Oh 2020

https://www.nature.com/articles/s41380-019-0384-6


 * Article
 * Published: 05 March 2019

= Hippocampal mossy cell involvement in behavioral and neurogenic responses to chronic antidepressant treatment =


 * Seo-Jin Oh,
 * Jia Cheng,
 * Jin-Hyeok Jang,
 * Jeffrey Arace,
 * Minseok Jeong,
 * Chang-Hoon Shin,
 * Jeongrak Park,
 * Junghee Jin,
 * Paul Greengard &
 * Yong-Seok Oh

Molecular Psychiatry volume 25, pages1215–1228(2020)Cite this article

Slow antidepressant actiion, immediate elevation of serotonin. suggests slow adaptive changes - maybe in multiple neuronal subtypes and neural circuits over time

Mossy cells: excitatory neurons, in the hilus, regulate DG

Neuronal activity of mossy cells increases with chronic SSI tx.

Chronic SSRI tx doesn't have behavioral or neurogenic effects when mossy cells are knocked out.

Activating Mossy cells by dreads increases proliferation/survival of neural stem cells.

Di dread impairs behavioral/neurogenic response o SSRIs

Mossy cells: in hilus, make connections with GCs and basket cells. integrate input from granule cells (multiple), CA3 pyramidal cells, and local GABAergic interneurons (basket cells, hilar perforant path-associated cells, propagate signal over llongitudinal axis of hippocampus.

Mossy cells are highly. enriched in monoamine receptors, type II glucocorticoid receptors (sense cortisol). Fire frequently (unlike GCs which fire sparsely).

Essential for proper HC function https://www.nature.com/articles/s41380-019-0384-6#ref-CR22

MC degeneration leads to dysregulation of GC excitation, leading to elevated anxiety and impaired pattern separation https://www.nature.com/articles/s41380-019-0384-6#ref-CR23.

p11 (S100A10): regulates affect, determines response to antidepressants

downregulated in depressed humans and rodents

induced by chronic but not actute amtidepresant tx in rodents

global abolition of 11 -> depressive state, removal of p11 from some neurons produces different kinds of depressive effects.

p11 forms a heterotetrameric complex with annexin A2 (AnxA2) in the DG upon SSRI tx. Also with SMARCA3. Crucial for behavioral and neurogenic responses to chronic antidepressant tx Chronic SSRI administration: increased trimeric complex (P11, AnxA2, SMARCA3). P11 and SMARCA are enriched in hilar mossy cells, basket cells. Cell type specific role of complex in antidepressant action not yet explored.

[p11 is a member of the S100 protein family. SMARCA3 is a helical transcrition factor of the SWI/SNF chromatin remodeling family. Annexin A2 is a calcium-relgulated membrane-binding protein)

Main results: complex of p11,/AnxA2, SMARCA3 regulates mossy cell activity.

RESULTS

Genetic deletion of p11 of Smarca3 in hipocampal mossy cells

Previously known: SMARCA3 is a binding parther of p11/Anxa2 complex, SMARCA has a crucial role in neurogenic and behavioral responses to chronic antidepressants. Also known that p11 and smarca3 are enriched in hippocampal mossy cells.

Main development is knocking out p11 or smarca specifically in mossy cells using cre driver line.

Dopamine D2 receptor gene is limited to the DG, validating in Fig 1a, 1b. Crossed with p11 or smarca3 flox lines to get conditional deletion (1c, 1f). After chronic SSRI tx, used NSF and found that latency to approach was decreased in control, but not p11 cKO or smarca3 cKO lines (demonstrating necessity of expression in mossy cells for antidepressant effet). Also validated with genetic deletion of smarca 3 in another mc specific cre line, with both hilar and DG and some pyramidal cell activity. consistent effect.

Fig 1: Selective deletion of p11 or Smarca3 in DG Mossy cells affects antidepressant response, suggesting the p11 and Smarca 3 in mossy cells are involved in mediating the behavioral responses to chronic tx.

2. Mossy cell-specific inhibition of p11/anxa2/smarca3 complex and neurogenic, and behavioral responses to chronic antidepressant tx.

To be sure it's not due to off target effects/developmental effects in cKO mice, made a model to test what happens when the trimer is disrupted.

Developed a recombinant construct to inhibit the assemply of the p11/Anxa2/SMARCA3 complex. Using a short peptide that occupies the binding site of the P11/ANXA2 complex, preventinng assembly of the P11/ANXA2/SMARCA3 trimer. (1a)

hypothesized that the peptide competes with smarca3, blocking the ternary complex (2b). In-vitro co=IP confimed that the Inhibitory peptide-AcGFP construct (ASIP-AcGFP1) blocked assembly of the P11/ANXA2/SMARCA3 complex (2c).

Cre-dependent aavs DELIVERED THE INHIBITORY or control constructs to hippocampal mossy cells via the hilus in MC-Cre mice [WHY this line?] (2d) Confirmed localization (S3C) ensured hippocampal specificity.

Fusion proteins were coparably expressed in control and inhibitory proteins, both in mossy cells but not in other GABA-ergic populations (2e) ensuring cell-type specificity.

After delivering AAV construct, examined neurogenic and behavioral outcome of chronic administration of ssri [dID THEY ever check if the construct remained inhibited? did they have a no-construct/sham control to make sure the constuct wasn't having deleterious effects]

2G/H: Fluoxetine increases brdU+ cells in the DG only in the mice with the ability to form the p11/Anxa/SMARCA3 complex. Delivering the inhibitory peptide abolishes the effect of fluoxetine.

2I/J. The same pattern is seen for DCX expression.

2K/I. The complex is necessary for a decreased latency to feed, but no difference is seen in food intake.

Anxiety-related behaviors were inconsistent. no difference in OFT (S5A) There was a difference in the light/dark box tezt [what is this?] (S5c), but not in the EPM(S5d) or thigmotaxis test [ANOTHER NAME FOR OFT, time in dark] (S5b).

Sig difference in TST between saline and fluoxetine treated groups whe given the control virus, but the difference was attenduated by the inhibitory virus. (S5e)

k/l: latency to feed was decreased in control mice, but home cage feeding did not change, suggesting the change was not due to different hunger levels.

Fig 3:effects of celltype specific inhbition of the complex on neuronal activity of mossy cells

mossy cells are normally spontaneous firing neurons in the hilus, and their activity modulates neuronal circuitry and function in the dg.

lABELED MOSSY CELLS WITH A FLUORESCENT PROTEIN (AcGFP1) by njecting with Cre-dependent AcGFP1 AAVs into MC-Cre mice (3a). AcGFP1 was shown to be restrictively expressed in MCs (3b) [why didn't they use staining for validation? maybe not enough mice?]

Mossy cells were conformed by morphology, tonic firing pattern, and membrane capacity (50-100 pF). Chronic fluoxetine increases spontaneous firing rate of MCs (8wo mice given 18 days of fluoxetine orally) (3C, 3D), but acute treatment (one dose of fluoxetine, injected- unclear how long before tracing) has no effect on activity (3e, f) [doesn't this contrast with the idea that mossy cells have receptors for serotonin?)]

Previous work shows taht acute restraint stress decreases c-fos staining in hilar mossy cells in the hippocampus. Chronic stress (S6a) showed decreased cFos in mossy cells. Chronic fluosetine increased cFos activity in mossy cells (SD,E,F), consistent with electrophysiology (3C, D). Neuronal activity of mossy cells is suppressed by chronic strss and is reversible by chronic fluoxetine administration

P11 CKO decreases firing of mossy cells (3g, h), Smarca cKO decreases firing rate of mossy cells (3I, J) compared to controls.

Previous studies show that p11/AnxA2/SMARCA3 comples induced in mossy cells by chronic fluoxetine, targeted to nucleus to regulate gene transcription.

Generated an inhibitory construct specifically tarrgeted to the nucleus by localization signals fused to teh C-terminus. Generated three Cre-dependent vectors: a control, an inhibitory vector (as usedb efore), and an inhibitory vector with the nuclear localization signal (S7A)

Ioinfection into cultured HEK293 cells or stereoinjection into DG of MC-Cre mice, verified nuclear specific expression of NLS inhibitory construct (S7B) and cell-type specificity for MCs.(F3L)

Firing rate of MCs decreases with injection of inhibitory or inhibitory-NLS compared to control (3M/3N). Further, c-Fos immunoreactivity is reduced by PASIP-AcGFP expression in mossy cells (S8a, Sc)- validation of firing.

Fig 3 summary:neural activity of mossy cells decreases during chronic stress-induced depression, reversed by antidepressant administration, possibly through the p11/anxa2/smarca3 complex.

4. Effects of MC stimulation on adult neurogenesis.

used viral vectrs expressing mCherry (control) or mCherry-hm3D-Gq into the hipus of MC-Cre transgenic mice (4A). c-fos staining confirms CNO-dependent activation (2h injection before sack) of mcs (4b).

Gq-DREADD stimulation of MCs, 2 hours prior to behavior, and one day prior to perfusion increased brdu+ PROLIFERATIN CELLS AND DCX (post-mitotic immature neurons) immunostaining in SGZ (4c-g). MC-dependent regulation of neurogenesis was consistent through rostrocaudal axis (S10a-S10d). [how is 1 day enough to increase DCX+ cells?]

Direct stimulation of mossy cells increased duration of time in open arms of EPM (S11d). Failed to induce significant differences in other anxiety-related behaviors (S11a-c). [Does noting the ones that didn't work make this stronger or weaker?]

Acute stimulation with CNO injection (2h) did not change NSF and TST. Acute stimulation of MCs triggers neurogenesis, but is not sufficient to achieve full improvement of anxiety and depressive-like behaviors.

[how does this fit with newborn neurons being sufficient to improve behavior? are DCX+ cells just too early? Maybe waiting a few more days would change this?]

Effect of slective inhibition of mossy cells on antidepressant actions in the hippocampus.

With cre-dependent AAVs to inject mCherry (control) and hM4Di-Gi-mCherry (Gi Dreadd) in MC-Cre mice, verified that MCs were silenced by CNO in mice expressing the Gi dreadd (5a, b)

CNO silencing of mossy cells in Gi dreadd aminals was confirmed by electrophysiology (5c, d).

Chronic SSRI tx increased BRDU cells in the sgz in control effect, but was significantly reduced Gi Dreadd mice.(5f, g). DCX intensity was increased by fluxetine treatment in control mice, but to a lower extent in Gi mice (5H, I). Chronic fluozetine promotes neural progenitory survival and differentiation. Inhibiting MC activity results in significant reduction in survival/differentiation of post-mitotic nwborn cells (5h, i) [this doesn't tell us anything about differentiation... no follow up on different markers]

Mossy cells may thus modulate chronic fluoxetine-induced proliferation of neural progenitorys, differentiation, and maturation into new-born granule cells. Consistent with KO/Inhibition of the p11/anxa2/smarca3 complex.

chronic fluoxetine-induced reduced latency to feed is abolished in Gi-dreadd mice (5j). Home cage feedling is unaffected, suggesting difference in hunger levels is unlikely (5k). In TST, effect of fluoxetine is seen in control mice but not in Gi-dreadd mice (s14e). MC activity is critical for neurogenic and behavioral responses to chronic fluoxetine administration.

Main take-aways

However, acute chemogenetic stimulation of mossy cells is unlikely to be sufficient to achieve full behavioral improvement as seen by chronic SSRI administration. Behavioral benefit after chronic SSRI administration may require cooperative neuroplastic changes in multiple neural cell types that include hippocampal mossy cells.

It has been unknown whether mossy cells are involved in adult neurogenesis, although a previous study suggested that mossy cells provide the first excitatory input to adult-born granule cells [60]. The current study is the first to show that mossy cells influence SSRI-induced neurogenesis

Several genetic and pharmacological approaches have demonstrated that 5-HT1A and 5-HT4 receptors, enriched in dentate granule cells, are crucial for SSRI-induced neurogenesis [61,62,63]. In contrast, the basal maintenance level of hippocampal neurogenesis is not altered in either 5-HT1A receptor-deficient or 5-HT4 receptor-deficient mice, which indicates the specific involvement of these serotonin pathways in SSRI-induced neurogenesis. Here we observed that the p11/AnxA2/SMARCA3 complex in mossy cells regulates SSRI-induced neurogenesis, but not the basal level. Given that mossy cells and granule cells make extensive reciprocal connections to construct the neural circuitry of the dentate gyrus, it is plausible that serotonergic regulation of both granule cells and mossy cells may cooperate to induce the neurogenesis and further behavioral response to chronic SSRI administration.

The present study contributes to a better understanding of the neural mechanisms mediating the beneficial effects of SSRI medication