α7 nicotinic acetylcholine receptor (nAChR) agonists are candidates for the treatment of cognitive deficits in schizophrenia. Selective α7 nAChR agonists, such as SSR180711, activate neurons in the medial prefrontal cortex (mPFC) and nucleus accumbens shell (ACCshell) in rats, regions important for cognitive function. However, the neural substrates involved in these effects remain elusive. Here we identify cortically projecting cholinergic neurons in the horizontal limb of the diagonal band of Broca (HDB) in the basal forebrain (BF) as important targets for α7 nAChR activation, as measured by c-Fos immunoreactivity, a marker of neuronal activation. Selective depletion of these cholinergic neurons abolishes the SSR180711-induced activation of the mPFC but not the ACCshell, demonstrating their critical importance for α7 nAChR-dependent activation of the mPFC. Contrarily, selective depletion of dopaminergic neurons in the ventral tegmental area abolishes the SSR180711-induced activation of the ACCshell but not the mPFC or HDB. These results demonstrate 2 distinct neural pathways activated by SSR180711. The BF and mPFC are important for attentional function and may subserve the procognitive effects of α7 nAChR agonists, whereas activation of the ACCshell is implicated in the beneficial effect of antipsychotics on the positive symptoms of schizophrenia.
α7 nicotinic acetylcholine receptor (nAChR) agonists improve attentional function in healthy volunteers and patients with schizophrenia (Kitagawa et al. 2003; Olincy et al. 2006; Freedman et al. 2008). The neural basis for this procognitive effect has not been elucidated, but animal studies have suggested several possible mechanisms.
We have previously reported that α7 nAChR agonists activate limbic brain regions in rats, including the prefrontal cortex (PFC) and shell subregion of the nucleus accumbens (ACCshell), as measured by induction of the immediate-early genes c-Fos and Arc (Hansen et al. 2007; Kristensen et al. 2007; Thomsen et al. 2008). This activation may be caused by local activation of α7 nAChRs or by activation of α7 nAChRs in regions projecting to these areas. The α7 nAChR is widely present in the rat and human brain, including the PFC, basal forebrain (BF), and ventral tegmental area (VTA), whereas the nucleus accumbens is practically devoid of α7 nAChR receptors (Clarke et al. 1985; Breese et al. 1997; Tribollet et al. 2004), suggesting that α7 nAChRs in the nucleus accumbens shell (ACCshell) are not responsible for activation of this region. Furthermore, activation of α7 nAChRs promote the release of acetylcholine and dopamine in the PFC (Biton et al. 2007; Tietje et al. 2008) and dopamine in the ACCshell (Schilstrom et al. 1998; Mansvelder and McGehee 2000), suggesting the involvement of cholinergic and dopaminergic pathways.
The BF cholinergic corticopetal system is important for attentional function (Sarter et al. 1999, 2005). In addition, BF neurons have a high expression of α7 nAChRs (Breese et al. 1997; Tribollet et al. 2004) and project strongly to the PFC (Rye et al. 1984; Gritti et al. 2006). The horizontal limb of the diagonal band of Broca (HDB), in particular, projects strongly to the prelimbic region of the PFC (Luiten et al. 1987; Gaykema et al. 1990; Hoover and Vertes 2007). Furthermore, it has been shown that SSR180711 increases the firing rate of neurons in the BF in anesthetized rats, although it is not known which cell types are activated or where they project (Biton et al. 2007).
Presynaptic α7 nAChRs on glutamatergic terminals in the VTA can activate dopaminergic neurons in the VTA (Schilstrom et al. 1998; Mansvelder and McGehee 2000), which send projections to the PFC, ACCshell, and BF (Fallon and Moore 1978; Albanese and Minciacchi 1983). In addition, local injection of the selective α7 nAChR antagonist methyllycaconitine (MLA) into the VTA blocks nicotine-induced dopamine release in the nucleus accumbens (Schilstrom et al. 1998), suggesting the involvement of α7 nAChRs in the VTA in the activation of the ACCshell.
Here we show that cortically projecting cholinergic neurons in the HDB are activated by the selective α7 nAChR agonist SSR180711 and are necessary for the SSR180711-induced activation of the medial prefrontal cortex (mPFC) but not the ACCshell. Furthermore, we show that dopaminergic neurons in the VTA are necessary for the SSR180711-induced activation of the ACCshell but not the mPFC.
Materials and Methods
Animals and Surgeries
A total of 115 adult male Wistar rats (200–300 g; Taconic Europe, Ll. Skensved, Denmark) were used in this study. The rats were housed under standardized conditions with free access to food and water for a minimum of 7 days prior to experiments. All experiments were conducted in accordance with guidelines of the National Institute of Health (NIH Publications No. 85-23, 1985) and the Animal Experimentation Inspectorate, Ministry of Justice, Denmark.
Four kinds of operations were performed where a total of 60 rats (housed 2 per cage) were anaesthetized with 0.8 ml hypnorm/dormicum (0.079 mg/ml fentanyl citrate, 2.5 mg/ml fluanisone, and 1.25 mg/ml midazolam) and placed in a stereotaxic frame (Kopf, Tujunga, CA), where a hole was drilled in the skull over the injection site(s) with a Micromotor™ drill (Stoelting, Wood Dale, IL). All stereotaxic injections were made with reference to Bregma, according to a standard rat brain stereotaxic atlas (Paxinos and Watson 1986), and the incisor bar was 2.7 mm below the ear bars.
For retrograde tracing (17 rats), glass micropipettes were cut to a tip-diameter of 15–25 μm and filled with a 0.4% cholera toxin subunit B (ChB) solution containing 0.08 M NaPO4, pH 7.5 (List Biological Laboratories, Campbell, CA). This solution was applied iontophoretically using positive current pulses (7 s on; 7 s off) of 10 μA for 10 min with a Midgard™ precision current source (Stoelting), where after the micropipette was left in place for 2 min to avoid a trace of ChB when retracting the syringe. The coordinates used for the mPFC were posterior 2.5 mm, lateral −0.5 mm, and ventral −3.5 mm.
For lesions with the immunotoxin 192 IgG-Saporin (SAP, 15 rats), injections were performed bilaterally with a 2 μL 25-gauge Hamilton syringe. SAP (MAB394; Millipore, Copenhagen, Denmark) was injected into the right hemisphere, and vehicle was injected into the left side. The syringe was filled with 0.5 μg/μL SAP in Dulbecco's phosphate-buffered saline (PBS) and was left in place 3–4 min before and after the injection, which was 0.6 μL at a rate of 0.2 μL/min. The coordinates used for the HDB were posterior −0.4 mm, lateral ±2.0 mm, and ventral −8.5 mm with reference to Bregma (Paxinos and Watson 1986).
For intracerebral injections of receptor modulators (18 rats), guide cannulae (22 gauge; Plastics One, Roanoke, VA) were implanted bilaterally into the HDB with the same coordinates as above. The tips of the guide cannulae were located 1 mm above the actual injection site to avoid tissue damage. The guide cannulae were then fixed to the skull with anchoring screws, super glue, and dental cement. Stainless steel dummy cannulae prevented the guide cannulae from clogging.
For 6-hydroxydopamine (6-OHDA, 18 rats) lesions, injections were performed bilaterally with a 10 μL 26-gauge Hamilton syringe. 6-OHDA (Sigma, St Louis, MO) was injected into the right hemisphere, and its vehicle was injected into the left side. The syringe was filled with 3.0 μg/μL 6-OHDA hydrochloride in 2% ascorbic acid and was left in place 3–4 min before and after the injection, which was 2.5 μL at a rate of 1 μL/min. The coordinates used for the VTA were posterior −4.0 mm, lateral ±0.8 mm, and ventral −8.0 mm with reference to Bregma (Paxinos and Watson 1986).
After surgery, a 1:1 mixture of lidocaine and bupivacaine was applied locally in the incision site for local anesthesia, and 5 mg/kg rimadyl was injected subcutaneously (s.c.) once a day for 3 days for pain relief. After a recovery period (7–14 days for tracing, 6–7 days for SAP lesion, and 22–23 days for 6-OHDA lesion), animals received a single s.c. injection of saline or SSR180711 (10 mg/kg) and were returned to their home cage for 60 min before transcardial perfusion.
For intracerebral injections, the animals were allowed a recovery period of 7 days after which 0.5 μL per side was administered over 30 s using a Hamilton pipette attached to an injection cannula. The injection cannulae were left in place for another 30 s, where after the animals were put back in their home cage for 60 min before transcardial perfusion. Animals received SSR180711 (0.5 or 5 μg) or MLA (10 μg) in one side and vehicle (saline) in the other. The intracerebral injection of MLA was combined with an s.c. injection of SSR180711 (10 mg/kg).
Thirty-two rats (housed ≤6 per cage) used for phenotypic neuron identification received a single s.c. injection of saline or SSR180711 (1, 3, or 10 mg/kg) with no prior surgery and were returned to their home cage for 60 min before transcardial perfusion. Similarly, 25 rats received a single s.c. injection of saline or MLA (10 mg/kg) 10 min before receiving saline or SSR180711 (10 mg/kg) s.c. and were returned to their home cage for 120 min before transcardial perfusion.
Transcardial Perfusion and Fixation
All rats were deeply anaesthetized with mebumal (50 mg/ml and 3 ml/kg), perfused transcardially with 0.9% saline for 2 min followed by fixation in ice-cold 4% paraformaldehyde in 0.1 M PBS for 8 min. Brains were dissected, immersed in fixative overnight at 4°C, and submerged in 30% sucrose in PBS at 4°C for 2–4 days. Forty-micrometer serial coronal sections were cut through the forebrain in 8 parallel series on a freezing microtome, and sections were stored at −20°C in antifreeze solution (0.01 M NaH2PO4, 0.03 M Na2HPO4, 30% ethylenglycol, and 30% glycerol) until use.
The sections were rinsed for 3 × 10 min in 0.01 M PBS, incubated for 10 min in 1% H2O2 in PBS to block endogenous peroxidase activity, and incubated for 20 min in 0.01 M PBS with 5% swine serum, 1% bovine serum albumin (BSA), and 0.3% Triton X-100 (TX) to block nonspecific binding sites. The sections were then incubated at 4°C for 18–24 h with the primary antiserum in 0.01 M PBS, 1% BSA, and 0.3% TX. After incubation in primary antiserum, the sections were washed for 3 × 10 min in 0.01 M PBS with 0.1% TX (PBS–TX), and incubated for 60 min in biotinylated secondary antibody diluted 1:1000 in PBS–TX with 1% BSA, washed for 3 × 10 min in PBS–TX, and transferred to an avidin–biotin complex solution (Vector Laboratories, Burlingame, CA) diluted 1:250 in PBS–TX for 60 min. After washing 10 min in PBS–TX, PBS, and Tris–HCl (pH 7.6), respectively, the sections were incubated in either 0.1% diaminobenzidine (DAB; Sigma) with 0.03% H2O2 in Tris–HCl buffer, for single staining, or 0.025% DAB with 0.015% H2O2, 0.075% nickel sulfate, and 0.005 M imidazole in Tris–HCl buffer, for double staining, for 10 min and then washed 3 × 10 min in PBS buffer. For double staining, the procedure was repeated with a different primary antiserum and finished with incubation in 0.1% DAB with 0.03% H2O2 in Tris–HCl buffer for 10 min. The sections were mounted on gelatinized glass slides, dried, and coverslipped in Pertex. For double staining, c-Fos was always used with the DAB-nickel reaction producing a black color localized to the nucleus, whereas the DAB reaction was orange–brown.
Primary antibodies used and their concentrations were as follows: polyclonal antiserum against c-Fos (1:4000 for single staining and 1:8000 for double staining), generated in a rabbit with a peptide corresponding to amino acids 2–17 of the rat c-Fos protein in our laboratory and characterized previously (Mikkelsen et al. 1998), polyclonal goat antiserum against ChB (1:2000 for single staining and 1:4000 for double staining, #703; List Biological Laboratories), polyclonal goat antiserum against choline acetyltransferase (ChAT, 1:1000, #AB144; Millipore), monoclonal mouse antibody against parvalbumin (PV, 1:80 000, #P3088; Sigma), monoclonal IgM mouse antibody against phosphate-activated glutaminase (PAG, 1:3000, kindly provided by Dr Takeshi Kaneko, Kyoto University) (Kaneko and Mizuno 1988; Kaneko et al. 1988), and monoclonal mouse antibody against tyrosine hydroxylase (TH, 1:10 000, MAB318; Millipore). Secondary antibodies were all raised in donkey, from Jackson Laboratories (Ben Harbor, ME) and used in a 1:1000 dilution.
Quantification of Immunoreactive Neurons
For double staining and ChAT staining, the number of ChB immunoreactive (IR), ChAT-IR, PV-IR, and PAG-IR neurons, respectively, and neurons, which were also c-Fos-IR, were counted in the area of interest on 3–4 sections spanning the anterior–posterior axis of the region of interest under light microscopy with ×60 magnification (Fig. 1A). As there were no clear visible anatomical boundaries, the area counted as HDB possibly also includes a small part of the magnocellular preoptic nucleus and the lateral preoptic area. The data are presented as the percentage of ChB-IR, ChAT-IR, PV-IR, and PAG-IR neurons, respectively, which were also c-Fos-IR. To ensure validity of the statistical tests for the tracing study, only animals with >15 neurons in the HDB were included in the data set.
For single staining, the number of c-Fos-IR cells in layer II–III in the prelimbic region of the mPFC and the ACCshell was counted using light microscopy with ×20 magnification using a counting grid (0.5 × 0.5 mm) placed over the region of interest (Fig. 1B,C), by an observer blind to the treatment of the animals as described earlier (Thomsen et al. 2008). The number of positive cells for each region was averaged from 2 adjacent sections of each animal.
Data are presented as mean ± SEM. Repeated measures 2-way analysis of variance (ANOVA), 1-way ANOVA followed by Tukey's post hoc test, and Bonferroni-corrected paired or unpaired t-test were used to compare groups, as appropriate, and a P value less than 0.05 was considered significant.
SSR180711 Activates BF Neurons Projecting to the mPFC
ChB was injected into the mPFC of rats and was allowed 7–14 days to be retrogradely transported to the soma of neurons projecting to this region. The site and extent of the injections are illustrated in Figure 2. The injections were centered in the prelimbic region of the mPFC and often filled most of this region but often extended into the infralimbic cortex and sometimes into cingulate cortex area 1. Many retrogradely traced ChB-IR somata were found in the contralateral mPFC and the ipsilateral, ventrolateral orbitofrontal cortex and dorsal endopiriform (dEP) cortex. In the BF, most ChB-IR soma were found in the HDB (Supplementary Fig. 1), which corresponds with earlier studies showing strong projections from this region to the prelimbic cortex (Luiten et al. 1987; Gaykema et al. 1990; Hoover and Vertes 2007).
In SSR180711-treated rats (10 mg/kg), ∼26% of the cortically projecting ChB-IR neurons in the HDB were also c-Fos-IR, corresponding to a ∼7-fold increase compared with saline controls (P < 0.0001, Fig. 3A,C). This induction was selective to the HDB, as no activation of ChB-IR cells was observed in the dEP, a region that also projected strongly to the mPFC (Fig. 3B).
The SSR180711-induced activation of cells in the HDB was accompanied by an increased expression of c-Fos in the mPFC and ACCshell (P < 0.01, Fig. 3D,E), indicating that the tracer injection did not affect SSR180711-induced activation of the mPFC, as reported earlier (Hansen et al. 2007).
SSR180711 Specifically Activates Cholinergic Forebrain Neurons
Double immunohistochemical staining was performed on brain sections from rats that received saline or SSR180711 (1, 3, and 10 mg/kg). SSR180711 induced a dose-dependent increase in the proportion of c-Fos-IR ChAT-IR cells in the HDB (Fig. 4A). This increase was significant for the 3 and 10 mg/kg groups, where 25% and 27% of the ChAT-IR cells were also c-Fos-IR, respectively, compared with 8% in saline-treated animals (F3,26 = 30.74, P < 0.001). In contrast, there was no significant effect of SSR180711 on the proportion of PV-IR or PAG-IR c-Fos-IR cells in the HDB (F3,28 = 0.34 and 0.57, respectively, Fig. 4B,C). Furthermore, single staining for c-Fos showed no significant increase in the number of c-Fos-IR cells in the HDB (data not shown), precluding the induction of c-Fos in a major cell type in the HDB that has not been identified in our study.
The induction of c-Fos in cholinergic neurons by SSR180711 was region selective, as there was no effect on cholinergic neurons in other regions of the BF, including the medial septum, ventral pallidum, magnocellular preoptic nucleus, substantia innominata, or the nucleus basalis of Meynert (Supplementary Fig. 2). In other regions, including the nucleus accumbens and the lateral preoptic nucleus, very few neurons were ChAT-IR, wherefore these data have not been included.
The involvement of the α7 nAChR in the effect of SSR180711 on HDB neurons was demonstrated in a separate experiment, where MLA was administered before SSR180711. Preadministration of MLA (10 mg/kg) completely blocked the SSR180711-induced increase in c-Fos/ChAT-IR cells in the HDB (F3,21 = 10.38, Fig. 5).
Partial Depletion of Cholinergic Neurons Abolishes the SSR180711-Induced Activation of the mPFC
A unilateral injection of SAP into the right HDB of rats caused a reduction in ChAT-IR neurons in this region (Fig. 6A). In some rats, the lesion also included other parts of the BF to some degree, primarily the magnocellular preoptic nucleus and the lateral preoptic nucleus. Unilateral SAP lesions caused a significant ∼50% reduction of ChAT-IR cells in the HDB of both SSR180711- and vehicle-treated rats compared with the vehicle-treated hemisphere (Fig. 6B), but did not affect the number of ChAT-IR neurons in the medial septum (Fig. 6C). In addition, SAP treatment did not affect the number of PV-IR cells in the HDB (Fig. 6D), indicating that the lesion was selective for cholinergic neurons.
The SAP-lesioned animals were treated with either SSR180711 or saline, and the number of c-Fos-IR neurons was counted in the mPFC and ACCshell. In the mPFC, a repeated measures 2-way ANOVA with treatment and hemisphere as the fixed factors revealed a significant treatment × hemisphere interaction (F1,11 = 18.38, P < 0.01, Fig. 7A). Subsequent Bonferroni-corrected paired t-tests showed a significant induction of c-Fos-IR in the sham-operated versus SAP-lesioned side of SSR180711-treated (10 mg/kg) animals (P < 0.01), whereas there was no side difference in saline-treated animals, suggesting that the lesion interfered with SSR180711-induced but not basal levels of c-Fos-IR in the mPFC. When data are expressed as the difference between c-Fos-IR in the SAP-lesioned and sham-operated side for each rat individually, the selective induction of c-Fos in the sham-operated side of SSR180711-treated animals became even clearer (P < 0.01, Fig. 7A, insert), due to a high interindividual variability of c-Fos-IR in saline-treated animals.
In the ACCshell, a repeated measures 2-way ANOVA with treatment and hemisphere as the fixed factors showed a significant main effect of treatment (F1,12 = 9.66, P < 0.01) but no treatment × hemisphere interaction (F1,12 = 0.46) and no effect of hemisphere (F1,12 = 4.10), indicating that the SAP lesion did not affect SSR180711-induced or basal levels of c-Fos-IR in the ACCshell (Fig. 7B).
BF α7 nAChRs Activate the mPFC
The placement of the injection cannulae for intracerebral injections was confirmed during sectioning of the tissue, and only animals with injections into the HDB were included.
A unilateral injection of 10 μg MLA into the HDB significantly attenuated the increase in c-Fos-IR in the mPFC induced by a systemic injection of SSR180711 (20% decrease, P < 0.05, Fig. 8A). Contrarily, intra-HDB injections of MLA did not affect c-Fos-IR in the ACCshell (Fig. 8B). MLA injections did not affect c-Fos levels in saline-treated rats (data not shown).
A unilateral injection of 5 μg SSR180711 into the HDB significantly increased c-Fos-IR in the mPFC. Thus, a repeated measures 2-way ANOVA with treatment and hemisphere as the fixed factors revealed a significant treatment × hemisphere interaction in the mPFC (F1,8 = 18.90, P < 0.01), and subsequent Bonferroni-corrected paired t-tests showed a significant induction of c-Fos-IR in the 5-μg group compared with the saline-treated hemisphere (P < 0.05), whereas 0.5 μg SSR180711 did not affect c-Fos-IR in the mPFC (Fig. 8C). Neither dose of SSR180711 affected c-Fos-IR in the ACCshell (Fig. 8D).
Depletion of Midbrain Dopaminergic Cells Abolishes the SSR180711-Induced Activation of the ACCshell
Unilateral 6-OHDA injections into the right medial forebrain bundle reduced TH-IR in the ipsilateral VTA (Fig. 9). Only rats with a virtually complete loss of TH-IR neurons in the VTA were included. The lack of TH-IR neurons in the VTA was reflected in a lack of TH-IR terminal staining in the ipsilateral accumbens. In most rats, the lesion also included the medial part of the substantia nigra to some degree.
After a 22- to 23-day recovery period, the 6-OHDA-lesioned animals were treated with either SSR180711 or saline, and the number of c-Fos-IR neurons was counted in the mPFC and ACCshell. In the ACCshell, a repeated measures 2-way ANOVA with treatment and hemisphere as the fixed factors revealed a significant treatment × hemisphere interaction (F1,15 = 11.93, P < 0.01, Fig. 10B). Subsequent Bonferroni-corrected paired t-tests showed a significant induction of c-Fos-IR in the sham-operated versus 6-OHDA–lesioned side of SSR180711-treated (10 mg/kg) animals (P < 0.01), whereas there was no side difference in vehicle-treated animals, suggesting that the lesion interfered with SSR180711-induced but not basal levels of c-Fos-IR in the ACCshell.
In the mPFC, a repeated measures 2-way ANOVA with treatment and hemisphere as the fixed factors showed a significant main effect of treatment (F1,15 = 2.78, P < 0.05) but no treatment × hemisphere interaction (F1,15 = 0.99) and no effect of hemisphere (F1,15 = 0.02), indicating that the 6-OHDA lesion did not affect SSR180711-induced or basal levels of c-Fos-IR in the mPFC (Fig. 10A).
Double staining for c-Fos and ChAT revealed that SSR180711 increased the proportion of c-Fos/ChAT-IR cells in the HDB of both the lesioned and the intact hemisphere (Fig. 11). Thus, a repeated measures 2-way ANOVA with treatment and hemisphere as the fixed factors showed a significant main effect of treatment (F1,15 = 33.19, P < 0.0001) but no treatment × hemisphere interaction (F1,15 = 1.11) and no effect of hemisphere (F1,15 = 0.23).
Here we demonstrate that the selective α7 nAChR agonist SSR180711 activates cortically projecting neurons in the HDB and that in the HDB, SSR180711 selectively activates cholinergic but not glutamatergic or PV-positive neurons. We further show that selective depletion of cholinergic neurons in the HDB abolishes SSR180711-induced increase in c-Fos-IR in the mPFC but not the ACCshell. Additionally, we show that depletion of dopaminergic neurons in the VTA abolishes the SSR180711-induced increase in c-Fos-IR in the ACCshell but not the mPFC. This demonstrates the functional importance of these cholinergic and dopaminergic neurons in the SSR180711-induced activation of the mPFC and ACCshell, respectively, and indicates that they comprise distinct neural pathways (Fig. 12).
Acute administration of SSR180711 induced c-Fos-IR in the mPFC and ACCshell of rats, which is in accordance with previous findings from our lab (Hansen et al. 2007; Thomsen et al. 2008). In addition, we used tracing with ChB, which is taken up by axon terminals and retrogradely transported to the soma (Luppi et al. 1990), to show that SSR180711 increases the proportion of c-Fos-IR neurons by ∼7-fold in neurons projecting to the mPFC from the HDB. This response was regionally selective, as no effect of SSR180711 was seen in mPFC-projecting cells in the dEP.
Both cholinergic and γ-aminobutyric acidergic (GABAergic) neurons in the BF express α7 nAChR messenger RNA (Azam et al. 2003). We therefore used a separate set of animals to phenotypically characterize the HDB neurons that were activated by SSR180711. SSR180711 produced a dose-dependent increase, up to ∼3-fold, in the number c-Fos/ChAT-IR neurons. Contrarily, there was no change in the number of c-Fos/PV-IR or c-Fos/PAG-IR neurons. The PV-positive subtype of GABAergic neurons comprise 90% of the GABAergic neurons projecting from the BF to the mPFC (Gritti et al. 2003). Thus, our results indicate that SSR180711 selectively activates cortically projecting cholinergic, but not GABAergic or glutamatergic, neurons in the HDB. Furthermore, the activation of cholinergic cells was inhibited by the selective α7 nAChR antagonist MLA, confirming the involvement of the α7 nAChR activation in this effect, and was regionally selective, as SSR180711 did not activate cholinergic neurons in other regions of the BF.
The regional selectivity of the response to SSR180711 may be caused by differential expression of α7 nAChRs or different combinations of nAChR subtypes. Functional α7β2 nAChR heteromers have recently been identified in the BF (Liu et al. 2009). This heteromer has longer decay rates, similar to those of α4β2 nAChRs, compared with homomeric α7 nAChRs (Liu et al. 2009). This may lead to a larger influx of Ca2+, which is a major signal for c-Fos expression (Herdegen and Leah 1998). In line with this, α4β2 nAChRs induce c-Fos more potently than α7 nAChRs (Seppa et al. 2001). Therefore, although α7 nAChRs have a high permeability to Ca2+ (Seguela et al. 1993), it is possible that α7β2 nAChRs induce c-Fos more potently than α7 nAChR homomers. Thus, different levels of α7β2 nAChRs may explain the regional selectivity of SSR180711-induced c-Fos-IR.
Cholinergic and GABAergic neurons comprise separate populations in the BF (Gritti et al. 1993), whereas somatic PAG staining overlaps extensively with both cholinergic and GABAergic neurons (Manns et al. 2001). Although, this somatic PAG staining reflects the ability to secrete glutamate locally from soma and dendrites rather than at the terminals (Gritti et al. 2006), this extensive overlap with other cell types precludes us from excluding an effect of SSR180711 on a subset of PAG-positive neurons in the BF. Furthermore, c-Fos is a surrogate marker for neuronal activation, and some neuronal populations that are activated by α7 nAChRs may not express c-Fos (Hoffman and Lyo 2002).
To test the functional importance of SSR180711-induced activation of cortically projecting cholinergic neurons in the HDB, we used a unilateral depletion of cholinergic cells in the HDB with SAP. SAP lesioning resulted in a 50% depletion of cholinergic cells in the HDB and abolished the SSR180711-induced increase in c-Fos-IR in the ipsilateral mPFC. This demonstrates that activation of cholinergic cells in the HDB is necessary for the SSR180711-induced activation of the mPFC.
The majority of BF corticopetal cholinergic terminals do not make identifiable synaptic contacts in the PFC, suggesting a diffuse activation of neurons by acetylcholine (Henny and Jones 2008). Therefore, activation of neurons in the PFC may rely on above-threshold ambient levels of acetylcholine, rather than direct synaptic transmission. This may explain why a partial depletion of cholinergic cells in the HDB is able to completely inhibit the SSR180711-induced increase in c-Fos-IR in the mPFC. However, whether acetylcholine primarily acts at classical synapses or through volume transmission is still debated (Sarter et al. 2009).
Selective lesions of cholinergic neurons in the BF, as well as removal of cholinergic input to the mPFC, using SAP induce deficits of attention and vigilance in rats (McGaughy et al. 1996, 2002; Newman and McGaughy 2008). In addition, second-to-second increases in acetylcholine release in the mPFC correlate with correct responding in a cue-detection task (Parikh et al. 2007). This suggests that the functional integrity of cortically projecting cholinergic BF neurons is important for attentional function. The present results demonstrate that the α7 nAChR agonist SSR180711 potently activates this system, which may underlie the procognitive effects of α7 nAChR agonists in animals. This is supported by the fact that SSR180711-induced release of acetylcholine and induction of c-Fos in the mPFC occurs in the same dose–interval that improves selective attention and working memory (Biton et al. 2007; Hansen et al. 2007; Pichat et al. 2007).
Notably, α7 nAChR agonists activate presynaptic receptors in the PFC (Rousseau et al. 2005; Dickinson et al. 2008), and α7 nAChRs are also present on cholinergic terminals in the mPFC (Sugaya et al. 1991; Duffy et al. 2009), as well as on somatodendrites in the BF (Liu et al. 2009). We show that direct injection of MLA into the HDB attenuates the increase in c-Fos-IR in the mPFC induced by systemic administration of SSR180711, suggesting that activation of α7 nAChRs in the HDB is partly responsible for this effect. Supporting this, we found that direct injection of SSR180711 into the HDB was sufficient to induce c-Fos-IR in the ipsilateral mPFC. The lower degree of inhibition by MLA after intra-HDB compared with systemic injection may be explained by incomplete diffusion throughout the HDB and an uneven concentration gradient from the site of injection. However, we cannot exclude that activation of presynaptic α7 nAChRs on BF cholinergic or thalamic glutamatergic nerve terminals in the mPFC contributes to the SSR180711-induced activation of the mPFC, such as that has been proposed for α4β2 nAChRs (Parikh et al. 2008). However, we have established the cholinergic cells in the HDB as an important target for α7 nAChR activation that likely mediates procognitive effects.
In addition to its importance for reward and reinforcement learning, the VTA has profound influence on cognitive performance (Tzschentke 2001), and blockade of dopamine D2 receptors in the striatum is critical for the effect of antipsychotic drugs (Kapur and Mamo 2003).
A unilateral lesion of the VTA with 6-OHDA abolished the SSR180711-induced increase in c-Fos-IR in the ipsilateral ACCshell. The VTA also projects to the PFC (Albanese and Minciacchi 1983) and to cholinergic neurons in the BF (Gaykema and Zaborszky 1996). However, the 6-OHDA lesion did not affect SSR180711-induced activation in these regions. Presynaptic α7 nAChRs on glutamatergic terminals in the VTA can activate dopaminergic neurons in this region (Schilstrom et al. 1998; Mansvelder and McGehee 2000; Schilstrom, Fagerquist, et al. 2000), and our results suggest that these neurons are necessary for SSR180711-induced activation of the ACCshell but not the mPFC or HDB. In accordance with this, nicotine-induced dopamine release in the nucleus accumbens is blocked by local injection of MLA in the VTA (Schilstrom et al. 1998). We therefore suggest that the induction of c-Fos-IR in the ACCshell is mediated by activation of α7 nAChRs on glutamatergic terminals in the VTA, which in turn stimulates dopamine release in the ACCshell.
SSR180711 has been shown to promote dopamine release in the PFC (Pichat et al. 2007; Tietje et al. 2008), although others have failed to reproduce this finding (Beyer et al. 2008). In addition, local injection of nicotine in the VTA does not induce c-Fos-IR in the mPFC (Schilstrom, De Villiers, et al. 2000). Our results suggest that a putative release of dopamine in the mPFC following α7 nAChR activation is not sufficient to cause neuronal activation in this region.
The ACCshell is consistently activated by acute treatment with both atypical and typical antipsychotic drugs, and this is believed to correlate with their antipsychotic effect (Arnt and Skarsfeldt 1998). Thus, the activation of the ACCshell by SSR180711 may suggest a potential antipsychotic effect of α7 nAChR activation (Hansen et al. 2007). This is supported by a recent study, showing that SSR180711 potentiates latent inhibition and reverses amphetamine-induced disruption of latent inhibition, 2 models considered to be predictive of activity against the positive symptoms of schizophrenia (Barak et al. 2009). In addition, a recent clinical trial with the weak α7 nAChR partial agonist GTS-21/DMXB-A has reported near-significant improvements in the brief psychiatric rating scale (Freedman et al. 2008), and Deutsch et al. (2008) reported a reduction in the positive and negative syndrome scale score in a small open-label study using cytidine-diphospho-choline in combination with galantamine. Notably, in both studies, α7 nAChR agonism was provided as an add-on to existing antipsychotic treatment, which may lead to an underestimation of the antipsychotic potential of α7 nAChR activation.
The results presented here suggest that 2 distinct neural pathways are activated by SSR180711, involving cholinergic-dependent activation of the mPFC and dopaminergic-dependent activation of the ACCshell, respectively. The BF and mPFC are important for attentional function (Sarter et al. 1999, 2005; Birrell and Brown 2000; Dalley et al. 2004), which is the cognitive domain predominantly affected by α7 nAChR agonists in clinical studies (Olincy et al. 2006; Freedman et al. 2008). Thus, activation of the HDB and mPFC may subserve the procognitive effects of α7 nAChR agonists, whereas activation of the ACCshell may confer beneficial effects on the positive symptoms of schizophrenia. Thus, these systems likely mediate different behavioral effects of α7 nAChR activation.
Danish Ministry of Science, Technology and Innovation (M.S.T.); Lundbeck Foundation; Doctor Sophus Carl Emil Friis and wife Olga Doris Friis (A.H.-S.)
The authors would like to thank Pia Rovsing Sandholm, Tine Ostergaard, Charlotte Holtoft, Charlotte Landholt Petersen, and Mia Billenberg Nielsen for skillful technical assistance and Helle Prudinsky for artwork assistance. Conflict of Interest: The authors declare that, except for income received from primary employers, no financial support or compensation has been received from any individual or corporate entity over the past 3 years for research or professional service, and there are no personal financial holdings that could be perceived as constituting a potential conflict of interest.
- antipsychotic agents
- cholinergic agents
- diagonal band of broca
- neural pathways
- nucleus accumbens
- prefrontal cortex
- nicotinic receptors
- ventral tegmental area
- normal saline
- cognitive ability
- dopaminergic neuron
- basal forebrain
- c-fos genes
- cholinergic neuron
- saline solution