Pet-1 is required across different stages of life to regulate serotonergic function

Abstract

Transcriptional cascades are required for the specification of serotonin (5-HT) neurons and behaviors modulated by 5-HT. Several cascade factors are expressed throughout the lifespan, which suggests that their control of behavior might not be temporally restricted to programming normal numbers of 5-HT neurons. We used new mouse conditional targeting approaches to investigate the ongoing requirements for Pet-1 (also called Fev), a cascade factor that is required for the initiation of 5-HT synthesis, but whose expression persists into adulthood. We found that Pet-1 was required after the generation of 5-HT neurons for multiple steps in 5-HT neuron maturation, including axonal innervation of the somatosensory cortex, expression of appropriate firing properties, and the expression of the Htr1a and Htr1b autoreceptors. Pet-1 was still required in adult 5-HT neurons to preserve normal anxiety-related behaviors through direct autoregulated control of serotonergic gene expression. These findings indicate that Pet-1 is required across the lifespan of the mouse and that behavioral pathogenesis can result from both developmental and adult-onset alterations in serotonergic transcription.

Figure 1

Conditional deletion of Pet-1 after 5-HT neuron fate specification. (a) Targeting strategy. From the top, schematic of the Pet-1 mRNA; wildtype Pet-1 allele (+); floxed Pet-1 allele (fl); and the conditionally deleted Pet-1 allele (Δ, bottom). (b–d) In situ hybridization to detect Pet-1 transcripts in the dorsal raphe (DRN) of mice heterozygous for the Pet-1 null allele and either the wild-type (b), floxed (c) or conditionally deleted Pet-1 alleles (d). (e) Time frame of Pet-1 expression in Pet-1 ^/− , Pet-1^eCKO (Pet-1^fl/−, ePet::Cre), and wildtype mice. (f–i) In situ hybridization to detect Pet-1 and Gata3 mRNAs in control (Pet-1^fl/+, ePet::Cre) and Pet-1^eCKO mice. (j–l) 5-HT immunostaining in control, Pet-1^eCKO, and Pet-1^−/−mice at E11.5. Scale bars are 100 μm in (i, l) and 200 μm in (d).

Figure 2

Continued Pet-1 function is required for maturation of serotonergic axonal innervation patterns. (a) Co-immunostaining of Yfp and Tph in adult Pet-1^eCKO mice; white arrows indicate untargeted 5-HT neurons; arrowheads indicate proximal axons extending from cell bodies of Pet-1 deficient 5-HT neurons. (b) Yfp immunostaining of Pet-1 deficient 5-HT neuron axon bundles at E14.5 in Pet-1^eCKO mice. Arrowheads mark axons that have crossed the MHB and entered the midbrain (dashed lines, midbrain-hindbrain boundary, MHB). (c) Schematic of the retrograde tracing experiment. Tracer injected into the somatosensory cortex labeled (d) Yfp⁺ Pet-1 deficient 5-HT neurons (e) in the dorsal raphe. (f), merge of (d) and (e). (g–l) Significantly fewer retrogradely labeled cells were found in the DRN of the Pet-1^eCKO brain (g, h; 49.9±3.1%, mean ± s.e.m, relative to control, n=6 for each genotype). Overlay of tracer signal with Yfp immunostaining (i, j) showed that 83.0±1.8%, mean ± s.e.m, of the retrogradely labeled DRN cells in control mice were ePet::Yfp⁺ 5-HT neurons (k), whereas only 19.3±1.7% Yfp⁺ Pet-1 deficient 5-HT neurons (l) were labeled by the same tracer injection in Pet-1^eCKO brain. *** p<0.001, two tailed t test. Scale bars are 20 μm in (a, f) and 200 μm in (l).

Figure 3

Continued Pet-1 function is required for 5-HT neuron firing properties and inhibitory autoreceptor function. (a, b) Whole cell current clamp recordings measuring spontaneous firing of Yfp⁺ neurons with indicated genotypes (c) Quantification of firing frequencies in (a, b), +/+, n=12; Pet-1^eCKO, n=19; * p<0.05, two tailed t-test. (d, e) Whole cell voltage clamp recordings measuring current changes of Yfp⁺ neurons induced with either 1 or 10 μM of 5-HT_1A agonist 8-OH-DPAT. A ramp voltage was applied at 200 mV/s. The intersection voltage, -87 mV, of the control and 8-OH-DPAT trace in (d) was close to the estimated K⁺ equilibrium potential (-99 mV), considering that recordings were not corrected for the liquid junction potential of around 10 mV. (f) Quantification of current changes at -110 mV in (d, e); * p<0.05, *** P<0.001; two tailed t-test (g–j) In situ hybridization of 5-HT_1A (g, h) and 5-HT_1B (i, j) in control and Pet-1^eCKO mice. Scale bar is 200 μm. Error bar is mean ± s.e.m.

Figure 4

Continued Gata3 expression is needed to maintain 5-HT gene expression but not autoreceptor function. (a) Yfp Immunostaining in adult DRN. (b) Tph immunostaining. (c) In situ hybridization of Pet-1 mRNA. (d)Counts of Tph⁺ cell bodies in control versus Gata3^eCKO mice in individual adult B nuclei, n=3 for each genotype. (e) RT-qPCR of _eCKO Aadc, Sert, Tph2, Vmat2, and 5-HT_1A mRNAs in control versus Gata3 mice, control n=7, normalized to 100%; Gata3^eCKO n=11, * p<0.05, ** p<0.01, two tailed t test. (f, g) HPLC analysis of 5-HT (f) and 5-HIAA (g) levels in forebrain and spinal cord of control (n=7) and Gata3^eCKO (n=5) mice. ** p<0.01, *** p<0.001, two tailed t test. (h) Whole cell current clamp recordings of spontaneous firing of R26R-Yfp⁺ Gata3 deficient cells. (i) Whole cell voltage clamp recordings of current changes in response to 5-HT_1A agonist, 8-OH-DPAT, in R26R-Yfp⁺ Gata3 deficient cells. Scale bars are 200 μm. Error bars represent s.e.m except for s.d in (d).

Figure 5

Stage-specific disruption of Pet-1 in the adult ascending 5-HT system. (a) Adult stage specific deletion of Pet-1 in Pet-1^aCKO mice. (b, c) Co-immunostaining of βgal and Tph in adult dorsal (b) and medullary (c) raphe. (d, e) Percentage of TPH⁺ cells expressing CreER activated βgal in individual adult B nuclei (d) and in 5-HT neurons of ascending versus descending pathways (e, 68.5±9.5% in the pons (B4-B9) and 12.1±6.8% in medullary nuclei (B1-B3). n=7, mean ± s.d, *** p<0.001, two tailed t test). (f–i) In situ hybridization of Pet-1 mRNA in coronal sections of adult Pet-1^aCKO mice treated witheither TM or vehicle (Veh). (j) RT-qPCR of Pet-1 mRNA in TM-treated adult Pet-1^aCKO mice (pons, n=30, 28.0±2.5% relative to control, n=35; medulla, n=12, 105.9±5.8% relative to the control, n=15). Each dot represents a sample from the indicated group, mean ± s.e.m, two tailed t test. Scale bars are 200 μm.

Figure 6

Disruption of Pet-1-dependent transcription in the adult ascending 5-HT system causes elevated anxiety-like behavior. Six to eight-week-old male Pet-1^aCKO mice (n=12) and their littermate controls (Pet-1^fl/−, n=12) were treated with TM for 5 consecutive days and then acclimated for another 4 weeks before behavioral testing. (a–e) Elevated plus maze, * p<0.05, ** p<0.01, *** p<0.001; two tailed t test. (f–h) Dark ↔ light exploration, * p<0.05, ** p<0.01; two tailed t test. (i, j) Open field, * p<0.05; two tailed t test. Error bars represent s.e.m.

Figure 7

5-HT synthesis and Sert expression is maintained in the adult ascending 5-HT system through positively autoregulated direct Pet-1 transactivation. (a, b) HPLC analysis of 5-HT and 5-HIAA levels in the forebrain of TM treated control (n=7) and Pet-1^aCKO (n=7) mice, *** p<0.001, two tailed t test. (c, d) Western blotting analysis of Tph protein in DRN of TM treated control (n=8) and Pet-1^aCKO mice (n=7, 50.3±4.0% relative to the control), ** p<0.01, two tailed test. (e) RT-qPCR analysis of serotonergic gene expression in Pet-1^aCKO or control mice either 5 days (control, n=11; Pet-1^aCKO, n=11) or 30 days (control n=14; Pet-1^aCKO, n=16) after TM or vehicle treatments, * p<0.05, ** p<0.01, *** p<0.001. (f–i) In situ hybridization to detect Tph2, Sert, Maob, and Lmx1b mRNAs in coronal sections of Pet-1^aCKO mice treated with either TM or vehicle. (j) Whole cell current clamp recordings of spontaneous firing of Yfp⁺ Pet-1 deficient cells in TM treated Pet-1^aCKO mice. (k) Whole cell voltage clamp recordings of current changes in response to 5-HT_1A agonist, 8-OH-DPAT, in Yfp⁺ Pet-1 deficient cells. (l) RT-qPCR analysis of chromatin immunoprecipitations. Values represent fold enrichment in binding to the indicated regions as compared to negative control region (Untr17). Untr, untranscribed genomic region; #, p<0.0001 for Tph2, Sert, Pet-1 versus Untr17 or Sert-intron, one-way ANOVA with Bonferroni's Multiple Comparison Test. (m, n) In situ hybridization to detect CreER^T2 mRNA in adult Pet-1^aCKO mice treated with either TM or vehicle. Scale bars are 200 μm. Error bars represent s.e.m except for s.d in (l).