UBR4/POE facilitates secretory trafficking to maintain circadian clock synchrony

To investigate the role of UBR4 in the murine SCN, we conditionally ablated the Ubr4 gene in GABAergic neurons by breeding Ubr4^fl/fl mice, which possess loxP sites flanking exon 1, with those that express the Cre recombinase under the control of the vesicular GABA transporter (Vgat) gene (Vgat-Cre). Given that all SCN neurons are GABAergic, the resulting Vgat-cre;Ubr4^fl/fl mice (hereafter referred to as Ubr4 cKO) are devoid of Ubr4 expression in neurons throughout the SCN, as shown by in situ hybridization (Fig. 1a). Western blot analysis confirmed the absence of UBR4 protein of the expected size, ~570 kDa, in the SCN of Ubr4 cKO mice (Fig. 1b). Interestingly, several bands between 100 and 400 kDa that were present in Ubr4 wild-type (control) SCN were not observed in the Ubr4 cKO samples (Fig. 1b), suggesting the existence of multiple protein isoforms of UBR4.

Next, we examined the effects of Ubr4 ablation on circadian rhythms of behavior by analyzing wheel-running activities of Ubr4 cKO mice and control animals (Ubr4^fl/fl and Vgat^cre/+ mice) under a fixed 12-h light:12-h dark (12:12 LD) schedule followed by constant darkness (DD) (Fig. 1c). Ubr4 cKO mice were similar to controls with respect to period, amplitude, and activity levels under LD and DD (Fig. 1c–f and Supplementary Table 1). Based on our previous observation that UBR4 is a light-inducible protein¹³, we assessed the ability of our mutant mice to acutely reset their clock phase in response to nocturnal light (Fig. 1c). A brief light pulse in the early subjective night (circadian time [CT] 15) triggered a phase delay of similar magnitude between Ubr4 cKO mice and controls (Fig. 1g). These data indicate that UBR4 is dispensable for light-induced phase shifts and period length determination under DD.

To test whether UBR4 is required for the disruptive and period-lengthening effects of constant light (LL), we exposed mice to a long-term LL paradigm in which the light intensity was increased in a stepwise fashion every 2–3 weeks (5, 10, 20, 40, 80, and 120 Lux) (Fig. 1h). Only 64% (7/11) of Ubr4 cKO mice were rhythmic at 80 Lux, in contrast to control mice that were all rhythmic (11/11) at that light intensity (Supplementary Table 1). Rhythmic control and Ubr4 cKO mice displayed increasingly longer free-running rhythms as the light intensity increased (Fig. 1h, i). However, the period of Ubr4 cKO mice under 10, 20, and 40 Lux LL was significantly shorter than that of controls (Fig. 1h, i). The amplitude of rhythms and daily activity levels in LL were also reduced in Ubr4 cKO mice (Fig. 1h, j–l). Collectively, these data suggest that UBR4 buffers the SCN clock against the disruptive effects of constant light.

Behavioral rhythmicity under LL has been shown to be positively correlated with intercellular synchrony in the SCN²⁵. Hence, the LL-induced behavior of Ubr4 cKO mice may be a consequence of altered SCN synchrony, which should also affect their ability to entrain to jetlag schedules. When subjected to an abrupt 7-h advance or delay of the LD schedule, both Ubr4 cKO and control mice were able to re-entrain to the shifted cycle (Fig. 2a). However, Ubr4 cKO mice re-entrained to an advanced LD schedule at an accelerated rate when compared to controls, requiring fewer days to reach stable entrainment (Fig. 2b, c). In contrast, the kinetics of re-entrainment to a delayed LD schedule were not different between genotypes (Fig. 2d, e).

Next, we investigated the consequence of Ubr4 ablation on an animal’s ability to entrain to repeated shifts of the LD schedule. The ChrA^6/2 chronic jetlag paradigm, in which the LD schedule is advanced by 6 h every 2 days, has been shown to trigger a state of internal desynchrony in a subset of C57Bl/6 mice, resulting in the emergence of two activity rhythms: a short-period component of ~21 h that signifies entrainment to the LD schedule and a long-period component of >24 h that is not entrained²⁶. Under the ChrA^6/2 protocol, 44% (4/9) of control mice displayed a single, entrained component with a period of 20.98 ± 0.02 h, and the remaining 56% (5/9) showed desynchronized behavior with two rhythmic components (period: 21.00 ± 0.00 h, 24.55 ± 0.20 h) (Fig. 2f, g). In contrast, 85% (11/13) of Ubr4 cKO mice had a single, entrained component (period: 21.00 ± 0.00 h); the remaining 15% (2/13) were desynchronized but with a “long-period” component that was less than 24 h (period: 21.00 ± 0.00 h, 23.42 ± 0.34 h) (Fig. 2f, g). The ChrA^6/2 paradigm did not trigger a change in the free-running period of Ubr4 cKO mice relative to controls upon their release into DD (Fig. 2h). However, the increase in circadian amplitude observed in control mice following exposure to the ChrA^6/2 paradigm was absent in Ubr4 cKO animals (Fig. 2i). Altogether, these data strongly indicate that the absence of UBR4 in the SCN increases the resistance to jetlag.

SCN neurons secrete a variety of neuropeptides within the tissue to regulate intercellular communication and synchrony⁸. The unusual circadian behavior of Ubr4 cKO mice suggested a potential defect in the synchronization process and motivated us to examine the expression of two prominent neuropeptides in the SCN. Vasoactive intestinal neuropeptide (VIP) is synthesized by ventrolateral (core) SCN neurons, which send dense projections to the dorsomedial (shell) SCN region (Fig. 3a)^27,28. In contrast, shell SCN neurons express arginine vasopressin (AVP) and project primarily to neurons within the same compartment (Fig. 3a)^27,28. Analysis of VIP and AVP immunoreactivity (IR) revealed no difference in mean expression intensity in the whole SCN of Ubr4 cKO mice relative to controls (Fig. 3b–d, i). However, the spatial distribution of both neuropeptides was noticeably altered in the mutant SCN in a manner that was independent of time-of-day (Fig. 3b, c). In control mice, VIP-IR was most prominent in the axons projecting to the shell, in contrast with Ubr4 cKO mice in which VIP-IR was highly concentrated in the cell soma situated in the core SCN (Fig. 3b). These distribution patterns were reflected in the mean expression intensity measurements in the core and shell regions, as well as the number of cell bodies expressing high levels of VIP (Fig. 3e–h). Similarly, Ubr4 cKO mice demonstrated intense AVP-IR in the cell soma along with a greater number of cell bodies with high AVP expression (Fig. 3c, j–l). Importantly, these effects could not be attributed to increased expression at the transcript level, as Vip mRNA was unaltered and Avp mRNA was actually reduced in the SCN of Ubr4 cKO mice (Supplementary Fig. 1a). These data suggest that UBR4 is critical for the proper trafficking of VIP and AVP neuropeptides within SCN neurons.

To determine whether Ubr4 ablation impacts the molecular clock, we analyzed the circadian expression profile of PERIOD2 (PER2) protein in the SCN of Ubr4 cKO and control mice (Fig. 3m). Relative to controls, Ubr4 cKO mice exhibited damped oscillations of PER2, as evident by the reduction in mean PER2-IR intensity and the number of cells with high PER2 expression in the SCN at the peak of the rhythm (Fig. 3n, o). Along these lines, knocking down Ubr4 expression in murine neuroblastoma (Neuro-2a) cells also reduced the abundance of ectopically expressed V5-tagged PER2 (Supplementary Fig. 1b). However, these differences at the protein level are not due to altered transcription or mRNA stability, as mRNA abundance of Per2, as well as of other clock genes, in the SCN at CT 9 was comparable between Ubr4 cKO and control mice (Supplementary Fig. 1a). Collectively, these findings reveal that UBR4 is crucial for neuropeptide trafficking within the SCN and modulation of PER2 protein rhythms. These functions may underlie the resistance of Ubr4 cKO mice to experimental jetlag and their susceptibility to the disruptive effects of constant light.

To determine whether UBR4 has a conserved function in circadian rhythms that extends beyond mammals, we investigated the effects of knocking down purity of essence (poe), the Drosophila homolog of UBR4, on invertebrate clock timing. Poe is widely expressed in the adult fly brain including the main pacemaker neurons, the pigment-dispersing factor (PDF)-positive small and large ventral lateral neurons (s-LN_v and l-LN_v, respectively) (Fig. 4a and Supplementary Fig. 3a). To assess the function of poe in circadian timekeeping, we used the clock-specific drivers, timeless (tim)-GAL4 and Pdf-GAL4, to induce the expression of poe^RNAi either in all clock neurons or specifically in the s-LN_v and l-LN_v neurons, respectively. The efficacy of the poe^RNAi line that we used in knocking down poe expression was confirmed by in situ hybridization analysis of s-LN_v and l-LN_v neurons from Pdf>Dcr2; poe^RNAi flies and by reverse transcription–quantitative polymerase chain reaction (qRT-PCR) analysis of whole-brain tissues from flies with pan-neuronal expression of poe^RNAi using the elav-GAL4 driver (Fig. 4a, b).

We assessed the consequence of poe knockdown (KD) in either tim- or Pdf-positive clock cells on circadian rhythms of locomotor activity. Under a 12:12 LD schedule, tim>Dcr2; poe^RNAi flies exhibited advanced evening and no morning anticipatory behavior (Supplementary Fig. 2a–d). Upon release into DD, all tim>Dcr2; poe^RNAi flies showed complete and immediate arrhythmicity, which was further indicated by a reduction in their rhythm strength when compared to UAS-poe^RNAi (+>poe^RNAi) and tim>Dcr2 controls (Fig. 4c, d, Supplementary Fig. 2a, and Supplementary Table 2). In contrast, Pdf>Dcr2; poe^RNAi flies displayed multiple phenotypes under free-running conditions (Fig. 4e and Supplementary Fig. 2e). The majority (72%) of Pdf>Dcr2; poe^RNAi flies, which we have termed “delayed arrhythmic,” remained rhythmic in the first ~3–6 days of DD before transitioning to complete arrhythmicity (Fig. 4e, f). This phenotype is consistent with the reduction in rhythm strength of these flies in the second week of DD (DD 7–12) compared to the first (DD 1–6) (Fig. 4g and Supplementary Table 3). In addition, 16% of Pdf>Dcr2; poe^RNAi flies exhibited immediate arrhythmicity (termed “arrhythmic”) and 12% displayed rhythms with fluctuating period lengths (termed “rhythmic-unstable period”) throughout DD (Fig. 4e, f). Pdf>Dcr2; poe^RNAi flies were indistinguishable from controls in terms of their ability to anticipate the onset of morning and evening under LD conditions (Supplementary Fig. 2e–h).

Collectively, these results reveal that poe, like its mammalian homolog, serves an important role in the control of circadian rhythms. Specifically, poe is critical for the generation or regulation of free-running rhythms in flies.

The behavioral phenotype of Pdf>Dcr2; poe^RNAi flies under DD partially mimics that of Pdf-null (Pdf¹) flies with respect to the delayed arrhythmicity⁴. Given the impaired axonal trafficking of VIP, the vertebrate functional homolog of PDF, in SCN neurons, we speculated that altered PDF trafficking may underlie the deficits in circadian behavior of these flies. To address this, we examined the abundance of PDF protein in cell bodies and axonal projections at CT2 and CT14 (Fig. 5a). The s-LN_v neurons extend projections to the dorsal region of the brain, whereas the l-LN_v neurons project to the ipsilateral and contralateral optic lobes (OL), the latter via the posterior optic tract (POT)^29,30. Pdf>Dcr2; poe^RNAi flies possessed the expected number of PDF-positive small and large LN_v neurons, indicating normal development of these cells (Supplementary Fig. 3b, c). At CT2, PDF expression within the OL, POT, and dorsal projections (DP) was drastically reduced in Pdf>Dcr2; poe^RNAi flies compared to +>poe^RNAi and Pdf>Dcr2 controls (Fig. 5b–d). However, there was no effect of poe KD on PDF expression in the dorsal terminals (DT) at CT2 (Fig. 5e). At CT14, Pdf>Dcr2; poe^RNAi flies continued to exhibit reduced PDF expression in the POT but not in the OL or DP (Fig. 5b–d). Intriguingly, we observed a significant accumulation of PDF in the DT of Pdf>Dcr2; poe^RNAi flies at CT14 when compared to controls (Fig. 5e). In terms of expression in the cell body, PDF levels were significantly greater in the soma of s-LN_v neurons of Pdf>Dcr2; poe^RNAi flies compared to controls, but were not different in the l-LN_vs, regardless of time point (Fig. 5f, g). Pdf mRNA expression in the small and large LN_vs of Pdf>Dcr2; poe^RNAi flies was not markedly different from that of +>poe^RNAi control flies (Supplementary Fig. 3d, e), indicating that post-transcriptional mechanisms underlie the observed changes in PDF distribution and abundance. Moreover, using a GAL4-driven membrane-targeted marker, CD2-HRP, we determined that poe KD in PDF neurons did not affect neuronal morphology, thus ruling out the possibility that altered PDF distribution in the Pdf>Dcr2; poe^RNAi flies was due to structural defects of the axonal projections (Supplementary Fig. 3f).

Next, we examined the effects of poe KD on the rhythmic expression of the core clock protein, dPER, in PDF-positive neurons (Fig. 5h, i and Supplementary Fig. 3g). In the s-LN_vs of control flies, dPER expression was strongly rhythmic with a peak in the late night-to-early day (Fig. 5j). In contrast, the expression of dPER was markedly reduced in the s-LN_vs of Pdf>Dcr2; poe^RNAi flies; furthermore, rhythms were severely damped and peaked in the mid-to-late day (Fig. 5j). dPER levels in the l-LN_vs were not affected by poe KD (Fig. 5k). We also observed drastically reduced and non-rhythmic expression of dPER in the small and large LN_vs of tim>Dcr2; poe^RNAi flies; however, interpretation of these results was complicated by the fact that these flies had fewer LN_vs than tim>Dcr2 and +>poe^RNAi controls (Supplementary Fig. 4). Collectively, these data argue for a conserved role of POE/UBR4 in the regulation of neuropeptide trafficking and PER expression within circadian clock neurons of flies and mice.

To understand the cellular basis for the neuropeptide transport defect arising from Ubr4/poe ablation, we conducted trafficking studies that utilized a human embryonic kidney (HEK) 293 T cell line in which the UBR4 gene had been knocked out (KO) using CRISPR/Cas9 technology²⁴. Immunocytochemistry revealed the punctate distribution of UBR4 protein in the cytoplasm, and Western blot analysis confirmed its absence in our UBR4 KO cell line (Supplementary Fig. 5a,b). UBR4 KO and wild-type (WT) cells were then transfected with a reporter construct expressing neuropeptide Y (NPY) fused to enhanced green fluorescent protein (GFP), and the subcellular localization of NPY-GFP was determined by immunofluorescence (Fig. 6a). NPY-GFP was distributed evenly throughout the cytoplasm of WT cells, whereas in UBR4 KO cells it was largely restricted to the Golgi apparatus, as shown by the high degree of co-localization with the cis-Golgi and trans-Golgi network (TGN) markers, GM130 and p230, respectively (Fig. 6a, c, d and Supplementary Fig. 5c). Additionally, the cis-Golgi and TGN compartments exhibited an abnormally distended morphology in UBR4 KO cells, which was reflected in an increase in their area and aspect ratio (H/W) relative to WT cells (Fig. 6b, e, f and Supplementary Fig. 5d, e). These results suggest that protein trafficking through the Golgi may be delayed or impaired in the absence of UBR4.

Next, we examined the kinetics of neuropeptide trafficking using the RUSH system to synchronize the transport process³¹. This system uses a Streptavidin “hook” to mediate ER retention of a cargo protein (in our case, NPY-GFP) that is fused to Streptavidin-binding peptide (SBP). The addition of biotin initiates the release of the cargo-SBP fusion protein from the hook and subsequent trafficking along the secretory pathway. Using this method, UBR4 WT and KO cells transfected with RUSH plasmids were harvested at several time points following the addition of biotin and processed by immunofluorescence (Fig. 6g). At t = 0’ (no biotin added), NPY-GFP was evenly dispersed in UBR4 WT and KO cells, indicating ER localization (Fig. 6g). ER-to-Golgi transport was accelerated in the absence of UBR4, as shown by the higher level of NPY-GFP expression in the Golgi relative to the rest of the cell and the greater proportion of cells with NPY-GFP expression in the Golgi in UBR4 KO cultures 15 min following the addition of biotin (Fig. 6h, i). Once NPY-GFP was fully trafficked to the Golgi (t = 30’), export from the Golgi was markedly slower in UBR4 KO cells compared to WT, as evident by the more gradual decline in relative NPY-GFP expression in the Golgi from t = 60’ to t = 120’ and the larger proportion of cells that retained NPY-GFP expression within the Golgi at t = 120’ (Fig. 6h, i). To confirm these findings, we combined the RUSH system with live-cell imaging to track the kinetics of NPY-GFP trafficking in individual cells (Fig. 6j and Supplementary Movie 1). The results are consistent with an impairment or delay in Golgi exit in the absence of UBR4, given the persistent presence of NPY-GFP signal within the Golgi of UBR4 KO cells as late as 144 min following addition of biotin (Fig. 6j and Supplementary Movie 1). Analysis of a constitutively secreted cargo, glycosylphosphatidylinositol-anchored enhanced GFP protein (GPI-GFP), using the RUSH assay showed a similar effect whereby ER-to-Golgi transport was faster but Golgi exit was slower in the absence of UBR4 (Supplementary Fig. 5f–h). Interestingly, large intracellular vesicles containing GPI-GFP were prevalent in the cytoplasm of UBR4 KO cells and preceded the appearance of GPI-GFP at the plasma membrane (Supplementary Fig. 5f).

Lastly, we determined whether the effects of UBR4 on cargo export from the Golgi are dependent on the UBR-box domain, which is required for the recognition of N-degrons³². Overexpression of full-length (FL), wild-type UBR4 in UBR4 KO HEK293T cells rescued the Golgi export phenotype, as evident by the broad distribution of NPY-GFP throughout the cytoplasm and reduced localization in the Golgi relative to the rest of the cell (Fig. 6k, l and Supplementary Fig. 5i). Surprisingly, the phenotype was also rescued by overexpression of two mutant forms of UBR4 in which the function of the UBR-box was abrogated: one mutant contained a partial deletion of the UBR-box domain (Del) while the other carried multiple His/Cys→Ala mutations in the UBR-box (Ala) (Fig. 6k, l and Supplementary Fig. 5i). The inefficient export of NPY-GFP from the Golgi of UBR4 KO cells ultimately impacted secretion, as shown by the reduced amount of NPY-GFP in the growth media of UBR4 KO cultures compared to WT controls (Supplementary Fig. 5j).

Collectively, our data strongly suggest that UBR4 has a critical role in controlling the kinetics of protein trafficking through the regulated and constitutive secretory pathways. Importantly, the absence of UBR4 impedes the rate of cargo exit from the Golgi and thus secretion from the cell, effects which we believe can have serious consequences on timely communication between cells within a network.

We next sought to identify the molecular mechanism by which UBR4 promotes the exit of cargo proteins from the Golgi. Based on our hypothesis that UBR4 regulates the abundance of a specific mediator of Golgi export, we used label-free quantitative mass spectrometry (MS) to analyze the proteomes of UBR4 KO and WT HEK293T cells. Out of 4,026 unique proteins that were accurately quantified, we identified 471 differentially expressed proteins (DEPs) (11.7%) based on a BH.q value cutoff of 0.05 (Fig. 7a and Supplementary Data 1). Of these, 191 were upregulated and 280 were downregulated in UBR4 KO cells relative to WT (Fig. 7a and Supplementary Data 1). As expected, Gene Ontology (GO) enrichment analysis of the upregulated and downregulated proteomes revealed significant enrichment of GO terms that are related to intracellular protein transport (Fig. 7c and Supplementary Data 2). The list of significantly enriched biological process (BP) GO terms for the upregulated DEPs included “intra-Golgi vesicle-mediated transport” and “Golgi vesicle transport”, whereas for the downregulated DEPs the list encompassed “regulation of protein targeting” and “positive regulation of intracellular protein transport” (Fig. 7c). Furthermore, the upregulated DEPs were significantly enriched for the cellular component (CC) GO terms “endoplasmic reticulum”, “Golgi apparatus”, and “endocytic vesicle lumen”, while the GO-CC terms that are associated with the downregulated DEPs included “extracellular vesicle”, and “membrane-bound vesicle” (Fig. 7c).

To identify a candidate mediator of the trafficking phenotype in UBR4 KO cells, we carefully evaluated the list of DEPs for proteins that had previously been implicated in Golgi export. CRN7, a protein that binds to the cytosolic side of Golgi complex membranes, has been shown to mediate cargo export from the Golgi and is required for normal Golgi morphology^33–35. Consistent with our quantitative MS data, Western blot analyses revealed a significant downregulation of CRN7 in UBR4 KO HEK293T cells and in SCN tissues extracted from Ubr4 cKO mice (MS: Log₂(fold-change (FC)) of CRN7 = −0.9; HEK293T WB: CRN7 FC = 0.62; SCN WB: CRN7 FC = 0.39) (Fig. 7b, d, e). Importantly, elevating the levels of CRN7 in UBR4 KO cells by transient transfection of a Flag-CRN7 construct rescued the deficit in Golgi export (Fig. 7f and Supplementary Fig. 6a). Compared to pcDNA-transfected controls, UBR4 KO cells expressing Flag-CRN7 showed a marked reduction in the accumulation of NPY-GFP within the cis-Golgi and TGN relative to the rest of the cell (Fig. 7f–h and Supplementary Fig. 6a). The distribution of NPY-GFP in CRN7-overexpressing UBR4 KO cells was indistinguishable from that of WT controls (Fig. 7f and Supplementary Fig. 6a). Overexpression of Flag-CRN7 in WT cells did not affect the localization of NPY-GFP, indicating that levels of endogenous CRN7 in these cells are sufficient to ensure proper neuropeptide trafficking (Fig. 7f and Supplementary Fig. 6a).

It does not appear that UBR4 promotes the accumulation of CRN7 by enhancing its gene transcription or mRNA stability, or by stabilizing the protein. First, transcript levels of CRN7 were not different between UBR4 WT and KO HEK293T cells (Supplementary Fig. 6b). Second, proteasomal inhibition using MG-132 for up to 18 h did not increase the level of CRN7 in UBR4 KO or WT HEK293T cells, consistent with reports that CRN7 is a highly stable protein³⁶ (Supplementary Fig. 6c). Even the combination of proteasomal and lysosomal inhibition, the latter using Bafilomycin A1 and E64d/Pepstatin A, failed to restore CRN7 abundance in UBR4 KO cells to WT levels (Supplementary Fig. 6d). These data suggest that UBR4 may instead be regulating CRN7 abundance at the translational level. Consistent with this idea, polysome profiling indicated that the sedimentation of CRN7 mRNA was shifted towards the lighter density polysome fractions in sucrose density gradient centrifugation in the UBR4 KO HEK293T lysates compared to the WT, while the distribution of GAPDH mRNA was not affected (Supplementary Fig. 6e,f). This result suggests that the rate of translation of CRN7 is selectively reduced in the absence of UBR4. Altogether, our findings show that UBR4 facilitates cargo export from the Golgi complex by positively regulating the abundance of CRN7.

So far, we have used HEK293T cells to identify a mechanism that potentially underlies the neuropeptide trafficking phenotype in central clock neurons of mice and flies. To show unequivocally that CRN7-mediated cargo export from the Golgi is impaired in Ubr4-deficient SCN neurons, we first examined the SCN of Ubr4 cKO mice for evidence of cargo retention within the Golgi using transmission electron microscopy (TEM). Unlike the controls, the SCN of Ubr4 cKO mice was characterized by the frequent presence of distended Golgi—a likely consequence of cargo retention—and a corresponding increase in the width of Golgi cisternae (Fig. 8a, b).

To demonstrate that cargo proteins are retained in SCN neurons in a cell-autonomous fashion, we prepared dissociated SCN neuronal cultures from Ubr4^fl/fl neonates and virally transduced them with Cre recombinase to ablate Ubr4 (Supplementary Fig. 7a). In Cre-transduced cultures, we could easily identify cells with endogenous VIP expression, as VIP was highly localized to the Golgi (Supplementary Fig. 7b). In contrast, we had difficulty finding VIP-expressing neurons in GFP-transduced control cultures, as the VIP signal was very weak and diffuse, presumably as a result of efficient trafficking and secretion (Supplementary Fig. 7b). To quantify VIP localization, we performed dual transduction using adeno-associated viral (AAV1) constructs that expressed Cre recombinase and the murine Vip gene under the control of the human Synapsin (hSyn) promoter (Fig. 8c). Similar to endogenous VIP, ectopically expressed VIP was highly concentrated in the Golgi of Cre-transduced neurons relative to GFP-transduced controls (Fig. 8c, d). Importantly, in triple transduction experiments, ectopically expressed VIP was no longer localized in the Golgi of Cre-transduced neurons when murine CRN7 was overexpressed (Fig. 8e, f and Supplementary Fig. 7a). These results demonstrate that the absence of UBR4 in SCN neurons impairs the export of cargo proteins from the Golgi through a CRN7-dependent mechanism.

Lastly, to determine whether defects in CRN7-mediated protein trafficking in the SCN underlie the resistance of Ubr4 cKO mice to jetlag, we examined the effects of CRN7 overexpression in the ChrA^6/2 jetlag paradigm. Ubr4 cKO mice received bilateral SCN injections of AAV1 constructs encoding CRN7-2A-mCherry under the control of the CMV promoter or a control AAV1-CMV-mCherry vector prior to activity monitoring. Post-mortem analysis revealed robust expression of mCherry in the majority of, but not all, SCN neurons (Supplementary Fig. 7c). All Ubr4 cKO mice (9/9) injected with the mCherry control vector showed a single, entrained component of ~21 h (period: 21.00 ± 0.01 h) under the ChrA^6/2 schedule (Fig. 8g, h). In contrast, only 50% of Ubr4 cKO mice (4/8) injected with the CRN7 vector were entrained to the ChrA^6/2 schedule (period: 21.02 ± 0.02 h), while the remaining 50% (4/8) exhibited desynchronized behavior with two rhythmic components (period: 20.96 ± 0.02 h and 23.69 ± 0.15 h) (Fig. 8g, h and Supplementary Fig. 7d). This dichotomy in the behavioral response to chronic jetlag is reminiscent of the behavior of wild-type mice, although the period of the long component is slightly shorter in the CRN7-overexpressing Ubr4 cKO mice, potentially as a result of the mosaicism in CRN7 transgene expression (Fig. 2g). Collectively, our findings demonstrate that UBR4 regulates the abundance of CRN7 in SCN neurons to facilitate neuropeptide trafficking from the Golgi. This UBR4-CRN7 pathway is crucial for maintaining appropriate behavioral rhythms under chronic jetlag.

Details of mouse strains used in this paper are listed in Supplementary Table. 4

All animal handling and experimental procedures were performed at the University of Toronto Mississauga (UTM) Animal Facility and were approved by the UTM Animal Care Committee, complying with guidelines established by the University of Toronto Animal Care Committee and the Canadian Council on Animal Care. The following mouse strains were purchased from The Jackson Laboratory (Bar Harbor, ME, USA) and bred in-house to generate the appropriate genotypes for this study: homozygous Ubr4^fl/fl mice in which the first coding exon of Ubr4/p600 is flanked by loxP sequences (Ubr4^tm1.2Nkt); homozygous Vgat-IRES-Cre (Vgat^cre/cre) knock-in mice in which the IRES-Cre recombinase cassette is inserted downstream of the stop codon of the endogenous vesicular GABA transporter (Vgat) gene (Slc32a1^tm2(cre)Lowl); and C57Bl/6 J mice. Vgat^cre/cre mice were bred to Ubr4^fl/fl mice, and a breeding colony was maintained by mating Ubr4^fl/fl and Vgat^cre/+;Ubr4^fl/+ animals. Littermate controls were used wherever possible for experiments. Vgat^cre/cre mice were also bred to C57BL/6 J mice to generate Vgat^cre/+ animals. For the acute jetlag (7-h LD advance) and PER2 immunostaining experiments, data from Vgat^cre/+ and Ubr4^fl/fl controls were pooled as they were phenotypically indistinguishable from one another. All other experiments utilized only Ubr4^fl/fl mice as controls. Unless otherwise specified, mice were bred and maintained on a fixed 12-h light:12-h dark (12:12 LD) schedule in which lights on and lights off corresponded to 7 a.m. and 7 p.m. Eastern Standard Time, respectively. The animals were maintained at 40–60% humidity and 20–24 °C. Lighting conditions in animal housing rooms were 100–150 Lux (white light) measured at room floor level. Lighting conditions in behavioral cabinets were as defined in each experiment. Air exchanges in the animal housing rooms and behavioral cabinets were maintained at 15–20 ACH. Vgat^cre/+;Ubr4^fl/fl mice were born at the expected Mendelian ratio, were viable, and appeared grossly normal despite having a slightly reduced body weight compared to their sex-matched littermates at the time of weaning.

Details of fly lines used in this paper are listed in Supplementary Table. 4

All fly strains were reared on food containing agar, glucose, sucrose, yeast, cornmeal, wheat germ, soya flour, molasses, propionic acid, and Tegosept on a 12:12 LD schedule at 25 °C, 50% relative humidity. UAS and GAL4 controls were used as heterozygotes after crossing to w¹¹¹⁸ as the wild-type strain. Unless otherwise stated, male flies were used for all experiments. The genotypes of experimental flies and controls are as follows:

w¹¹¹⁸;UAS-poe^RNAi/+;+/+ (+>poe^RNAi),

UAS-Dcr2; tim-GAL4/+;+/+ (tim>Dcr2),

UAS-Dcr2; tim-GAL4/UAS-poe^RNAi;+/+ (tim>Dcr2; poe^RNAi),

w¹¹¹⁸; Pdf-GAL4/+; UAS-Dcr2/+ (Pdf>Dcr2),

w¹¹¹⁸; Pdf-GAL4/UAS-poe^RNAi; UAS-Dcr2/+ (Pdf>Dcr2; poe^RNAi),

elav-GAL4; UAS-Dcr2/+; +/+ (elav>Dcr2),

elav-GAL4; UAS-Dcr2/UAS-poe^RNAi;+/+ (elav>Dcr2; poe^RNAi),

w¹¹¹⁸; Pdf-GAL4/UAS-CD2-HRP; UAS-Dcr2/+ (Pdf>Dcr2; CD2-HRP),

w¹¹¹⁸; Pdf-GAL4/UAS-poe^RNAi, UAS-CD2-HRP; UAS-Dcr2/+ (Pdf>Dcr2; poe^RNAi, CD2-HRP).

Details of AAV1 constructs used in stereotaxic injections are listed in Supplementary Table. 4

Eight–12-week-old Ubr4 cKO male mice were anesthetized with ~2% (vol/vol) isoflurane in O₂. Toe pinch responses (or the lack of a paw withdrawal reflex) were frequently assessed to ensure that a surgical depth of anesthesia was achieved. A thin layer of eye ointment (Systane Nighttime Lubricant) was applied to the eyes to prevent them from drying. The scalp was cut open and a craniotomy of ~300–500 μm diameter was made with a 0.3 mm Tungsten bur (Gesswein 122–2008). Through this cranial window, a beveled glass pipette (tip size ~30 μm) was inserted into the SCN using the following stereotaxic coordinates (AP: 0.4 mm posterior from bregma, ML: ±0.2 mm from the midline, DV: 5.03 mm from dura with head level), with minor adjustments based on the distance between bregma and lambda. A viral particle suspension of AAV1-CMV-m-CORO7-2A-mCherry or AAV1-CMV-mCherry (control) was injected at a speed of 55 nl/min in a total volume of 350 nl into each unilateral SCN using an UltraMicroPump (UMP3, WPI) to which the glass pipette was attached. The scalp was then sutured with 6–0 silk suture (CP Medical 667S). Mice were set up in wheel cages for activity recordings two weeks after the viral injection. At the end of the ChrA^6/2 schedule, SCN tissues were harvested from the injected mice at 16–20 weeks of age.

Details of software used for behavioral analyses in this paper are listed in Supplementary Table. 4

Unless otherwise specified, at 5–8 weeks of age, male mice were individually housed in running-wheel cages within ventilated, light-tight cabinets under computer-controlled light schedules (Phenome Technologies). Cage-level light intensity was set at 40 Lux (0.5 μE), except where indicated, using white light-emitting diodes (LEDs) installed within the cabinets. Wheel revolutions were recorded and analyzed using Clocklab software (Actimetrics). In all experiments, mice were initially entrained to a 12:12 LD cycle for at least 2 weeks prior to the indicated changes in the light schedule. To assess free-running rhythms, entrained mice were released into constant darkness (DD) for 2 weeks. To determine the effect of a brief, early-night light pulse, mice received a 15-min light pulse (LP: 40 Lux [0.5 μE]) at CT 15 after ~2 weeks in DD. Mice were maintained in DD for an additional 2 weeks following the LP to measure the phase shift through extended fitted regression lines before and after the LP. To determine the effects of constant light (LL), mice were subjected to LL in which the light intensity was increased in a stepwise fashion (5 Lux [0.06 μE], 10 Lux [0.13 μE], 20 Lux [0.25 μE], 40 Lux [0.5 μE], 80 Lux [1 μE], 120 Lux [1.5 μE]) every ~2–3 weeks. For the acute jetlag paradigm, mice were subjected to an abrupt 7-h advance of the 12:12 LD schedule for ~3 weeks, and for some mice this was followed by an abrupt 7-h delay of the LD schedule for an additional 3 weeks. To assess the effects of chronic jetlag (ChrA^6/2), mice were subjected to a 6-h advance in the 12:12 LD schedule every 2 days for 4 weeks, after which they were released into DD. Times of activity onset were generated by Clocklab and corrected where necessary after visual inspection of the actograms. The onset of activity was considered as the first bin where 20% of maximum activity during that cycle was displayed, and the offset was the last 5-minute bin with 5% of peak activity. Depending on the light paradigm assessed, the mean of the total daily wheel revolutions was calculated by taking the average of daily wheel revolutions for 7 (LD, DD) or 12 (LL) consecutive days. The period was measured by fitting a regression line through daily activity onsets or by using the χ² periodogram. The DD period reported in Fig. 1d was measured from the 10 days preceding the CT15 LP. Amplitude was measured by χ² periodogram and Fast Fourier Transformation (FFT). Arrhythmicity in LL was determined by χ² periodogram. The phase angle of entrainment was measured by calculating the difference between the time of activity onset and lights-off for 6 consecutive days.

Locomotor activity of individual male flies was recorded using the Drosophila Activity Monitoring (DAM) System (Trikinetics). Briefly, 7-day-old male flies were individually aspirated into glass tubes containing 4% sucrose/2% bactoagar media on one end and capped with a cotton plug on the other end. The tubes were then placed in DAM monitors in a 25 °C incubator, and the flies were allowed to entrain to a 12:12 LD cycle for ~5 days before being released into DD for an additional 1–2 weeks. Activity data were recorded for individual flies as the number of infrared-beam crosses per minute (1-min bins). The fly toolbox⁵⁶ implemented in Matlab (Mathworks) was used to generate actograms, and Clocklab software (Actimetrics) was used to analyze rest:activity rhythms. The free-running period under DD was analyzed using χ² periodogram, and FFT power was used to measure the rhythm strength of individual flies. A fly was deemed to be rhythmic if the χ² periodogram showed a peak above the 95% confidence interval and the FFT power was >0.01. Behavioral scoring (Fig. 4f) was achieved using the abovementioned criteria as well as by visual assessment of the actograms by an observer blind to the genotype of the flies. LD activity line plots (Supplementary Fig. 2c, d, g, h) were generated by averaging the minute-by-minute activity of individual flies during the last 3 days of LD, converting the activity data of individual flies into 1 h bins, and then calculating the average activity of flies within each genotype. For an optimal depiction of LD anticipation, the activity levels were normalized to the maximum activity level in each genotype and presented as “relative activity.” LD anticipation indices (AI) were calculated as follows: Morning AI = “sum of activity in the 3 h preceding lights-on/sum of activity in the 6 h preceding lights-on,” and evening AI = “sum of activity in the 3 h preceding lights-off/sum of activity in the 6 h preceding lights-off.” AI values that are significantly >0.5 represent anticipatory behavior⁴⁴.

Unless otherwise specified, mice (both sexes, 5–8 weeks of age) were maintained on a fixed 12:12 LD schedule and released into DD for two consecutive cycles prior to further treatment and tissue harvest on day 3 of DD. Mice were killed by cervical dislocation and their brains were rapidly dissected under dim red light. Brains were sectioned in ice-cold oxygenated media or diethyl pyrocarbonate-treated phosphate-buffered saline (PBS), pH 7.4 (for qRT-PCR and RNAscope experiments) with an oscillating tissue slicer (Electron Microscopy Sciences) to obtain an 800-μm-thick coronal slice containing the SCN. For immunostaining experiments, tissue slices were fixed in 4% paraformaldehyde (PFA) in PBS (pH 7.4) for 6 h at room temperature (RT), cryoprotected in 30% sucrose in PBS at 4 °C overnight (O/N), cut into 30-μm thin sections using a freezing microtome (Leica Microsystems), and stored in 30% sucrose at 4 °C until further use. For electron microscopy, the 800-μm coronal slice was fixed in 4% PFA and 1% glutaraldehyde in PBS (pH 7.2) for 1 h at RT, followed by an O/N incubation in fixative at 4 °C. For Western blotting and qRT-PCR experiments, the 800-μm coronal slice was placed on a dry ice-chilled glass slide and the SCN was microdissected using a sharp scalpel blade. SCN tissues were frozen immediately on dry ice and stored at –80 °C until further processing. For Western blotting and qRT-PCR experiments, each sample represented SCN tissues from 1 or 2 mice, respectively. For RNAscope in situ hybridization (ISH) experiments, tissue slices were fixed in 4% PFA in RNase-free PBS (pH 7.4) for 24 h at 4 °C, cryoprotected in 30% sucrose in PBS at 4 °C O/N, and then frozen in Tissue Tek Optimal Cutting Temperature (OCT) embedding media with dry ice. Tissue blocks were cut into 14-µm thin sections using CryoStar NX50 Cryostat (ThermoFisher Scientific) and thaw-mounted on SuperFrost Plus slides. Slides were air-dried for 2 h and stored at −80 °C until further use.

For adult fly brain dissections, all flies were maintained under a 12:12 LD cycle at 25 °C, released into DD for two consecutive cycles (if required), and dissected at the indicated Zeitgeber time (ZT) or CT (on day 3 of DD). Adult flies were anesthetized with CO₂, pinned down, and dissected in a dish containing ice-cold PBS (pH 7.4, RNase-free). For qRT-PCR experiments, dissected brains from ~2–4-day-old female flies were immediately transferred to cell lysis buffer for RNA isolation and stored at −80 °C until further processing. Each sample contained a pool of ~15 brains. For ISH experiments, dissected whole brains from ~17–18-day-old male and female flies were fixed in 4% PFA in PBS (pH 7.4, RNase-free) at RT for 30 min, and then immediately processed for ISH. Each sample contained ~7–10 brains for whole-mount ISH. For immunostaining and CD2-HRP detection experiments, whole brains from ~17–18-day-old male flies were fixed in 4% PFA in PBS (pH 7.4) for 20 min on ice, followed by an additional 20 min at RT in fresh fixative. Brains were rinsed in PBS (pH 7.4), and subsequently processed for immunostaining. Each sample contained ~8–15 brains for whole-mount staining. For each experiment, all brains were processed at the same time and treated in an identical manner.

Details of the RNAscope kit and ISH probes used in this paper are listed in Supplementary Table. 4

All ISH probes were purchased from Advanced Cell Diagnostics, and the RNAscope Multiplex Fluorescent Reagent Kit v2 was used according to the manufacturer’s instructions. For Ubr4 ISH, slides with mouse SCN sections were washed in PBS for 5 min, baked at 60 °C for 30 min, fixed with 4% PFA in PBS for 90 min at RT, dehydrated in increasing concentrations of ethanol (50%, 70%, 2 × 100%), and baked again at 60 °C for 10 min. Slides were treated with hydrogen peroxide for 10 min at RT. Antigen retrieval was performed by placing slides into mildly boiling target retrieval reagent for 5 min and then washed in distilled water for 15 sec and 100% ethanol for 3 min. Slides were dried and permeabilized with Protease III for 30 min at 40 °C. Next, slides were rinsed with distilled water followed by probe incubation for 2 h at 40 °C. Afterwards, slides were incubated with a series of amplifier probes and the appropriate horseradish peroxidase (HRP) solution at 40 °C (AMP1, 30 min; AMP2, 30 min; AMP3, 15 min; HRP-C1, 15 min) with 2 × 2 min washes in wash buffer in between each amplification step. Slides were incubated with Opal 570 dye diluted in TSA buffer for 30 min at 40 °C. After blocking HRP activity, sections were counterstained with DAPI and coverslipped with Fluorescence Mounting Medium.

For poe and Pdf ISH, we followed a modified protocol from the manufacturer that was optimized for whole-mount Drosophila brains. Following fixation, brains were washed 2 × 10 min in PBST (PBS, pH 7.4, with 0.1% Tween-20) at RT, dehydrated in increasing concentrations of methanol (25%, 50%, 75%, 100%) in PBST, incubated in 0.2 M hydrochloric acid in 100% methanol for 30 min at RT, and rehydrated in decreasing concentrations of methanol (75%, 50%, 25%) in PBST. Antigen retrieval was performed by adding mildly boiling target retrieval solution to the samples for 15 min. Brains were then immediately transferred to PBST with 1% bovine serum albumin (BSA) for 1 min, rinsed in 100% methanol for 1 min, and washed with PBST + 1% BSA for 5–10 min. Brains were permeabilized with Protease III for 20 min at 40 °C, rinsed in probe diluent, and incubated O/N in the appropriate probes at 40 °C. The next day, brains were washed 2 × 10 min in wash buffer, followed by incubation with a series of amplifier probes and the appropriate HRP solutions at 40 °C (AMP1, 30 min; AMP2, 30 min; AMP3, 15 min; HRP-C1 and HRP-C3, 15 min), with 2 × 10 min washes in wash buffer between every step. Brains were then incubated with Opal 570 and Opal 520 dyes diluted in TSA buffer for 30 min at 40 °C. HRP activity was blocked, and brains were mounted onto glass microscope slides in Fluorescence Mounting Medium.

Details of antibodies, including working dilutions, and other immunostaining reagents are listed in Supplementary Table. 4

For IHC, coronal brain sections were washed 5 × 5 min in PBS (pH 7.4) with 0.1% Triton X-100 (PBST), treated with 0.3% H₂O₂ in PBS for 20 min at RT, washed 5 × 5 min in PBST, and incubated for 1 h at RT in blocking solution (10% horse serum in PBST). Tissues were incubated O/N at 4 °C in primary antibody diluted in fresh blocking solution. The next day, sections were washed 5 × 5 min in PBST, incubated for 2 h at RT with the appropriate biotinylated secondary antibody diluted in blocking solution, washed 5 × 5 min in PBST, and incubated for 45 min with Avidin-Biotinylated HRP Complexes (ABC). After washing 5 × 5 min in PBST, sections were developed with the 3,3’-diaminobenzidine (DAB) HRP substrate according to the manufacturer’s instructions. For each experiment, all sections were processed concurrently and with the same DAB development time. Sections were mounted onto gelatin-coated glass microscope slides, dehydrated, and coverslipped with Permount Mounting Medium.

To visualize endogenous CD2-HRP expression in flies, following fixation, adult fly brains were rinsed in PBS (pH 7.4), washed 4 × 20 min in PBST (PBS, pH 7.4, with 0.3% Triton X-100) at RT, and then developed with the DAB HRP substrate according to the manufacturer’s instructions. For each experiment, all brains were processed concurrently and with the same DAB development time. Brains were then dehydrated in a graded series of ethanol (30%, 50%, 70%, 90%, 95%, 100%, 100%), cleared in xylene, and mounted with Permount Mounting Medium.

For ICC, HEK293T cells and primary neurons were plated onto poly-D-lysine-coated glass coverslips in 24-well plates. Cells were washed with ice-cold PBS (pH 7.4), fixed in 4% PFA in PBS for 15 min at RT, washed 2 × 5 min in PBS, washed 5 × 5 min in PBS (pH7.4) with 0.1% Triton X-100 (PBST), incubated for 1 h in blocking solution, and incubated O/N at 4 °C in fresh blocking solution containing the primary antibody. The next day, coverslips were washed 5 × 5 min in PBST and incubated with the appropriate AlexaFluor secondary antibody (1:1000) diluted in blocking solution, in the dark for 2 h at RT. Subsequently, coverslips were washed 5 × 5 min with PBST, incubated for 10 min in DAPI (1:10,000) diluted in PBS where applicable, and washed 2 × 5 min in PBS. Coverslips were mounted onto glass microscope slides with Fluorescence Mounting Medium and sealed with nail polish.

For IF of mouse tissues, following fixation, coronal mouse brain sections were washed 5 × 5 min in PBS (pH 7.4) with 0.1% Triton X-100 (PBST), and incubated for 1 h at RT in blocking solution (10% horse serum in PBST). Tissues were incubated O/N at 4 °C in primary antibody diluted in fresh blocking solution. The next day, tissues were washed 5 × 5 min with PBST and incubated for 2 h at RT, protected from light, with the appropriate secondary AlexaFluor antibody diluted in blocking solution. Sections were washed 5 × 5 min with PBST, incubated for 10 min in DAPI (1:10,000) diluted in PBS, washed twice in PBS, and mounted onto glass microscope slides. Slides were coverslipped with Fluorescence Mounting Medium and sealed with nail polish.

For IF of fly tissues, following fixation, adult fly brains were rinsed in PBS (pH 7.4), washed 4 × 20 min in PBST (PBS, pH 7.4, with 0.3% Triton X-100) at RT, and incubated in blocking solution (5% horse serum in PBST) for 2 h at RT. Subsequently, brains were incubated in a fresh blocking solution containing primary antibodies for 2 h at RT, followed by a 4 °C incubation for 40–48 h. After washing 5 × 20 min in PBST, brains were incubated O/N with the appropriate AlexaFluor secondary antibody (1:100 or 1:1000) diluted in blocking solution, in the dark at 4 °C. The next day, brains were washed 4 × 20 min in PBST followed by 2 × 10 min in PBS (pH 7.4). All incubations were done on a rocker set at low speed. Brains were mounted onto glass microscope slides in Fluorescence Mounting Medium.

Fixed coronal slices were washed 3 × 20 min in PBS (pH 7.2) to remove residual aldehydes. Sections were post-fixed in 1% osmium tetroxide in PBS (pH 7.2) for 90 min at RT and subsequently washed 2 × 30 min in PBS. Slices were dehydrated in increasing concentrations of ethanol at RT (50% ethanol for 20 min, 70% ethanol for 30 min, 90% ethanol for 60 min, 100% ethanol O/N). SCN tissues were then microdissected, embedded in Epon resin, and polymerized for 48 h at 40 °C. Ultrathin sections (90 nm) were cut, collected on single-slot formvar-coated copper grids, and stained in uranyl acetate and lead citrate. Ultrathin SCN sections were imaged using a Hitachi HT7700 transmission electron microscope at 80 kV.

Details of reagents and AAV1 constructs used in cell culture experiments are listed in Supplementary Table. 4

Human embryonic kidney (HEK) 293T and mouse Neuro-2a cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% heat-inactivated Fetal Bovine Serum (FBS) and 1% penicillin-streptomycin at 37 °C and 5% CO₂. Cells were transfected with plasmid constructs or siRNAs using Lipofectamine 3000 in DMEM and returned to complete growth media 4–6 h after transfection. Unless otherwise specified, subsequent experimental manipulations or cell harvests were performed 24–48 h post-transfection. For pharmacological studies, HEK293T cells were incubated in growth media supplemented with DMSO (vehicle), 2–20 μM MG132 for 6–18 h, 500 nM Bafilomycin A1 for 6 h, or 10 μg/mL E64D along with pepstatin A for 10 h.

For primary SCN neuronal cultures, SCN tissues from P0 Ubr4^fl/fl mouse pups were rapidly dissected and placed in dissection media (HBSS [Ca²⁺ and Mg²⁺ free] supplemented with 1 mM sodium pyruvate, 0.1% glucose, and 10 mM HEPES [pH 7.3]), dissociated with 0.125% trypsin for 20 min in a 37 °C water bath, and triturated with fire-polished glass pipets. Cells were plated onto glass coverslips placed inside the wells of a 24-well plate or onto the plastic wells directly, in plating media (MEM Eagle with Earle’s BSS supplemented with 5% FBS, 2 mM glutamine, 1 mM sodium pyruvate, 0.45% glucose, and penicillin/streptomycin) as described by Beaudoin et al.⁵⁷. Glass or plastic surfaces were coated with 1 mg/ml or 100 μg/ml poly-D-lysine, respectively, prior to plating. Unless otherwise stated, neurons were plated at densities of 2 × 10⁵ cells per well of a 24-well plate. Three to 6 h later, the plating media was removed and replaced with neuronal growth media (Neurobasal media supplemented with 2% B27, 1 mM glutamine, and penicillin-streptomycin). A half media change was performed every 2 days. Primary neuronal cultures were transduced with AAV1 constructs on day in vitro (DIV) 3. The following multiplicities of infection (MOIs) were used: 5000 MOI for hSyn-CRE, hSyn-GFP, CMV-m-CORO7-2A-mCherry, and CMV-mCherry, and 500 MOI for hSyn-m-VIP. Unless otherwise specified, subsequent experimental manipulations or cell harvests were performed on DIV 10.

Details of plasmids purchased or generated in this paper are listed in Supplementary Table. 4

To generate a CMV-driven Streptavidin-KDEL/SBP-EGFP-NPY construct for the RUSH experiments, human NPY was PCR-amplified from the NPY-GFP plasmid template (Addgene plasmid #74629), and cloned into the Streptavidin-KDEL/SBP-EGFP-GPI plasmid (Addgene plasmid #65294), using the FseI and XhoI sites to place the NPY insert in-frame with the EGFP cassette. Primer sequences can be found in Supplementary Table. 5

Expression plasmids encoding wild-type or mutant human UBR4 fused to an IRES-mCherry sequence were generated using Gateway LR cloning (Supplementary Table 5). Mutant UBR4 harbors the Ubr box^15,32 encoding either multiple Ala mutations or a partially deleted sequence. To construct pENTR3C_hUBR4_ubrbox_Ala and pENTR3C_hUBR4_ubrbox_del entry clones, the previously reported entry clone pENTR3ChUBR4³² was digested with KpnI and CsiI and ligated with a commercially synthesized DNA fragment (Genscript, Piscataway, NJ, USA) (sequences in Supplementary Table 5).

Details of antibodies, including working dilutions, and other Western blotting reagents are listed in Supplementary Table. 4

SCN tissues or cultured cells were lysed in ice-cold RIPA buffer (10 mM Tris-HCl pH 8.0, 1 mM EDTA, 1% NP-40, 0.1% sodium deoxycholate, 0.1% SDS, 140 mM NaCl, Protease Inhibitor Cocktail). Lysates were centrifuged at 13,000 × g for 20 min, and the protein concentration of the supernatant was determined using the Bradford method. Protein samples were resolved by SDS-PAGE using 4–20% Mini-PROTEAN TGX precast gels and transferred onto PVDF membranes. Membranes were cut at approximately the position of the 63 kDa band of the ladder; the top half of the membrane was probed for UBR4/CRN7/V5 and the bottom half for actin. Membranes were washed 5 × 5 min in TBST (20 mM Tris base, 150 mM NaCl, 0.1% Tween-20) and incubated for 1 h at RT in blocking solution (5% skim milk in TBST). Membranes were incubated O/N at 4 °C with primary antibodies diluted in blocking solution. The next day, membranes were washed 5 × 5 min with TBST, incubated for 2 h at RT in HRP-conjugated secondary antibodies, and washed 5 × 5 min in TBST. Signals were detected by chemiluminescence using the SuperSignal West Femto Maximum Sensitivity Substrate and developed on X-ray film. To quantify relative protein expression, the polygon selection tool on the ImageJ software was used to delineate individual bands, and the “measure” function was used to determine band intensity. Results are presented as values normalized to the loading control, actin. Uncropped blots for all experiments are shown in the main figures or, alternatively, in the Source Data file.

Details of kits and plasmids used for ELISA assays are listed in Supplementary Table. 4

WT and UBR4 KO HEK293T cells were co-transfected with NPY-GFP and CMV-mCherry constructs. Culture media was replaced with fresh growth media 48 h post-transfection, and collected 2 h later along with the cell lysates. GFP and mCherry ELISA kits (Abcam) were used to quantify the amount of NPY-GFP in the culture media and the amount of mCherry in the cell lysates, respectively. For every experimental well, relative NPY-GFP secretion was calculated by normalizing the amount of NPY-GFP in the culture media to the amount of mCherry in the cell lysate.

WT and UBR4 KO HEK293T cells were replated and grown overnight in growth media. The next day, cells were treated with 100 µg/ml cycloheximide (Sigma) for 5 min prior to lysis in a hypotonic buffer containing 2.5 mM MgCl₂, 5 mM Tris-HCl (pH 7.5), 1× protease inhibitor cocktail (EDTA-free; Roche), 1.5 mM KCl, 2 mM DTT, 200 U/ml RNaseIn, 0.5% (v/w) sodium deoxycholate, and 0.5% (v/w) Triton X-100. Four hundred micrograms of the isolated ribonucleoproteins were sedimented on a 10–50% sucrose gradient using ultracentrifugation at 250,000 × g in a SW40 rotor (Beckman Coulter) at 4 °C for 2 h. The sediments were then fractionated using an ISCO gradient fractionation system while continuously recording the optical density at 260 nm with a FOXO JR Fractionator. RNA was then isolated from each fraction using TRIzol (Invitrogen) according to the manufacturer’s instructions and subsequently used for qRT-PCR analysis.

Details of reagents used for RNA extraction and qRT-PCR experiments are listed in Supplementary Table. 4

Total RNA was extracted from SCN tissues and HEK293T cells using the RNeasy Micro Kit and TRIzol Reagent, respectively, according to the manufacturer’s instructions. For the polysome profiling experiment, TRIzol-extracted RNA from 14 individual fractions (#1 to #14) were pooled to generate 5 fractions (‘A’ to ‘E’) for subsequent cDNA synthesis: A (#1 to #3), B (#4 and #5), C (#6 and #7), D (#8 to #10), and E (#11 to #14). cDNA was synthesized using SuperScript IV Reverse Transcriptase, and qPCR was performed using SsoFast EvaGreen Supermix on a Bio-Rad CFX384 Real-Time PCR Detection System. Except for the polysome profiling study, relative transcript abundance was calculated using the delta-delta CT method and normalized to Gapdh. For the polysome profiling experiment, the amount of CRN7 or GAPDH transcripts in each pooled fraction was calculated as a percentage of the total amount of the transcript detected in all fractions.

Total RNA was extracted from fly brain tissues using the RNeasy Micro Kit. cDNA was synthesized using iScript cDNA Synthesis Kit, and qPCR was performed using iTaq Universal SYBR Green Supermix. Rp49 was used to normalize relative transcript expression. All primer sequences used for qPCR analyses are listed in Supplementary Table 5.

Details of reagents and software used for time-lapse image acquisition and processing are listed in Supplementary Table. 4

HEK293T cells seeded on 4-chamber glass slides were transduced with Golgi-RFP BacMam 2.0 one day after transfection with the Streptavidin-KDEL/SBP-EGFP-NPY plasmid. The next day, cells were supplemented with fresh phenol red-free growth media, and the chamber slide was transferred to a stage-top incubator (Live Cell Instrument Inc.) where the cells were maintained at 37 °C and supplied with humidified air and 5% CO₂ throughout the time-lapse imaging session. At recording time = 0’, the media was removed, and replaced with fresh media supplemented with cycloheximide (100 μg/mL) and D-biotin (40 μM). Confocal imaging was performed using a Leica DMi8 inverted microscope (Leica) equipped with X-Light V2 spinning disk confocal (CrestOptics), laser diode illuminator-7 (89 North), and a prime HS:sCMOS camera (Photometrics). The multipoint acquisition was achieved through the MS-2000 motorized XY stage (Applied Scientific Instrumentation), while axial sectioning was obtained with the MS-2000 piezo top-plate (Applied Scientific Instrumentation). Images were acquired with a 63X/1.40 NA oil objective and Metamorph software (Molecular Devices). Design of the acquisition journals and system integration were done by Quorum Technologies. All subsequent image processing was performed on ImageJ.

Details of software used for image acquisition and analysis are listed in Supplementary Table. 4

Cell processing for MS, LC-MS analyses, and database search was performed by the Donnelly Mass Spectrometry Center at the University of Toronto. Details of resources and software used in MS experiments and data analyses are listed in Supplementary Table. 4

The detail of the software used for statistical analyses in this paper is listed in Supplementary Table. 4

Data were analyzed using one- and two-sample t test, one-way analysis of variance (ANOVA), two-way ANOVA, linear mixed-effects modeling, and chi-square test. Post hoc significance of pairwise comparisons was assessed using the Bonferroni test with α set at 0.05. When assumptions of normality were not met as assessed by Shapiro–Wilk test, Mann–Whitney U test, or Kruskal–Wallis test with Dunn’s post hoc were used (α set at 0.05). Where applicable, tests were two-tailed. Control and KO/KD groups that lack an asterisk (*) notation were not significantly different. GraphPad Prism (version 8.3.1) was used to perform all statistical analyses. Sample sizes reported for all experiments represent “n” following the exclusion of outliers. All experiments reported in this paper were repeated at least twice with similar results.

Further information on research design is available in thelinked to this article. Nature Research Reporting Summary

Nature Communications thanks the anonymous reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available.

The mass spectrometry proteomics data generated in this study have been deposited to the ProteomeXchange Consortium (http://proteomecentral.proteomexchange.org↗) via the PRIDE⁵⁸ partner repository with the dataset identifier PXD020630↗. The human UniProt database used in this study is publicly available at https://www.uniprot.org/help/uniprotkb↗. All data generated during this study that support our findings are available within the article and its Supplementary Information files. Further information and requests for data, resources, and reagents should be directed to and will be fulfilled by the corresponding authors. Source data are provided with this paper.