Abstract
Genetic variation accounts for much of the risk for developing a
substance use disorder, but the underlying genetic factors and their
genetic effector mechanisms are mostly unknown. Inbred mouse strains
exhibit substantial and heritable differences in the extent of
voluntary cocaine self-administration. Computational genetic analysis
of cocaine self-administration data obtained from twenty-one inbred
strains identified Nav1, a member of the neuron navigator family that
regulates dendrite formation and axonal guidance, as a candidate gene.
To test this genetic hypothesis, we generated and characterized Nav1
knockout mice. Consistent with the genetic prediction, Nav1 knockout
mice exhibited increased voluntary cocaine intake and had increased
motivation for cocaine consumption. Immunohistochemistry,
electrophysiology, and transcriptomic studies were performed as a
starting point for investigating the mechanism for the Nav1 knockout
effect. Nav1 knockout mice had a reduced inhibitory synapse density in
their cortex, increased excitatory synaptic transmission in their
cortex and hippocampus, and increased excitatory neurons in a deep
cortical layer. Collectively, our results indicate that Nav1 regulates
the response to cocaine, and we identified Nav1 knockout induced
changes in the excitatory and inhibitory synaptic balance in the cortex
and hippocampus that could contribute to this effect.
Subject terms: Genome-wide association studies, Addiction
__________________________________________________________________
Mice lacking neuron navigator 1 (Nav1) exhibit increased voluntary
cocaine intake, potentially due to altered excitatory and inhibitory
synaptic balance in the cortex and hippocampus.
Introduction
While a substantial proportion of the risk for developing a SUD is
genetic^[54]1,[55]2, the genes and alleles that influence
susceptibility to a SUD remain largely unknown. Just as in the human
population, inbred mouse strains exhibit substantial and heritable
differences in their responses to commonly abused drugs, which has
enabled murine models for many human SUD behaviors to be
generated^[56]3. Inbred mouse strains have been analyzed to identify
genetic factors affecting addiction-related behaviors and their
responses to drugs with high abuse potential^[57]4–[58]8. While we do
not expect that murine studies will be likely to identify the same
genetic factors that are operative in humans, the neurobiological
mechanisms that underlie SUD risk are likely to converge in mice and
humans^[59]3. Hence, characterizing genetic factors that affect the
responses of inbred mouse strains to these drugs could (i) increase our
understanding of the neurobiological pathways that are impacted by
them; (ii) reveal how addiction-related behaviors are generated; and
(iii) could help to generate new approaches for preventing drug
addiction in humans^[60]9. Among the various rodent assays, the cocaine
self-administration assay (CSA) is considered a gold-standard assay for
studying cocaine-related behavior, neurobiology, and genetics in rodent
populations^[61]10–[62]12. Mice are fitted with a jugular catheter and
placed in an operant conditioning box where they actuate a lever to
trigger cocaine infusions, which enables the extent of CSA to be
measured. The rate of CSA reflects the reinforcing potential of
cocaine^[63]13, and inter-strain differences in the magnitude of CSA
reflect the propensity that a strain could misuse
cocaine^[64]14–[65]16.
We analyzed a murine genetic model for CSA and identified a candidate
gene (Nav1) that could regulate cocaine responses. Analysis of Nav1 KO
mice confirmed that Nav1 had a strong effect on the level of voluntary
cocaine consumption; and that it also impacted food reinforcement,
which is another addiction-related trait. Some potential insight into
the mechanism for this was provided by finding that the Nav1 KO altered
the excitatory/inhibitory synaptic balance in hippocampal and cortical
brain regions.
Results
Identification of Nav1 as a candidate gene affecting CSA
Voluntary CSA was examined over a 10-day period (0.5 mg/kg/infusion) in
adult male mice of 21 inbred mouse strains. During the last 3 days of
testing, the strains exhibited very different levels of CSA (range 0–65
infusions), and the heritability for the inter-strain differences in
this measurement was 0.68 (Fig. [66]1a). HBCGM analysis of the CSA data
for the 21 inbred strains revealed that the allelic patterns within
multiple genes that were co-localized within a region on chromosome 1
(134–137 MB) were most strongly associated with the inter-strain
differences in CSA (Fig. [67]1). Among the genes within the chromosome
1 region (134–137 MB) with allelic patterns that correlated with
inter-strain differences in CSA, only two (Nav1, Etnk2) had SNP alleles
that altered the predicted amino acid sequence. However, Nav1 was the
only one of these genes with a high level of mRNA expression in key
brain regions, and its allelic pattern was not associated with
population structure^[68]17 (Supplementary Table [69]2). There were
2163 SNPs within or near (±10 kB) the Nav1 gene, which included 20
synonymous SNPs, and 3 cSNPs (D198E, P1366L, A911V) that altered the
Nav1 amino acid sequence (Supplementary Table [70]3). Also, two SNPs
(D198E, A911V) were within predicted sumoylation (193–202 -EAAVSDDGKS)
and phosphorylation (910–917 -TAPSEEDT) motifs. While we certainly do
not know the specific Nav1 alleles that contribute to the different
levels of CSA exhibited by the inbred strains, it is noteworthy that
the level of CSA was correlated with the Nav1 D198E allele of an inbred
strain (p = 0.007) (Fig. [71]1d). Several other factors suggested that
genetic variation within Nav1 could affect CSA. (i) Nav1 is expressed
predominantly in the nervous system^[72]18. (ii) Nav1 plays a role in
neuronal development and in the directional migration of neurons^[73]19
by regulating neurite outgrowth through effects on cytoskeletal
remodeling^[74]20, which are processes that have been associated with
addiction-induced changes in brain^[75]21. (iii) Multiple Nav1 mRNA
isoforms are produced by alternative splicing^[76]18, which provides
another mechanism by which allelic differences could impact a
phenotype. (iv) A prior HBCGM analysis led to the discovery that
alleles within another axonal guidance protein (Netrin 1) affected
multiple maladaptive responses to opioids^[77]7. (v) Analysis of
striatal Nav1 mRNA expression levels in a BXD recombinant inbred strain
panel indicated that cis acting alleles regulate its expression
(Supplementary Fig. [78]1). (vi) When the Nav1 mRNA expression QTL was
evaluated for correlation with the entire database of phenotypes
measured in this panel of recombinant inbred strains, the highest
correlation was an inverse one with striatal dopamine receptor D2
(Drd2) mRNA levels. This correlation is specific in that striatal Drd1
mRNA expression is positively associated with Nav1 mRNA levels
(Supplementary Fig. [79]1), and this is noteworthy since low Drd2
expression and function is a biomarker for addiction vulnerability in
animals and humans^[80]22–[81]26. All of these features suggest that
Nav1 allelic variation could impact Drd2 dopamine-dependent signaling
within the striatum, and thus, responses to potentially addictive
drugs.
Fig. 1. Identification of Nav1 as a candidate gene effecting CSA.
[82]Fig. 1
[83]Open in a new tab
a The average number of cocaine infusions earned during the last 3 days
of a 10-day CSA session was measured for each of 21 inbred strains.
Each bar shows the strain mean ± SEM (n = 1–6 mice per strain). Scatter
points represent infusions by individual mice. b A diagram of the CSA
assay. c The HBCGM analysis output is shown. The top panel shows the
average number of cocaine infusions for each strain, and the bottom
panel shows the top 7 genes identified by HBCGM analysis of this data.
The gene symbol, genetic effect size, chromosomal position, and the p
values for the genetic association are shown. If the box surrounding
the gene symbol has an orange color, it indicates that the haplotype
block has a SNP allele causes a significant change in the protein
sequence. In the haplotype box, the haplotypic pattern is shown as
colored rectangles that are arranged in the same order as the input
data shown above; strains with the same-colored rectangle have the same
haplotype. d The mean ± SEM for the number of cocaine infusions
(y-axis) self-administered by strains with different D198E cSNP alleles
in Nav1 are shown. There is a statistically significant increase
(p = 0.007) in number of cocaine infusions self-administered by the 12
strains (C57BL/6, CE/J, LP/J, Balbc/J, C3H/HeJ, CBA/J/ NOD/J, RIIIS,
SMJ, FVB, MaMy, BTBR) with the C57BL/6J allele (D198) (average of
32.5 ± 5.2 infusions) relative to the 8 strains (NZW, AKR, KK, SJL,
DBA2, DBA1, MRL, 129Sv1) with the 129Sv1 allele (E198) (average of
11.3 ± 2.8 infusions).
Generation and characterization of a Nav1 knockout (KO) mouse
Due to its large size, the presence of many SNPs within Nav1, and the
multiple mechanisms by which Nav1 alleles could impact a genetic trait,
the genetic hypothesis was tested by producing a homozygous Nav1 KO
mouse on a strain background (C57BL/6J) that exhibited a moderate level
of CSA^[84]27 (Supplementary Fig. [85]2). Single molecule FISH analysis
indicated that full length Nav1 mRNA is expressed in C57BL/6J (but not
in Nav1 KO) brain tissue, while the Nav1 KO transcript is exclusively
expressed in Nav1 KO mice (Supplementary Fig. [86]3, Supplementary
Table [87]5). Proteomic analysis confirmed that Nav1 protein was absent
in Nav1 KO brain tissue (see Methods). MRI brain scans revealed that
Nav1 KO mice did not have gross neuroanatomical abnormalities. Nav1 KO
mice had a slightly smaller overall brain volume, which was consistent
with their smaller body size (vs. C57BL/6J mice). After normalization
of hippocampal volume to overall brain size, there was no significant
difference in their relative hippocampal volume (vs. C57BL/6J mice)
(Supplementary Fig. [88]4).
The Nav1 KO effect on voluntary CSA
A between-subjects dose-response design was used to compare the level
of CSA between wildtype, heterozygous Nav1 KO (HET) and homozygous Nav1
KO mice. The data were analyzed by mixed ANOVA with session as a
repeated measure and genotype/dose/sex as between subjects factors.
Analysis of this data indicated that there was a clear effect of
genotype [F[2,113) = 8.5, p < 0.001], and sex [F[1,113) = 7.2,
p = 0.008] on the number of cocaine infusions earned. A
session-genotype interaction [F[8.6,483.3) = 2.8, p = 0.003], along
with a main effect of session [F[4.3, 483.3) = 4.4, p = 0.001] were
also found. Post hoc, pairwise comparisons indicated that the Nav1 KO
mice earned more infusions than HET Nav1 KO mice (p < 0.05) following
session 2 and more infusions than wildtype mice following session 3
(except for session 7 (p = 0.052)). These results indicate that Nav1 KO
mice self-administered more cocaine infusions and this effect was not
impacted by dose. Furthermore, their increased level of consumption
appeared during the early period of cocaine IVSA acquisition and was
maintained throughout the 10 days of testing (Fig. [89]2a, b). The sex
effect indicated that males consumed more cocaine than females across
all doses and genotypes tested (Supplementary Fig. [90]5).
Fig. 2. CSA dose responses for wildtype, HET, and homozygous Nav1 KO mice.
[91]Fig. 2
[92]Open in a new tab
All graphs include the mean ± SEM. a Infusions earned in 10 sessions of
CSA across genotype and across/within dose (n = 13–16 mice per
genotype/dose). Nav1 KO mice self-administered more cocaine than
wildtype and HET mice from session 3 on (upper graph represents
genotype means across dose; *p < 0.05 between KO and wildtype/HET for
post hoc comparisons across dose and within session). Lower graphs
visualize responding within each dose (however no genotype-by-dose
interaction was detected). b A visualization of the dose response curve
for the number of cocaine infusions during last 3 sessions, which are
averaged for each dose for mice of the indicated genotype. Nav1 KO mice
displayed an upward shift in the dose response curve. c Visualization
of the dose response curve for cocaine intake (mg/kg) during the last 3
sessions for each cocaine dose. Similar to the number of infusions,
Nav1 KO mice displayed an upward shift in the dose-response curve for
cocaine intake. d Active lever preference (i.e., the % of correct lever
presses) during the last 3 sessions are shown for each of the 3 doses
of cocaine. There was no difference in active lever presses among the
groups of mice with different genotypes. e CSA was measured at two
different doses (0.1 and 1.0 mg/kg) in Wt, Het and Nav1 KO mice
(n = 13–16 mice per group) using a progressive ratio test, where the
number of active lever presses required for receiving the next cocaine
infusion is progressively doubled from that required for the previous
infusion. The number of lever presses required to earn the last cocaine
infusion (i.e., the last ratio) provides an assessment of their
motivation for cocaine consumption. Nav1 KO mice achieved a
significantly greater last ratio than wildtype or Het mice (*p < 0.05;
main effect of genotype).
Cocaine intake was analyzed by ANOVA with session as a repeated measure
and genotype/dose/sex as between subject factors. This analysis
indicated that Nav1 KO mice self-administered a larger amount of
cocaine than HET and wildtype mice at all doses (session-genotype
interaction [F(7.7,434.5) = 3.0, p = 0.003], and that there was a main
effect of session [F(3.8,434.5) = 7.5, p < 0.001], and genotype
[F(2,113) = 9.5, p < 0.001]). Nav1 KO mice self-administered more
cocaine than HET mice after session 2 and more than wild-type mice
after session 3 (p < 0.05). Mice of all three genotypes
self-administered more cocaine as the dose increased (i.e. there was a
main effect of dose [F(2,113) = 10.3, p < 0.001]) (Fig. [93]2c). Also,
a genotype-sex interaction [F(2,113) = 3.5, p = 0.034] suggested that
the sex differences were dependent on genotype. A post-hoc analysis
indicated that there were significant sex differences (males > females)
in wildtype (p < 0.001) but not in Nav1 KO (p = 0.912) or HET
(p = 0.218) mice. Similarly, when the preference for the active lever
was assessed during the last 3 days of CSA, a genotype-sex interaction
was detected [F(2,99) = 3.3, p = 0.041], and wildtype female mice had a
lower preference than males (p = 0.017) (Supplementary Fig. [94]5). We
found no significant differences in lever preference between the
genotypes (Fig. [95]2d).
We then assessed their motivation for self-administering cocaine using
a progressive ratio schedule of reinforcement (0.1 and 1.0 mg/kg
doses), where the number of active lever presses required for receiving
the next cocaine infusion was progressively doubled from the response
requirement of the previous infusion. Analysis of the last ratio of
lever presses required to earn a cocaine infusion indicated that there
was a genotype-dose-sex [F(2,73) = 3.8, p = 0.029] and genotype-sex
[F(2,73) = 8.3, p = 0.001] interaction, and a main effect of genotype
[F(2,73) = 3.9, p = 0.025]. Analysis within sex revealed that there was
a genotype-dose interaction in males [F(2,38) = 4.7, p = 0.015] but not
in females [F(2,35) = 1.3, p = 0.292]. Males of the wildtype genotype
earned a higher last ratio under the 1 mg/kg dose relative to the
0.1 mg/kg dose (p = 0.043) (Supplementary Fig. [96]5). Overall, these
data indicate that Nav1 KO mice maintained their greater level of
cocaine intake under a progressive ratio schedule (Fig. [97]2e). In
contrast, the acute locomotor responses of Nav1 KO mice after
experimentally administered cocaine were not different from those of
C57BL/6J mice (Supplementary Fig. [98]6).
The Nav1 KO also impacts food self-administration (FSA)
To determine whether the Nav1 KO impacted self-administration of a
non-drug reinforcer, mice were tested using a FSA procedure that was
similar to that of CSA testing except that a chocolate solution was
delivered as the reinforcer. Analysis of the number of food reinforcers
earned indicated that there was a genotype-session interaction
[F(5.1,66.3) = 2.5, p = 0.036] and a main effect of genotype
[F(2,26) = 6.5, p = 0.005] and session [F(2.6,234) = 26.2, p < 0.001]
(Fig. [99]3a). The pairwise comparisons, which examined the effect of
different genotypes on the FSA results within each session, revealed
that Nav1 KO mice earned a greater number of food reinforcers than HET
(from session 3 on) or wildtype (from session 4 on except for session
8) (p = 0.059) mice (Fig. [100]3a). Assessment of active lever
preference over the last 3 sessions did not indicate any significant
genotypic effect (Fig. [101]3a). We observed that KO mice weighed less
at baseline (bodyweight mean ± SD, KO male 26.6 ± 1.2, female
19.3 ± 1.1; HET male 27.4 ± 1.6, female 23.0 ± 0.8; wildtype male
28.5 ± 0.3, female 22.9 ± 1.2). Nav1 KO mice also lost more weight than
HET or wildtype mice during the period of food restriction, which may
be due to their baseline increase in locomotor activity (percent of
baseline bodyweight after food restriction (mean ± SD), KO 86% ± 6%;
HET 94% ± 2%; wildtype 92% ± 2%). Therefore, the FSA test was repeated
without food restriction. These results confirmed that Nav1 KO mice
earned more food reinforcers: there was a session-genotype interaction
[F(5.5,35.6) = 3.0, p = 0.02], and there was a main effect of genotype
[F(2,13) = 4.0, p = 0.045]) on the FSA results. These results indicate
that Nav1 KO effect on FSA was not due to food-deprivation. After 10
sessions on a fixed ratio1 (FR1) schedule, the mice underwent a
progressive ratio test. Analysis of the last ratio achieved also
indicated that there was a main effect of genotype [F(2,27) = 8.1,
p = 0.002]. Post hoc, pairwise comparisons indicated that Nav1 KO mice
achieved a greater last ratio than wildtype (p = 0.026) and Het
(p = 0.001) mice (Fig. [102]3c). Taken together, the FSA results
indicate that the Nav1 KO also altered food reinforcement.
Fig. 3. Food self-administration (FSA) by wildtype, HET, and Nav1 KO mice
(n = 11 mice per group) was measured in 10 daily sessions.
Fig. 3
[103]Open in a new tab
All graphs include the mean ± SEM, and the response of an individual
mouse is shown as a dot. a Nav1 KO mice had a continuously higher level
of FSA than wildtype or HET mice beginning from session 4 (*p < 0.05).
b The percentage of active lever presses (Active Lever Preference) was
measured during the last 3 sessions. No differences in active lever
preference was detected between mice with the three different
genotypes. c FSA was measured in male and female Wt, Het, and Nav1 KO
mice (n = 5–6 mice of each sex per group) using a progressive ratio
test where the number of active lever presses required for receiving
the next food reinforcer is progressively doubled from that required
for the previous one. The number of lever presses required to earn the
last food reinforcer (i.e., the last ratio) provides an assessment of
their motivation for food consumption. Male and female Nav1 KO mice
achieved a significantly greater last ratio than Het or Wt mice
(p < 0.5).
Nav1 KO mice have alterations in learning, memory, and in other behaviors
Since it was postulated that addictive drugs coopt the reward pathways
used for learning and memory^[104]28, we investigated whether learning
and memory were impacted by the Nav1 KO. To do this, we first assessed
whether the ability to recognize a novel object was altered in Nav1 KO
mice. While C57BL/6J mice spent significantly more time exploring a
novel object relative to a familiar one, homozygous and HET Nav1 KO
mice spent a similar amount of time exploring the two objects
(Supplementary Fig. [105]7a) (novelty: p = 0.0008; genotype × novelty:
p = 0.45). There were significant differences in the time that C57BL/6J
mice spent with novel vs familiar objects (p = 0.045), but this
difference was not manifested by HET (p = 1.0) or homozygous Nav1 KO
(p-value = 0.87) mice. When spatial learning and memory capabilities
were assessed using the Barnes maze test, Nav1 KO mice had a
significantly reduced ability to correctly recognize the escape hole
relative to HET (p < 0.01) or C57BL/6J (p < 0.0001) mice (Supplementary
Fig. [106]7b). There were no significant differences in the number of
primary errors made (p = 0.36) or in the total distance traveled
(p = 0.16) between the three groups of mice. We next investigated
whether the Nav1 KO impacted anxiety or locomotor activity. Nav1 KO
mice exhibited altered exploratory behavior in the elevated plus maze;
they spent more time in the open arms relative to HET (p = 0.005) or
C57BL/6J mice (p = 0.006) (Supplementary Fig. [107]8a). Open field
testing revealed that Nav1 KO mice had a significant increase in the
total distance traveled relative to HET mice (p = 0.01); and spent a
decreased percentage of time in the open field relative to C57BL/6J
mice (p = 0.04) (Supplementary Fig. [108]8b). However, Nav1 KO mice had
normal motor coordination in the rotarod test (Supplementary
Fig. [109]8c). These results indicate that Nav1 plays an important role
in learning and memory capabilities, and that it also effects the level
of anxiety and exploratory behavior.
Increased excitatory hippocampal synaptic transmission in Nav1 KO mice
Since Nav1 is involved in neuronal development and
migration^[110]19,[111]20, we quantitatively examined the impact of the
Nav1 KO on synapse formation. Nav1 KO mice had a significant increase
in the number of excitatory synapses in the dentate gyrus of the
hippocampus (vs C57BL/6J mice, p = 0.03), but inhibitory synapse
density was not changed (p > 0.99) (Supplementary Fig. [112]9a). In
vitro electrophysiological recordings from granule cells in the dentate
gyrus of the hippocampus showed a significant increase in the frequency
(but not in the amplitude) of miniature excitatory postsynaptic
currents (mEPSCs) in Nav1 KO vs C57BL/6J mice (Supplementary
Fig. [113]9a), which indicates that there is increased excitatory
synaptic transmission in the dentate gyrus of Nav1 KO mice. In
contrast, there was no significant difference in the frequency or
amplitude of the miniature inhibitory postsynaptic currents (mIPSCs)
measured in Nav1 KO vs C57BL/6J mice. There was also a significant
decrease in the number of inhibitory synapses formed in the prefrontal
cortex (PFC) of Nav1 KO mice (vs. C57BL/6J, p = 0.008), while
excitatory synapse density was not altered (Supplementary
Fig. [114]10). Since the nucleus accumbens (NAc) is a critical center
for responses to addictive drugs^[115]29, we compared the cells in the
NAc of Nav1 KO and C57BL/6J mice by immunohistochemical staining. There
was a similar density of neuronal cells in the NAc of Nav1 KO and
C57BL/6J mice, as shown by the equal number of neuronal nuclei marker^+
and neurofilament H^+ cells. However, glial fibrillary acidic protein
(GFAP) staining indicated that there was a slight trend toward an
increase in the number of glial cells in the NAc of Nav1 KO mice
(Supplementary Fig. [116]10). Thus, Nav1 KO mice have an increase in
hippocampal excitatory synapses and synaptic transmission, and a
decrease in the number of inhibitory synapses formed in the PFC.
Nav1 KO-induced transcriptomic changes
The PFC regulates higher cognitive functions (i.e., learning, memory,
and decision making) and forms connections with multiple other brain
regions^[117]30,[118]31, and we identified synaptic changes in the PFC
of Nav1 KO mice. Therefore, we examined Nav1 KO-induced transcriptomic
changes in the PFC. snRNA-Seq was performed on 28,686 high quality PFC
cells obtained from Nav1 KO mice (14,056 cells; 68,637 reads per cell)
and from age-matched, isogenic C57BL/6J mice (14,630 cells, 56,980
reads per cell with a mean of 1438 genes per cell). The PFC cells were
separated into 14 clusters, which based upon their pattern of marker
mRNA expression^[119]32, were derived from 4 lineages: neurons,
astrocytes, oligodendrocytes, and microglia (Fig. [120]4, Supplementary
Figs. [121]11–[122]13). The eight neuronal clusters were separated into
six excitatory (0–2, 4, 5, and 10) and two (3, 8) inhibitory neuronal
clusters. Consistent with prior findings^[123]32, excitatory neurons
were 5 to 7-fold more abundant than inhibitory neurons (52–75% vs 7–14%
of the total). The cell clusters and their distribution in our C57BL/6J
PFC data were very similar to that of a prior analysis^[124]32
(Supplementary Fig. [125]14). Clusters 10 and 5 had the highest levels
of Nav1 mRNA (Supplementary Fig. [126]11). Cluster 10 was of particular
interest since its abundance was most increased (14-fold) in Nav1 KO
(vs C57BL/6J) mice. Cluster 10 uniquely expressed an excitatory
neuronal marker (Tshz2) that is characteristic of cortical layer 5 (L5)
cells (Supplementary Fig. [127]11), which form projections to
extra-cortical brain regions and are in the cortical layer whose
transcriptome was most altered during cocaine withdrawal^[128]32.
Pathway enrichment analysis of the differentially expressed genes
(DEGs) in cluster 10, revealed that many of the enriched pathways were
associated with neuronal guidance, synapse or signaling functions; and
7 DEGs (Gria3, Grin2b, Grin2a, Camk2a, Ppp3ca, Gria4, Prkcb) were
associated with an amphetamine addiction pathway (p = 0.000025)
(Supplementary Fig. [129]15). Clusters 5 and 10 also had the highest
levels of expression of two mRNAs (Rab3a^[130]33, Cck^[131]34), whose
expression levels were altered during cocaine withdrawal^[132]32
(Supplementary Fig. [133]16). Cluster 5 uniquely expressed a
transcription factor (Etv1, aka Er8) that is required for generating
habitual behaviors^[134]35. The inhibitory neurons in cluster 8 were of
interest because their abundance was 3-fold decreased in the PFC of
Nav1 KO mice, and they uniquely expressed Drd1 mRNA (Supplementary
Fig. [135]12). In contrast, Drd2 mRNA was not expressed at detectable
levels in any cluster. Of importance, the mRNA levels measured in PFC
obtained from C57BL/6J and Nav1 KO mice are not necessarily related to
the Drd1-2 mRNA levels in measured in the striatum of the recombinant
inbred strains. In summary, the transcriptomic analysis indicates that
the Nav1 KO affected neurons located in a deep cortical layer, and some
of these cells had transcriptomic changes that were similar to those
associated with cocaine withdrawal.
Fig. 4. snRNA-Seq analysis of Nav1 KO and isogenic C57BL/6J PFC tissue.
[136]Fig. 4
[137]Open in a new tab
UMAP plots show the data for the C57BL/6J and Nav1 KO cells. Each dot
represents an individual cell. The table indicates the dot color for
each cluster, the percentage of cells in each cluster, and the ratio of
the cell percentages (Nav1 KO/C57BL/6J) in C57BL/6J and Nav1 KO PFC.
This dataset analyzed the expression of 21,892 genes in 28,686 nuclei.
Based upon total transcriptomic differences, the PFC cells were
separated into 14 different clusters. Based upon canonical marker
expression, the clusters are derived from the five lineages: astrocytes
(6, 11), microglia (12), oligodendrocytes (7); and inhibitory (3, 8)
and excitatory (0, 1, 2, 4, 5, 10) neurons. The lineages for two
clusters (9, 11) were undefined (UD) because they expressed markers
from different cell lineages. Of particular interest, the number of
cells in an inhibitory neuron cluster (cluster 8) is three-fold
decreased in Nav1 KO mice, and there is a 14-fold increase in the
number of cells in an excitatory neuron cluster (cluster 10) in Nav1 KO
mice.
Discussion
Our analysis of a murine genetic model identified Nav1 as a candidate
gene that influences voluntary cocaine consumption, and this genetic
finding was confirmed by the increased level of CSA exhibited by Nav1
KO mice across a broad range of cocaine doses. The effect of the Nav1
KO on voluntary cocaine consumption occurred early; it was maintained
throughout a 10-day period of CSA testing. Moreover, subsequent testing
revealed that Nav1 KO mice exhibited an increased motivation for
consuming cocaine^[138]36. Since Nav1 KO and wildtype mice did not
exhibit differences in lever preference or in cocaine-induced acute
locomotor responses; neither indiscriminate lever pressing nor
differential locomotor effects could be responsible for their increased
cocaine-intake. Although HET mice did not significantly differ from
wildtype mice, a potential trend towards lower cocaine intake in HET
mice is present in the data, suggesting that the heterozygous allele
state alters cocaine intake in the opposite direction of the homozygous
knockout state. Overall, these results indicate that the reinforcer
efficacy of cocaine is increased in Nav1 KO mice. The Nav1 KO had a
similar effect on food reinforcement, which suggests that the effect of
the Nav1 KO could be generalized across drug and non-drug reinforcers.
It has been postulated that a common neural circuitry may underlie food
and drug addictions^[139]37, and genetic variation in Cyfip1 had
effects on both cocaine and food reward^[140]5,[141]38. Nav1 KO mice
also exhibited reduced anxiety-like behavior. An inverse relationship
between anxiety-like behavior and CSA^[142]39 and ethanol
drinking^[143]40 was previously observed in rats. While the Nav1
results also indicate that the neural circuits mediating drug reward,
food reward and anxiety may have some overlap, additional research is
necessary to further explore this possibility.
After finding that the Nav1 KO affected CSA and FSA, we also found that
Nav1 KO mice also exhibited reduced spatial learning and memory.
Therefore, we performed a series of studies to identify potential
mechanism(s) for these Nav1 KO-effects. (i) Brain MRI demonstrated that
the Nav1 KO did not cause gross structural changes. (ii) Since Nav1 is
involved in neuronal development and migration, we examined the Nav1 KO
effect on neurons and synapses. Nav1 KO mice had a significant increase
in excitatory synapse density in the dentate gyrus and a significant
decrease in inhibitory synapse density in the PFC. Consistent with
these results, EPS studies revealed that cells in the dentate gyrus of
Nav1 KO mice had an increased frequency of mEPSCs. (iii) The PFC was
analyzed by scRNA-Seq because it regulates higher cognitive functions
that were impacted by the Nav1 KO (i.e., learning and memory) and
because of the observed change in excitatory and inhibitory synapse
density in the Nav1 KO PFC. The Nav1 KO-induced alterations in the
balance between excitatory and inhibitory synapses in the cortex and
hippocampus provides a potential mechanism by which Nav1 could impact
learning, memory, and possibly the response to an addictive drug. This
mechanism is consistent with the known role of Nav1 in regulating
neuronal development, directional migration^[144]19, and neurite
outgrowth^[145]20. While various cortical and subcortical brain regions
affect the response to addictive drugs, the PFC integrates
reward-seeking and decision-making, through the inhibitory control it
exerts^[146]14,[147]41 through the dense reciprocal connections it
forms with virtually all neuromodulatory centers within cortical and
subcortical regions^[148]42. For example, PFC glutamatergic neurons
connect to the NAc core, which can provide a top-down control mechanism
for preventing food addiction behaviors^[149]43. Disruptions to
hippocampal synaptic balance could cause the learning/memory deficits
observed in Nav1 KO mice, since these tests (particularly the Barnes
maze) are hippocampal-dependent^[150]44,[151]45. Since the hippocampus
also has a role in cocaine responses^[152]46, Nav1-induced disruptions
of hippocampal function could also affect CSA.
In summary, we demonstrated that Nav1 plays a role in in
addiction-related traits that include voluntary cocaine consumption and
food reinforcement. Moreover, Nav1 also altered anxiety-like behavior,
learning/memory, and the excitatory/inhibitory synaptic balance in
hippocampal and PFC brain regions. However, further characterization of
neural adaptations occurring in Nav1 KO mice could enable the role of
Nav1 and the pathways mediating addiction-associated behaviors to be
more fully characterized. Moreover, generation and characterization of
mice with brain a region-specific conditional Nav1 knockout or of
knockin mice with alterations in specific Nav1 alleles are also needed
to improve our understanding of the role that Nav1 plays in SUDs.
Methods
Mouse strains
All mouse procedures were approved by the Institutional Animal Care and
Use Committees at Binghamton or Stanford University; and were conducted
in accordance with the National Institute of Health Guide for Care and
Use of Laboratory Animals, Eighth Edition. All mice were originally
obtained from Jackson Laboratories, and the results are reported
according to the ARRIVE guidelines^[153]47. The strains utilized for
genetic mapping were bred for 7 generations or less prior to testing.
Nav1 KO mice (on C57BL/6J background) were maintained at Binghamton and
at the Stanford University School of Medicine. Mice were housed in
their home cage and maintained on ad libitum mouse chow (5L0D, Purina
Lab Diet) and water. Mice were individually housed in polycarbonate
cages (30 × 8 cm) with wood-chip bedding (SANI-CHIPS), a paper nestlet
and a red polycarbonate hut.
Cocaine self-administration (CSA)
Chronic indwelling jugular catheters were inserted and managed as
described^[154]27. Twenty-one strains (n = 1–6 mice per strain, male
mice, age range: 12–18 weeks) were tested for CSA acquisition in 10
consecutive daily sessions (Fixed-ratio-1 schedule of reinforcement, 0
or 0.5 mg/kg body weight of cocaine per infusion) that ran until 65
infusions were earned or 2 h passed, whichever came first. Cocaine
hydrochloride (Sigma Aldrich; St Louis MO) was dissolved in sterile
saline at a concentration of 0.17, 0.84, or 1.68 mg/mL to produce a
freebase dose of 0.1, 0.5, or 1.0 mg/kg/infusion (infusion volume was
0.67 mL/kg/infusion). Testing occurred at the same time each day,
during the light phase of a 12/12 h cycle. The animals were tested in
Med Associates mouse self-administration chambers
(55.69 × 38.1 × 35.56 cm, MED-307W-CT-D1, Med Associates, VT) that were
fitted with 2 retractable ultrasensitive levers and that were housed
within sound-attenuating cubicles. Assignment of the active infusion
lever (right or left side of the box) was counterbalanced across
strains/sex. Assignment of testing chamber minimized testing multiple
mice from a given strain in the same chamber. Test sessions began with
the activation of the white noise and the illumination of 5 stimulus
lights on the back wall of the chamber. No priming infusion(s) were
delivered. When a subject actuated the active lever, an infusion was
delivered, the house light flashed, and the aperture lights turned off
for 20 s. During this time-out period, contacts on the active lever
were recorded but had no programmed consequence. Actuation of the
inactive lever had no programmed consequence. The number of infusions
earned, cocaine intake (mg/kg) (number of infusions X dose) and active
lever preference (active lever presses/total lever presses) were key
variables of interest. Active lever preference is incalculable if the
mouse fails to press either lever, and consequently analysis of this
variable by session leads to excessive missing data points. Therefore,
the last 3 CSA sessions were collapsed by calculating preference across
these days, within individual mice.
Haplotype based computational genetic mapping (HBCGM)
The SNP database was generated by analysis of the genomic sequences of
47 classical (C57BL/6J, 129P2, 129S1, 129S5, AKR, A_J, B10, BPL, BPN,
BTBR, BUB, BALB, C3H, C57BL10J, C57BL6NJ, C57BRcd, C57LJ, C58, CBA,
CEJ, DBA, DBA1J, FVB, ILNJ, KK, LGJ, LPJ, MAMy, NOD, NON, NOR, NUJ,
NZB, NZO, NZW, PJ, PLJ, RFJ, RHJ, RIIIS, SEA, SJL, SMJ, ST, SWR,
TALLYHO, RBF, MRL) and 6 wild-derived (CAST, MOLF, PWD, PWK, SPRET,
WSB) inbred strains as described in^[155]48–[156]50. HBCGM was
performed as described^[157]51 using modifications described
in^[158]52. The chromosomal location and potential codon-changes for a
SNP are annotated using predictive gene models from Ensembl version 65.
The methods for calculation of the genetic effect size (η^2) and for
other mapping methods are provided elsewhere^[159]53,[160]54.
Generation and characterization of Nav1 KO mice at the Stanford University
School of Medicine
C57BL/6J female mice were super-ovulated by intraperitoneal injection
of pregnant mare’s serum gonadotropin and human chorionic gonadotropin.
These mice were then paired with C57BL/6J males to generate fertilized
embryos, and pronucleus (PN) stage embryos were collected. S. pyogenes
Cas9 protein (IDT) and guide RNAs (target 1 crRNA:
CAAACCTAGCCGGATTCCTC, target 2 crRNA: GCACGGTAACCACAAGCTCG, in the form
of crRNA:tracrRNA duplex, from IDT) were then electroporated into PN
embryos using a NEPA21 electroporation system (Supplementary
Fig. [161]2) with a CUY505P5 electrode. The guide RNAs were designed to
delete a 178 bp region at the end of exon 1 and to also introduce an
early stop codon in exon 1. This region was deleted because exon 1 is
expressed in all 7 known isoforms of Nav1 mRNAs. Healthy embryos were
transferred into the oviducts of pseudo-pregnant recipient females.
Genomic DNA from the pups were screened by PCR amplification using the
strategy shown in Supplementary Fig. [162]2. Mice with genomic DNA that
generated diagnostic amplicons were subsequently sequenced to
characterize the deleted region. To reduce the possibility of off
target editing, both selected guide RNAs had very high specificities
for the target sites. The target 1 crRNA has an MIT specificity score
of 92 and a cutting frequency determination (CFD) specificity score of
96, and target 2 crRNA has an MIT specificity score of 90 and a CFD
specificity score of 95^[163]55,[164]56. Both guide RNA evaluation
programs generate scores that range from 0 to 100, with 100 being the
most specific. It has been reported that the total number of mismatched
base-pairs is a key determinant of Cas9 cleavage efficiency. Two
mismatches, particularly those occurring in a PAM-proximal region,
considerably reduce Cas9 activity, irrespective of whether they are
concatenated or interspaced, and this effect is further magnified for
three and four mismatches. Three or more interspaced mismatches
eliminates detectable Cas9 cleavage at the vast majority of
loci^[165]57. Of importance, no off-target sites for either crRNA had
0, 1, or 2 base mismatches. The target 1 crRNA has 51 off-target sites
with 4 base mismatches and the target 2 crRNA has 41 sites with 4
mismatches; and none are in an exon on chromosome 1. There was one
off-target site with 3 base mismatches for both crRNAs; but both sites
are in intergenic sequences that were not present on chromosome 1
(where the Nav1 is located), and at least one mismatch is in the
PAM-proximal region. To further minimize the chance of an off-target
effect of CRISPR engineering, a Nav1 KO mouse was backcrossed to
C57BL/6J mice for two generations. The resulting Nav1 het KO mice were
intercrossed to generate homozygous Nav1 KO mice. Since Nav1 KO females
are poor mothers, homozygous Nav1 KO mice were generated by breeding
homozygous Nav1 KO males with heterozygous Nav1 females for colony
maintenance. To generate mice used in the CSA and FSA experiments,
heterozygous Nav1 mice were intercrossed to generate WT, heterozygous,
or homozygous mutant mice. Single molecule fluorescence in situ
hybridization (smFISH) was performed according to^[166]58. In brief,
frozen brain tissue sections (20 μm) were pre-treated with 0.01% pepsin
in 0.1 M HCl for 2 min at room temperature followed by washing in 0.05%
Tween 20 in 1 x diethyl pyrocarbonate-treated (DEPC)-PBS. mRNA was
reverse transcribed to cDNA in a room temperature buffer containing
0.5 mM dNTP, 0.2 μg/μl BSA, 1 μM cDNA primer, 1 U/μl RNaseIn
(Clonetech, 2313B) and 20 U/μl RT (Maxima, Thermo Scientific™, EP0752)
for 3 h at 50 °C in a securely sealed chamber. After three brief washes
in 0.05% Tween 20 in 1 x DEPC-PBS, the sections were post-fixed in 4%
paraformaldehyde for 30 min at room temperature and washed three times
in phosphate-buffered saline with Tween 20 (PBST). Hybridization and
ligation with T4 DNA Ligase were performed in T4 ligase buffer with
0.2 μg/μl BSA, 100 nM padlock probe and 0.1 U/μl T4 DNA Ligase (New
England Biolabs) for 30–45 min in 37 °C. This was followed by washing
with 2× SSC with 0.05% Tween-20 at 37 °C for 5 min, and then rinsing in
PBST. Rolling circle amplification (RCA) was performed with 1 U/μl Φ29
DNA polymerase (New England Biolabs) using the reaction buffer supplied
by the manufacturer with 250 μM dNTPs, 0.2 μg/μl BSA and 10% glycerol.
The incubation was carried out for 60–150 min at 30 °C, which was
followed by a washing in TBST. Stranded RCA products (RCP) were
hybridized in 250 nM of Cy5 and FAM fluorescence-labeled
oligonucleotide probes in a solution of 2× SSC, 20% formamide for
30 min at 37 °C. Slides were then washed in TBST. Dried slides were
mounted with VECTASHIELD® PLUS Antifade Mounting Medium (vectorlabs,
H-1900-10). Images were acquired using an SP8 Confocal microscope
(Leica). For quantification, the number of RCPs and cell nuclei in the
images were counted digitally using Fiji software (version
1.53C)^[167]59. All oligos listed in Supplementary Table [168]5 were
synthesized by Integrated DNA Technologies, Inc (Coralville, Iowa).
For proteomic mapping, proteins in brain tissue obtained from C57BL/6J
and Nav1 KO (N93) mice were extracted and separated by SDS-PAGE.
Protein bands corresponding with the molecular weight of Nav1 were
excised, and trypsin digested. The digested peptide mixtures were z-tip
purified and run on an Orbitrap Fusion™ Lumos™ Tribrid™ Mass
Spectrometer (ThermoFisher, San Jose), which was equipped with an
Acquity UPLC M-class system (Waters, MA). The peptide data were
searched against the mouse proteome database. Two different Nav1
peptides were found in the C57BL/6J brain sample (MW 202.423;
calculated pI 8.06, Score Sequest HT 2.00323, n = 2 peptides; and MW
252.876, calculated pI 8.76, Score Sequest 2.131383, n = 1 peptide). In
contrast Nav1 peptides were completely absent in brain tissue obtained
from the Nav1 KO. Thus, proteomic mapping confirmed that Nav1 protein
was absent from brain tissue obtained from homozygous Nav1 KO mice,
while Nav1 peptides were detected in C57BL/6J brain tissue.
Food self-administration (FSA) assays
The FSA procedure utilized the same testing conditions as CSA, except
that 20 µl of Chocolate Boost (Nestle) was delivered as the reinforcer.
Additionally, the number of reinforcers per session was not limited to
avoid any ceiling effects; the mice had no prior surgical procedures
and were not tethered to an infusion line. All sessions were terminated
after 2 h of testing. To facilitate acquisition of FSA, home-cage food
was removed ~16 h before the 1st FSA session. Following the 1st
session, the mice were fed 2.5 g of food in the home-cage, in order to
maintain food restriction through the 2nd session. Following the 2nd
session, mice were returned to ad lib feeding for the remaining of the
testing period. The following numbers of mice of the indicated sex were
tested in the FSA assay: Nav1 KO – 6 F, 5 M; Het – 6 F, 5 M; and Wild
type – 6 F, 5 M. The average age of the Nav1 KO, Het and wild-type mice
tested in the FSA assay was 15.8, 15.6 and 14.4 weeks (Supplementary
Table [169]1). Following the 1st FSA experiment, we noted that
homozygous Nav1 KO mice lost more body weight, relative to HET and
wildtype mice, in response to food deprivation. Therefore, we tested a
second cohort of mice under no food restriction. This cohort was
instead subjected to magazine training prior to FSA sessions. This
training involved placing mice in the operant chambers and activating a
program that turned on the white noise and illuminated the nose-poke
apertures and house light. After 10 s, 20 µl of Boost was delivered to
the magazine, the aperture lights turned off and the house light
flashed (1 s on, 1 s off). These conditions continued until the mouse
entered the magazine, as detected by infrared beam break. Entry to the
magazine stopped the flashing house light and illuminated the
apertures. 30 s following the magazine entry, 20 µl of Boost was again
delivered to the magazine, the aperture lights turned off and the house
light flashed until another magazine entry. The sessions were limited
to 50 Boost rewards or 2 h. Mice were required to earn at least 30
rewards before moving on to FSA testing.
After 10 days of FR1 testing, mice in the CSA and FSA studies were
tested under a progressive ratio schedule of reinforcement in one
session. The first reinforcer required one press of the active lever,
and the response requirement was doubled every reinforcer thereafter.
The session ended when the mouse failed to earn a reinforcer in 30 min,
or 2 h total elapsed. The last ratio achieved for each mouse and was
utilized as an indicator of performance.
Analysis of the CSA and FSA data for assessing the effect of the Nav1 KO
C57BL/6J (Wt), heterozygous Nav1 KO (Het) and homozygous Nav1 KO mice
(KO), which were all naive to any prior experimentation, were tested
for CSA using a between-subjects dose-response test with cocaine doses
of 0.1, 0.5, and 1.0 mg/kg. To determine if sex impacted the Nav1 KO
effect on CSA, male and female mice were tested. For all CSA and FSA
studies involving Nav1 KO mice, het-het breeding pairs were used to
generate offspring of all three genotypes, including the wild-type mice
that were utilized as controls. The following numbers of mice of the
indicated sex were tested at each of the following indicated doses:
Nav1 KO – 0.1 Dose: 7 F, 7 M; 0.5 Dose: 8 F, 8 M; 1.0 Dose: 6 F, 8 M;
Het – 0.1 Dose: 7 F, 8 M; 0.5 Dose: 7 F, 8 M; 1.0 Dose: 6 F, 7 M; and
Wildtype – 0.1 Dose: 7 F, 8 M; 0.5 Dose: 7 F, 9 M; 1.0 Dose: 8 F, 6 M.
The average age of the Nav1 KO, Het, and wild-type mice tested in the
CSA assay was 16.2, 16.2, and 16.3 weeks. For the FR1 CSA data, the
results for the effect of a variable (i.e., mouse genotype, sex, or the
effect of an individual session) on the number of cocaine infusions are
reported as an F-statistic [F(variation among sample means / variation
within sample)] and as a p-value. Variables that had a significant
effect were further evaluated by performing pairwise comparisons of
that variable (i.e., genotype or sex) with the CSA results obtained at
each session, and the results are reported as a p-value for the effect
of that variable on the number of cocaine infusions.
Acute locomotor effects of cocaine
Mice, which were naïve to any prior experimentation, were tested in an
acute locomotor dose-response procedure (Nav1 KO – 6F, 6M; Het – 6F,
5M; and Wild type – 6F, 6M). The average age of the Nav1 KO, Het and
wild-type mice tested in the cocaine locomotor assay was 14.7, 14.2,
and 14.3 weeks. Testing occurred in Med Associates open field boxes
(43.2 × 43.2 × 30.5 cm, Med-Associates MED-OFAS-RSU; St Albans VT),
housed individually in sound-attenuating chambers. All testing sessions
were preceded by a 30-min acclimation period in the open field box with
no prior injection. Following this 30-min session, mice were briefly
removed and received an intraperitoneal (ip) injection of either saline
(sessions 1 and 2) or cocaine (sessions 3, 4, and 5) at a dose volume
of 10 mL/kg body-weight, and returned to the open field box for 1 h.
Cocaine was administered at 3 doses (5, 10, and 20 mg/kg body weight)
in a within-subjects dose-response design. All 6 possible dose orders
were utilized and balanced across genotype groups. Sessions 1, 2, and 3
occurred over 3 consecutive days. Sessions 4 occurred 2 days following
session 3 and session 5 occurred 3 days following session 4, in order
to limit any potential effects of prior cocaine exposures.
Locomotor behavior was assessed by distance traveled, as determined by
infrared beam breaks. Distance traveled under no injection, saline
injection, and cocaine injection was assessed. The acute locomotor
effect of cocaine was calculated by subtracting the average of the
distance traveled after both saline injection sessions from the
distance traveled after cocaine injection. The data were binned into
two, 30-minute bins and assessed within bin in addition to the full 1-h
session.
MRI analyses
The brains of age-matched adult male C57BL/6J and Nav1 KO mice
(n = 4/group, age 3–4 months) were examined by in vivo MRI using a
high-field 7 T MRI scanner (Bruker, Billerica, MA) at the Stanford
Center for Innovation in In vivo Imaging (SCi^3) facility. All mice
were anesthetized under 1.0–1.5% isoflurane that was administered by
nose cone throughout the session. Their body temperatures were
supported with warm air, while their respiratory rates were
continuously monitored. Anatomical images were acquired using
T2-weighted turbo rapid acquisition with relaxation enhancement (T2
TurboRARE) with the following parameters: repetition time
(TR) = 2500 ms, echo time (TE) = 40 ms, flip angle = 90 degrees, slice
thickness = 0.5 mm. Slices were obtained in the axial view (coronal in
the mouse) with the first slice starting at the rostral-most extension
of the prefrontal/motor cortex, while the olfactory bulb was excluded.
The DICOM files obtained were processed using Osirix software (Pixmeo
SARL, Bernex, Switzerland). The cortical thickness, hippocampal volume,
and brain volume were manually labeled by an experimenter that was
blinded to the genotype of mice. Measurements were obtained from a
continuous series of slices (for cortical thickness: from +0.63 to
−0.67 mm; for hippocampal volume: −0.77 to −3.37 mm; for brain volume:
+3.33 to −4.87 mm; locations relative to Bregma) that were aligned
across mice groups. The normalized hippocampal volume was calculated by
dividing the absolute hippocampal volume by the total brain volume. The
MRI data were analyzed using Prism 9.1.0 (GraphPad Software, Inc. La
Jolla, CA) with unpaired student t-test.
Whole-cell patch-clamp recording
Brain slices were prepared from anesthetized male mice (3–4 months of
age) using previously described techniques^[170]60. In brief, coronal
slices (~300 μm) were prepared from excised brains that were sectioned
with a vibratome in cold (4 °C) buffer (ACSF) used for tissue slicing
that contains: 126 mM NaCl, 2.5 mM KCl, 1.25 mM NaH[2]PO[4], 1 mM
CaCl[2], 2 mM MgSO[4], 26 mM NaHCO[3], and 10 mM glucose; pH 7.4, when
saturated with 95% O[2]/5% CO[2]. Slices were then transferred to an
incubation chamber filled with standard ACSF buffer containing: 126 mM
NaCl, 2.5 mM KCl, 1.25 mM NaH[2]PO[4], 2 mM CaCl[2], 1 mM MgSO[4],
26 mM NaHCO[3], and 10 mM glucose. The slices were incubated at
33 ± 1 °C for 1 h, and then at room temperature before use. After
incubation, slices were transferred to a recording chamber where they
were minimally submerged (32 ± 1 °C) and perfused at the rate of
2.5–3 mL/min with standard ACSF buffer. Patch electrodes pulled from
borosilicate glass tubing (1.5 mm OD) and had impedances of 4–6 MΩ when
filled with Cs-gluconate based intracellular solution containing:
120 mM Cs-gluconate, 10 mM KCl, 11 mM EGTA, 1 mM CaCl[2], 2 mM MgCl[2],
10 mM HEPES, 2 mM Na[2]ATP, 0.5 mM NaGTP. The osmolarity of the pipette
solution was adjusted to 285–295 mOsm and the pH to 7.35–7.4 with CsOH
and E[Cl]^− was −70 mV calculated from the Nernst equation. Whole-cell
voltage clamp recordings of miniature (m) IPSCs were obtained from the
granule cells in the dentate gyrus of right hippocampus at a holding
potential (V[h]) of +20 mV in the presence of 1 μM tetrodotoxin (TTX,
Ascent Scientific) without application of glutamate receptor blocking
agents^[171]60. Miniature (m) EPSCs were recorded from the granule
cells at V[h] = −70 mV, the estimated E[Cl]^− with the Cs-gluconate
internal solution^[172]60. All recordings were made with a Multiclamp
700 A amplifier, sampled at 10 kHz, filtered at 4 kHz with a Digidata
1320 A digitizer, and analyzed using Clampfit 9.0 (Molecular Devices,
Sunnyvalle, CA), Mini Analysis (Synaptosoft, Decatur, GA), and Prism
(GraphPad software). Only recordings with a stable access resistance
<20 MΩ that varied <15% during the recording were accepted for
analysis. One or two neurons were recorded per slice, and no more than
three slices were used per mouse.
Tissue collection and sectioning
After isoflurane anesthesia, transcardial perfusion (Harvard Apparatus
p-70, Holliston, MA) was performed with a 0.15 M NaCl solution was
followed by fixation in 4% paraformaldehyde (Aldrich Chemistry,
Darmstadt, Germany). The brains were then extracted and post-fixed in
4% paraformaldehyde for 4 h. After fixation, the brains were
transferred into a 30% sucrose solution, and then stored at 4 °C until
sectioning. The brain tissue slices used for immunohistochemistry were
sectioned at 30 µm, and then stored in a cryo-protective solution (PBS,
20 g PVP-40, 600 ml ethylene glycol, 600 g sucrose) at −20 °C.
Immunohistochemistry
The immunohistochemical analyses were performed as described in ref.
^[173]61. In brief, brain sections were rinsed 5 times in PBS
(Sigma-Aldrich P5368-10pak) for 5 min, and then blocked in 10% normal
donkey serum and 0.3% Triton X-100 in PBS to minimize nonspecific
binding. The sections were then incubated with the primary antibodies
(Supplementary Table [174]4) in the blocking solution for 48 h at 4 °C.
The sections were then rinsed in PBS and incubated with a corresponding
secondary antibody for 4 h. After the last rinse, the sections were
mounted, air dried overnight; and were then sealed with a cover slip in
Dako Fluorescence Mounting Medium (S3023, Dako North America, Inc.,
Carpinteria, CA) and stored at −20 °C.
Image capture
Images were captured and analyzed using a Leica SP8 white light pulsed
laser confocal microscope and Leica LAS X Premium software at the Cell
Sciences Imaging Facility at Stanford. Channels were selected and the
exposure times were adjusted to optimize the images. Imaris software
(Bitplane Inc., Concord, MA) was used for image capture^[175]62.
Analysis of learning, memory, and exploratory behaviors
Pre-experiment habituation was performed on all mice used in the
behavioral tests described below; this involved daily handling and
habituation that was initiated a week before the behavioral testing
began. Mice were picked up by hand, stroked and touched for
approximately 2 min per session. A quick assessment was made during the
last handling session; if a mouse exhibited high levels of anxiety-like
behavior (incontinence, hyperactivity, etc.), 1–2 additional handling
sessions were conducted before experimental testing.
Open field test
The open field test was conducted as described^[176]63 using a
40 × 40 × 40 cm (l × w × h) square arena with white plastic boards.
Each mouse was habituated within the procedure room for 30 min and was
then placed in the center of the open field arena. The total distance
traveled and duration of time in the center (~25% of total area) or
edge over a period of 15 min were recorded and analyzed using Viewer
III software (BIOBSERVE, Bonn, Germany). The total duration of the open
field test was 15 min.
Novel object recognition test
The novel object recognition test was conducted as described^[177]64 in
the apparatus used for the open field tests. A test mouse was given
free access to the entire chamber for a 5 min habituation period. When
the training session started, two identical objects (Lego blocks or a
flask filled with bedding) were placed in diagonally opposite corners
of the arena (6 cm from the wall), and the test mouse was allowed to
freely explore for 10 min. After 24 h, the test mouse was returned to
the center of the arena and habituated for 5 min. When the testing
session started, one familiar and one novel object were presented at
the same positions in the arena. The test mouse was then given 10 min
to explore the objects, while exploratory behaviors (sniffing, rearing
against the objects, and head within 2 cm of the object) were recorded.
Videos were processed using Viewer III tracking software by experienced
personnel. The first 5 min of the training and testing sessions were
used for analysis. The discrimination index (DI) was calculated as:
[MATH: (Explorationtimewithnovelobject−Explorationtimewithfamiliarobject)/Explorationtimewithnovelobject+Explorationtimewithfamiliarobject :MATH]
1
Elevated plus maze test
The elevated plus maze test was conducted as described^[178]63 in an
arena with two open arms and two closed arms that are raised 50 cm
above the floor. All arms were 30 × 5 cm (l × w) with white walls
(15 cm height) and floors. The test started 2 h after the end of the
dark cycle. Thirty minutes after acclimation in the behavioral room,
the test mouse was placed in the center of the maze and allowed to
freely explore the arena. The total test duration was 5 min. The
duration of time spent in exploratory behaviors, the number of entries
and the distance traveled in the open and closed arms (excluding
center) were analyzed using Viewer III tracking software.
Rotarod test
Motor coordination was evaluated using methods that are described in
the Standard Operating Procedures of the Jackson Laboratory Mouse
Neurobehavioral Phenotyping Facility using a five-station rotarod
treadmill (ENV-575M, Med Associates, St. Albans, VT). Mice were first
acclimated to the behavioral room for 1 hr before testing. The testing
session consisted of three trials; in each trial, the speed was
increased from 4 to 40 rpm; and each trial was separated by an interval
of 1 min. A trial was terminated when a mouse fell off, clung to the
rod and completed full passive rotation, or after 300 s. The duration
and end speed on rotarod were recorded and averaged from the last two
trials for each mouse.
Barnes maze test
Barnes maze tests were conducted as described in ref. ^[179]65. The
experimental protocol consisted of a habituation session (day 1); 12
training sessions with 3 trials per day and a 15 min intertrial
interval on days 2–5; and the testing session on day 6. During the
training sessions, the mice were released into the middle of the maze,
and they learned to enter the open escape hole to avoid exposure to a
strong light. Three visual cues were placed at distinct positions
outside of the maze to facilitate learning. The maze was cleaned with
70% ethanol thoroughly between trials to eliminate olfactory cues.
Twenty-four hours after the training sessions, mice were tested in the
arena for 90 s with all holes closed. The results evaluated include
primary errors (errors made before reaching the escape hole), latency
(the time elapsed before reaching the escape hole), track length (the
total length traveled), and target hole preference (percentage of time
spent adjacent to the escape hole). Their performance was recorded and
analyzed using Viewer III tracking software.
Statistics and reproducibility
Prism 9.1.0 was used for analysis of the MRI data and Prism 8.4.1 for
Windows (GraphPad Software, Inc. La Jolla, CA) was used for analyzing
the behavioral data. Other data were analyzed using Statistica 8.0
(TIBCO Software, Inc. Palo Alto, CA) or SPSS (IBM Corp. Released 2020.
IBM SPSS Statistics for Windows, Version 27.0. Armonk, NY: IBM Corp).
The CSA, FSA, and acute cocaine locomotor behavior results were
assessed by ANOVA. Interactions and main effects were decomposed by
simple main effects and pairwise comparisons with Sidak correction for
multiple comparisons. Since CSA data tends to depart from normal
distributions, all CSA data was log[10] transformed prior to analysis.
For the acute locomotor response, Barnes maze test, and novel object
recognition test, the statistical significance was determined using a
two-factor ANOVA with repeated measures. Stimulus (familiar vs novel)
or genotype (control vs. KO) were the two factors assessed. Post-hoc
analysis was conducted using a Bonferroni post-test. Exploratory
behaviors were evaluated using a one-way ANOVA with Tukey’s post-test.
In the figures, the significance levels are indicated by: *p < 0.05,
**p < 0.01, ***p < 0.001, ****p < 0.0001.
snRNA-Seq analysis
The protocol used for nuclear isolation was modified from that
developed by 10X Genomics. Brain tissue was obtained from age-matched
adult male Nav1 KO (n = 4) and C57BL/6J (n = 5) mice. The PFC was
quickly dissected from the freshly extracted brain tissue and it was
place in chilled Hibernate AB Complete (HEB) medium (BrainBits LLC,
Springfield, IL) at 4 °C. Freshly dissected PFC from each group were
pooled in separate 50 ml conical tubes with a minimum amount of chilled
Hibernate AB Complete (HEB) medium (BrainBits LLC, Springfield, IL).
Then, 5 ml of chilled lysis buffer (2 mM Tris-HCl, 2 mM NaCl, 0.6 mM
MgCl[2] and 0.02% Nonidet^TM P40 Substitute in nuclease-free water) was
added to the tubes, which were then incubated at 4 °C for 10 min. The
amount of lysis buffer and the incubation time were optimized
determining the amount of buffer and incubation time that produced
high-quality nuclei. After adding 5 more ml of HEB medium to the tubes,
the tissues were triturated with a fire-polished silanized Pasteur
pipette for 10–15 passes and then strained with 30 μm MACS strainer
(Miltenyi Biotec Inc, Auburn CA). Nuclei were pelleted by
centrifugation at 500 × g for 5 min at 4 °C and were then resuspended
in a chilled PBS with 1% BSA (Invitrogen AM2618, Thermo Fisher
Scientific, Pittsburgh PA) wash buffer with 0.2 U/ul RNase inhibitor
(#3335402001, Millipore Sigma, Darmstadt Germany). The nuclei were
pelleted and resuspended twice, strained using a 30 μm MACS strainer,
and then centrifuged and resuspended in a chilled wash buffer. Mouse
anti-NeuN antibody (MAB377, Millipore Sigma, Darmstadt Germany) was
added at a 1:500 dilution, and the samples were incubated for 40 min at
4 °C. Samples were then centrifuged at 400 × g for 5 min at 4 °C and
the pelleted nuclei were resuspended with chilled wash buffer. Alexa
647 chicken anti-mouse antibody (A21463, Life technologies, Pittsburgh
PA) was added to the preparations at a 1:500 dilution and the samples
were incubated for 40 min at 4 °C. After incubation, nuclei were
pelleted and resuspended, and Hoechst 33342 (Invitrogen H3570, Thermo
Fisher Scientific, Pittsburgh PA) was added to a final concentration of
0.25 μg/ml.
Single nuclei sorting was conducted using a 6-laser BD Influx sorter
(BD Biosciences, San Jose CA) with a 100-um nozzle in the Stanford
Shared FACS Facility. Various control samples (unstained,
Hochest33342^+ or NeuN^+) were examined to optimize the gating
strategy. Nuclei were gated based on size, scatter properties and
staining for Hoechst and NeuN. To ensure that we were able to analyze
different types of cells, ~60% of the sorted nuclei were collected as
NeuN^+ and 40% were NeuN^-. Single nuclei were sorted into collecting
tubes and then visually inspected under a microscope for quality
control; they were then pelleted and resuspended to produce a solution
with 600 nuclei/ul in the final volume. Single nuclei were then
captured in droplets with barcodes using the 10x Genomics Chromium
system and cDNA libraries were produced using Chromium Next GEM Single
Cell 3’ Reagent Kits v3.1 (10x Genomics, Pleasanton, CA) according to
the manufacturer’s instructions. The Nav1 KO and C57BL/6J samples were
sequenced on the NovaSeq platform (Illumina, San Diego CA).
FASTQ files with the snRNA-Seq data were processed using Cell Ranger
software (v6.0.0) and the cellranger count pipeline to generate a
filtered feature-barcode matrix (the gene expression matrix). The reads
within the C57BL/6J and Nav1 KO FASTQ files were aligned using the Cell
Ranger built-in mouse reference (mm10). Two gene expression matrices
were produced: 23445 features × 15402 cells for C57BL/6; and 23737
features × 14623 cells for Nav1 KO. The C57BL/6 J and Nav1 KO gene
expression matrices were then analyzed using the R/Seurat (v4.0.1)
package. Low-quality cells with unique gene counts <200 and >10%
mitochondrial counts were filtered out. The high-quality C57BL/6J
(20955 features × 14,630 cells) and Nav1 KO (21,310 features × 14,056
cells) matrices were merged into one Seurat object (21,892 features ×
28,686 cells) that was used for subsequent analyses (data
normalization, variable feature identification, dimensional reduction,
etc.).
The gene counts for each cell within the global matrix was normalized
by dividing it by the total counts in that nucleus; and this number was
multiplied by 10,000 and were then natural log transformed to become
the normalized values. We also identified 2000 variable features, which
exhibited high cell-to-cell variability in the matrix. The matrix was
then further scaled for linear dimensional reduction purposes. For the
scaled gene expression matrix, the mean expression of a gene across
different cells is set to 0 and the variance is set to 1. Principal
component analysis (PCA) was performed on the scaled data, and the
first 10 PCs were chosen to represent the dimensionality of the global
matrix. The default K-nearest neighbor (KNN) graph-based method and
Louvain algorithm for single-cell clustering was used; and the
resolution parameter was set to 0.3. In total, 14 clusters (cluster 0
to 13) were classified; and of these, cluster 0 contained the most
cells while cluster 13 contained the fewest. To visualize the cells in
low-dimensional space (to aid interpretation), the non-linear
dimensional reduction program (UMAP)^[180]66 was performed using the
first 10 PCs.
The differentially expressed genes (DEG) for each cluster were detected
using a minimum percentage of 0.25 in either of the two clusters and an
average log[2] fold-change (FC) ≥ 0.25. Cell type-specific canonical
markers were used to determine the cell type identity of the 14
clusters. Non-neuron cells were assigned as follows: Astrocyte (Gja1,
Aqp4), Microglia (C1qa), and Oligodendrocyte (Apsa, Mbp). Of note, no
endothelial cell markers (Flt1, Cldn5, Nostrin) were highly expressed
in any of 14 clusters. The type of neuronal cells was determined by
whether they expressed mRNA markers for excitatory (Slc17a7, Tshz2,
Thsd7a) or inhibitory (Gad1, Gad2, Slc32a1, Meis2) neuronal markers. To
verify the cell type assignments, our C57BL/6J gene expression matrix
was compared with that of a published reference
([181]GSE124952)^[182]32 data set for PFC cells obtained from saline
control C57BL/6J mice, which generated an expression matrix with 20718
features × 11886 cells. The canonical correlation analysis (CCA)
between this reference and our C57BL/6J dataset was used to remove
batch effects before matrix integration. We then projected the
reference single cells with their cell type labels onto the UMAP plot
for comparison with our C57BL/6J dataset.
Reporting summary
Further information on research design is available in the [183]Nature
Portfolio Reporting Summary linked to this article.
Supplementary information
[184]Supplementary Information^ (5.3MB, pdf)
[185]Description of Supplementary Files^ (14.6KB, docx)
[186]Supplementary Data 1^ (118.6KB, xlsx)
[187]Reporting Summary^ (1.2MB, pdf)
Acknowledgements