Abstract
Sulf1 is an extracellular sulfatase that regulates cell signaling by removing 6-O-sulfates from heparan sulfate. Although the roles of Sulf1 in neural development have been studied extensively, its functions in the adult brain remain largely unknown. Here, we report the effects of Sulf1 disruption on the neuronal properties of the medium spiny neurons (MSNs) in the nucleus accumbens (NAc) shell, one of the regions highly expressing Sulf1. We separately labeled MSNs expressing dopamine D1 receptors (D1-MSNs) or D2 receptors (D2-MSNs) by injecting adult male Drd1-Cre and Drd2-Cre mice with a Cre-dependent AAV vector expressing a red fluorescent protein, mCherry, and examined their electrophysiological properties by means of whole-cell patch–clamp recording. In the D2-MSNs, Sulf1 disruption led to drastic changes in neural firing responses to depolarizing current injections: in the Sulf1 knock-out mice, the rheobase was smaller than in the wild-type mice, but the number of action potentials elicited by depolarization did not increase at larger current injections. In the D1-MSNs, Sulf1 disruption resulted in more depolarized resting membrane potentials and increase in the AMPA/NMDA ratio. These results suggest that Sulf1 is essential for regulation of neuronal excitability and glutamatergic transmission of NAc MSNs in adult mice and implicate the potential roles of Sulf1 in NAc circuit activity, reward-aversion behaviors, and psychiatric disorders such as schizophrenia and drug addiction.
Significance Statement
Heparan sulfate (HS) plays critical roles in neural differentiation, axon guidance, synaptogenesis, and neurotransmission. Sulf1 is an extracellular sulfatase that removes 6-O-sulfate from HS, thereby regulating various cellular functions. Although its roles during development have been studied extensively, its functions in the adult brain remain largely unknown. Here, we examined the electrophysiological properties of medium spiny neurons (MSNs) in the nucleus accumbens shell of adult mice by means of whole-cell patch–clamp recording. We found that Sulf1 disruption led to changes in neuronal excitability and glutamatergic transmission in MSNs. This study demonstrates the roles of the Sulf1 gene in neuronal activities at the cellular level, providing an important clue toward understanding the functions of Sulf1 in the adult brain.
Introduction
Heparan sulfate proteoglycans (HSPGs) are glycoproteins ubiquitously present on the cell surface and in the extracellular matrix. They consist of a core protein and heparan sulfate (HS) sugar chains covalently attached to it. HSPGs interact with a large number of signaling molecules through HS, thereby regulating a wide variety of biological functions (Perrimon and Bernfield, 2000; Bishop et al., 2007). A number of previous studies have revealed that HSPGs play critical roles in cell differentiation, migration, axon guidance, synapse development, and synaptic functions in the nervous system (Holt and Dickson, 2005; Condomitti and de Wit, 2018; Zhang et al., 2018; Kamimura and Maeda, 2021). Genetic or enzymatic ablation of HS impaired synaptic transmission and plasticity (Irie et al., 2012; Minge et al., 2017), suggesting that HS is involved in the regulation of neurotransmission, but the underlying mechanisms remain largely unknown.
HS is a linear polysaccharide that consists of repeating disaccharide units composed of uronic acid and glucosamine. It has many sulfate groups necessary for interaction with its binding proteins. Sulf1 and Sulf2 are extracellular sulfatases that remove 6-O-sulfates from mature HS chains. They modulate various cellular functions by changing the interaction between HS and signaling molecules through 6-O-desulfation of HS (Dhoot et al., 2001; Morimoto-Tomita et al., 2002; Lamanna et al., 2007; El Masri et al., 2017). Studies using knock-out (KO) mice have revealed critical roles of Sulfs in normal development: although Sulf1 or Sulf2 single KO mice appear largely normal, Sulf1/Sulf2 double KO mice die neonatally (Ai et al., 2007; Holst et al., 2007; Okada et al., 2017). Sulf1/Sulf2 double KO mice have defects in esophageal innervation and axon guidance of the corticospinal tract (Ai et al., 2007; Okada et al., 2017). In both Sulf1 and Sulf2 single KO mice, development of oligodendrocyte precursor cells in the embryonic spinal cord is impaired (Touahri et al., 2012; Jiang et al., 2017). As compared with these well-studied roles of Sulf1/Sulf2 genes in development, the functions of Sulf1/2 in the adult brain remain largely unknown, with only one report published on synaptic and behavioral abnormalities in Sulf1 KO mice (Kalus et al., 2009).
The nucleus accumbens (NAc) is a part of the basal ganglia located in the ventral forebrain. It plays important roles in appetitive and aversive responses (Floresco, 2015; Castro and Bruchas, 2019; Cox and Witten, 2019). The NAc receives glutamatergic inputs from the prefrontal cortex, ventral hippocampus, and basolateral amygdala, and the glutamatergic transmission is modulated by neurotransmitters such as dopamine from the ventral tegmental area (Britt et al., 2012; Floresco, 2015; Castro and Bruchas, 2019; Cox and Witten, 2019; Christoffel et al., 2021). More than 90% of neurons in the NAc are GABAergic medium spiny neurons (MSNs), which are subdivided into two groups: those expressing dopamine D1 receptors (D1-MSNs) and those expressing D2 receptors (D2-MSNs). A small number of MSNs express both D1 and D2 receptors. Many lines of previous clinical as well as animal model studies have revealed that the NAc is implicated in psychiatric disorders including schizophrenia, depression, obsessive–compulsive disorders, bipolar disorders, and drug addiction (Francis and Lobo, 2017; McCutcheon et al., 2019). Therefore, investigating the molecular mechanisms that regulate NAc activity is important from both basic and clinical perspectives.
In our previous study, we examined the expression patterns of Sulf1/Sulf2 mRNA in the adult mouse brain systematically (Miya et al., 2021). Although Sulf1 and Sulf2 expressions overlap largely in the embryonic brain (Okada et al., 2017), they are mostly segregated in the adult brain. We found that Sulf1 was highly expressed in the NAc, tail of the striatum, paraventricular nucleus of the thalamus, and prefrontal cortex (Miya et al., 2021). In addition, we showed that Sulf1 expression was detected in most Drd1- or Drd2-expressing neurons in the NAc and other brain regions mentioned above (Miya et al., 2021). We thus thought it might be possible to clarify the roles of Sulf1 in the adult brain by studying the effects of Sulf1 KO on NAc MSNs in which Sulf2 is not detected (Miya et al., 2021).
In this study, we examined the electrophysiological properties of MSNs in the NAc shell of adult mice by means of whole-cell patch–clamp recording. We show that Sulf1 disruption leads to changes in neuronal excitability and glutamatergic transmission in MSNs. These data suggest that Sulf1 is essential for regulation of neuronal activity and synaptic transmission of NAc MSNs in the mature brain.
Materials and Methods
Animals
Two- to twenty-seven-week-old male mice were used in this study. In situ hybridization data were obtained from three mice, and electrophysiological experiment data, from three to six mice. The experimenters were not blinded to genotype or cell identity. No statistical methods were used to predict the sample size before the study. All the animal experiments were approved by and performed according to the guidelines of the Animal Care and Use Committee of the University of Tsukuba (22-206, 22-211, 23-160, 23-165, 24-249, and 24-254) and of the Jikei University School of Medicine (2021-048C2 and 2022-009C1).
Sulf1 KO mice, Sulf1tm1Mmas (MGI 5318489), were generated by insertion of a cassette of stop-IRES-lacZ-poly(A) into the Sulf1 gene using homologous recombination in ES cells, and Sulf1 mRNA was shown to be completely abolished in the KO mice (Nagamine et al., 2012). Drd1-Cre, B6.FVB(Cg)-Tg(Drd1-cre)EY262Gsat/Mmucd (MGI 4358480), and Drd2-Cre, B6.FVB(Cg)-Tg(Drd2-cre)ER44Gsat/Mmucd (MGI 5003554; Gong et al., 2003; Heintz, 2004), maintained on a C57BL/6J background, were purchased from Mutant Mouse Resource and Research Centers. All these mice were backcrossed to C57BL/6N for 10 successive generations. Sulf1−/−;Drd1-Cre and Sulf1−/−;Drd2-Cre mice were generated by mating of mice carrying the Sulf1 KO allele with Drd1-Cre or Drd2-Cre mice. Genotypes were determined by the use of genomic PCR with tail biopsy.
Multiplex fluorescence in situ hybridization
Wild-type (WT) mice (15–19 weeks old) were deeply anesthetized with isoflurane and transcardially perfused with 4% paraformaldehyde in phosphate-buffered saline (PBS). The extracted brains were postfixed with the same solution at 4°C overnight. The brains were cryoprotected in 30% sucrose/PBS, embedded in Tissue-Tek OCT compound (Sakura Finetek), and stored at −80°C. Coronal brain sections (14 μm thick) were cut with a cryostat (CM 1850; Leica Biosystems) and collected onto MAS-coated slide glasses (Matsunami Glass Industry). Multiplex fluorescence in situ hybridization was performed by the use of an RNAscope Multiplex Fluorescent v2 Assay kit (Advanced Cell Diagnostics) according to the manufacturer's instructions. Specific probes for Sulf1 (Mm-Sulf1 #495411), Drd1 (Mm-Drd1a #406491-C2), and Drd2 (Mm-Drd2 #406501-C3) were used. Signals were developed by means of a TSA Vivid Fluorophore kit 650 for Sulf1, 570 for Drd1, and 520 for Drd2 (Tocris Bioscience). The sections were mounted with CC/Mount (Diagnostic BioSystems).
Images were acquired by the use of a laser scanning confocal microscope (LSM 700; Carl Zeiss) with a 10× objective lens for low-magnification images. Two regions of interest were selected from each of the three mice and analyzed by the use of a 20× objective lens and Z-stacking (1 μm intervals). Cells positive for Sulf1, Drd1, or Drd2 were marked manually, and their overlapping was analyzed as described previously (Miya et al., 2021).
Viral vectors
Adeno-associated virus (AAV) vectors were generated by use of the AAV Helper Free Expression System (Cell Biolabs) as reported previously (Sano et al., 2020). Briefly, HEK293T cells were transfected with the packaging plasmids (pAAV-RC5 and pHelper) and pAAV-hSyn-DIO-mCherry by means of a calcium phosphate method. The purified virus particles were concentrated with an Amicon 10K MWCO filter (Merck Millipore), and the copy number of the viral genome was determined by the use of PCR. In some experiments, the same vector purchased from Addgene (50459-AAV5; https://www.addgene.org/50459/) was used.
Stereotactic surgery
For stereotactic surgery, adult male mice (11–24 weeks old) were anesthetized with a mixture of midazolam, medetomidine, and butorphanol (4, 0.75, and 5 mg/kg body weight, respectively) and head-fixed on a stereotaxic frame (David Kopf Instruments). To label D1-MSNs and D2-MSNs, AAV5-hSyn-DIO-mCherry (1.1 × 1013 viral genome/ml, 0.5 μl) was injected into the bilateral NAc shell of the Drd1-Cre and Drd2-Cre mice, respectively, at a rate of 200 nl/min through a burr hole by the use of a pressure microinjector (KDS 101; KD Scientific). The stereotactic coordinates for the NAc shell (in mm, relative to the bregma) were AP 1.5, ML ±0.5, DV 4.5. The injection needle was kept in place for 5 min and then slowly retracted. Two to thirteen weeks after the virus injection, the mice were subjected to electrophysiological recordings. To study juvenile mice, male mice were injected at Postnatal Day (P)20 and analyzed at P28–34 by means of the same above-described method used in the adult mice.
Slice preparation
Brain slices were obtained from 17- to 27-week-old mice as reported previously (Suzuki and Momiyama, 2021; Nishijo et al., 2022). For juvenile mice, brain slices were obtained at P28–34 (4–5 weeks). Briefly, after the mice had been decapitated under deep isoflurane anesthesia and their brains quickly dissected, coronal brain slices (300 μm thick) were cut by means of a microslicer (LinearSlicer PRO-7; Dosaka) in an ice-cold oxygenated modified cutting solution with the following composition (in mM): 92 choline chloride; 2.5 KCl; 30 NaHCO3; 1.2 NaH2PO4; 20 HEPES; 25 d-glucose; 5 ascorbic acid; 2 thiourea; 3 sodium pyruvate; 12 N-acetyl-l-cysteine; 0.5 CaCl2; and 10 MgCl2, pH 7.2, adjusted with N-methyl-d(-)-glucamine. The slices containing the NAc region were transferred to a recovery chamber containing the modified cutting solution at 32–34°C for 10 min and then incubated at room temperature in a holding chamber containing standard Krebs’ solution with the following composition (in mM): 124 NaCl; 3 KCl; 26 NaHCO3;, 1 NaH2PO4; 2.4 CaCl2; 1.2 MgCl2; and 10 d-glucose, pH 7.4, when bubbled with 95% O2 and 5% CO2.
Whole-cell patch–clamp recording
The electrophysiological experiment data in each section were obtained from 7 to 14 neurons of 3–6 adult mice. For the juvenile mice, data in each strain were obtained from 12 to 13 neurons of three mice. The numbers of animals and cells used for each experiment are described in the corresponding figure legends or text. For recording, a brain slice was transferred to a recording chamber, held submerged, and superfused with standard Krebs’ solution (bubbled with 95% O2 and 5% CO2, 32–34°C) at a rate of 2–3 ml/min. The mCherry-positive cells in the NAc shell region were identified by means of an appropriate fluorescence filter (U-MWIG3; Olympus) and a 40× water immersion objective lens attached to an upright microscope (BX51WI; Olympus). Images were detected with a CCD camera (IR-1000; DAGE-MTI) and displayed on a video monitor (VU-17; DAGE-MTI). Considering the potential cytotoxicity associated with high levels of mCherry expression, cells with low to moderate fluorescence intensity were selected for recording. We performed whole-cell patch–clamp recordings from D1- and D2-MSNs in the NAc shell using a patch-clamp amplifier (MultiClamp 700B; Molecular Devices). Patch electrodes were pulled from Standard Wall Borosilicate Glass with Filament (1.5 mm outer diameter; Sutter Instrument) by means of a pipette puller (P-97; Sutter Instrument). They had resistances of 3–6 MΩ when filled with the internal solution. Data were low-pass filtered at 5 kHz and stored at 10–20 kHz. We used the pCLAMP system for data acquisition and Igor Pro 8 (WaveMetrics) with NeuroMatic (Rothman and Silver, 2018) and OriginPro2025 (OriginLab) for the analysis.
For current-clamp recording, patch pipettes were filled with a solution containing the following (in mM): 135 potassium gluconate; 6 NaCl; 10 KCl; 10 K-HEPES; 2 Mg-ATP; 0.3 Na2-GTP; and 0.1 K-EGTA, pH 7.3, adjusted with 1 M KOH. A liquid junction potential (LJP) of 15.5 mV was corrected offline after the experiments by means of pCLAMP software (version 10; Molecular Devices). For voltage-clamp recording, patch pipettes were filled with a solution containing the following (in mM): 140 CsCl; 9 NaCl; 1 Cs-EGTA; 10 Cs-HEPES; and 2 Mg-ATP, pH 7.3, adjusted with 1 M CsOH. The LJP was 5 mV, which was not compensated. The range of series resistance was 4.1–30 MΩ (11.18 ± 1.30 MΩ) in juvenile MSNs and 6.3–35 MΩ (17.50 ± 1.48 MΩ) in adult MSNs in the current-clamp mode and 4.87–28.3 MΩ (12.66 ± 0.89 MΩ) in adult MSNs in the voltage-clamp mode. These values were not compensated. Data were excluded from the analysis if the series resistance changed by >20% of the initial value.
Membrane resistance was obtained from the steady-state voltage response to a −200 pA current injection, with bridge balance compensation performed during current-clamp recordings. Increasing step currents from +20 to +440 pA in increments of 20 pA were injected for 500 ms to assess membrane potential responses under the current-clamp mode. The action potential amplitude was determined by measuring the difference between the peak membrane potential and the threshold potential. The rheobase was defined as the minimum depolarizing current required to evoke an action potential during step current injections. We performed phase plot analysis (Trombin et al., 2011) and determined the threshold for the action potential to be the point at which the membrane potential changed by >10 mV within 1 ms. Latency was determined by measuring the time from the start of current injection to the point at which the membrane potential reached the action potential threshold. The afterhyperpolarization (AHP) amplitude was defined as the difference between the threshold of the action potential and the peak of the hyperpolarization.
To evoke glutamatergic excitatory postsynaptic currents (EPSCs), we placed a glass pipette filled with 1 M NaCl 100–200 µm from the recorded neuron, and we delivered electrical stimulations every 5 s (100 μs duration, 0.2 Hz) in the presence of 10 μM bicuculline (Tocris Bioscience) and 0.5 μM strychnine (Sigma-Aldrich) to block GABAA and glycine receptors, respectively. The stimulation intensity was adjusted to evoke EPSCs with an amplitude of approximately −100 and +100 pA at holding potentials of −80 and +40 mV, respectively. The AMPA current component of the evoked EPSCs was obtained by applying 25 μM d-AP5 (Tocris Bioscience) at a holding potential of +40 mV. The NMDA current was calculated by subtracting the AMPA current component from the EPSC recorded at +40 mV. The AMPA/NMDA ratio was then determined by dividing the peak amplitude of the AMPA current component by that of the NMDA current component. Bicuculline, strychnine, and d-AP5 were stored as frozen stock solutions and dissolved in the perfusing solution just before application in the final concentrations indicated.
Statistical analysis
Data were analyzed by the use of Igor Pro 8, and statistical analyses were performed with OriginPro2025. For comparison of the electrophysiological parameters between the WT and Sulf1 KO mice, statistical analyses were done with the nonparametric Mann–Whitney U test. To compare the data of multiple current stimulus intensities (Figs 2C, 3B,G, 4B,G), statistical analyses were done with two-way mixed–model ANOVA followed by the Bonferroni’s post hoc test. Statistical differences were considered significant at p < 0.05. All the statistical values are described in the corresponding figure legends or text and summarized in Table 1.
Results from statistical analyses
Results
Sulf1 expression in the NAc shell
We previously reported that Sulf1 mRNA is highly expressed in the NAc of the adult mouse brain (Miya et al., 2021). We also showed coexpression of Sulf1 and dopamine D1 and D2 receptors by the use of combination of immunohistochemistry and AAV vector-mediated labeling of D1- and D2-MSNs. These experiments revealed that Sulf1 is expressed abundantly in most of the D1- and D2-MSNs in the NAc shell (Miya et al., 2021). In this study, to directly confirm the overlapping expression of Sulf1, Drd1, and Drd2 mRNAs, we performed multiplex fluorescence in situ hybridization. When observed at low magnification, strong signals for Drd1 and Drd2 were observed throughout the NAc and the Sulf1 signals mainly in the NAc shell, with higher density in its dorsomedial portion (Fig. 1A). At high magnification, both the Drd1 and the Drd2 signals overlapped with the Sulf1 signals (Fig. 1B). The vast majority of Drd1+ and Drd2+ cells coexpressed Sulf1: 98.9% of Drd1+ cells (n = 264) and 99.5% of Drd2+ cells (n = 215) were positive for Sulf1. Among the Sulf1-expressing cells, when Drd1/Drd2 double-positive cells were included, 56.9% were positive for Drd1 and 46.3% for Drd2 (Fig. 1C). These data were close to the reported fractions of D1-MSNs and D2-MSNs in the NAc shell (Bertran-Gonzalez et al., 2008; Kupchik et al., 2015; Miyasaka et al., 2025). About 4% were positive for both Drd1 and Drd2, similar to the percentages reported previously (Kupchik et al., 2015; Miyasaka et al., 2025). These results indicate that almost all the MSNs in the NAc shell express Sulf1.
Expression of Sulf1 mRNA in D1- and D2-MSNs. A, Representative images of triple fluorescence in situ hybridization for Drd1 (red), Drd2 (green), and Sulf1 (magenta). Broken lines indicate the border of the NAc shell. Asterisks indicate the major island of Calleja. ac, anterior commissure; NAcC, nucleus accumbens core; NAcSh, nucleus accumbens shell. B, Magnified images of the boxed region in A. White and yellow arrowheads indicate the representative cells coexpressing Sulf1 and Drd1, and Sulf1 and Drd2, respectively. C, Percentages of Drd1/Drd2-expressing cells in the Sulf1-expressing cells in the NAc shell. The pie chart shows percentages of Drd1 single-positive (Drd1+, red), Drd2 single-positive (Drd2+, green), Drd1/Drd2 double-positive (Drd1+/Drd2+, yellow), and Drd1/Drd2 double-negative (Drd1−/Drd2−, dark magenta) cells in the Sulf1-expressing cells (n = 464, 6 slices from 3 mice). Scale bars, A, 400 μm; B, 50 μm.
Differences of electrophysiological properties in D1-MSNs and D2-MSNs
To study the roles of Sulf1 in the NAc, we decided to examine the electrophysiological properties of MSNs by means of whole-cell patch–clamp recording. Considering the possibility that Sulf1 KO may have different effects on D1-MSNs and D2-MSNs, we adopted a strategy of recording them separately. To this end, we labeled D1-MSNs and D2-MSNs separately by stereotactic injection of a Cre-dependent AAV vector expressing a red fluorescent protein, mCherry, into Drd1-Cre and Drd2-Cre mice homozygous for the WT or KO allele of the Sulf1 gene. Two to thirteen weeks after virus injection, we identified mCherry-positive cells in the medial shell region of the NAc in 300 μm-thick slices and performed whole-cell patch–clamp recordings.
Before starting the analysis of Sulf1 KO mice, we compared the membrane properties of D1-MSNs and D2-MSNs in the adult WT mice because several studies have reported electrophysiological differences between D1-MSNs and D2-MSNs in the dorsal striatum (Ma et al., 2012; Schier et al., 2017; Willett et al., 2019) and NAc (Francis et al., 2015; Al-Muhtasib et al., 2018; Cao et al., 2018). When the resting membrane potentials were measured under the current-clamp mode, the D1-MSNs were significantly more hyperpolarized than were the D2-MSNs, whereas the membrane resistances did not differ significantly between them (Fig. 2A,B). We next examined the neuronal firing responses to depolarizing current injections. The D1-MSNs showed a higher spike frequency than that of the D2-MSNs at low injected currents (100–240 pA, not statistically significant) but did not show an increase in the spike number at currents larger than 240 pA (Fig. 2C). In contrast, in the D2-MSNs, the spike numbers increased linearly as the current increased and the spike frequency was higher than in the D1-MSNs at large injected currents (Fig. 2C; D1-MSN, n = 10 neurons from four mice; D2-MSN, n = 7 neurons from four mice; main effect of the cell type, F(1,15) = 8.21; p = 0.012; main effect of current, F(21,315) = 17.9; p < 0.0001; interaction, F(21,315) = 9.05; p < 0.0001; two-way mixed–model ANOVA). No differences were observed between the D1-MSNs and D2-MSNs in the rheobase, the threshold of the first action potential during current injection, and the AHP amplitude at the rheobase (Fig. 2D–F). These results suggest that D1-MSNs and D2-MSNs in the NAc shell of the adult WT mice have distinct membrane properties and excitabilities.
Comparison of electrophysiological properties between D1-MSNs and D2-MSNs of the WT mice. A, Resting membrane potential (V rest). The V rest in the D1-MSNs (−96.09 ± 1.49 mV, n = 10 neurons from 4 mice) was significantly more hyperpolarized than that in the D2-MSNs (−87.70 ± 1.76 mV, 7 neurons from 4 mice; U = 8; p = 0.0097; Mann–Whitney U test). Open circles and squares indicate D1-MSNs and D2-MSNs, respectively (same as below). B, Membrane resistance (Rm). The Rm were 113.64 ± 15.54 MΩ in the D1-MSNs (n = 10 neurons from 4 mice) and 94.71 ± 11.41 MΩ in the D2-MSNs (n = 7 neurons from 4 mice). The Rm did not differ significantly between the groups (U = 30; p = 0.66; Mann–Whitney U test). C, Left, action potentials of D1-MSNs (top) and D2-MSNs (bottom) in WT mice. Membrane potential changes were elicited by depolarizing currents (+20 to +440 pA in 20 pA increments, 500 ms duration). Scale bars, 100 ms and 20 mV. The traces when action potentials were generated at a minimum injection current are shown in color (magenta in D1-MSNs and green in D2-MSNs). The top-right panel shows a part of the current injection protocol from +20 to +440 pA in 60 pA increments (scale bars, 100 ms and 60 pA). Right, relationship between injected currents and number of action potentials. At 360–440 pA current injections, the numbers of action potentials were significantly higher in the D2-MSNs (n = 7 neurons from 4 mice) than in the D1-MSNs (10 neurons from 4 mice; main effect of cell type, F(1,15) = 8.21; p = 0.012; main effect of current, F(21,315) = 17.9; p < 0.0001; interaction; F(21,315) = 9.05; p < 0.0001; Bonferroni’s post hoc test, at 360 pA, t = 4.86; p = 0.0018; at 380 pA; t = 5.60; p < 0.0001; at 400 pA, t = 6.15; p < 0.0001; at 420 pA, t = 6.64; p < 0.0001; and at 440 pA, t = 6.77; p < 0.0001; two-way mixed–model ANOVA). D, Rheobase in D1-MSNs and D2-MSNs. The rheobases were 172.00 ± 29.84 pA in the D1-MSNs (n = 10 neurons from 4 mice) and 188.57 ± 22.62 pA in the D2-MSNs (n = 7 neurons from 4 mice). They did not differ significantly between the groups (U = 25.5; p = 0.38; Mann–Whitney U test). E, Threshold of the first action potential elicited by current injection. The thresholds were −54.35 ± 1.81 mV in the D1-MSNs (n = 10 neurons from 4 mice) and −59.54 ± 1.54 mV in the D2-MSNs (n = 7 neurons from 4 mice). They did not differ significantly between the groups (U = 17; p = 0.088; Mann–Whitney U test). F, Amplitude of AHP of the first action potential elicited by the current injection. The AHP amplitudes were −14.36 ± 1.01 mV in the D1-MSNs (n = 10 neurons from 4 mice) and −10.88 ± 2.00 mV in the D2-MSNs (n = 7 neurons from 4 mice). They did not differ significantly between the groups (U = 24; p = 0.31; Mann–Whitney U test). These data are a reanalysis of the WT mouse data presented in Figures 3 and 4. Data are presented as means ± SEMs. **p < 0.01; †p < 0.0001. Extended Data Figure 2-1 shows comparison of electrophysiological properties between D1-MSNs and D2-MSNs of juvenile WT mice. Extended Data Figure 2-2 shows the results of statistical analyses of juvenile mouse studies in Extended Data Figure 2-1.
Figure 2-1
Comparison of electrophysiological properties between D1-MSNs and D2-MSNs of juvenile WT mice. A, Resting membrane potential (V rest). The V rest did not differ between the D1-MSNs (−91.61 ± 1.21 mV, n = 12 neurons from 3 mice) and D2-MSNs (−90.30 ± 1.89 mV, 13 neurons from 3 mice; U = 71, p = 0.72; Mann-Whitney U test). Open circles and squares indicate D1-MSNs and D2-MSNs, respectively (same as below). B, Membrane resistance (Rm). The Rm did not differ between the D1-MSNs (140.73 ± 19.78 MΩ, n = 12 neurons from 3 mice) and D2-MSNs (105.58 ± 11.76 MΩ, n = 13 neurons from 3 mice; U = 52, p = 0.17; Mann-Whitney U test). C, Left, action potentials of D1-MSNs (top) and D2-MSNs (bottom) in juvenile WT mice. Membrane potential changes were elicited by depolarizing currents (+20 to +440 pA in 20 pA increments, 500-ms duration). Scale bars, 100 ms and 20 mV. The traces when action potentials were generated at a minimum injection current are shown in color (magenta in D1-MSNs and green in D2-MSNs). The upper right panel shows a part of the current injection protocol from +20 pA to +440 pA in 60-pA increments (scale bars, 100 ms and 60 pA). Right, relationship between injected currents and number of action potentials. They did not differ significantly between the groups (main effect of cell type, F(1,23) = 0.42, p = 0.53; main effect of current, F(21,483) = 33.8, p < 0.0001; interaction; F(21,483) = 0.31, p = 0.999). Download Figure 2-1, TIF file.
Figure 2-2
Results of statistical analyses of juvenile mouse studies in Figure 2-1. Download Figure 2-2, DOCX file.
The excitability of D1-MSNs and D2-MSNs in the adult NAc shell described above differs significantly from those reported in the dorsal striatum (Ma et al., 2012; Schier et al., 2017; Willett et al., 2019) and NAc core (Al-Muhtasib et al., 2018; Cao et al., 2018): in previous reports, D1-MSNs were generally harder to excite than were D2-MSNs. Because those previous studies used juvenile mice, we considered that this might be due to differences in the age of the mice analyzed. We thus performed electrophysiological recording of the D1-MSNs and D2-MSNs in the NAc shell of the juvenile mice. For this purpose, we injected Drd1-Cre and Drd2-Cre mice with AAV-DIO-mCherry at P20 and performed slice patch-clamp recording at P28–34 by means of the same method used in the adult mice. As a result, we found no significant differences in the resting membrane potential and membrane resistance between the D1-MSNs and D2-MSNs (Extended Data Figs. 2-1A,B, 2-2). In addition, the numbers of action potentials after current injection did not differ significantly between the D1-MSNs and D2-MSNs (Extended Data Figs. 2-1C, 2-2).
Effects of Sulf1 disruption on the intrinsic electrophysiological properties and neuronal excitability of NAc D1-MSNs
Next, we compared the intrinsic electrophysiological properties of D1-MSNs between the adult WT and Sulf1 KO mice. The resting membrane potentials were significantly more depolarized in the Sulf1 KO mice than in the WT mice (Fig. 3A). No difference in the membrane resistance was observed between the WT and the Sulf1 KO mice (Fig. 3A). We then examined the effects of Sulf1 KO on the neuronal excitability of D1-MSNs. The number of action potentials in the D1-MSNs to depolarizing current injections from +20 to +440 pA did not differ significantly between the WT and the Sulf1 KO mice (Fig. 3B; WT, n = 10 neurons from four mice; KO, n = 11 neurons form four mice; main effect of genotype, F(1,19) = 0.41; p = 0.53; main effect of current, F(21,399) = 8.51; p < 0.0001; interaction, F(21,399) = 0.12; p = 1; two-way mixed-model ANOVA). However, we noticed that in some D1-MSNs, the action potentials evoked by large current injections exhibited attenuation of the amplitude, and it seemed that the tendency was pronounced in the KO mice. To investigate this point in detail, we compared the amplitudes of the first and second action potentials evoked by a 440 pA current injection in the WT and KO mice. In both genotypes, the amplitude of the second action potentials was lower than that of the first action potentials, and there was no significant difference in the reduction ratio between the genotypes (Fig. 3C; U = 37; p = 0.22; Mann–Whitney U test). No significant differences were observed between the WT and the KO mice in the rheobase (Fig. 3D) nor in the latency and threshold of the first action potential evoked by a 440 pA current injection (Fig. 3E,F). We next compared the AHP amplitude of the first action potentials at the rheobase and at a 440 pA current injection. In both the WT and the KO mice, the AHP amplitude of the first action potentials at the 440 pA current injection was smaller than that at the rheobase (Fig. 3G; main effect of genotype, F(1,19) = 0.75; p = 0.40; main effect of AHP, F(1,19) = 37.2; p < 0.0001; interaction, F(1,19) = 0.000001; p = 0.999; two-way mixed–model ANOVA). These results suggest that Sulf1 disruption resulted in changes in the resting membrane potential of the D1-MSNs.
Electrophysiological properties and neuronal excitability of D1-MSNs in WT and Sulf1 KO mice. A, Top, resting membrane potential (V rest) of D1-MSNs. Open and closed circles indicate control WT and Sulf1 KO mice, respectively (same as below). The V rest in the KO mice (−88.63 ± 1.87 mV, n = 11 neurons from 4 mice) was significantly more depolarized than that in the WT mice (−96.09 ± 1.49 mV, n = 10 neurons from 4 mice; U = 19; p = 0.012; Mann–Whitney U test). Bottom, membrane resistance (Rm) of D1-MSNs. The Rm were 113.64 ± 15.54 MΩ in the WT mice (n = 10 neurons from 4 mice) and 156.40 ± 21.54 MΩ in the KO mice (n = 11 neurons from 4 mice). The Rm did not differ significantly between the groups (U = 27; p = 0.053; Mann–Whitney U test). B, Left, action potentials of D1-MSNs of the WT (top) and KO (bottom) mice. Membrane potential changes were elicited by depolarizing currents (+20 to +440 pA in 20 pA increments, 500 ms duration). Scale bars, 100 ms and 20 mV. The traces when action potentials were generated at a minimum injection current are shown in magenta. The top-right panel shows a part of the current injection protocol from +20 to +440 pA in 60 pA increments (scale bars, 100 ms and 60 pA). Right, relationship between injected currents and number of action potentials. The number of action potentials did not differ between the WT and KO mice (main effect of genotype, F(1,19) = 0.41; p = 0.53; main effect of current, F(21,399) = 8.51; p < 0.0001, interaction, F(21,399) = 0.12; p = 1). C, Comparison of the first and second action potentials. Left, action potentials evoked by a 440 pA current injection (500 ms duration) in the WT (top) and KO (bottom) mice. Right, the ratio of the second and first action potential amplitudes. There was no significant difference in the reduction ratio between the groups (U = 37; p = 0.22; Mann–Whitney U test). D, Rheobase in WT and KO mice. The rheobases were 172.00 ± 29.84 pA in the WT mice (n = 10 neurons from 4 mice) and 141.82 ± 16.94 pA in the KO mice (n = 11 neurons from 4 mice). They did not differ significantly between the groups (U = 48; p = 0.64; Mann–Whitney U test). E, Latency of the first action potential evoked by a 440 pA current injection. The latencies were 26.90 ± 10.30 ms in the WT mice (n = 10 neurons from 4 mice) and 8.69 ± 2.10 ms in the KO mice (n = 11 neurons from 4 mice). They did not differ significantly between the groups (U = 47; p = 0.60; Mann–Whitney U test). F, Threshold of the first action potential evoked by a 440 pA current injection. The thresholds were −53.40 ± 3.28 mV in the WT mice (n = 10 neurons from 4 mice) and −54.48 ± 2.30 mV in the KO mice (n = 11 neurons from 4 mice). They did not differ significantly between the groups (U = 54; p = 0.97; Mann–Whitney U test). G, AHP amplitude of the first action potential at the rheobase and a 440 pA current injection. The AHP amplitudes at the rheobase were −14.39 ± 1.03 mV in the WT mice (n = 10 neurons from 4 mice) and −12.39 ± 1.77 mV in the KO mice (n = 11 neurons from 4 mice). The AHP amplitudes at a 440 pA current injection were −5.70 ± 2.49 mV in the WT mice (n = 10 neurons from 4 mice) and −3.69 ± 2.06 mV in the KO mice (n = 11 neurons from 4 mice). The AHP amplitudes of the first action potentials evoked by a 440 pA current injection were smaller than those at the rheobase (main effect of genotype, F(1,19) = 0.75; p = 0.40; main effect of AHP, F(1,19) = 37.2,; p < 0.0001; interaction, F(1,19) = 0.000001; p = 0.999; two-way mixed–model ANOVA). *p < 0.05.
Effects of Sulf1 disruption on the intrinsic electrophysiological properties and neuronal excitability of NAc D2-MSNs
Next, we compared the intrinsic electrophysiological properties of the D2-MSNs of the adult WT and the Sulf1 KO mice. The D2-MSNs showed no significant differences in resting membrane potentials and membrane resistances between the WT and the Sulf1 KO mice (Fig. 4A). We then analyzed the effects of Sulf1 disruption on the excitability of D2-MSNs. When the membrane potential responses to depolarizing current injections (from +20 to +440 pA) were measured, marked differences were observed in the firing patterns between the WT mice and the Sulf1 KO mice (Fig. 4B). In the Sulf1 KO mice, the action potentials occurred at lower currents and were more frequent than in the WT mice between currents of 100 and 280 pA (Fig. 4B, not statistically significant). However, at a higher current (>300 pA), the number of action potentials did not increase in the Sulf1 KO mice, whereas that in the WT mice increased as the stimulus currents became larger (Fig. 4B). These differences between the WT mice and the Sulf1 KO mice were statistically significant (Fig. 4B; WT, n = 7 neurons from four mice; KO, n = 14 neurons from five mice; F(1,19) = 1.01; p = 0.33; main effect of current, F(21,399) = 26.2; p < 0.0001; interaction, F(21,399) = 5.83; p < 0.0001; two-way mixed–model ANOVA). The Bonferroni’s post hoc analysis revealed that the D2-MSNs in the WT mice generated more spikes than those in the Sulf1 KO mice at 420 and 440 pA current injections (Fig. 4B; at 420 pA; t = 4.41; p = 0.013, at 440 pA; t = 4.41; p = 0.013). We then compared the amplitudes of the first and second action potentials evoked by a 440 pA current injection (Fig. 4C). In both the WT and the Sulf1 KO mice, the amplitudes of the second action potentials were smaller than those of the first action potentials, and the reduction ratio was significantly higher in the KO mice than in the WT mice (U = 22; p = 0.048; Mann–Whitney U test). The rheobase in the D2-MSNs of the Sulf1 KO mice was significantly smaller than that in the WT mice (Fig. 4D). No differences were observed between the genotypes in the latency and threshold of the first action potential evoked by a 440 pA current injection (Fig. 4E,F). We next compared the AHP amplitude of the first action potentials at the rheobase and at a 440 pA current injection. In both the WT and the KO mice, the AHP amplitudes of the first action potentials at a 440 pA current injection appeared to be smaller than those at the rheobase (Fig. 4G; main effect of genotype, F(1,19) = 0.14; p = 0.71; main effect of AHP, F(1,19) = 39.2; p < 0.0001; interaction, F(1,19) = 17.9; p = 0.00046; two-way mixed–model ANOVA). However, the Bonferroni’s post hoc test revealed that the degree of the decrease was significant in the KO mice but not in the WT mice (Fig. 4G; WT, t = 1.03; p = 1; KO, t = 7.52; p < 0.0001). These results suggest that the D2-MSNs of Sulf1 KO mice are more easily activated by weak stimuli but tend to be inactivated by strong excitations.
Electrophysiological properties and neuronal excitability of D2-MSNs in WT and Sulf1 KO mice. A, Top, resting membrane potentials (V rest) of D2-MSNs. The open and closed squares indicate control WT and Sulf1 KO mice, respectively (same as below). The V rests were −87.70 ± 1.76 mV in the WT mice (n = 7 neurons from 4 mice) and −92.65 ± 1.88 mV in the KO mice (n = 14 neurons from 5 mice). They did not differ significantly between the groups (U = 30; p = 0.17; Mann–Whitney U test). Bottom, membrane resistances (Rm) of D2-MSNs. The Rm were 94.7 ± 11.41 MΩ in the WT mice (n = 7 neurons from 4 mice) and 115.43 ± 10.93 MΩ in the KO mice (n = 14 neurons from 5 mice). They did not differ significantly different between the groups (U = 37; p = 0.39; Mann–Whitney U test). B, Left, action potentials of D2-MSNs in WT (top) and KO (bottom) mice. Membrane potential changes were elicited by depolarizing currents (+20 to +440 pA in 20 pA increments, 500 ms duration). Scale bars, 100 ms and 20 mV. The traces when action potentials were generated at a minimum injection current are shown in green. The top-right panel shows a part of the current injection protocol from +20 to +440 pA in 60 pA increments (scale bars, 100 ms and 60 pA). Right, relationship between injected currents and number of action potentials. At 420 and 440 pA current injections, the numbers of action potentials were significantly higher in the WT mice (n = 7 neurons form 4 mice) than those in the KO mice (n = 14 neurons form 5 mice; main effect of genotype, F(1,19) = 1.01; p = 0.33; main effect of current, F(21,399) = 26.2; p < 0.0001; interaction, F(21,399) = 5.83; p < 0.0001; Bonferroni’s post hoc test, at 420 pA, t = 4.41; p = 0.013; at 440 pA, t = 4.41; p = 0.013; two-way mixed–model ANOVA). C, Comparison of the first and second action potentials. Left, action potentials evoked by a 440 pA current injection (500 ms duration) in the WT (top) and KO (bottom) mice. Right, The ratio of the second and first action potentials. The reduction ratio was significantly higher in the KO mice than in the WT mice (U = 22; p = 0.048; Mann–Whitney U test). D, Rheobase in WT and KO mice. The rheobase in the KO mice (130.00 ± 9.55 pA; n = 14 neurons from 5 mice) was significantly smaller than that in the WT mice (188.57 ± 22.62 pA; n = 7 neurons from 4 mice; U = 17, p = 0.011; Mann–Whitney U test). E, Latency of the first action potential evoked by a 440 pA current injection. The latencies were 19.77 ± 6.46 ms in the WT mice (n = 7 neurons from 4 mice) and 10.21 ± 3.19 ms in the KO mice (n = 14 neurons from 5 mice). They did not differ significantly between the groups (U = 32.5; p = 0.23; Mann–Whitney U test). F, Threshold of the first action potential evoked by a 440 pA current injection. The thresholds were −58.37 ± 1.20 mV in the WT mice (n = 7 neurons from 4 mice) and −58.54 ± 2.23 mV in the KO mice (n = 14 neurons from 5 mice). They did not differ significantly between the groups (U = 42; p = 0.63; Mann–Whitney U test). G, AHP amplitude of the first action potential evoked at the rheobase and 440 pA current injection. The AHP amplitudes at the rheobase were −10.88 ± 2.00 mV in the WT mice (n = 7 neurons from 4 mice) and −15.43 ± 0.81 mV in the KO mice (n = 14 neurons from 5 mice). The AHP amplitudes of the first action potential evoked by a 440 pA current injection were −8.35 ± 2.99 mV in the WT mice (n = 7 neurons from 4 mice) and −2.43 ± 1.09 mV in the KO mice (n = 14 neurons from 5 mice). The AHP amplitude of the first action potential at a 440 pA current injection was significantly smaller than that at the rheobase in the KO mice but not in the WT mice (main effect of genotype, F(1,19) = 0.14; p = 0.71; main effect of AHP, F(1,19) = 39.2; p < 0.0001; interaction; F(1,19) = 17.9; p = 0.00046; Bonferroni’s post hoc test, AHP within WT, t = 1.03; p = 1; AHP within KO, t = 7.52; p < 0.0001; two-way mixed–model ANOVA). Data are presented as means ± SEMs. *p < 0.05; †p < 0.0001.
Glutamatergic synaptic transmission in Sulf1 KO mice
We finally examined the effects of Sulf1 disruption on excitatory synaptic transmission onto MSNs in the NAc of adult mice. The cells were voltage clamped at −80 mV, and glutamatergic EPSCs were evoked by focal electrical stimulation at 0.2 Hz in the presence of bicuculline (10 µM) and strychnine (0.5 µM). The EPSC amplitudes in the D1-MSNs were −137.59 ± 33.19 pA in the WT mice (n = 10 neurons from three mice) and −158.50 ± 30.29 pA in the KO mice (n = 11 neurons from six mice), and those in the D2-MSNs were −119.33 ± 16.61 in the WT mice (n = 10 neurons from four mice) and −218.21 ± 28.43 pA in the KO mice (n = 11 neurons from six mice). These data showed that normal glutamatergic transmission was maintained in the Sulf1 KO mice.
Next, the holding potential was changed to +40 mV, and the EPSCs, which contained both AMPA and NMDA receptor-mediated components, were measured. Subsequently, AMPA receptor-mediated components were isolated by application of an NMDA receptor antagonist (d-AP5, 25 µM), and NMDA receptor-mediated components were obtained by electrical subtraction of AMPA components from the mixed EPSCs (Fig. 5A,B). To obtain insights into the synaptic functions of MSNs, the AMPA/NMDA ratio was calculated. In the D1-MSNs, the AMPA/NMDA ratio was significantly higher in the KO mice (1.75 ± 0.21; n = 11 neurons from five mice) than in the WT mice (Fig. 5A; 0.85 ± 0.10, n = 10 neurons from three mice; U = 10; p = 0.0017; Mann–Whitney U test). In the D2-MSNs, the AMPA/NMDA ratio appeared to be slightly higher in the KO mice (2.32 ± 0.38; n = 11 neurons from six mice) than in the WT mice (1.46 ± 0.18, n = 10 neurons from four mice), but the decrease was not significant (Fig. 5B; U = 27; p = 0.053; Mann–Whitney U test). These results suggest that Sulf1 disruption altered the synaptic strength in the D1-MSNs.
Glutamatergic transmission in D1-MSNs and D2-MSNs of WT and Sulf1 KO mice. A, Left, AMPA and NMDA EPSCs recorded under a voltage clamp at +40 mV in D1-MSNs. Scale bars, 100 ms and 20 pA. Middle, amplitude of AMPA EPSCs and NMDA EPSCs in D1-MSNs. The amplitudes of AMPA and NMDA EPSCs in the WT mice (n = 10 neurons from 3 mice) were 35.62 ± 5.71 pA and 44.37 ± 5.40 pA, respectively. The amplitudes of AMPA and NMDA EPSCs in the KO mice (n = 11 neurons from 5 mice) were 78.72 ± 10.72 pA and 48.44 ± 6.36 pA, respectively. Statistical analyses were not performed on the amplitudes of AMPA and NMDA EPSCs because these responses were elicited by stimuli of different sizes for each cell to obtain ∼100 pA responses (same as in B). Right, The AMPA/NMDA ratio. The AMPA/NMDA ratio in the D1-MSNs of the KO mice (1.75 ± 0.21; n = 11 neurons from 5 mice) was significantly higher than that in the WT mice (0.85 ± 0.10; n = 10 neurons from 3 mice; U = 10; p = 0.0017; Mann–Whitney U test). B, Left, AMPA and NMDA EPSCs recorded under a voltage clamp at +40 mV in D2-MSNs. Scale bars, 100 ms and 20 pA. Middle, Amplitude of AMPA and NMDA EPSCs in D2-MSNs. The amplitudes of AMPA and NMDA EPSCs in the WT mice (n = 10 neurons from 4 mice) were 57.09 ± 7.16 pA and 40.16 ± 3.10 pA, respectively. The amplitudes of AMPA and NMDA EPSCs in the KO mice (n = 11 neurons from 6 mice) were 73.45 ± 6.25 pA and 37.56 ± 4.22 pA, respectively. Right, The AMPA/NMDA ratio in the D2-MSNs of the KO mice (2.32 ± 0.38; n = 11 neurons from 6 mice) did not differ significantly from that of the WT mice (1.46 ± 0.18; n = 10 neurons from 4 mice; U = 27; p = 0.053; Mann–Whitney U test). Data are presented as means ± SEMs. **p < 0.01.
Discussion
In the present study, we have clarified the physiological roles of Sulf1 in the NAc shell by examining the electrophysiological properties of MSNs in Sulf1 KO mice. Sulf1 disruption caused changes in neuronal excitability and glutamatergic transmission in the MSNs of the adult mice. These data clearly demonstrate the functions of Sulf1 in central neuronal activities and synaptic transmission in the adult brain.
We first performed multiplex in situ hybridization and showed the overlapping expression of Sulf1, Drd1, and Drd2 in the NAc shell. As shown in Figure 1, most of the Drd1- or Drd2-expressing cells expressed Sulf1 mRNA. Among the Sulf1-expressing cells, ∼60 and 40% expressed Drd1 and Drd2, respectively, which is similar to the percentages reported previously (Bertran-Gonzalez et al., 2008; Thibault et al., 2013; Kupchik et al., 2015; Gagnon et al., 2017). The percentage of Drd1/Drd2 double-positive cells in the NAc shell was ∼4% in the present study, which is lower than the percentages in some studies (17% in Bertran-Gonzalez et al., 2008; 15% in Gagnon et al., 2017) but similar to those in other studies (1.5% in Thibault et al., 2013; 1.8% in Kupchik et al., 2015; 2.6% in Miyasaka et al., 2025). The discrepancy in the coexpression percentages is likely derived from the differences in the methods and mice used in the studies as well as from the difference in the analyzed subregion of the NAc shell (Gagnon et al., 2017).
D1-MSNs and D2-MSNs in the dorsal striatum and NAc differ in their electrophysiological properties (Ma et al., 2012; Francis et al., 2015; Schier et al., 2017; Al-Muhtasib et al., 2018; Cao et al., 2018; Willett et al., 2019) in addition to their gene expression, projection, and functions. In this study, we compared the intrinsic membrane properties and excitability between the D1-MSNs and D2-MSNs of adult WT mice because we have successfully established a method for preparing brain slices containing healthy neurons from adult mice by modifying the composition of a cutting Krebs’ solution (Suzuki and Momiyama, 2021). The D1-MSNs showed more hyperpolarized resting membrane potential than that of the D2-MSNs. In addition, the D1-MSNs generated more action potentials than those of the D2-MSNs at low current injections (100–240 pA), but the number of action potentials did not increase as a consequence of the injected currents being increased. In contrast, in the D2-MSNs, the number of action potentials increased as the stimulation currents increased. The D2-MSNs showed a higher rheobase but stronger activity at large depolarizing current injections than those of the D1-MSNs. The relationship between the injection currents and action potential numbers in our present study differs from those in most of the previous studies reporting that D2-MSNs in the dorsal striatum and NAc generate action potentials at a lower threshold and are more excitable than D1-MSNs [reviewed in Gertler et al. (2008) and Kreitzer (2009)]. We wondered whether this might be due to the differences in the age of the analyzed mice because juvenile mice aged 2–4 weeks were used in most slice patch-clamp recording studies, whereas we used adult mice aged 17–27 weeks. We thus performed the analysis of juvenile mice at P28–34 and found no significant differences in membrane properties and excitability between D1-MSNs and D2-MSNs. These data imply that excitability of D1-MSNs and D2-MSNs in the NAc shell is developmentally regulated and that the differences in excitability between D1-MSNs and D2-MSNs exist in the dorsal striatum and NAc core, but not in the NAc shell. The distinct electrophysiological properties of MSNs between the dorsal striatum/NAc core and NAc shell are a novel and interesting finding that should be examined in detail in future studies.
Disruption of the Sulf1 gene resulted in changes in neuronal excitability in the adult brain, which was more pronounced in D2-MSNs. In D2-MSNs, the rheobase was significantly smaller, and more action potentials were generated at low currents in the Sulf1 KO mice than in the WT mice, However, at larger current injections, the number of action potentials did not increase in proportion to the magnitude of the stimulation currents in the D2-MSNs of Sulf1 KO mice. In addition, in the D2-MSNs, attenuation of the action potentials during current injections was more pronounced in the KO mice than in the WT mice. Meanwhile, in the D1-MSNs, attenuation of the action potentials was also observed in the recordings of the Sulf1 KO mice, though the reduction ratio was not significantly higher than in the WT mice. These data indicate that Sulf1 disruption affects firing properties in both D1-MSNs and D2-MSNs. Reduction of the AHP amplitude at 440 pA current injection as compared with that at the rheobase in the D2-MSNs of the Sulf1 KO mice may contribute to the slow recovery from inactivation of voltage-dependent Na+ channels and the consequent low-frequency firing (Cloues and Sather, 2003; Duménieu et al., 2015). The AHP is generally mediated by multiple types of potassium channels. Previous studies have reported that small-conductance calcium–activated potassium channels (SK channels) regulate the AHP and excitability of MSNs in the NAc shell (Zhang et al., 2023). Furthermore, the strong depolarization lasting during step current injections is likely due to modification of voltage-gated outward–rectifying K+ channels with a slower time course. Actually, phosphorylation of voltage-gated potassium channels (KCNQ channels) inhibits KCNQ-mediated currents and increases excitability in NAc MSNs (Faruk et al., 2022; Tsuboi et al., 2022). In cardiomyocytes, voltage-gated potassium currents are modulated by one of the HSPGs, glypican 1 (Souza et al., 2022). Therefore, Sulf1-mediated changes in signal transduction pathways may affect the activities of potassium channels, thereby modulating the neuronal excitability of MSNs in the NAc shell. It is also possible that Sulf1 regulates MSN activity by affecting the activities of other ion channels as well. Future studies on ionic currents in the NAc MSNs of Sulf1 KO mice will be required to understand the mechanisms underlying the changes in excitability.
Interestingly, our data on the injected current-firing rate relationship indicate that Sulf1 disruption eliminates the differences in excitability between D1-MSNs and D2-MSNs (compare Fig. 2C and Fig. 4B). Given that Sulf1 KO does not alter the percentages of Drd1+/Drd2+ cells in the NAc (K.M. and M.M., unpublished data), it is likely that Sulf1 disruption changes the cellular properties that play critical roles in the regulation of neuronal excitability. Several recent studies in mice have reported that D1-MSNs and D2-MSNs in the NAc shell have distinct physiological roles in the responses to appetitive or aversive stimuli (Domingues et al., 2025), hedonic eating behaviors (Guillaumin et al., 2023), and itch signal processing (Liang et al., 2022). Therefore, Sulf1 gene disruption may lead to impairment of NAc-mediated behaviors.
The AMPA/NMDA ratio, a metric for synaptic strength and maturation, is generally used to examine changes in excitatory synaptic transmission (Kauer and Malenka, 2007). In the NAc, it is correlated with experience-dependent synaptic plasticity and changed by means of various signals including glutamate, dopamine, serotonin, opioids, and endocannabinoid (Turner et al., 2018). To explore any changes in the excitatory synapses of the NAc of Sulf1 KO mice, we examined glutamatergic EPSCs and calculated the AMPA/NMDA ratio. In Sulf1 KO mice, the AMPA/NMDA ratio was significantly higher in the D1-MSNs than that in the WT mice, suggesting that Sulf1 affects glutamatergic transmission into D1-MSNs. The AMPA/NMDA ratio was slightly higher in the D2-MSNs of the Sulf1 KO mice than in those of the WT mice, but the difference was not statistically significant. The increase in the AMPA/NMDA ratio may be related to the changes in cell-surface expression and subunit composition of AMPA receptors in MSNs (Turner et al., 2018). Because MSNs in the NAc receive glutamatergic innervation from the ventral hippocampus, amygdala, prefrontal cortex, and thalamus (Britt et al., 2012) and because they are distinctly modulated by dopamine and serotonin (Christoffel et al., 2021), pathway-specific analysis of synaptic transmission by means of optogenetic techniques will be necessary in future studies to reveal the roles of Sulf1 in the NAc circuit. Given that the changes in the AMPA/NMDA ratio are associated with synaptic plasticity in learning and behavioral adaptations (Turner et al., 2018), NAc-dependent behaviors to reward and aversion may be affected in Sulf1 KO mice.
HS plays important roles in neural development, axon guidance, synaptogenesis, and neuronal functions (Holt and Dickson, 2005; Condomitti and de Wit, 2018; Zhang et al., 2018; Kamimura and Maeda, 2021). In conditional Ext1 KO mice, in which HS is not synthesized in the excitatory neurons in the forebrain, autism-like deficits were observed as a result of attenuation of EPSCs in the amygdala, presumably owing to the reduction in the number of postsynaptic AMPA receptors (Irie et al., 2012). Acute enzymatic digestion of HS resulted in impairment of long-term potentiation in hippocampal slices and reduced excitability of CA1 pyramidal neurons (Minge et al., 2017). These data suggest that HS is essential for neurotransmission and plasticity. It was also reported that Sulf1 KO mice show a reduced hippocampal spine density and impaired long-term potentiation evoked by theta burst stimulation of Schaffer collaterals in hippocampal slices, while Sulf1 mRNA is not detected in the adult hippocampus, and thus developmental defects are discussed as causes of these abnormalities (Kalus et al., 2009). Sulf1/2 double KO mice have defects in the axon guidance of the corticospinal tract, but no brain defects are found in single Sulf1 KO mice as far as we have investigated. However, because it is impossible to exclude the possibility that undetected abnormalities are the cause of the changes in the excitability of the adult NAc MSNs, mice in which the Sulf1 gene is conditionally disrupted in the adult brain should be used in future studies.
Although we have provided compelling evidence for the involvement of Sulf1 in neuronal functions in the adult mammalian brain, the molecular and cellular mechanisms underlying the changes observed in Sulf1 KO mice remain to be clarified. As for the changes in synaptic transmission, SK channels regulate the AMPA/NMDA ratio as well as surface AMPA receptor expression in MSNs (Zhang et al., 2023). In addition, in a mouse model of Sanfilippo syndrome, in which HS accumulates as a result of the defects of an HS-degrading enzyme, HS from the cerebral cortex of mutant mice showed enhanced ability to increase the AMPA receptor GluA2 subunit on the cell surface (Dwyer et al., 2017). Furthermore, because diffusible secreted proteins, including Wnts, fibroblast growth factors, netrins, semaphorins, and sonic hedgehog, which are also well known axon guidance molecules (Yuzaki, 2018; Rennich et al., 2023), interact with HS, their signaling may be affected in Sulf1 KO mice. Therefore, Sulf1 disruption may alter AMPA receptor trafficking and synapse formation, thereby affecting the AMPA/NMDA ratio. It will be intriguing to examine the possible behavioral abnormalities that can be caused as a consequence of the changes in MSN excitability and glutamatergic transmission, for example, deficits in reward and aversive learning, in Sulf1 KO mice in future studies.
Footnotes
The authors declare no competing financial interests.
We are grateful to Drs. Toshiya Manabe, Yoshihiro Kubo, Taro Ishikawa, and Yukihiro Nakamura and to Flaminia Miyamasu for critical reading of the manuscript. This work was supported by KAKENHI grants 25293065 (to M.M.), 22K06488 (to T.M.), and 23K05981 (to E.S.) from MEXT and JSPS and by grants from the Takeda Science Foundation, Naito Foundation and SENSHIN Medical Research Foundation (to M.M.).
↵*K.M. and E.S. contributed equally to this work.
This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International license, which permits unrestricted use, distribution and reproduction in any medium provided that the original work is properly attributed.
References
In this issue
