Blue light, yellow light, and the role of parvalbumin-positive neurons in the pathophysiology of schizophrenia
Parvalbumin (PV)-positive cells are a prominent subtype of GABA neuron that via perisomatic synapses may strongly inhibit pyramidal cell activity (however, see Szabadics et al., 2006). In schizophrenia, PV neurons have reduced levels of mRNA for PV and for GAD67, the 67 kilodalton form of the GABA-synthesizing enzyme glutamate decarboxylase. The functional consequences of the PV reduction in schizophrenia are poorly understood, but one possibility is that decreased PV partially compensates for a deficit in GABA release caused by the GAD67 reduction. PV is a slow Ca2+ buffer, and so decreasing PV in nerve terminals may facilitate GABA release during repetitive PV cell firing (for a review, see Gonzalez-Burgos and Lewis, 2008).
Why is PV cell-mediated inhibition significant to brain function? What deficits in cortical circuit function may be compensated for (at least partially) by a decrease of PV in schizophrenia? The answers to these questions depend on our knowledge of the functional role of PV neurons in cortical circuits. A leading hypothesis in this regard suggests that PV cells are essential for the production of synchronized oscillations in cortex, particularly in the γ frequency band (Bartos et al., 2007). γ oscillations are thought to be important for the transmission of information within and between neocortical areas (Salinas and Sejnowski, 2001). If so, then γ activity must be important for cognition, which is critically dependent on the flow of information across the neocortex (Singer, 1999; Fuster, 2004). Therefore, cognitive deficits (a key feature of schizophrenia) may result from the impairment of γ oscillations reported in the illness, which in turn could be a consequence of deficits in PV cell-mediated inhibition.
How is PV cell-mediated inhibition linked to the production of γ oscillations? PV neurons alone are not sufficient to produce network oscillations because, although PV cell membranes resonate at γ frequency (Pike et al., 2000), PV neurons actually lack intrinsic pacemaker activity. Therefore, whereas PV cells may be necessary to produce γ oscillations, synaptic interactions with other cell types in the cortical circuit must be necessary as well.
Studying the link between PV neuron activity and γ oscillations is complicated by the lack of tools available to selectively manipulate PV cell activity. Interestingly, two recent studies (Sohal et al., 2009; Cardin et al., 2009) employed novel “optogenetic” methods in mice, to further advance our understanding of how PV neurons are involved in γ oscillations.
Using viral vectors to drive cell type-specific expression of recombinant genes, Sohal and colleagues, as well as Cardin et al., produced expression of microbial opsins (which are light-sensitive ion channel proteins) selectively in specific populations of cortical neurons. Briefly described, modulation of neuronal activity using such optogenetic techniques works as follows: in cells expressing channelrhodopsin-2 (ChR2), illumination with blue light produces a depolarizing current that has excitatory effects (increases cell firing). On the other hand, cells expressing halorhodopsin (eNpHR) respond to yellow light with a hyperpolarizing current that has inhibitory effects (decreases cell firing). In this way, light of different wavelengths can be used to either inhibit or excite PV cells or pyramidal (PYR) neurons in the mouse cortex (in vivo or in vitro) in different experimental designs.
Sohal and colleagues first demonstrated that inhibiting the activity of eNpHR-expressing PV neurons with yellow light suppresses γ oscillations generated in vivo by rhythmic flashes of blue light applied to stimulate nearby ChR2-expressing PYR neurons. Furthermore, they show that non-rhythmic excitation of PYR cells produces non-rhythmic PYR cell firing. However, if the non-rhythmic PYR spikes are used to trigger feedback inhibition by PV neurons (driven by blue light flashes that stimulate ChR2-expressing PV cells), the addition of feedback inhibition induces a γ rhythm in PYR cell output. These results show that by means of recurrent interactions with pyramidal cells (consistent with the so-called PING models of γ rhythms; Whittington et al., 2000), PV cell activity is crucial for γ oscillations.
Finally, the experiments performed by Sohal et al. suggest that γ activity may selectively enhance the flow of information in cortical circuits. For example, their experiments demonstrate that the gain of the neuronal input-output relation (that is, the slope of the curve describing the transformation of inputs into outputs) is specifically enhanced when excitatory inputs onto a PYR cell are modulated rhythmically at γ frequency. Moreover, the amount of information flowing across synapses during interactions between PYR and PV neurons was estimated using concepts derived from information theory. The estimates from Sohal et al. showed that when network activity was driven by trains of blue flashes delivered at γ frequency, information transmission was markedly enhanced.
Using somewhat different genetic engineering approaches, Cardin and colleagues produced cell type-specific expression of ChR2 in PV cells or PYR neurons. Expression of ChR2 in PV cells of the mouse barrel (somatosensory) cortex allowed the activation of PV neurons with rhythmic flashes of light. Such manipulation produced rhythmic population activity (as detected recording local field potentials) more strongly when PV cells were driven at γ frequency compared with other frequencies. They report, in addition, some data showing that manipulation of PV cell activity has an impact on γ oscillations intrinsically generated by the cortical circuits, as opposed to γ rhythms induced by rhythmic stimuli applied by the investigators. For example, brief flashes of blue light applied to stimulate firing of ChR2-expressing PV neurons during spontaneous γ activity were able to reset the phase of the γ rhythm.
Cardin et al. also demonstrate that activation of ChR2-expressing PV neurons can suppress the somatosensory response (to whisker stimulation) of nearby PYR cells. Then they go on to test an important functional role predicted for γ oscillations: that cells in a local network engaged in γ oscillations may respond differently to incoming inputs, depending on the timing of the incoming inputs relative to the phase of the γ cycle (Fries et al., 2007). In an elegant experiment, the investigators paired brief whisker stimulation with rhythmic flashes of blue light which, by stimulating ChR2-expressing PV cells, generate a γ rhythm locally. The crucial finding from this experiment is that the excitatory power of whisker stimulation was strongly dependent on the γ cycle phase at which whisker stimulation was delivered.
The data from these two recent studies briefly summarized above further consolidate the notion that PV-positive GABA neurons are key players in the mechanisms of γ synchrony in cortex. The data from these studies confirm that rhythmic PV neuron firing at γ frequency is sufficient to generate a γ rhythm in the population of postsynaptic neurons. This is not the same, however, as saying that PV neurons alone are sufficient to generate γ. Indeed, the data from these two studies point to the idea that PV neurons work in close interaction, via feedback loops, with nearby pyramidal cells. This makes sense, given that PV neurons are not intrinsic pacemakers. γ rhythms, therefore, seem to originate in complex network interactions that require the coordinated activity of pyramidal neurons, PV cells, and possibly other GABA neuron subtypes as well.
These two studies also highlight the similar importance of PV cell-dependent γ oscillations across different regions of cortex primarily involved in very different functions: Sohal and colleagues studied the role of PV neurons in frontal cortical areas, whereas Cardin et al. manipulated PV cell activity in the primary somatosensory cortex. The similarity of the findings regarding PV neuron function suggests that these neurons probably play a very similar role, at the microcircuit level, in these two very different cortical areas. Interestingly, deficits in GABA transmission, as assessed in postmortem brain studies, appear to be found in multiple areas of cortex simultaneously (Hashimoto et al., 2008). If this is indeed the case, then an impairment of γ oscillations may be present in most cortical areas. Thus, deficits in PV cell-dependent γ rhythms may explain the impairment of not only complex cognitive functions (for example, working memory), but also of more basic sensory processing which, as reviewed elsewhere (Javitt, 2009), is also impaired in schizophrenia. Finding a global deficit of GABA transmission and γ oscillations, as opposed to a deficit restricted to a single cortical area, increases the probability of success in developing pharmacological treatments.
A better understanding of the functional role of PV neurons in normal cortical circuits is crucial to developing better models that could explain how alterations of PV cells (and of other GABA neurons subtypes) originate in schizophrenia. As highlighted elsewhere (Lewis and Gonzalez-Burgos, 2008), any given alteration observed in schizophrenia may represent 1) cause, an upstream factor related to the disease pathogenesis; 2) consequence, a deleterious effect of a cause; 3) compensation, a response to either cause or consequence that helps restore homeostasis; or 4) confound, a product of factors frequently associated with, but not a part of, the disease process, or an artifact of the approach used to obtain the measure of interest. A major challenge for schizophrenia research is therefore determining to which of the four “C” categories each alteration belongs (Lewis and Gonzalez-Burgos, 2008). Certainly, information from basic neuroscience research studies such as those of Sohal et al. and Cardin et al. and many other past and future studies is extremely helpful in achieving this goal.
Szabadics J, Varga C, Molnar G, Olah S, Barzo P, Tamas G: Excitatory effect of GABAergic axo-axonic cells in cortical microcircuits. Science 311:233-235, 2006. Abstract
Gonzalez-Burgos G, Lewis DA: GABA Neurons and the Mechanisms of Network Oscillations: Implications for Understanding Cortical Dysfunction in Schizophrenia. Schizophr Bull 34:944-961, 2008. Abstract
Bartos M, Vida I, Jonas P: Synaptic mechanisms of synchronized gamma oscillations in inhibitory interneuron networks. Nat Rev Neurosci 8:45-56, 2007.Abstract
Salinas E, Sejnowski TJ: Correlated neuronal activity and the flow of neural information. Nat Rev Neurosci 2:539-550, 2001. Abstract
Singer W: Neuronal synchrony: a versatile code for the definition of relations? Neuron 24:111-25, 1999. Abstract
Fuster JM: Upper processing stages of the perception-action cycle. Trends Cogn Sci 8:143-145, 2004. Abstract
Pike FG, Goddard RS, Suckling JM, Ganter P, Kasthuri N, Paulsen O: Distinct frequency preferences of different types of rat hippocampal neurones in response to oscillatory input currents. J Physiol 529 Pt 1:205-213, 2000. Abstract
Sohal VS, Zhang F, Yizhar O, Deisseroth K: Parvalbumin neurons and gamma rhythms enhance cortical circuit performance. Nature 2009 Jun 4;459(7247):698-702. Abstract
Cardin JA, Carlen M, Meletis K, Knoblich U, Zhang F, Deisseroth K, Tsai LH, Moore CI: Driving fast-spiking cells induces gamma rhythm and controls sensory responses. Nature 2009 Jun 4;459(7247):663-7. Abstract
Whittington MA, Traub RD, Kopell N, Ermentrout B, Buhl EH: Inhibition-based rhythms: experimental and mathematical observations on network dynamics. Int J Psychophysiol 38:315-336, 2000. Abstract
Fries P, Nikolic D, Singer W: The gamma cycle. Trends Neurosci 30:309-316, 2007. Abstract
Hashimoto T, Bazmi HH, Mirnics K, Wu Q, Sampson AR, Lewis DA: Conserved Regional Patterns of GABA-Related Transcript Expression in the Neocortex of Subjects With Schizophrenia. American Journal of Psychiatry 162:479-489, 2008. Abstract
Javitt DC: When doors of perception close: bottom-up models of disrupted cognition in schizophrenia. Annu Rev Clin Psychol 5:249-275, 2009. Abstract
Lewis DA, Gonzalez-Burgos G: Neuroplasticity of neocortical circuits in schizophrenia. Neuropsychopharmacology 33:141-165, 2008. Abstract
View all comments by Guillermo Gonzalez-Burgos