Daily rhythms in mammals are programmed by the circadian timekeeping system (Mohawk et al., 2012), which ensures that behavior and physiology are well matched to environmental conditions over the solar day. In nearly every biological system, cell physiology is modulated by autoregulatory genetic feedback loops controlling circadian rhythms in gene expression (Buhr and Takahashi, 2013). At the system level, clock tissues in the body are coordinated by a central clock in the suprachiasmatic nucleus (SCN), which is necessary for daily rhythms in behavior and physiology (Hastings et al., 2018). As the central pacemaker, the SCN processes photic inputs from the retina, sustains tissue-level rhythms through local communication, and provides outputs to coordinate cellular rhythms in downstream targets. Neural network mechanisms that support SCN timekeeping are essential for achieving internal and external coordination of the circadian system in an ever-changing environment.

The SCN is a heterogenous network of cellular clocks that displays self-sustained circadian rhythms in metabolism, electrical activity, gene/protein expression, and peptide release (Hastings et al., 2018). SCN neurons express the neurotransmitter GABA and can be distinguished into different subpopulations based on peptide expression (Antle et al., 2003). Two types of SCN neurons have been studied in mammals in depth (Abrahamson and Moore, 2001; Moore et al., 2002; Ono et al., 2021). Located in the SCN shell and core respectively, AVP and VIP neurons provide network signals that regulate daily rhythms in behavior and physiology (Vosko et al., 2007; Kalsbeek et al., 2010; Ono et al., 2021). In addition to regional patterns of peptide expression, SCN neurons display cellular rhythms with spatial gradients that repeat across the network each circadian cycle (Hamada et al., 2004; Evans et al., 2011; Enoki et al., 2012; Brancaccio et al., 2013). Spatial gradients in clock function are stereotyped across individual animals, are evident in a variety of cellular processes, and can be modulated by experience (Inagaki et al., 2007; Evans et al., 2013). How neural identity maps onto differences in cellular function in the SCN network is a key question in the field.

The importance of the SCN clock during adulthood is well established, but the process by which SCN circuits form is not fully understood (Landgraf et al., 2014; Bedont and Blackshaw, 2015; Carmona-Alcocer et al., 2020). Across mammalian species, SCN neurogenesis occurs over the third to fourth quarter of gestation (Shimada and Nakamura, 1973; Altman and Bayer, 1978; Davis et al., 1990; Antle et al., 2005; Kabrita and Davis, 2008). The onset of daily rhythms in SCN activity has been detected as early as the end of neurogenesis and as late as the first few days after birth (Reppert, 1992; Shimomura et al., 2001; Sladek et al., 2004; Ohta et al., 2006; Wreschnig et al., 2014; Carmona-Alcocer et al., 2018). Despite these early milestones, postnatal development is critical for SCN circuit formation (Landgraf et al., 2014; Bedont and Blackshaw, 2015; Carmona-Alcocer et al., 2020). Both Avp and Vip transcripts are detected in the mouse SCN during late embryonic development (Okamura et al., 1983; Vandunk et al., 2011), but transcript and peptide levels increase over the first 3 weeks after birth (Hyodo et al., 1992; Ban et al., 1997; Herzog et al., 2000). Previous work suggests that the roles of AVP and VIP in the regulation of SCN function vary over development (Wreschnig et al., 2014; Ono et al., 2016; Carmona-Alcocer et al., 2018; Mazuski et al., 2020), but how these peptide circuits mature remains unclear.

One outstanding question concerns spatial patterning of SCN circuits during development. Spatiotemporal gradients in SCN neurogenesis have been reported, with SCN core neurons appearing before those in the SCN shell in mice, rats, and hamsters (Altman and Bayer, 1978; Davis et al., 1990; Antle et al., 2005; Kabrita and Davis, 2008). In the mouse, SCN shell neurons are generated in the middle-posterior regions before those in the anterior pole (Okamura et al., 1983; Kabrita and Davis, 2008). These studies suggest that SCN neurons in different regions of the network develop at different times, but it remains unclear if spatial patterning occurs for other milestones in cellular development (e.g., differentiation). Interestingly, previous work suggests that the onset of Vip transcription occurs in two distinct subclusters of SCN neurons that differ in spatial location and cellular function in adulthood (Ban et al., 1997). One obstacle in understanding SCN peptide development is that rhythms in SCN transcripts can change as the network matures (Isobe and Muramatsu, 1995; Ban et al., 1997; Shimomura et al., 2001; Houdek and Sumova, 2014). The resulting need to conduct a circadian time course at each developmental age has limited insight into spatial patterning during SCN development.

Here we use a genetic approach to test if SCN neurons display spatial patterning of peptide transcription during development. This approach uses Cre to permanently label cells with a fluorescent reporter at the time of Avp and Vip transcription (Harris et al., 2014; Taniguchi, 2014), thus circumventing the need to conduct a circadian time course to detect expression of the peptide itself. Using this genetic approach, we tracked Avp and Vip transcription across the entire SCN at key stages of pre-and post-natal development. We find that genetically labeled cells in each peptide class appear in spatially distinct subclusters over development. In addition, we find that biological sex influences developmental patterning of Avp and Vip labeling in a manner that differs for each SCN peptide class. Collectively, these data suggest that SCN neurons can be distinguished into further subclasses based on developmental patterning of neuropeptide transcription.

Hypothalamus anatomy is conserved across vertebrates, guided by molecular mechanisms that determine nuclei that contain a large diversity of neuron subtypes (Xie and Dorsky, 2017; Benevento et al., 2022). Relative to early induction, less is known about how these peptide circuits are built and remodeled. In the SCN, AVP and VIP neurons regulate the timing of sleep, stress, and reproductive rhythms. How peptide circuits in the SCN network are patterned over development may have profound impacts on clock function in adulthood. Using a genetic approach to track SCN development of peptide circuits, our results suggest that SCN patterning varies by cell type, regional subcluster, and sex. The genetic and/or hormonal factors that guide spatial patterning of SCN peptide circuits warrant further research.

Genetic labeling provides insight into cellular appearance over development without the need for surgical or chemical interventions that could interfere with gestation and rearing, but this approach is not without limitations. Both Avp-tdT and Vip-tdT expression were Cre-dependent consistent with previous work describing these genetic models (Harris et al., 2014; Taniguchi, 2014). We find that 70% of Avp-tdT+ neurons were AVP-IHC+, but less than 30% of Vip-tdT+ neurons were VIP-IHC+. Low VIP co-expression could reflect transient Vip transcription over development in a large subset of these cells, which would suggest that the VIP cell population may expand and contract over development. However, this observation could also reflect threshold limits of IHC and/or expression of VIP-related peptides that are not recognized by the antibody used here (Lee et al., 2013; Southey et al., 2014). On the other hand, over 80% of VIP+ cells were labeled with Vip-tdT, but only 43% of AVP+ cells were labeled by Avp-tdT. This has been noted in previous work using this genetic model (Jamieson, 2020), and our data indicate that the dorsal region of the middle SCN displays the highest rate of AVP and Avp-tdT discordance. Failure of Cre-mediated recombination may reflect cellular variation in epigenetic landscape or genetic history (i.e., loss of Cre during cell division), as suggested in previous work (Jamieson, 2020). Interestingly, low co-expression of AVP and Avp-tdT labeling also occurs in this genetic model after adult-specific viral transduction (Jamieson, 2020), which suggests that this observation is not a developmental artifact. Further, we used colchicine to visualize total protein expression over the daily cycle, yet our estimates of colocalization are similar to this previous work (Jamieson, 2020). Collectively, these data indicate that each mouse model used here does not fully capture peptide expression in adulthood, thus limiting the ability to comprehensively map each peptide population during development. However, these validation data also provide an interesting complement to our developmental results by suggesting that neurons in each peptide class may be divided into subclusters. Another known caveat of Cre models is that the transgene can interfere with native peptide expression (Cheng et al., 2019; Joye et al., 2020; Rohr et al., 2020). Importantly, peptide levels in heterozygous Avp-tdT and Vip-tdT mice do not differ from wildtype mice during early development, and circadian behavior does not differ between these two groups during adulthood (Joye et al., 2020; Rohr et al., 2020). However, it is difficult to dismiss that a non-significant decrease in peptide expression could alter SCN patterning. With these caveats in mind, we decided to employ this genetic approach to study SCN peptide development because it avoids the need to conduct a circadian time course at every age.

Consistent with previous work, we find developmental differences in the appearance of SCN neurons in these two peptide classes. Avp-tdT+ and Vip-tdT+ neurons were detected as early as E18, and cell number for each peptide class increased over the first 10 days after birth. Our results align well with previous work characterizing peptide development, despite the likely delays between transcription onset and tdT labeling. Avp and Vip transcripts are first detected in the mouse SCN at E17-18 and E18-19, respectively (Vandunk et al., 2011). AVP and VIP peptide levels increase over the first 2 days after birth (Hyodo et al., 1992; Carmona-Alcocer et al., 2018). SCN AVP cell numbers are stable after P06, but AVP peptide levels continue to increase from P06-P30 (Herzog et al., 2000), which would not be captured with the present approach. Interesting, VIP cell numbers increase between P06 and P30 (Herzog et al., 2000), which we also observed in the present work. It remains unclear what molecular factors drive ontogenetic patterning in these two different SCN peptide classes. The expression of transcription factors during embryogenesis (e.g., Lhx1, Shh, Six3, Six6) is important in early SCN specification (Bedont and Blackshaw, 2015; Carmona-Alcocer et al., 2020). Lhx1 and Foxd1 deletion decreases both Avp and Vip expression (Vandunk et al., 2011; Newman et al., 2018), suggesting common genetic programs direct cellular differentiation in both classes. Differences in the timing of developmental patterning across these two cell types may be linked to intrinsic and/or extrinsic factors. Over the first week of life, terminal differentiation, synaptogenesis, gliogenesis, and retinal innervation occurs in the SCN (Bedont and Blackshaw, 2015; Carmona-Alcocer et al., 2020), and later maturation of the VIP population may be linked to postnatal maturation of retinal inputs (McNeill et al., 2011). Both AVP and VIP influence SCN circuit formation and function during development (Ono et al., 2016; Bedont et al., 2018; Mazuski et al., 2020); thus, the timing and patterning of these peptides may have important consequences for pacemaker function.

Notably, we find that SCN peptide classes can be further subdivided based on spatial patterning during development. For Avp-tdT+ neurons, we find that there is a posterior–anterior gradient when analyzed by anatomical division based on slice position. Consistent with these results, k means clustering detected that the posterior Avp-tdT+ cluster matured faster than other clusters in each sex, with development of one of the more anterior Avp-tdT+ clusters delayed in a sex-influenced manner. For Vip-tdT+ neurons, we observed a rapid increase of cells in the posterior cluster. Consistent with this result, VIP neurons have been reported to increase in the middle and posterior SCN over P06 to P30 in the mouse SCN (Herzog et al., 2000). In rats, two developmental waves of Vip expression have been reported, with Vip transcription occurring in medial SCN cells earlier than lateral SCN cells (Ban et al., 1997; Kawamoto et al., 2003). These two spatially defined subclusters displayed different patterns of clock gene expression and photic sensitivity in adulthood (Ban et al., 1997; Kawamoto et al., 2003). Further, VIP neurons in adulthood can be divided into two subpopulations based on electrical firing (Mazuski et al., 2018) and Grp expression (Todd et al., 2020; Wen et al., 2020). Interestingly, recent work in the mouse has found two subsets of Avp cells that differ in the expression of Cck or Nms (Moffitt et al., 2018). Whether functional subclasses of SCN Vip and Avp cells map onto the regional subclusters found here would be interesting to examine in future work. It is also unclear how the current spatial gradients may relate to those found for SCN neurogenesis (Okamura et al., 1983; Kabrita and Davis, 2008). In the mouse, SCN neurogenesis occurs over embryonic days 11–16 (E11–16), with a peak at E14 (Shimada and Nakamura, 1973; Kabrita and Davis, 2008). SCN core neurons are generated at an earlier age (peak at E12) than shell neurons (peak E13.5) in the mouse (Kabrita and Davis, 2008). In the hamster, AVP neurons are generated over a longer period of gestation than VIP and GRP neurons in the SCN core, with posterior-to-anterior gradients (Antle et al., 2005). The degree to which onset of neuropeptide expression is timed by neurogenesis and/or extrinsic cues present in the microenvironment (Xie and Dorsky, 2017; Benevento et al., 2022) remains an open question.

Last, our results suggest that developmental patterning of peptide circuits is influenced by sex. The male SCN displayed postnatal increases in Avp-tdT+ and Vip-tdT+ cell number and density that were not maintained into adulthood. The number of Vip-tdT+ cells at P05–P10 exceeded adult levels in male SCN by 20%, and the number of Avp-tdT+ neurons at P05 exceeded adult levels in male SCN by 14%. Given that tdT labeling is permanent, this observation suggests a loss of SCN cells in males. The majority of SCN apoptosis occurs over P01-P07 in mice, but an estimated 20% of cells are lost between P07 and adulthood (Ahern et al., 2013; Bedont et al., 2014; Mosley et al., 2017). In contrast, the female SCN displayed a more linear patterning of Avp-tdT+ and Vip-tdT+ cell appearance, with increasing numbers of both cell types between P10 and P84. In addition, the more anterior Avp-tdT+ cluster that matured last differed by sex, with the lateral cluster appearing last in females and the medial cluster appearing last in males. Interestingly, the lateral cluster had significantly more Avp-tdT+ neurons in adult females than males due to the increase in cell number after P10. Likewise, the Vip-tdT+ cluster that matured last differed by sex, again being the lateral cluster in females and the medial cluster in males. Cell number in the lateral Vip-tdT+ cluster also differed by sex due to post-P10 increases in cell numbers in females. Collectively, these data indicate that SCN development is not complete by P10, raising the possibility that puberty represents another time of SCN development (Bakker and Baum, 2008).

Whether these sex differences are driven by genetic and/or hormonal factors remains to be tested, but it is tempting to speculate that sex steroids are organizing development of SCN circuits. The critical period in the mouse is E18-P01, with testosterone levels decreasing rapidly at birth and the sensitive period in females extending to P07 (McCarthy et al., 2018). The mouse SCN expresses receptors for sex steroids that are regionally clustered in adulthood (Joye and Evans, 2021), potentially contributing to the spatial gradients in peptide development observed here. Sex differences in SCN neurogenesis have been reported in mice (Abizaid et al., 2004). Specifically, females display more SCN neurogenesis at E18, and testosterone administration to pregnant dams reduces SCN neurogenesis at this late stage of gestation (Abizaid et al., 2004). This suggests that neurogenesis closes earlier in males (Shimada and Nakamura, 1973; Kabrita and Davis, 2008) due to sex steroid signaling. Last, the peak in SCN apoptosis occurs at P03 in males and P05 in females with equivalent postnatal SCN volume (Ahern et al., 2013). To our knowledge, sex steroid receptor expression over early SCN development has not been examined in the mouse, but androgen receptors are expressed later in life in the slow-maturing diurnal rodent Octogon degus (Lee et al., 2004). Overall, our work suggests that SCN shape and peptide expression is influenced by sex, as reported in humans (Swaab et al., 1985, 1994). Future work is warranted to further explore how adult clock function in both sexes is influenced by SCN patterning during development and how this process is impacted by postnatal conditions, such as light exposure (Cambras et al., 1998, 2015; Ohta et al., 2006; Ciarleglio et al., 2011; Ono et al., 2013; Madahi et al., 2018).

A central question here concerned the spatial patterning of SCN maturation, which we have represented in 3D maps for two different peptide classes at critical developmental time points spanning mid-embryonic age to adulthood in both sexes. Both AVP and VIP neurons provide local and long-range cues that regulate daily rhythms (Vosko et al., 2007; Kalsbeek et al., 2010; Mieda et al., 2016; Rohr et al., 2020; Shan et al., 2020). In addition to shell-core compartmentalization, cellular differences in clock function are also evident across the anteroposterior axis (Hamada et al., 2004; Yan et al., 2007; Evans et al., 2011; Yoshikawa et al., 2015, 2017). Do regional and sex differences in peptide development relate to differences in cellular clock function in adulthood? Does the spatial patterning of the SCN circuit extend to other developmental milestones (e.g., gliogenesis, axonal projections)? Do sexual dimorphisms in spatial patterning relate to sex differences in peptide expression and clock function in adulthood? More broadly, how do changes in the conditions present during the postnatal SCN developmental period influence circuit formation to modulate adult clock function? Future work investigating these questions may provide insight into the long-lasting effects of perinatal light exposure on health (Torrey et al., 2000; Madahi et al., 2018).

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The animal study was reviewed and approved by the Institutional Animal Care and Use Committees at Marquette University.

JAE wrote the manuscript with contributions and approval of submitted version by all authors. All authors contributed to experimental design, performed research and analyzed data.

This work was supported by the National Institutes of Health, R01GM143545.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.