RICERCA ITALIANA     

Plasticity of cortical networks in normal human and in neurological patients: The influence on local sleep processes

Università degli Studi di Roma "La Sapienza"

Abstract

A recently formulated hypothesis, the Synaptic Homeostasis Hypothesis links sleep with synaptic homeostasis, argueing that the specific function of sleep is to downscale the weight of cortical synapses (Tononi and Cirelli, 2006). Among others, the hypothesis states that during wakefulness, learning and plasticity processes, such as long-term potentiation (LTP) bring to a net increment of synaptic weight in several cortical circuits. Globally, the strength of intracortical connections reaches a maximum toward the end of the day.
In fact, several lines of evidence support the notion that plastic changes occurring during wakefulness in human brain could be directly linked to the extent of synaptic potentiation mediated by mechanisms of LTP in specific neural networks. The expression of this synaptic reorganization during sleep, mainly during slow-wave sleep (SWS), should allow an active synaptic downscaling, essential for the appropriate cellular functions and associated to improvements in performances after a night-sleep. In other words, an increased LTP during wakefulness would need a synaptic homeostasis during sleep, expressed by an increased amount of slow-wave activity (SWA). This homeostasis also has a local component, that is specific brain areas involved in synaptic potentiation should show a higher amount of SWA during subsequent sleep than other areas; this has been clearly shown by a recent experiment, published in Nature by the director of Unit III (Huber et al., 2004).
What is lacking is a demonstration that 1) local homeostasis of SWA is direct consequence of LTP-like plastic changes in human brain, 2) duration of wakefulness affects cortical plasticity and connectivity, 3) also patients affected by neurological diseases (i.e., with unilateral stroke and with Alzheimer’s disease) show a homeostatic local response of SWA as a consequence of waking plastic changes, and that there is a bi-directional relationship between beficial changes due to cortical plasticity during waking and the local increase of SWA during sleep.
More in detail, the current program is aimed to assess:
1) the effect of LTP-like plastic changes in human brain on local homeostasis of slow-wave activity (SWA) during sleep (Unit I –De Gennaro-).
2) the effect of high-frequency electrophysiological stimulation, associated with synaptic potentiation, during waking of patients with chronic unilateral stroke on SWA homeostatic response during sleep (Unit II –Pizzella-).
3) the relationship between regional levels of SWA during sleep of patients with chronic unilateral stroke and the extent of LTP-like plastic changes during waking (Unit II –Pizzella-).
4) the relationship between regional levels of SWA during sleep of Alzheimer’s disease patients (AD) and of normal elderly (Nold) and the extent of LTP-like plastic changes during waking (Unit II –Pizzella-).
5) the effect of sleep deprivation on cortical excitability and connectivity (Unit III –Massimini-). <<<

Principal Investigator

Luigi De Gennaro Università degli Studi di ROMA "La Sapienza"

Research Objectives

Several lines of evidence support the notion that plastic changes occurring during wakefulness in human brain could be directly linked to the extent of synaptic potentiation mediated by mechanisms of long-term potentiation (LTP) in specific neural networks. The expression of this synaptic reorganization during sleep, mainly during slow-wave sleep (SWS), should allow an active synaptic downscaling, essential for the appropriate cellular functions and associated to improvements in performances after a night-sleep (Tononi &amp; Cirelli, 2003; Tononi &amp; Cirelli, 2006). In other words, an increased LTP during wakefulness would need a synaptic homeostasis during sleep, expressed by an increased amount of slow-wave activity (SWA). As shown by a recent experiment on Nature by the scientific director of Unit III, this homeostasis also has a local component, that is specific brain areas involved in synaptic potentiation should show a higher amount of SWA during subsequent sleep than other areas (Huber et al., 2004).
In the theorethical framework of this Synaptic Homeostasis Hypothesis (Tononi e Cirelli, 2006), some specific predictions on normal human sleep and on some neurological disorders will be tested.
1) Effect of LTP-like plastic changes in human brain on local homeostasis of slow-wave activity (SWA) during sleep (Unit I –De Gennaro-).
The hypothesis that local homeostasis of SWA is direct consequence of LTP-like plastic changes in humans will be assessed by a specific Transcranial Magnetic Stimulation (TMS) paradigm (Paired Associative Stimulation, PAS) that induces LTP-like plastic changes in human motor cortex during wakefulness (Stefan et al., 2000; Stefan et al., 2002; Wolters et al., 2003; Wolters et al., 2005).
2) Effect of high-frequency electrophysiological stimulation, associated with synaptic potentiation, during waking of patients with chronic unilateral stroke on SWA homeostatic response during sleep (Unit II –Pizzella-).
This will be tested by inducing in these patients during waking plastic changes changes with non-invasive transcranial direct current stimulation (tDCS) (Hummel et al., 2004), and measuring subsequent SWA homeostatic response during sleep. According to the hypothesis, the beneficial effect of non-invasive cortical stimulation on a set of hand functions that mimic activities of daily living in the paretic hand of patients should be mirrored by a subsequent increase of SWA in the affected than in the unaffected hemisphere.
3) Relationship between regional levels of SWA of patients with chronic unilateral stroke and the extent of cortical plastic chages during waking (Unit II –Pizzella-).
The extent of plastic changes in motor cortex of patients with chronic unilateral stroke, as shown by changes in the excitability of the motor by the coupling of a single electric stimulus delivered on a peripheral nerve with a single TMS pulse on M1 (Paired Associative Stimulation, PAS: Stefan et al., 2000; Stefan et al., 2002; Wolters et al., 2003; Wolters et al., 2005), will be correlated with an increased SWA in the same motor cortex.
4) Relationship between regional levels of SWA during sleep of Alzheimer’s disease patients (AD) and of normal elderly (Nold) and the extent of deterioration during waking of specific functions (Unit II –Pizzella-).
The extent of plastic changes in motor cortex of Nold and AD patients, as shown by changes in the excitability of the motor by the coupling of a single electric stimulus delivered on a peripheral nerve with a single TMS pulse on M1 (Paired Associative Stimulation, PAS: Stefan et al., 2000; Stefan et al., 2002; Wolters et al., 2003; Wolters et al., 2005), will be correlated with SWA changes in the same motor cortex.
5) Effect of sleep deprivation on cortical excitability and connectivity (Unit III –Massimini-).
This will be tested in humans by recording cortical EEG potentials evoked by TMS single pulses. According to the hypothesis, the amplitude of TMS-evoked potentials on the scalp should be higher at the end of a day of wakefulness and lower after a night of sleep, and a further increase of EEG potentials should be recorded after a prolonged sleep deprivation. Furthermore, TMS responses should increase selectively in brain areas where local slow-wave homeostasis has been induced. <<<

Timescale

24 months

National and international background

General background
For almost a century, several studies showed beneficial effects of sleep on memory function in animals and humans for different types of learning materials (Jenkins &amp; Dallenbach, 1924; Smith, 1995; Peigneux et al., 2001; Walker &amp; Stickgold, 2006). Recent studies in molecular genetics, neurophysiology, cognitive and behavioral neuroscience have strengthened the idea that sleep may play an important role in learning and memory, although the extent of this role remains hostly debated (Siegel, 2001; Stickgold &amp; Walker, 2005; Vertes &amp; Siegel, 2005). Undoubtedly the reason this issue continues to be ‘hot’ in the sleep field, and possibly in the neurosciences in general, is that it speaks both to the function of sleep and to the nature of memory processing.
In fact, functions of human sleep remain unclear. During much of sleep, cortical neurons undergo slow oscillations in membrane potential, which appear in electroencephalograms (EEG) as slow wave activity (SWA) of &lt;4.5 Hz (Steriade, 2000). SWA is the most pronounced EEG feature of non-rapid eye movement (NREM) sleep, is also a reliable predictor of sleep intensity. An important feature of slow-wave activity during sleep is that it increases as a function of previous wakefulness, and it gradually decreases in the course of sleep (Borbèly &amp; Achermann, 2000). This homeostatic regulation suggests that slow-wave activity may be linked to some restorative aspect of sleep. However, also the mechanisms and functions of slow-wave homeostasis are still unclear. A recently formulated hypothesis, the Synaptic Homeostasis Hypothesis links sleep with synaptic homeostasis, argueing that the specific function of sleep is to downscale the weight of cortical synapses (Tononi and Cirelli, 2006).
Synthetically the hypothesis states that:
1) During wakefulness, learning and plasticity processes, such as long-term potentiation (LTP) bring to a net increment of synaptic weight in several cortical circuits (Cirelli and Tononi, 2004). Globally, the strength of intracortical connections reaches a maximum toward the end of the day.
2) Upon falling asleep, the increased connectivity determines a high degree of synchronization among cortical neurons (Bazhenov et al., 2002). This is reflected in the occurrence, especially at the beginning of the night, of high-amplitude slow oscillations that are more prominent where synapses have been potentiated during previous wakefulness (Huber et al., 2004).
3) The rhythmic and repeated occurrence of depolarization/hyperpolarization sequences, in association with a lack of noradrenergic modulation (Cirelli and Tononi, 2000) brings to an homogeneous reduction in the strength of all cortical synapses (through mechanisms of spike timing-dependent plasticity). This process reduces all synaptic weights still preserving the relative differences and, thus, preserving learning. The process is self-limiting.
4) Synaptic downscaling plays a fundamental homeostatic role with respect to the metabolic and space requirements of the brain. In addition, by eliminating weakly (incidentally) potentiated synapses, it increases the signal-to-noise ratio and the specificity within cortical circuits. In this way downscaling may promote both memory consolidation and performance improvement.
Since the synaptic reorganization could be directly linked to the extent of synaptic potentiation mediated by mechanisms of LTP during wakefulness in specific neural networks, an increased LTP during wakefulness would need of a synaptic homeostasis during sleep (Tononi &amp; Cirelli, 2006), expressed by an increased amount of SWA during sleep. Plastic changes occur in human brain through much of waking life (not necessarely when engaged in specific learning tasks), whether strong presynaptic firing is accompanied by postsynaptic depolarization or firing in the presence of appropriate levels of neuromodulators, which should be a frequent occurrence during alert wakefulness. Plastic changes occurring during wakefulness, at least in the adult, would result more often in LTP than in long-term depression (LTD), thus resulting in a net potentiation of synaptic strength. Direct evidence supporting this hypothesis comes from anatomical studies demonstrating a net and diffuse increase in synaptic density in animals exposed to enriched environments likely to induce LTP-like molecular changes (Klintsova &amp; Greenough, 1993). Local increases in synaptic density have also been observed by stimulating a whisker for 24 h that produces a selective net increase of synaptic density (by 35%) on cortical neurons in the corresponding barrel field (Knott et al., 2002). Evidence for synaptic potentiation comes from the finding that spontaneous wakefulness is regularly associated with the diffuse induction of molecular changes usually associated with LTP (Cirelli et al., 2000; Cirelli et al., 2004), including the phosphorylation of CREB and the induction of genes such as Arc, BDNF, NGFI-A, Homer, and Narp (e.g., Ying et al., 2002; Silva et al., 2003; Wallace at al., 1995). This induction of LTP-related genes during wakefulness can increase further if animals are kept awake longer by gentle handling, or if they engage in extensive exploration of their environment (Tononi &amp; Cirelli, 2006). During sleep, by contrast, the expression of LTP-related genes is severely reduced or abolished (Cirelli et al., 2000; Cirelli et al., 2004). Support for the notion that synaptic strength may increase during wakefulness also comes from experiments in humans (Braun et al., 1997) and in rats (Vyazovskiy et al., 2004), showing that brain metabolism, which is mostly due to synaptic activity, increases from early to late wakefulness.
The hypothesis states that amount of synaptic potentiation that has occurred during wakefulness should directly affect the homeostatic regulation of SWA, and specific brain areas involved in synaptic potentiation should show a higher amount of SWA during subsequent sleep than other areas.
A direct demonstration that sleep homeostasis has a local component, which can be triggered by a learning task involving specific brain regions, has been provided by a recent study (Huber et al., 2005). The induction of local plastic changes associated with practicing a visuomotor task (a rotation adaptation task) was associated with a local induction of SWA in subsequent sleep. The increase in power, as measured by high definition EEG, was mostly in the slow wave frequency range, and it declined over time, just like the global homeostatic response of SWA. Moreover, the increase of SWA (by 27%) was localized exactly at a small cluster of electrodes in right parietal cortex (areas 40 and 7), that is brain regions activated during wakefulness by the same rotation adaptation task (Ghilardi et al., 2000). Furthermore, the local increase in SWA after learning correlated with improved performance after sleep.
An indirect support to this local sleep homeostasis also comes from studies documenting that unilateral somatosensory stimulation during waking of humans (Kattler et al., 1994) and rats (Vyazovskiy et al., 2000) is associated to a local increase of SWA. Furthermore, the topographic differences in SWA homeostasis, with frontal regions showing an especially strong response to sleep deprivation (Ferrara et al., 2002; Finelli et al., 2001) also are coherent with topographic differences in the susceptibility to plastic changes (Trachtenberg et al., 2002). Finally, after visual deprivation during the critical period—a procedure associated with synaptic depression (Heynen et al., 2003), SWA decreased by 40% in the absence of changes in sleep architecture (Miyamoto et al., 2003).
In this theorethical and empirical framework, some specific predictions on normal human sleep and on some neurological disorders can be formulated.
1) LTP-like plastic changes in human brain should directly affect local homeostasis of SWA.
Since specific experimental paradigms by means of Transcranial Magnetic Stimulation (TMS) provides a viable method to induce LTP-like plastic changes in human motor cortex during wakefulness, the hypothesis that local homeostasis of SWA is direct consequence of LTP-like plastic changes in human brain can be directly assessed.
2) Brain metabolism, cortical excitability and connectivity should increase during wakefulness and decrease after sleep, and sleep deprivation should induce a further increase.
This can be tested in humans by recording cortical EEG potentials evoked by transcranial magnetic stimulation (TMS). According to the hypothesis, the amplitude of TMS-evoked potentials on the scalp should be higher at the end of a day of wakefulness and lower after a night of sleep, and a further increase of EEG potentials should be recorded after a prolonged sleep deprivation. Furthermore, TMS responses should increase selectively in brain areas where local slow-wave homeostasis has been induced, for example, with the visuomotor task discussed above (Huber et al., 2004).
3) Waking behaviors of patients with chronic stroke associated with synaptic potentiation should be followed by an increased SWA homeostatic response during sleep. This can be tested for example using learning tasks or high-frequency electrophysiological stimulation [A. transcranial Direct Current Stimulation (tDCS); B. Paired Associative Stimulation (PAS)].
? Plastic changes changes induced in these patients with non-invasive tDCS (Hummel et al., 2004), should increase SWA homeostatic response. In other words, a beneficial effect of non-invasive cortical stimulation on a set of hand functions that mimic activities of daily living in the paretic hand of patients with chronic monohemispheric stroke should be mirrored by a subsequent increase of SWA in the affected than in the unaffected hemisphere.
? Similarly, the extent of plastic changes in motor hand of patients with chronic monohemispheric stroke, as shown by changes in the excitability of the motor by the coupling of a single electric stimulus delivered on a peripheral nerve with a single TMS pulse on M1 (Paired Associative Stimulation, PAS: Stefan et al., 2000; Stefan et al., 2002; Wolters et al., 2003; Wolters et al., 2005), should correlate with an increased SWA in the affected than in the unaffected hemisphere.
4) Regional levels of SWA during sleep of Alzheimer’s disease patients (AD) and of normal elderly (Nold) should be associated to the extent of deterioration during waking of specific functions, as a model of altered synaptic plasticity.
This prediction is coherent with the well-known reduction of SWS in elderly (for a meta-analysis, Ohayon et al., 2004) and in AD patients (Loewenstein et al., 1982; Feinberg et al., 1967; Prinz et al., 1982; Blois et al., 1983; Prinz et al., 1982; Bliwise et al., 1989; Vitiello et al., 1990), with the decreased frontal SWA in old healty subjects (Landolt &amp; Borbely, 2001) and with the decreased frontal predominance of SWA after sleep loss (Munch et al., 2004).
Hence, the extent of plastic changes in motor cortex of Nold and AD patients, as shown by changes in the excitability of the motor by the coupling of a single electric stimulus delivered on a peripheral nerve with a single TMS pulse on M1 (Stefan et al., 2000; Stefan et al., 2002; Wolters et al., 2003; Wolters et al., 2005), should correlate with an increased SWA in the same motor cortex. <<<