Correction

NEUROSCIENCECorrection for “Cortically projecting basal forebrain parvalbu-min neurons regulate cortical gamma band oscillations,” by TaeKim, Stephen Thankachan, James T. McKenna, James M. McNally,Chun Yang, Jee Hyun Choi, Lichao Chen, Bernat Kocsis, KarlDeisseroth, Robert E. Strecker, Radhika Basheer, Ritchie E.Brown, and Robert W. McCarley, which appeared in issue 11,March 17, 2015, of Proc Natl Acad Sci USA (112:3535–3540; firstpublished March 2, 2015; 10.1073/pnas.1413625112).The authors note that the following statement should be added

to the Acknowledgments: “This work was supported by the De-partment of Veterans Affairs (VAMerit Awards to R.W.M., R.B.,and R.E.S., and VA Career Development Award to J.M.M.).”

www.pnas.org/cgi/doi/10.1073/pnas.1507465112

E2848 | PNAS | May 26, 2015 | vol. 112 | no. 21 www.pnas.org

Dow

nloa

ded

by g

uest

on

May

20,

202

1 D

ownl

oade

d by

gue

st o

n M

ay 2

0, 2

021

Dow

nloa

ded

by g

uest

on

May

20,

202

1 D

ownl

oade

d by

gue

st o

n M

ay 2

0, 2

021

Dow

nloa

ded

by g

uest

on

May

20,

202

1 D

ownl

oade

d by

gue

st o

n M

ay 2

0, 2

021

Dow

nloa

ded

by g

uest

on

May

20,

202

1 D

ownl

oade

d by

gue

st o

n M

ay 2

0, 2

021

Cortically projecting basal forebrain parvalbuminneurons regulate cortical gamma band oscillationsTae Kima,b,1,2, Stephen Thankachana,b,2, James T. McKennaa,b, James M. McNallya,b, Chun Yanga,b, Jee Hyun Choic,Lichao Chena,b, Bernat Kocsisb,d, Karl Deisserothe,f, Robert E. Streckera,b, Radhika Basheera,b, Ritchie E. Browna,b,3,and Robert W. McCarleya,b,3,4

aDepartment of Psychiatry, VA Boston Healthcare System, Brockton, MA 02301; bDepartment of Psychiatry, Harvard Medical School, Boston, MA 02115;cCenter for Neuroscience, Korea Institute of Science and Technology, 130-722 Seoul, South Korea; dDepartment of Psychiatry, Beth Israel Deaconess MedicalCenter, Boston, MA 02215; and Departments of ePsychiatry and Behavioral Sciences and fBioengineering, Stanford University, Stanford, CA 94305

Edited by Nancy Kopell, Boston University, Boston, MA, and approved February 2, 2015 (received for review July 17, 2014)

Cortical gamma band oscillations (GBO, 30–80 Hz, typically ∼40 Hz)are involved in higher cognitive functions such as feature binding,attention, and working memory. GBO abnormalities are a featureof several neuropsychiatric disorders associated with dysfunctionof cortical fast-spiking interneurons containing the calcium-bindingprotein parvalbumin (PV). GBO vary according to the state ofarousal, are modulated by attention, and are correlated with con-scious awareness. However, the subcortical cell types underlyingthe state-dependent control of GBO are not well understood. Herewe tested the role of one cell type in the wakefulness-promotingbasal forebrain (BF) region, cortically projecting GABAergic neuronscontaining PV, whose virally transduced fibers we found apposedcortical PV interneurons involved in generating GBO. Optogeneticstimulation of BF PV neurons in mice preferentially increased cor-tical GBO power by entraining a cortical oscillator with a resonantfrequency of ∼40 Hz, as revealed by analysis of both rhythmic andnonrhythmic BF PV stimulation. Selective saporin lesions of BF cho-linergic neurons did not alter the enhancement of cortical GBOpower induced by BF PV stimulation. Importantly, bilateral optoge-netic inhibition of BF PV neurons decreased the power of the 40-Hzauditory steady-state response, a read-out of the ability of thecortex to generate GBO used in clinical studies. Our results aresurprising and novel in indicating that this presumptively inhibitoryBF PV input controls cortical GBO, likely by synchronizing the activ-ity of cortical PV interneurons. BF PV neurons may represent a pre-viously unidentified therapeutic target to treat disorders involvingabnormal GBO, such as schizophrenia.

ArchT | arousal | auditory steady-state response | channelrhodopsin2 |optogenetics

Gamma band oscillations (GBO) (30–80 Hz) recorded in thecortical EEG have generated intense interest in recent

years. Theoretical and experimental work suggests that GBOplay a crucial role in feature binding (1), attention (2), andconsciousness (3). Furthermore, GBO abnormalities have beenreported in severe neuropsychiatric disorders, such as Alzheimer’sdisease (4) and schizophrenia (5, 6). Animal studies have shownthat GBO are generated by fast-spiking cortical interneuronscontaining the calcium-binding protein parvalbumin (PV) (7, 8),which postmortem clinical studies have shown to be abnormal inschizophrenia (9). Animal models of schizophrenia have shownthat alterations in the function of cortical PV interneurons arestrongly correlated with deficits in GBO and cognition (10, 11).Thus, understanding the regulation of cortical PV interneuronsand GBO is essential for developing treatments for schizophreniaand other disorders in which GBO are abnormal.The ability of the cortex to generate GBO fluctuates according

to the level of arousal and consciousness. GBO are reduced innon-rapid eye movement (NREM) sleep (12), deep anesthesia(13), and vegetative state (14). Thus, cortical GBO are regulatedby the ascending reticular activating system (ARAS), originallydescribed by Moruzzi and Magoun (15). Electrical stimulation

of the origin of the ARAS in the brainstem reticular formationelicited cortical low-voltage fast activity (15) and enhanced GBO(16), suggesting that increased activity in the ARAS is responsiblefor behavioral state-dependent changes in cortical GBO.Lesion experiments in rodents suggested that cortically pro-

jecting neurons in the basal forebrain (BF), the final node of theventral arm of the ARAS, may be particularly important incontrolling cortical activation and GBO (17–19). Although thesestudies indicated the importance of the BF, they did not revealwhich specific BF cell types are involved. Although selective BFcholinergic lesions cause a modest reduction in the amplitude ofcortical fast oscillations, they do not abolish them (17), sug-gesting that additional neuronal subtypes are involved.Among the BF noncholinergic neurons, GABAergic neurons

are of particular interest because a significant minority dischargesat high rates (20–60 Hz) in association with cortical activationduring wakefulness and REM sleep (20). Many cortically projec-ting BF GABAergic neurons contain the calcium-binding proteinPV, and a large majority of BF PV neurons are GABAergic (21,22). Recordings from identified BF PV neurons in anesthetizedanimals revealed they increase their discharge rate during corticalactivation induced by sensory stimulation (23). Furthermore,

Significance

When we are awake, purposeful thinking and behavior re-quire the synchronization of brain cells involved in differentaspects of the same task. Cerebral cortex electrical oscillationsin the gamma (30–80 Hz) range are particularly important insuch synchronization. In this report we identify a particularsubcortical cell type which has increased activity during wak-ing and is involved in activating the cerebral cortex and gen-erating gamma oscillations, enabling active cortical processing.Abnormalities of the brain mechanisms controlling gammaoscillations are involved in the disordered thinking typical ofneuropsychiatric disorders such as schizophrenia. Thus, thesefindings may pave the way for targeted therapies to treatschizophrenia and other disorders involving abnormal corticalgamma oscillations.

Author contributions: T.K., S.T., J.T.M., J.M.M., C.Y., R.E.S., R.B., R.E.B., and R.W.M. de-signed research; T.K., S.T., J.T.M., J.M.M., C.Y., L.C., and R.B. performed research; K.D.contributed new reagents/analytic tools; T.K., S.T., J.T.M., J.M.M., C.Y., J.H.C., B.K., R.E.S.,R.B., R.E.B., and R.W.M. analyzed data; and T.K., S.T., J.T.M., J.M.M., C.Y., B.K., R.E.S., R.B.,R.E.B., and R.W.M. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

Freely available online through the PNAS open access option.1Present address: Department of Psychiatry, Kyung Hee University Hospital at Gangdong,134-727 Seoul, Korea.

2T.K. and S.T. contributed equally to this work.3R.E.B. and R.W.M. contributed equally to this work.4To whom correspondence should be addressed. Email: robert_mccarley@hms.harvard.edu.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1413625112/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1413625112 PNAS | March 17, 2015 | vol. 112 | no. 11 | 3535–3540

NEU

ROSC

IENCE

reduced cortical activation following BF lesions was correlatedwith the extent of PV neuronal loss (17). Interestingly, BFGABAergic projections to cortical GABAergic interneuronscontaining PV (24, 25) have been demonstrated, although directBF PV→cortical PV connections remain to be shown. Whenconsidered together, the physiology and anatomical connections ofBF GABAergic/PV neurons suggest that they are ideally posi-tioned to control cortical GBO (Fig. 1A). However, this hypothesishas not been tested directly. Thus, here we tested the effect ofselective excitation of BF PV neurons on the cortical EEG and theeffect of selective inhibition on the 40-Hz auditory steady-stateresponse (ASSR), a test of the ability of the cortex to generateGBO (26) widely used in clinical studies of anesthesia (27),hearing loss (28), and schizophrenia (5, 29).

ResultsOptogenetic Excitation of BF PV Neurons Preferentially EnhancesCortical Power at 40 Hz. To determine if optical stimulation of BFPV neurons affects cortical GBO activity (Fig. 1B), mice consti-tutively expressing Cre-recombinase in PV neurons (PV-Cre mice)

received unilateral BF injections of double-floxed adeno-associatedviral (AAV) vectors expressing channelrhodopsin2 (ChR2) andenhanced YFP (EYFP) fusion protein (7, 8). Subsequently,transduced BF PV neurons were stimulated in vivo at 40 Hz witha 5-s train of 10-ms light pulses (Fig. 1 C, E, and G and Fig. S1).Control mice received the same optical stimulation but withoutprior viral injection (Fig. 1 D, F, and H). In each mouse thisoptical stimulation protocol was repeated 20 times, and the EEGtraces were averaged; a grand average for ChR2-expressing mice(n = 5) and control mice (n = 3) is presented in Fig. 1. Strikingly,in ChR2-expressing mice, time–frequency analysis (Fig. 1C) andpower spectra (Fig. 1G) of averaged EEG traces (Fig. 1E) allshowed a pronounced increase in 40-Hz power during opticalstimulation. In contrast, controls did not show this response (Fig.1 D, F, and H).These results showed that 40-Hz stimulation of BF PV neu-

rons strongly modulates the cortical EEG at 40 Hz. However, itwas not clear from these experiments whether BF PV neuronshave a specific role in controlling GBO or simply drive corticalactivity at the stimulation frequency. Thus, we next examined ifthere were preferred BF stimulation frequencies for eliciting acortical response. The BF was optically stimulated at seven fre-quencies within the physiological firing range of BF GABAergicneurons (2, 10, 20, 30, 40, 50, and 60 Hz) (20) in time-of-day–matched trials in which each stimulation frequency was given 20total times in a pseudorandom order for each mouse, indepen-dent of ongoing behavioral state. For each mouse and eachfrequency the EEG was averaged, and the ratio of the power ateach frequency ±2 Hz during the 5-s stimulation epoch to the 5-sprestimulation epoch was calculated. Fig. 2A shows a grand av-erage of the five animals for the ratio of power during stimulationto the prestimulation power at the same frequency (geometricmean and SEM). A one-way repeated-measures ANOVA showedsignificant differences among the frequencies (F = 6.498, df = 6,P < 0.001). Compatible with the visual impression in Fig. 2A, thepower of stimulation frequencies at 40 Hz or the frequenciesclosest to it (30 and 50 Hz) was significantly different from pre-stimulation baseline (paired t test, P < 0.05), but the power of otherfrequencies did not differ significantly from prestimulation baseline.The preferential enhancement of EEG power at and near

40 Hz in our experiments suggests that BF PV stimulation wasnot simply driving cortical EEG at any frequency but instead wasentraining a cortical oscillator with a resonant frequency of ∼40 Hz.Four additional observations supported this interpretation. First, theEEG power response at a stimulation frequency of 40 Hz increasedgradually over time during the course of stimulation, indicative ofentrainment of the oscillation (Fig. 2B; r = 0.675, P < 0.001, linear fit;shown is the average of data from five animals, with each dot rep-resenting power at 40 ± 5 Hz for successive 200-ms epochs). A sta-tistically significant increase was observed only at 40 Hz (Fig. S2).Second, the relative power of the harmonic response was signifi-cantly higher with 20-Hz stimulation than with 30-Hz stimulation(n = 5; t = 3.445, df = 8, P = 0.009, unpaired t test) (Fig. 2 C,D, andE), as would be expected if the oscillator is tuned at 40 Hz. Third,the oscillation in response to 40-Hz stimulation developed overtime and persisted for four or more 40-Hz cycles following cessationof stimulation (Fig. 2F), another feature expected of an oscillator.Stimulation at other frequencies, e.g., 10 Hz, also induced 40-Hzoscillations that persisted at the end of the stimulation and lastedlonger than those at the stimulation frequency (Fig. S3), as isconsistent with 40 Hz being the resonant frequency. Finally, in aseparate group of mice (n = 3), nonrhythmic stimulation of BF PVin a frequency band spanning the natural firing frequencies of BFGABA neurons during wakefulness and REM sleep (20–60 Hz)(20) preferentially enhanced cortical power at 40 Hz (Fig. S4).

Validation of Selective Optogenetic Excitation of BF PV Neurons.Several lines of evidence supported our conclusion that selec-tive optogenetic stimulation of BF PV neurons was responsiblefor the entrainment of cortical gamma oscillations. First, in micewith a high GBO response (>4-fold power increase relative to

Fig. 1. BF PV neurons control cortical GBO, typically at ∼40 Hz. (A) Theexperimental model tested in this study. Cortical GBO are generatedthrough the interaction of cortical GABAergic PV interneurons and pyra-midal neurons (PYR). BF PV projection neurons control cortical GBO bysynchronizing the activity of cortical PV interneurons. Red and green linesshow inhibitory and excitatory connections, respectively. (B) Schematic ofthe experiment. The frontal cortex EEG (red) is recorded while opticalstimulation is delivered to the BF through an optical fiber (blue). Modifiedwith permission from ref. 52. (C, E, and G) Response in mice expressing ChR2in PV neurons (n = 5). (D, F, and H) Response in mice without viral injection(n = 3). (C and D) Time–frequency spectrograms of grand averaged EEGrecordings before, during, and after stimulation (stim). Increased power at40 Hz during optical stimulation (between the white lines) was seen only inChR2-expressing mice. (E and F) EEG traces grand-averaged over 20 trialswith and without high-pass filtering. (Scale bars: 10 μV, 100 ms.) (G and H)Power spectra of 5-s prestimulation (G) and 5-s stimulation (H) periods reveala prominent increase at 40 Hz in the ChR2 group.

3536 | www.pnas.org/cgi/doi/10.1073/pnas.1413625112 Kim et al.

baseline), the guide cannula tip was well targeted to the BF (Fig.3A). BF site specificity was indicated by the lack of a change inEEG power when the injection site was located outside the BF(n = 4). Furthermore, there was no response to optical stimulationin mice 1 wk after viral injection (before robust viral expressionbegan); however, subsequent stimulation of the same animalsand same sites 2 wk or more after injection showed a robustincrease in EEG power at 40 Hz. As in previous studies (30),injections of a control virus, which expressed only EYFP, con-firmed that EYFP was localized selectively to PV-expressingneurons (Fig. 3 D–F). Of the EYFP-transduced cells, 76.2 ±2.7% stained positively for PV (n = 5). In the remaining EYFPneurons, PV staining was close to background, possibly becauseof difficulties in antibody penetration in the fiber-dense BF re-gion and low levels of PV antigen in the cell bodies of BF neu-rons. Previous anterograde tracings of GABAergic projectionsfrom BF have shown that they target cortical PV interneurons(24, 25). However, direct BF PV→cortical PV connections hadnot been shown. In our study, dense cortical ChR2-EYFP–labeledfibers were observed following transduction of BF PV neurons,especially in superficial cortical layers II–III and in layer V (Fig.3G). Confocal images showed varicosities of these fibers locatedclose to cortical PV interneurons (Fig. 3H) with appositionssuggestive of synaptic contacts (Fig. 3I).To confirm that optical stimulation activated BF PV neurons,

we performed in vivo immunohistochemical and in vitro elec-trophysiology experiments (next section). In our immunohisto-chemical experiments, we performed 2 h of 40-Hz stimulation(5-s optical stimulation at 40 Hz every 60 s for 2 h) and theneuthanized the animals (n = 3) for immunohistochemical stain-ing of the protein product (cFos) of the immediate early genecFos, a marker of neuronal activation. In animals receiving op-tical stimulation, cFos was observed within the nuclei of virallytransduced neurons expressing ChR2-EYFP (Fig. 3C). In vivo, werecorded unit activity close to the optical fiber in virally transduced

mice. Consistent short-latency (<1 ms) action potential firing wasobserved, indicating the direct excitation of PV neurons by opticalstimulation (viz., the two neurons shown in Fig. S5A). Furthermore,these optically identified BF PV neurons discharged at higher rates,including at GBO frequencies, during wakefulness and REM sleep(Fig. S5B).Our conclusion that optical stimulation of BF PV neurons

shows selectivity for 40 Hz assumes that optical stimulation re-liably elicits action potentials at all the stimulation frequencies.To confirm this assumption, we made cell-attached recordingsfrom BF PV neurons expressing ChR2-EYFP in vitro. BF brainslices were prepared either from PV-Cre mice 2 wk followingviral injection or from mice generated by crossing PV-Cre micewith a Cre-reporter strain expressing ChR2-EYFP (31). Resultswere identical and have been pooled. Optical stimulation of PVneurons using the stimulation parameters used in vivo revealedthe neurons reliably followed the optical stimulation up to andincluding 60 Hz (n = 3) (Fig. S6B). Each 10-ms light pulse (re-gardless of frequency) elicited two to four action potentials, andthe number of action potentials per light pulse did not vary

Fig. 2. Optogenetic stimulation of BF PV neurons preferentially entrains a40-Hz endogenous cortical oscillator. (A) Stimulation at or near 40 Hz elicitsthe largest power increase; n = 5, *P < 0.05. (B) EEG power near 40 Hzincreases gradually during the stimulation period. The Inset shows linear fit;n = 5. (C and D) Grand-average spectrograms of responses to stimulation at20 (C) and 30 (D) Hz show a harmonic response at twice the stimulationfrequency; n = 5. (E) Relative power of the harmonic response is significantlyhigher during 20-Hz stimulation (40-Hz harmonic) than during 30-Hz stim-ulation (60-Hz harmonic). **P < 0.01. (F) The 40-Hz response develops overtime and outlasts the optical stimulation period (500 ms, between verticalsolid bars, n = 1). (Scale bars: 2 μV, 50 ms.)

A

B C

D E F

G H I

Fig. 3. BF PV neurons selectively transduced with ChR2-EYFP project ontocortical PV interneurons. (A) Stimulation sites in the BF (stars), depicted onone coronal section [anterior–posterior (AP) −0.10; actual AP coordinates0.26 to −0.34)]. Aca, anterior commissure, anterior part; acp, anterior com-missure, posterior part; CPu, caudate putamen, HDB, horizontal limb of thediagonal band; LPO, lateral preoptic area; MCPO, magnocellular preopticnucleus; SI, substantia innominata; VP, ventral pallidum. Modified withpermission from ref. 52. (B) Virally transduced (EYFP, green) BF cells andfibers for one case (red star in A). (C) cFos nuclear labeling (red) indicatedneuronal activation in optically stimulated ChR2-EYFP–expressing BF PV neu-rons (green). (D–F) Anti-PV immunohistochemistry shows AAV-EYFP controlinjections selectively transduce PV somata in the BF. (G–I) Anti-GFP–stained BFPV fibers transduced with ChR2-EYFP (green) innervated layers II–III and V of thesomatosensory cortex, and fiber varicosities formed appositions onto cortical PVinterneurons (red). (G) Low-magnification image. (H) High-magnification (100×)confocal immunofluorescence Z-stack image (36 optical sections, 1-μm thick-ness) of the boxed area in G. Arrowheads indicate colocalization of BF PV fibersand cortical PV interneurons, suggesting apposition. (I) High-magnification ofthe boxed area in H showing one 1-μm optical section. Apposition of the BF PVfiber to the cortical PV interneuron, suggestive of a synaptic contact, wasconfirmed by the presence of a yellow (merge of red and green)-labeled axonalvaricosity (arrowheads) on the PV cell in the orthogonal planes of the x–z(Upper) and y–z (Right) projection images. (Scale bars: 1 mm in A, 500 μm in B,50 μm in C, 100 μm in F and G, and 10 μm in H and I.)

Kim et al. PNAS | March 17, 2015 | vol. 112 | no. 11 | 3537

NEU

ROSC

IENCE

significantly as the stimulation frequency was increased (Fig. S6C).Thus, the gradually diminishing cortical EEG response in re-sponse to stimulation at frequencies above 40 Hz is not caused byan inability of BF PV neurons to follow higher frequency stimu-lation. Our data also raise the question of whether the BF itselfhas a preferred resonant frequency of 40 Hz or whether the GBOresults from a 40-Hz cortical oscillator that is entrained by BFinput. Local field potential data recorded ipsilateral to the BFstimulation site support the latter possibility, because they did notshow a preferred 40-Hz response in the BF (n = 3) (Fig. S6 D–F).

Cholinergic Neurons Are Not Required for the Cortical 40-Hz ResponseMediated by Stimulation of BF PV Neurons. Our immunohisto-chemical experiments revealed a dense plexus of ChR2-EYFP–labeled fibers in the BF (Fig. 3). Thus, effects of BF PV stimu-lation could be mediated through interaction with other BFneurons, in particular, cholinergic neurons (32). To determineif BF cholinergic neurons are required for increased corticalGBO produced by BF PV stimulation, we bilaterally injectedthe selective cholinergic toxin murine-p75NTR-saporin (mu p75-saporin) (33) into the lateral ventricles and tested the corticalresponse 3–4 wk after saporin-induced destruction of BF cho-linergic neurons. Successful lesioning of cholinergic neurons wasconfirmed by immunohistochemistry for the selective markerof cholinergic neurons, choline acetyltransferase (ChAT) (Fig.S7A). Cell counting indicated an extensive (70.2 ± 2.1%, n = 4)lesion of BF cholinergic neurons, a percentage consistent withthe literature reports using mu p75-saporin (33). These extensivelesions did not impair the ability of BF PV optogenetic stimu-lation to facilitate cortical GBO (control, 15.49 ± 6.43 fold-increase in power at 40 Hz; saporin-lesioned, 12.06 ± 1.95 in-crease, n = 6; n.s., Mann–Whitney u test) (Fig. S7B).

BF PV Neuronal Control of Cortical GBO Is Likely Mediated by DirectCortical Projections. In addition to the cortical fiber labeling ob-served following transduction of BF PV neurons (Fig. 3 G–I),labeled fibers also were seen densely innervating the thalamicreticular nucleus (TRN), a known projection site of BF PV neu-rons (34). The strong projection of BF PV neurons to the TRNsuggested that this could also be a pathway that modulates corticalGBO via entrainment of thalamic relay neurons. Therefore, wetested whether optogenetic stimulation of TRN PV neurons wouldmimic the effect of stimulation of BF PV neurons. However, incontrast to the effect of stimulation of BF PV neurons, bilateralstimulation of ChR2-transduced TRN PV neurons preferentiallyenhanced cortical power at 10 Hz, consistent with the known roleof TRN neurons in the generation of spindle (8–14 Hz) activityduring NREM sleep (n = 4) (Fig. S8). Thus, the most parsimo-nious explanation for the BF PV control of cortical GBO is viatheir direct projections to cortical PV interneurons.

Bilateral Inhibition of BF PV Neurons Impaired the 40-Hz AuditorySteady-State Cortical Response. To assess further a physiologicalrole of BF PV neurons in the control of cortical GBO, we testedthe effect of BF PV inhibition on the ASSR, a paradigm com-monly used to test GBO in clinical studies of schizophrenia (5)and anesthesia (27). In this paradigm, presentation of a trainof auditory clicks elicits a steady-state oscillation in the EEGrecorded above auditory and frontal cortices (5, 26). In ourASSR paradigm, mice were exposed to 1-s trains of auditorystimuli at 40 Hz with parameters similar to those used in ASSRstudies of schizophrenia patients (5). To inhibit BF PV neurons,AAV vectors with Cre-dependent expression of the inhibitoryopsin, Archaerhodopsin from Halorubrum strain TP009 (ArchT)(35), and a fluorescent marker of expression (AAV-ArchT-GFP)were bilaterally injected into the BF of PV-Cre mice (n = 8) (Fig.4A and Fig. S9A).Under control conditions each 1-s 40-Hz train elicited an

auditory evoked potential, associated with a pronounced low-frequency response recorded in the frontal EEG electrode, fol-lowed by a sustained steady-state 40-Hz response (Fig. 4B). In

the same animals the mean power at 40 ± 2 Hz was calculatedduring the 1-s click train with and without bilateral ArchTinhibition of BF PV neurons. For the ArchT trials, laser light(532 nm) was applied continuously, beginning 2 s before the firsttrial and continuing throughout the block of trials. When BF PVneurons were bilaterally inhibited by ArchT stimulation, thepower of the 40-Hz EEG response recorded above frontal cortexwas reduced by 33.1 ± 4.4% (ASSR alone, 0.52 ± 0.10 μV2;ASSR + ArchT, 0.34 ± 0.06 μV2; P < 0.05, n = 8, Wilcoxonsigned-rank test) (Fig. 4 B–D), indicating that inhibition of BFPV neurons reduced the ability of the cortex to generate GBO inresponse to a sensory stimulus. Post hoc histological examinationof BF slices from AAV-ArchT-GFP–injected animals (n = 8)(Fig. S9A) revealed an extensive plexus of transduced neurons/fibers in the BF. As with injections of AAV-ChR2-EYFP (Fig. 3G–I), extensive cortical fiber labeling was observed, particularlyin layers II–III and V (Fig. S9B). In vitro whole-cell patch-clamprecordings confirmed that optical activation of ArchT caused astrong hyperpolarization and inhibited BF PV neuronal dis-charge (Fig. S9 C–E).

DiscussionHere we found that selective optogenetic excitation of BF PVneurons preferentially enhanced EEG power in the gammarange. Both rhythmic and nonrhythmic stimulation of BF PVneurons led to entrainment of a frontal cortex oscillator tuned at∼40 Hz. Conversely, optogenetic inhibition of BF PV neuronsreduced the power of the 40-Hz ASSR. Together, these resultssuggest a role for BF PV projection neurons in the state-dependentcontrol of cortical GBO activity, most likely mediated by a directprojection of BF PV neurons to cortical PV interneurons.Previous anatomical tracing and unit recording studies (cited

in the Introduction) had hinted at a possible role for BF GABA/PV neurons in control of cortical GBO. Until now, however, this

Fig. 4. Bilateral optogenetic inhibition of BF neurons with ArchT reducesthe power of the 40-Hz ASSR. (A) Location of the center of transductionwithin the BF for all eight mice, depicted on one side of one coronal section(AP −0.10; actual AP coordinates 0.26 to −0.34). The red star indicates themouse whose response is illustrated in B and C. Modified with permissionfrom ref. 52. (Scale bar: 1 mm.) (B) Representative time–frequency spectro-gram showing that inhibition of BF PV neurons with ArchT leads to a re-duction of the power of the 40-Hz ASSR. (C) Power spectrum of the responseshown in B. (D) Mean data for each of eight animals and grand means forthe ASSR 40-Hz cortical response without and with ArchT. Solid symbolsindicate the case illustrated in B and C.

3538 | www.pnas.org/cgi/doi/10.1073/pnas.1413625112 Kim et al.

idea had not been tested using selective stimulation and inhibition.In our first set of experiments we showed that selective opto-genetic stimulation of BF PV neurons preferentially increasedcortical EEG power at 40 Hz. Our validation experiments con-firmed that viral injections localized to the BF were required toelicit cortical GBO responses and that viral transduction wasselective for BF PV neurons. We also confirmed that opticalstimulation activated BF PV neurons and that PV neurons couldfollow the optical stimulation at the frequencies tested. Onepossible caveat is that optogenetic stimulation may cause anunphysiologically synchronous activation of the target neuronalpopulation. However, our recent in vitro data (21) suggest thatBF GABAergic/PV neurons are electrically coupled; thus ournearly synchronous optogenetic stimulation of a subset of theseneurons may parallel synchronous firing of a subset of these neu-rons in vivo. In fact, one group of neurochemically unidentifiednoncholinergic BF neurons recorded in vivo exhibited synchronousfiring in response to salient stimuli, correlated with a local fieldpotential recorded in the frontal cortex (36).

Evidence for BF PV Entrainment of a Cortical Gamma Band Oscillator.The preferential enhancement of EEG power at and near 40 Hzby optogenetic stimulation of BF PV neurons suggests theincrease in cortical EEG power was caused not by a passiveresponse to optogenetic BF stimulation but rather by an en-trainment of a cortical oscillator with a resonant frequency of∼40 Hz. Similar conclusions regarding the existence of a 40-Hzcortical oscillator were reached by Gray and Singer (1) in theirstudy of GBO in visual cortex and Cardin and colleagues (8)using optical stimulation of fast-spiking cortical PV interneurons.In our study, four additional observations supported this in-terpretation. First, the EEG power response at a stimulationfrequency of 40 Hz increased gradually over time during thecourse of stimulation, indicative of entrainment of the oscilla-tion. Secondly, the relative power of the harmonic response wassignificantly higher with 20-Hz stimulation than with 30-Hzstimulation. Third, examination of the oscillation in response to40-Hz stimulation at a millisecond timescale revealed the os-cillation both increased in time and persisted for several cyclesfollowing cessation of stimulation. Finally, nonrhythmicstimulation at frequencies (20–60 Hz) approximating the nat-ural firing frequencies of BF GABA neurons during wakefulness/REM sleep (20) significantly enhanced cortical power only at40 Hz. Recent studies using transcranial magnetic stimulation inhumans have shown that the frontal cortex has an intrinsic res-onance frequency in the gamma range (37), which is reduced inschizophrenic patients (38). As shown by our anatomical data(Fig. 3 and Fig. S9), BF GABA/PV neurons may have a privi-leged access to this cortical oscillator through their projections tocortical PV interneurons.ChR2-EYFP and ArchT-GFP are targeted specifically to the

membrane and distribute throughout the neuron, allowing theanalysis of axonal projections. Previous anterograde tracingexperiments in rats showed that GABAergic BF neurons projectto the neocortex and preferentially target cortical interneuronscontaining PV and somatostatin (24, 25). Here, in the mouse,fibers of BF PV neurons, labeled with ChR2-EYFP or ArchT-GFP, innervated both deep and superficial layers of the neo-cortex, and varicosities apposed cortical PV interneurons, sug-gestive of a direct BF PV→cortical PV synaptic connection. Inaddition, there were strong projections to the TRN. However,stimulation of TRN PV neurons preferentially enhanced corticalEEG power at 10 Hz. Thus, the direct cortical projection is amore likely mediator of BF PV control of cortical GBO.Previous studies suggested a prominent role for BF cholinergic

neurons in the control of cortical activation (32, 39, 40). How-ever, noncholinergic BF neurons are likely to be equally or moreimportant. Selective lesions of cholinergic neurons using IgG192-saporin have relatively minor effects on the cortical EEG (17,19). In contrast, ibotenate lesions of BF which preferentiallytarget noncholinergic neurons cause more profound reductions

in cortical low-voltage fast activity than cholinergic lesions alone(17, 19). Here, we found that saporin lesions of BF cholinergicneurons, resulting in a 70% loss of this cell type, did not signifi-cantly affect the ability of BF PV stimulation to induce a cortical40-Hz response, suggesting that cholinergic neurons are notrequired. However, under normal conditions, cholinergic andGABA/PV neurons likely work together to control cortical acti-vation and GBO activity because cholinergic neurons stronglyexcite identified BF GABA and PV neurons in vitro (41).Further strong evidence supporting our conclusion that BF PV

neurons control cortical GBO came from our loss-of-functionexperiment using ArchT to inhibit BF PV neurons bilaterally.These experiments showed a dramatic reduction in the 40-HzASSR. Validation experiments confirmed that ArchT transductionwas targeted correctly to the BF and that light activation of ArchTinhibited BF PV neurons. Cortical fiber labeling was observedfollowing ArchT-GFP transduction, similar to our findings withChR2-EYFP, suggesting a direct BF PV→cortical PV projectionas a likely mediator of the effect.

Functional and Clinical Implications of BF PV Modulation of CorticalGBO. Spontaneously occurring gamma oscillations vary accordingto behavioral state, being higher during wakefulness than duringNREM sleep (12) and reduced in anesthesia (13). Similarly, theASSR power is reduced by about 50% during sleep (28) and isabolished by deep anesthesia (42). Although other parts of theARAS, such as the thalamus (3), likely are important, our resultssuggest that BF PV neurons are involved in state-dependentcontrol of GBO. Juxtacellularly identified BF PV neurons inanesthetized rats increased their firing rate with cortical activa-tion caused by tail pinch (23). Our preliminary recordings fromtwo optically identified BF PV neurons in mice confirmed thattheir discharge is highest in states that show more GBO activity,i.e., wakefulness and REM sleep. Furthermore, our results showedthat stimulation of BF PV neurons increases GBO, and inhibitionreduces the power of the ASSR. GBO are strongly reduced andabnormal in coma and vegetative states (14). Increasing the ac-tivity of BF PV neurons may be an interesting strategy to improvecortical function in traumatic brain injury, because recent brainimaging studies in traumatic coma revealed a complete disruptionof the axonal pathways connecting brainstem arousal centers tothe BF, whereas BF axonal projections to the cerebral cortex arepartially preserved (43).The BF also plays a critical role in sleep homeostasis (39).

Adenosine levels rise in the BF in association with prolongedwakefulness (44). In vitro, the glutamatergic inputs to BF PVneurons are inhibited by adenosine (45). Thus, adenosinergic in-hibition of BF PV neurons may contribute to the reduced corticalGBO associated with drowsiness (46).The BF is one of the earliest brain areas to be affected in

Alzheimer’s disease, and abnormalities of cortical oscillationsare a feature of this disease (4, 47). Atrophy of BF cholinergicneurons has been widely replicated both in this disease and innormal aging (48). Cholinergic neurons normally excite BFGABA/PV neurons (41), and loss of BF GABA/PV neurons hasbeen noted in aged mice which have accumulation of extracellularamyloid plaques and episodic-like memory impairments (49).Thus, reduced activity caused by the loss of cholinergic neuronsor the direct loss of BF GABA/PV neurons may be partially re-sponsible for cortical oscillation abnormalities and associatedcognitive impairments in aging and Alzheimer’s disease.With respect to schizophrenia, one of the most widely repli-

cated findings is a reduced power of the 40-Hz ASSR (5, 29).Cortical PV neurons show reductions in GAD67 and PV ex-pression in schizophrenia (11). BF PV neurons are derived fromthe same developmental pathway as cortical PV interneurons(50) and thus may be affected also. Unfortunately, to date, nopostmortem study has examined BF PV neurons in schizophrenia.Our findings suggest that modulating the activity of BF PVneurons may be an effective strategy to restore abnormal GBOin schizophrenia.

Kim et al. PNAS | March 17, 2015 | vol. 112 | no. 11 | 3539

NEU

ROSC

IENCE

ConclusionsOur results showing that BF PV neurons regulate and entraincortical GBO have important implications for our understandingof the physiological and pathophysiological control of arousal.We have identified, for the first time to our knowledge, a specificcell type (PV neurons) important in the BF control of corticalGBO. Because most BF PV neurons are GABAergic (21, 22), ourresults are surprising in indicating that this presumptively inhibitoryinput controls cortical activation, likely by synchronizing the activityof cortical inhibitory neurons. Furthermore, we identified a fre-quency-specific cortical response to a subcortical stimulation, sug-gesting that the cortex contains an endogenous oscillator tunedpreferentially to GBO frequencies. Our results identify BF PVneurons as a possible therapeutic target to treat disorders involvingabnormal GBO, such as schizophrenia (51), Alzheimer’s disease (4,49), and vegetative state (14).

MethodsPV-Cre mice (strain 008069; Jackson Laboratories) received BF injections ofAAV-ChR2-EYFP or AAV-ArchT-GFP followed by implantation of opticalfibers. Cortical EEG recordings, immunohistochemistry, and in vitro record-ings were performed as described previously (21, 41, 44). Details of the ex-perimental procedures are provided in SI Materials and Methods. Allexperiments conformed to US Veterans Administration, Harvard University,and US National Institutes of Health guidelines. All experimental procedureswere approved by the Institutional Animal Care and Use Committee of theVA Boston Healthcare System.

ACKNOWLEDGMENTS. This work was supported by Department of VeteransAffairs Awards (to R.W.M., R.B., and R.E.S.); by National Institutes of MentalHealth Grants R01 MH039683 (to R.W.M.), R21 MH094803 (to R.E.B.), and R01MH100820 (to B.K.); National Institute of Neurological Disorders and StrokeGrants R21 NS079866-01 (to R.B.) and P01 HL095491 (to R.W.M. and B.K.); andGlobal Frontier Grant 2011-0031525 (to J.H.C.).

1. Gray CM, Singer W (1989) Stimulus-specific neuronal oscillations in orientation col-umns of cat visual cortex. Proc Natl Acad Sci USA 86(5):1698–1702.

2. Tiitinen H, et al. (1993) Selective attention enhances the auditory 40-Hz transientresponse in humans. Nature 364(6432):59–60.

3. Llinás R, Ribary U (1993) Coherent 40-Hz oscillation characterizes dream state in hu-mans. Proc Natl Acad Sci USA 90(5):2078–2081.

4. Herrmann CS, Demiralp T (2005) Human EEG gamma oscillations in neuropsychiatricdisorders. Clin Neurophysiol 116(12):2719–2733.

5. Kwon JS, et al. (1999) Gamma frequency-range abnormalities to auditory stimulationin schizophrenia. Arch Gen Psychiatry 56(11):1001–1005.

6. Spencer KM, et al. (2004) Neural synchrony indexes disordered perception and cog-nition in schizophrenia. Proc Natl Acad Sci USA 101(49):17288–17293.

7. Sohal VS, Zhang F, Yizhar O, Deisseroth K (2009) Parvalbumin neurons and gammarhythms enhance cortical circuit performance. Nature 459(7247):698–702.

8. Cardin JA, et al. (2009) Driving fast-spiking cells induces gamma rhythm and controlssensory responses. Nature 459(7247):663–667.

9. Lewis DA, Hashimoto T, Volk DW (2005) Cortical inhibitory neurons and schizophre-nia. Nat Rev Neurosci 6(4):312–324.

10. Carlén M, et al. (2012) A critical role for NMDA receptors in parvalbumin interneuronsfor gamma rhythm induction and behavior. Mol Psychiatry 17(5):537–548.

11. McNally JM, McCarley RW, Brown RE (2013) Impaired GABAergic neurotransmissionin schizophrenia underlies impairments in cortical gamma band oscillations. CurrPsychiatry Rep 15(3):346.

12. Maloney KJ, Cape EG, Gotman J, Jones BE (1997) High-frequency gamma electroen-cephalogram activity in association with sleep-wake states and spontaneous behav-iors in the rat. Neuroscience 76(2):541–555.

13. John ER, et al. (2001) Invariant reversible QEEG effects of anesthetics. Conscious Cogn10(2):165–183.

14. Schiff ND, et al. (2002) Residual cerebral activity and behavioural fragments can re-main in the persistently vegetative brain. Brain 125(Pt 6):1210–1234.

15. Moruzzi G, Magoun HW (1949) Brain stem reticular formation and activation of theEEG. Electroencephalogr Clin Neurophysiol 1(4):455–473.

16. Munk MH, Roelfsema PR, König P, Engel AK, Singer W (1996) Role of reticular acti-vation in the modulation of intracortical synchronization. Science 272(5259):271–274.

17. Kaur S, Junek A, Black MA, Semba K (2008) Effects of ibotenate and 192IgG-saporinlesions of the nucleus basalis magnocellularis/substantia innominata on spontaneoussleep and wake states and on recovery sleep after sleep deprivation in rats. J Neurosci28(2):491–504.

18. Buzsaki G, et al. (1988) Nucleus basalis and thalamic control of neocortical activity inthe freely moving rat. J Neurosci 8(11):4007–4026.

19. Fuller PM, Sherman D, Pedersen NP, Saper CB, Lu J (2011) Reassessment of thestructural basis of the ascending arousal system. J Comp Neurol 519(5):933–956.

20. Hassani OK, Lee MG, Henny P, Jones BE (2009) Discharge profiles of identifiedGABAergic in comparison to cholinergic and putative glutamatergic basal forebrainneurons across the sleep-wake cycle. J Neurosci 29(38):11828–11840.

21. McKenna JT, et al. (2013) Distribution and intrinsic membrane properties of basalforebrain GABAergic and parvalbumin neurons in the mouse. J Comp Neurol 521(6):1225–1250.

22. Gritti I, Manns ID, Mainville L, Jones BE (2003) Parvalbumin, calbindin, or calretinin incortically projecting and GABAergic, cholinergic, or glutamatergic basal forebrainneurons of the rat. J Comp Neurol 458(1):11–31.

23. Duque A, Balatoni B, Detari L, Zaborszky L (2000) EEG correlation of the dischargeproperties of identified neurons in the basal forebrain. J Neurophysiol 84(3):1627–1635.

24. Freund TF, Meskenaite V (1992) gamma-Aminobutyric acid-containing basal forebrainneurons innervate inhibitory interneurons in the neocortex. Proc Natl Acad Sci USA89(2):738–742.

25. Henny P, Jones BE (2008) Projections from basal forebrain to prefrontal cortex com-prise cholinergic, GABAergic and glutamatergic inputs to pyramidal cells or inter-neurons. Eur J Neurosci 27(3):654–670.

26. Galambos R, Makeig S, Talmachoff PJ (1981) A 40-Hz auditory potential recordedfrom the human scalp. Proc Natl Acad Sci USA 78(4):2643–2647.

27. Plourde G (1999) Auditory evoked potentials and 40-Hz oscillations. Anesthesiology91(5):1187–1189.

28. Linden RD, Campbell KB, Hamel G, Picton TW (1985) Human auditory steady stateevoked potentials during sleep. Ear Hear 6(3):167–174.

29. O’Donnell BF, et al. (2013) The auditory steady-state response (ASSR): A translationalbiomarker for schizophrenia. Suppl Clin Neurophysiol 62:101–112.

30. Adamantidis AR, Zhang F, Aravanis AM, Deisseroth K, de Lecea L (2007) Neural sub-strates of awakening probed with optogenetic control of hypocretin neurons. Nature450(7168):420–424.

31. Madisen L, et al. (2012) A toolbox of Cre-dependent optogenetic transgenic mice forlight-induced activation and silencing. Nat Neurosci 15(5):793–802.

32. Metherate R, Cox CL, Ashe JH (1992) Cellular bases of neocortical activation: Modu-lation of neural oscillations by the nucleus basalis and endogenous acetylcholine.J Neurosci 12(12):4701–4711.

33. Nag N, Baxter MG, Berger-Sweeney JE (2009) Efficacy of a murine-p75-saporin im-munotoxin for selective lesions of basal forebrain cholinergic neurons in mice. Neu-rosci Lett 452(3):247–251.

34. Jourdain A, Semba K, Fibiger HC (1989) Basal forebrain and mesopontine tegmentalprojections to the reticular thalamic nucleus: An axonal collateralization and immu-nohistochemical study in the rat. Brain Res 505(1):55–65.

35. Han X, et al. (2011) A high-light sensitivity optical neural silencer: Development and ap-plication to optogenetic control of non-human primate cortex. Front Syst Neurosci 5:18.

36. Lin SC, Gervasoni D, Nicolelis MA (2006) Fast modulation of prefrontal cortex activity bybasal forebrain noncholinergic neuronal ensembles. J Neurophysiol 96(6):3209–3219.

37. Rosanova M, et al. (2009) Natural frequencies of human corticothalamic circuits.J Neurosci 29(24):7679–7685.

38. Ferrarelli F, et al. (2012) Reduced natural oscillatory frequency of frontal thalamo-cortical circuits in schizophrenia. Arch Gen Psychiatry 69(8):766–774.

39. Brown RE, Basheer R, McKenna JT, Strecker RE, McCarley RW (2012) Control of sleepand wakefulness. Physiol Rev 92(3):1087–1187.

40. Détári L, Rasmusson DD, Semba K (1999) The role of basal forebrain neurons in tonicand phasic activation of the cerebral cortex. Prog Neurobiol 58(3):249–277.

41. Yang C, et al. (2014) Cholinergic neurons excite cortically projecting basal forebrainGABAergic neurons. J Neurosci 34(8):2832–2844.

42. Plourde G (1996) The effects of propofol on the 40-Hz auditory steady-state responseand on the electroencephalogram in humans. Anesth Analg 82(5):1015–1022.

43. Edlow BL, et al. (2013) Disconnection of the ascending arousal system in traumaticcoma. J Neuropathol Exp Neurol 72(6):505–523.

44. Porkka-Heiskanen T, et al. (1997) Adenosine: A mediator of the sleep-inducing effectsof prolonged wakefulness. Science 276(5316):1265–1268.

45. Yang C, Franciosi S, Brown RE (2013) Adenosine inhibits the excitatory synaptic inputsto basal forebrain cholinergic, GABAergic, and parvalbumin neurons in mice. Fron-tiers in neurology 4:77.

46. Makeig S, Jung TP (1996) Tonic, phasic, and transient EEG correlates of auditoryawareness in drowsiness. Brain Res Cogn Brain Res 4(1):15–25.

47. Verret L, et al. (2012) Inhibitory interneuron deficit links altered network activity andcognitive dysfunction in Alzheimer model. Cell 149(3):708–721.

48. Grothe M, Heinsen H, Teipel SJ (2012) Atrophy of the cholinergic Basal forebrain over theadult age range and in early stages of Alzheimer’s disease. Biol Psychiatry 71(9):805–813.

49. Madhusudan A, Sidler C, Knuesel I (2009) Accumulation of reelin-positive plaques isaccompanied by a decline in basal forebrain projection neurons during normal aging.Eur J Neurosci 30(6):1064–1076.

50. Marín O (2012) Interneuron dysfunction in psychiatric disorders. Nat Rev Neurosci13(2):107–120.

51. Woo TU, Spencer K, McCarley RW (2010) Gamma oscillation deficits and the onset andearly progression of schizophrenia. Harv Rev Psychiatry 18(3):173–189.

52. Franklin KBJ, Paxinos G (2008) The Mouse Brain in Stereotaxic Coordinates (Academic,New York), 3rd Ed.

3540 | www.pnas.org/cgi/doi/10.1073/pnas.1413625112 Kim et al.