R E S E A R CH A R T I C L E

Mapping mesoscale cortical connectivity in monkeysensorimotor cortex with optical imaging and microstimulation

Robert M. Friedman1 | Katherine A. Morone2 | Omar A. Gharbawie3 |

Anna Wang Roe1,4

1Division of Neuroscience, Oregon National

Primate Research Center, OHSU, Beaverton,

Oregon

2Department of Neurology, Vanderbilt

University Medical Center, Nashville,

Tennessee

3Department of Neurobiology, Center for

Neural Basis of Cognition, University of

Pittsburgh, Pittsburgh, Pennsylvania

4Interdisciplinary Institute of Neuroscience

and Technology, School of Medicine, Zhejiang

University, Hangzhou, China

Correspondence

Anna Wang Roe, Interdisciplinary Institute of

Neuroscience and Technology, School of

Medicine, Zhejiang University, Hangzhou

310029, China.

Email: annawang@zju.edu.cn

Robert M. Friedman, Division of Neuroscience,

Oregon National Primate Research Center,

OHSU, 505 NW 185th Ave, Beaverton,

OR 97006.

Email: friedmro@ohsu.edu

Funding information

NIH, Grant/Award Numbers: NS044375,

NS093998, TL1TR000447

Peer Review

The peer review history for this article is

available at https://publons.com/publon/10.

1002/cne.24918.

Abstract

To map in vivo cortical circuitry at the mesoscale, we applied a novel approach to

map interareal functional connectivity. Electrical intracortical microstimulation (ICMS)

in conjunction with optical imaging of intrinsic signals (OIS) was used map functional

connections in somatosensory cortical areas in anesthetized squirrel monkeys. ICMS

produced activations that were focal and that displayed responses which were stimu-

lation intensity dependent. ICMS in supragranular layers of Brodmann Areas 3b, 1, 2,

3a, and M1 evoked interareal activation patterns that were topographically appropri-

ate and appeared consistent with known anatomical connectivity. Specifically, ICMS

revealed Area 3b connections with Area 1; Area 1 connections with Areas 2 and 3a;

Area 2 connections with Areas 1, 3a, and M1; Area 3a connections with Areas M1,

1, and 2; and M1 connections with Areas 3a, 1, and 2. These somatosensory connec-

tivity patterns were reminiscent of feedforward patterns observed anatomically,

although feedback contributions are also likely present. Further consistent with ana-

tomical connectivity, intra-areal and intra-areal patterns of activation were patchy

with patch sizes of 200–300 μm. In summary, ICMS with OIS is a novel approach for

mapping interareal and intra-areal connections in vivo. Comparisons with

feedforward and feedback anatomical connectivity are discussed.

K E YWORD S

intrinsic optical imaging, intracortical microstimulation, somatosensory cortex, cortical

connections, Saimiri sciureus, RRID:SCR_001622 (MATLAB), RRID:SCR_008291 (Oregon

National Primate Research Center), RRID:SCR_009665 (Oregon Health and Science

University; OR), RRID:SCR_006659 (University of Pittsburgh; PA), RRID:SCR_000224

(Zhejiang University; Zhejiang; China), RRID:SCR_011756 (Vanderbilt University; TN)

1 | INTRODUCTION

The challenge of understanding connections in cerebral cortex has

been addressed traditionally with anatomical tract tracing techniques.

These gold standard methods have defined interareal connectivity, as

well as distinguished feedforward and feedback circuits. Furthermore,

they have revealed extensive intrinsic intra-areal horizontal connec-

tions. However, traditional anatomical techniques are limited to the

study of only a few targeted sites per animal and involve extensive

postmortem tissue processing. We seek a method that is capable of

revealing anatomical connections in vivo, can be rapidly evaluated,

and can achieve a spatial resolution capable of revealing anatomical

patches.

We have previously shown that the combination of electrical

intracortical microstimulation (ICMS) and the optical imaging of intrin-

sic signals (OIS) permit the visualization of the evoked activity (Brock,

Friedman, Fan, & Roe, 2013; Stepniewska et al., 2011). The advan-

tages of this combined approach are twofold. First, by measuring

Received: 6 February 2020 Revised: 23 March 2020 Accepted: 24 March 2020

DOI: 10.1002/cne.24918

J Comp Neurol. 2020;1–13. wileyonlinelibrary.com/journal/cne © 2020 Wiley Periodicals, Inc. 1

changes in the concentration of deoxygenated hemoglobin correlated

with local cortical activation and without external imaging agents, OIS

reveals sites of cortical activation with a spatial resolution on the

order of tens of micrometers (Bonhoeffer & Grinvald, 1996). Second,

ICMS and OIS can be conducted in vivo without animal sacrifice, and

can be done quickly, and repeatedly, over periods of weeks and

months. Third, ICMS of the cerebral cortex is capable of generating or

altering primate perception and motor behavior in a precise manner

(Graziano, Aflalo, & Cooke, 2005; Murphey & Maunsell, 2007; Romo,

Hernández, Zainos, Brody, & Lemus, 2000; Salzman, Britten, &

Newsome, 1990; Stepniewska, Fang, & Kaas, 2009; Tehovnik &

Slocum, 2009). This raises the possibility of pairing an induced behav-

ioral effect with underlying circuitry.

In this study, we investigated whether ICMS in combination with

OIS can reveal cortical connections observed with traditional anatomi-

cal approaches. We chose primate somatosensory cortex as a testbed.

The anatomical feedforward and feedback interareal circuitry patterns

between the Brodmann somatosensory areas are well known as dia-

gramed in Figure 1 (Kaas, 1983; Shanks et al., 1985). By applying

ICMS at locations of single digit representation in Brodmann Areas

3b, 1, 2, 3a, and M1, local and interareal evoked activity was mapped

with OIS. We observed that the patterns of interareal activations

were consistent with known anatomical connectivity patterns. In con-

clusion, we find that ICMS paired with OIS can be used in vivo to map

connections between cortical areas at mesoscale.

2 | METHODS

2.1 | Surgical procedures

Squirrel monkeys (Saimiri sciureus N = 3) were sedated with ketamine

(15 mg/kg). After placement of an endotracheal tube, the animals were

anesthetized with isoflurane (0.9–1.5%) and given an analgesic (0.01 mg/

kg buprenorphine, i.m.) to maintain a light surgical plane of anesthesia,

where cortical responses to vibrotactile stimulation and ICMS were read-

ily recorded. During the anesthesia procedure, animals were artificially

ventilated, secured in a stereotaxic frame for surgery and to prevent

hypothermia wrapped in a heating blanket, and monitored continuously

for end-tidal CO2, heart rate, temperature, and respiratory function. A

craniotomy and durotomy were performed to expose the somatosensory

and motor cortices. The brain was stabilized with 4% agar. To perform

optical imaging with ICMS, an electrode was positioned at an angle into a

desired site, and a cover glass was placed immediately next to the elec-

trode and covered the area to be imaged. All procedures were conducted

in accordance with National Institutes of Health guidelines and were

approved by Institutional Animal Care and Use Committees.

2.2 | Somatosensory stimulation

For functional mapping with tactile stimulation, the contralateral fore-

arm and hand were shaved and secured outstretched in a supine

position. Two-millimeter-diameter probes attached to piezoceramic

benders (Noliac North America, Alpharetta, GA) were placed with a

slight indentation on the palmar side of a distal phalanx of one or

many digits. Tactile stimulation consisted of a train of pulses delivered

at 8 Hz lasting for 3 s, with each pulse having a displacement of

250 μm and lasting 20 ms.

2.3 | Electrical microstimulation

After targeting a cortical site for stimulation, a microdrive was used to

lower a 0.9- to 1.5-MΩ Parylene C-coated tungsten microelectrode

(Microprobes, Gaithersburg, MD) through the agar and penetrate cor-

tex. An electrode was typically inserted 350–500 μm in depth into

cortex of single digit domains. Electrical microstimulation (Biphasic

Stimulus Isolator, Bak Electronics Inc., Umatilla, FL) consisted of a train

of biphasic pulses (250 Hz, 0.4-ms pulse duration) that varied in

F IGURE 1 Schematic of primary interareal connections ofprimate somatosensory cortex and primary motor cortex. Uniquefeedforward (solid lines) and feedback (dotted lines) intracorticalprojection patterns exist between the four cytoarchitecturalsubdivisions (Areas 3a, 3b, 1, and 2) of primate somatosensory cortexand primary motor cortex (M1). M1 projects to Areas 3a, 1, and 2 andreceives reciprocal feedback connections from Areas 3a and 2. Area3a projects to and receives reciprocal connections with M1, and Areas1 and 2. The major projection of Area 3b is to Area 1 and receivesfeedback connections from Areas 1 and 2. Area 1 projects to Areas 3aand 2, and receives feedback connections from Area 2. Area2 receives feedforward input from Areas 3b and 1, and projects toAreas 3a, 1, and M1. Schematic is adapted from Shanks, Pearson, andPowell (1985) and is based on anatomical tracer studies (see

Section 4) [Color figure can be viewed at wileyonlinelibrary.com]

2 FRIEDMAN ET AL.

current amplitude (15–300 μA) with a train duration of 250 ms

(63 pulses). These stimulation parameters were based on prior func-

tional ICMS studies (cf. Graziano, Taylor, & Moore, 2002; Stepniewska

et al., 2011; Tolias et al., 2005). Electrode impedances were evaluated

before and after ICMS. In most cases, impedances were within the ini-

tial range. Electrodes were discarded for future use if significantly

lower impedances were obtained.

To determine the cortical depth of stimulation we measured the

angle of electrode penetration from the cortical surface with a pro-

tractor. We monitored, both visually and by listening to an audio mon-

itor, the moment when the electrode entered cortex; there is very

little dimpling, as we use very fine-tipped microelectrodes. Thus, we

were confident to within 100 μm of the recorded depth. The vertical

depth of the electrode was based on the traverse along the electrode

penetration multiplied by the sine of the electrode angle. The depth

reported in this study was the estimated vertical depth.

2.4 | Optical imaging

Intrinsic optical images were obtained with a CCD video camera

(504 × 504 pixels, 12 × 12-mm field of view) connected to an imaging sys-

tem (Imager 3001, Optical Imaging, Ltd., Rehovot, Israel). Cortex was illu-

minated with red light (630 nm) for signal acquisition and green light

(578 nm) to record vessel maps. Images were collected for either 3 or 4 s

at a frame rate of 5 Hz with an interstimulus interval of 5–8 s. A pre-

stimulus period of 300 ms (1.5 frames) was included in each trial of image

acquisitions. An experimental run consisted of 15–30 stimulus replications.

Stimuli were presented in a block format in a randomly selected order.

2.5 | Image analysis

Images were first frame subtracted using the first frame of each trial.

Single stimulus condition maps were constructed according

dR/R = (Fx-y–F1)/F1, where dR/R is the percent reflectance change, Fx-

y is the average raw reflectance in Frames X-Y, and F1 is the raw reflec-

tance in the first, blank, frame. Signal-to-noise ratio was improved by

averaging across frames and trials. The gray value of each pixel repre-

sents the change in light reflectance relative to prestimulus baseline.

In the image maps, dark pixels represent stimulus-induced signal

change, since with 630-nm light an increase in cortical activity leads

to an initial decrease in reflectance (−0.01 to −0.2%). A high-pass

mean filter (100 μm) and a low-pass median filter (2.4 mm) were

applied to all images. For presentation purposes, images were clipped

with an intensity range of 0.02–0.05%. Time courses were calculated

for regions of interest (ROIs) to characterize the temporal develop-

ment of the hemodynamic signal. ROIs of diameter (140 μm) were

manually selected. All images and time courses were calculated with

custom software written in MATLAB (MathWorks, Natick, MA).

T-maps, generated by pixel-by-pixel Student's t-tests, were used

to identify significant pixels activated by the ICMS. This analysis was

used to identify locations of stimulus-related response, and to identify

ROIs for time course analysis. T-maps were generated with MATLAB

and were visualized by creating binary maps thresholded for a specific

p-value (i.e., p < .01). Isolated statistically significant pixels, for exam-

ple, along blood vessels, were not included in subsequent analysis as

the significance of these signals was ambiguous.

2.6 | Demarcation of the areal borders and digitrepresentations

The somatotopic and functional organization of the hand area is well

characterized in squirrel monkeys where both optical imaging and

electrophysiological methods have been used to delineate the

somatotopic organization of the hand region of somatosensory cortex

(Areas 3a, 3b, 1, and 2; Sur, Nelson, & Kaas, 1982; Chen, Friedman,

Ramsden, LaMotte, & Roe, 2001; Friedman, Chen, & Roe,

2004, 2008). Borders between cortical areas were estimated based on

electrophysiological mapping, receptive field characteristics, and opti-

cal imaging. For electrophysiological mapping, the responsive skin area

of the unit activity was identified by palpating areas on the contralat-

eral hand while listening to the audio amplifier for spiking activity.

Area 3b is characterized by single and multiunit response properties

having small receptive fields (restricted to a single finger) and brisk

responsiveness to light touch. In comparison, Area 1 units have larger

receptive fields that often covered more than one finger. The cortical

region between the representations of the distal finger pads in Areas

3b and 1 is marked by the representations of the middle phalanges and

palm. Thus, the Areas 3b and 1 border was estimated to lie in the cen-

ter of this region. The 3a/3b border is characterized by a tip-to-tip

organization of the distal finger pads and a change in activation from

cutaneous receptors in 3b to predominantly deep receptors that are

activated by joint movement in 3a was used to estimate the Area

3a/3b border. Similarly, the Areas 1 and 2 border is also characterized

by a tip-to-tip organization of the distal finger pads, where the approxi-

mate border was placed where there is a sharp reduction in responsive-

ness to cutaneous stimuli, characteristic of Area 2 under anesthesia.

With optical imaging, the cortical somatotopy of the hand region was

mapped by presenting vibrotactile stimulation to multiple distal finger

pads. Digit tip activations provided both the medial to lateral orienta-

tions of Areas 3b and 1 as well as the anterior and posterior borders of

Areas 3b and 1, respectively, as neurons in Areas 3a and 2 respond

poorly to the vibrotactile stimulus used for optical imaging mapping.

3 | RESULTS

To examine whether electrical microstimulation in combination with

OIS can reveal functional and anatomical connectivity patterns, we

compared our results with the well described anatomical connections

of the digit representations in somatosensory cortex of squirrel mon-

keys (Ashaber et al., 2014; Cerkevich & Kaas, 2019; Liao, Gharbawie,

Qi, & Kaas, 2013; Négyessy et al., 2013; Pálfi et al., 2018). Stimulation

was applied in cortical Areas 3a, 3b, 1, 2, and M1 in 16 stimulation

FRIEDMAN ET AL. 3

sites in four hemispheres of three squirrel monkeys (S, C, M) con-

ducted in six experiments (Table 1). Areal boundaries were estimated

from mapping digit activations using OIS and electrophysiological

characterization of neuronal receptive field properties (see Figure 2a,

Section 2). Microstimulation was administered in the superficial layers

(at a cortical depth of ~400 μm), at a range of current amplitudes

(0.4 ms biphasic pulses, at 250 Hz, duration 250 ms). These parame-

ters were based on our previous functional ICMS studies (Brock

et al., 2013; Stepniewska et al., 2011).

3.1 | Electrical microstimulation elicits focal anddistant areas of activation

Our approach first involved functional mapping of the digit represen-

tation of somatosensory cortex with OIS and electrophysiology in

response to cutaneous stimulation. We then mapped intracortical

functional connectivity by targeting a microelectrode into cortical

superficial layers of a selected distal finger pad representation and

imaged the effects of focal cortical microstimulation. An example of

this approach is shown in Figure 2 for a squirrel monkey (SQM S, Expt

2, D2) under isoflurane anesthesia. Figure 2a (left panels) illustrates

the OIS response to cutaneous stimulation of D2 and D4 in Area 3b

with activations in Areas 3b and 1 outlined. As is sometimes the case,

weak activation was evident in Area 1, due to greater sensitivity to

anesthetics in higher cortical areas. Electrophysiological mapping of

the response properties of units along with OIS allowed for the

demarcation of somatosensory Brodmann areas (right panel).

A microelectrode (red arrowhead, Figure 2b) was then placed in

the superficial layers of Area 3b in the D2-responsive location and a

100 μA stimulus was applied. Figure 2b illustrates, for a single stimula-

tion trial, the temporal development of the OIS response. At the elec-

trode tip site, a focal millimeter-sized activation became apparent

after ~600 ms and remained throughout the recording period (4 s

total). This single trial response demonstrates that such ICMS parame-

ters evoke focal and robust cortical activity. Averaging the response

over multiple trials (n = 27, Figure 2c, t-maps shown in 2d) improves

the signal to noise ratio and further reveals the spatiotemporal devel-

opment of the activation. Spatially, we observed an increase in

response size from ~1 mm in diameter at ~0.5 s to ~2–3 mm at ~2 s,

followed by a gradual shrinkage in size. Temporally, the signal peaked

around 2 s and achieved a reflectance change magnitude of about

0.2% (Figure 2e, blue line).

We also observed other secondary activation sites. Posterior to

the site of stimulation were two adjacent 0.5-mm-sized activations,

consistent with D2 digit tip representations in Areas 1 and 2, reflected

across the Area 1/2 border (Figure 2c,d, light dotted line). Similar to

the stimulation site, these focal activations increased in size over time

until around 2 s, after which they decreased in size. The signal magni-

tudes were much weaker than at the stimulation site (Figure 2e, red

and orange lines vs. blue line), and were comparable to magnitudes

observed with cutaneous vibrotactile stimulation (Chen, Friedman, &

Roe, 2003; Friedman et al., 2008). These timecourses are consistent

with that of the well-established intrinsic signal response, which,

under anesthesia, typically peaks at around 2–3 s and is followed by a

slow decline to baseline over several seconds and reaches a magni-

tude in the range of 0.1–0.2% (Bonhoeffer & Grinvald, 1996; Chen

et al., 2003; Friedman et al., 2004).Furthermore, consistent with intra-

areal interdigit connectivity revealed anatomically (Négyessy

et al., 2013; Wang et al., 2013), significant activations were also

observed medial and lateral to the stimulation site (Figure 2d, white

arrows). Responses at control regions were not significant (Figure 2e,

green line).

In sum, with the stimulation parameters used, ICMS evoked a

focal activation at the stimulation site and several associated activa-

tions. The locations of these secondary activations were consistent

with anatomically connected sites shown previously (Négyessy

et al., 2013; Wang et al., 2013; Xu et al., 2019). The largest amplitude

of activity was at the site of stimulation (in Area 3b). Those at

TABLE 1 Experimental summary of observed inter arealconnectivity patterns. Anesthetized experiments were performed inthree squirrel monkeys (S: three recovery experiments and oneterminal T; sessions; C: one terminal T; M: one terminal T). ICMS wasperformed in primary motor cortex (M1) and somatosensory Areas 3a,3b, 1, and 2. X indicates the presence of any significant pixels

Area Cases D M1 3a 3b A1 A2

3b S

Expt 1 D4 X X X X X

Expt 2 D2 X

T D4 X X X

C

T D1 X X

M

T (L hemi) D2 X X

T (R hemi) D2 X X

3a S

Expt 3 D4 X X+ X

T D4 X X+ X

C

T D1 X X X

M

T (L hemi) D2 X X X

T (R hemi) D2 X X

A1 M

T (L hemi) D2 X X X

A2 M

T (L hemi) D2 X X X

M1 S

Expt 3 X X X

T X X

M

T (L hemi) X X X

Abbreviation: ICMS, intracortical microstimulation; D, stimulated digit.

4 FRIEDMAN ET AL.

connected sites (within Area 3b, in Area 1, and in Area 2) were also

focal and reproducible, although smaller in size and in amplitude.

These observations suggest that ICMS evokes physiological responses

at anatomically connected sites.

3.2 | Activation is dependent on ICMS intensity

The magnitude of the cortical response reported by OIS was depen-

dent on the amount of current delivered during electrical ICMS (see

also Brock et al., 2013). Figure 3 shows the OIS and t-maps and intrin-

sic signal timecourses to four different ICMS current levels of stimula-

tion in Area 3b (same case and electrode placement as Figure 2).

At 15 μA, a very small (100–200 μm) but significant activation is evi-

dent at the site of the electrode (Figure 3a, first row). This activation

size increased as stimulation increased from 25 μA (second row) to

50 μA (third row) to 100 μA (fourth row). The distant spots of activa-

tion in Areas 1 and 2 appeared with stimulation intensity of 50 μA

and increased in size with 100 μA. Secondary sites in Area 3b were

also seen at 50 and 100 μA. These increases in activation were evi-

dent in the amplitudes of reflectance change in Area 3b (Figure 3b,

top) from roughly 0.01% at 15 μA (black), to 0.05% at 25 μA (blue),

to 0.1% at 50 μA (orange), to 0.2% at 100 μA (red). Amplitudes in Area

1 (Figure 3b, bottom, left) and Area 2 (Figure 3b, bottom, right) also

exhibited intensity dependence, although amplitudes were overall

lower. Thus, at both the stimulation site and connected sites, cortical

activation increased with ICMS stimulus intensity. Connected cortical

sites showed higher thresholds of activation.

3.3 | Consistency of cortical activation patternsevoked by ICMS

In all the experimental cases (Table 1), ICMS in Area 3b elicited focal

activation at the site of the electrode and at additional posterior sites

in Area 1. In some cases, additional activations within Area 3b could

be seen as well. Figure 4 shows examples from three different animals

((a): SQMM left hemi, D2; (b): S Expt T, D4; (c): C, Expt T, D1) to ICMS

F IGURE 2 Electrical microstimulation in Area 3b evokes focal activations at stimulated and connected sites. (a) Cutaneous stimulation of D2and D4 elicits OIS activations in Areas 3b and 1. Left: optical imaging of intrinsic signal (OIS) maps. Evoked OIS are the dark outlined activations.Grayscale bar: reflectance change in units of 10−4. Right: Panel of a blood vessel map overlaid with outlines of OIS activation to D2 (orange), D3(green) and D4 (blue) vibrotactile stimulation, electrophysiological map of unit responses (deep/cutaneous, squares), and estimated bordersbetween Brodmann areas (dashed lines). M = medial. P = posterior. (b) Image frames from a single trial showing the OIS temporal response tointracortical microstimulation (ICMS) in the D2 distal finger pad representation of 3b (times shown at top). ICMS: 100 μA, 0.4 ms biphasic pulses,250 Hz, 250 ms, depth: 350 μm. Red arrowhead: Microelectrode, applies to all frames in (b) and (c). (c) Timecourse of activation after averagingacross trials (n = 27). (d) T-maps indicating significance of maps shown in (c). Colored scale bar: p < 10−2 to 10−7 (uncorrected). Color dots: regions

of interest (ROIs) for timecourses shown in (e). Heavy and light dotted lines: Approximate border between Areas 3b and 1, and between Areas1 and 2, respectively. White arrows: Activations at connected sites within Area 3b. All images are blank subtracted. Scale bar = 1 mm, applies toall panels. (e) Time course of reflectance change for selected ROIs shown in (d). Blue: Area 3b. Red: Area 1. Orange: Area 2. Black: Control site(gray dot in d). Shading: SEM. Red bar: Electrical stimulation period. Case: SQM S, Expt 2 [Color figure can be viewed at wileyonlinelibrary.com]

FRIEDMAN ET AL. 5

of Area 3b using the same stimulation parameters (100 μA). Focal acti-

vations (2–4 mm in size) were seen in Area 3b (white arrows), and dis-

tant activations (1–2 mm in size) were observed in Area

1. Importantly, the pattern of activation elicited by ICMS in Area 3b

was comparable to patterns of connectivity shown in the previous

case (Figures 2 and 3), as well as to patterns revealed by anatomical

tracing studies of projections from Area 3b to Area 1 (Cerkevich &

Kaas, 2019; Négyessy et al., 2013; Pálfi et al., 2018; Shanks

et al., 1985).

3.4 | Mapping connectivity across multiple areas insomatosensory cortex

Given that focal ICMS could elicit interareal and intra-areal connec-

tions, we wanted to know whether this method could be used to

study focal connectivity of the different Brodmann-somatosensory

cortical areas. Previous anatomical methods have delineated the pat-

tern of feedforward connections between different areas of somato-

sensory and motor cortex (Jones, Coulter, & Hendry, 1978; Shanks

et al., 1985). That is, as shown in Figure 1, Area 3b projects to Area

1, Area 1 projects to Areas 2 and 3a, Area 2 projects to Areas 3a and

M1, and Area 3a projects to Areas M1, 1, and 2. Could these projec-

tions be revealed with ICMS? If so, then this would be the first dem-

onstration of mapping focal cortical networks in vivo and in a digit

specific manner.

To examine this, we conducted focal ICMS in each of Areas M1,

3a, 3b, 1, and 2 and mapped functional connectivity with OIS. To

focus primarily on feedforward connection patterns, we stimulated in

the superficial and middle cortical layers. Although it is likely that elec-

trical stimulation activates both feedforward and feedback projec-

tions, we reasoned that feedforward projections, which are known to

be more clustered, are likely to result in stronger focal signal in an

imaged population response; feedback connections are generally con-

sidered more diffuse and perhaps less likely to reveal clear activations.

This is supported by a previous study examining results from stimulat-

ing different cortical depths in squirrel monkey somatosensory cortex

(Brock et al., 2013).

As shown in Figure 5, we conducted focal ICMS in Areas M1 (a),

3a (b), 3b (c), 1 (d), and 2 (e). Each panel, with the contrast emphasized,

illustrates the activations elicited by the electrode inserted into the

superficial layers of cortex ((a): SQM S Expt 3, (b): C Expt T, D1; (c): S,

Expt 1, D4; (d) M, left hemi, D2; (e) M, left hemi, D2). The left column

is a wide field of view of Areas 3a, 3b, 1, 2, and, in some cases, M1.

This column is redisplayed with annotations in the middle column.

The right column is a magnified view of the stimulation region (region

in red box of middle column), shown to illustrate the intra-areal activa-

tions (orange arrows) and nearby interareal activations (white and

green arrows) and the patchiness of activations (colored circles). The

most prominent activations in these images reveal connectivity pat-

terns similar to previously identified feedforward connections

between the different areas. As indicated by the white arrows,

(1) stimulation of Area 3b predominantly activates Area 1 (Figure 5c),

(2) stimulation of Area 1 activates Areas 2 and 3a, as well as patchy

domains within Areas 1 and 2 (Figure 5d), (3) stimulation of Area

2 activates Area 3a as well as patchy domains within Area 2 (Figure 5e),

(4) stimulation of Area 3a activates Areas M1, 1, and 2, as well as pat-

chy domains within 3a (Figure 5b), and (5) stimulation of M1

F IGURE 3 Intensity dependence of intracortical microstimulation (ICMS) activations. (a) Stimulation of distal digit D2 in Area 3b revealsstimulus intensity dependence. Activation (left column) and T-maps (right column) in response to ICMS at 15 μA, 25 μA, 50 μA, and 100 μA. Forboth stimulated site in Area 3b and connected sites in Areas 1 and 2, the size of activation increases with increases in stimulus intensity.Conventions as Figure 2; same case as Figure 2. Scale Bar =1 mm. (b) Time course of reflectance change of select regions of interest (ROIs) fordifferent stimulus intensities. Top graph: Response of stimulation site in Area 3b. Bottom graphs: Response of Area 1 (left) and Area 2 (right)activations. Shading: SE of the mean. Red bar: Electrical stimulation period [Color figure can be viewed at wileyonlinelibrary.com]

6 FRIEDMAN ET AL.

(Figure 5a) resulted in many patches of activation surrounding the stimu-

lation site. We noted that both the activations within M1 (Figure 5a) and

the M1 activations following stimulation of 3a (Figure 5b) appeared less

focal and spanned a larger area, several millimeters in extent (Figure 5a,

right panel; Figure 5b, green encircled region).

Over the many cortical areas, patchy activations were elicited.

Patches were typically less than 500 μm in size, consistent with previ-

ous reports that anatomical patches in Areas 3b and 1 are typically

200–300 μm in size (Négyessy et al., 2013). Moreover, interareal acti-

vations exhibit a reasonable degree of digit specificity, as evidenced

by the topographic reflection across areal borders. This seen in the

connections from Area 3b to Area 1 (Figure 5c), from Area 1 to Areas

2 and 3a (Figure 5d), from Area 2 to Area 3a (Figure 5e), and from

Area 3a to Areas 1 and 2 (Figure 5b). Thus, the elicited connectivity

patterns appear consistent with feedforward connections between

different areas of somatosensory and motor cortex (compare with

Figure 1, Shanks et al., 1985).

This consistency is further illustrated in Figure 6 (same cases as

Figure 5). The diagram in Figure 6a (right) summarizes the basic

feedforward connections of M1 as shown by previous anatomical

tracer studies (Darian-Smith, Darian-Smith, Burman, & Ratcliffe, 1993;

Gharbawie, Stepniewska, & Kaas, 2011; Stepniewska, Preuss, &

Kaas, 1993). As shown in Figure 6a (left), the stimulation of M1 evokes

several focal sites of activation within M1, 3a, and Area 1 that

F IGURE 4 Consistent activation patterns across cases. (a–c)Optical imaging of intrinsic signal (OIS) image (left) and correspondingT-maps (right) for three different cases of Area 3b stimulation ((a):SQM M left hemi, D2; (b): S Expt T, D4; (c): C, Expt T, D1).Intracortical microstimulation (ICMS): 100 μA. Red arrowhead:Microelectrode. White dashed lines: Approximate border betweenAreas 3b and 1. White arrows: Activations at connected sites withinArea 3b. Scale bar = 1 mm. p = posterior, l = lateral. Grayscale bar: inunits of 10−4 [Color figure can be viewed at wileyonlinelibrary.com]

F IGURE 5 Mapping connectivity across multiple Brodmannareas. Focal electrical stimulation reveals connectivity networksacross multiple areas in squirrel monkey somatomotor cortex. Each

panel illustrates patchy activation resulting from focal stimulation in(a) M1 (100 μA), (b) Area 3a (150 μA), (c) Area 3b (100 μA), (d) Area1 (100 μA), and (e) Area 2 (200 μA). Cases: (a): SQM S Expt 3, (b): CExpt T, D1; (c): S, Expt 1, D4; (d) M, left hemi, D2; (e) M, left hemi, D2.Images in left column are a wide field of view of Areas 3a, 3b, 1, 2,and, in some cases, M1. Middle column is the left column redisplayedwith annotations and cortex outlined in yellow (dura, bone excluded).Right column is a magnified view of the stimulation region (region inred box of middle column). Patchy activations indicated by circles.White arrows: Activations consistent with feedforward connections.Orange arrows: Intra-areal connections. Green arrows: Motor cortexconnections. Each image sum of 15–30 trials. For all, scale bars:1 mm. Red arrowhead: Microelectrode [Color figure can be viewed atwileyonlinelibrary.com]

FRIEDMAN ET AL. 7

reproduce this pattern (Area 2 out of field of view). The thresholded

t-maps are shown in Figure 6a (middle). For each site that was stimu-

lated (Figure 6b: Area 3a, Figure 6c: Area 3b, Figure 6d: Area 1, Figure

6e: Area 2), we observe that the known anatomical connections

(Figure 6a, right column schematics) are replicated in the stimulation

patterns (Figure 6a–e, left and middle columns, see legend). We note,

however, that feedback connections may also have contributed to

these activation patterns, something that we cannot rule out.

In sum, we systematically probed each area in somatosensory cor-

tex and primary motor cortex to examine whether this focal stimula-

tion method could elicit activations consistent with known anatomical

connections. We found that activation patterns exhibited remarkable

similarities. In fact, (a) interareal activations patterns paralleled known

feedforward connection hierarchies (Shanks et al., 1985), although

feedback contributions may also be present, (b) interareal connectivity

largely replicated topographic digit specificity in the anteroposterior

axis and interdigit connectivity in the mediolateral axis, and

(c) activations often appeared patchy, consistent with anatomical trac-

ing studies.

3.5 | Signal amplitude at different distal sites is notuniform

Connections from a single cortical site may elicit response from mul-

tiple interareal and intra-areal sites. However, their response ampli-

tudes are not the same. Two examples are shown in Figure 7. For

example, in Figure 7a (same case as Figure 6c), stimulation of Area

3b produces a robust activation in Area 1 (approximately 0.09%

reflectance change), a well-established primary cortical target of

Area 3b. The other activations are much reduced in amplitude:

Area 3a ~ (0.03%), Area 2 (~0.02%), M1 (~0.02%), with control site

<0.01% in amplitude. A second example is shown in Figure 7b

(same case as Figure 6a). Stimulation of M1 leads to strongest acti-

vation in Areas M1 (blue, orange, gray, ~0.03–0.06%) and in 3a (red,

green, ~0.03–0.06%). Reflectance changes in Area 1 (peach, 0.01%)

and control site (dark gray, 0%) are substantially weaker. This range

is not unexpected, and may be attributed to differences in anatomi-

cal projection strengths, and feedforward and feedback

contributions.

F IGURE 6 Activations are similar to anatomical feedforward connectivity patterns. (a–e) Image maps show the response to intracorticalmicrostimulation (ICMS) in (a) M1 (100 μA), (b) Area 3a (150 μA), (c) Area 3b (100 μA), (d) Area 1 (100 μA), and (e) Area 2 (200 μA). Same cases asFigure 5. The connected sites of activation are outlined in green. The microelectrode can be seen in each of these images pointing towards thesite of stimulation. T-maps indicate the significance of evoked activations (p < .01). To emphasize the projection patterns, the significant opticalimaging of intrinsic signals (OIS) activity at the site of stimulation is not shown. Schematic on right shows the feedforward projections of M1 andAreas 3a, 3b, 1, and 2, revealed by published anatomical tracer studies. Note that the OIS activities induced by cortical microstimulation parallelthose predicted by known feedforward connectivity. However, feedback contributions cannot be excluded white dashed lines: Approximateborder between cortical areas. Red arrowhead: Microelectrode. Scale bar = 1 mm. p = posterior, l = lateral. Grayscale bar: in units of 10−4 [Colorfigure can be viewed at wileyonlinelibrary.com]

8 FRIEDMAN ET AL.

4 | DISCUSSION

We have demonstrated that combining ICMS with OIS can be a useful

approach for examining neuronal circuitry within and between cortical

areas. In somatosensory cortex and MI, this novel approach revealed

the functional connectivity between Areas 3a, 3b, 1 and 2, and M1.

ICMS in different somatosensory Brodmann areas resulted in different

activation patterns. These activations were focal, exhibited reflec-

tance change timecourses typical of that induced by sensory induced

neuronal response (Figure 2; cf. Chen et al., 2001, 2003), and were

intensity dependent (Figure 3). Furthermore, we found that activation

patterns were reproducible across trials within single animals and

across animals (Figure 4).

Most importantly, our results replicated known feedforward

interareal connectivity patterns between areas in somatosensory cor-

tex and M1 (Figures 5 and 6). ICMS in Area 3b showed functional acti-

vation sites predominantly in Area 1 and weaker connectivity in Area

2. ICMS in Area 1 showed activation in Area 3a and 2, and that in

Area 2 showed activation in Areas 1, 3a, and M1. Stimulation in Area

3a activation revealed activations in Areas M1, 1, and 2, and that in

M1 showed activation predominantly in Areas 3a and 1. These areal

specific activation patterns parallel the cortico-cortical projections

previously mapped with gold standard anatomical methods. In addi-

tion, there were many examples of activations consistent with local

intra-areal connectivity. Patch spacing and size were consistent with

previous reports of anatomical patches (e.g., Négyessy et al., 2013)

and functionally imaged domains (e.g., Chen et al., 2003, Friedman

et al., 2004).

In light of these results, OIS in combination with ICMS provides

an additional tool for studying functional circuitry in vivo. The possi-

bility of revealing connections without sacrifice of the animal opens

doors for studying cortical connections more rapidly and/or with

repeated assessment of connections (e.g., changing connectivity over

development).

4.1 | Interareal connections revealed by ICMS-OISparallel known feedforward connections of SI

Feedforward and feedback connectivity patterns have been revealed

anatomically in somatosensory cortex of the macaque monkey

(Burton & Fabri, 1995; Darian-Smith et al., 1993; Jones et al., 1978;

Pons & Kaas, 1986; Shanks et al., 1985), marmoset (Krubitzer &

Kaas, 1990), squirrel monkey (Gharbawie et al., 2011; Liao

et al., 2013), and galago (Liao et al., 2013). It was unknown whether

combination of ICMS and OIS would reveal feedforward, feedback

functional connections, or anatomically specific activations. To

address this question we applied ICMS combined with OIS to reveal

F IGURE 7 Signal amplitudes differ at connected sites. (a,b) Two examples of activation amplitudes at connected sites in response tointracortical microstimulation (ICMS) in Area 3b (a) and M1 (b). ICMS: 100 μA. Right: Optical imaging of intrinsic signal (OIS) maps. White dashedlines: Approximate borders between areas. Color dots: Locations of significant response; response timecourses shown in graphs on left. In (a), thegraph illustrates that stimulation of Area 3b more strongly activates Area 1 than Areas M1, 3a, and 2. Gray line: Response at control site. In (b),M1 stimulation leads to strongest activation one M1 site (blue) and one Area 3a site (red), strong activation in other M1 (orange, gray) and oneArea 3a site (green), and weak activation in Area 1 site (peach). Shading: SE of the mean. Scale bar = 1 mm. p = posterior, l = lateral. Grayscale bar:in units of 10−4. Red bar: Electrical stimulation period [Color figure can be viewed at wileyonlinelibrary.com]

FRIEDMAN ET AL. 9

the functional connections of squirrel monkey somatosensory

cortex. Previous studies have shown that electrical microstimulation

produces patchy patterns of both focal and distant activations

(Sawaguchi, 1994; Stepniewska et al., 2011). Brock et al. (2013) had

shown that ICMS can achieve patchy label close to (within 1–2 mm)

of the stimulation site. Here, to examine interareal connections, our

focus was on stimulating primarily feedforward connection patterns.

To accomplish this goal, stimulation was applied in the superficial cor-

tical layers.

Although it is likely that ICMS activated both feedforward and

feedback projections, the imaged signal was likely biased towards

feedforward projections. Feedforward projections are known to be

more clustered and feedback connections more diffuse, which would

result in stronger focal signal in a population response revealed by

OIS. In studies of SI, Shanks et al. (1985) defined feedforward fibers

as originating in Layer III and terminating in all layers but most com-

monly terminating in Layer IV and deep Layer III. Feedforward projec-

tions were convergent and mediated by axons that were larger and

therefore would be expected to produce coherent bursts of neural

activity leading to strong OIS signals. In somatosensory cortex

feedforward, fibers were observed for Area 3b to Areas 3a, 1, and 2;

Area 1 to Areas 3a and 2; and Area 2 to Area 3a. In contrast, feedback

fibers were observed as less dense connections originating in Layers

V and VI and terminating in Layer I, II, top of Layer III and Layers V

and VI. Feedback fibers were observed for Area 1 to Area 3b; and

Area 2 to Areas 3b and 1. As the connectivity revealed following

ICMS of superficial layers are more consistent with a feedforward pat-

tern of connectivity (Figure 1 solid lines, Figure 4) and less so with

feedback (Figure 1, dotted lines), we suggest that feedforward net-

works were readily identified by this method.

We also note that the activations from a single stimulation site

differ in magnitude. This raises the question of whether this method

could be further refined to indicate the strength or the density of

feedforward projections from an area. As shown in the two examples

in Figure 7, the strongest responses were obtained in sites with

known robust connections. That is, stimulation of Area 3b led to a

response magnitude in Area 1 that was at least three times greater

than that of other activation sites (Areas 2, 3a, M1); stimulation of M1

resulted in strongest activations in two of nearby patches in M1 and

3a, slightly weaker signals in the other more lateral patches in M1 and

Area 3a, and weakest signals in Area 1. However, this will require fur-

ther examination using different stimulation intensities, electrophysi-

ology, modeling, and anatomical study.

4.1.1 | Functional connections of Area 3a

Only a few studies have performed functional anatomical mapping of

Brodmann Area 3a (Burton & Fabri, 1995; Huerta & Pons, 1990;

Huffman & Krubitzer, 2001). It is located deep in the central sulcus of

Old World primates, small in width, and neurons exhibit broad recep-

tive fields (in comparison to the size and receptive fields for Areas 3b

and 1), making it far more tedious to study (Kaas, 1983). The

functional maps obtained through the use of OIS in combination with

ICMS replicated the anatomically projections seen in previous studies

and showed significant M1 activation in response to stimulation of

Area 3a. Furthermore, the activity imaged displayed horizontal con-

nections in Area 3a (away from focal site of stimulation) and posterior

in Area 3b. Projections from Area 3a to Area 3b have been identified

in squirrel monkeys and galago monkeys, specifically in the

somatotopic representation of hand digits (Huffman &

Krubitzer, 2001; Liao et al., 2013). Huffman and Krubitzer (2001), in a

very thorough anatomical tracing map in marmoset, showed 3a

received input from somatosensory areas (especially dense input from

Area 3b), M1, SMA, and premotor areas ipsilaterally. Furthermore,

Guldin, Akbarian, and Grüsser (1992) reported in squirrel monkeys

very similar inputs to Area 3a. That our study was able to replicate

previously identified connectivity and identify other functionally con-

nected areas supports the novel combined use of OIS with ICMS is a

highly sensitive method of determining functional connectivity in the

cerebral cortex.

4.2 | Functional connections of MI at themesoscale

With ICMS of MI, we found activation in Area 3a and significant acti-

vation throughout M1 at both the site of stimulation as well as at

numerous surrounding intrinsic projection patches. These interareal

connections in M1 were patchy and typically less than 500 μm in size.

A previous study using microelectrode to functionally map of M1

zones involved in reach, defense, and grasp in squirrel monkeys to

identify sites for tracer injections revealed there is a number of paral-

lel networks between M1 and somatosensory cortex that likely act as

a sensory-effector modules (Gharbawie et al., 2011). The patchy con-

nections within MI that we observed might be the underpinning of

parallel networks within MI and indicate that ICMS with OIS has the

potential to be a useful approach for unraveling the functional cir-

cuitry of MI.

4.3 | Other methods of functional tract tracing

A challenge addressed in this study was to evaluate whether com-

bined usage of OIS with ICMS would reveal functional connectivity

in vivo at the level of cortical columns with the long-term goal of using

this novel approach to study the relationship between neuronal activ-

ity and cortical organization. Other in vivo methods such as fMRI uses

blood volume and oxygenation level dependent signals to monitor

neuronal activity, and have the advantages of being completely nonin-

vasive as well as being able to provide whole brain coverage. How-

ever, fMRI signals are slow and do not provide the spatial resolution

of OIS. Furthermore, MRI studies are expensive and cumbersome for

applications involving intracortical mapping (although cf. Xu

et al., 2019). EEG lacks the spatial resolution required for detailed

analysis of anatomy, yet does provide the temporal resolution useful

10 FRIEDMAN ET AL.

to monitor neuronal activity. Single and multiunit electrophysiology

have the best spatial and temporal resolution but only record from small

numbers of neurons. Two-photon imaging has the advantage of imaging

many neurons at single unit resolution and, when combined with

optogenetics, can reveal functional circuits with exquisite cell type speci-

ficity (e.g., Yamawaki, Radulovic, & Shepherd, 2016). We have also

shown that when optogenetic stimulation is applied in a column-targeted

fashion and combined with OIS, functionally specific local networks can

be revealed (Chernov, Friedman, Chen, Stoner, & Roe, 2018); however,

for primate studies, optogenetics requires additional, time-consuming

and cumbersome steps. When considering the advantages and disadvan-

tages of other approaches, the combined use of OIS with ICMS fills a

useful niche, by providing the visualization of functional connectivity in

neural circuits on the order of tens of millimeters with the spatial resolu-

tion of cortical domains while also providing capability of modulating cor-

tical activity in functionally precise manner through ICMS.

4.4 | Limitations of OIS and ICMS

While OIS with ICMS is a powerful novel technique for mapping local cir-

cuits, these two techniques do have some inherent limitations. Our

method of ICMS involved providing cortical stimulation through the pas-

sage of micro currents through a microelectrode inserted into cortex.

This approach is not cell type or neural structure specific because of its

nature of stimulating a volume of tissue (Tehovnik, Tolias, Sultan, Slo-

cum, & Logothetis, 2006). Some of these limitations can be overcome by

using other methods of cortical stimulation, like optogenetics or infrared

neural stimulation (Cayce et al., 2014; Wells et al., 2005; Xu et al., 2019).

Even with these limitations, ICMS combined with OIS revealed functional

cortical interareal projections similar to that reported with anatomical

methods. OIS cannot be used to image deep cortical areas and regardless

of how neural activity is evoked requires invasive neurosurgery to pro-

vide an optical window. Furthermore, OIS visualizes and only indirectly

reports on neural activity by measuring the hemodynamic response asso-

ciated with increased neural activity (Bonhoeffer & Grinvald, 1996).

Increased neural activity causes an increased consumption of oxygen

and therefore an initial increase in deoxyhemoglobin followed by a

rebound in oxyhemoglobin as local capillaries respond with increased

blood flow. Consequently, the temporal resolution of OIS is too slow to

study cortical processing dynamics and the spatial resolution is more dif-

fuse than the underlying neural activity as it incorporates the summed

activity of astrocytes, spiking neurons, and subthreshold neuronal

changes in synaptic potential. Despite these limitations, OIS readily

images cortical domains at the mesoscale (Roe, 2010).

4.5 | Conclusions: An approach to study functionalmicrocircuitry

To examine whether ICMS can reveal cortical connectivity at the level

of cortical domains, we conducted OIS during ICMS at functionally

identified locations. While OIS paired with ICMS can successfully

reveal cortico-cortical projections, it would be of even greater value to

be able to apply this approach to study functional microcircuitry. Previ-

ous studies in macaque primary visual cortex have shown that focal

injections of anatomical tracers reveal connections with focal patches

nearby (e.g., Malach, Amir, Harel, & Grinvald, 1993) and that such con-

nectivity is functionally specific (e.g., Ts'o & Gilbert, 1988; Chernov

et al., 2018). Use of ICMS could be used to image the functional pro-

jections of single functional domains such as V1 specific orientation

domains that project to like orientation domains in V1 or in V2. It is

thought that modularity is a common organizational feature of cerebral

cortex and this approach could be used to study the circuitry of senso-

rimotor networks (e.g., Stepniewska et al., 2011). One way to causally

investigate microcircuity would be to combine ICMS and OIS with nat-

ural stimulation and compare the influence of ICMS on behavior and

cortical activity.

ACKNOWLEDGMENTS

U.S. National Institutes of Health NS044375 and NS093998 (to A. W.

R. and R. M. F.), CTSA NIH Grant TL1TR000447 (to K. K. M.),

NS105697 (to O.A.G.). The following funds to A. W. R. from Chinese

funding sources were NOT used to support this research: National

Key R&D Program of China 2018YFA0701400, National Natural Sci-

ence Foundation 31627802, Zhejiang Provincial Department of Sci-

ence and Technology 2020C03004, China NSFC-US NIH Cooperative

Biomedical Grant NSFC Grant 8191101288. There is no overlap.

A. W. R.'s acknowledgment: I give enormous thanks to Jack Pettigrew

for showing me that “you've got to trust your instincts.” I met Jack at

a raucous party at SFN, where after 20 min of knowing me, he offered

me a research position. As I would soon learn, this was typical Jack,

someone with no qualms about making decisions because he felt good

about it. I ended up accepting this wonderful offer and flew off with

my son and a giant suitcase to Brisbane. So many wonderful and

memorable adventures followed…marmosets and flying foxes…

bushwalking, Glass Mountains, Heron Island, diving to study fish

vision, and unending Jack enthusiasm. Jack's example taught me to be

bold and to take advantage of the Vision, Touch and Hearing Research

Centre opportunities that feel right. His concept of interdisciplinary

pursuits at VTHRC influenced me greatly and planted the seeds for

my own Interdisciplinary Institute of Neuroscience and Technology in

China (www.ziint.zju.edu.cn). I will be forever grateful to Jack for help-

ing me to think big.

CONFLICT OF INTEREST

The authors declare that they have no conflict of interest.

AUTHOR CONTRIBUTIONS

Anna Wang Roe and Robert M. Friedman designed the study.

Katherine A. Morone, Omar A Gharbawie, Robert M. Friedman, and

Anna Wang Roe performed research. Katherine A. Morone and

Robert M. Friedman analyzed the data. Robert M. Friedman and Anna

Wang Roe wrote the manuscript.

FRIEDMAN ET AL. 11

DATA AVAILABILITY STATEMENT

Data that support the findings of this study are available from the

corresponding author upon request.

ETHICS STATEMENT

All applicable international, national, and/or institutional guidelines for

the care and use of animals were followed. Animal care and surgeries

were performed according to NIH (National Institute of Health) regu-

lations and were in compliance with and approved by the Institutional

Animal Care and Use Committee of Vanderbilt University.

ORCID

Robert M. Friedman https://orcid.org/0000-0002-7391-8352

Omar A. Gharbawie https://orcid.org/0000-0002-2744-9305

Anna Wang Roe https://orcid.org/0000-0003-4146-9705

REFERENCES

Ashaber, M., Pálfi, E., Friedman, R. M., Palmer, C., Jákli, B., Chen, L. M., …Négyessy, L. (2014). Connectivity of somatosensory cortical area

1 forms an anatomical substrate for the emergence of multifinger

receptive fields and complex feature selectivity in the squirrel monkey

(Saimiri sciureus). Journal of Comparative Neurology, 522(8),

1769–1785.Bonhoeffer, T., & Grinvald, A. (1996). Optical imaging based on intrinsic

signals: The methodology. In A. W. Toga & J. C. Mazziotta (Eds.), Brain

mapping: The methods (Vol. 1996, pp. 55–97). San Diego, CA: Aca-

demic Press.

Brock, A. A., Friedman, R. M., Fan, R. H., & Roe, A. W. (2013). Optical imag-

ing of cortical networks via intracortical microstimulation. Journal of

Neurophysiology, 110(11), 2670–2678.Burton, H., & Fabri, M. (1995). Ipsilateral intracortical connections of phys-

iologically defined cutaneous representations in areas 3b and 1 of

macaque monkeys: Projections in the vicinity of the central sulcus.

Journal of Comparative Neurology, 355(4), 508–538.Cayce, J. M., Friedman, R. M., Chen, G., Jansen, E. D., Mahadevan-

Jansen, A., & Roe, A. W. (2014). Infrared neural stimulation of primary

visual cortex in non-human primates. NeuroImage, 84, 181–190.Cerkevich, C. M., & Kaas, J. H. (2019). Corticocortical projections to area

1 in squirrel monkeys (Saimiri sciureus). European Journal of Neurosci-

ence, 49(8), 1024–1040.Chen, L. M., Friedman, R. M., Ramsden, B. M., LaMotte, R. H., & Roe, A. W.

(2001). Fine-scale organization of SI (area 3b) in the squirrel monkey

revealed with intrinsic optical imaging. Journal of Neurophysiology, 86

(6), 3011–3029.Chen, L. M., Friedman, R. M., & Roe, A. W. (2003). Optical imaging of a tac-

tile illusion in area 3b of the primary somatosensory cortex. Science,

302(5646), 881–885.Chernov, M. M., Friedman, R. M., Chen, G., Stoner, G. R., & Roe, A. W.

(2018). Functionally specific optogenetic modulation in primate visual

cortex. Proceedings of the National Academy of Sciences of the United

States of America, 115(41), 10505–10510.Darian-Smith, C., Darian-Smith, I., Burman, K., & Ratcliffe, N. (1993). Ipsi-

lateral cortical projections to areas 3a, 3b, and 4 in the macaque mon-

key. Journal of Comparative Neurology, 335(2), 200–313.Friedman, R. M., Chen, L. M., & Roe, A. W. (2004). Modality maps

within primate somatosensory cortex. Proceedings of the National

Academy of Sciences of the United States of America, 101(34),

12724–12729.Friedman, R. M., Chen, L. M., & Roe, A. W. (2008). Responses of areas 3b

and 1 in anesthetized squirrel monkeys to single- and dual-site stimu-

lation of the digits. Journal of Neurophysiology, 100(6), 3185–3196.

Gharbawie, O. A., Stepniewska, I., & Kaas, J. H. (2011). Cortical connec-

tions of functional zones in posterior parietal cortex and frontal cortex

motor regions in new world monkeys. Cerebral Cortex, 21(9),

1981–2002.Graziano, M. S., Aflalo, T. N., & Cooke, D. F. (2005). Arm movements

evoked by electrical stimulation in the motor cortex of monkeys. Jour-

nal of Neurophysiology, 94(6), 4209–4223.Graziano, M. S., Taylor, C. S., & Moore, T. (2002). Complex movements

evoked by microstimulation of precentral cortex. Neuron, 34(5),

841–851.Guldin, W. O., Akbarian, S., & Grüsser, O. J. (1992). Cortico-cortical con-

nections and cytoarchitectonics of the primate vestibular cortex: A

study in squirrel monkeys (Saimiri sciureus). Journal of Comparative Neu-

rology, 326(3), 375–401.Huerta, M. F., & Pons, T. P. (1990). Primary motor cortex receives input

from area 3a in macaques. Brain Research, 537(1–2), 367–371.Huffman, K. J., & Krubitzer, L. (2001). Area 3a topographic organization

and cortical connections in marmoset monkeys. Cerebral Cortex, 11(9),

849–867.Jones, E. G., Coulter, J. D., & Hendry, S. H. (1978). Intracortical connectiv-

ity of architectonic fields in the somatic sensory, motor and parietal

cortex of monkeys. Journal of Comparative Neurology, 181(2),

291–347.Kaas, J. H. (1983). What, if anything, is SI? Organization of first somato-

sensory area of cortex. Physiological Reviews, 63(1), 206–231.Krubitzer, L. A., & Kaas, J. H. (1990). The organization and connections of

somatosensory cortex in marmosets. Journal of Neuroscience, 10(3),

952–974.Liao, C. C., Gharbawie, O. A., Qi, H., & Kaas, J. H. (2013). Cortical connec-

tions to single digit representations in area 3b of somatosensory cor-

tex in squirrel monkeys and prosimian galagos. Journal of Comparative

Neurology, 521(16), 3768–3790.Malach, R., Amir, Y., Harel, M., & Grinvald, A. (1993). Relationship between

intrinsic connections and functional architecture revealed by optical

imaging and in vivo targeted biocytin injections in primate striate cor-

tex. Proceedings of the National Academy of Sciences of the United

States of America, 90(22), 10469–10473.Murphey, D. K., & Maunsell, J. H. (2007). Behavioral detection of electrical

microstimulation in different cortical visual areas. Current Biology, 17

(10), 862–867.Négyessy, L., Pálfi, E., Ashaber, M., Palmer, C., Jákli, B., Friedman, R. M., …

Roe, A. W. (2013). Intrinsic horizontal connections process global tac-

tile features in the primary somatosensory cortex: Neuroanatomical

evidence. Journal of Comparative Neurology, 521(12), 2798–2817.Pálfi, E., Zalányi, L., Ashaber, M., Palmer, C., Kántor, O., Roe, A. W., …

Négyessy, L. (2018). Connectivity of neuronal populations within and

between areas of primate somatosensory cortex. Brain Structure and

Function, 223(6), 2949–2971.Pons, T. P., & Kaas, J. H. (1986). Corticocortical connections of area 2 of

somatosensory cortex in macaque monkeys: A correlative anatomical

and electrophysiological study. Journal of Comparative Neurology, 248

(3), 313–335.Roe, A. W. (2010). Imaging the brain with optical methods. New York, NY:

Springer-Verlag.

Romo, R., Hernández, A., Zainos, A., Brody, C. D., & Lemus, L. (2000). Sens-

ing without touching: Psychophysical performance based on cortical

microstimulation. Neuron, 26(1), 273–278.Salzman, C. D., Britten, K. H., & Newsome, W. T. (1990). Cortical

microstimulation influences perceptual judgments of motion direction.

Nature, 346(6280), 174–177.Sawaguchi, T. (1994). Modular activation and suppression of neocortical

activity in the monkey revealed by optical imaging. Neuroreport, 6(1),

185–189.Shanks, M. F., Pearson, R. C., & Powell, T. P. (1985). The ipsilateral cortico-

cortical connexions between the cytoarchitectonic subdivisions of the

12 FRIEDMAN ET AL.

primary somatic sensory cortex in the monkey. Brain Research, 356(1),

67–88.Stepniewska, I., Fang, P. C., & Kaas, J. H. (2009). Organization of the pos-

terior parietal cortex in galagos: I. functional zones identified by

microstimulation. Journal of Comparative Neurology, 517(6), 765–782.Stepniewska, I., Friedman, R. M., Gharbawie, O. A., Cerkevich, C. M.,

Roe, A. W., & Kaas, J. H. (2011). Optical imaging in galagos reveals

parietal-frontal circuits underlying motor behavior. Proceedings of the

National Academy of Sciences of the United States of America, 108(37),

E725–E732.Stepniewska, I., Preuss, T. M., & Kaas, J. H. (1993). Architectonics,

somatotopic organization, and ipsilateral cortical connections of the

primary motor area (MI) of owl monkeys. Journal of Comparative Neu-

rology, 330(2), 238–271.Sur, M., Nelson, R. J., & Kaas, J. H. (1982). Representations of the body sur-

face in cortical areas 3b and 1 of squirrel monkeys: Comparisons with

other primates. Journal of Comparative Neurology, 211(2), 177–192.Tehovnik, E. J., & Slocum, W. (2009). Background luminance affects the

detection of microampere currents delivered to macaque striate cor-

tex. European Journal of Neuroscience, 30(2), 263–271.Tehovnik, E. J., Tolias, A. S., Sultan, F., Slocum, W. M., & Logothetis, N. K.

(2006). Direct and indirect activation of cortical neurons by electrical

microstimulation. Journal of Neurophysiology, 96(2), 512–521.Tolias, A. S., Sultan, F., Augath, M., Oeltermann, A., Tehovnik, E. J.,

Schiller, P. H., & Logothetis, N. K. (2005). Mapping cortical activity

elicited with electrical microstimulation using FMRI in the macaque.

Neuron, 48(6), 901–911.

Ts'o, D. Y., & Gilbert, C. D. (1988). The organization of chromatic and spa-

tial interactions in the primate striate cortex. Journal of Neuroscience, 8

(5), 1712–1727.Wang, Z., Chen, L. M., Négyessy, L., Friedman, R. M., Mishra, A.,

Gore, J. C., & Roe, A. W. (2013). The relationship of anatomical and

functional connectivity to resting-state connectivity in primate

somatosensory cortex. Neuron, 78(6), 1116–1126.Wells, J., Kao, C., Mariappan, K., Albea, J., Jansen, E. D., Konrad, P., &

Mahadevan-Jansen, A. (2005). Optical stimulation of neural tissue

in vivo. Optics Letters, 30(5), 504–506.Xu, A. G., Qian, M., Tian, F., Xu, B., Friedman, R. M., Wang, J., … Roe, A. W.

(2019). Focal infrared neural stimulation with high-field functional

MRI: A rapid way to map mesoscale brain connectomes. Science

Advances, 5(4), eaau7046.

Yamawaki, N., Radulovic, J., & Shepherd, G. M. (2016). A corticocortical

circuit directly links retrosplenial cortex to M2 in the mouse. Journal of

Neuroscience, 36(36), 9365–9374.

How to cite this article: Friedman RM, Morone KA,

Gharbawie OA, Roe AW. Mapping mesoscale cortical

connectivity in monkey sensorimotor cortex with optical

imaging and microstimulation. J Comp Neurol. 2020;1–13.

https://doi.org/10.1002/cne.24918

FRIEDMAN ET AL. 13