Skip to main content
NCBI home page

This is a preprint.

It has not yet been peer reviewed by a journal.

The National Library of Medicine is running a pilot to include preprints that result from research funded by NIH in PMC and PubMed.

bioRxiv logoLink to bioRxiv
[Preprint]. 2025 Aug 1:2025.07.30.667705. [Version 1] doi: 10.1101/2025.07.30.667705

Sleep pressure is differentially regulated by molecularly distinct subtypes of Lhx6-positive and Lhx6-negative neurons of the zona incerta

Parris Washington Chandler 1, Sang Soo Lee 2, Leighton H Duncan 1, Dong Won Kim 1,6, Mark N Wu 1,2, Seth Blackshaw 1,2,3,4,5,*
PMCID: PMC12324396  PMID: 40766484

Abstract

Sleep pressure, the accumulating drive towards sleep during wakefulness, is shaped by Lhx6-positive GABAergic neurons in the zona incerta (ZI). Here, we show that these neurons are broadly activated both by spontaneous and experimentally-elevated sleep pressure, and remain active for hours into recovery sleep. Activation is stronger in anterior ZI and varies across molecularly defined subgroups: Nkx2-2-positive neurons respond vigorously, whereas Calb2-positive neurons respond weakly. We also identify Lhx6-negative/Slc32a1-positive GABAergic ZI neurons with distinct sleep pressure responses. Selective genetic ablation of Nkx2-2 in Lhx6-positive neurons reduces the number of Lhx6-positive neurons, alters their distribution, blunts sleep pressure-induced activation of both Lhx6-positive and negative cells, and increases sleep duration. These findings show that developmentally specified, molecularly heterogeneous Lhx6-positive ZI neurons form a key hub for regulating sleep homeostasis, and offer new insight into the circuitry that controls sleep pressure.

Introduction:

Sleep is an evolutionarily-conserved state that is essential for the survival of all organisms examined to date (Keene and Duboue 2018; Freiberg 2020; Miyazaki, Liu, and Hayashi 2017). In recent years, multiple neuronal subtypes that rapidly regulate bistable sleep-wake transitions have been identified (Eban-Rothschild, Appelbaum, and de Lecea 2018; Xu et al. 2015; Yu et al. 2019; Vanini and Torterolo 2021). The drive to sleep is controlled by an activity/time-dependent component referred to as sleep pressure, that progressively accumulates during wakefulness and dissipates during sleep, as well as by circadian factors (Skeldon and Dijk 2025; Borbély 2022). However, the molecular basis of this activity/time-dependent drive, and thus of sleep homeostasis, remains unknown (Thomas et al. 2020; Duhart, Inami, and Koh 2023).

Recent evidence from both Drosophila and mammals has suggested that activity-dependent changes in rapid-acting wake and sleep-promoting neurons may drive the gradual accumulation of sleep pressure (Sawada et al. 2024; Thomas et al. 2020). However, studies in mice have identified discrete neuronal subpopulations that are activated selectively in response to both natural or induced changes in sleep pressure, and are distinct from those controlling rapid sleep-wake transitions. One such neuronal population is Lhx6-positive GABAergic neurons of the zona incerta (ZI) (Liu et al. 2017a; Lee et al. 2025; Kim et al. 2021a), a small yet highly complex region that broadly regulates sensory integration, innate behaviors, and both motivational and emotional states (Mitrofanis 2005; Venkataraman et al. 2019; Li, Rizzi, and Tan 2021; Wang et al. 2020; Monosov et al. 2022; X. Zhang and van den Pol 2017). Although these neurons are immediately presynaptic to several arousal-promoting cell types, their chemogenetic activation induces sleep with notably delayed and prolonged kinetics, ranging from 2-8 hours following CNO administration (Liu et al. 2017a). Recent work has shown that glutamatergic neurons of the thalamic reuniens nucleus (ReN), which are both selectively responsive to induced sleep deprivation and project to anterior Lhx6-positive ZI neurons, promote sleep with similar delayed kinetics (Lee et al. 2025). Elevated sleep pressure both structurally and functionally strengthens the synaptic connection between ReN and Lhx6-positive ZI neurons, and Lhx6-positive neurons are essential for the sleep-promoting function of ReN neurons (Lee et al. 2025). Together, these findings suggest that this ReN-ZI connection forms a dedicated neural circuit for sensing and conveying sleep pressure, distinct from the circuitry that mediate rapid bistable sleep-wake transitions (Han et al. 2014; Kroeger et al. 2018; Kashiwagi et al. 2024).

This raises the question of exactly how sleep pressure is regulated by Lhx6-positive ZI neurons. Previous studies indicate substantial functional heterogeneity within this population. Using Fos expression as a readout for activity, at least 25% of Lhx6-positive ZI neurons are active even in the early evening, when sleep pressure is the lowest, while many remain inactive even under high levels of experimentally-induced sleep pressure (Liu et al. 2017a). These neurons also remain active even during early stages of recovery sleep (Liu et al. 2017a). However, it is not known whether the same subpopulations respond to natural versus induced sleep pressure, or during recovery sleep.. Other studies have demonstrated substantial molecular heterogeneity among these cells (Kim et al. 2021a, 2020, 2025). This raises the possibility that molecularly distinct Lhx6-positive ZI subtypes may serve different functions in sleep homeostasis.

In this study, we systematically applied multiplexed single-molecule fISH (smfISH) analysis for Lhx6, Fos, and molecular markers that define distinct subtypes of Lhx6-positive ZI neurons. We confirm that these neurons are broadly activated by both naturally occurring and sleep deprivation-induced increases in sleep pressure, and remain strongly activated for at least three hours after the onset of recovery sleep. Anterior Lhx6 mRNA-positive neurons show relatively stronger sleep pressure-induced activation, broadly coinciding with patterns of ReN innervation. Distinct patterns of Fos mRNA induction were observed in molecularly distinct subpopulations of Lhx6-positive ZI neurons, with Nkx2-2, Nfia, and Calb1 mRNA-positive cells showing the strongest responses to sleep pressure, whereas Calb2-positive cells respond weakly. We also identified subpopulations of Lhx6-negative, Slc32a1-positive GABAergic ZI neurons that were differentially activated by sleep pressure. Finally, using intersectional genetics, we show that loss of function of the homeodomain transcription factor Nkx2-2 markedly reduces the number of Lhx6-positive neurons, redistributes them along the anterior-posterior axis, diminishes their sleep-pressure-induced activity, and increases both daytime and nighttime sleep. These results indicate that Lhx6-positive ZI neurons play a central but complex role in regulating sleep homeostasis.

Results:

Multiplexed smfISH analysis of Lhx6-positive ZI neurons

To map spatial, temporal, and subtype-dependent activation of Lhx6-positive ZI neurons, we used the HiPlex platform (ACD Bio) to simultaneously profile up to 12 different genes, under a range of both naturally occurring and induced elevated sleep pressure. Core probes targeted Lhx6 and Fos, as well as Slc32a1, which broadly labels GABAergic neurons in the ZI (Watson, Lind, and Thomas 2014; Li, Rizzi, and Tan 2021). Additional probes marked subtypes previously defined in the ZI (Kim et al. 2021a; Watson, Lind, and Thomas 2014; Zhao et al. 2019; M. Zhang et al. 2023; Mitrofanis 2005): Nkx2-2, Calb1, Calb2, Nfia, Cck, Penk, Pvalb, Gal, Nos1, and Pnoc. Slc17a6 was used in some studies to detect glutamatergic neurons. Pilot analyses showed that Nkx2-2, Calb1, Calb2, Nfia, and Cck gave the strongest and most robust signal, and these were used for subsequent analysis (Table 1).

Table 1: Hiplex probes used in the study.

Probes used in each iteration of the Hiplex analysis. Nfix, Nos1, and Pnoc were removed due to low, inconsistent signals. Pvalb, Penk, and Gal were not included in the final analysis because of data loss due to tissue damage (TD) by the 3rd round of staining. Slc17a6 was only in a small subset of experiments and was not included in the final analysis.

Round/Channel Fluorophore 1st Iteration 2nd Iteration 3rd Iteration
R1:T1 488 (Green) Lhx6 In Data Lhx6 In Data Lhx6 In Data
R1:T2 550 (Orange) Nkx2-2 In Data Fos In Data Nkx2-2 In Data
R1:T3 647N (Far Red) Nfia In Data Nkx2-2 In Data Nfia In Data
R1:T4 750 (Near Red) Calb1 In Data Slc17a6 Not in data analysis Calb1 In Data
R1:T5 488 (Green) Calb2 In Data Slc32a1 In Data Calb2 In Data
R2:T6 550 (Orange) Nfix Removed Nfia In Data Slc32a1 In Data
R2:T7 647N (Far Red) Fos In Data Nfix Removed Fos In Data
R2:T8 750 (Near Red) Cck In Data Cck In Data Cck In Data
R3:T9 488 (Green) Nos1 Removed Nos1 Removed Nos1 Removed
R3:T10 550 (Orange) Pnoc Removed Pvalb TD * Pvalb TD
R3:T11 647N (Far Red) Penk TD Penk TD Penk TD
R3:T12 750 (Near Red) Gal TD Gal TD Gal TD
Open in a new tab
*

TD - Samples incurred severe tissue damage by R3 of imaging rendering foci uncountable

Using these probe sets, we first analyzed the distribution of Lhx6-positive cells along the anterior-posterior axis of the ZI at Zeitgeber Time 6 (ZT6), which corresponds to 6 hours following lights on, which is a mild to moderate sleep pressure state in nocturnal mice (Fig. 1A, Table 2) (Mitler et al. 1977). Lhx6 mRNA-positive cells comprised 35.2% of all ZI cells (Fig. 1B), consistent with estimates from Lhx6 immunohistochemistry and Lhx6-eGFP reporter lines (Liu et al. 2017a; Kim et al. 2021a). Most of these cells (33.7% of all ZI cells) showed relatively low levels of Lhx6 expression, ranging from 1-5 puncta associated with each DAPI-positive cell (Fig. 1A, B, Table 2), whereas only 1.5 % showed high expression. The density of Lhx6-positive cells decreased from anterior to posterior ZI (Fig. 1C): anterior (−0.955 to −1.554 mm Bregma) 2.7 % high / 53.2 % low, medial (−1.555 to −2.154 mm Bregma) 1.5 % / 31 %, and posterior (−2.155 to −2.780 mm Bregma) 0.8 % / 21.8 %.

Figure 1: Lhx6 expression decreases along the anterior-posterior axis of the zona incerta.

Figure 1:

Open in a new tab

A. Top row: Spatial diagram depicting the anterior and posterior limits of the zona incerta (green). Bottom row: In situ hybridization showing Lhx6 (green) expressing cells (blue) decreasing along the anterior-posterior axis, covering a region spanning −0.955 mm to −2.780 mm Bregma. Scale bar=20 μm for all images. B. Bar graph depicting the fraction of Lhx6-positive cells at ZT6. Lhx6-positive cells compose 35.2% of total cells in the zona incerta (ZI), with high-expressing Lhx6-positive cells (red) comprising 1.5±1.3% of cells, low-expressing (blue) Lhx6-positive cells comprising 33.7±18.7% of cells, and the remaining 64.7±19.9% of cells do not express Lhx6 (grey). C. Bar graph depicting the fraction of Lhx6-positive cells at ZT6 along the anterior-posterior axis. The anterior region contains the most Lhx6-positive expressing cells, with high-expressing cells comprising 2.6±1.5% of total cells and low-expressing cells comprising 53.2±14.3%. The medial region contains 1.5±1.4% high-expressing cells and 30.9±13.2% low-expressing cells. The posterior region contains the least Lhx6-positive cells, with high comprising 0.75±0.77% and low comprising 21.8±16.5%. There are significant differences in the numbers of low-expressing cells between the anterior and medial (p = 0.0187, alpha = 0.05) and the anterior and posterior (p = 0.0012, alpha=0.05) regions.

Table 2: Puncta number distribution among probes tested.

The table describes the composition of cell types amongst each marker at ZT6 (left) and amongst all experimental groups in wildtype mice (right). At ZT6, the majority of cells are negative for markers (61.6±10.6%), followed by low-expressing cells (33.6±9.1%). High expressing cells are the least abundant (4.5±3.8%). Similar trends exist across all experimental groups in wildtype animals (marker negative cells: 63.6±7.4%, low expressing cells: 31.3±7.5%, high expressing cells: 4.8±2.6%).

ZT6 All WT Control & SD Conditions
Probe High (6+ Foci) Low (1-5 Foci) Neg (0 Foci) Probe High (6+ Foci) Low (1-5 Foci) Neg (0 Foci)
Lhx6 0.01 0.31 0.67 Lhx6 0.04 0.31 0.65
Calb1 0.13 0.34 0.53 Calb1 0.10 0.30 0.60
Calb2 0.05 0.28 0.68 Calb2 0.02 0.22 0.74
Cck 0.06 0.53 0.41 Cck 0.05 0.31 0.63
Nfia 0.03 0.39 0.57 Nfia 0.03 0.48 0.49
Nkx2-2 0.03 0.25 0.72 Nkx2-2 0.03 0.32 0.65
Fos 0.02 0.33 0.65 Fos 0.07 0.30 0.63
Slc32a1 0.03 0.26 0.70 Slc32a1 0.04 0.26 0.70
Avg 0.045 0.336 0.616 Avg 0.048 0.313 0.636
SD 0.038 0.091 0.106 SD 0.026 0.075 0.074
Open in a new tab

We performed the same analysis for the other markers. Calb1 (11.9% high/33% low), Cck (4.8% high/42.5% low), and Nfia (3.9% high/38.8% low) labeled the largest fractions of ZI cells, although no probe labeled fewer than 25.8% of ZI cells, with Nkx2-2 (3.7% high/22.5% low) showing the lowest overall prevalence (Fig. S3A, B). We observed only non-significant trends in the distribution of many of these markers along the anterior-posterior axis of the ZI, with Fos and Nkx2-2 distribution paralleling the anterior enrichment seen for Lhx6, and Calb1 exhibiting a medial bias (Fig. S3C).

Responses of Lhx6-positive and Lhx6-negative ZI cells to sleep pressure

To profile activity changes under both naturally occurring and induced sleep pressure, we quantified Fos expression in both Lhx6-positive and Lhx6-negative ZI cells along the anterior-posterior axis of the ZI (Table 3; Supplemental Dataset 1). No significant changes in the total number of Lhx6-positive cells were detected between conditions (Fig. S4). Across the ZI, Fos levels were higher at high (ZT0) and mild-moderate (ZT6) sleep pressure than at low pressure (ZT12) (Fig. 2A,E) in both Lhx6-positive (Fig. 2B) and Lhx6-negative (Fig. 2C) populations, indicating that many ZI subtypes track naturally fluctuating sleep pressure.

Table 3: Samples examined by time and sleep deprivation conditions.

The table lists the collection points during experimentation in standard time (left column), zeitberger time (middle column), and the corresponding sleep deprivation and recovery sleep conditions (right column).

Time ZT Time SD Condition
6:30 AM ZT0 -
12:30 PM ZT6 SD
1:30 PM ZT7 SDRS1
3:30 PM ZT9 SDRS3
6:30 PM ZT12 SDRS6
8:30 PM ZT14 SDRS8
Open in a new tab

Figure 2: Induced sleep deprivation induces Fos expression in both Lhx6-positive and Lhx6-negative cells.

Figure 2:

Open in a new tab

A. Bar graph depicting the Fos-positive fraction of total cells in the undisturbed control and after sleep deprivation. Significant differences were found between: (alpha = 0.05) Control ZT0 v. Control ZT12 (p = 0.0117), Control ZT0 v. Control ZT14 (p = 0.0415), SD ZT0 v. SD ZT12 (SDRS6, p = 0.0117), SD ZT0 V. ZT14 SD (p = 0.0415), Control ZT6 v. Control ZT12 (p = 0.0274), SD ZT6 v. SD ZT12 (SDRS8, p = 0.0274), SD ZT7 (SDRS1) v. Control ZT9 (p = 0.0256), SD ZT7 (SDRS1) v. Control ZT12 (p = 0.0006), SD ZT7 (SDRS1) v. SD ZT12 (SDRS6) (p = 0.0393), SD ZT7 (SDRS1) v. Control ZT14 (p = 0.0022). B. Bar graph depicts the Fos-positive fraction of Lhx6-positive cells in control and sleep deprivation groups. Significant differences were found between: (alpha = 0.05) ZT0:Control v. ZT12:Control (p = 0.0068), ZT0:SD vs. ZT12:SD (p = 0.0068), ZT6:Control vs. ZT7:SD (p = 0.0221), ZT6:Control vs. ZT12:Control (p = 0.0261), ZT6:SD vs. ZT12:SD (p = 0.0261), ZT7:SD vs. ZT9:Control (p = 0.0033), ZT7:SD vs. ZT12:Control (p ≤ 0.0001), ZT7:SD vs. ZT12:SD (p = 0.0035), ZT7:SD vs. ZT14:Control (p = 0.0004), ZT7:SD vs. ZT14:SD (p = 0.0185), ZT9:SD vs. ZT12:Control (p = 0.0443). C. Bar graph depicting the Fos-positive fraction of Lhx6-negative cells in control and sleep deprivation groups. Significant differences were found: ZT0:Control vs. ZT12:Control (p = 0.0056), ZT0:Control vs. ZT14:Control (p = 0.0152), ZT0:SD vs. ZT12:SD (p = 0.0056), ZT0:SD vs. ZT14:SD (p = 0.0152), ZT6:Control vs. ZT12:Control (p = 0.0219), ZT6:SD vs. ZT12:SD (p = 0.0219), ZT7:SD vs. ZT9:Control (p = 0.0280), ZT7:SD vs. ZT12:Control (p = 0.0022), ZT7:SD vs. ZT14:Control (p = 0.0060). D. In situ hybridization showing Lhx6 (green) and Fos (red) expression (DAPI. blue) in response to control (ZT6, scale bar=20 μm) and sleep deprivation conditions: sleep deprivation (SD, scale bar=20 μm), sleep deprivation with 1 hour recovery sleep (SDRS1, scale bar=50 μm), sleep deprivation with 3 hours recovery sleep (SDRS3, scale bar=20 μm). Table 1 describes standard time and the corresponding zeitgeber time, and sleep deprivation conditions.

To examine the effects of induced sleep pressure, mice were sleep deprived by continuous gentle brushing from ZT0 to ZT6 (Lemons, Saré, and Beebe Smith 2018; Liu et al. 2017a). One hour into recovery sleep (ZT7), Fos expression rose markedly in Lhx6-positive cells relative to ZT6 controls (Fig. 2B,E), but not in Lhx6-negative cells (Fig. 2C). After three hours of recovery (ZT9) Fos remained elevated, with a larger Fos-positive fraction in the Lhx6-positive cohort (Fig. 3A). By six hours of recovery (ZT12), Fos had declined in both groups (Fig. 2B,C). Notably, more Lhx6-positive neurons displayed high Fos intensity after sleep deprivation than their Lhx6-negative counterparts (Fig. 3C,F), demonstrating that Lhx6-positive cells mount a stronger and more persistent response as pressure dissipates (Fig. 3A–F).

Figure 3: Sleep pressure-dependent Fos induction is stronger in Lhx6-positive than in Lhx6-negative cells.

Figure 3:

Open in a new tab

A. Bar graph depicts the Fos-positive fraction of Lhx6-negative cells (left, grey) and the Fos-positive fraction of Lhx6-positive cells (right, green) across ZT points. B. Bar graph depicts the Fos-positive fraction of Lhx6-negative cells (left, grey) and the Fos-positive fraction of Lhx6- positive cells (right, green) aggregated across all control timepoints under natural circadian conditions (p < 0.0001). C. Bar graph depicts the Fos-positive fraction of Lhx6-negative cells (left, grey) and the Fos-positive fraction of Lhx6-positive cells (right, green) aggregated across all control timepoints under induced sleep deprivation (p<0.0001). D. In situ hybridization illustrates examples of highly expressing Fos-positive cells (red arrow), low-expressing Fos cells (orange), and Fos-negative cells (orange). 20 μm scale bar. E. Bar graph depicts the high expressed Fos-positive fraction of Lhx6 negative cells (left, grey) and the Fos-positive fraction of Lhx6-positive cells (right, green) aggregated across all control timepoints under natural circadian conditions (p < 0.01). F. Bar graph depicts the high-expressing Fos-positive fraction of Lhx6-negative cells (left, grey) and the Fos-positive fraction of Lhx6-positive cells (right, green) aggregated across all control timepoints under induced sleep deprivation (p < 0.01).

Spatial analysis revealed that natural pressure rarely drove high Fos expression (> 5 puncta per cell) along the anterior–posterior axis, whereas early sleep deprivation induced numerous high-Fos/Lhx6-positive neurons, where Fos expression peaked after deprivation and one hour into recovery, then falling sharply after three hours of recovery sleep (Fig. S5). We conclude that induced sleep deprivation elicits much higher cellular levels of Fos in Lhx6-positive ZI neurons than does naturally arising sleep pressure changes.

Responses of molecularly distinct subtypes of Lhx6-positive and Lhx6-negative ZI cells to sleep pressure

We next examine Fos induction in molecularly distinct subtypes of Lhx6-positive and Lhx6-negative ZI cells to sleep pressure (Table 3; Supplemental Dataset 1). As with Lhx6 itself, none of the subtype markers showed any significant sleep pressure-dependent changes in gene expression (Fig. S6). Cck expression, however, did show time of day-dependent changes, with the highest expression observed at ZT0 and ZT6, and significantly reduced expression at later time points. However, the same pattern persisted during induced sleep pressure, indicating that while Cck expression may be regulated by circadian time and/or naturally occurring changes in sleep pressure, rather than pressure-dependent regulation.

All Lhx6-positive subtypes showed fewer Fos-positive cells under low sleep pressure (ZT12) than under high sleep pressure (ZT0) in unstimulated mice; all but the Lhx6-positive/Calb2-positive cells also differed between moderate (ZT6) and low sleep pressure (Fig. 4A–E). Following experimentally-induced sleep deprivation, significant decreases in Fos were observed between samples at ZT6 (end of SD treatment) and/or ZT7 (SD with 1 hour of recovery sleep) relative to ZT12 (SD with 6 hours of recovery sleep). This decreased Fos expression was observed in every Lhx6-positive subtype except for Lhx6-positive/Calb2-positive cells (Fig. 4A–E). Lhx6-negative neurons bearing these markers largely mirrored these kinetics (Fig. S7A–E), with two exceptions: Cck- and Calb2-expressing Lhx6-negative cells followed a circadian-like pattern, indicating weaker responsiveness to induced deprivation (Fig. S7B–C).

Figure 4: Lhx6-positive zona incerta cells expressing cell subtype-specific markers show similar changes in Fos induction in response to naturally-occurring and experimentally-induced changes in sleep pressure.

Figure 4:

Open in a new tab

Bar graphs depicting the Fos-positive fraction of marker-positive, Lhx6-positive cells under naturally occurring (grey) and induced sleep pressure (colored). A. Calb1 (blue). Significant differences: ZT0:Control vs. ZT12:Control (p = 0.0020), ZT0:SD vs. ZT12:SD (p = 0.0020), ZT6:Control vs. ZT12:Control (p = 0.0012), ZT6:SD vs. ZT12:SD (p = 0.0012), ZT7:SD vs. ZT9:Control (p = 0.0086), ZT7:SD vs. ZT12:Control (p ≤ 0.0001), ZT7:SD vs. ZT12:SD (p = 0.0008), ZT7:SD vs. ZT14:Control (p = 0.0012), ZT7:SD vs. ZT14:SD (p = 0.0224), ZT9:SD vs. ZT12:Control (p = 0.0132). B. Calb2 (red). Significant differences: ZT0:Control vs. ZT6:Control (p =0.0063), ZT0:Control vs. ZT12:Control (p = 0.0032), ZT0:Control vs. ZT14:Control (p =0.0078), ZT0:SD vs. ZT6:SD (p = 0.0063), ZT0:SD vs. ZT12:SD (p = 0.0032), ZT0:SD vs. ZT14:SD (p = 0.0078), ZT7:SD vs. ZT12:Control (p = 0.0431). C. Cck (yellow). Significant differences: ZT0:Control vs. ZT9:Control (p = 0.0171), ZT0:Control vs. ZT12:Control (p = 0.0003), ZT0:Control vs. ZT14:Control (p = 0.0045), ZT0:SD vs. ZT9:SD (p=0.0171), ZT0:SD vs. ZT12:SD (p = 0.0003), ZT0:SD vs. ZT14:SD (p = 0.0045), ZT6:Control vs. ZT12:Control (p = 0.0117), ZT6:SD vs. ZT12:SD (p = 0.0117), ZT7:SD vs. ZT9:Control (p = 0.0072), ZT7:SD vs. ZT12:Control (p = 0.0002), ZT7:SD vs. ZT12:SD (pp = 0.0106), ZT7:SD vs. ZT14:Control (p = 0.0023) D. Nfia (purple). Significant differences: ZT0:Control vs. ZT12:Control (p=0.0181), ZT0:SD vs. ZT12:SD (p = 0.0181), ZT6:Control vs. ZT7:SD (p = 0.0328), ZT6:Control vs. ZT12:Control (p = 0.0137), ZT6:SD vs. ZT12:SD (p = 0.0137), ZT7:SD vs. ZT9:Control (p = 0.0069), ZT7:SD vs. ZT12:Control (p < 0.0001), ZT7:SD vs. ZT12:SD (0.0034), ZT7:SD vs. ZT14:Control (p = 0.0004), ZT7:SD vs. ZT14:SD (p = 0.0199) E. Nkx2-2 (orange) ZT0:Control vs. ZT12:Control (p = 0.0215), ZT0:SD vs. ZT12:SD (p = 0.0215), ZT6:Control vs. ZT12:Control (p = 0.0169), ZT6:SD vs. ZT12:SD (p=0.0169).

Nkx2-2 is essential for the development and function of Lhx6-positive ZI neurons

These data indicate that Lhx6-positive/Nkx2-2-positive cells may be particularly sensitive to sleep pressure. Previous studies from our group have also suggested that Nkx2-2 may be important for the specification and/or differentiation of a subset of hypothalamic Lhx6-positive neurons, including those in the ZI (Kim et al. 2021b). To test this directly, we generated Lhx6-Cre;Nkx2-2lox/lox mice, thereby selectively inactivating Nkx2-2 in Lhx6-expressing cells (Fragkouli et al. 2009; Mastracci, Lin, and Sussel 2013; Gross et al. 2016) (Fig. 5A). Mouse litters contained the expected Mendelian ratios of Lhx6-Cre;Nkx2-2lox/+ and Lhx6-Cre;Nkx2-2lox/lox offspring (Fig. S10).

Figure 5: Lhx6-Cre;Nkx2-2lox/lox mice show reduced expression of both Lhx6 and Nkx2-2 in zona incerta.

Figure 5:

Open in a new tab

A. Diagram depicting the expression of Lhx6 (green) in the coronally sliced mouse brain (left). Diagram depicting Lhx6-Cre-mediated deletion of Nkx2-2 (right). B. Bar graph depicting the fraction of Lhx6-Nkx2-2 positive cells of total cells amongst experimental groups: Wildtype (WT, blue), Lhx6-Cre;Nkx2-2lox/+ (HET, red), and Lhx6-Cre;Nkx2-2lox/lox(HOMO, green). Significant differences: WT vs. HOMO (p = 0.0038). C. Bar graph depicting the fraction of Slc32a1/Nkx2-2 positive cells of total cells amongst experimental groups: Wildtype (WT, blue), Lhx6-Cre;Nkx2-2lox/+ (HET, red), and Lhx6-Cre;Nkx2-2lox/lox(HOMO, green). Significant differences: HET vs. HOMO (p = 0.0194). D. Bar graph depicting the fraction of Lhx6-positive cells of total cells in the cortex (CTX) amongst experimental groups: Wildtype (WT, blue), Lhx6-Cre;Nkx2-2lox/+ (HET, red), and Lhx6-Cre;Nkx2-2lox/lox(HOMO, green). E. Bar graph depicting the fraction of Slc32a1/Nkx2-2 positive cells of total cells in the lateral hypothalamus (LH) amongst experimental groups: Wildtype (WT, blue), Lhx6-Cre;Nkx2-2lox/+ (HET, red), and Lhx6-Cre;Nkx2-2lox/lox(HOMO, green). F-I. In situ hybridization depicting the expression of Lhx6 (green), Nkx2-2 (purple), Slc32a1 (white) in cells (DAPI, blue) in the Zona Incerta (ZI, F-G), lateral hypothalamus (LH, H), and the Cortex (CTX, I) in wildtype (WT) mice. J-M. In situ hybridization depicting the expression of Lhx6 (green), Nkx2-2 (purple), Slc32a1 (white) in cells (DAPI, blue) in the ZI (J-K), LH (L), and cerebral cortex (M) in Lhx6-Cre;Nkx2-2lox/+ (HET) mice. N-Q. In situ hybridization depicting the expression of Lhx6 (green), Nkx2-2 (purple), Slc32a1 (white) in cells (DAPI, blue) in the ZI (J-K), LH (L), and cortex (M) in Lhx6-Cre;Nkx2-2lox/lox (HOMO) mice. Scale bar=20 μm for all images.

Mutant animals were indistinguishable from controls in appearance, body weight, and general locomotor activity. In the ZI, however, Lhx6-Cre;Nkx2-2lox/lox mice showed significant reductions in both Lhx6-positive/Nkx2-2-positive and Slc32a1-positive/Nkx2-2-positive GABAergic neurons, indicating the efficiency of the intersectional loss of function mutants generated here (Fig. 5B, C). Specificity was confirmed by the finding of unchanged numbers of both Lhx6-positive cortical interneurons, which never express Nkx2-2, and Slc32a1-positive/Nkx2-2-positive GABAergic neurons in the lateral hypothalamus, where Lhx6-Cre is inactive (Fig. 5D).

A non-significant trend towards reduced numbers of Lhx6-positive/Nkx2-2-positive cells was also observed in heterozygous Lhx6-Cre;Nkx2-2lox/+ mice, indicating a possible dose-dependent requirement for Nkx2-2 in the development of Lhx6-positive ZI neurons (Fig. 5B, C). Loss of both Lhx6-positive and Nkx2-2-positive cells was evident across all positions along the anterior-posterior axis of the ZI (Fig. S11).

Loss of Nkx2-2 also disrupted expression of other subtype markers of Lhx6-positive ZI neurons (Fig. 6). Relative to wild-type animals, both heterozygous Lhx6-Cre;Nkx2-2lox/+ and homozygous Lhx6-Cre;Nkx2-2lox/lox mutants displayed significant reductions in Calb1-, Calb2-, and Cck-positive Lhx6 neurons. in. We also observe a trend towards reduced expression of all probes in the medial ZI, and a corresponding relative increase in both the anterior and, in particular, the posterior ZI. In heterozygous Lhx6-Cre;Nkx2-2lox/+ mice, a similar redistribution was observed for Lhx6. Although the total number of Slc32a1-positive cells in the ZI was unchanged across all genotypes examined, the relative number of cells showing strong (>5 puncta/cell) Slc32a1 expression was reduced in homozygous Lhx6-Cre;Nkx2-2lox/lox mutants, relative to Lhx6-Cre;Nkx2-2lox/+ heterozygotes.

Figure 6: Lhx6-Cre;Nkx2-2lox/lox mice show reduced expression of cell subtype-specific markers in zona incerta.

Figure 6:

Open in a new tab

A. Bar graph depicting the marker (from left to right: Calb1, Calb2, Cck, Nfia, Slc32a1, Fos) positive fraction of total cells across experimental groups: Wildtype (WT, blue), Lhx6-Cre;Nkx2-2lox/+ (HET, red), Lhx6-Cre;Nkx2-2lox/lox (HOMO, green). Significant differences were found between wildtype and het in the expression of Calb1 (p = 0.0065), Calb2 (p = 0.0175), Cck(p = 0.0001), and Fos (p = 0.0320); as well as between wildtype and homo in the expression of Calb1(p=0.0015), Calb2(p = 0.0098), Cck(p = <0.0001), Nfia(p = 0.0078), and Fos(p = 0.0032). B-G. Bar graphs depicting the fraction of marker-positive cells (high expressing, red; low expressing, blue) of total cells, along the anterior to posterior axis, across experimental groups. Right panel: Wildtype (WT), Middle panel: Lhx6-Cre;Nkx2-2lox/+ (HET), left panel: Lhx6-Cre;Nkx2-2lox/lox (HOMO). No statistical analysis conducted due to HET and HOMO samples containing N < 2 per location.

We next used Xenium-based spatial transcriptomic analysis to more comprehensively analyze gene expression in wildtype mice using a 347-gene panel, which consisted of the Xenium Mouse Brain probeset (Ma et al. 2024) and 100 additional probes selected for enriched expression in hypothalamic cell types (Kim et al. 2020). After segmenting cells from the ZI and adjacent tissues and performing UMAP analysis on their expression profiles, we resolved discrete clusters of glutamatergic and GABAergic neurons, and non-neuronal cell types such as astrocytes, oligodendrocyte precursor cells, mature oligodendrocytes, endothelial cells, and microglia (Fig. 7C). Subclustering the GABAergic Lhx6-expressing neuronal population identified a subset corresponding to Lhx6-positive ZI neurons (Fig. 7E–G). This analysis confirmed the anterior-posterior gradient in the relative number of Lhx6-expressing cells in wildtype animals observed using HiPlex analysis (Fig. 7H). We identified additional genes enriched in Nkx2-2-positive and Nkx2-2-negative cell types, and further subclusters within these groups (Fig. S12). Nkx2-2-positive cells comprised four clusters (clusters 0-3), and expressed Lypd6, Nkx2-1, Foxp2, Ly6a, Adgrl4, Paqr5, Penk, and Dlx1. In contrast, Nkx2-2-negative cell types (clusters 4 and 5) lack expression of these markers. Instead, Nkx2-2-negative cell types have more intense and abundant expression of Col6a1 and Tmem132d, and more abundant expression of Calb2.

Figure 7: Xenium-based analysis of Lhx6-positive ZI cells.

Figure 7:

Open in a new tab

A. UMAP depicts the clustering of wild-type anterior, medial, and posterior samples. B. UMAP identifies 17 clusters in wildtype anterior, medial, and posterior samples. C. UMAP depicts major cell types. D. Visualization of Seurat-identified clusters in the spatial context of the hypothalamus. E. Visualization of Seurat identified clusters in the spatial context of the hypothalamus with Lhx6-expressing ZI neurons highlighted (red with red arrow). F. UMAP identifying the GABAergic Lhx6 population. G. UMAP depicts subclusters of GABAergic Lhx6-positive cells after additional filtration of cells expressing astrocyte, glial, and glutamatergic markers. H. UMAP depicts subclusters of GABAergic Lhx6 neurons distributed along the anterior-posterior axis.

We next tested whether the reduction in the number of Lhx6-positive cells was reflected in defective activation of the remaining ZI cells to elevate sleep pressure. In both heterozygous Lhx6-Cre;Nkx2-2lox/+ and homozygous Lhx6-Cre;Nkx2-2lox/lox mutants, we observed dramatic reductions in the number of total Fos-positive neurons under conditions of both moderate (ZT6), moderately high (SD+3 hours recovery sleep) and high (ZT6+6 hours of SD) sleep pressure (Fig.8). The few remaining Lhx6-positive/Nkx2-2-positive ZI cells likewise showed reduced levels of overall activation, and no clear changes in activity in response to altered sleep pressure. High levels of Fos expression were likewise not observed in cells at any point along the anterior-posterior axis of the ZI (Fig. S13).

Figure 8: Lhx6-Cre;Nkx2-2lox/lox mice fail to induce Fos in response to experimentally-induced sleep deprivation.

Figure 8:

Open in a new tab

A-C. In situ hybridization depicts Fos (red) and Lhx6 (green) expression in wildtype (WT) at ZT6 (A), after sleep deprivation (SD, B), and after sleep deprivation with 3 hours of recovery sleep (SDRS3, C). D. Bar graph depicts the Fos-positive fraction of total cells in wildtype (WT, blue), Lhx6-Cre;Nkx2-2lox/+ (HET, red), and Lhx6-Cre;Nkx2-2lox/lox(HOMO, green) at ZT6, SD, and SDRS3. Significant differences were found between the wildtype and HET at ZT6(p=0.0354), SD(p<0.0001), and SDRS3(p=0.0037); as well as the wildtype and HOMO at ZT6(p=0.0045), SD(p<0.0001), and SDRS3(p<0.0001). E-G. In situ hybridization depicts Fos (red) and Lhx6 (green) expression in Lhx6-Cre;Nkx2-2lox/+ (HET) mice at ZT6 (E), after sleep deprivation (SD, F), and after sleep deprivation with 3 hours of recovery sleep (SDRS3, G). H. Bar graph depicts the Lhx6-Nkx2-2 positive fraction of total cells in wildtype (WT, blue), Lhx6-Cre;Nkx2-2lox/+ (HET, red), and Lhx6-Cre;Nkx2-2lox/lox(HOMO, green) at ZT6, SD, and SDRS3. Significant differences were found between wildtype and HET at SD (p=0.0015) and SDRS3 (p<0.0001); as well as the wildtype and HOMO at ZT6(0.0235), SD(p=0.0009), and SDRS3(p<0.0001). I-K. In situ hybridization depicts Fos (red) and Lhx6 (green) expression in Lhx6-Cre;Nkx2-2lox/lox(HOMO) mice at ZT6 (I), after sleep deprivation (SD, J), and after sleep deprivation with 3 hours of recovery sleep (SDRS3, K). L. Bar graph depicts the Fos positive fraction of Lhx6-Nkx2-2 positive cells in wildtype (WT, blue), Lhx6-Cre;Nkx2-2lox/+ (HET, red), and Lhx6-Cre;Nkx2-2lox/lox(HOMO, green) at ZT6, SD, and SDRS3. Significant differences were found between the wildtype and HET at SD(p=0.0007) and SDRS3(p=0.0008); as well as the wildtype and HOMO at ZT6(p=0.0259), SD(p=0.0012), and SDRS3(p=0.0001).

Finally, we examined sleep patterns in control Lhx6-Cre mice and heterozygous Lhx6-Cre;Nkx2-2lox/+ and homozygous Lhx6-Cre;Nkx2-2lox/lox mutants using the Piezo system (Yaghouby et al. 2016) (Fig. 9). Relative to control animals, heterozygous Lhx6-Cre;Nkx2-2lox/+ mice showed increased sleep time during the day, but no change in total nighttime sleep, or either daytime or nighttime sleep bout length. In contrast, homozygous Lhx6-Cre;Nkx2-2lox/lox showed significantly increased sleep time and sleep bout length during both day and night relative to both heterozygotes and, except for nighttime sleep bout length, to controls.

Figure 9: Total sleep is increased in Lhx6-Cre;Nkx2-2lox/+ and Lhx6-Cre;Nkx2-2lox/lox mice.

Figure 9:

Open in a new tab

A Bar graph depicts the percent of time spent asleep during the day in Lhx6-Cre (white), Lhx6-Cre;Nkx2-2lox/+ (grey), and Lhx6-Cre;Nkx2-2lox/lox mice (blue). Significant differences found between Lhx6-Cre and Lhx6-Cre;Nkx2-2lox/+ mice (p≤0.05)as well as between Lhx6-Cre and Lhx6-Cre;Nkx2-2lox/lox mice (p≤ 0.01). B Bar graph depicts the percent of time spent asleep at night in Lhx6-Cre (white), Lhx6-Cre;Nkx2-2lox/+ (grey), and Lhx6-Cre;Nkx2-2lox/lox mice (blue). Significant differences were found between Lhx6-Cre and Lhx6-Cre;Nkx2-2lox/lox mice (p≤0.05), as well as between Lhx6-Cre;Nkx2-2lox/+ and Lhx6-Cre;Nkx2-2lox/lox mice (p≤0.05). C Box and whisker plot depicts the duration of sleep bouts during the day in Lhx6-Cre (white), Lhx6-Cre;Nkx2-2lox/+ (grey), and Lhx6-Cre;Nkx2-2lox/lox mice (blue). Significant differences were found between Lhx6-Cre and Lhx6-Cre;Nkx2-2lox/lox mice (p≤0.01), as well as between Lhx6-Cre;Nkx2-2lox/+ and Lhx6-Cre;Nkx2-2lox/lox mice (p≤0.05). C Box and whisker plot depicts the duration of sleep bouts at night in Lhx6-Cre (white), Lhx6-Cre;Nkx2-2lox/+ (grey), and Lhx6-Cre;Nkx2-2lox/lox mice (blue). A significant difference was found between Lhx6-Cre;Nkx2-2lox/+ and Lhx6-Cre;Nkx2-2lox/lox mice.

Discussion:

The ZI was first studied for its role in sensorimotor integration (Deschênes et al. 2005; Chometton, Barbier, and Risold 2021; Hormigo et al. 2023; Schäfer and Hoebeek 2018). More recently, it has emerged as a hub for regulating internal states and innate behaviours, including feeding, thirst, and anxiety and aggression-related behaviors (de Git et al. 2021; Zhao et al. 2019; Li, Rizzi, and Tan 2021; Walsh and Grossman 1976; Chou et al. 2018) (de Git et al. 2021; Zhao et al. 2019; Li, Rizzi, and Tan 2021; Walsh and Grossman 1976; Chou et al. 2018). Importantly, the ZI also contributes to sleep-wake regulation by detecting and signaling sleep pressure (Liu et al. 2017b; Vidal-Ortiz, Blanco-Centurion, and Shiromani 2024; Lee et al. 2025; Blanco-Centurion et al. 2021). Lhx6-positive ZI neurons are critical for these functions: developmental disruption of Lhx6 function causes apoptotic death, reduces their numbers, and decreases overall sleep (Kim et al. 2021a; Liu et al. 2017a). Glutamatergic neurons of the thalamic reuniens nucleus, which mediate the effects of induced sleep pressure, directly activate anterior Lhx6-positive ZI neurons (Lee et al. 2025). However, activation patterns within Lhx6-positive neurons are heterogeneous, and responses of Lhx6-negative ZI neurons to sleep pressure remain poorly characterized.

Sleep pressure-dependent activation of Lhx6-positive and Lhx6-negative ZI neurons

Using multiplex smFISH, we quantified Fos induction in Lhx6-positive and Lhx6-negative ZI neurons across molecular subtypes and examined developmental and behavioral effects of selectively disrupting Nkx2-2 in Lhx6-positive neurons. Consistent with prior reports (Liu et al. 2017a), Lhx6-positive neurons robustly induced Fos following naturally occurring and experimentally-induced sleep pressure, and sustained elevated expression for at least 3 hours into recovery sleep before gradually declining. Induced sleep pressure initially triggered higher Fos levels than naturally-regulated sleep pressure changes, though this difference disappeared after several hours of recovery sleep. Lhx6-negative neurons showed similar but weaker Fos responses, with fewer Fos-positive cells overall and lower cellular expression levels. Both Lhx6-positive and -negative anterior ZI neurons exhibited higher Fos induction than posterior ZI neurons, paralleling the pattern of reuniens inputs (Lee et al. 2025). We conclude that while Lhx6-positive neurons show the strongest responses to elevated sleep pressure, multiple ZI subtypes participate in sleep pressure signaling.

Molecular diversity of sleep pressure-responsive ZI neuronal subtypes

ZI neurons are highly heterogeneous (M. Zhang et al. 2023), yet activity-dependent differences among subtypes remain largely unexplored. Using a panel of molecular markers, we observed no sleep pressure-dependent changes in their expression, though Cck levels rose at the start of the circadian day. Nkx2-2 and Nfia were enriched anteriorly, while Calb1, Calb2, and Cck were more medial/posterior. Fos induction across these molecularly defined subtypes largely paralleled patterns in Lhx6-positive neurons, with Lhx6-negative counterparts showing reduced numbers of Fos-positive cells and lower expression levels. These findings confirm that sleep pressure broadly activates diverse ZI cell types, with Lhx6-positive neurons showing the strongest responses.

Nkx2-2 and developmental specification of Lhx6-positive ZI neurons

To define Nkx2-2’s role in ZI development, we disrupted it selectively in Lhx6-positive neurons and analyzed molecular and behavioral consequences using both HiPlex smFISH and Xenium spatial transcriptomics. Wildtype ZI neurons segregated into Nkx2-2-positive and -negative groups with distinct molecular markers. Nkx2-2 loss decreased expression of markers for Nkx2-2-positive neurons and increased those for Nkx2-2-negative neurons, with residual Nkx2-2-positive populations displaced anteriorly and posteriorly. Calb1, Calb2, and Cck expression was reduced in both heterozygous and homozygous mutants, while overall Lhx6-positive neuron numbers declined significantly in homozygotes. These results imply that Nkx2-2 acts during development to specify both Lhx6-positive and -negative ZI neurons.

Consequences of Nkx2-2 loss for sleep pressure signaling and sleep behavior

Developmental Nkx2-2 loss of function produced profound effects on sleep pressure signaling and behavior. Fos induction in Lhx6-positive neurons was markedly reduced under all conditions, especially in homozygous mutants. Paradoxically, these animals exhibited increased sleep time and bout length, with homozygotes showing stronger effects. Thus, Nkx2-2 disruption compromises ZI signaling of sleep pressure and alters sleep architecture, with effects opposite to those observed following complete Lhx6 loss.

Coordinated sleep pressure coding in the ZI

These results underscore a central role for Lhx6-positive ZI neurons in signaling levels of both naturally occurring and induced sleep pressure, but also show that many Lhx6-negative ZI cells show broadly similar responses to elevated sleep pressure. Since Lhx6-positive ZI neurons are essential for the accumulation of induced sleep pressure (Liu et al. 2017a; Lee et al. 2025), and selective disruption of Lhx6-positive ZI neuron development by the loss of function of Nkx2-2 severely disrupts sleep pressure-induced Fos expression in Lhx6-negative neurons, this implies that Lhx6-positive neurons coordinate these broader sleep pressure-dependent changes in activity across the ZI. This conclusion is further supported by the extensive reciprocal connections that exist among ZI neurons (Londei et al. 2024). In the case of induced sleep pressure, this may reflect the fact that glutamatergic neurons of the thalamic reuniens nucleus selectively project to Lhx6-positive ZI neurons and show sleep pressure-dependent increases in synaptic strength. However, this cannot account for the effects of naturally induced changes in sleep pressure, as this does not induce changes in the activity of reuniens neurons (Lee et al. 2025).

Open questions and future directions

This raises the broader question of exactly how Lhx6-positive ZI neurons are selectively responsive to naturally occurring changes in sleep pressure. Our findings do not identify a molecularly distinct neuronal subpopulation that is responsive to any type of sleep pressure change, although Calb2-positive cells/Lhx6-positive cells appear to be less responsive. SnRNA-Seq from Fos-trapped neurons in the thalamic reuniens nucleus failed to identify any other molecular markers of specific neuronal subtypes responsive to induced sleep pressure (Lee et al. 2025), and the response properties of individual neurons in the ZI may likewise be determined by connectivity patterns that do not generally correlate with their gene expression profiles. Lhx6-positive ZI neurons may selectively receive synaptic input from an as yet unidentified neuronal subpopulation that is activated by naturally occurring changes in sleep pressure, or may sense these changes directly through as yet uncharacterized mechanisms. A more detailed analysis of the presynaptic inputs to these neurons that builds on previous work using rabies-based viral tracers (Liu et al. 2017a) and analysis of sleep pressure-induced transcriptomic changes will help address this.

The divergent sleep phenotypes seen in mutants that selectively disrupt the development of Lhx6-positive ZI neurons imply a more complex role for the ZI in regulating sleep pressure than has been previously hypothesized. Loss of function of Lhx6 in early hypothalamic neuroepithelium leads to a complete loss of Lhx6-positive ZI neurons that likely results from selective apoptosis (Kim et al. 2021a), and in turn leads to increased wake and decreased sleep (Liu et al. 2017a). In contrast, loss of function of Nkx2-2 in Lhx6-positive neural precursors reduces but does not completely eliminate Lhx6-positive ZI neurons, instead broadly disrupting the development and distribution of multiple subtypes of ZI neurons. This both dramatically reduces the induction of Fos that is induced by both naturally occurring and induced sleep pressure, and leads to significantly increased sleep. This implies that while Lhx6-positive ZI neurons are essential for sensing and signaling appropriate levels of sleep pressure both to Lhx6-negative ZI neurons and to arousal-promoting neurons in other brain regions, broad disruptions in the composition and organization of neuronal subtypes within the ZI like those that result from Nkx2-2 loss of function may instead result in excessive activation of inhibitory projections to arousal-promoting neurons. This, in turn, underscores the importance of connectivity among specific neuronal populations of the ZI in maintaining appropriate levels of sleep homeostasis. A systematic in vivo analysis of changes in the activity and connectivity of both Lhx6-positive and Lhx6-negative neurons of the ZI will help clarify the dynamic regulation of these neural circuits, sensing, and signaling changes in sleep pressure.

Materials and Methods:

Mouse colony maintenance

All experimental animal procedures were approved by the Johns Hopkins University Institutional Animal Care and Use Committee. All mice were housed in a climate-controlled facility (14-h light and 10-h dark cycle) with ad libitum access to food and water.

C57BL/6 were ordered from the Jackson Laboratory. Lhx6-iCre (B6;CBA-Tg(Lhx6-icre)1Kess/J, JAX#026555) and Nkx2-2lox/lox (Mastracci, Lin, and Sussel 2013) were crossed to generate Lhx6-Cre;Nkx2-2lox/+ mice. Lhx6-Cre;Nkx2-2lox/l+ mice were then bred to generate Lhx6-Cre;Nkx2-2lox/l+ or Lhx6-Cre;Nkx2-2lox/lox mice.

Sleep deprivation

Between 8-12 weeks of age, mice were moved to cages in a climate-controlled room, given food and water ad libitum at 4:30 AM, and allowed an hour and a half to acclimate to the new environment before experimentation. Sleep deprivation began at 6:30 AM (ZT0) and was facilitated by gentle brushing until 12:30 PM for a total of 6 hours. Control mice were undisturbed and allowed to sleep freely. Mice were sacrificed after six hours (ZT6) using cervical dislocation, corresponding to the collection of sleep deprivation and control groups, and after one, three, six and eight hours of recovery sleep; corresponding to seven hours (ZT7), nine hours (ZT9), twelve hours (ZT12) and fourteen hours (ZT14) after lights on, respectively. Brains were collected within 15 minutes of sacrifice and stored in OCT at −80°C until sectioning.

Tissue sectioning

Sections were coronally sliced between 10 and 14 μm using Leica Cryotstat CM3050S and placed on Superfrost Plus Microscope Slides or Xenium Slides. For Hiplex analysis, sections were collected every 30-42 μm. For Xenium analysis, sections were collected every 100 μm.

Hiplex analysis

Multiplex in situ hybridization was performed according to the fresh-frozen protocol listed by the RNAscope HiPlex v2 assay (ACDBio) using the probes (also provided by ACDBio) found in Table 1.

Xenium analysis

Xenium In Situ Gene expression was performed according to the fresh-frozen protocol listed by 10x Genomics (Ma et al. 2024).

Imaging

Slides were imaged using the Olympus widefield microscope at 20x, 40x, and 60x Oil magnifications. Images were captured using z-stacked montages using DAPI, GFP, TRITC, Cy5, and Cy7 filters. Images were further processed using Slidebook Reader (Intelligent Imaging Innovations, Denver, CO, USA) and ImageJ (Schneider, Rasband, and Eliceiri 2012; Blattner et al. 2014; Chalfoun et al. 2017).

Hiplex: Cell counting, segmentation & data assembly

Cells were segmented, counted, and assigned puncta using pipelines developed in Cell Profiler (Stirling et al. 2021). From these assignments, cells were further categorized into high, low, and negative expressing cell types based on the number of puncta expressed per probe (Table 2). With this information, a master sheet was created describing each cell, its assigned puncta, experimental group, and cell type (Supplemental Dataset 1).

Xenium: Analysis, Segmentation & Visualization

Data visualization was performed using the Xenium Explorer 1.3 (Ma et al. 2024) and the Seurat package in R (Satija et al. 2015; Hao et al. 2024).

Piezo analysis of sleep/wake behavior

The PiezoSleep system (Signal Solutions, LLC) was used to record and analyze sleep of mice under the light/dark (12h/12h) condition as described previously (Yaghouby et al. 2016). Briefly, Piezoelectric motion sensor film detects the sleep and wakefulness states of individual mice. The signal was amplified via a PiezoSleep in-line amplifier and collected via a Calamari SAS (8-channel) data acquisition system connected to a computer with PiezoSleep v.2.18 software installed. Sleep amount and bout duration were analyzed offline using SleepStats v.4.

Statistical analysis

Paired t-tests, Fisher’s exact tests, one-way ANOVAs, two-way ANOVAs, Kruskal-Wallis tests (in cases of non-parametric data), mixed-effect analysis, Tukey’s Multiple Comparison tests, and Dunn’s Multiple Comparison tests were performed using GraphPad Prism version 10.0.0 for Windows (GraphPad Software, Boston, Massachusetts, USA). The Seurat “FindAllMarkers” function with assay “SCT” and default parameters was used for analyzing differential gene expression, using the number of total mRNAs and genes as a variable. All bar graphs show mean and standard deviation (SD), with individual data points plotted. Significance was determined using an alpha of 0.05.

Resource Availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Seth Blackshaw (sblack@jhmi.edu).

Materials availability

All unique/stable reagents generated in this study are available from the Lead Contact upon request.

Supplementary Material

Supplement 1

Supplemental Dataset 1: Hiplex data for all samples.

The dataset contains all of the cells identified in the zona incerta across all experimental conditions and genotypes. Each cell is delineated by a unique cell ID name and contains corresponding information regarding experimental group, genotype, location, markers expressed, and cell type.

media-1.xlsx (31.7MB, xlsx)
2

Acknowledgements:

We thank W. Yap for comments on the manuscript. This work was supported by a National Science Foundation graduate fellowship to P.W.C and a grant from the NIH (R01MH126676) to S.B.

Declaration of interests

S.B. receives research support from Genentech and is a co-founder and shareholder in CDI Labs, LLC.

References:

  1. Blanco-Centurion Carlos, Luo Siwei, Vidal-Ortiz Aurelio, Swank Colby, and Shiromani Priyattam J.. 2021. “Activity of a Subset of Vesicular GABA-Transporter Neurons in the Ventral Zona Incerta Anticipates Sleep Onset.” Sleep 44 (6). 10.1093/sleep/zsaa268. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Blattner Timothy, Keyrouz Walid, Chalfoun Joe, Stivalet Bertrand, Brady Mary, and Zhou Shujia. 2014. “A Hybrid CPU-GPU System for Stitching Large Scale Optical Microscopy Images.” In 2014 43rd International Conference on Parallel Processing. IEEE. 10.1109/icpp.2014.9. [DOI] [Google Scholar]
  3. Borbély Alexander. 2022. “The Two-Process Model of Sleep Regulation: Beginnings and Outlook.” Journal of Sleep Research 31 (4): e13598. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Chalfoun Joe, Majurski Michael, Blattner Tim, Bhadriraju Kiran, Keyrouz Walid, Bajcsy Peter, and Brady Mary. 2017. “MIST: Accurate and Scalable Microscopy Image Stitching Tool with Stage Modeling and Error Minimization.” Scientific Reports 7 (1): 4988. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Chometton Sandrine, Barbier Marie, and Risold Pierre-Yves. 2021. “The Zona Incerta System: Involvement in Attention and Movement.” Handbook of Clinical Neurology 180:173–84. [DOI] [PubMed] [Google Scholar]
  6. Chou Xiao-Lin, Wang Xiyue, Zhang Zheng-Gang, Shen Li, Zingg Brian, Huang Junxiang, Zhong Wen, Mesik Lukas, Zhang Li I., and Tao Huizhong Whit. 2018. “Inhibitory Gain Modulation of Defense Behaviors by Zona Incerta.” Nature Communications 9 (1): 1151. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Deschênes Martin, Timofeeva Elena, Lavallée Philippe, and Dufresne Caroline. 2005. “The Vibrissal System as a Model of Thalamic Operations.” Progress in Brain Research 149:31–40. [DOI] [PubMed] [Google Scholar]
  8. Duhart José Manuel, Inami Sho, and Koh Kyunghee. 2023. “Many Faces of Sleep Regulation: Beyond the Time of Day and Prior Wake Time.” The FEBS Journal 290 (4): 931–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Eban-Rothschild Ada, Appelbaum Lior, and Lecea Luis de. 2018. “Neuronal Mechanisms for Sleep/Wake Regulation and Modulatory Drive.” Neuropsychopharmacology: Official Publication of the American College of Neuropsychopharmacology 43 (5): 937–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Fragkouli Apostolia, van Wijk Nicole Verhey, Lopes Rita, Kessaris Nicoletta, and Pachnis Vassilis. 2009. “LIM Homeodomain Transcription Factor-Dependent Specification of Bipotential MGE Progenitors into Cholinergic and GABAergic Striatal Interneurons.” Development (Cambridge, England) 136 (22): 3841–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Freiberg Andrew S. 2020. “Why We Sleep: A Hypothesis for an Ultimate or Evolutionary Origin for Sleep and Other Physiological Rhythms.” Journal of Circadian Rhythms 18 (March):2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Git Kathy C. G. de, Hazelhoff Esther M., Nota Minke H. C., Schele Erik, Luijendijk Mieneke C. M., Dickson Suzanne L., van der Plasse Geoffrey, and Adan Roger A. H.. 2021. “Zona Incerta Neurons Projecting to the Ventral Tegmental Area Promote Action Initiation towards Feeding.” The Journal of Physiology 599 (2): 709–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Gross Stefanie, Garofalo Diana C., Balderes Dina A., Mastracci Teresa L., Dias José M., Perlmann Thomas, Ericson Johan, and Sussel Lori. 2016. “The Novel Enterochromaffin Marker Lmx1a Regulates Serotonin Biosynthesis in Enteroendocrine Cell Lineages Downstream of Nkx2.2.” Development (Cambridge, England) 143 (14): 2616–28. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Han Yong, Shi Yu-Feng, Xi Wang, Zhou Rui, Tan Zhi-Bing, Wang Hao, Li Xiao-Ming, et al. 2014. “Selective Activation of Cholinergic Basal Forebrain Neurons Induces Immediate Sleep-Wake Transitions.” Current Biology: CB 24 (6): 693–98. [DOI] [PubMed] [Google Scholar]
  15. Hao Yuhan, Stuart Tim, Kowalski Madeline H., Choudhary Saket, Hoffman Paul, Hartman Austin, Srivastava Avi, et al. 2024. “Dictionary Learning for Integrative, Multimodal and Scalable Single-Cell Analysis.” Nature Biotechnology 42 (2): 293–304. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Hormigo Sebastian, Zhou Ji, Chabbert Dorian, Sajid Sarmad, Busel Natan, and Castro-Alamancos Manuel. 2023. “Zona Incerta Distributes a Broad Movement Signal That Modulates Behavior.” eLife 12 (December). 10.7554/eLife.89366. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Kashiwagi Mitsuaki, Beck Goichi, Kanuka Mika, Arai Yoshifumi, Tanaka Kaeko, Tatsuzawa Chika, Koga Yumiko, et al. 2024. “A Pontine-Medullary Loop Crucial for REM Sleep and Its Deficit in Parkinson’s Disease.” Cell 187 (22): 6272–89.e21. [DOI] [PubMed] [Google Scholar]
  18. Keene Alex C., and Duboue Erik R.. 2018. “The Origins and Evolution of Sleep.” The Journal of Experimental Biology 221 (Pt 11). 10.1242/jeb.159533. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Kim Dong Won, Duncan Leighton H., Xu Zheng, Chang Minzi, Sejer Sara, Terrillion Chantelle E., Kanold Patrick O., Place Elsie, and Blackshaw Seth. 2025. “Decoding Gene Networks Controlling Hypothalamic and Prethalamic Neuron Development.” Cell Reports 44 (6): 115858. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Kim Dong Won, Liu Kai, Wang Zoe Qianyi, Zhang Yi Stephanie, Bathini Abhijith, Brown Matthew P., Lin Sonia Hao, et al. 2021a. “Gene Regulatory Networks Controlling Differentiation, Survival, and Diversification of Hypothalamic Lhx6-Expressing GABAergic Neurons.” Communications Biology 4 (1): 95. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. ——— 2021b. “Gene Regulatory Networks Controlling Differentiation, Survival, and Diversification of Hypothalamic Lhx6-Expressing GABAergic Neurons.” Communications Biology 4 (1): 95. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Kim Dong Won, Washington Parris Whitney, Wang Zoe Qianyi, Lin Sonia Hao, Sun Changyu, Ismail Basma Taleb, Wang Hong, Jiang Lizhi, and Blackshaw Seth. 2020. “The Cellular and Molecular Landscape of Hypothalamic Patterning and Differentiation from Embryonic to Late Postnatal Development.” Nature Communications 11 (1): 4360. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Kroeger Daniel, Absi Gianna, Gagliardi Celia, Bandaru Sathyajit S., Madara Joseph C., Ferrari Loris L., Arrigoni Elda, et al. 2018. “Galanin Neurons in the Ventrolateral Preoptic Area Promote Sleep and Heat Loss in Mice.” Nature Communications 9 (1): 4129. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Lee Sang Soo, Liu Qiang, Cheng Alexandra H. R., Kim Dong Won, Boudreau Daphne M., Mehta Anuradha, Keles Mehmet F., et al. 2025. “Sleep Need-Dependent Plasticity of a Thalamic Circuit Promotes Homeostatic Recovery Sleep.” Science (New York, N.Y.) 388 (6753): eadm8203. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Lemons Abigail, Saré R. Michelle, and Smith Carolyn Beebe. 2018. “Chronic Sleep Deprivation in Mouse Pups by Means of Gentle Handling.” Journal of Visualized Experiments: JoVE, no. 140 (October). 10.3791/58150. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Liu Kai, Kim Juhyun, Kim Dong Won, Zhang Yi Stephanie, Bao Hechen, Denaxa Myrto, Lim Szu-Aun, et al. 2017a. “Lhx6-Positive GABA-Releasing Neurons of the Zona Incerta Promote Sleep.” Nature 548 (7669): 582–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. ——— 2017b. “Lhx6-Positive GABA-Releasing Neurons of the Zona Incerta Promote Sleep.” Nature 548 (7669): 582–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Li Zhuoliang, Rizzi Giorgio, and Tan Kelly R.. 2021. “Zona Incerta Subpopulations Differentially Encode and Modulate Anxiety.” Science Advances 7 (37): eabf6709. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Londei Fabrizio, Arena Giulia, Ferrucci Lorenzo, Russo Eleonora, Ceccarelli Francesco, and Genovesio Aldo. 2024. “Connecting the Dots in the Zona Incerta: A Study of Neural Assemblies and Motifs of Inter-Area Coordination in Mice.” iScience 27 (1): 108761. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Mastracci Teresa L., Lin Chyuan-Sheng, and Sussel Lori. 2013. “Generation of Mice Encoding a Conditional Allele of Nkx2.2.” Transgenic Research 22 (5): 965–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Ma Xiaokuang, Chen Peng, Wei Jing, Zhang John, Chen Chang, Zhao Hanqiu, Ferguson Deveroux, McGee Aaron W., Dai Zhiyu, and Qiu Shenfeng. 2024. “Protocol for Xenium Spatial Transcriptomics Studies Using Fixed Frozen Mouse Brain Sections.” STAR Protocols 5 (4): 103420. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Mitler M. M., Lund R., Sokolove P. G., Pittendrigh C. S., and Dement W. C.. 1977. “Sleep and Activity Rhythms in Mice: A Description of Circadian Patterns and Unexpected Disruptions in Sleep.” Brain Research 131 (1): 129–45. [DOI] [PubMed] [Google Scholar]
  33. Mitrofanis J. 2005. “Some Certainty for the ‘Zone of Uncertainty’? Exploring the Function of the Zona Incerta.” Neuroscience 130 (1): 1–15. [DOI] [PubMed] [Google Scholar]
  34. Miyazaki Shinichi, Liu Chih-Yao, and Hayashi Yu. 2017. “Sleep in Vertebrate and Invertebrate Animals, and Insights into the Function and Evolution of Sleep.” Neuroscience Research 118 (May):3–12. [DOI] [PubMed] [Google Scholar]
  35. Monosov Ilya E., Ogasawara Takaya, Haber Suzanne N., Heimel J. Alexander, and Ahmadlou Mehran. 2022. “The Zona Incerta in Control of Novelty Seeking and Investigation across Species.” Current Opinion in Neurobiology 77 (December):102650. [DOI] [PubMed] [Google Scholar]
  36. Satija Rahul, Farrell Jeffrey A., Gennert David, Schier Alexander F., and Regev Aviv. 2015. “Spatial Reconstruction of Single-Cell Gene Expression Data.” Nature Biotechnology 33 (5): 495–502. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Sawada Takeshi, Iino Yusuke, Yoshida Kensuke, Okazaki Hitoshi, Nomura Shinnosuke, Shimizu Chika, Arima Tomoki, et al. 2024. “Prefrontal Synaptic Regulation of Homeostatic Sleep Pressure Revealed through Synaptic Chemogenetics.” Science (New York, N.Y.) 385 (6716): 1459–65. [DOI] [PubMed] [Google Scholar]
  38. Schafer Carmen B., and Hoebeek Freek E.. 2018. “Convergence of Primary Sensory Cortex and Cerebellar Nuclei Pathways in the Whisker System.” Neuroscience 368 (January):229– 39. [DOI] [PubMed] [Google Scholar]
  39. Schneider Caroline A., Rasband Wayne S., and Eliceiri Kevin W.. 2012. “NIH Image to ImageJ: 25 Years of Image Analysis.” Nature Methods 9 (7): 671–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Skeldon Anne C., and Dijk Derk-Jan. 2025. “The Complexity and Commonness of the Two-Process Model of Sleep Regulation from a Mathematical Perspective.” Npj Biological Timing and Sleep 2 (1): 24. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Stirling David R., Swain-Bowden Madison J., Lucas Alice M., Carpenter Anne E., Cimini Beth A., and Goodman Allen. 2021. “CellProfiler 4: Improvements in Speed, Utility and Usability.” BMC Bioinformatics 22 (1): 433. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Thomas Christopher W., Guillaumin Mathilde Cc, McKillop Laura E., Achermann Peter, and Vyazovskiy Vladyslav V.. 2020. “Global Sleep Homeostasis Reflects Temporally and Spatially Integrated Local Cortical Neuronal Activity.” eLife 9 (July). 10.7554/eLife.54148. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Vanini Giancarlo, and Torterolo Pablo. 2021. “Sleep-Wake Neurobiology.” Advances in Experimental Medicine and Biology 1297:65–82. [DOI] [PubMed] [Google Scholar]
  44. Venkataraman Archana, Brody Natalia, Reddi Preethi, Guo Jidong, Rainnie Donald Gordon, and Dias Brian George. 2019. “Modulation of Fear Generalization by the Zona Incerta.” Proceedings of the National Academy of Sciences of the United States of America 116 (18): 9072–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Vidal-Ortiz Aurelio, Blanco-Centurion Carlos, and Shiromani Priyattam J.. 2024. “Unilateral Optogenetic Stimulation of Lhx6 Neurons in the Zona Incerta Increases REM Sleep.” Sleep 47 (3). 10.1093/sleep/zsad217. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Walsh L. L., and Grossman S. P.. 1976. “Zona Incerta Lesions Impair Osmotic but Not Hypovolemic Thirst.” Physiology & Behavior 16 (2): 211–15. [DOI] [PubMed] [Google Scholar]
  47. Wang Xiyue, Chou Xiao-Lin, Zhang Li I., and Tao Huizhong Whit. 2020. “Zona Incerta: An Integrative Node for Global Behavioral Modulation.” Trends in Neurosciences 43 (2): 82–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Watson Charles, Lind Christopher R. P., and Thomas Meghan G. 2014. “The Anatomy of the Caudal Zona Incerta in Rodents and Primates.” Journal of Anatomy 224 (2): 95–107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Xu Min, Chung Shinjae, Zhang Siyu, Zhong Peng, Ma Chenyan, Chang Wei-Cheng, Weissbourd Brandon, et al. 2015. “Basal Forebrain Circuit for Sleep-Wake Control.” Nature Neuroscience 18 (11): 1641–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Yaghouby Farid, Donohue Kevin D., O’Hara Bruce F., and Sunderam Sridhar 2016. “Noninvasive Dissection of Mouse Sleep Using a Piezoelectric Motion Sensor.” Journal of Neuroscience Methods 259 (February):90–100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Yu Xiao, Li Wen, Ma Ying, Tossell Kyoko, Harris Julia J., Harding Edward C., Ba Wei, et al. 2019. “GABA and Glutamate Neurons in the VTA Regulate Sleep and Wakefulness.” Nature Neuroscience 22 (1): 106–19. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Zhang Meng, Pan Xingjie, Jung Won, Halpern Aaron R., Eichhorn Stephen W., Lei Zhiyun, Cohen Limor, et al. 2023. “Molecularly Defined and Spatially Resolved Cell Atlas of the Whole Mouse Brain.” Nature 624 (7991): 343–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Zhang Xiaobing, and van den Pol Anthony N. 2017. “Rapid Binge-like Eating and Body Weight Gain Driven by Zona Incerta GABA Neuron Activation.” Science (New York, N.Y.) 356 (6340): 853–59. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Zhao Zheng-Dong, Chen Zongming, Xiang Xinkuan, Hu Mengna, Xie Hengchang, Jia Xiaoning, Cai Fang, et al. 2019. “Zona Incerta GABAergic Neurons Integrate Prey-Related Sensory Signals and Induce an Appetitive Drive to Promote Hunting.” Nature Neuroscience 22 (6): 921–32. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplement 1

Supplemental Dataset 1: Hiplex data for all samples.

The dataset contains all of the cells identified in the zona incerta across all experimental conditions and genotypes. Each cell is delineated by a unique cell ID name and contains corresponding information regarding experimental group, genotype, location, markers expressed, and cell type.

media-1.xlsx (31.7MB, xlsx)
2
NIHPP2025.07.30.667705v1-supplement-2.pdf (20MB, pdf)

Articles from bioRxiv are provided here courtesy of Cold Spring Harbor Laboratory Preprints

RESOURCES