Simultaneous Fmri and Fast-Scan Cyclic Voltammetry Bridges Oxygenation and Neurotransmitter Dynamics Across 2 Spatiotemporal Scales

Home , Reduction potential

bioRxiv preprint doi: https://doi.org/10.1101/2021.05.28.446169; this version posted May 30, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

1 Simultaneous fMRI and fast-scan cyclic voltammetry bridges oxygenation and neurotransmitter dynamics across 2 spatiotemporal scales

3 Lindsay R Waltona,b,c*

4 Matthew Verbera,b,d

5 Sung-Ho Leea,b,c

6 Tzu-Hao Chaoa,b,c

7 R. Mark Wightmand

8 Yen-Yu Ian Shiha,b,c*

9 aCenter for Animal MRI, University of North Carolina at Chapel Hill, Chapel Hill, NC USA

10 bBiomedical Research Imaging Center, University of North Carolina at Chapel Hill, Chapel Hill, NC USA

11 cDepartment of Neurology, University of North Carolina at Chapel Hill, Chapel Hill, NC USA

12 dDepartment of Chemistry, University of North Carolina at Chapel Hill, Chapel Hill, NC USA

13 *Corresponding authors contact information: [email protected] (LR Walton); [email protected] (YY 14 Shih)

15 Abstract:

16 The vascular contributions of neurotransmitters to the hemodynamic response are gaining more attention in 17 neuroimaging studies, as many neurotransmitters are vasomodulatory. To date, well-established electrochemical 18 techniques that detect neurotransmission in high magnetic field environments are limited. Here, we propose an 19 experimental setting enabling simultaneous fast-scan cyclic voltammetry (FSCV) and blood oxygenation-dependent 20 functional magnetic imaging (BOLD fMRI) to measure both local tissue oxygen and dopamine responses, and global 21 BOLD changes, respectively. By using MR-compatible materials and the proposed data acquisition schemes, FSCV 22 detected physiological analyte concentrations with high spatiotemporal resolution inside of a 9.4 T MRI bore. We 23 found that tissue oxygen and BOLD correlate strongly, and brain regions that encode dopamine amplitude 24 differences can be identified via modeling simultaneously acquired dopamine FSCV and BOLD fMRI time-courses. 25 This technique provides complementary neurochemical and hemodynamic information and expands the scope of 26 studying the influence of local neurotransmitter release over the entire brain.

28 Introduction

29 Coupling between neuronal activity and cerebral hemodynamic changes ensures that healthy energetic 30 homeostasis is maintained within the active brain1,2. Classic blood oxygenation-level dependent functional magnetic 31 resonance imaging (BOLD fMRI) data interpretation assumes that hemodynamic responses to stimuli are 32 proportional to neuronal activity, such that hemodynamic response functions (HRF)1 can be derived to translate 33 neuronal activity into BOLD signal predictions3. However, these assumptions are not always valid. Neurotransmitters 34 have been receiving more attention as contributors to the hemodynamic response, as many are vasomodulatory 35 (e.g., dopamine, nitric oxide, norepinephrine)1,4 and dysregulated in pathologies where neurovascular coupling is 36 also dysregulated, such as schizophrenia and Parkinson’s1,5. It is paramount to understand how specific 37 neurotransmitters influence vascular responses and accurately interpret neuroimaging data throughout the brain. bioRxiv preprint doi: https://doi.org/10.1101/2021.05.28.446169; this version posted May 30, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

38 Dopamine is a vasomodulatory neurotransmitter concentrated in the striatum4,6, and plays a strong role in 39 motivation, movement, addiction, learning, and reward prediction7. Dopamine receptors are located on 40 microvasculature within the striatum, thalamus, and cortex, as well as on astrocytes6,8. Though dopamine has been 41 studied intensely for decades, few studies have delved into its vascular modulation properties. Amphetamine and 42 phencyclidine challenges show a linear relationship between evoked striatal hemodynamic changes and dopamine9– 43 11, and D1- and D2-like dopamine receptor agonists show diametrically opposed vascular responses6. Though fMRI 44 responses to cocaine and amphetamine challenge have been identified10, existing HRF models do not consider 45 dopaminergic influence from shorter release events where higher affinity, presynaptic receptor binding would have 46 more of an impact9. Dissecting how dopamine influences striatal vasculature is an important first step into 47 understanding how whole-brain hemodynamics are affected by dopaminergic neurotransmission.

48 Popular methods of monitoring dopamine or other neurotransmitter kinetics in vivo are electrochemical, 49 such as fast-scan cyclic voltammetry (FSCV) and amperometry12, though advances are being made with fluorescent 50 sensor and magnetic resonance-based detection13–16, among others17. FSCV is especially desirable because it is 51 chemically selective and quantifiable, and multiple, different neurotransmitters can be detected using well- 52 established modifications12,17. It has high spatiotemporal resolution and minimal tissue damage, especially versus 53 competing techniques like positron emission tomography and microdialysis17. FSCV translates directly to human 54 use18–20, a challenge for fluorescent sensor-based techniques, and is therefore a promising method for studying 55 neurotransmission during human behavior19, improving therapeutic deep-brain stimulation21, and interpreting 56 human pharmacological fMRI studies22. Though FSCV can detect both tonic23 and phasic dopamine changes, and 57 oxygen changes24–27, its high spatial resolution limits the scope of interpretation. To reveal the influence of local 58 dopamine release on whole brain circuits requires pairing FSCV with fMRI to acquire neurotransmitter and brain- 59 wide hemodynamic data simultaneously.

60 In this work, we develop and characterize a simultaneous FSCV-fMRI technique. We address the challenges 61 of developing an MR-compatible FSCV recording system, circumventing high-frequency electronic noise from MR 62 imaging gradients, and synchronizing multimodal data acquisition. We perform in vivo FSCV-fMRI during an oxygen 63 inhalation challenge using an oxygen-sensitive FSCV waveform and electrical deep brain stimulation using both 64 oxygen- and dopamine-sensitive FSCV waveforms at an electrode implanted into the nucleus accumbens (NAc). To 65 compare hemodynamic measurements at different spatial scales, we collected tissue oxygen and BOLD fMRI data 66 concurrently. We derive an HRF from high-resolution dopamine data acquired during simultaneous fMRI, and for 67 the first time, we demonstrate that this simultaneous FSCV-fMRI recording platform can identify brain regions that 68 encode dopamine amplitude changes. This method should contribute significantly to the understanding of local 69 neurotransmission on dynamic changes of brain activity.

70 71 Results

72 fMRI and FSCV compatibility

73 A major step to achieve simultaneous FSCV-fMRI is ensuring that FSCV material components do not produce 74 imaging artifacts. Traditional glass capillary microelectrodes, like those used for FSCV, produce minor artifacts28, but 75 polyimide-fused silica further enhances MR-compatibility29 and is also used with FSCV microelectrodes (Fig.1A)30. 76 Both in agarose phantoms and in vivo, polyimide-fused silica electrodes showed minimal MR artifacts (Fig.1A-C); 77 however, modifications were necessary to maximize MR compatibility. Standard fabrication connects the carbon 78 fiber directly to a connection pin via silver epoxy, which provides structural support within the headcap and makes 79 the electrode suitable for chronic implantation30. Unfortunately, commercially available pins contain trace amounts 80 of nickel, a magnetic material with large susceptibility artifacts31. Here, we silver epoxied the carbon fiber directly 81 onto silver wires. Following implantation, 1 cm of silver wire from both working and Ag/AgCl reference electrodes 82 was left exposed above the headcap during recovery (Fig.1D). To protect the free-standing wires, plastic shields were 83 cemented to the headcaps (Fig.1D). We performed flow-through analysis to assess whether these modifications 84 affected sensitivity. In our hands, the calibration factors for oxygen and dopamine on the oxygen-sensitive waveform 85 were -0.19 nA/µM/100 µm and 4.8 nA/µM/100 µm, respectively (Supplemental Fig.1). Both calibration factors fall 86 between values reported for glass microelectrodes (normalized to 100 µm)24. On the dopamine-sensitive waveform, 87 dopamine sensitivity was 34 nA/µM/100 µm, in agreement with the literature30 (Supplemental Fig.1). These data 88 indicate that our modified polyimide-fused silica electrodes are MR compatible and can detect analytes of interest 89 with high sensitivity.

A B Hz C D 50 0 -50 40 Stimulating Electrode Working Electrode SE (RARE) SE (RARE) Hz

Ag/AgCl

GE (FLASH) Hz 0 -40 50

50 GE (EPI) 90 - 91 Fig.1. Modifying FSCV materials improves MRI compatibility. (A) Left: Polyimide-fused silica capillary 92 microelectrode (scale bar=100 µm). Right: Electrode axial cross-section scanned inside an agarose phantom using 93 spin echo (SE) rapid acquisition with relaxation enhancement (RARE), and gradient echo (GE) fast low angle shot 94 (FLASH) sequences (scale bar=1 mm). (B) Axial view of the magnetic field map showing inhomogeneities induced 95 by the microelectrode materials in an agar phantom. Scale bar=200 µm. Frequency profiles are taken from the 96 dashed line cross-section of the same color (horizontal=blue, vertical=red). (C) Microelectrodes implanted in the 97 rat NAc in vivo, indicated with arrows, during a SE-based RARE anatomical scan and a GE-based echo planar 98 imaging (EPI) functional scan. (D) Rat headcaps were fitted with plastic shields to protect loose silver wire 99 connections to FSCV working and reference electrodes.

100 FSCV is a differential technique that detects electrooxidation and reduction reactions on the scale of nA or 101 less (Supplementary Fig.2), so electrical noise must be minimized to maximize FSCV sensitivity. First, we placed a 102 dummy cell (vide infra) inside the MRI bore to quantify ambient electrical noise. Sub-nA noise was achieved using 103 minimal grounding and braided cables with the gradients disabled, but a 4th order Bessel low-pass filter with a 2 kHz 104 cut-off frequency was necessary to remove the high amplitude, 50 kHz power amplifier noise present with enabled 105 gradients (Supplementary Fig.3). To ensure that noise levels were low enough to detect physiologically relevant 106 concentrations of analyte, we performed a flow-injection experiment using dopamine boluses. A flow-injection cell 107 was placed inside the MRI bore on a 3D printed holder attached to the animal sled (Fig.2A). With repeated boluses, 108 there was a significant main effect of the recording environment on the signal-noise-ratio (SNR; F(1.8, 5.3)=40.9, 109 p=0.0007, Fig.2B). Boluses recorded with the gradient enabled and without low-pass filtering had significantly lower 110 SNR (6.46±2.63) than those recorded with low-pass filtering (disabled with filter: 135.9±13.9, enabled with filter: 111 89.5±16.0; Tukey’s post-hoc, p=0.003 and p=0.03, respectively; Fig.2B). Our chosen filter did not distort bolus 112 amplitudes or shapes (Fig.2C-D), making it an effective method of removing the majority of electrical noise within 113 the MRI bore.

114 Simultaneous fMRI and FSCV

115 It is well-documented with other electrical recoding modalities (e.g., electrophysiology) that echo planar 116 imaging (EPI)-based fMRI produces intrusive, high-amplitude noise during encoding, which can be removed via 117 principal component regression, algorithms, data acquisition interleaving, or deleting encoding portions outright3,32– 118 34. FSCV detects electrochemically active analytes via periodic electrooxidation/reduction voltage sweeps 119 (Supplementary Fig.2). As EPI-based fMRI also delivers RF and gradient pulses periodically, ideally FSCV waveforms 120 could interleave with active RF and gradient pulsing. A dummy cell was placed inside the bore during fMRI acquisition 121 to observe the encoding artifact, which was approximately 2 VP-P, 55 ms, and repeated with each slice acquisition 122 (Fig.2E-F). Any temporal overlap between the waveform application and EPI encoding resulted in high amplitude 123 interference (Fig.2E). To interleave FSCV and EPI recordings, MRI protocols were set to wait for a TTL trigger per slice 124 and all timings were coordinated within the FSCV data acquisition software. A 10-slice EPI scan using a 1000 ms 125 repetition time (TR; i.e., 100 ms between EPI artifacts) could interleave with FSCV waveforms applied at 10 Hz (i.e., 126 100 ms between waveform scans) for 10:1 FSCV:fMRI temporal resolution acquisition. When FSCV data acquisition 127 begins, the software sends a high TTL voltage to the MRI. After each FSCV waveform is applied, the TTL remains high 128 for an additional 3 ms before a 10 ms, low voltage TTL triggers EPI slice acquisition (Fig.2F). By using TTL trigger- 129 dependent interleaving, one or more FSCV scans can be collected per EPI slice without encoding gradient 130 interference. A Working B 200 ✱✱ Electrode 6-valve flow injector Ag/AgCl ✱ Injection Loop 150 reference Buffer Waste 100 SNR 50

0 Flow

*Instrument sizes are not to scale Disabled Enabled Enabled +Filter

C No filters No filters Low-Pass Filtered D Gradient Disabled Gradient Enabled Gradients Enabled -0.4 20 nA 100 nA 20 nA

-0.4 V +1.3

+1.3 Disabled Enabled

(V) vs. Ag/AgCl Enabled + Filters app

E -0.4 -13 nA -67 nA -13 nA 0 140 s 0 140 s 0 140 s 0 60 120 s E -0.4 40 nA 30 Encoding FSCV Waveform Start

(V) +1.3 15 app E Current (nA) 0 -0.4 -25 nA 0 30 s 0 10 20 30 s 0 50 100 150 200 250 ms F -0.4 15 nA 30 Raw Input Voltage TTL Trigger

(V) +1.3 15 app E

Current (nA) 0 -0.4 -10 nA 0 30 s 0 10 20 30 s 131 0 50 100 150 200 250 ms 132 Fig.2. Software and hardware modifications for simultaneous FSCV-fMRI. (A) Within-bore flow-injection analysis 133 requires the 6-port injection valve to be placed outside of the magnet at a safe distance. A fresh dopamine bolus is 134 injected and flowed through to the electrode inside of the MRI bore for detection. (B) SNR for a physiologically 135 relevant dopamine concentration. Averaged triplicate measures are shown (n=4 electrodes; Tukey’s post-hoc 136 *p<0.05, **p<0.01). (C) Low-pass digital Bessel filters remove the high-frequency noise from enabled MRI gradients 137 without distorting FSCV signals. Color plots indicate flow-injection dopamine boluses within the MRI bore. Cyclic 138 voltammograms obtained from the maximum bolus current are overlaid in white. (D) Dopamine bolus injection time- 139 courses from the dopamine oxidation potential in C. Scale bar=15 nA. (E) Dummy cell data collected inside the MRI 140 bore. From left to right: Color plots, current time-courses acquired at 10 Hz, and raw voltage time-courses acquired 141 at 100 kHz (scale bar= 4V) with corresponding magnification insets (vertical scale bar=1 V, horizontal scale bar=20 142 ms). Large artifacts appear when FSCV waveform applications and EPI encoding overlap temporally, but (F) 143 interleaving the signals with TTL triggers avoids this interference. Time-courses are taken from the dopamine 144 oxidation potential, indicated by a white dashed line in the color plots. Red bars indicate EPI scan duration.

145 In Vivo Oxygen Challenge in the NAc

146 Previous FSCV studies have shown that pharmacologically manipulating neurotransmission affects oxygen 147 dynamics, which are taken to represent hemodynamic changes24,25,27; however, the spatial resolution of FSCV limits 148 the interpretation to highly localized microenvironments. Using an oxygen-sensitive waveform (Fig.3A), we 149 simultaneously acquired oxygen-related signals from FSCV and fMRI, in vivo, at spatial scales of different magnitudes 150 (Fig.3B). We first administered a 100% oxygen breathing challenge (Fig.3A) and compared evoked responses within 151 subject-level regions of interest (ROI) at the electrode tip implanted in the NAc (Supplementary Fig.4). Both fMRI 152 and FSCV oxygen changes increased during hyperoxia and fell to baseline afterwards (Fig.3C, n=6 subjects). The 40 s 153 delay before oxygen levels change corresponds to the transit time between the gas flow source and the ventilated 154 subject. FSCV data, binned to match fMRI temporal resolution (TR=1000 ms), had significantly higher SNR during 155 hyperoxia (FSCV: 47.7±13.8, fMRI: 4.4±0.4; t(17)=8.24, p<0.0001, n=18 events; Fig.3D). The averaged event time- 156 courses highly correlated (r=0.85, p<0.0001, n=18 events; Fig.3E) in agreement with the literature35; however, there 157 were differences between the observed oxygen kinetics (Fig.3F, n=18 events). The FSCV response preceded the 158 highest correlated BOLD responses by 5±2 s, according to time shift analysis (n=18 events; Fig.3F). The clearance 159 time from the maximum response to half-max was significantly longer in BOLD fMRI versus oxygen FSCV (n=6 160 subjects, FSCV: 20.7±3.0 s, fMRI: 33.7±3.4 s; t(5)=4.48, p=0.007; Fig.3G), but there was no difference in half-max rise 161 times. The data trended towards a longer fMRI versus FSCV full-width at half-max (FWHM; n=6 subjects, t(5)=2.21, 162 p=0.08; Fig.3G), but did not reach statistical significance. Though FSCV and fMRI detect highly correlated changes 163 during an oxygen inhalation challenge, the difference in clearance kinetics indicate that these data are 164 complementary. As expected of a systemic, uniform increase in oxygen from an inhalation challenge, induced BOLD 165 responses and FSCV oxygen highly correlated spatially across the brain (one sample t-test, p=0.05 compared to other 166 voxels, followed by group statistics with n=6 subjects; Fig.3H). This analysis demonstrates the feasibility of using 167 local FSCV signals for correlation/regression analysis at a whole brain scale.

✱✱✱✱ A B C D E 3 Pearson’s r = 0.85 0.8 V BOLD 300 p<0.0001 fMRI 100 2 0.0 V m μ

1000 1 -1.4 V % BOLD fMRI Oxygen SNR 50

100% O2 m FSCV μ 0 60 s

320 FSCV 60 s 180 s X 3 0 -1 320 μm 0 200 400 600 800 s FSCV fMRI -20 0 20 40 60 80 FSCV Oxygen (µM) F G H 3 100 0.8 100 100 FSCV fMRI fMRI leads leads 75 -1 2 10 75 FSCV

Oxygen ( 0.6

P-value ✱✱ 50 10-2 2.4 mm 1.4 mm 1 0.4 50

25 µ -3 M)

10 (s) Time % BOLD 0 0 0.2 25 10-4 Correlation Coefficient -25 0.4 mm -0.4 mm -1 0.0 10-5 0 0 100 200 s -20 -10 0 10 20 FWHM Half-MaxHalf-Max 0 0.7 Rise Fall Time Shift (s) Pearson’s r 168 169 Fig.3. BOLD fMRI and oxygen FSCV time-courses in the rat NAc highly correlate during oxygen breathing challenges. 170 (A) Paradigm schematic. Top: An oxygen-sensitive waveform was repeated throughout the recording session to 171 detect oxygen changes. Red indicates the oxidation potential sweep, green indicates the reduction potential sweep. 172 Bottom: Oxygen inhalation event paradigm. When not inspiring pure oxygen, subjects inspired medical air. (B) A 173 single BOLD fMRI voxel compared to the volume sampled by FSCV, to scale. (C) BOLD fMRI and oxygen FSCV time- 174 courses show signal increases during oxygen inhalation periods (n=6 subjects). Blue scale bar=2% signal change. 175 Black scale bar=50 µM oxygen. (D) SNR from FSCV and fMRI NAc ROI oxygen inhalation time-courses (n=18 events, 176 ROIs=36±3 voxels). (E) Averaged BOLD fMRI and oxygen FSCV time-course correlation (n=18 events). (F) Averaged 177 oxygen challenge events from C (left) and corresponding time shift-correlation plot (right). (G) Kinetic parameters 178 derived from subject-averaged time-courses, where half-max rise is the time from t=0 to half-max, and half-max fall 179 is the time from t=30 s after the 100% oxygen ends (indicated in f as a blue dashed line) to the half-max clearance 180 value (n=6 subjects). (H) Group-level correlation between BOLD fMRI and concurrently acquired oxygen FSCV time- 181 courses (p=0.05). Voxels with correlation coefficient <0.3 are excluded. Grey boxes in A, C, and F indicate times 182 where gas inhalation lines were switched to 100% oxygen. Error bars are ±SEM or +SEM, for visual clarity. Ratio- 183 paired t-test **p<0.01, ****p<0.001.

184 In Vivo Electrical VTA Stimulation

185 BOLD fMRI and oxygen FSCV map stimulus-evoked responses in the brain on different scales2,24,25. To 186 demonstrate how FSCV-fMRI can study specific stimulus-evoked responses, we electrically stimulated the ventral 187 tegmental area (VTA; Fig.4A). Evoked responses had higher SNR with FSCV versus fMRI responses within the NAc 188 ROI (FSCV: 17.4±3.2, fMRI: 3.3±0.6; t(17)=5.40, p<0.0001, n=18 events; Fig.4B). Stimulations evoked positive oxygen 189 FSCV and BOLD fMRI changes, which are characteristic of electrical VTA stimulation24 (Fig.4C). The averaged time- 190 courses highly correlated (r=0.83, p<0.0001, n=18 events; Fig.4D). FSCV time-courses lagged the maximally 191 correlated fMRI time-courses by 0.8±0.8 s, according to time shift analysis (Fig.4E). While traditional block-design 192 general linear model (GLM) analysis extracted broad areas of VTA-induced BOLD responses (Supplementary Fig.5), 193 similar to those reported in the literature7,36,37, the brain areas showing the highest oxygen FSCV time-course 194 correlations were more tightly localized to clusters in ipsilateral NAc, claustrum, and insular cortex, bilateral 195 prelimbic and cingulate cortices, and septum (one sample t-test cutoff at p=0.05 compared to other voxels in the 196 brain, followed by group GLM analysis with n=6 subjects; Fig.4F). FSCV oxygen measurements could be used to 197 identify subsets of traditional GLM-identified voxels that most closely match the hemodynamic response of an ROI. A B 60 C 4 50 Stim 0.8 V ✱✱✱✱ 40 FSCV 50 3 Ag/AgCl FSCV

0.0 V 30 Oxygen ( 40 fMRI 2 20 NAc 30 µ

VTA SNR -1.4 V 1 M) % BOLD 10 20 0 0 10 2 s 30 s 210 s -10 X 3 0 -1 -20 0 20 40 60 80 100 s FSCV fMRI D E F 0 4 1.0 FSCV fMRI 10 0.7 leads leads -1 2 10 Pearson’sr 0.5 2.4 mm 1.4 mm 10-2

0 P-value % BOLD 10-3 -2 r: 0.83 0.0 p<0.0001 Correlation Coefficient 0 10-4 -10 0 10 20 30 40 -10 -5 0 5 10 -10 -5 0 5 10 0.4 mm -0.4 mm Oxygen (µM) Time Shift (s) Time Shift (s)

1.3 V 5 70 8 150 G H I J ✱✱✱✱ 60 4 ✱✱✱✱ ✱✱ -0.4 V -0.4 V 50 6 60 Dopamine (nM) 3 40 100 50 2 -5 0 5 10 15 30 4 40

% BOLD 20 30 1 50

SNR 10 2 20 NAc Max. % BOLD 0 0 10 NAc Max. Dopamine (nM) -10 0 -1 0 0 FSCV -20 0 20 40 60 80 100 s 300 500 700 µA 300 500 700 µA

K B L A-B: 7 s 25 A-C: 22 s A-D: 34 s

D 2.4 mm 1.4 mm C F A

0 0 20 40 60 80 100 s 0.4 mm -0.4 mm 198 199 Fig.4. Simultaneous FSCV-fMRI recordings during electrical VTA deep-brain stimulation, using both tissue oxygen 200 FSCV and dopamine FSCV waveforms (2 s, 700 µA, 60 Hz, 2 ms pulse width). (A) Paradigm schematic for electrical 201 stimulations. Top: Cartoon of implant locations (left), and the oxygen-sensitive FSCV waveform used (right). Red 202 indicates the oxidation potential sweep, green indicates the reduction potential sweep. Bottom: Electrical 203 stimulation paradigm. (B) SNR values from FSCV and fMRI ROIs during individual stimulation time-courses (n=18 204 events, ROIs=36±3 voxels; ratio-paired t-test ****p<0.001). (C) Time-courses from NAc recording sites during 205 ipsilateral electrical VTA stimulation (n=18 events). (D) Pearson’s correlation of averaged BOLD fMRI and FSCV 206 oxygen time-courses from c. (E) BOLD fMRI and oxygen FSCV time shift correlations (n=18 events). Individual event 207 time shifts are grey, averaged responses are blue. (F) Group-level correlation maps between FSCV oxygen and BOLD 208 fMRI time-courses (p=0.05). Voxels with correlation coefficients <0.3 are excluded. (G) Dopamine-sensitive FSCV 209 waveform used for the second set of electrical stimulations (top), and the SNR for FSCV dopamine (bottom; n=18 210 events). (H) Time-courses from NAc recording sites during the second set of ipsilateral electrical VTA stimulations. 211 Stimulation duration is indicated with a red bar. Inset is a magnification; blue scale bar=2% BOLD response, purple 212 scale bar=20 nM dopamine. (I) Maximum evoked NAc BOLD fMRI values and (J) FSCV dopamine release from 213 different amplitude stimuli (one-way RM-ANOVA, Tukey’s post-hoc **p<0.01, ****p<0.001). (K) Group-level 214 dopamine-derived HRF. Scale bar=2% of BOLD response. Averaged HRF is grey, with smoothed HRF overlaid in black. 215 (L) Linear mixed-effects model maps derived from BOLD fMRI using the dopamine-derived HRF. Group-level main 216 effect of maximum evoked dopamine amplitude (one-way ANOVA, p<0.05 threshold). Error bars are ±SEM.

217 Next, we recorded and analyzed simultaneous dopamine and BOLD data using FSCV-fMRI. The dopamine- 218 sensitive FSCV waveform recorded high SNR dopamine measurements during fMRI (29.2±7.1; Fig.4G). Evoked 219 dopamine release preceded the hemodynamic response, in agreement with the literature24 (Fig.4H), and fMRI 220 responses were consistent (i.e., fMRI responses during dopamine FSCV resembled those obtained during oxygen 221 FSCV in Fig.4C). One-way repeated-measures (RM) analysis of variance (ANOVA) revealed that the maximum evoked 222 BOLD fMRI in the NAc did not significantly change with stimulation amplitude (Fig.4I), but evoked dopamine release 223 did (F(1.3,22.3)=39.2, p<0.0001; Tukey’s post-hoc, 300 vs 500 and 700 µA, p<0.0001; 500 vs 700 µA, p=0.0014; Fig.4J). 224 Deconvolution analysis was performed for each simultaneously acquired pair of BOLD and dopamine time-courses 225 to derive subject-specific HRFs (Supplementary Fig.6). The average of all subject-specific HRFs has an initial rise 226 peaking at 7 s, undershoot peaking at 22 s, and a second increase above baseline peaking at 34 s (Fig.4K). Because 227 graded stimulation increased maximum dopamine release but not BOLD amplitudes in the NAc, we used this unique 228 FSCV-fMRI dataset to identify spatial locations where the BOLD response covaried with the released dopamine 229 amplitude. Each dopamine time-course (n=18 events) was convolved with our derived HRF to form a GLM, then used 230 to estimate subject-level contrast maps. The main effect of peak dopamine amplitude on the BOLD response 231 identified significant BOLD activations in the prefrontal cortex and septum, where dopamine neurons also project38, 232 as having encoded dopamine amplitude dependence (p<0.05; Fig.4L). Building upon the traditional neuroimaging 233 analysis pipeline, neurotransmitter data from FSCV can provide an additional ground-truth regressor for fMRI data 234 that is crucial for understanding how local neurotransmission can impact brain network dynamics.

235 Discussion

236 FSCV is a minimally invasive electrochemical technique used to detect neurotransmitters such as dopamine 237 at a carbon-fiber microelectrode, and fMRI is a noninvasive way of mapping hemodynamic changes across the entire 238 brain. A platform capable of monitoring dopamine and fMRI simultaneously has been desirable because how 239 dopamine affects brain-wide activity dynamics is poorly understood22 and advancing this knowledge could make a 240 tremendous impact on understanding dopaminergic drug mechanisms22, human behaviors19, and how to improve 241 deep-brain stimulation therapies21. Existing FSCV and fMRI studies have used these techniques separately7,19,20,39, 242 but one study indicated that simultaneous recordings could be possible40. In this study, we develop and characterize 243 simultaneous FSCV-fMRI for the first time, in vitro and in vivo, within a 9.4 T MRI. Highly resolved oxygen changes 244 via FSCV and concurrent BOLD fMRI detected complementary information about hemodynamics across multiple 245 spatial and temporal scales in response to electrical stimulation and oxygen inhalation challenges (i.e., with and 246 without induced neuronal activity increases). Electrically stimulated dopamine release and simultaneous BOLD fMRI 247 data were collected, and we propose a way to use dopamine time-course information to identify specific brain 248 regions that encode dopamine amplitudes. This technique bridges a critical gap between neurotransmitter activity 249 and neuroimaging data and can help isolate the effect of chemical signaling on brain-wide hemodynamics.

250 We took advantage of the oxygen-sensing capabilities of FSCV and fMRI to cross-validate evoked 251 hemodynamic changes at multiple scales. The hemodynamic response lags increased neuronal activity by up to 252 several seconds2, so it can be observed using conventional FSCV and fMRI sampling rates (10 and 1 Hz, respectively). 253 The spatial resolutions of FSCV and fMRI, however, differ by orders of magnitude (105 µm3 and 108 µm3 for a 100 µm 254 FSCV microelectrode sensing volume and single fMRI voxel, respectively)25. This work required voxel clusters of 36±3 255 per ROI to obtain detectable fMRI responses, which further increased the spatial mismatch. We observed strong 256 correlations between FSCV and fMRI time-courses in the NAc following oxygen challenges (Fig.3E), in agreement 257 with the only known study to perform simultaneous amperometric tissue oxygen detection and BOLD fMRI35; 258 however, FSCV oxygen cleared faster (Fig.3F). Extended hyperoxia decreases venous deoxyhemoglobin percentages, 259 slows cerebral blood flow41, and pushes excess dissolved oxygen into surrounding tissues35. BOLD fMRI is weighted 260 towards deoxyhemoglobin in venous blood vessels2, and FSCV detects tissue oxygen (i.e., oxygen diffused from blood 261 vessels), so these techniques are sensitive to different oxygen dynamics. Thus, slower blood flow extends 262 oxygenated hemoglobin clearance time from vessels within the fMRI ROI relative to tissue oxygen clearance from 263 the much smaller FSCV sampling volume. Despite FSCV’s faster temporal resolution, the onset kinetics of electrically 264 evoked FSCV oxygen responses slightly lagged those of fMRI (Fig.4C,H), in agreement with literature that performed 265 these measurements separately39. These data indicate that with respect to tissue oxygen, hemoglobin oxygenation 266 changes more rapidly in response to evoked neuronal activity than oxygen inhalation. It is possible that processes 267 downstream of stimulated neuronal activity may more stringently regulate oxygen delivery within highly localized 268 environments. Combining FSCV with more advanced, vessel-resolution fMRI techniques42, or using FSCV arrays with 269 fMRI, could further connect the knowledge gap between hemodynamic responses observed at these two different 270 spatial scales.

271 Our simultaneous FSCV-fMRI experiments were performed in the NAc because the striatum has the highest 272 concentration of dopamine terminals, and dopamine is a known vasomodulatory neurotransmitter4,6. Though the 273 oxygen-sensitive waveform can detect dopamine24,26 (Supplemental Fig.1), we quantified dopamine release using 274 data from the dopamine-sensitive waveform12 (Fig.4G). The dopamine oxidation potential was shifted anodically 275 during the scan versus during implantation (Supplementary Fig.7), an indication of reference and/or recording 276 electrode biofouling seen with chronically implanted glass carbon-fiber microelectrodes43. The shift was not related 277 to the magnetic field or active gradients, as it did not occur in vitro (Fig.2B) and did occur in Faraday cage recordings 278 from a subject given the same surgery and recovery period (Supplementary Fig.7). Though post-implantation 279 electrodes showed decreased sensitivity in our hands, others saw consistent longitudinal responses30. Strategies to 280 mitigate chronic implantation sensitivity problems are being researched44,45, but our data indicate that reliable, high 281 SNR analyte detection during fMRI can still be achieved days after electrode implantation. 282 283 The ability to observe dopamine dynamics and whole-brain imaging simultaneously would enormously 284 benefit research into global brain circuitry changes from drug abuse, pain, or disorders implicated in dopamine 285 dysfunction7,10,19,22. By measuring and quantifying evoked dopamine release, FSCV-fMRI can generate data-driven 286 HRFs (Fig.4K) and help identify brain regions sensitive to different dopamine release amplitudes (Fig.4L). These data 287 can be used to study effects of neuronal or pharmacological manipulations and tease apart the role of dopaminergic 288 signaling on local and global hemodynamic responses in healthy and pathological brains10,22. Our study chose an 289 electrical stimulation protocol for FSCV to induce reliable dopamine changes in the NAc24; however, neurochemicals 290 other than dopamine are released in the NAc following electrical or optogenetic VTA dopaminergic neuron 291 stimulation46, so these results do not establish dopamine as the sole neurochemical source responsible for the 292 observed oxygen changes. Future studies that selectively drive dopamine release with confounding factors 293 eliminated (e.g., suppressing neuronal activity and/or unwanted neurotransmission) will further isolate the role of 294 dopamine and its receptors7,36. 295 296 Though our electrical stimulations lack specificity, the analysis framework proposed herein should set the 297 foundation for neurotransmitter-based HRF modeling in the future. Interestingly, our dopamine-modeled HRF has 298 two notable peaks, different from the single-peak canonical HRF47. The first peak is the initial rise, and an additional 299 peak occurs after the first post-stimulus undershoot. This “second peak” is well documented in the oxygen FSCV 300 literature using both VTA and median forebrain bundle stimulations24,26,27, supporting our approach towards deriving 301 an HRF to describe the relationship between concurrently measured neurochemical release and BOLD fMRI. 302 Literature attributes these oxygen increases to cerebral blood flow27, and that each peak is mediated differently; for 303 example, an adenosine receptor antagonist decimated the second peak response but had no effect on the first24. 304 Interestingly, blocking dopamine synthesis had no effect on the either oxygen peak27. More research is needed to 305 isolate dopamine and dopamine receptors’ influence over HRFs. 306 307 Our method using a ground-truth output metric (i.e., quantified dopamine release amplitudes) lends 308 additional neurochemical context to response map interpretations and can pull out relationships that traditional 309 input models (e.g., stimulation paradigms) would miss. For instance, though a traditional stimulus paradigm model 310 with GLM pulled wide-spread activation around ipsilateral NAc (Supplementary Fig.5) similar to literature7,36,37, maps 311 of dopamine amplitude as the main effect isolated prefrontal cortex and septum (Fig.4L). Our method utilizing 312 ground truth data to guide models is more effective than relying on assumptions that the evoked response would 313 scale with the given stimulus inputs, and can account for inter-subject response variability. The ability to identify 314 locations where brain activity co-varies with quantitative neurotransmitter amplitudes could have far-reaching 315 benefits for research studying local neurotransmission influence over brain hemodynamics and behavior. 316 317 Though FSCV has been a common method of detecting in vivo neurotransmitters for decades, an increasing 318 number of optical detection alternatives exist that could also pair with fMRI17,48–50. A dopamine-sensitive protein- 319 based MRI contrast agent has been used with BOLD13, and dopamine dynamics have been explored using near- 320 infrared fluorescent sensors15 and genetically encoded fluorescent sensors14,16. These techniques require either 321 infusion cannulae or optical fibers for contrast agent dispersal or fluorescence detection, respectively, which use 322 fMRI-compatible materials. These alternative dopamine detection methods have high spatiotemporal resolution, 323 even exceeding FSCV, making them exciting potential multimodality expansions. These methods show qualitative 324 percent changes in signal, so quantitative FSCV measurements could provide useful cross-validation. FSCV-fMRI 325 remains a strong multimodality for studying dopaminergic influence on neurovascular coupling, as it requires a single 326 implantation surgery, no genetic modifications or injections, and is the only current method available to measure 327 electroactive neurotransmitter activity and whole-brain hemodynamics simultaneously. Importantly, FSCV is the 328 fastest-resolution technique to quantify neurotransmitter release in humans12,18–20, which is crucial for detecting 329 rapid neurotransmission events during behavior. Expanding this multimodality to include other neurotransmitter 330 sensors or other techniques is a realistic possibility that could hone in on the role of dopamine in the hemodynamic 331 response at multiple spatiotemporal scales in both preclinical22 and clinical21 experiments.

332 Materials and Methods 333

334 Fast Scan Cyclic Voltammetry (FSCV). Carbon-fiber microelectrodes were fabricated as previously described30, with 335 modifications for MRI compatibility. Briefly, a 5-7 µm-diameter carbon fiber was manually threaded through a saline- 336 filled, polyimide-fused silica capillary. The capillary was sealed with clear epoxy at the recording end such that 100- 337 200 µm of the fiber was left exposed, and the other end of the fiber was sealed to a silver wire with silver epoxy and 338 silver paint. The silver wire/carbon fiber junction was encapsulated with clear epoxy for stability.

339 FSCV acquisition hardware was custom-built within MR-compatible aluminum casing (UNC Electronics Facility, 340 Chapel Hill, NC, USA), and data were collected and analyzed with the High Definition Cyclic Voltammetry (HDCV) 341 computer program51. The standard coaxial cables used to carry signal from the electrodes to the headstage were 342 replaced with MR-compatible triaxial cables. Triaxial cables were chosen to prevent current leakage via a driven 343 shield, as well as to compensate for the additional cable capacitance introduced by the length of the cables necessary 344 to connect the headstage to the electrodes through the back of the MRI bore (approximately 2 m). Output 345 digitization was smoothed by passing digital waveform signals through a 2 kHz low-pass analog filter before being 346 applied to the electrode. Input signals were split via T adapter and collected as raw current at one input channel (for 347 analysis) and passed through an analog 2 kHz low-pass filter prior to collection on a second input channel (for real- 348 time signal viewing). Ferrites were placed on both input and output voltage coaxial and triaxial cables to further 349 eliminate high-frequency noise (Fair-Rite Products Corp., Wallkill, NY, USA). A 4th order Bessel low-pass filter (2 kHz 350 cutoff frequency) was digitally applied to input currents post-acquisition.

351 Waveforms were applied to the microelectrodes to selectively oxidize and/or reduce dopamine and/or oxygen 352 (Supplementary Fig.2). The oxygen-sensing waveform is an 11 ms waveform that holds at 0 V, scans up to +0.8 V, 353 down to –1.4 V, then back to 0 V24–26. The dopamine-sensitive waveform is an 8.5 ms, triangular waveform with 354 a -0.4 V holding potential that scans to +1.3 V and back to -0.4 V24,52. Waveforms were applied at 5 or 10 Hz, voltages 355 were scanned at 400 V/s, and data were sampled at 100 kHz. All voltages are versus Ag/AgCl. In both calibration and 356 in vivo studies, the oxygen-sensitive waveform was used prior to the dopamine-sensitive waveform.

357 A “dummy cell” was fabricated to characterize noise within the MRI bore, which consisted of a 30 kΩ resistor and 358 0.33 nF capacitor connected in series on a circuit board, to mimic solution resistance and carbon fiber electrode 359 double-layer capacitance, respectively. Dopamine stock solutions and oxygen calibration solutions were freshly 360 prepared in room temperature phosphate-buffered saline (pH 7.4). Flow-injection analysis was used to collect the 361 current amplitude responses to an analyte bolus of known concentration53. Concentrations from 100-2000 nM were 362 injected in random order, and each concentration was injected in triplicate. The within-bore MRI flow-injection 363 datasets were acquired with the gradient disabled with low-pass filtering, enabled without low-pass filtering, and 364 enabled with low-pass filtering. These datasets were acquired in random order to ensure that noise from the enabled 365 gradient had no detrimental effect on electrode sensitivity. Flow injection analysis data from within the MRI bore 366 were not used to calibrate electrodes, as the tubing length required to keep the magnetic 6-port injector a safe 367 distance from the MR heavily diluted the bolus concentrations. Flow-through analysis was repeated inside a Faraday 368 cage to obtain calibration factors for pre- and post-implantation electrodes. Only oxygen waveform calibration data 369 where the oxygen waveform was applied first are included (i.e., those that used the dopamine waveform first were 370 excluded).

371 Animal Care. All animal protocols were approved by the Institutional Animal Care and Use Committee of the 372 University of North Carolina at Chapel Hill (UNC) in accordance with the Guide for Care and Use of Laboratory 373 Animals (eighth edition). Nine male Long-Evans rats (300–450 g, Charles River, Wilmington, MA, USA) were pair- 374 housed prior to surgery at UNC animal facilities, given food and water ad libitum, and kept on a 12 h light/dark cycle. 375 Care was taken to reduce the number of animals used. Three animals were excluded: Two animals had prohibitively 376 large imaging artifacts, likely from the connective leads plugged into the stimulating electrode pedestal in the slices 377 of interest, and one animal’s electrode stopped functioning before the experiment finished.

378 Surgery. Rats were deeply anaesthetized with 4% isoflurane, placed in a rodent stereotaxic frame (Kopf, Tujunga, 379 CA, USA), and maintained at 2% isoflurane. Depth of anesthesia was confirmed via toe pinch and eye blink responses. 380 Coordinates are relative to skull at bregma54. Holes were drilled over the ventral striatum (1.3 mm A-P, 2.0 mm M- 381 L, 6.8-7.8 mm D-V) for the carbon-fiber electrode and the VTA (-5.2 mm A-P, +1.2 mm M-L, −8.2 to -9.2 mm D-V) for 382 the bipolar, MR-compatible twisted tungsten stimulating electrode (PlasticsOne, VA, USA); to avoid the sagittal sinus, 383 the stimulating electrode was implanted at a 4 degree angle. The stimulating electrode was untwisted enough to 384 separate the tips by 1 mm55. The D-V coordinates of the stimulating and working electrodes were optimized to evoke 385 maximal dopamine release. The dopamine-sensitive waveform was used during implantation, so oxygen responses 386 to stimulation were unknown prior to fMRI scanning. An additional 3-4 burr holes were drilled for MR-compatible, 387 brass skull screws. An Ag/AgCl reference electrode wire was placed in the contralateral hemisphere. Silicone 388 adhesive was applied around the working electrode to prevent cortical tissue damage from the heat of curing dental 389 cement (World Precision Instruments, Sarasota, FL, USA). Skull screws, electrodes, and plastic shield were held in 390 place using dental cement. Subjects were singly-housed for at least 4 days to recover before scanning.

391 Phantom MRI. All MRI data were acquired using a Bruker 9.4 T MRI. In vitro scans used a Bruker 72 mm volume coil. 392 An FSCV electrode was embedded within an agarose phantom and scanned to examine whether the materials 393 induced artifacts in RARE (BW=312.5 kHz, TR=2000 ms, TE=40 ms, in-plane resolution= 100 µm, slice thickness=100 394 µm) or FLASH (BW=312.5 kHz, TR=50 ms, TE=3.52 ms, in-plane resolution= 100 µm, slice thickness=100 µm). 395 FSCV-fMRI in vivo. In vivo BOLD fMRI data were acquired using a homemade surface coil, and subjects were 396 prepared as described previously36,56. Briefly, subjects were anesthetized and paralyzed by infusing a cocktail of 397 dexmedetomidine and pancuronium bromide (0.1 mg/mL and 1.0 mg/mL respectively, intraperitoneal) to prevent 398 motion artifacts, supplemented with low dose isoflurane (0.5%), and artificially ventilated via endotracheal 399 intubation. Experiments did not begin until rectal temperature stabilized at 37±0.5 °C, end-tidal CO2 was 2.6-3.2%, 400 and oxygen saturation was ≥95%. Physiology was monitored throughout the experiment and maintained within 401 normal ranges. Five-slice echo planar imaging (EPI) data were acquired with BW=333.33 kHz, TR=1000 ms, TE=15 402 ms, 80x80 matrix, FOV=2.56 cm2, slice thickness=1 mm. One slice was discarded prior to data analysis due to artifacts 403 from the stimulating electrode. EPI slice acquisition and electrical stimulations were both controlled via HDCV 404 software. To interleave fMRI and FSCV data acquisition, HDCV sent a 10 ms, high-to-low TTL pulse to trigger EPI per- 405 slice acquisition 3 ms after each applied FSCV waveform.

406 Oxygen challenges were administered and repeated in triplicate, where 100% medical air was inhaled for 60 s, then 407 100% oxygen for 60 s, and finally 100% medical air for 180 s. For deep brain stimulations, biphasic square-wave 408 stimulations were administered 180 s apart, at 60 Hz with current amplitudes of 300, 500, and 700 µA, a 2 ms pulse- 409 width, and total duration of 2 s. Each stimulation amplitude was administered in triplicate and in random order with 410 simultaneously acquired oxygen FSCV, followed by a repeated set of randomly ordered triplicate stimulations per 411 amplitude with simultaneously acquired dopamine FSCV. Finally, a powerful stimulation was given to evoke large 412 responses for principal component analysis (PCA) training sets needed for FSCV analysis (60 Hz, 900 µA, 4 ms pulse- 413 width, 4 s duration). A high-resolution RARE scan was obtained at the end of each scan to verify each electrode 414 location (BW=183.1 kHz, TR=2500 ms, TE=33 ms, in-plane resolution=100 µm, slice thickness=1 mm).

415 Data Analysis. FSCV data were analyzed within the HDCV analysis software, which can be licensed to academic users 416 through the University of North Carolina Electronics Facility (University of North Carolina at Chapel Hill, Chapel Hill, 417 NC, USA). FSCV data were background subtracted and filtered with a 4th order, 2 kHz cutoff low-pass Bessel filter. 418 FSCV background currents were taken prior to each gas challenge or stimulation event. Post-implantation calibration 419 factors from flow-through analysis, normalized to the length of the exposed carbon fiber, were used to quantify 420 oxygen and dopamine concentrations. To verify that all faradaic currents resulted from oxygen or dopamine, data 421 sets were analyzed using principal component analysis (PCA) within the HDCV analysis software51. Training sets for 422 PCA consisted of 4-8 different concentrations per analyte of interest, per electrode. FSCV SNRs were calculated by 423 subtracting a 10 s baseline average prior to stimulation from the maximum peak oxidative (dopamine) or reductive 424 (oxygen) current, then dividing the difference by the standard deviation of the same 10 s baseline period. FSCV data 425 were binned to 1 s increments for all fMRI time-course comparisons.

426 BOLD fMRI data were pre-processed and analyzed in Analysis of Functional NeuroImages (AFNI) software (National 427 Institute of Mental Health, Bethesda, MD, USA). Prior to preprocessing, all fMRI data were zero-padded to ensure 428 that the data had a consistent and sufficient matrix size to be used as inputs for the AFNI image registration software. 429 T2 anatomical and EPI fMRI data were manually aligned to a template using ITK-Snap. The first 8 TRs were excluded 430 to account for steady state, and the first 8 s of FSCV time-courses were discarded to match. Data were interpolated 431 in order to follow AFNI pre-processing pipeline slice requirements. After pre-processing, data matrices were 432 truncated to their original size. BOLD time-courses were extracted using an ROI mask unique to each subject, based 433 on the location of the tip of the FSCV microelectrode (Supplementary Fig.4).

434 Head motion and baseline drift were removed by linear regression prior to first-level statistical analyses and ROI 435 time-course extractions. Subject-level correlation maps were calculated using BOLD fMRI and oxygen FSCV time- 436 courses from oxygen challenge and electrical stimulation experiments. Fisher z-transformations were performed on 437 subject-level oxygen challenge and electrical stimulation correlation maps for group statistics. A one-sample t-test 438 was performed using linear regression to identify voxel clusters that fit with the intercept line representing the 439 average, at p<0.05. Benjamini/Hochberg false discovery rate (FDR) correction was used to correct the voxel-level 440 multiple comparison problem, with a q<0.05 threshold. Then, a group average correlation map was generated using 441 the surviving clusters. Subsequently, the inverse Fisher's Z-transform was performed to restore Fisher Z to Pearson's 442 R correlation coefficients.

443 Traditional functional activation maps were obtained to compare with our hemodynamic response function (HRF) 444 derived maps. The BOLD response was first predicted by convolving the electrical stimulus (i.e., 2 sec block) and the 445 following gamma function (i.e., the default HRF in AFNI): HRF(t) = int(g(t-s), s=0..min(t,d)), where g(t) = t^q*exp(-t) 446 /(q^q*exp(-q)), where t=time and q=5. This stimulus-derived HRF predictor was used as a regressor for subject-level 447 GLMs to identify significantly activated voxels from individual response maps. Group-level analyses used a one-tailed 448 t-test to test mean subject-level regression coefficient maps against zero, with a q=0.05 FDR threshold.

449 Oxygen ROI time-courses were low-pass filtered with a 0.05 Hz cutoff to remove physiological noise prior to deriving 450 kinetic parameters (oxygen challenge) and subject-level HRFs (electrical stimulation). The HRF was derived using the 451 linear system: BOLD(t) = DA(t) * HRF + baseline + linear drift(t), where DA is dopamine. DC offset and linear drift 452 were removed from BOLD time-courses during HRF deconvolution analysis (Supplementary Fig.6). The averaged HRF 453 was smoothed with a 3rd order Savitzky Golay filter. To identify brain regions that encode dopamine or stimulation 454 amplitude, the smoothed HRF was convolved with dopamine time-courses to form Bayesian Ridge GLM regressors, 455 which were subsequently used to estimate subject-level BOLD fMRI contrast maps. Second-order polynomial curves 456 were added to detrend the BOLD signal prior to statistical group-level analysis.

457 Code Availability. Standard AFNI codes were used for BOLD fMRI pre-processing and traditional functional activation 458 map analysis. MATLAB (The Mathworks, Inc., Natick, MA, USA) codes for the time shift analysis and HRF convolution 459 will be made available upon request, as will Python codes used to obtain statistical response maps and correlation 460 maps.

461 Statistical Information. Statistics were performed using GraphPad Prism 8 (GraphPad Software, San Diego, CA, USA), 462 AFNI, and Python. RM-ANOVA was used to compare in vitro SNRs between multiple electrodes, with gradient and 463 filter status as the within-subject factor. Maximum evoked BOLD and dopamine RM-ANOVA analyses were 464 performed using stimulation amplitude as the within-subject factor. Significant ANOVA interactions were subjected 465 to Tukey’s post-hoc multiple comparisons analysis. FSCV and fMRI values for in vivo SNR data and oxygen challenge 466 kinetic parameters were compared using two-tailed, ratio-paired t-tests.

467 One-way ANOVAs were used to test the main effects of the maximum evoked dopamine amplitude on HRF-derived 468 contrast maps, using a permutation method (5000 times iteration) with the maximum T method for multi-voxel, 469 multiple-comparison correction. A p<0.05 threshold was used to indicate statistical significance in all analyses, and 470 all statistical clusters were limited to clusters >50 voxels to exclude potential false positives.

471

472 Acknowledgements

473 We acknowledge the help of Dr. Weiting Zhang for her analysis code, Dr. Domenic Cerri for helpful discussions, and 474 members of the UNC Center for Animal MRI (CAMRI) and Biomedical Research Imaging Center (BRIC) for technical 475 assistance. Research reported in this publication was supported by the National Institutes of Health under Award 476 Number F32MH115439, RF1MH117053, R01NS091236, R01MH111429, R41MH113252, P60AA011605, 477 U01AA020023, R01AA025582, and P50HD103573. The content is solely the responsibility of the authors and does 478 not necessarily represent the official views of the National Institutes of Health.

479 Author Contributions

480 L.W. designed and ran all experiments, preprocessed and analyzed data, and prepared the manuscript. M.V. helped 481 characterize the multimodality and troubleshoot electronic problems, and fabricated custom hardware. S.L. 482 performed the majority of the fMRI statistical analyses, and, along with T.C., assisted in the design, execution, and 483 interpretation of the HRF analysis. L.W., R.M.W., and Y.Y.S. conceived the research idea. R.M.W. and Y.Y.S. provided 484 advice on experimental designs throughout the project. All authors have read, commented on, and approved this 485 manuscript.

486 Competing Interests

487 The authors declare no competing interests.

488 References

489 1. Iadecola, C. The Neurovascular Unit Coming of Age: A Journey through Neurovascular Coupling in Health 490 and Disease. Neuron 96, 17–42 (2017). 491 2. Buxton, R. B. The physics of functional magnetic resonance imaging (fMRI). Reports Prog. Phys. 76, (2013). 492 3. Logothetis, N. K., Pauls, J., Augath, M., Trinath, T. & Oeltermann, A. Neurophysiological investigation of the 493 basis of the fMRI signal. Nature 412, 150–157 (2001). 494 4. Krimer, L. S., Christopher Muly, E., Williams, G. V. & Goldman-Rakic, P. S. Dopaminergic regulation of 495 cerebral cortical microcirculation. Nat. Neurosci. 1, 286–289 (1998). 496 5. Meyer-Lindenberg, A. et al. Reduced prefrontal activity predicts exaggerated striatal dopaminergic 497 function in schizophrenia. Nat. Neurosci. 5, 267–271 (2002). 498 6. Choi, J.-K., Chen, Y. I., Hamel, E. & Jenkins, B. G. Brain hemodynamic changes mediated by dopamine 499 receptors: Role of the cerebral microvasculature in dopamine-mediated neurovascular coupling. 500 Neuroimage 30, 700–712 (2006). 501 7. Brocka, M. et al. Contributions of dopaminergic and non-dopaminergic neurons to VTA-stimulation 502 induced neurovascular responses in brain reward circuits. Neuroimage 177, 88–97 (2018). 503 8. Miyazaki, I., Asanuma, M., Diaz-Corrales, F. J., Miyoshi, K. & Ogawa, N. Direct evidence for expression of 504 dopamine receptors in astrocytes from basal ganglia. Brain Res. 1029, 120–123 (2004). 505 9. Ren, J., Xu, H., Choi, J.-K., Jenkins, B. G. & Chen, Y. I. Dopaminergic response to graded dopamine 506 concentration elicited by four amphetamine doses. Synapse 63, 764–772 (2009). 507 10. Mandeville, J. B. et al. A receptor-based model for dopamine-induced fMRI signal. Neuroimage 75, 46–57 508 (2013). 509 11. Paasonen, J. et al. Dose-response effect of acute phencyclidine on functional connectivity and dopamine 510 levels, and their association with schizophrenia-like symptom classes in rat. Neuropsychopharmacology 511 119, 15–25 (2017). 512 12. Bucher, E. S. & Wightman, R. M. Electrochemical Analysis of Neurotransmitters. Annu. Rev. Anal. Chem. 8, 513 239–61 (2015). 514 13. Li, N. & Jasanoff, A. Local and global consequences of reward-evoked striatal dopamine release. Nature 515 580, 239–244 (2020). 516 14. Patriarchi, T. et al. Ultrafast neuronal imaging of dopamine dynamics with designed genetically encoded 517 sensors. Science (80-. ). 360, (2018). 518 15. Beyene, A. G. et al. Imaging striatal dopamine release using a nongenetically encoded near infrared 519 fluorescent catecholamine nanosensor. Sci. Adv. 5, (2019). 520 16. Sun, F. et al. A Genetically Encoded Fluorescent Sensor Enables Rapid and Specific Detection of Dopamine 521 in Flies, Fish, and Mice. Cell 174, 481–496 (2018). 522 17. Ganesana, M., Lee, S. T., Wang, Y. & Venton, B. J. Analytical Techniques in Neuroscience: Recent Advances 523 in Imaging, Separation, and Electrochemical Methods. Anal. Chem. 89, 314–341 (2017). 524 18. Kishida, K. T. et al. Sub-second dopamine detection in human striatum. PLoS One 6, (2011). 525 19. Lohrenz, T., Kishida, K. T. & Montague, P. R. BOLD and its connection to dopamine release in human 526 striatum: a cross-cohort comparison. Philos. Trans. R. Soc. B Biol. Sci. 371, 20150352 (2016). 527 20. Chang, S. Y. et al. Wireless fast-scan cyclic voltammetry to monitor adenosine in patients with essential 528 tremor during deep brain stimulation. Mayo Clin. Proc. 87, 760–765 (2012). 529 21. Price, J. B. et al. Clinical applications of neurochemical and electrophysiological measurements for closed- 530 loop neurostimulation. Neurosurg. Focus 49, E6 (2020). 531 22. Bruinsma, T. J. et al. The relationship between dopamine neurotransmitter dynamics and the blood- 532 oxygen-level-dependent (BOLD) signal: A review of pharmacological functional magnetic resonance 533 imaging. Front. Neurosci. 12, 238 (2018). 534 23. Oh, Y. et al. Tracking tonic dopamine levels in vivo using multiple cyclic square wave voltammetry. Biosens. 535 Bioelectron. 121, 174–182 (2018). 536 24. Venton, B. J., Michael, D. J. & Wightman, R. M. Correlation of local changes in extracellular oxygen and pH 537 that accompany dopaminergic terminal activity in the rat caudate-putamen. J. Neurochem. 84, 373–381 538 (2003). 539 25. Walton, L. R., Boustead, N. G., Carroll, S. & Wightman, R. M. Effects of Glutamate Receptor Activation on 540 Local Oxygen Changes. ACS Chem. Neurosci. 8, 1598–1608 (2017). 541 26. Zimmerman, J. B. & Wightman, R. M. Simultaneous Electrochemical Measurements of Oxygen and 542 Dopamine in Vivo. Anal. Chem. 63, 24–28 (1991). 543 27. Zimmerman, J. B., Kennedy, R. T. & Wightman, R. M. Evoked Neuronal Activity Accompanied by 544 Transmitter Release Increases Oxygen Concentration in Rat Striatum In Vivo But Not In Vitro. J. Cereb. 545 Blood Flow Metab. 12, 629 (1992). 546 28. Shyu, B.-C., Lin, C.-Y., Sun, J.-J., Sylantyev, S. & Chang, C. A method for direct thalamic stimulation in fMRI 547 studies using a glass-coated carbon fiber electrode. J. Neurosci. Methods 137, 123–131 (2004). 548 29. Yang, P.-F. et al. Comparison of fMRI BOLD Response Patterns by Electrical Stimulation of the 549 Ventroposterior Complex and Medial Thalamus of the Rat. PLoS One 8, (2013). 550 30. Clark, J. J. et al. Chronic microsensors for longitudinal, subsecond dopamine detection in behaving animals. 551 Nat. Methods 7, 126–129 (2010). 552 31. Schenck, J. F. The role of magnetic susceptibility in magnetic resonance imaging: MRI magnetic 553 compatibility of the first and second kinds. Med. Phys. 23, 815–850 (1996). 554 32. Dong, L. et al. Neuroscience Information Toolbox: An Open Source Toolbox for EEG–fMRI Multimodal 555 Fusion Analysis. Front. Neuroinform. 12, 56 (2018). 556 33. Jaime, S., Cavazos, J. E., Yang, Y. & Lu, H. Longitudinal observations using simultaneous fMRI, multiple 557 channel electrophysiology recording, and chemical microiontophoresis in the rat brain. J. Neurosci. 558 Methods 306, 68–76 (2018). 559 34. Huttunen, J. K., Gröhn, O. & Penttonen, M. Coupling between simultaneously recorded BOLD response and 560 neuronal activity in the rat somatosensory cortex. Neuroimage 39, 775–785 (2008). 561 35. Lowry, J. P. et al. Real-time electrochemical monitoring of brain tissue oxygen: A surrogate for functional 562 magnetic resonance imaging in rodents. Neuroimage 52, 549–555 (2010). 563 36. Decot, H. K. et al. Coordination of Brain-Wide Activity Dynamics by Dopaminergic Neurons. 564 Neuropsychopharmacology 42, 615–627 (2017). 565 37. Ferenczi, E. A. et al. Prefrontal cortical regulation of brainwide circuit dynamics and reward-related 566 behavior. Science (80-. ). 351, (2016). 567 38. Bjö, A. & Dunnett, S. B. Dopamine neuron systems in the brain: an update. Trends Neurosci. 30, 194–202 568 (2007). 569 39. Hosford, P. S. et al. Electrochemical carbon fiber-based technique for simultaneous recordings of brain 570 tissue PO2, pH, and extracellular field potentials. Biosens. Bioelectron. X 3, (2019). 571 40. Kimble, C. J. et al. Wireless Instantaneous Neurotransmitter Concentration Sensing System (WINCS) for 572 intraoperative neurochemical monitoring. in Proceedings of the 31st Annual International Conference of 573 the IEEE Engineering in Medicine and Biology Society: Engineering the Future of Biomedicine, EMBC 2009 574 vol. 2009 4856–4859 (IEEE Computer Society, 2009). 575 41. Bulte, D. P., Chiarelli, P. A., Wise, R. G. & Jezzard, P. Cerebral perfusion response to hyperoxia. J. Cereb. 576 Blood Flow Metab. 27, 69–75 (2007). 577 42. Yu, X. et al. Sensory and optogenetically driven single-vessel fMRI. Nat. Methods 13, 337–340 (2016). 578 43. Rodeberg, N. T., Sandberg, S. G., Johnson, J. A., Phillips, P. E. M. & Wightman, R. M. Hitchhiker’s Guide to 579 Voltammetry: Acute and Chronic Electrodes for in Vivo Fast-Scan Cyclic Voltammetry. ACS Chem. Neurosci. 580 8, 221–234 (2017). 581 44. Hashemi, P. et al. Chronically Implanted, Nafion-Coated Ag/AgCl Reference Electrodes for Neurochemical 582 Applications. ACS Chem. Neurosci 2, 658–666 (2011). 583 45. Seaton, B. T., Hill, D. F., Cowen, S. L. & Heien, M. L. Mitigating the Effects of Electrode Biofouling-Induced 584 Impedance for Improved Long-Term Electrochemical Measurements in Vivo. Anal. Chem. 92, 6334–6340 585 (2020). 586 46. Stuber, G. D., Hnasko, T. S., Britt, J. P., Edwards, R. H. & Bonci, A. Dopaminergic Terminals in the Nucleus 587 Accumbens But Not the Dorsal Striatum Corelease Glutamate. J. Neurosci. 30, 8229–8233 (2010). 588 47. Lu, Y. et al. Using voxel-specific hemodynamic response function in EEG-fMRI data analysis. Neuroimage 589 32, 238–247 (2006). 590 48. Albers, F., Wachsmuth, L., van Alst, T. M. & Faber, C. Multimodal Functional Neuroimaging by Simultaneous 591 BOLD fMRI and Fiber-Optic Calcium Recordings and Optogenetic Control. Mol. Imaging Biol. 20, 171–182 592 (2018). 593 49. Liang, Z., Ma, Y., Watson, G. D. R. & Zhang, N. Simultaneous GCaMP6-based fiber photometry and fMRI in 594 rats. J. Neurosci. Methods 289, 31–38 (2017). 595 50. Chen, X. et al. Mapping optogenetically-driven single-vessel fMRI with concurrent neuronal calcium 596 recordings in the rat hippocampus. Nat. Commun. 10, 5239 (2019). 597 51. Bucher, E. S. et al. Flexible Software Platform for Fast-Scan Cyclic Voltammetry Data Acquisition and 598 Analysis. Anal. Chem. 85, 10344–10353 (2013). 599 52. Robinson, D. L., Venton, B. J., Heien, M. L. A. V. & Wightman, R. M. Detecting subsecond dopamine release 600 with fast-scan cyclic voltammetry in vivo. Clin. Chem. 49, 1763–1773 (2003). 601 53. Kristensen, E. W., Wilson, R. L. & Wightman, R. M. Dispersion in Flow Injection Analysis Measured with 602 Microvoltammetric Electrodes. Anal. Chem. 58, 986–988 (1986). 603 54. Paxinos, G. & Watson, C. The rat brain in stereotaxic coordinates. (Elsevier, 2007). 604 55. Fortin, S. M., Cone, J. J., Ng-Evans, S., McCutcheon, J. E. & Roitman, M. F. Sampling phasic dopamine 605 signaling with fast-scan cyclic voltammetry in awake, behaving rats. Curr. Protoc. Neurosci. 2015, 7.25.1- 606 7.25.20 (2015). 607 56. Albaugh, D. L. et al. Functional Magnetic Resonance Imaging of Electrical and Optogenetic Deep Brain 608 Stimulation at the Rat Nucleus Accumbens. Sci. Rep. 6, 1–13 (2016). 609