Tag Archives: learning and memory

Psychiatric illness ‘explained’: Disorders of CNS Connectivity

Posted on by

The power of the nervous system:

network-of-cortical-neurons

The astonishing power of the nervous system does not reside in a single neuron. (That said, an advanced supercomputer is required for the task of modelling the processing power of even a single neuron).

Nervous tissue is immensely powerful because of the rich connectivity between neurons. A 1mm voxel of cerebral cortex (a standard fMRI unit), contains ~300 million synaptic connections and ~50 thousand neurons [ref].  Scaled up to the whole human brain, there are estimated to be several hundred trillion synaptic connections within a total pool of ~100 billion neurons. Neuronal networks are the foundation of, perception, movement, thinking, memory and the personality.

Network learning

A crucial property of neuronal networks is that they learn from experience. Experience may stem from the external world (sensation) or the inner world. Learning is achieved by adjusting the strength of the connections between neurons. New connections can form, and weak connections wither away – essentially a process of re-wiring. Taking up a musical instrument or a new language, for example, constitutes a major re-wiring exercise, although higher, more mysterious faculties – such as selfhood, agency and individual identity – are already wired-up in infancy, and remain a foundation throughout life, except if threatened by the most severe psychiatric disorders.

Alzheimer’s disease is the prototypical example of a network illness. Progressive       shrivelling of the network mirrors the decline of the faculties, from initial problems with memory right up to the disintegration of selfhood.

Network health

Network health is vital for mental health. The stabilisation of essential connections, the formation of new connections and the controlled elimination of redundant connections involves many components.

  • There are components which span the gap between nerve terminals and dendritic spines to ensure that connections remain tightly bound [link].
  • There are signalling pathways which control the dynamic, flexible actin scaffold which give terminals and spines their anatomical structure.
  • There is, ready-to-hand, protein-synthesis machinery for making additional spines as learning proceeds.
  • Finally, and most recently explored, there are mechanisms for ‘clearing up’ the debris when connections are no longer required. Such components (microglia, complement proteins) are much more familiar in their role as immune cells and immune signals, but their role extends beyond inflammation. Microglia and complement are now recognised as key components in the wiring of the brain as it learns and develops.

Major psychiatric illness

dendritic spine

Where those components involved in the function and structure of synaptic connections are defective, psychiatric illness can result. Mutations in the components which bind the nerve terminal and dendritic spine are a cause of autism. The cause of many learning disability cases, hitherto unknown, are mutations in proteins which control the actin scaffold. The psychiatric manifestations of Fragile X syndrome (intellectual deficits / autistic features / hyperactivity) result from abnormal protein synthesis in dendritic spines and subsequent abnormal local wiring.

dendritic-spine

Microglia & complement proteins

pink-eatme-cake-topperThe latest components to receive attention, as pertains to psychiatric illness are the microglia and their signalling pathways, specifically complement proteins.

Complement proteins function as a tag, essentially an ‘eat-me’ signal, on synapses destined for elimination. The tag is recognised by the phagocytic microglia which engulf and clear the redundant synaptic elements [link].

Although the role of immune components in psychiatric illness has become a hot topic, many researchers are still accustomed to regard microglia and complement in the context of inflammation rather than CNS re-wiring. Both major depression and schizophrenia, have been linked with abnormal immune components, but neither disorder is inflammatory in the same sense as encephalitis or meningitis. The main histological finding in schizophrenia is decreased connectivity between neurons, not inflamed nervous tissue. Similarly, an anatomical correlate of depression is impoverished connectivity in the hippocampus, not inflammation.

A major development in Alzheimer’s research has been the recognition of up-regulated complement proteins and microglial phagocytosis commensurate with the loss of neuronal connections. The crucial observation is that such changes occur prior to amyloid deposition and tangle formation [link]. Alzheimer’s appears to be a disorder of runaway synaptic loss. Drug discovery efforts are aimed at blocking complement protein receptors to protect synapses [link].

Schizophrenia has been associated with changes in the genes coding for a specific complement protein (C4A). Knockout of the C4A gene in an animal model causes a marked alteration in the pruning of synaptic connections in later life [link]. Schizophrenia, albeit to a far less extent than Alzheimer’s, is regarded as a disorder of impoverished connectivity, (whereas Autism is associated with increased dendritic spines and increased connectivity) [link].

Hold on –  what about the ‘dominant’ wet-ware hypotheses?

hoovers

An older generation of psychiatric researchers may ask where dopamine [link]] and perhaps glutamate [link] fit into a model of psychiatric illness in which abnormal connectivity between neurons appears to carry robust explanatory power. Earlier models posited that an excess or deficiency of neurotransmitter or receptors lay at the root of major depression and schizophrenia. Such models stemmed from the relatively primitive knowledge of the synapse available at the time (circa 1965-1975). Then, the hot topics in neuroscience were; the nature of neurotransmitter release (Sir Bernard Katz, UCL) and the ‘visualisation’ of receptors (Solomon Snyder, John Hopkins).

The answer (to the question of how glutamate and dopamine are accommodated) is fairly straightforward: Glutamate (finally admitted to the neurotransmitter club circa 1983-87) is the fast neurotransmitter between nerve terminals and dendritic spines, throughout nervous tissue. Dopamine determines the strength of the connection between the glutamate terminal and the dendritic spine within specific CNS structures. Dopamine functions as a teaching signal; adjusting connectivity and promoting learning in higher centres.

Frontier psychiatry

hippocampus

The obvious strategy of searching for molecules which can impact on connectivity is well underway.

That said, existing psychiatric treatments, such as antidepressants, lithium and dopamine antipsychotics have an impact upon connectivity to the extent that structural changes can already be detected, albeit in a population of patients rather than the individual, with routine MRI scans. Drugs impact upon plasticity: Drugs impact upon CNS structure.

A more basic question goes back to the very roots of modern psychiatry. The question is whether, for some, the neuronal networks are destined to be unwell from the outset (endogenous psychiatric illness), or if, for others, adverse experiences during development cause the network to wire-up pathologically (exogenous psychiatric illness). Then again, there is the third position, in which the choreography between the neuronal hardware and the external environment determines who will succumb to psychiatric syndromes. Whatever the proximal cause(s), endogenous or exogenous, major psychiatric illness appears to stem from abnormal connectivity within neuronal networks.

Why NMDA drugs keep failing in schizophrenia.

Posted on by
nmda receptor

The NMDA receptor. Glutamate and glycine are required for NMDA receptor activation. Activation involves the opening of a channel allowing calcium and sodium ions to flow into the neuron. Recent attempts to translate NMDA pharmacology into the clinic have focussed on the glycine site.

Twenty years ago it all looked so promising. The model was as follows: Learning and memory were clearly being driven by activity at the glutamate NMDA receptor. Boost the NMDA receptor by pharmacological means, and perhaps intellectual performance could be improved above baseline. The hope was that an NMDA enhancer might work in schizophrenia, which many had come to regard as a disorder of cognition. Yet the story has not played out as anticipated. The latest generation of NMDA enhancers, like their predecessors, has failed in schizophrenia [link]. And it is looking increasingly likely that the basic model [boost NMDA -> boost intellectual functioning] was overtly simplistic.

long term potentiation

Long Term Potentiation (LTP) is induced by NMDA receptor activation. The mechanism of early-phase LTP involves the enhancement of AMPA receptor conductances and insertion of new AMPA receptors into the post-synaptic membrane.

An recent review article by Collingridge and colleagues is worthy of study. Back in 1983, Collingridge had shown that activation of the glutamate NMDA receptor was the initial catalyst for the process of LTP (long-term-potentiation). At that time glutamate was only just gaining entry to the neurotransmitter club, whereas LTP [a process in which excitatory synapses become and remain stronger] had achieved fame ten years earlier as a likely substrate for learning and memory in nervous systems.

The discovery of NMDA-dependent LTP, as the phenomena came to be known, was the stimulus for an enormous, worldwide research effort into glutamate neurobiology. Since then, our knowledge of NMDA receptors has advanced, to the point where the complexity can be overwhelming [figure below]. But the medicines have not materialised. The biology appears to be several orders more complex than the model. Is that why the drugs have failed? In any case, the model [boost NMDA -> boost intellectual functioning] can now be safely abandoned with little risk of missing a major therapeutic breakthrough.

Intracellular modulation of NMDA receptors

Sites of intracellular modulation of NMDARs. Schematic representation of the distribution of selected posttranslational regulatory sites on the intracellular C-terminal domains of NMDAR subunits. Properties such as channel gating, receptor desensitisation and receptor shuttling are modulated by phosphorylation at key residues. Collingridge et al 2013

POSTSCRIPT

Recently the NIMH (National Institute of Mental Health], the main funder of mental health research in the world, announced that they would no longer support clinical trials of new drugs unless there was a clear mechanistic advance at the same time:

“a positive result will require not only that an intervention ameliorated a symptom, but that it had a demonstrable effect on a target, such as a neural pathway implicated in the disorder or a key cognitive operation.”

The NMDA receptor story calls the logic of this approach into question. That story is the archetypal case in which a mechanism was clearly defined, and well supported after decades of preclinical research. Indeed the mechanism [the model] had become so appealing that many were reluctant to abandon it, even as it was becoming obvious that the therapeutics were not going to work. An overhaul of drug discovery in psychiatry is needed, but it will require to be more realistic than solving mechanism and efficacy problems concurrently. Pulling back the bureaucracy, the inflated costs and the micromanagement could be a more fruitful intervention.

Modafinil to boost academic performance: Effective, Addictive, Cheating?

Posted on by

Originally marketed as a wake-promoting agent, modafinil is a prescription drug that is said to boost cognition in healthy subjects. As such it’s use has spread amongst college students cramming for dreaded examinations. Anecdotal reports are of enhanced focus, clarity of thought and intellectual stamina; attractive properties for those hoping to secure a competitive edge for themselves.

But how do the pro-cognitive effects of modafinil stack up in proper scientific studies? Is modafinil addictive? And what ethical stance should be taken on the use of performance-enhancing agents in academic life?

Does modafinil enhance cognitive performance?

The first laboratory-based study of modafinil (single dose 100 or 200mg) in 2003 showed that it had clear pro-cognitive properties. Since then a further six studies have been in agreement, with performance enhancements in the domains of working memory, cognitive flexibility and planning.

A recent and elegant study carried out in Cambridge involving 64 healthy participants between the ages of 19-36 is illustrative [Muller et al 2012]. Participants were randomly allocated to receive modafinil (200mg) or placebo under experimental conditions, two hours ahead of a cognitive challenge. In addition to the usual measures of memory performance, task enjoyment was rated.

Performance in planning/problem solving under modafinil v placebo

The modafinil group achieved success with fewer choices in a task requiring cognitive planning. Performance enhancement was most apparent at the highest level of difficulty. Error bars are SEM.
From: Muller et al Neuropharmacology 64 (2013) 490-495.

The main findings were that the modafinil group out-performed the placebo group on tests of working memory, planning and pattern recognition memory. These improvements were more prominent as the cognitive tasks became more difficult.

And for the first time, it was shown that modafinil boosted enjoyment during the testing.

The authors postulated that the enjoyment could have arisen from the sense of satisfaction at task mastery or instead be the result of heightened motivation as a direct effect of the drug – surely now a topic for further study.

Is modafinil addictive?

The behavioural pharmacology of modafinil appears to stem from inhibition of the dopamine re-uptake transporter (DAT), akin to the mechanism of the classic [and addictive] stimulants, cocaine and amphetamine. However modafinil is a relatively weak inhibitor of DAT.

raclopride PET following modafinil

PET images of the human brain showing that compared to placebo, modafinil reduces raclopride binding in the striatum. The reduction in raclopride binding is indicative of dopamine release. Volkow et al (2009) JAMA 2009 301:1148-54

There are a number of behavioural differences between modafinil and the classical stimulants. Perhaps most notably, modafinil has a very low propensity for abuse (Wisor 2013). Indeed there was some hope that modafinil might actually constitute a treatment for cocaine/amphetamine addiction, but the findings to date in clinical trials have been disappointing.

Does the use of modafinil for exam revision constitute cheating?

Modafinil certainly confers a cognitive advantage, at least in the short term. And the downside in terms of addiction appears to be negligible, despite the pharmacological similarities of modafinil to ‘hard drugs’ such as cocaine and amphetamine.

The differences in cognitve performance under modafinil may be slight, and only apparent as the demands of the task increase. But isn’t this similar to the highest levels of sport, in which performance enhancing substances confer a critical edge as the competition reaches a climax.

The ethics of ‘smart drugs’ is complex [unlike the pharmacological questions above, which in contrast, can be settled by experiment, as well as reason]. One could argue that personal choice is all that matters. Surely the individual student should make their own judgement on whether to use, or abstain from, cognitive enhancers?  But is it only a personal matter? A decision to use smart drugs has a potential impact on the competition, the rest of the field. Is the use of modafinil, and the like, nothing other than cheating?

Psychosis Research. Where have we been & where are we going?

Posted on by
 
phenotype and genotype

The Institute of Psychiatry at The Maudsley is the largest centre for psychiatric research in Europe. Recently a group of leading researchers were tasked with summarising an area of research as it pertains to psychosis and psychopharmacology.

The outcome was a series of short lectures, delivered to a lively audience of psychiatrists, mental health workers and psychologists at The Maudsley. The lecture slides and audio are now available below and constitute a unique training resource for those who treat patients.

1. Sir Robin Murray,
“Psychosis research: Deconstructing the dogma“
2. David Taylor,
“Current Psychopharmacology: Facts & Fiction“
3. Oliver Howes,
“How can we Treat psychosis better?“
4. Marta DiForti,
“An idiot's guide to psychiatric genetics“
5. Sameer Jauhar,
“Ten psychosis papers to read before you die!”
6. Paul Morrison,
“Future antipsychotics“

 

Neurophysiology can free psychiatry from it’s dependence on metaphor.

Posted on by
el Greco

For psychiatry to progress, it can take as it's starting point the most up to date thinking on how the nervous system operates. This necessitates an appreciation of how neurons communicate with each other, how circuits emerge and how CNS tissue is sculpted in the very act of processing information. A short synopsis of some of the main themes in contemporary neurophysiology is presented here. First we shall consider the two main theories of how information is processed in the here-and-now. Then we shall look briefly at spike-timing dependent plasticity, the latest and arguably the most elegant form of plasticity within the brain, which synthesises many strands.

Information Processing

Special gnostic cells

There are two major theoretical accounts of how neural tissue “performs its computations”. The first account postulates the existence of ‘special cells’ at the top of a processing hierarchy. These cells are less ‘concerned’ by the raw ‘building blocks’ of sensory experience – orientation, brightness, colour, pitch etc. Instead, they respond (‘fire’) to whole objects (Gestalts), regardless of perspective, illumination and all the other idiosyncrasies that make up a perceptual scene. The metaphor of the ‘grandmother cell’ captures the idea. “Each time my grandmother comes into consciousness, via any of the sensory channels or in imagination, a ‘special’ cell, somewhere in the brain, is “active”.

The main criticism of the ‘grandmother cell’ hypothesis [aside from its prioritising of perception over thought & movement] is that there are far more potential percepts, than available neurons. Another criticism is that by focusing exclusively on feed-forward pathways, the hypothesis ignores the anatomical 'reality’ of extensive feedback pathways. Nevertheless, in-vivo electrophysiological work in humans undergoing neurosurgical procedures has provided evidence that there are neurons in the medial temporal lobe, which have the characteristics of grandmother cells.

Dynamic Assemblies

The second account prioritizes flexible, dynamic assemblies of neurons over ‘special’ cells. An assembly is defined as a constellation of neurons, which are firing action-potentials within the same narrow time-window (synchronously). Here, processing is a more ‘democratic affair’, and no special cells are required. Feedback and feed-forward connections are equally important, as the network (the assembly) reaches a consensus. Assemblies are transient entities, emerging for a period before ‘dissolving’, perhaps to ‘reappear’ at a later instant. A temporarily ‘dominant assembly' may ‘recruit’ other ‘partners’. Allegiances are flexible, with co-operation at one instant and competition at another. And over longer periods of time, assemblies can become – stronger; by virtue of sheer repetition and the ‘rules’ of long-term-potentiation (LTP), particularly if monoamine systems are co-active – or weaker; if the ‘content’ is fleeting or insignificant. Network oscillations (rhythms) provide a metronome, to ensure that the right cells fire in synchrony. Gamma (30–200 Hz) rhythms ‘bind’ local assemblies, whereas lower frequencies (theta, alpha, and beta) sub-serve long-distance communication between brain areas.

Of course, it is entirely feasible that the CNS makes use of both schemes described above [special cells & dynamic assemblies]. Processing power may reach grand heights when special [gnostic] cells come together as an assembly.

Sculpting CNS tissue

Spike-timing-dependent plasticity (STDP) depends on the conjunction of pre and post-synaptic events, within a narrow time envelope, of the order of tens of milliseconds or so. In the most straightforward version, a synapse is strengthened if a pre-synaptic input occurs immediately prior to a post-synaptic action potential (AP). If on the other hand, the input arrives in the immediate aftermath of a post-synaptic AP, the synapse is weakened. Pre and post-synaptic events beyond the critical time-window (i.e. unpaired ‘events’) leave synaptic strength unchanged. This shows how the precise timing of neuronal firing impacts upon the network. [And this impact is structural, as well as biochemical, Link]. Two aspects of STDP are notable:

1. Conventional neuromodulators appear to ‘tweak’ STDP. Actually ‘tweak’ is an understatement. The presence of a modulator such as dopamine can transform a normal pre-> post strengthening into a depression instead. More succinctly, dopamine can determine the direction of plasticity (+ or -).

2. The critical time window of STDP (tens of milliseconds) is in exactly the same ‘ballpark’ as network oscillations in the gamma band (period ~25ms).

The elegance of STDP is that it begins to reveal how apparently unconnected phenomena [brain-oscillations and neuromodulator systems], are integrated within a fundamental CNS function – how synapses and circuits are sculpted over time.