Pmch

Spatial tier: RESTRICTED · Class: OTHER · Top region: PSTN (Parasubthalamic nucleus) · Positive regions: 15/554

48.9Discovery Score
40.9Targetability Score
CEvidence grade C
What this means: Pmch: fine CCF profile is restricted toward PSTN (top regions: PSTN, TMv, LHA, STN, ME; 15 positive regions); direct spatial validation is still required.

Function in PSTN: evidence, prediction, innovation

Gene function
Pmch encodes pro-melanin-concentrating hormone, precursor of the 19-amino-acid cyclic peptide MCH plus NEI and NGE. MCH signals through the Gi/Go-coupled receptor MCHR1 (rodents lack functional MCHR2), inhibiting cAMP and mobilising Ca2+. It is made by a large lateral hypothalamic/zona incerta population and is a canonical orexigenic and REM-sleep-promoting peptide.
Region function
The parasubthalamic nucleus (PSTN) is a small nucleus at the lateral edge of the posterior hypothalamus whose Tac1- and CRH-expressing neurons are meal- and threat-activated and suppress appetite, delay feeding initiation and mediate fear-induced anorexia.
Published in this region
expression only No region-specific literature found for Pmch acting within PSTN. The closest anatomical result is Chometton et al. 2016, defining a premammillary lateral hypothalamic complex that contains PSTN and the calbindin nucleus embedded in MCH-rich posterior LHA territory and responds to palatable food. The two literatures run in parallel but have never been joined: PSTN-Tac1 neurons suppress appetite (Kim et al. 2022) and mediate fear-induced feeding suppression (Nagashima et al. 2022), while MCH neurons promote feeding, reward and REM sleep (Vetrivelan et al. 2016).
Predicted function
Prediction: MCH released by surrounding LHA/incerto-hypothalamic fibres acts on MCHR1 to provide Gi-mediated inhibition of anorexigenic PSTN-Tac1/CRH neurons, forming a local push-pull node where hunger-state MCH tone releases the PSTN brake on feeding. Intra-PSTN MCH or an MCHR1 agonist should increase intake and shorten latency to feed, whereas the MCHR1 antagonists SNAP-94847 or GW803430 delivered into PSTN should exaggerate meal-induced and fear-induced anorexia.
Innovation
★★★☆☆ 3/5 Both MCH biology and PSTN feeding control are individually well developed with ready tools, so only the specific intersection is untested.
Tools
Pmch-Cre BAC and Pmch-CreER lines (note the Cre/Pmch dissociation reported by Beekly et al. 2020), Pmch-null mice, MCH antisera; MCHR1 antagonists SNAP-94847 and GW803430; MCHR1-knockout mice. For the target nucleus: Tac1-Cre and Crh-Cre PSTN drivers with Cre-dependent hM3Dq/hM4Di and caspase AAVs.
References

Direct evidence located at the region

Pmch Allen ISH at the CCF centroid of PSTN

Allen ISH experiment 55, section image 101306490 (section 217): the section that contains the CCF centroid of Parasubthalamic nucleus according to the Allen image-to-atlas alignment (reference_to_image), with a 2 mm window around that point. Left: whole section, red box = window. Middle: ISH signal. Right: Allen's expression-mask view of the same window. open this image in the Allen viewer

Direct spatial evidence

MERFISH REGION LOCATOR
PSTN MERFISH region locator
Real single-cell MERFISH coordinates for the PSTN (Parasubthalamic nucleus) section -- our own ground-truth data, showing exactly where this region sits, independent of whether the Allen ISH section below is easy to read.

Regional expression figure

Pmch regional expression profile

Predicted WMB × fine-CCF profile; not direct full-transcriptome spatial measurement.

Spatial profile

PSTN
0.911
TMv
0.806
LHA
0.696
STN
0.684
ME
0.639
TU
0.627
PeF
0.513
DMH
0.485
VMH
0.463
ARH
0.407
PVHd
0.391
PH
0.328
PMv
0.289
SO
0.247
RCH
0.234
PVi
0.227
AHN
0.175
LPO
0.164
FF
0.16
PVH
0.153

Evidence and specificity

Evidence status
PREDICTED_FINE
Direct sources
None; predicted localization only
Exclusive threshold
12 positive regions out of 554
Spatial specificity
0.406
Top-region fraction
0.087
Filter status
PASS

Interpretation limits