Bernhard-hommel.eu

European Journal of Neuroscience, Vol. 21, pp. 591–595, 2005 ª Federation of European Neuroscience Societies SHORT COMMUNICATIONCaffeine, but not nicotine, enhances visual feature binding Lorenza S. Colzato,1 Sabrina Fagioli,2 Vicki Erasmus1 and Bernhard Hommel11Leiden University, Department of Psychology, Cognitive Psychology Unit, Postbus 9555, 2300 RB Leiden, the Netherlands2University of Rome ‘La Sapienza’, Department of Psychology, Rome, Italy Keywords: binding problem, caffeine, cholinergic system, neuronal synchronization, nicotine, visual perception The distributed organization of the human visual cortex calls for a mechanism that integrates and binds the features of a perceivedevent, and neural synchronization is a prime candidate to serve that purpose. Animal studies suggest that synchronization in thevisual cortex is enhanced by the muscarinic cholinergic system. Here we show that in healthy humans the binding of shape andcolour, and of shape and location, of visual objects is increased by stimulating the muscarinic cholinergic system (caffeineconsumption) but not by stimulating the nicotinic cholinergic system (nicotine consumption). Binding across perception and action isunaffected by either manipulation, suggesting a specific link between the visual system and the muscarinic cholinergic system.
When we are facing multiple objects, a red apple and a yellow banana, cognitive processes such as the planning of multifeatured actions say, the features of these objects are coded and processed in different cortical areas, yet what we perceive are well integrated objects and not Although the available evidence points to a link between neural mere bundles of attributes. The human brain thus seems to employ synchronization and feature binding, it remains unclear how and under some kind of feature-binding mechanism which integrates the neural which circumstances synchronization emerges, which makes it patterns coding the features belonging to a given event and makes us difficult to appreciate its true functional role. A potentially important validly perceive that bananas are yellow but not red (Treisman, 1996).
observation in this context is the finding that, in the cat, neocortical One candidate for such a mechanism are conjunction detectors, that is, synchrony in the gamma band ( 30–70 Hz) is enhanced by neural units that are selective for the presence of particular feature muscarinic cholinergic agonists and disrupted by antagonists combinations (Mozer, 1991; Riesenhuber & Poggio, 1999). Coding by (Rodriguez et al., 2000, 2001, 2004). Empirical evidence for this means of conjunction detector makes sense for processing highly link is consistent with Colzato et al. (2004), who observed in humans probable, evolutionarily important feature conjunctions. However, that alcohol, which is known to cause hypoactivity of the cholinergic excessive numbers of conjunction detectors would be necessary to system (Sanna et al., 1994; Minami et al., 1997), impairs feature code any arbitrary feature combination, suggesting that frequently binding in visual perception but not binding across perception and changing feature relations are processed in a different way (Singer, action. This fits with the hypothesis that the binding of visual features 1994; Colzato et al., 2004; Hommel, 2004). Better suited for this case is driven by the muscarinic cholinergic system (Metherate et al., 1992; seems to be the neural coupling of cell populations (Abeles, 1991; Rodriguez et al., 2000, 2001, 2004).
Engel & Singer, 2001), which is assumed to be achieved by The aim of the present study was to set up a specific test of this coordinating and synchronizing the firing rates of cells referring to hypothesis in healthy human subjects: can it be demonstrated that, the same event, i.e. feature conjunctions may be coded spontaneously first, cholinergic agonists increase the binding of visual features but through the temporal coherence of their neural codes.
not of other features and that, second, this increase is specifically The possible role of neural synchronization in binding features driven by agonists of the muscarinic but not the nicotinic cholinergic within the visual and the motor system has been implicated in many system? The two agonists we compared were caffeine and nicotine.
studies. Apart from numerous animal studies (Abeles, 1991; Engel & The behavioural effects of caffeine have been attributed to several Singer, 2001; Roelfsema et al., 1997), transient increases in neuromodulatory systems, including dopaminergic, GABA, serotin- synchronization in the gamma frequency range have been observed ergic, cholinergic and noradrenergic pathways (Nehlig et al., 1992).
in healthy humans in visual tasks such as figure–ground discrimination However, there is evidence for a direct muscarinic cholinergic link (Engel & Singer, 2001), switching between bistable visual figures between caffeine and the processing and short-term memory of visual (Keil et al., 1999), and the retention of visual patterns in short-term memory (Tallon-Baudry et al., 1999) Motor tasks have revealed First, caffeine impacts the muscarinic cholinergic but not the nicotinic cholinergic system (Sorimachi et al., 1992). Cholinergicsystems are under tonic inhibitory control by endogeneous adenosine,as suggested by observations of an increase of adenosine extracellular Correspondence: Dr Bernhard Hommel, as above.
concentrations in the basal forebrain cholinergic areas of the behaving cat during prolonged wakefulness (Porkka-Heiskanen et al., 1997), Received 17 July 2004, revised 5 November 2004, accepted 9 November 2004 and of dramatic decreases in waking induced by the perfusion of adenosine into the same regions (Portas et al., 1997). Also, the state of defined); (ii) the nonspatial stimulus features (e.g. shape and ⁄ or prolonged sustained wakefulness can be mimicked by increasing colour); and (iii) the response (for an overview, see Hommel, 2004).
adenosine levels in basal forebrain cholinergic regions but not by These interactions all follow the same pattern: performance is increasing those levels in noncholinergic areas (Porkka-Heiskanen impaired in partial-repetition trials, that is, if one stimulus feature et al., 1997). Given that caffeine (and other xanthines) are adenosine (or the response) is repeated (e.g. S1 shape ¼ S2 shape), while the antagonists, it thus makes sense to assume that they unfold their other is not (e.g. S1 colour „ S2 colour or R1 „ R2). This arousing impact by reducing the amount of inhibition adenosine exerts demonstrates that the mere co-occurrence of a feature–feature or on cholinergic pathways. Moreover, evidence that the cholinergic feature–response conjunction is sufficient to create a temporary impact on cortical states is mainly muscarinic (Lamour et al., 1982) binding of the respective feature codes, a kind of ‘event file’ suggests a central role of muscarinic cholinergic pathways in linking (Hommel, 1998, 2004; Hommel & Colzato, 2004). Reactivating one member of this binding (as with feature repetition) spreads activation Second, a number of studies suggest that the impact caffeine exerts to other members, which calls for a time-consuming rebinding process on the muscarinic cholinergic system eventually targets visual in partial-repetition trials. Most important for our purposes, these processes. Monkey studies have shown that systemic injections of partial-repetition costs can be taken to indicate feature–feature and scopolamine, a muscarinic cholinergic antagonist, impair the encoding feature–response binding, which is why we chose them as behavioural of new visual objects but have little effect during recognition (Aigner markers. Along the lines described above we thus expected caffeine & Mishkin, 1986; Aigner et al., 1991). In humans, the intake of and nicotine consumption to decrease partial-repetition costs for caffeine attenuates the scopolamine-induced impairment of, among shape–colour conjunctions and for shape–location conjunctions (i.e.
other things, perceptual sensitivity in visual search, visual short-term visual–visual bindings), but not for shape–response or location– memory, and reading (Riedel et al., 1995). At the same time, caffeine does not modulate the impact of scopolamine on simple and choicereaction time, suggesting that caffeine has a specific effect on visualprocessing via muscarinic cholinergic pathways (cf. Smith et al., 2003). Consistent with this interpretation, a recent PET study in Eighteen volunteers took part in each study and served in two humans using a visual task provides evidence that muscarinic experimental sessions separated by 3–7 days. Informed consent was cholinergic effects modulate visual attribute processing (Mentis obtained from all participants after the nature and possible conse- et al., 2001). In particular, muscarinic action was found to predom- quences of the study were explained to them; the protocol was inate in striate cortex (Brodmann Area 17) and lateral visual approved by the local ethical committee (Leiden University, Faculty of association areas (18 and 19), whereas nicotinic action predominated Social and Behavioural Sciences). Subjects in the caffeine group in the thalamus and inferior parietal regions (areas 39 and 40). In (habitual coffee consumers: 3–5 units per day on average) and in the contrast to caffeine, nicotine is well known to induce cholinergic nicotine group were in the age range of 20–30, healthy nonsmokers, facilitation via nicotinic but not muscarinic receptors.
not on medication or drugs, and without neurological or psychiatric In view of this evidence, we assumed that, first, caffeine would act history according to self-report. All experimental sessions were held in via muscarinic cholinergic pathways and nicotine via nicotinic the morning to prevent time-of-day effects and the possible occurrence cholinergic pathways and that, second, enhancing muscarinic but not of withdrawal symptoms. Participants were asked to refrain from all nicotinic action would affect visual feature binding. Accordingly, we caffeine-containing foods and beverages for 12 h prior to the expected caffeine but not nicotine intake to affect the behavioural experimental sessions, not to consume alcohol on the night before measure of feature integration we used, and this effect to be specific to the experimental session and to have a normal night’s rest. A double- blind, placebo-controlled, randomized cross-over design with coun- We adopted the task from Hommel (1998), which involves the terbalancing of the order of conditions was used to avoid expectancy repetition of task-related and -unrelated visual features and of the effects. Treatments were deceptive: subjects were led to believe that response (see Fig. 1). The standard findings are interactions involving they were drinking regular coffee or keeping a nicotine patch each (i) the task-related stimulus features (e.g. shape, if and only if it signals experimental session. Subjects’ compliance was encouraged by taking the response; location, if and only if the responses are spatially a saliva sample (not further analysed) at the beginning of each Fig. 1. Sequence of events in the present experiments (Hommel, 1998). A response cue signaled a left or right key press (R1) that was to be delayed untilpresentation of S1, a red or green, vertical or horizontal line in a top or bottom box. S2 appeared 1 s later: another red or green, vertical or horizontal line in the top orbottom box. S2 shape signaled R2, also a speeded left or right key press. R2 speed and accuracy were analysed as functions of the repetition vs. alternation ofstimulus shape, colour and location, and of the response.
ª 2005 Federation of European Neuroscience Societies, European Journal of Neuroscience, 21, 591–595 ª 2005 Federation of European Neuroscience Societies, European Journal of Neuroscience, 21, 591–595 experimental session. Placebo and dose caffeine and nicotine quan- and caffeine exert (at least part of) their muscarinic cholinergic tities corresponded to 250 mg lactose and 250 mg caffeine ( 3 cups influence on visual processing by increasing and decreasing the of coffee) and 0 and 7 mg nicotine ( 1 cigarette), respectively.
variability of firing rates, respectively (cf. Borchard et al., 2000); this Subjects completed a version of the task adapted from Hommel again might affect higher synchronization frequencies more than lower (1998); see Fig. 1. They faced three grey, vertically arranged boxes in frequencies (Kopell et al., 2000) and, hence, local, intra-area bindings the middle of a monitor and carried out two responses per trial. R1 was more than long-range interarea bindings.
a delayed simple reaction with the left or right key, as indicated by a100%-valid response cue (three left- or right-pointing arrows in themiddle box) that preceded the trigger stimulus S1 by 3000 ms. S1 varied randomly in shape (a thin vertical or horizontal line), colour Abeles, M. (1991) Corticonics: Neural Circuits of the Cerebral Cortex.
(red or green) and location (top or bottom box). R1 was to be carried Cambridge University Press, Cambridge.
out as soon as S1 appeared, independent of its shape, colour or Aigner, T.G. & Mishkin, M. (1986) The effects of physostigmine and location; i.e. subjects were encouraged to respond to the mere onset of scopolamine on recognition memory in monkeys. Behav. Neurosci., 45, S1. R2 was a binary-choice reaction to the shape of S2 (vertical or Aigner, T.G., Walker, D.L. & Mishkin, M. (1991) Comparison of the effects of horizontal orientation), which also appeared in red or green and in the scopolamine administered before and after acquisition in a test of visual top or bottom box, 1000 ms after S1 onset. Responses to S1 and to S2 recognition memory in monkeys. Behav. Neural Biol., 55, 61–67.
were made by pressing the left or right shift-key of the computer Borchard, F., Singer, W. & Munk, M. (2000) Electrical stimulation of the keyboard with the corresponding index finger. Each session was substantia innominata leads to an increase in fast oscillatory multiunit composed of a factorial combination of the two possible shapes, activity in cat visual cortex. Soc. Neurosci Abstr., 30, 736.20.
Colzato, L.S., Erasmus, V. & Hommel, B. (2004) Moderate alcohol colours and locations of S2, the repetition vs. alternation of shape, consumption impairs feature binding in visual perception but not across colour, location, and the response, and three replications per condition.
perception and action. Neurosci. Lett., 360, 103–105.
Engel, A.K. & Singer, W. (2001) Temporal binding and the neural correlates of sensory awareness. Trends Cogn. Sci., 5, 16–25.
Fellous, J.-M. & Sejnowski, T.J. (2000) Cholinergic induction of oscillations in the hippocampal slice in the slow (0.5–2 Hz), theta (5–12 Hz), and gamma After excluding trials with missing (< 1500 ms) or anticipatory (35–70 Hz) bands. Hippocampus, 10, 187–197.
Gross, J., Schmitz, F., Schnitzler, I., Kessler, K., Shapiro, K., Hommel, B. & (< 200 ms) responses, mean reaction times (RTs) and proportions Schnitzler, A. (2004) Long-range neural synchrony predicts temporal of errors for R2 (i.e. the response to S2) were analysed as a function of limitations of visual attention in humans. Proc. Natl Acad. Sci. USA, 101, treatment (placebo vs. treatment) and the repetition vs. alternation of response (R1 fi R2), stimulus shape, colour, and location Hari, R. & Salenius, S. (1999) Rhythmical corticomotor communication.
(S1 fi S2). Anovas were performed by using a five-way design for Hommel, B. (1998) Event files: Evidence for automatic integration of stimulus repeated measures. Figure 2 shows the results from Caffeine and response episodes. Vis. Cognition, 5, 183–216.
Nicotine groups comparing treatment with placebo conditions. We Hommel, B. (2004) Event files: Feature binding in and across perception and replicated the earlier finding (Hommel, 1998; Hommel & Colzato, action. Trends Cogn. Sci., 8, 494–500.
2004) of a shape–colour and a shape–location interaction in both Hommel, B. & Colzato, L.S. (2004) Visual attention and the temporal dynamics of feature integration. Vis. Cognition, 11, 483–521.
groups: repeating the shape of a stimulus but not its colour or its Keil, A., Muller, M.M., Ray, W.J., Gruber, T. & Elbert, T. (1999) Human gamma location incurred a partial-repetition cost (see Fig. 2A and B).
band activity and perception of a Gestalt. J. Neurosci., 19, 7152–7161.
However, these two interactions were only affected by caffeine intake Kopell, N., Ermentrout, G.B., Whittington, M.A. & Traub, R.D. (2000) Gamma but not by nicotine. We also replicated the common interactions rhythms and beta rhythms have different synchronization properties. Proc.
Natl Acad. Sci. USA, 97, 1867–1872.
between response and shape, and between response and location in Lamour, Y., Dutar, P. & Jobert, A. (1982) Excitatory effect of acetylcholine on both groups (see Fig. 2C and D). Importantly, none of these effects different types of neurons in the first somatosensory neocortex of the rat: Laminar distribution and pharmacological characteristics. Neuroscience, 7,1483–1494.
Mentis, M.J., Sunderland, T., Lai, J., Connolly, C., Krasuki, J., Levine, B., Friz, J., Sobti, S., Schapiro, M. & Rapoport, M.D. (2001) Muscarinic versus nicotinic modulation of a visual task: a PET Study using drugs probes.
Our findings show that intake of caffeine, but not of nicotine, Neuropsychopharmacology, 25, 555–564.
Metherate, R., Cox, C.L. & Ashe, J.H. (1992) Cellular bases of neocortical increased the binding of visual features while both drugs spared cross- activation: modulation of neural oscillations by the nucleus basalis and domain bindings between visual features and manual responses. This endogenous acetylcholine. J. Neurosci., 12, 4701–4711.
observation lends support to the hypothesis that feature binding in Minami, K., Vanderah, T.W., Minami, M. & Harris, R.A. (1997) Inhibitory visual cortex is associated with neural synchronization enhanced by effects of anesthetics and ethanol on muscarinic receptors expressed inXenopus oocytes. Eur. J. Pharmacol., 339, 237–244.
the muscarinic cholinergic system (Rodriguez et al., 2000, 2001, Mozer, M.C. (1991) The Perception of Multiple Objects: a Connectionist 2004). Whereas the apparently selective link between synchronization and muscarinic receptors is consistent with animal studies on both Nehlig, A., Daval, J. & Debry, G. (1992) Caffeine and the central nervous neocortical and hippocampal synchronization (e.g. Fellous & system: Mechanisms of action, biochemical, metabolic and psychostimulant Sejnowski, 2000; Rodriguez et al., 2004), it remains to be determined effects. Brain Res. Rev., 17, 139–170.
Porkka-Heiskanen, T., Strecker, R.E., Thakkar, M., Bjorkum, A.A., Greene, why this link is selective for local visual feature binding. Interestingly, R.W. & McCarley, R.W. (1997) Adenosine: a mediator of the sleep-inducing local integration processes within visual and motor cortex are effects of prolonged wakefulness. Science, 276, 1265–1268.
commonly associated with synchronization frequencies in the gamma Portas, C.M., Thakkar, M., Rainnie, D.G., Greene, R.W. & McCarley, R.W.
band (Engel & Singer, 2001), while interarea synchronization, such as (1997) Role of adenosine in behavioral state modulation: a microdialysisstudy in the freely moving cat. Neuroscience, 79, 225–235.
in visuomotor integration (Roelfsema et al., 1997) or visual target Riedel, W., Hogervorst, E., Leboux, R., Verhey, F., van Praag, H. & Jolles, J.
selection (Gross et al., 2004), relies on lower frequencies in the beta (1995) Caffeine attenuates scopolamine-induced memory impairment in band ( 13–30 Hz). It thus seems possible that drugs such as alcohol humans. Psychopharmacology, 122, 158–168.
ª 2005 Federation of European Neuroscience Societies, European Journal of Neuroscience, 21, 591–595 Riesenhuber, M. & Poggio, T. (1999) Hierarchical models of object recognition linked receptors in Xenopus oocytes expressing brain messenger RNA—role in cortex. Nature Neurosci., 2, 1019–1025.
of protein kinase. Mol. Pharmacol., 45, 1004–1012.
Rodriguez, R., Kallenbach, U., Singer, W. & Munk, M.H. (2000) Effects of Singer, W. (1994) The organization of sensory motor representations in the cholinergic stimulation on gamma oscillations during visual responses in cat Neocortex: A hypothesis based on temporal coding. In: Umilta`, C. & visual cortex. Eur. J. Neurosci. Suppl., 12, 221.15.
Moscovitch, M. (eds), Attention and Performance XV: Conscious and Rodriguez, R., Kallenbach, U., Singer, W. & Munk, M.H. (2001) Reduction of Nonconscious Information Processing. MIT Press, Cambridge, MA, pp.
gamma-oscillatory responses after scopolamine application in cat visual cortex. Soc. Neurosci. Abstr., 27, 123.3.
Smith, A., Brice, C., Nash, J., Rich, N. & Nutt, D.J. (2003) Caffeine and central Rodriguez, R., Kallenbach, U., Singer, W. & Munk, M.H. (2004) Short- and noradrenaline: Effects on mood, cognitive performance, eye movements and long-term effects of cholinergic modulation on gamma oscillations and cardiovascular function. J. Psychopharmacol., 17, 283–292.
response synchronization in the visual cortex. J. Neuroscience, 24, 10369– Sorimachi, M., Yamagami, K. & Nishimura, S. (1992) A muscarinic receptor agonist mobilizes Ca2+ from caffeine and inositol-1, 4, 5-trisphosphate- Roelfsema, P.R., Engel, A.K., Koenig, P. & Singer, W. (1997) Visuomotor sensitive Ca2+ stores in cat adrenal chromaffin cells. Brain Res., 571, 154–158.
integration is associated with zero time-lag synchronization among cortical Tallon-Baudry, C., Kreiter, A. & Bertrand, O. (1999) Subtained and transient oscillatory responses in the gamma and beta bands in a visual short-term Sanna, E., Dildy-Mayfield, J.E. & Harris, R.A. (1994) Ethanol inhibits the memory task in humans. Vis. Neuroscience, 16, 449–459.
function of 5-hydroxytryptamine type-1C and muscarinic-M1, G-protein Treisman, A. (1996) The binding problem. Curr. Opin. Neurobiol., 6, 171–178.
ª 2005 Federation of European Neuroscience Societies, European Journal of Neuroscience, 21, 591–595

Source: http://www.bernhard-hommel.eu/CaffeineNicotine.pdf