist, provides the molecular trigger of X-inactivation.

Now, a new UNC study appearing online Wednesday July 1^st in the journal Nature disputes the current dogma by showing that this process can occur even in the absence of this gene. 

“Our study contradicts what is written in the textbooks,” said senior study author Terry Magnuson, Ph.D., Sarah Graham Kenan Professor and chair of genetics, director of the Carolina Center for Genome Sciences and a member of the UNC Lineberger Comprehensive Cancer Center. “Everybody thought that Xist triggers X-inactivation, but now we have to rethink how this important process starts.”

Previous studies showed that the Xist gene was active or “turned on” early in the course of X-inactivation and that disruptions in the gene resulted in irregular X-inactivation, eventually leading to the accepted assumption that Xist was the trigger. But it wasn’t clear in the literature if this genetic phenomenon would initiate if Xist isn’t present, said lead study author Sundeep Kalantry, Ph.D., postdoctoral fellow in the UNC department of genetics.

Kalantry used three different molecular techniques to look at X-inactivation in the embryos of mice that were genetically engineered to contain a defective Xist gene on their future inactive X-chromosome. He discovered that the genes on this X-chromosome could be silenced regardless of whether they produced Xist. But while Xist was not absolutely required to start X-inactivation, without it genes along the X-chromosome eventually became active again. Thus, Xist appears to stabilize silencing of the X-chromosome over the long term.

Unlike most genes, the Xist gene doesn’t code for a protein. Rather, it acts at the level of RNA – a copy of the DNA genetic sequence – which serves to recruit protein complexes through a process known as epigenetics. These proteins then form a molecular scaffold along the inactive-X chromosome that can stably silence the genes contained within it. The UNC researchers are now actively investigating how this chromosomal remodeling begins in the first place.

“If we can figure out the mechanism that triggers X-inactivation, we can potentially apply this knowledge to diseases that have an epigenetic component,” Kalantry said. “So it can have implications not only in fundamentally understanding X-inactivation but also to gain insight into the increasing array of illnesses where the epigenetic machinery has gone awry – such as in prostate and breast cancers.”

Study outlines schizophrenia puzzle

Scientists have long recognized that schizophrenia can run in families and  that it also has a genetic component.  But one might find any number of studies that point to other explanations for the disease, including dysfunctional communication patterns within families,  child abuse, drug abuse, or, as in the 1960s,  the “anti-psychiatry” view of Thomas Szasz and others that saw schizophrenia as a “social construct. “

However, only recently has science begun to pinpoint the exact spots in our genetic material that contribute to the illness.  Last year, the International Schizophrenia Consortium found that rare chromosomal structural variants elevate the risk of developing schizophrenia.

Now, a multi-national group of investigators, including a scientist at the University of North Carolina at Chapel Hill, has discovered that nearly a third of the genetic basis of schizophrenia may be attributed to the cumulative actions of thousands of common genetic variants.

The effects of each of these genetic changes, innocuous on its own, add up to a significant risk for developing both schizophrenia and bipolar disorder.

The finding, published online July 1, 2009, in the journal Nature, suggests that schizophrenia is much more complex than previously thought, and can arise not only from both rare genetic variants but also from a significant number of common ones.

“This is an enormous first for our field,” said co-author Patrick Sullivan, M.D., Ray M. Hayworth and Family Distinguished Professor of Psychiatry in the department of genetics at the UNC School of Medicine. “You could say that we now have the outline of the puzzle, and we just need to take all of these pieces that we have identified and see how they fit them together.”

In this study, Sullivan and other investigators in the Consortium used “genechip” technology to identify 30,000 genetic variants (single nucleotide polymorphisms or “SNPs”) that were more common in 3,000 individuals with schizophrenia than in 3,000 comparison subjects without schizophrenia.

This pattern was found in three separate samples of individuals with schizophrenia and two samples with bipolar disorder – indicating a previously unrecognized overlap between the two diseases. These risk variants were not present in patients with other non-psychiatric diseases, such as hypertension or diabetes.

The researchers are also investigating how genes and environment interact to cause the disease. One additional finding of their study was the identification of the human leukocyte antigen (HLA) locus as a possible risk factor. Because this region plays an important role in immune response to infection, it could suggest that exposure to an infectious agent increases risk of developing psychiatric disease.