MicroRNAs: thinking globally, acting locally

Neurons are consummate multi-taskers. They respond to and memorise a barrage of signals by making subtle changes at individual synapses and basically learn from experience.

This ability, which enables the nervous system to learn and to retain memories, is referred to as neuronal or synaptic plasticity.

The changes may be macroscopic or occur at the molecular and physiological level. Some forms of plasticity at the level of an individual synapse rely on the neuronal cellular machinery making protein available quickly and in the right amounts.

According to Professor Kenneth Kosik of the University of California Santa Barbara, who visited Australia this year to address the 7th Discovery Science and Biotechnology meeting, this requirement seems particularly well suited to a microRNA-regulated system of local regulation.

MicroRNAs (miRNAs), discovered just over a decade ago, revealed a completely new and unsuspected level of gene regulation.

These non-coding RNA molecules regulate gene expression post-transcriptionally via RNA silencing. Similar to the inhibitory RNA (RNAi) system, the 21-nucleotide miRNAs bind to complementary regions in target mRNAs to regulate translation via mRNA degradation or translational repression.

First identified in C. elegans, we now know that miRNAs are expressed widely in metazoan cells, with the potential to modulate around one-third of coding mRNA expression in mammals.

In the nervous system, miRNAs have key roles in development, and possibly in mature neurons as mediators of plasticity. Around 500 human miRNA sequences are known, although their mRNA targets, and therefore, their functional roles in specific cells remain largely a mystery.

Historical perspective

Kosik has been interested for a number of years in neuronal plasticity and its impairment in neurodegeneration. More specifically, he wants to know how translation is regulated at neuronal dendrites, which of course can be a long way from 'central control' in the nucleus.

It is well established that neuronal synapses are subject to individual, very local control, and that this local mechanism involves translational regulation.

Although not involved at the discovery end of the miRNA story, Kosik was quick to realise how these groundbreaking discoveries were extremely relevant to the phenomena of neurons and plasticity.

"It struck me as an elegant way that nature would do the kinds of things that we were interested in," he says.

At that time, there was very little known about miRNAs in the nervous system. So, before deciding on the miRNA species to study in their specific research context, Kosik's group needed to ascertain how many and which miRNAs are present in neurons.

Kosik says he was lucky in this endeavour - a very talented postdoc in his lab at the time, Anna Krivchevsky, devised a method to profile miRNAs. This was one of the first techniques established for profiling many miRNAs in a single experiment, he says.

The method involved making the complementary sequences of known miRNAs into trimers using their antisense sequences. She then put these onto a nylon filter together with RNA purified from brain or brain cultures.

"This technique turned out to be a goldmine for us - particularly for getting started in looking at miRNAs in neurons. There are a lot more advanced ways now to attack the same problems, but her method gave us and other groups an early look at the particular miRNAs that might be important for say neuronal plasticity."

Soon after, Kosik's group established that miRNAs are turned on during synaptogenesis and that some are found in neuronal compartments called polysomes - these are the actively translating pools of mRNAs.

"From there, we started profiling everything we could get our hands on - neurons of course, but also tumours and stem cells." Kosik's group soon identified an miRNA called mir21 that was extremely elevated in glioblastomas and also seemed to work in apoptosis suppression pathways, one of the critical ways in which cancer progresses.

It turned out that mir21 is markedly upregulated in many tumour types, and is now recognised as perhaps the most commonly disregulated miRNA in cancer. Work on this project remains ongoing in Kosik's lab in addition to work in the nervous system, particularly on the role of mir21 in apoptosis more generally.

Although some neurological diseases have now been linked to miRNAs, most interest is still in the field of cancer biology. The term oncomirs was even coined to describe cancer-related miRNAs like mir21.

---PB---

Firing by remote control

Kosik has continued to profile neuronal miRNAs to identify those potentially involved in neuronal plasticity, the original motivation for this work.

"One of the current interests in the field is to determine and understand the functions that are unique to dendrites," he says. "The ongoing issue of mRNA distribution between dendrites and the cell body also applies to the miRNAs."

To address this, Kosik's group combined their miRNA profiling with a laser-capture technique to look at specific cells or parts of cells. "We were able to quantify miRNAs in the dendrite and cell body compartments separately by laser capturing just the dendrites or just the cell bodies and profiling the contained set of microRNAs."

Most neuronal miRNAs were distributed throughout the neuron. However, as is the case with mRNAs, a subset of miRNAs were either relatively enriched or depleted out in the dendrites.

"We then had a repertoire of miRNAs that are probably involved in regulating protein translation locally, due to their specific and restricted location near the synapse. We also had supremely good candidates for targeting miRNAs involved in plasticity."

Home delivery system

For those involved in RNA research there is a great deal of interest in not just what miRNAs are doing in dendrites but how they get there, he says.

A number of years ago, his group reported structures in neurons they called RNA granules. Kosik describes this granule as like a translocation vehicle, carrying RNA molecules to their required destinations nearer the synapse when needed to regulate local translation.

In further studies, Kosik's team found these RNA granules to be devoid of transfer RNA and other elements of the translation system needed to turn mRNA into proteins. It seemed then that this intracellular compartment is not merely a delivery system, but also acts as a holding unit for mRNA.

On stimulation, the contents would presumably be released for an efficient synaptic response (sort of like home delivery instead of take-away).

Recent evidence suggests another type of intracellular compartment that holds RNA called the P body or processing body.

"We are really excited right now to know how the RNA granules, which contain certain categories of mRNAs, are getting to their destinations, and then how these P bodies, which are perhaps involved in actual storage of mRNAs or maybe even their degradation, are talking to the granules. How are these different subcellular units dynamically related?"

And now of course, the same issue applies to miRNAs. "The real challenge for neurons is to get everything they need to the right place when they need it, including miRNAs."

One possibility is that the processed mature miRNA hitches a ride on a target mRNA containing dendritic localisation signals. "Alternatively, dendritic miRNAs might be processed locally from precursor molecules, in which case the cell must also move the miRNA biogenesis machinery out there. This is going to be a really exciting direction in the field."

Kosik's research now covers many topics, including the more recent and exciting work on miRNAs in the nervous system. "I like to think my research encompasses a common theme - that we are really talking about problems of synaptic plasticity with all the various issues I research.

"With Alzheimer's disease it is the loss of plasticity and how you can help people who are losing their synaptic efficacy, and miRNAs will definitely be important there. In fact, recent work has shown particular miRNAs targeting the amyloid precursor protein, so they may soon become directly relevant in that field also.

"So, my ideal for the lab would be to work towards making all these connections even stronger and eventually link up, but if we don't and all we do is get insights into synaptic plasticity I would be very happy."

---PB---

Alzheimer's expert

Ken Kosik is renowned internationally for almost two decades of research into the biology of Alzheimer's disease.

After completing a medical degree, Kosik did his neurology residency at Tufts New England Medical Centre in 1979, before embarking on a purely research career at the Harvard Medical School, becoming professor of neurology and neuroscience in 1996.

He joined a group there trying to understand the pathobiology of Alzheimer's disease, and exactly what it is that slowly tangles components of the brain to gradually but inevitably dement the person's thoughts and personality.

Kosik and colleagues discovered that the neurofibrillary tangles that characterise Alzheimer's are composed of a protein called tau and that loss of this protein function inhibits neuronal plasticity. Subsequent studies of the tau protein and its pathological fate as a neurofibrillary tangle have been a pillar of Alzheimer's disease research.

In parallel to this work, Kosik is very interested in the underlying cellular mechanisms by which plasticity is lost in the course of neurodegeneration. In 2004, Kosik moved to the University of California Santa Barbara to co-direct the Neuroscience Research Institute and take up a chair in neuroscience research.

Interestingly, Ken Kosik began his university education as a literature major, but then decided that studying literature was not the path to his dream of becoming a writer. Over 25 years later, Kosik has certainly made an indelible mark not only on the scientific 'literature', but ultimately, through his work, on the human mind.

In his spare time, Kosik has embarked on setting up a centre for patients with Alzheimer's disease, the Centre for Innovative Therapy (cFIT). "I describe it like research is my day job, that is what I get paid for, but I continue to have a long-standing interest in clinical problems, and particularly Alzheimer's-type conditions," he says.

"So, I spend probably five to 10 per cent of my time and effort working with other people in my community in Santa Barbara in developing a new approach to clinical aspects of AD.

"I have always been a little troubled by the fact that we were not doing a very good job for people with this disease. And that problem was not just the absence of a cure, but the fact that people were very frustrated with the whole system they had to deal with - a very sophisticated and complicated medical system, big hospitals, and tertiary care centres, which are set up for people that require cardiac transplants or advanced surgery and high-level clinical cases, but not for patients with diseases like Alzheimer's.

"The problem does not require surgical or significant drug therapy, it is more about cognition, and so is not a problem from which hospitals can ever expect to make money, and precious hospital real estate is not readily handed over to physicians interested in that problem.

"It occurred to me that we needed to approach the problem in a novel way, to move the issue of AD and cognitive impairment out of the hospital and into a setting that was more friendly - sort of like a living room, but that had all the medical expertise needed for that specific disease.

"It also has to be a place where we are comfortable with genetic risk factors, the latest research - whether you want to take supplements or be engaged in clinical trials going on somewhere - so we want to reshuffle the deck a bit and 'de-medicalise' the problem, and move it into an information-rich rather than a procedure-oriented environment.

"This place does not exist yet anywhere -- a place where you can take a problem that doesn't fit into the traditional medical model of treatment, such as AD, and in a single location, provide comprehensive services; a place that combines both sophisticated information for patients with the level of medical and other care that is really needed."