(C) PLOS One
This story was originally published by PLOS One and is unaltered.
. . . . . . . . . .
Twelve-hour rhythms in transcript expression within the human dorsolateral prefrontal cortex are altered in schizophrenia [1]
['Madeline R. Scott', 'Translational Neuroscience Program', 'Department Of Psychiatry', 'Center For Neuroscience', 'University Of Pittsburgh', 'Pittsburgh', 'Pennsylvania', 'United States Of America', 'Wei Zong', 'Department Of Bioinformatics']
Date: 2023-02
Twelve-hour (12 h) ultradian rhythms are a well-known phenomenon in coastal marine organisms. While 12 h cycles are observed in human behavior and physiology, no study has measured 12 h rhythms in the human brain. Here, we identify 12 h rhythms in transcripts that either peak at sleep/wake transitions (approximately 9 AM/PM) or static times (approximately 3 PM/AM) in the dorsolateral prefrontal cortex, a region involved in cognition. Subjects with schizophrenia (SZ) lose 12 h rhythms in genes associated with the unfolded protein response and neuronal structural maintenance. Moreover, genes involved in mitochondrial function and protein translation, which normally peak at sleep/wake transitions, peak instead at static times in SZ, suggesting suboptimal timing of these essential processes.
Funding: This work was supported by the National Institutes of Health (R01MH111601 to CMC; T32MH016804 to MRS; T32HL082610 to MRS; DP2GM140924 to BZ; K01MH128763 to KDK, DA051390 to MS) and the Brain & Behavior Research Foundation (NARSAD Independence Award to CMC). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Data Availability: All data are within the paper and its Supporting Information files or are freely available through the CommonMind Consortium through an application process. The Common Mind also provides ethics statements from it’s various brain banks that are part of the consortium.
https://www.nimhgenetics.org/resources/commonmind .
In the current study, we use DLPFC data previously analyzed for circadian rhythms, allowing us to compare both 12 and 24 h rhythms within the same subjects. Multiple convergent analyses identify transcripts that have measurable 12 h rhythms in human DLPFC, with distinct abnormalities in the identity and timing of these transcripts in SZ.
Schizophrenia (SZ) is a chronic neuropsychiatric illness that affects over 20 million people worldwide and is a leading cause of disability [ 2 ]. Many SZ patients experience disturbances in the rhythmicity of sleep/wake cycles, peripheral gene expression, and daily hormones [ 3 , 4 ]. Molecular rhythm patterns, however, have only just begun to be directly measured in the human brain. To explore these rhythms in human postmortem brain tissue, our lab and others have utilized a “time of death” (TOD) analysis, in which gene expression data are organized across a 24 h clock based on the time of day of the subject’s death, to identify significant changes in gene expression rhythm patterns associated with specific brain regions [ 5 ], age [ 6 ], and psychiatric illnesses [ 7 – 9 ]. In SZ subjects, rhythmic analysis of RNA sequencing (RNA-seq) data collected by the CommonMind Consortium [ 10 ] from the dorsolateral prefrontal cortex (DLPFC) identified a loss of diurnal rhythmicity in a number of transcripts. Notably, SZ subjects exhibited 24 h rhythmicity in a set of transcripts that were not rhythmic in subjects with no psychiatric diagnosis (NP) [ 8 ]. Genes with enhanced 24 h rhythmicity in the SZ cohort were associated with mitochondria dysfunction and GABA-ergic signaling [ 8 ], consistent with previous work that finds differential expression of these pathways in subjects with SZ [ 11 , 12 ]. These studies demonstrate that circadian rhythms in gene expression can be reliably measured in human brain tissue and are severely disrupted in the DLPFC of SZ subjects. However, no study to date has attempted to measure 12 h rhythms in transcript expression in human brain or determine if there are changes to these ultradian rhythms in subjects with SZ.
Twelve-hour (12 h) ultradian rhythms have long been observed in coastal marine animals, whose behavior aligns with ocean tides [ 1 ]. Recent studies have confirmed 12 h transcriptional rhythms in other organisms including C. elegans, mice, and olive baboons [ 1 ]. Various aspects of human behavior (sleep patterns, cognitive performance) and physiology (body temperature, blood pressure, migraine onset, circulating hormone levels) also exhibit 12 h rhythms [ 1 ]. In cases like blood pressure and body temperature, these 12 h rhythms are secondary to a dominant 24 h rhythm, suggesting the presence of multiple superimposed rhythms. However, as 12 h rhythms in transcript expression have not been identified in human tissue, it is unknown whether these processes are related to and/or regulated by molecular ultradian rhythms. Therefore, characterization of the human brain ultradian transcriptome will expand our understanding of transcript expression rhythms in the brain and their contribution to dysfunction in subjects with abnormalities in brain function.
Finally, we applied motif enrichment analysis to these groups (mNP morning/evening, mNP afternoon/night, SZ morning/evening, SZ afternoon/night) to determine whether there may be differences in upstream regulators associated with 12 h rhythms in SZ ( S9 Fig and S8 File ). Additionally, we split transcripts that significantly lost/gained 12 h rhythms by peak time and performed a motif enrichment analysis ( S9 Fig and S8 File ). We found that ETS domain, PAR bZIP, POU class homeobox, PRD class homeobox, Forkhead box, and SRY-box families were associated with motifs enriched in the mNP morning/evening group of transcripts, but none of these families were enriched in the SZ morning/evening group ( S9A, S9B and S9D Fig ). Consistent with this, these families were associated with the morning/evening peaking transcripts that lost rhythmicity ( S9A , S9C , and S9E Fig ). The ETS domain family, however, was enriched in the SZ afternoon/night group and in transcripts that gained rhythmicity in the afternoon/night, suggesting that, like the mitochondria pathways observed earlier, this protein family was associated with transcripts that have altered timing in SZ ( S9A–S9C Fig ). Similarly, the KLF, SP domain family, and BHLH domain family were associated with motifs enriched in transcripts that peak in the afternoon/night in the mNP cohort, but with the morning/evening group in the SZ cohort ( S9A and S9F Fig ). These distinct patterns in enrichment give us insight into which pathways may be regulating the multifaceted differences observed in 12 h rhythms in SZ.
( A ) Histogram of 12 h rhythm peak times in mNP and SZ. ( B ) IPA of transcripts with 12 h rhythms split by peak time. ( C - E ) Cytoscape visualization [ 20 ] of a Metascape analysis demonstrating the biological processes enriched for 12 h rhythms that peak during the morning/evening or afternoon/night. ( F ) Radar plot showing number of transcripts that peak at each time of day. Outer scale is ZT (h) time, radial scale is number of transcripts, solid lines indicate first peak, and dashed lines indicate second peak. Results of the NLR, IPA, and Metascape analyses can be found in S1 , S7 , and S11 Files, respectively. IPA, Ingenuity Pathway Analysis; mNP, match NP; NLR, nonlinear regression; SZ, schizophrenia; ZT, Zeitgeiber time.
The SZ and mNP cohorts had similar timing patterns for transcripts with 12 h rhythms, with one population of transcripts that peaked in expression in the morning/evening, and the other that peaked during the afternoon/night ( Fig 5A and S1 File ). However, IPA and Metascape analyses indicated that mitochondria, EIF2 signaling, and protein ubiquitination pathways were associated with morning/evening 12 h rhythmic transcripts in the mNP cohort, but with afternoon/night transcripts in the SZ cohort (Figs 5B , 5D , and S8A and S7 and S11 Files). Transcripts associated with the UPR also peaked during the morning/evening in the mNP cohort but, consistent with the loss of rhythmicity analysis, were not associated with either population in the SZ cohort ( Fig 5B and S7 and S11 Files). Similarly, transcripts associated with the cytoskeleton and synaptogenesis (Synaptogenesis Signaling Pathway, Reelin Signaling in Neurons, Actin Cytoskeleton Signaling, RhoA Signaling) peaked in expression during the afternoon/night in the mNP cohort but were not associated with either time point in the SZ cohort (Figs 5B , 5C , and S8A and S7 and S11 Files). While most transcripts with 12 h rhythms peaked in the morning/evening in SZ, there was little distinct pathway recognition in either the IPA or Metascape analyses ( S7 and S11 Files). In both analyses, histone/chromosome regulation (histone h3 k36, positive regulation heterochromatin, regulation transcription initiation) was identified as a top pathway/biological process, though at a notably lower level of enrichment than the other groups (Figs 5B , 5E , and S8B and S7 and S11 Files).
We next utilized the eigenvalue/pencil analysis, which allowed us to identify transcripts with gene expression that were best explained by either a single 12 h RC, a single 24 h RC, or a combination of both 12 and 24 h RCs ( Fig 4B ). Notably, the eigenvalue/pencil analysis identifies RCs—not overall rhythms in expression—and does not incorporate p-values [ 19 ]. This results in far more transcripts identified as having a 12 h RC by the eigenvalue/pencil analysis than a 12 h rhythm by the NLR analysis. Despite this, a very similar pattern in pathway enrichment emerged between the two analyses. In the mNP cohort, transcripts with 12 and 24 h RCs were found in distinct pathways ( Fig 4C and S7 and S12 Files), while both 12 and 24 h RCs were linked to mitochondria-related pathways in SZ ( Fig 4D and S7 and S13 Files). Intriguingly, the number of transcripts with both 12 and 24 h RCs increased from 1,710 (30% and 40% of 24 and 12 h RCs, respectively) in the mNP cohort to 3,974 (58% and 67% of 24 and 12 h RCs, respectively) in the SZ cohort ( Fig 4C and 4D and S5 File ). As an example of what these data show, expression of transcripts in the electron transport chain complexes across time indicated mitochondria-associated genes had 12 h rhythms in the mNP cohort, while in SZ, these transcripts had a combination of 12 and 24 h RCs ( Fig 4E and 4F ). We propose that this may explain why the NLR analysis identifies this group of transcripts as either a 12 h or 24 h rhythm, or both, in SZ subjects.
( A ) Metascape-derived heatmaps of biological processes enriched for 12 and 24 h rhythms in mNP and SZ cohorts. Full Metascape results can be found in S11 File . ( B ) Illustrations of how transcripts could have no rhythms, a 24 h rhythm, a 12 h rhythm, or a combined 12 and 24 h rhythm. Results of eigenvalue/pencil analysis for mNP and SZ subjects can be found in S12 and S13 Files, while a summary of the 12 and 24 h RCs is detailed in S5 File . ( C - D ) IPA of transcripts with 12 h and 24 h RCs in ( C ) mNP subjects and ( D ) subjects with SZ. IPA results can be found in S7 File . ( E - F ) Gene expression heatmaps of subunits from the MT electron transport chain complexes (CI-CV) across time of day from ( E ) mNP subjects, which show a 12 h rhythm, and ( F ) subjects with SZ, which have a 12 + 24 h rhythm. IPA, Ingenuity Pathway Analysis; mNP, match NP; MT, mitochondria; RC, rhythmic component; SZ, schizophrenia.
Our previous work found a surprising gain of 24 h rhythmicity in mitochondria-related transcripts in subjects with SZ compared to the NP group [ 8 ]. Here, we expanded upon those results using Metascape [ 18 ] to determine the biological processes implicated in the top transcripts with significant 12 and 24 h rhythms in mNP subjects and subjects with SZ (Figs 4A and S7 and S11 File ). While each group had unique aspects of biological process enrichment, mitochondria- and translation-associated biological processes were enriched for both 12 and 24 h rhythms in SZ, but only 12 h rhythms in the mNP cohort (Figs 4A and S7 and S11 File ).
We next performed a loss/gain analysis of 12 h rhythmicity in SZ, as described previously for 24 h rhythms [ 8 ], to confirm our finding of fewer 12 h rhythms in SZ and to determine if transcripts experience ultradian reprogramming in SZ. A total of 800 transcripts significantly lost 12 h rhythmicity, and 276 transcripts gained rhythmicity in SZ ( Fig 3D-3H and S1 File ). Overall, transcripts that lost rhythmicity were associated with the UPR and the Protein Ubiquitination Pathway, while the smaller number of transcripts that gained rhythmicity did not fall into clear pathways ( Fig 3I and 3J and S7 File ).
NLR analysis identified 12 h rhythms in the ( A ) mNP and ( B ) SZ cohorts. ( C ) IPA of transcripts with significant 12 h rhythms. ( D ) Number of transcripts identified as having a significant 12 h rhythm or a significant difference in goodness of fit (R 2 ) between the two cohorts (deltaR 2 ). ( E - H ) Examples of genes that ( E - F ) lose or ( G - H ) gain 12 h rhythms in SZ. Expression values for these graphs can be found in S6 File . Gene expression over time is shown for both the ( E , G ) mNP and ( F , H ) SZ cohorts. ( I - J ) IPA of transcripts that significantly ( I ) lose or ( J ) gain rhythmicity in SZ. Results of the NLR and IPA can be found in S1 and S7 Files, respectively. IPA, Ingenuity Pathway Analysis; mNP, match NP; NP, no psychiatric diagnosis; NLR, nonlinear regression; SZ, schizophrenia; ZT, Zeitgeiber time.
( A ) Motif enrichment analysis of transcripts with 12 h rhythms in the human DLPFC. Dashed lines separate three groups: (1) All motifs tested; (2) Protein families previously implicated in regulating 12 h rhythms; (3) Protein families previously implicated in regulating 24 h rhythms. ( B ) Examples of transcription factors from ETS domain (Elk1), Kruppel-like factor (KLF7), and SP transcription factor (SP1) families that have 12 h rhythms in human DLPFC. Expression values for these graphs can be found in S6 File . ( C - E ) Comparison of motif enrichment analysis of transcripts with 12 h rhythms that peak in the morning/evening and those that peak in the afternoon/night. ( C ) Protein families previously implicated in regulating 12 h rhythms. Individual motifs are marked in blue for each family. ( D ) Protein families previously implicated in regulating 24 h rhythms. Individual motifs are marked in yellow for each family. ( E ) Other top protein families associated with motif enrichment. Motif analysis results used to create this figure can be found in S8 File . BHLH, Basic Helix–Loop–Helix; bZIP, basic leucine zipper; DLPFC, dorsolateral prefrontal cortex; KLF, Kruppel-like factor; TOD, time of death; ZT, Zeitgeiber time.
We next determined potential sites of regulation and predicted upstream regulators for 12 h rhythms using a motif enrichment analysis [ 14 ] ( Fig 1H and 1I and S8 File ). Motifs enriched in transcripts with 12 h rhythms are bound by transcription factors from the ETS domain family, Kruppel-like factors (KLFs), and the SP domain family (Figs 1H , 2A , and 2B ). Motifs associated with the ETS domain family remained enriched when we separately analyzed transcripts that peak in the morning/evening and afternoon/night, but both the KLF and SP domain families were associated only with motifs enriched in the afternoon/night (Figs 1I and 2C ). Motifs associated with the circadian-related Basic Helix–Loop–Helix (BHLH) domain family were not enriched in the analysis of transcripts with 12 h rhythms but were strongly enriched when analyzing the afternoon/night group separately ( Fig 2D ). Alternatively, motifs associated with PAR domain containing basic leucine zipper (bZIP) proteins, which have also been implicated in regulating circadian rhythms [ 15 ], were enriched both in the overall analysis (Figs 1H and 2A ) and in the morning/evening group ( Fig 2D ). Other families associated with motifs enriched in the morning/evening include the homeobox POU and PRD classes, while motifs associated with the SMAD family and peroxisome proliferator-activated receptors (PPARs) were enriched in the afternoon/night group (Figs 1I and 2E ).
The timing of 12 h rhythms revealed two distinct populations of transcripts, one which peaked in expression in the morning/evening (Zeitgeiber time (ZT) 2 to 3 and 14 to 15; approximately 9 AM/PM) and the other that peaked during the afternoon/night (ZT 8 to 9 and 20 to 21; approximately 3 AM/PM) ( Fig 1F ). Transcripts associated with mitochondria and the proteasome (Polyamine Regulation in Colon Cancer) were enriched in the morning/evening population, while those associated with the cytoskeleton and calcium signaling (RAR Activation, Insulin Secretion Pathway, VDR/RXR Activation) peaked in the afternoon/night ( Fig 1G and S7 File ).
( A - E ) 12 h rhythms identified by a sinusoidal NLR. ( A ) Heatmap of top 100 transcripts with 12 h rhythms, ordered by phase. ( B - D ) Gene expression over TOD scatterplots for the top 3 transcripts with 12 h rhythms showing the sinusoidal curve, goodness of fit (R 2 ), and p-value. Expression values for these graphs can be found in S6 File . ( E ) Plot comparing NLR-derived amplitude and p-values for all transcripts. Transcripts with significant 12 h rhythms in expression are denoted in blue (p < 0.01). ( F ) Radar plot showing peak expression times (ZT) of transcripts with 12 h rhythms (p < 0.01). Outer scale in ZT (h), radial scale represents number of transcripts, solid lines indicate first peak, and dashed lines indicated second peak. ME peak times indicated in dark blue, AN indicated in light blue. (G) IPA of transcripts with 12 h rhythms separated by peak time. ( H - I ) Motif enrichment analysis of transcripts with 12 h rhythms ( H ) all together and ( I ) separated by phase. Results of the NLR, IPA, and motif analyses can be found in S1 , S7 , and S8 Files, respectively. AN, afternoon/night; DLPFC, dorsolateral prefrontal cortex; IPA, Ingenuity Pathway Analysis; ME, morning/evening; NLR, nonlinear regression; TOD, time of death; ZT, Zeitgeiber time.
Discussion
Biological rhythms allow organisms ranging from bacteria to humans to anticipate changes in the environment across the light/dark cycle and adapt accordingly. These rhythms occur on many scales, from seasons to days to hours, and the importance of circadian rhythms in health and disease has become increasingly clear over the past few decades, particularly in the context of psychiatric illnesses [21,22]. Far less is known about ultradian rhythms, including how prevalent they are in the human brain, and whether they are disrupted in subjects with psychiatric disorders. In this study, we characterized the 12 h transcriptome within the DLPFC of NP and SZ subjects. These rhythms have distinct timing patterns, splitting them largely into two populations of transcripts. One of these peaks in the morning/evening (ZT 2 to 3/14 to 15; approximately 9 AM/PM) and is largely associated with mitochondria, EIF2 signaling, and the UPR. The other peaks during the afternoon/night (ZT 8 to 9/20 to 21; approximately 3 AM/PM) and is associated with cytoskeleton dynamics and the processes necessary to build and maintain neuronal connections.
These findings indicate that 12 h rhythms in the brain are associated with processes necessary for essential cellular functions—and may be fundamental for timing processes to maximize resources and reserve energy when not needed. This is consistent with previous analyses in mouse liver, which uncovered 12 h rhythms in metabolism-related transcripts and processes fundamental to transcription, RNA splicing, translation, and proteostasis [23,24]. An exploratory analysis comparing 12 h rhythms in the mouse liver to our findings identified 620 overlapping genes (S5 Fig and S14 File), suggesting a high degree of similarity between the two studies despite differences in technique, species, and region [25]. In the liver, the timing of these processes coincides with sleep/wake transition times, leading to the proposal of a vehicle-cargo hypothesis for 12 h rhythms, in which 12 h rhythmicity accommodates increased demand for gene expression/processing at biological “rush hours” (i.e., sleep/wake transitions) by up-regulating expression of factors that facilitate protein and energy production [23]. Our observation that mitochondria- and translation-associated transcripts in human brain peak in expression in the morning/evening—when individuals are likely transitioning from wake to sleep or sleep to wake—strongly supports this hypothesis.
Strikingly similar to the mouse liver, our motif enrichment analysis implicated the ETS domain, KLF, and SP domain families in ultradian rhythm regulation (Fig 2A). Individual members of these transcription factor families also had significant 12 h rhythms, marking them as potential transcription factor regulators of these rhythms (Fig 2B). Consistent with our pathway analyses, separation of the data into two populations by peak time resulted in increased clarity. We found that transcripts that peak in the morning/evening are regulated by the POU class of homeoboxes and PAR bZIP family, while transcripts that peak in the afternoon/night are regulated by KLF, SP domain, and BHLH domain families (Fig 2C). The PAR bZIP and BHLH families are both implicated in circadian rhythms [15], suggesting that at least some proportion of what we have identified as 12 h rhythms could be regulated by the canonical circadian clock; however, the BHLH family is very large and the circadian clock regulators are not the top hits within this group.
We observed far fewer transcripts with 12 h rhythms in SZ subjects than in the mNP subjects. Intriguingly, the UPR is the top pathway associated with transcripts that no longer have 12 h rhythms in SZ. The UPR has been shown to regulate 12 h rhythms in in vivo culture models and mouse liver [23,25]. Consistent with this, differential expression of UPR proteins and markers of UPR activity have been found in the DLPFC of subjects with SZ [26]. SZ subjects also do not display rhythmicity in transcripts associated with the actin cytoskeleton and synaptogenesis, which peak during the afternoon/evening in the mNP subjects (Fig 5). Genomic analyses strongly implicate synaptogenesis and synaptic plasticity processes, while neuroanatomical studies have demonstrated reduced dendritic spine density in various regions, including the DLPFC, in SZ [27,28]. Of particular note, several voltage-gated calcium channels (CACNAs), a couple of which are top SZ risk factor genes [27], are among this group with 12 h rhythms in mNP subjects but no rhythms in SZ.
The NLR analysis showed relatively little direct overlap in the transcripts that are identified as rhythmic in the mNP and SZ cohorts (Fig 3 and S1 File). However, a threshold-free approach (S6 Fig) and pathway analyses indicate some degree of overlap. This is best exemplified by 12 h rhythms in mitochondria-associated pathways, which are not lost but instead switch from peaking during the morning/evening to afternoon/night time points in SZ. This could have profound effects on mitochondrial energy production at times of day when this is most needed. Various studies have implicated circadian rhythms in mitochondria biology, including gene expression in the mouse SCN, oxygen consumption and mitochondria respiration in isolated mitochondria from rat brains, and fusion/fission states in mouse macrophages [29–31]. Interestingly, mitochondrial-related pathways also gained 24 h rhythms in the DLPFC of subjects with SZ [8] (Fig 4), resulting in a convergence of rhythmicity dysregulation in mitochondrial function transcript expression. Gaining a 24 h rhythm may be a compensatory measure to account for suboptimal timing of 12 h rhythms or reflect changes associated with diminished neuronal activity specifically at night. Our findings are consistent with a robust literature of mitochondrial abnormalities in SZ, from genetics to function, number, location, and shape [32]. It will be interesting in future studies to determine changes in mitochondrial morphology, number, function, and location in subjects with SZ as a function of time of day, and how this relates to transcript expression.
This study’s findings fit into a growing body of literature attempting to characterize mammalian 12 h rhythms. 12 h rhythms were first observed as circatidal rhythms, in which coastal and estuarine animals show behavior aligned with the approximately 12.4 h ebb and flow of tides [23]. As researchers studied these behaviors, two theories on the molecular mechanisms emerged: The first suggested that there are two circadian clocks acting in antiphase to produce the 12 h peaks in behavior, while the second proposed a dedicated 12 h molecular clock. Several studies have reported that disrupting light and the circadian clock does not impact circatidal rhythms, providing support for the second hypothesis [1]. The study of 12 h rhythms in mammals, is much newer, with the first study showing 12 h rhythms in transcript expression in mouse liver and other tissues occurring in 2009 [33]. Multiple studies have since confirmed the existence of 12 h rhythms in mouse liver [25,33–36]. The same debate over the molecular mechanism that generates these rhythms has occurred within this group, with some evidence for two circadian antiphasic regulators [33,36] and others demonstrating that 12 h rhythms are dependent on the UPR and demonstrating involvement of specific transcription factors [35,37,38]. Our findings best align with the idea of a dedicated 12 h clock, but future work in cell and animal models will be necessary to confirm this.
The superchiasmatic nucleus (SCN) is the master pacemaker of circadian rhythms in the brain, but whether ultradian rhythms are regulated across regions by a dedicated system is unknown. While not specific to 12 h rhythms, levels of dopamine in the striatum fluctuate in synchrony with ultradian locomotor activity cycles and dopaminergic transmission directly regulates ultradian cycle length in mice [39]. Interestingly, these dopamine-driven ultradian cycles in locomotor activity harmonize with circadian rhythms coordinated by the SCN, but if this relationship is disrupted, it can lead to altered patterns of arousal and disrupted sleep/wake cycles when desynchronized [39], which are commonly observed in subjects with SZ. SZ has long been associated with altered dopaminergic transmission in cortico-striatal pathways [40]. Abnormalities in dopaminergic signaling, therefore, may be a potential mechanism by which ultradian rhythms are disrupted across the brain in SZ.
Due to the limited sample size and novel nature of the study design, many of the analyses in our study are exploratory and use p-value cutoffs for determining statistical significance. As such, we chose to focus throughout on the overall patterns and themes that the study illuminated, rather than individual genes identified as rhythmic. Additionally, human postmortem brain tissue research presents a variety of limitations. We have addressed several through our study design, which includes exclusion of subjects older than 65, strict TOD criteria, and cohorts matched for a number of important biological, clinical, and technical factors. However, factors like disrupted sleep, antipsychotic medication, and nicotine use are common in the SZ cohort, with low to no prevalence in the NP cohort. As such, we are unable to determine whether our findings are a component of SZ pathophysiology, due to one of these factors, or an interaction between the two.
Sleep disturbances are frequently observed in patients with SZ, with prevalence rates reported to be approximately 80% [41,42]. Higher rates of sleep disruption, particularly insomnia, are noted in both chronic and first-episode populations [42]. Additionally, sleep disruption is associated with symptom severity of positive and cognitive symptoms [42]. Intriguingly, sleep disturbances often precede episodes of psychosis and/or worsening of symptoms and have been observed in young populations at high risk for psychosis [42]. Similar to other psychiatric illnesses like bipolar disorder and major depression, sleep architecture abnormalities like reduced slow-wave sleep, increased sleep onset latency, decreased total sleep time, and decreased sleep efficiency have been repeatedly shown in subjects with SZ [42]. Notably, SZ patients also have decreased sleep spindle density, which is not found in other psychiatric populations [42], suggesting that there may be a unique relationship between SZ and sleep. While we are not looking directly at sleep in this study, altered/sleep wake cycles can influence molecular rhythms [43–45], and as such, we do not know whether our findings are due to the sleep disturbances experienced by these patients, a component of the disease pathology, or a combination of the two. While the TOD information for these patients has allowed us to perform the analyses we report here, we do not have information on each subject’s sleep patterns, which may be important for interpreting our findings. Future work, using both animal and human postmortem models, defining how chronotype and sleep disruptions impact the 24 and 12 h transcriptomes in the prefrontal cortex will be necessary to begin teasing apart how these dynamics impact molecular rhythms in SZ.
Antipsychotic medications are also important to consider, as the majority of the SZ cohort is on them while none of the NP cohort is. Antipsychotic medications could be having an impact from multiple directions, as they can directly impact gene expression as well as improve sleep in some patients [41,46,47]. Antipsychotics can improve total sleep time and efficiency, though this is variable dependent on the type of medication, with atypical antipsychotics being more effective than typical [41]. A few studies have attempted to look at the impact of antipsychotics on circadian gene expression in rodents but have had mixed results [41]. Future work in populations taking antipsychotics that are not diagnosed with psychoses and in rodent models will be necessary to determine what impact antipsychotics are having on gene expression, and whether this is indirectly through their sedative properties.
In contrast, SZ subjects use nicotine, which has stimulant properties, at much higher rates than the general population [48]. This use typically starts before disease onset, and higher smoking rates are associated with poorer quality of life, worse prognosis, and increased disease severity [48]. It is possible that the disproportionate nicotine use and disruptions in the cholinergic system may be contributing to our findings. This could be through many mechanisms, as this system is incredibly complex and heavily interwoven into both cognition and sleep [49].
In conclusion, to our knowledge, this study is the first to identify 12 h rhythms in transcript expression in the human brain. These rhythms are associated with fundamental cellular processes. However, in SZ, there is a strong reduction in the number of transcripts with 12 h rhythms, along with altered timing of transcripts important in mitochondrial function. Future studies will determine the functional consequences of these findings to optimal brain health and the pathophysiology of brain disorders.
[END]
---
[1] Url:
https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.3001688
Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.
via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/
