The Genesis of the BOLD Signal

Although neuronal activity is accompanied by increased energy utilization, it is not the increased energy use itself that directly triggers the associated increase in blood flow. Instead, the increased blood flow is a direct consequence of presynaptic neurotransmitter action (Attwell and Iadecola,

2002) and thus reflects local neuronal signaling. Increases in the BOLD signal are correlated electrophysiologically most strongly with the local field potential rather than the neuronal firing rate (Logothetis et al., 2001). The volume over which blood flow increases associated with neuronal activity is determined by the perfusion territory of local arterioles.

There may be multiple mediators of the arteriolar response, but nitric oxide (NO) and eicosanoids clearly are important under normal circumstances (Buerk et al., 2003; St. Lawrence et al., 2003). Binding of glutamate to receptors on astrocytes triggers NO release. Neuronal-hemodynamic coupling thus may change with disease (D'Esposito et al.,

2003). For example, early after ischemic infarction, regional cerebral blood flow (rCBF) may be uncoupled (lack of response) from regional cerebral metabolic rate for glucose consumption (rCMRglu) (persistent response) as a consequence of ischemia-induced inhibition of nitric oxide syn-

thetase (NOS). Even chronically there may be changes in brain neuronal-hemodynamic coupling related to this or other mechanisms (Pineiro et al., 2002).

In MS there is widespread upregulation of NOS (Smith and Lassmann, 2002). We have made an initial assessment of potential changes in hemodynamic coupling in patients with relapsing-remitting MS (Saini et al., 2004). Hemo-dynamic responses to motor activation reassuringly appear well preserved, suggesting that there are no major confounds to inferring functional localization on the basis of fMRI in studies of MS (Fig. 2).

III. Motor Learning by the Healthy Brain Provides a Basis for Understanding Adaptive Functional Reorganization of the Brain after Injury

The brain arguably has not evolved so much to adapt to injury as to adapt its responses to changing external or internal environments. It is attractive to hypothesize that the broad range of examples of functional reorganization of the brain in response to altered external stimuli or internal state rely on common mechanisms. Studies in animal models have shown that similar functional changes in sensorimotor cortical areas are associated with peripheral nerve injury (Seil, 1997), amputation (Florence et al., 1998), or repeated direct cortical stimulation (Nudo et al., 1990), for example. Peripheral

Signal Intensity (%)

■ Controls A MS patients -

3.5 -i 3.02.52.01.51.00.5 0.0 -0.5-1.0-1.5-

M f m j[f

\ rv t\

■ i i i i i i i i i i i i i i i i i i i 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60

Time(s)

Figure 2 Group mean hemodynamic responses are shown for the single most highly activated voxel in the primary sensorimotor cortex of healthy controls and patients with clinically stable relapsing-remitting multiple sclerosis during a repetitive writing task performed with the right hand (see Saini et al., 2004). No significant differences in the rate of signal intensity rise or maximum increase in blood oxygenation level dependent (BOLD) signal are between control and patient groups. These results suggest that the relationship between neuronal activity and the BOLD response is not altered in these patients with MS.

Figure 2 Group mean hemodynamic responses are shown for the single most highly activated voxel in the primary sensorimotor cortex of healthy controls and patients with clinically stable relapsing-remitting multiple sclerosis during a repetitive writing task performed with the right hand (see Saini et al., 2004). No significant differences in the rate of signal intensity rise or maximum increase in blood oxygenation level dependent (BOLD) signal are between control and patient groups. These results suggest that the relationship between neuronal activity and the BOLD response is not altered in these patients with MS.

deafferentation in monkeys leads to a reduction of the area of cortical responsiveness for the deafferented limb (Merzenich et al., 1983; Garraghty and Kaas, 1991) that mirrors enlargements of cortical responsiveness with motor-skill learning (Kleim et al., 1998) or repetitive intracortical microstimulation (Nudo et al., 1990). Changes in functional organization of the cortex related to altered afferent or efferent activity must share at least some common mechanisms, whatever the specific cause of the altered activity; however, the most common manifestation of this type of change is for learning. Most generally formulated learning involves changing the brain representations for the relationship between a stimulus and its consequences. Studies of learning therefore may provide important lessons to help understand responses of the brain to injury.

Motor learning can occur over both the short and long term. Different patterns of behavioral and brain functional activity changes are associated with the two time frames for learning. With short-term learning, there is a rapid improvement in performance associated with a decrease in the specific attentional demands of the task (Floyer-Lea and Matthews, 2003). This is associated with decreasing activity in prefrontal cortex and a progressive "focusing" of activity in the primary motor cortex contralateral to the limb moved (Fig. 3) (see for example, Ramnani et al., 2000; Toni et al., 2001a; Grafton et al., 1992; Karni et al., 1995, 1998) as performance becomes more automatic with increasing practice. It is intriguing to note the substantial similarities between the dynamic change in activation patterns illustrated in Fig. 3 and the dynamic changes in brain activity with recovery after acute stroke (see for example, Ward et al., 2003).

The acquisition of motor skills of many types appears to be associated with altered primary motor cortex representations;

Higjit Flint ary Motor Cortex

1234567SS1Û Block

Frontal Pole

1234567SS1Û Block

Frontal Pole

Higjit Flint ary Motor Cortex

R

L

r

ñ

Left Print ary Motor Cortes

12Ï4567SS 1Q Bledi

Pre-S up pi em entary Motor Area a i-

Dorsolateral Prefrontal C ortex

Dorsolateral Prefrontal C ortex

1334-3678» 10

Intra-Parietal Sulcus

Figure 3 Brain activation correlated with improvement in performance for a group of healthy subjects when learning to exert finger flexion pressure with the (dominant) right hand according to a visually-presented pattern, together with their time courses (group random effects image, z < 2.3, P < 0.01, corrected). Areas that showed an increase in activation are shown in red; those areas that showed a decrease in activation are shown in blue. At the start of the experiment the level of activation is similar in the right (A) and left (B) motor cortex. Although both sides show a decrease in activation over the course of the experiment, the right hemisphere shows a more significant decrease than the left hemisphere, so that by the end of the experiment, the activation is almost completely left-lateralized. The dorsolateral prefrontal cortex (DLPFC) and the presupplementary motor area (pre-SMA) show contrasting time courses. The DLPFC (E) shows initially high activation that drops off sharply, reaching baseline levels before the early learning phase is complete. The pre-SMA (D) shows an initial increase in activation before decreasing, reaching a peak at the time when activation in the DLPFC has declined. The frontal pole (C) has a time course closer to that of the behavioral learning itself, whereas the intraparietal sulcus (IPS) (F) shows a more sustained decrease that continues into the postlearning phase.

for example, qualitatively similar phenomena are found after learning an isometric tracking task (Floyer-Lea et al., unpublished) or after development of skills in a sport such as badminton (Pearce et al., 2000). Analogous changes have been described in the nonhuman primate primary motor cortex based on invasive cortical mapping. For example, as monkeys learn to draw peanuts from smaller wells, there is reorganization of motor representations for limb movement in the primary motor cortex (Nudo et al., 1996a). It is important to note, however, that both human and primate studies have emphasized that these changes in patterns of brain activity associated with learning are task specific (Floyer-Lea and Matthews, 2003; Karni et al., 1995; Kleim et al., 1998). Even two similar (but distinct) finger movement sequences do not use precisely the same representation (Karni et al., 1995).

This suggests that the substrate for general skill learning (e.g., being able to play the piano, as opposed to being able to play any specific piece on the piano) is not simply a change in motor representation in the primary motor cortex. One may speculate that this form of learning, which possibly is central to understanding recovery, is mediated by changes that are much more widely distributed. A potentially important part of such a network for change involves frontoparietal circuits. Activity in the dorsolateral prefrontal cortex is associated with problem solving and shows performance-related increases in activity (Duncan et al., 2000). Parietal activity is involved in evaluating the potential significance of stimuli (Driver and Mattingley. 1998; Mort et al., 2003). The posterior parietal cortex is also concerned with aspects of visual and spatial stimulus selection (Fink et al., 1996). Consistent with this, the preparation to move is associated with frontoparietal activation.

It has been more difficult to study subcortical activity, but subcortical regions clearly also contribute substantially to the brain circuits involved in learning. Subcortical circuits appear to be involved particularly for learning implicit tasks (Doyon et al., 2002; Laforce, Jr. and Doyon, 2001) or when tasks become more automatic (when performance is less affected by distracters) (Floyer-Lea and Matthews, 2003).

Although lessons can be drawn from learning in the healthy brain, however, it is unlikely that motor recovery after brain injury simply represents learning redux. A number of special mechanisms may alter the potential for functional reorganization in the context of brain injury (Witte, 1998). Ischemic brain injury triggers complex changes in cortical excitability (Witte et al., 2000): local to the infarct, excitability is decreased (Neumann-Haefelin and Witte, 2000), but more distantly, excitability is increased. The latter occurs in association with downregulation of inhibitory y-aminobu-tyric acid (GABA) receptor levels (Redecker et al., 2002). Studies in the healthy brain suggest that functional reorganization in motor cortex could be strongly affected by GABAergic activity (Butefisch et al., 2000), as well as the excitatory glutaminergic system acting through N-methy-

D-aspartate (NMDA) receptors (Hess et al., 1994). Deafferentation induces at least transiently increased excitability in the contralateral hemisphere (Werhahn et al., 2002, 2003). Loss of sensory afferents alone reduce local GABA concentrations, potentially helping to drive short-term functional reorganization. These factors may even contribute to the structural and functional changes that occur in the brain many months after the injury, when primary repair has slowed or stopped (Taub et al., 2002, 2003). Although similar phenomena are much less well characterized in MS, TMS data, for example, suggest that such changes may be important (Ho et al., 1999).

Another distinction between simple motor learning and recovery after brain injury is that the latter represents relearning to perform tasks in different ways. Either compensatory strategies must be developed or new pathways must be recruited adaptively. In general, this must involve some systems distinct from those engaged for naive motor learning. Exciting work has begun to explore the more general mechanisms by which feedback relevant to the action plan is processed that emphasizes these differences. Ramnani et al. used a simple visuomotor association learning task and distinguished between activity time-locked to positive, negative, and neutral (control) feedback (Ramnani and Matthews, 2003). Specific brain regions (including the dorsal prefrontal cortex, cingulate cortex, anterior temporal cortex, ventral striatum/pallidum, thalamus, and the amygdala) showed increased activity with meaningful feedback. When the contingencies between stimuli and actions were changed to create a relearning trial, meaningful feedback activated the supramarginal gyrus (left-lateralized for negative feedback, right-lateralized for positive feedback) to a significantly greater degree than with the initial learning. Two conclusions relevant to the current discussion can be drawn from this. First, specific mechanisms are responsible for responding to the feedback necessary to modify behavior. Second, the specific pathways engaged with performance feedback are determined in part by the context in which they are engaged.

IV. Direct Observations on Patients After Brain Injury: Common Brain Mechanisms Contribute to Adaptive Changes After Corticospinal Tract Injury

0 0

Post a comment