Circadian profiling of the transcriptome in immortalized rat SCN cells

Gus J. Menger2,3, Kim Lu3, Terry Thomas2,3, Vincent M. Cassone2,3 and David J. Earnest1,2,3

1 Department of Human Anatomy and Medical Neurobiology, Texas A & M University Health Science Center, College of Medicine, Texas
2 Center for Biological Clocks Research, Texas A & M University, College Station, Texas
3 Department of Biology, Texas A & M University, College Station, Texas


    ABSTRACT
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 GRANTS
 REFERENCES
 
Endogenous oscillations in gene expression are a prevalent feature of the circadian clock in the mammalian suprachiasmatic nucleus (SCN) and similar timekeeping systems in other organisms. To determine whether immortalized cells derived from the rat SCN (SCN2.2) retain these intrinsic rhythm-generating properties, oscillatory behavior of the SCN2.2 transcriptome was analyzed and compared with that found in the rat SCN in vivo using rat U34A Affymetrix GeneChips. In SCN2.2 cells, 116 unique genes and 46 ESTs or genes of unknown function exhibited circadian fluctuations with a 1.5-fold or greater difference in their mRNA abundance for two cycles. Many (35%) of these rhythmically regulated genes in SCN2.2 cells also exhibited circadian profiles of mRNA expression in the rat SCN in vivo. Functional analyses and cartography indicate that a diverse set of cellular pathways are strategically regulated by the circadian clock in SCN2.2 cells and that the largest categories of rhythmic genes are those involved in cellular and systems-level communication or in metabolic processes like cellular respiration, fatty acid recycling, and steroid synthesis. Because many of the same genes or nodes within these functional categories were rhythmically expressed in both SCN2.2 cells and the rat SCN, the circadian regulation of these pathways may be important in modulating input to or output from the SCN clock mechanism. In summary, global expression and circadian regulation of the SCN2.2 transcriptome retain many SCN-like properties, suggesting that genes displaying rhythmic profiles in both experimental models may be integral to their function as both circadian oscillators and pacemakers.

rhythm; suprachiasmatic nucleus; pacemaker; clock; oscillator; SCN2.2


    INTRODUCTION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 GRANTS
 REFERENCES
 
MANY ORGANISMS, ranging from cyanobacteria to humans, have evolved circadian clock mechanisms that provide for the rhythmic organization of internal biological processes so as to approximate the 24-h time course of the daily solar cycle (25, 56). In mammals, the master clock responsible for the regulation of circadian rhythms throughout the body is located in the suprachiasmatic nuclei (SCN) of the anterior hypothalamus (34). Complete lesions of the SCN abolish circadian rhythmicity in a wide array of molecular, biochemical, endocrine, and behavioral processes (49), and transplantation of fetal SCN tissue restores behavioral rhythmicity in SCN-lesioned hosts (59). Consistent with the timekeeping function of the SCN, many of its cellular activities and functions oscillate independently of environmental or external input. Circadian rhythms of gene expression, cellular metabolism, neuropeptide secretion, and neural activity are intrinsic to the SCN in vivo and persist after isolation of SCN cells in vitro (19, 23, 52, 61, 66).

The SCN clock mechanism consists of interlocked transcription-translation feedback loops in which the gene products of core components rhythmically suppress their own transcription or the expression of other clock genes. Clock, Period1 (Per1), Per2, cryptochrome1 (Cry1), Cry2, and Bmal1 (Mop3) have been identified as core elements of the molecular clockworks in the SCN, based on evidence indicating that mutation or knockout of these genes in mice alters or abolishes circadian behavior (62). In turn, these core elements of the molecular clockworks coordinate downstream rhythmicity in the expression of other genes controlling diverse biochemical, cellular, and physiological processes within SCN cells. Some of these clock-controlled genes presumably couple the clock mechanism to SCN-specific output signals that regulate rhythmicity in other neural substrates or endocrine organs.

Despite our current understanding of the molecular organization of the mammalian circadian clock, important questions remain with regard to how the core clock mechanism regulates clock-controlled genes in SCN cells and how these genes are configured in biochemical pathways so as to generate SCN outputs that synchronize or coordinate rhythmicity in other cells and tissues. We used an immortalized line of rat SCN cells (SCN2.2) to explore these issues, because these cells retain many of the circadian properties of the SCN in vitro and in vivo. SCN2.2 cells are capable of endogenous, self-sustained rhythmicity and of functioning as a pacemaker by imposing rhythmic properties on cocultured cells and restoring behavioral rhythmicity when transplanted into SCN-lesioned hosts (3, 20, 21). In the absence of other neural and endocrine inputs, the SCN2.2 model also provides an opportunity to ultimately examine how the molecular clock imposes oscillatory properties on SCN cellular physiology. With the use of Affymetrix GeneChips, the transcriptome in SCN2.2 cells collected at 6-h intervals over two circadian cycles was profiled and then compared with the temporal patterns of gene expression observed in the rat SCN. This profiling of rhythmic gene expression was used to determine 1) whether global properties of the rat SCN transcriptome are conserved in SCN2.2 cells, 2) the extent to which the transcriptome is rhythmically regulated by the circadian clock in SCN2.2 cells, and 3) whether rhythmically regulated genes in SCN2.2 cells show similar patterns of expression in the rat SCN. Bioinformatic tools were applied to characterize the functional distribution of clock-controlled genes in SCN2.2 cells and their configuration within metabolic and signaling pathways.


    MATERIALS AND METHODS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 GRANTS
 REFERENCES
 
SCN2.2 cultures and RNA extraction.
For each of three biological replicates, SCN2.2 cells (passages 15 and 16) were seeded on laminin-coated culture dishes (60 mm; Corning, Corning, NY) and maintained at 37°C and 5% CO2 in minimum essential medium (MEM) supplemented with 20% FBS, 3,000 µg/ml glucose, and 292 µg/ml L-glutamine. At 48-h intervals, the medium was changed and cultures were split at 1:3 or 1:4. Before experimental analysis, cultures were plated on multiple six-well plates (35 mm) coated with poly-D-lysine and laminin. At ~24 and 44 h after plating, the culture medium was changed so as to, respectively, reduce the FBS concentration to 10% and then to 5%. To facilitate cell cycle and circadian clock synchronization across cultures, cells were subjected to medium replacement and exposed to serum-free medium (neurobasal medium supplemented with 3,000 µg/ml glucose, 292 µg/ml L-glutamine, and 1x B-27 serum-free supplement) containing 15 µM forskolin (Calbiochem, La Jolla, CA) for 2 h. Cultures were rinsed and thereafter maintained in serum-free neurobasal medium with supplements. At the conclusion of the forskolin treatment, cells were harvested from individual six-well plates (approximate density = 9.5 x 104 cells/cm2) at 6-h intervals for 48 h, and total cellular RNA was extracted using RNeasy Midi-Kit protocols (Qiagen, Valencia, CA). RNA extracts from individual samples were treated with on-column DNase I digestion and concentrated with sequential ethanol precipitations.

Rat SCN and RNA extractions.
Adult male Long-Evans rats (175–200 g, n = 45; Harlan Laboratories, Indianapolis, IN) were housed two to three per cage and maintained in the vivarium at the Texas A & M University System Health Science Center under a standard 12:12-h light-dark photoperiod (LD 12:12; lights on at 0600). At 1800 [circadian time 12 (CT 12)] animals were exposed to constant darkness (DD), and 12 h later (0600 or CT 0) they were killed under isoflurane anesthesia by decapitation at 6-h intervals (n = 5) for 48 h using an infrared viewer. After the eyes were removed in the dark, SCN tissue was immediately dissected as described previously (19) under dim light, frozen in liquid nitrogen, and stored at –80°C. SCN tissue from individual animals was separately homogenized in TRIzol reagent (Invitrogen, Carlsbad, CA) by aspiration through a 25-gauge needle, and then extracted total cellular RNA for all five animals at each time point was pooled into a single sample. RNA samples were subjected to on-column treatment with DNase I to digest genomic DNA and then stored at –80°C.

Affymetrix GeneChip analysis.
Before microarray analysis, the quality of all SCN2.2 and rat SCN RNA samples was assessed by electrophoresis on 1% agarose gels containing 0.1 µg/ml ethidium bromide. Experimental procedures including double-stranded cDNA synthesis and biotinylated cRNA preparation were conducted according to protocols described in the Affymetrix GeneChip Expression Analysis Technical Manual (39). Briefly, first-strand cDNA was reverse transcribed from 15 µg of total RNA using 600 U of SuperScript II RT (Invitrogen) in a 20-µl total volume containing T7-(dT)24 primer, 10 mM DTT, 500 µM of each dNTP, and 1x first-strand cDNA buffer for 1 h at 42°C. Second-strand cDNA was synthesized in 1x reaction buffer containing 200 µM of each dNTP, 10 U of DNA ligase, 40 U of DNA polymerase I, and 2 U of RNase H (final volume = 150 µl) for 2 h at 16°C. Samples were then treated with 20 U of T4 DNA polymerase for 5 min at 16°C and incubated in 10 µl of 0.5 M EDTA. Double-stranded cDNA was purified with phase lock gel electrophoresis (Eppendorf Scientific, Westbury, NY) and phenol-chloroform extraction, followed by ethanol precipitation. Biotin-labeled cRNA was subsequently produced with the use of a BioArray HighYield RNA transcript labeling kit (Affymetrix, Santa Clara, CA). Labeled cRNA was spin purified (Qiagen) and fragmented. To assure the quality of labeling and fragmentation efficiency, unfragmented and fragmented cRNA products were analyzed on an Agilent 2100 bioanalyzer before hybridization on arrays. Fragmented biotinylated cRNA (15 µg) was hybridized on Affymetrix GeneChip rat U34A arrays at 45°C and 60 rpm in a GeneChip Hybridization Oven 640 (Affymetrix) for 16 h.

After hybridization, arrays were washed and stained, using Affymetrix protocols for antibody amplification staining on a GeneChip Fluidics Station 400 in conjunction with Affymetrix Microarray Suite 5.0 software. After a brief wash with a nonstringent buffer, the stained signals on the array were then amplified with a solution containing 3 µg/ml anti-streptavidin biotinylated antibody (Vector Laboratories, Burlingame, CA), 1x morpholine ethane sulfonic (MES) buffer, 2 mg/ml acetylated BSA, and 0.1 mg/ml normal goat IgG for 10 min, followed by a second staining with streptavidin-phycoerythrin (SAPE) for 10 min at 25°C. After a final wash with a stringent buffer, the probe array was scanned at the excitation wavelength of 570 nm, using an Agilent GeneArray Scanner (Palo Alto, CA).

After scanning, each image was first checked for major chip defects or abnormalities during hybridization as a quality control. Arrays were scanned using a global scaling strategy in which the average absolute signal intensity of all arrays was set to an arbitrary target signal intensity of 500 before being uploaded into GeneSpring 6.1 software (Silicon Genetics, Redwood City, CA).

Data processing and global analyses.
GeneChip signal intensity data from three biological replicates of SCN2.2 cells and two technical replicates of the rat SCN derived from separate aliquots of pooled RNA samples were uploaded into GeneSpring 6.1 and filtered in an identical fashion. For each GeneChip probe set, signal intensities were converted to log base 2 values and then normalized to the 50th percentile of all measurements. Experimental averages of normalized data were calculated for each of nine time points in SCN2.2 cells and for each of eight time points in the rat SCN. The data were then sequentially filtered at three levels. To verify gene expression, we filtered the probe sets according to two criteria: 1) detection of a "present" flag in 55% or more of the time points and 2) a raw signal intensity value of ≥50, which was above the average experimental background signal in 55% or more of the time points. In both SCN2.2 cells and the rat SCN, the temporal profiles of genes surpassing these expression criteria were filtered to isolate those showing stable expression. Stable genes were distinguished by temporal expression profiles that did not differ statistically (standard correlation value: P > 0.995) from a flat line created with the GeneSpring 6.1 "Draw Gene" tool. Finally, cycling genes with a peak-to-trough difference of 1.5-fold or greater and periodicities of 18–30 h were identified among the remaining transcripts, using methods similar to those described previously (13, 43). The normalized data were cross-correlated (P > 0.90) with cosine waves of specific phase and period using the Prism software package (GraphPad, San Diego, CA). For SCN2.2 cells, we imposed an additional requirement that circadian-regulated transcripts show at least one pair of nonoverlapping standard errors bars between time points with the highest and lowest values. The analytical and amplitude criteria used to identify cycling transcripts in both SCN2.2 cells and the rat SCN are consistent with those applied in recent microarray studies profiling circadian gene expression (2, 13, 43, 65). Because our filtering parameters excluded some transcripts that were readily detectable in our and/or other analyses of the rat SCN, we reanalyzed the normalized data using secondary expression criteria that identified genes with a raw signal intensity of ≥50, which was above the average experimental background signal in at least four of the time points in SCN2.2 cells and three or more of the time points in the rat SCN. Microarray data have been deposited in the National Center for Biotechnology Information (NCBI) Gene Expression Omnibus (GEO) database (accession nos. GSE1654 and GSE1673).

To assess the false positive rate for detecting rhythmically expressed genes in our analysis, data files were randomized by first converting each file to a two-dimensional data matrix and then assigning each element of the array a mapped integer key stored in a one-dimensional array. The one-dimensional array of keys was randomized by a perl module (List::Utils) utilizing a Fisher-Yates algorithm. The shuffled keys were read from the array and converted to their corresponding location in the data matrix. Data values were then read from the data matrix and sequentially written to the output file, producing a randomized version of the original data file while maintaining the numerical composition of the file. These randomized data were then subjected to the same analytic and amplitude criteria used to identify cycling transcripts in the original data sets for SCN2.2 cells and the rat SCN. On the basis of a comparison of rhythmic data sets identified in this randomization analysis with the total number of circadian-regulated probe sets in the raw data, we estimated false-positive rates of 8–17% for SCN2.2 cells and 7–19% for the rat SCN.

Bioinformatics and validation.
Because some genes surpassing circadian expression filters were represented by multiple probes on the RG-U34A GeneChip, we used GeneSpring 6.1, GenMAPP 1.0 and 2.0 freeware (14, 16) and bioinformatic tools available as links from the NCBI (51, 58), including Basic Alignment Search Tool (BLAST) (5), to identify unique genes with rhythmic patterns in SCN2.2 cells. To corroborate rhythmicity in SCN2.2 cells, cycling genes were compared with those found in the rat SCN. For a limited set of clock and clock-controlled genes, quantitative PCR (qt-PCR) was also used to validate rhythmic expression in SCN2.2 cells. Per2, Bmal1, Nos2, and the calcium channel subunit-{alpha}1C are primary examples of cycling genes that were validated by qt-PCR. Quantification of relative mRNA abundance was performed using TaqMan or SYBR Green real-time PCR technology [Applied Biosystems (ABI), Foster City, CA]. To generate single-strand cDNAs, total cellular RNA (1–2 µg) was reverse transcribed using Superscript II (Invitrogen) and a primer mixture of oligo-dTs and random hexamers. With the use of the cDNA equivalent of 50 ng of total RNA, triplicate aliquots of each sample were then PCR amplified in an ABI Prism 7700 sequence detection system (28). The following probes and primers were designed, using PrimerExpress software (ABI): rPer2, forward 5'-TTCGACTACCTGCATCCAAAAG-3', reverse 5'-AAGTCCAGTCTTCGCATCGAT-3'; rBmal1, forward 5'-CCAAGAAAGTATGGACACAGAC-3', reverse 5'-GCATTTTTGATCCTTCCTTGGT-3'; rCry1, forward 5'-CTGGCGTGGAAGTCATCGT-3', reverse 5'-CTGTCCGCCATTGAGTTCTATG-3'; Id-1, forward 5'-TGGTCTGTCGGAGCAAAGC-3', reverse 5'-TCCTTGAGGCGTGAGTAGCA-3'; Nos2, forward 5'-ACCCGACTGAAGCACTTTGG-3', reverse 5'-TCGTTGGGAGTGGACGAAG-3'; {alpha}1C (Rob2), forward 5'-CCGGAAGCCAGTGCATTTT-3', reverse 5'-TGGTGAAGATCGTGTCATTGAC-3'; {alpha}1C (Rob2) probe, 5'-FAM-CCAAACAACAGGTTCCGCCTGCAGT-TAMRA-3'; {alpha}1A, forward 5'-GGATGACAACACCGTTCACTTC-3', reverse 5'- CCACCCTTTGCGATTTTGAT-3'; {alpha}1A probe, 5'-FAM-TGGCTCTGATCCGAACCGCCC-TAMRA-3'; {alpha}1G, forward 5'-CCTGCCTGTTGCCGAGAG-3', reverse 5'-CAGGAGACGAAACCTTGACTGA-3'; {alpha}1G probe, 5'-FAM-CGGCCTATATCTTTCCTC-TAMRA-3'; CypA, forward 5'-TGTGCCAGGGTGGTGACTT-3', reverse 5'-TCAAATTTCTCTCCGTAGATGGACTT-3'.

To control for differences in sample RNA content, 18S ribosomal RNA (rRNA) was amplified and multiplexed in the same reaction with target genes or amplified in separate cDNA aliquots (50 ng) from the same RNA samples using a ribosomal RNA control reagent kit (VIC Probe, ABI). For normalization of the clock genes, Per2, Bmal1, and Cry1, rRNA or cyclophilin A (CypA) mRNA was amplified using the cDNA equivalent of 1 ng of total RNA. In both multiplex and conventional reactions, the relative mRNA abundance for a given target gene was calculated by normalization to corresponding rRNA or cyclophilin levels in each sample using the comparative cycle threshold (CT) method described in the ABI Prism 7700 Sequence Detection System User Bulletin 2 (PE-ABI). Values for the same target gene (amplicon) were adjusted for interassay variation according to ABI protocols. Relative abundance of target mRNA was represented as a percentage of the maximal value obtained within an individual experiment.


    RESULTS AND DISCUSSION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 GRANTS
 REFERENCES
 
Global properties.
We separately examined temporal regulation of the transcriptome in SCN2.2 cells and in the rat SCN so as to compare global and specific facets of gene expression between these two experimental models. The breadth of gene expression was similar in SCN2.2 cells and the rat SCN. In SCN2.2 cells, 3,993 of the 8,900 probe sets on the rat U34A GeneChip and 2,929 probe sets in the rat SCN surpassed stringent expression criteria (Table 1). Comparison of probe sets detected in both models with those in only the rat SCN indicates that SCN2.2 cells express 83% of the genes detected in the rat SCN (2,430 of 2,929). The similarity in the extent and level of common gene expression across these experimental models suggests that SCN2.2 cells retain SCN-like properties with regard to the regulation of genomic activity and the functionality of in vivo cellular pathways. In addition, a wide array of genes was regulated in a similar fashion in SCN2.2 cells and the rat SCN. Many of the genes common to both models shared analogous patterns of stable or variable expression. Cyclin D3, brain-specific protein gene Rb109, and thioredoxin (Txn2) represent functionally diverse genes with stable expression in both models (Fig. 1, A and B). Prominent examples of genes with variable expression in SCN2.2 cells and the rat SCN include nerve growth-induced factor A (Ngfi-A), metallothionein (Mt1a), and ubiquitin carboxy-terminal hydrolase I (Uchl1). For many of these genes, relative levels of mRNA abundance were also comparable between these models.


View this table:
[in this window]
[in a new window]
 
Table 1. Comparison of gene expression and function in SCN2.2 cells and the rat SCN

 


View larger version (23K):
[in this window]
[in a new window]
 
Fig. 1. Temporal profiles of representative genes that are stably or variably expressed in both SCN2.2 cells (A) and the rat suprachiasmatic nuclei (SCN) (B). Symbols denote determinations of mRNA abundance (n = 3) at 6-h intervals for cyclin D3 (squares), Rb109 (circles), Txn2 (diamonds), and Ngfi-A (triangles). In this and subsequent figures, these values were derived by normalizing GeneChip log base 2 signal intensity to the 50th percentile of all measurements per sample. The raw values for the 50th percentile of all measurements per sample are plotted as a black line in relation to the secondary ordinant in each panel. Thus the normalized values between 3 and 60 for these genes indicate that their raw values were ~3- to 60-fold greater than those for the 50th percentile.

 
Global circadian regulation.
Recent studies profiling the transcriptome in the murine SCN have identified {approx}100–365 circadian-regulated genes that mediate a wide range of cellular functions (54, 70). In the present study, a comparable number of genes in SCN2.2 cells were distinguished by circadian profiles. Specifically, 162 genes in SCN2.2 cells showed circadian rhythms of mRNA abundance that surpassed stringent expression criteria (Table 1). The amplitude of these SCN2.2 rhythms in gene expression ranged from 1.5- to 12-fold differences between peak and minimum levels of mRNA abundance. By comparison, 301 genes in the rat SCN displayed circadian oscillations that were identified using the same expression and rhythm criteria. Because 35% of the unique genes with rhythmic patterns in SCN2.2 cells surpassing stringent expression criteria exhibited similar oscillations in the rat SCN, some of these genes may be critical for the circadian clock and pacemaker properties of SCN cells. To screen additional genes in SCN2.2 cells with rhythmic profiles that passed secondary expression criteria (see Supplemental Table S2; the Supplemental Material for this article is available at the Physiological Genomics web site),1 we used three different approaches: 1) validation of rhythmicity for a limited set of these genes using qt-PCR, 2) comparative analysis to determine whether these genes were both expressed based on stringent criteria and rhythmically controlled in the rat SCN, and 3) establishment of a limited homology table between the rat U34A and mouse U74A (version 2) GeneChips using GeneSpring 6.1 (see Supplemental Table S4). Some of the rhythmic genes validated by qt-PCR in this supplementary screening analysis include inducible nitric oxide synthase (iNos or Nos2), calcium channel Rob2, and Per2. Our homology screen identified 50 homologs or functionally related genes that were rhythmically expressed in murine tissues and SCN2.2 cells. Circadian-regulated genes that were identified in SCN2.2 cells and in the murine SCN include pericentriolar material 1 (Pcm1), DNA directed RNA polymerase II, the putative chloride channel Clcn4–2, Id3a, Hsp 70 protein 5 (Hspa5), and Snap25 (Supplemental Table S4, Supplemental Fig. S2). An interesting facet of our homology screen is that some genes with circadian patterns in both SCN2.2 cells and the rat SCN, such as nerve growth-induced factor A (Ngfi-A) and ornithine decarboxylase antizyme inhibitor (Oazi) were rhythmically regulated in the murine liver but not in the SCN (17, 54, 70).

To examine the phase of rhythmic genes expressed in SCN2.2 cells, we compared their first peak of cyclic mRNA abundance (maxima of normalized signal intensity) in relation to the first sampling interval and the circadian expression profiles of core clock genes. For example, circadian-regulated genes exhibiting maximal mRNA abundance during the first sampling interval (hour 0) and recurrent 24-h peaks thereafter were assigned to phase group I (PG I). In a similar fashion, rhythmic genes with peak mRNA expression during the second (hour 6), third (hour 12), or fourth (hour 18) sampling interval were assigned to PG II, PG III, and PG IV, respectively. On the basis of both gene array and qt-PCR analyses, peak mRNA expression in SCN2.2 cells was observed at hours 0 and 24 for Per2 and at hours 12 and 36 for Bmal1 (Mop3) (Fig. 2), so the rhythms for these clock genes were, respectively, assigned to PG I and to PG III. With the use of these comparisons, phase was unambiguously assigned to 149 of 162 genes with rhythmic profiles in SCN2.2 cells (Supplemental Table S1). Circadian phase could not be clearly assigned to the rhythmic patterns of 13 genes due to inconsistencies between experimental replicates and probe sets. In SCN2.2 cells, 56 genes displayed PG I rhythms in which peak mRNA abundance coincided with the zenith of Per2 expression, whereas 37 genes exhibited PG III oscillations in which peak levels were concurrent with the crest of Bmal1 expression. Thirty-eight and 18 circadian-regulated genes were identified in PG II and PG IV, respectively. The predominant distribution of circadian-regulated genes in SCN2.2 cells within phase groups coinciding with peak Per2 or Bmal1 expression is compatible with phase cluster analyses of rhythmic genes in the murine SCN (70). Similarly, 53 and 141 genes with cycling profiles in the rat SCN showed phase distributions that were, respectively, coincident with the maxima of Per2 or Bmal1 expression.



View larger version (31K):
[in this window]
[in a new window]
 
Fig. 2. Rhythmic patterns of Per2 (circles) and Bmal1 (triangles) expression in SCN2.2 cells. Symbols denote average values (n = 3) for mRNA abundance at 6-h intervals based on normalized GeneChip determinations (solid lines) or on quantitative PCR (qt-PCR) validation analysis (dashed lines). The profile of Cry1 (squares) mRNA expression is depicted for comparison. The qt-PCR profiles are derived from triplicate analyses of RNA samples at each time point (n = 3), and the plotted values correspond to the ratios of species-specific Per2, Bmal1 or Cry1/rRNA or CypA mRNA signal in which the maximal value for each experiment was set at 100%. GeneChip and qt-PCR values are, respectively, represented on the primary and secondary ordinant.

 
Functional analyses of clock-controlled genes in SCN2.2 cells.
Consistent with recent studies demonstrating that eukaryotic circadian clocks rhythmically regulate the expression of genes mediating diverse aspects of cell physiology (6, 12, 13, 18, 54, 69, 70, 71), SCN2.2 cells exhibited circadian fluctuations in genes with a wide range of functions, such as metabolism, immunity, neurotransmission, and development. To examine the breadth of cellular processes regulated by the circadian clock, rhythmic genes in SCN2.2 cells that surpassed stringent expression criteria were organized according to one of seven broad physiological categories: 1) energetics, 2) cellular and systems-level communication, 3) protein dynamics, 4) cellular development, 5) defense and detoxification, 6) cytoskeleton and adhesion, and 7) unknown or expressed sequence tags (ESTs) (Table 1, bottom, and Supplemental Table S1). On the basis of their application to the 162 rhythmic genes surpassing stringent expression criteria in SCN2.2 cells, 15% regulate energetic processes such as cellular respiration, fatty acid recycling, and steroid synthesis; 30% are involved in cellular or systems-level communication; 12% mediate the dynamics of gene expression by controlling protein translation, degradation, or trafficking; 8% regulate aspects of cellular development; 3% are associated with defense and detoxification; 3% affect cytoskeletal elements, cellular adhesion, or components of the extracellular matrix; and 28% are ESTs or genes with uncharacterized functions (Table 1, bottom). The prevalence of rhythmic genes in the cellular and systems-level communication category may reflect their degree of representation on the rat U34A GeneChip (i.e., {approx}45% of annotated genes with known functions). However, the distribution of circadian-regulated genes across other functional categories was not related to a preferential or exclusive bias on the array. On the basis of analysis using Simplified Gene Ontology lists in GeneSpring 6.1, genes involved in energetics, protein dynamics, cellular development, defense and detoxification, and cytoskeleton and adhesion were, respectively, represented by {approx}7, 7, 20, 15, and 6% of the probe sets on the array. Within the first four categories, rhythmic genes in SCN2.2 cells were further segregated into three or more functional clusters.

Energetics.
The rhythmic regulation of metabolic activity is a hallmark circadian property of the SCN. For example, the circadian utilization of 2-deoxyglucose is a well-documented marker of endogenous rhythmicity in both the rat SCN and SCN2.2 cells (3, 66). In addition, the SCN exhibits circadian regulation of genes encoding transporters for energy metabolites and metabolic enzymes (27, 64). Consistent with this evidence for the circadian control of SCN metabolism, 24 genes involved in energetic processes showed rhythmic expression profiles in SCN2.2 cells. These genes with energetic functions were subdivided into four functional clusters: 1) glucose metabolism and mitochondrial energy transduction (n = 7), 2) lipid and fatty acid metabolism (n = 7), 3) transporters of energy metabolites (transporters) (n = 6), and 4) miscellaneous metabolism (n = 4) (Supplemental Table S1). Circadian-regulated genes involved in glucose metabolism include malic enzyme 1 (Me1), hexokinase 2 (Hk2), and glyoxylate reductase/hydroxypyruvate reductase (GenBank no. AA892799), an enzyme that mediates the conversion of serine to glucose (31). The circadian clock in SCN2.2 cells also impacts mitochondrial energy transduction through the rhythmic expression of mitochondrial ATP synthase-8 (mt-Atp8). Importantly, circadian regulation of Glut-1 (Slc2a1), the primary facilitative transporter of D-glucose across the blood-brain barrier (44, 74), and Mct1 (Slc16a1), a major transporter of ketone bodies and lactate in glial cells (73), was observed in both SCN2.2 cells and the rat SCN. Rhythmically expressed constituents of the miscellaneous metabolism cluster include two genes involved in steroidogenesis, 3-hydroxy-3-methylglutaryl-CoA reductase (Hmgcr) and Hmgc synthase-1 (Hmgcs1).

To address the question of how clock-regulated genes impose circadian rhythms of metabolism on SCN cells, we constructed a GenMAPP of circadian-regulated genes identified using stringent expression criteria (14, 16). In SCN2.2 cells, many of the rhythmic genes in the energetics category are critically positioned so as to provide for the coupling of metabolic pathways associated with fatty acid recycling and the synthesis of cholesterol and polyunsaturated fats (Fig. 3A) (24). For instance, fatty acid reservoirs fueling ß-oxidation in the mitochondria are functionally linked to the citric acid cycle by means of acetyl-CoA. In turn, citrate generated in the citric acid cycle can be used to form fatty acid acyl-CoAs that can replenish fatty acid reservoirs or contribute to the formation of polyunsaturated fats, cholesterol, and ultimately hormone production. Within these pathways, circadian regulation of gene expression was observed in SCN2.2 cells and validated in the rat SCN at four key points: carnitine palmitoyltransferase-1b, Hmgc synthase-1, and fatty acid synthase. Because these genes widely impact the regulation of synthetic, anabolic, and catabolic processing of energy substrates, this finding suggests that circadian clock regulation of a small number of genes in crucial positions may provide an effective strategy for transmitting temporal information throughout different networks of cellular metabolism. It is interesting that circadian profiles of these strategically positioned genes share a common "phase group" assignment with that for Per2.



View larger version (30K):
[in this window]
[in a new window]
 
Fig. 3. GenMAPP analysis of circadian-regulated genes in SCN2.2 cells depicting their distribution in relation to information flow within pathways mediating fatty acid and steroid metabolism (A); G protein-coupled receptor, second messenger, and mitogenic signaling cascades (B); and protein dynamics (C). Boxes indicate the relative positions of circadian-regulated genes surpassing stringent expression criteria. Phase group assignments of rhythmic genes and their coincidence with clock gene oscillations are denoted as I (Per2), II, III (Bmal1b), and IV. In some instances, no phase group was assigned due to inconsistencies between experimental replicates and probe sets. Solid lines denote probable interactions among rhythmically expressed genes or other pathway components and/or products arising from these interactions. Dashed lines signify indirect or predicted functional interactions between circadian-regulated genes and critical elements of these pathways.

 
Despite ample evidence in both the present study and previous reports for the circadian regulation of genes involved in energetic processes and redox status (27, 64), the functional significance of these and other metabolic oscillations in the SCN clock is unknown. According to a model developed by Pellerin and Magistretti (55), circadian clock regulation of transporters for energy metabolites may provide a basis for metabolic coupling between glial and neuronal cells within the SCN. During the subjective day, increased electric activity in SCN neurons is associated with the release of an excitatory neurotransmitter such as glutamate, which subsequently induces glucose uptake and glycolysis in astrocytes. Thus clock-regulated expression of Glut-1 (Slc2a1) and Mct1 (Slc16a1) may play a critical role in the rhythmic transportation of glucose across the blood-brain barrier and of lactate produced by astrocytes, so as to fulfill the circadian metabolic requirements of SCN neurons.

Cellular and systems-level communications.
Communication of temporal information to the rest of the brain and to peripheral tissues is a critical function for the pacemaker of the SCN. Rhythms of mRNA expression were identified in SCN2.2 cells for 49 genes mediating cellular or systems-level communication. Circadian-regulated genes in the communication category were subdivided into five functional clusters: 1) neurotransmission (n = 11), 2) extracellular factors (n = 3), 3) G protein-coupled receptors and associated proteins (n = 9), 4) cytosolic signal transduction (n = 8), and 5) nuclear factors (n = 18) (Supplemental Table S1).

To understand how specific aspects of neurotransmission in SCN cells are modulated by the circadian clock, we further classified the rhythmic 11 genes within this cluster according to the following subgroups: 1) glutamatergic metabolism and signaling (n = 4), 2) synaptic function and maintenance (n = 4), and 3) miscellaneous neurotransmission (n = 3) (Supplemental Table S1). The inositol 1, 4, 5-trisphosphate receptor-coupling element and immediate early gene Homer1 is an example of a clock-regulated gene associated with glutamatergic signaling in SCN2.2 cells and the rat SCN. Two ionotropic {alpha}-amino-3-hydroxy-5-methyl-4-isoxasole proprionic acid receptor genes, Gria1 and Gria4, that passed secondary expression criteria (Supplemental Table S2) displayed circadian fluctuations of mRNA abundance in SCN2.2 cells. Circadian regulation of these glutamatergic signaling genes may contribute to the function of this neurotransmitter in circadian photoentrainment. Glutamate involvement in the transmission of entraining light signals by the retinohypothalamic tract to the SCN is supported by immunohistochemical localization of glutamate (11) and different subtypes of glutamate receptors within the retinorecipient subfield of the SCN (22, 72) and by physiological studies demonstrating that, similar to light, glutamate elicits time-dependent phase shifts in circadian rhythms in vivo and in vitro (46, 68). In a similar fashion, glutamate has been shown to phase shift the SCN2.2 rhythmicity in luciferase-reported Ca2+/cAMP response element (CRE) activity (32). In addition to the oscillations in these glutamatergic signaling elements, the rhythmic expression of Glut-1 (glutamate transporter), Gclc, and Got2 may be important in processes by which this neurotransmitter mediates the circadian regulation of the SCN clock by light, because these genes serve to replenish presynaptic reservoirs of L-glutamate or to regulate glutamate-based carbon fixation and amino acid metabolism.

Rhythmic genes in the synaptic function and maintenance cluster are associated with pre- and postsynaptic membranes and include Ykt6, Stx4a, Snap29, and the PSD-95/SAP90-associated gene Dlgap1. Interestingly, Dlgap1, which possesses guanylate kinase activity when functionally expressed, has been implicated in the postsynaptic organization of voltage-gated potassium channels and N-methyl-D-aspartate (NMDA) receptors (53). Consequently, the rhythmic expression of Dlgap1 mRNA may reflect a divergent point of circadian gene regulation affecting multiple facets of postsynaptic function. Such divergent regulation may provide a platform for the intrinsic rhythm of SCN electric activity in vivo and in vitro (23). Circadian rhythms of mRNA expression in SCN2.2 cells were also observed for genes that directly affect neuronal electric properties and neurotransmitter production, such as the voltage-gated potassium channel Kcng1, the inwardly rectifying potassium channel Kcnj14, and A-2 arylamine N-acetyltransferase (Nat1) (Supplemental Table S1). Because rectifying potassium channels are thought to contribute to the modulation of cardiac pacemaking (29), circadian regulation of a voltage-sensitive channel like Kcnj14 may have some significance in the endogenous capacity of SCN cells to generate rhythmic patterns of electric activity.

Of the remaining clock-controlled genes in the cellular and systems-level communication category, most are G protein-coupled receptors (n = 9), cytosolic signaling factors and transducers (n = 8), or nuclear factors (n = 18). To examine how temporal information may be transmitted from the cell membrane to the nucleus via these clock-regulated genes, we analyzed their functional positions within a GenMAPP encompassing G protein-coupled receptor, second messenger, or mitogenic signaling cascades (Fig. 3B) (14, 16). G protein-coupled receptors or associated proteins with rhythmic patterns of mRNA expression in SCN2.2 cells included G protein-coupled receptor 48 (Gpr48), guanine nucleotide-binding protein-{alpha}12 (Gna12), muscle and microspikes RAS (M-ras), and the olfactory-like receptor (Ol1/Olf-r1). With the use of secondary expression criteria, circadian profiles were also observed in SCN2.2 expression of somatostatin receptor subtype (Sstr4), cholecystokinin A receptor (Cckar), angiotensin II receptor (Agtr2), orexin receptor-1 (Hcrtr1), 5-hydroxytryptamine-5A receptor (Htr5a), and the olfactory-like receptors Hfl-vn1 and Tpcr13 (Supplemental Table S2). The rhythmic regulation of monoamine receptors in SCN2.2 cells is consistent with observations on the function of this neurotransmitter system in the SCN in vivo and in vitro. The monoamine serotonin is thought to play a role in photic regulation of the SCN clock, because 5-HT agonists directly induce phase shifts or modulate the resetting action of light (40, 57, 60).

Downstream of these G protein-signaling elements, circadian regulation of moieties involved in cyclic AMP and MAP kinase signaling pathways in SCN2.2 cells may have important functional implications, because these pathways are thought to modulate the transmission of photic input to the SCN (for review, see Ref. 45). Circadian-regulated genes in these signaling pathways, such as adenylyl cyclase-3 (Adcy3) and MAP kinase phosphatase-3 (Dusp6 or Mpkp3), may impact the regulation of cyclic AMP and phosphorelay transduction in SCN2.2 cells (Fig. 3B). By affecting cyclic AMP and MAP kinase signaling, the specific distribution of rhythmically regulated genes within these pathways may differentially influence their interactions with other processes and thus provide an effective strategy for the circadian gating of SCN responses to input signals. CCAAT/enhancer binding protein-ß (Cebpb), CREBBP/EP300 inhibitory protein 1, and activating transcription factor 3 (Atf3), provide representative examples of genes that may be circadian regulated as a function of their responsiveness to cyclic AMP.

Protein dynamics.
The regulation of gene expression at the protein level is critical for the circadian clock function of the SCN. For instance, cyclical expression of translated gene products and posttranslational modification are required for the appropriately timed progression and function of the self-sustained feedback loops that comprise the core clock mechanism (4, 77). Thus cellular processes influencing protein dynamics, such as protein synthesis, folding, modification, sorting, trafficking, and degradation, are likely to have critical roles in the oscillatory properties of the SCN clock. Circadian regulation of mRNA levels in SCN2.2 cells was observed for 20 genes mediating protein dynamics. Clock-controlled genes in the protein dynamics category were further subdivided into three functional clusters: 1) degradation and synthesis (n = 6), 2) protein sorting and trafficking (n = 4), and 3) protein modification and folding (n = 10) (Supplemental Table S1). Ubiquitin carboxy-terminal hydrolase L1 (Uchl1), ribosomal protein L6 and L21 (Rpl6 and Rpl21), vacuolar protein sorting homolog (Vps33b), the nucleoporin p58/p45, and Hsp 27 protein 1 (Hsp27/Hspb1) are representative examples of circadian-regulated genes within these functional clusters.

Circadian regulation of genes involved in protein dynamics may ultimately serve to resonate temporal information across cellular pathways that are not subject to clock control at the mRNA level. To examine this possibility, we used GenMAPP 2.0 to determine whether a few circadian-regulated genes involved in protein synthesis, degradation, and processing could effectively modulate the flow of temporal information in SCN2.2 cells (Fig. 3C) (14, 16). Circadian oscillations of mRNA expression in SCN2.2 cells were observed in processes associated with the 60S ribosomal subunit (Rpl6 and Rpl21) and tRNA synthesis [aspartyl-tRNA synthetase (Dars)]. Because of their location at the apex of pathways in the protein dynamics GenMAPP, rhythmic expression of these translation regulatory genes may have significant implications in clock control of gene function. Both of the 60S ribosomal genes, Rpl6 and Rpl21, interact with the initiation complex to form the eukaryotic 80S ribosomal complex (24, 75), and aspartyl-tRNA synthetase functions as a dimeric enzyme or part of a multi-enzyme complex comprising aminoacyl-tRNA synthetases for several amino acids, including methionine (47). Because >4% of the amino acids in rPer2 are aspartyl residues, the rhythmic RNA expression of Dars, Rpl6, and Rpl21 could modulate the availability of cognate tRNAs as well as the actual ribosomal reading of the mRNA sequences for clock and even nonclock genes, thereby imposing oscillations on the activity or function of their protein products.

In the protein dynamics GenMAPP (Fig. 3C), seven genes associated with chaperone-like activities or posttranslational processing and four genes involved in protein turnover and/or metabolism showed circadian expression profiles in SCN2.2 cells. Hsp701b/Hspa1b, Hsp105, Hsp27/Hspb1, Hspa5/Grp78, and pre-mtHsp70 (mitochondrial type) are rhythmically regulated genes that encode heat shock proteins involved in the folding and targeted destruction of proteins damaged by stress. Circadian regulation of Pmpca and Tcp1 may also contribute to clock control of protein dynamics, because these genes are involved in posttranslational processing of nacent polypeptides (33, 42).

SCN2.2 cells also exhibited circadian oscillations in the expression of two genes involved in the proteasome targeting pathway, polyubiquitin and an ubiquitin recycling enzyme, Uchl1. Circadian regulation of Uchl1 could modulate reservoirs of ubiquitin and thus influence the proteasome targeting of proteins. Because Hsp 70 proteins are involved in targeting damaged proteins for proteasome-mediated degradation (9), the rhythmic regulation of Hsp701b/Hspa1b may contribute in parallel to the regulation of protein degradation. The possible role of these and other proteasome-targeting genes in circadian clock function is supported by the recent finding that mutations of the Slimb gene, which functions in the ubiquitin-proteasome pathway, disrupt the circadian rhythm of activity in Drosophila (26, 35). It is also interesting that Vsp33b is rhythmically expressed in SCN2.2 cells, because this gene is thought to facilitate vesicle-mediated protein trafficking to lysosomes and participate in membrane docking and fusion between late endosomal and lysosomal compartments (24). The specific functions of Vsp33b and genes involved in proteasome targeting in SCN cells are not known, but the circadian regulation of these genes may enable the clock to both impose its influence across diverse cellular pathways and sustain the feedback loops comprising its core.

Circadian regulation of translational, posttranslational, and protein degradation elements suggests that the modulation of functional gene expression beyond the transcriptional level may contribute to dissemination of temporal information across cellular networks. For example, by regulating the expression of nuclear pore genes such as the nucleoporin p58/p45 and ribosomal genes such as Rpl6 and Rpl21, the circadian clock could directly modulate the efficiency of information flow from the nucleus and of ensuing polypeptide synthesis. This strategy could establish rhythms of bandwidth necessary to transmit large amounts of information across physiological nodes during certain times of the day. Furthermore, circadian regulation of posttranslational and proteasomal elements could provide node-specific control over the packaging and duration of such information volleys. In particular, the rhythmic expression of genes encoding heat shock proteins, components within the ubiquitin-proteasome pathway, and endosomal/lysosomal elements could impose the influence of time, respectively, on the folding dynamics of newly synthesized (or damaged) polypeptides, half-life of proteins targeted for ubiquitin-mediated degradation, and the degradation or recycling of extracellular and intracellular constituents. In this regard, it is noteworthy that peak expression of several heat shock-related genes, including Hsp105, Hspa5, and Hsp701b/Hspa1b, occurred at times that would be optimal for influencing PG I rhythms, which represent the majority of the circadian-regulated genes in SCN2.2 cells. Another interesting aspect of this GenMAPP analysis is that protein-processing elements responsive to internal stimuli such as glucose could play a role in adjusting how information is packaged and presented to specific physiological nodes. To exemplify this possibility, rhythms of glucose metabolism could harmonize information flow by modulating the induction of glucose-responsive heat shock proteins such as pre-mtHsp70/Grp75 and Hspa5/Grp78 (Fig. 3C) (48, 63). Collectively, the rhythmic regulation of translational, posttranslational, and protein degradation elements is likely to have a broad impact on the functional expression of rhythmic and stable gene transcripts.

Cellular development.
In SCN2.2 cells, rhythmic mRNA expression was observed in 13 genes that mediate cellular development. Circadian-regulated genes in this category were subdivided into three functional clusters: cell cycle (n = 3), 2) DNA/chromatin related (n = 3), and 3) growth and differentiation (n = 7).

Organisms ranging from cyanobacteria to mammals display rhythms of cell division that are "gated" by a circadian oscillator (10, 41, 50). The possible impact of the circadian clock in the indigenous control of the cell cycle in SCN cells is supported by the observation that SCN2.2 cells show rhythmic fluctuations of mRNA abundance in three genes involved in regulating the cell cycle. In SCN2.2 cells, the oscillations of mRNA expression observed in rat cyclin E, serine/threonine protein kinase (Pctaire-2) and the cell cycle protein p55CDC (Cdc20) (Supplemental Table S1) may serve to couple the clock mechanism to the cell cycle. In addition, another element of cell cycle regulation, cyclin L1 (Ccnl1), displayed rhythmic RNA expression in the rat SCN.

Rhythmic genes within the differentiation and growth cluster in SCN2.2 cells included the developmental genes, ornithine decarboxylase antizyme inhibitor (Oazi), and interferon-related developmental regulator 1 (Ifrd). These genes were also rhythmically expressed in the rat SCN. Another developmental gene, Bmp4, was rhythmically regulated in SCN2.2 cells. Although the specific functions of this gene within the SCN are unknown, Bmp4 has been shown to regulate transcriptional factors, some of which are also rhythmically expressed in both SCN2.2 cells and the rat SCN. In a variety of cell lines including embryonic stem cells, Bmp4 has been shown to induce mRNA expression for immediate early genes such as Ngfi-A (Egr1) and members of the Id gene family that inhibit the binding of basic helix-loop-helix (bHLH) transcription factors to DNA (30). The oscillations of these genes in SCN2.2 cells were validated either by parallel microarray analysis of the rat SCN or by qt-PCR.

Defense and detoxification.
SCN2.2 cells exhibited rhythmic patterns of mRNA expression for five genes involved in defense and detoxification (Table 1). Circadian-regulated genes in this category include the cell surface antigen (RT1-Aw2), p105 coactivator (U83883), an EST displaying similarity to mucin, prostaglandin-endoperoxidase synthase-2 (Ptgs2), and interleukin-6 receptor (Il6r). Inducible nitric oxide synthase (iNos or Nos2) was also categorized in the defense and detoxification category and represented a prominent example of a rhythmic gene that only surpassed secondary expression criteria (Supplemental Table S2). Nonetheless, circadian expression of iNos mRNA was verified in SCN2.2 cells by qt-PCR (Supplemental Fig. S1A) and was observed in a parallel analysis of the rat SCN. On the basis of evidence indicating that a competitive inhibitor of all three isoforms of nitric oxide synthase, nitro-L-arginine methyl ester (L-NAME), blocks glutamate- and NMDA-induced phase shifts of the SCN rhythm in neuronal firing rate in vitro (15, 76), the gaseous neurotransmitter nitric oxide (NO) is thought to be involved in the pathway by which glutamate mediates the phase-resetting action of light signals on the SCN clock. However, in vivo analyses using mutant mice do not appear to support this function for NO in the SCN, because animals lacking neuronal or endothelial isoforms of Nos (nNos and eNos, respectively) show normal circadian entrainment to light-dark cycles and phase-shifting responses to light (36, 37). The present evidence for the circadian regulation of iNos in both SCN2.2 cells and the rat SCN suggests that this isoform merits further analysis to address conflicting observations on NO function in the circadian resetting of the SCN clock mechanism by light.

Cytoskeleton and adhesion.
Five genes with cytoskeletal or cellular adhesion functions, integrin-{alpha}E1 (Itgae), CD44 antigen (Cd44), tissue inhibitor of metalloproteinase-1 (Timp 1), tropomyosin (Tmp1), and dynamin-like protein (Dnml1) exhibited circadian expression profiles in SCN2.2 cells. Circadian regulation in SCN2.2 cells was observed in the expression of 10 additional genes in this physiological category that surpassed secondary expression criteria (Supplemental Table S2), including troponin I type 3 (Tnni3), vascular cell adhesion molecule 1 (Vcam1), and neural cell adhesion molecule 1 (Ncam1). Rhythmic expression of cytoskeletal and cellular adhesion constituents within SCN cells is consistent with anatomical and functional evidence for SCN oscillations in cellular and synaptic plasticity and their role in regulation of circadian rhythms. In the hamster SCN, the distribution of astrocytes expressing glial fibrillary acidic protein fluctuates over the circadian cycle, and this structural rhythmicity may participate in the regulation of extracellular glutamate levels (38). Ncam has been specifically implicated in the regulation of SCN circadian function by studies demonstrating that transgenic mutant mice lacking different isoforms of this gene and an associated glycoprotein that regulates plastic interactions between nerve cells, polysialic acid, are distinguished by activity rhythms with decreased free-running periods and altered patterns of light-dark entrainment (67).

Summary.
The gene profiling analyses in the present study indicate that SCN-like global and temporal patterns of gene expression are conserved in the SCN2.2 transcriptome. SCN2.2 cells and the rat SCN show similar properties with regard to relative levels of mRNA expression for many different genes, oscillations in the expression of the core clock genes, Per2, Bmal1 (Mop3), and Cry1, and circadian regulation of many clock-controlled genes. Many of these circadian-regulated genes in SCN2.2 cells and the rat SCN were homologs or functionally related genes that exhibit rhythmic expression profiles in the murine SCN and/or liver (54, 70). Despite the similarities in circadian regulation of the transcriptome in SCN2.2 cells and the rat SCN, there was not complete overlap in the rhythmically regulated genes between these experimental models. Furthermore, the number of genes with circadian profiles in SCN2.2 cells was lower than that observed in the rat SCN and described previously for the murine SCN (54). It is possible that this inequality in the extent of circadian gene expression may be related to comparative differences in the cellular heterogeneity of SCN2.2 cells and the SCN in vivo. Although SCN2.2 cells were developed as a heterogeneous, rather than clonal, cell line to provide adequate representation of most SCN phenotypes, especially those with pacemaker properties, the immortalization strategy used to establish this line may have inherently selected for or against certain cell types. Consequently, circadian, and even global, gene expression may be more restricted in SCN2.2 cells, because this line does not express the full complement of cellular phenotypes found in the SCN due to selection bias in immortalization procedures. This possibility is supported by molecular and antigenic analyses indicating that many SCN2.2 cells exhibit SCN-like peptidergic phenotypes, but the overall proportion of peptidergic neurons in the line is less than that normally found in the SCN in situ (20). Alternatively, the diminished extent of circadian-regulated gene expression in SCN2.2 cells may reflect the absence of other neural and endocrine inputs that normally influence SCN rhythmicity in vivo. Nonetheless, comparison of circadian gene expression in SCN2.2 cells and the rat SCN provides a robust filter for identifying genes that are rhythmically regulated by the circadian clock mechanism indigenous to SCN cells.

In SCN2.2 cells, genes with circadian profiles were functionally diverse but were most frequently associated with metabolic, cellular and systems-level communication, and protein dynamics. The prevalence of rhythmic genes in these categories is consistent with studies documenting their importance in the circadian photoentrainment and pacemaker functions of the SCN. Furthermore, GenMAPP analysis of rhythmically regulated genes involved in fatty acid and steroid metabolism, G protein-coupled receptors, second messengers and mitogenic cascades, and protein dynamics has yielded insight into how SCN2.2 cells manage information flow so as to modulate input to or output from the clock mechanism. In particular, circadian regulation of genes involved in protein synthesis and degradation, sorting and trafficking, and modification or folding may extend clock control to processes that are not rhythmically orchestrated at the transcriptional level.

In addition to profiling the extent of circadian expression in SCN cells, the present analysis may have implications for understanding what molecular elements are necessary for the endogenous rhythm-generating and pacemaker properties of the SCN. Similar to the SCN in vivo, SCN2.2 cells are distinguished by the capacity to generate self-sustained rhythmicity in their own molecular or physiological processes and to restore behavioral rhythmicity to the entire animal when transplanted into SCN-lesioned hosts (21). Furthermore, these cells are capable of driving rhythms of clock gene expression and glucose metabolism in NIH/3T3 cells via the secretion of an unknown diffusible signal (3). Although circadian expression of various clock and clock-controlled genes can be induced in cultures of the rat-1 and NIH/3T3 fibroblasts by serum shock treatment or activation of various signal transduction pathways (1, 7, 8), these cell lines are nonrhythmic in the absence of this stimulatory input or SCN2.2-derived signals and incapable of conferring this induced oscillatory behavior to other cells (3). A possible explanation for the distinctions between the circadian properties of SCN2.2 cells and the rhythmic behavior of stimulated fibroblast lines is that fibroblasts may not express critical circadian output signals found in SCN cells. Consequently, comparison of circadian gene expression in SCN2.2 cells (or the rat SCN) with that in cells like NIH/3T3 fibroblasts or with other peripheral tissues, which contain all of the known components of the canonical clockworks, may provide an opportunity to identify candidate signals that mediate the pacemaking function of the SCN.


    GRANTS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 GRANTS
 REFERENCES
 
This study was conducted using the Texas A & M University Core Genomics Facility and supported by National Institutes of Health Program Project Grant P01-NS-39546 (D. J. Earnest, V. M. Cassone, and T. Thomas).


    ACKNOWLEDGMENTS
 
We thank Rodney Walline, Gregg C. Allen, and Sang Nahm for valuable technical contributions, and we thank Phil Beremand, Andrew Tag, and David Reed for assistance with analytical software and hardware.


    FOOTNOTES
 
Article published online before print. See web site for date of publication (http://physiolgenomics.physiology.org).

Address for reprint requests and other correspondence: D. J. Earnest, Texas A & M Univ. Health Science Center, Dept. of Human Anatomy and Medical Neurobiology, 238 Reynolds Medical Bldg., College Station, TX 77843-1114 (e-mail: dearnest{at}tamu.edu).

10.1152/physiolgenomics.00224.2004.

1 The Supplemental Material (Supplemental Tables S1–S4 and Supplemental Figs. S1 and S2) for this article is available online at http://physiolgenomics.physiology.org/cgi/content/full/00224.2004/DC1. Back


    REFERENCES
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 GRANTS
 REFERENCES
 

  1. Akashi M and Nishida E. Involvement of the MAP kinase cascade in resetting of the mammalian circadian clock. Genes Dev 14: 645–649, 2000.[Abstract/Free Full Text]
  2. Akhtar RA, Reddy AB, Maywood ES, Clayton JD, King VM, Smith AG, Gant TW, Hastings MH, and Kyriacou CP. Circadian cycling of the mouse liver transcriptome, as revealed by cDNA microarray, is driven by the suprachiasmatic nucleus. Curr Biol 12: 540–550, 2002.[CrossRef][ISI][Medline]
  3. Allen G, Rappe J, Earnest DJ, and Cassone VM. Oscillating on borrowed time: diffusible signals from immortalized suprachiasmatic nucleus cells regulate circadian rhythmicity in cultured fibroblasts. J Neurosci 21: 7937–7943, 2001.[Abstract/Free Full Text]
  4. Allen GC, Farnell Y, Bell-Pedersen D, Cassone VM, and Earnest DJ. Effects of altered clock gene expression on the pacemaker properties of SCN2.2 cells and oscillatory properties of NIH/3T3 cells. Neuroscience 127: 989–999, 2004.[CrossRef][ISI][Medline]
  5. Altschul SF, Gish W, Miller W, Myers EW, and Lipman DJ. Basic local alignment search tool. J Mol Biol 215: 403–410, 1990.[CrossRef][ISI][Medline]
  6. Bailey MJ, Beremand PD, Hammer R, Bell-Pedersen D, Thomas TL, and Cassone VM. Transcriptional profiling of the chick pineal gland, a photoreceptive circadian oscillator and pacemaker. Mol Endocrinol 17: 2084–2095, 2003.[Abstract/Free Full Text]
  7. Balsalobre A, Damiola F, and Schibler U. A serum shock induces circadian gene expression in mammalian tissue culture cells. Cell 93: 929–937, 1998.[CrossRef][ISI][Medline]
  8. Balsalobre A, Brown SA, Marcacci L, Tronche F, Kellendonk C, Reichardt HM, Schutz G, and Schibler U. Resetting of circadian time in peripheral tissues by glucocorticoid signaling. Science 289: 2344–2347, 2000.[Abstract/Free Full Text]
  9. Bercovich B, Stancovski I, Mayer A, Blumenfeld N, Laszlo A, Schwartz AL, and Ciechanover A. Ubiquitin-dependent degradation of certain protein substrates in vitro requires the molecular chaperone Hsc70. J Biol Chem 272: 9002–9010, 1997.[Abstract/Free Full Text]
  10. Carbajo-Perez E, Carbajo S, Orfao A, Vicente-Villardon JL, and Vazquez R. Circadian variation in the distribution of cells throughout the different phases of the cell cycle in the anterior pituitary gland of adult male rats as analysed by flow cytometry. J Endocrinol 129: 329–333, 1991.[Abstract]
  11. Castel M, Belenky M, Cohen S, Ottersen OP, and Storm-Mathisen J. Glutamate-like immunoreactivity in retinal terminals of the mouse suprachiasmatic nucleus. Eur J Neurosci 5: 368–381, 1993.[ISI][Medline]
  12. Ceriani MF, Gre JB, Yanovsky M, Panda S, Straume M, and Kay SA. Genome-wide expression analysis in Drosophila reveals genes controlling circadian behavior. J Neurosci 22: 9305–9319, 2002.[Abstract/Free Full Text]
  13. Correa A, Lewis ZA, Greene AV, March IJ, Gomer RH, and Bell-Pedersen D. Multiple oscillators regulate circadian gene expression in Neurospora. Proc Natl Acad Sci USA 100: 13597–13602, 2003.[Abstract/Free Full Text]
  14. Dahlquist KD, Salomonis N, Vranizan K, Lawlor SC, and Conklin BR. GenMAPP, a new tool for viewing and analyzing microarray data on biological pathways. Nat Genet 31: 19–20, 2002.[CrossRef][ISI][Medline]
  15. Ding JM, Chen D, Weber ET, Faiman LE, Rea MA, and Gillette MU. Resetting the biological clock: mediation of nocturnal circadian shifts by glutamate and NO. Science 266: 1713–1717, 1994.[ISI][Medline]
  16. Doniger SW, Salomonis N, Dahlquist KD, Vranizan K, Lawlor SC, and Conklin BR. MAPPFinder: using Gene Ontology and GenMAPP to create a global gene-expression profile from microarray data. Genome Biol 4: R7, 2003.[CrossRef][Medline]
  17. Duffield GE. DNA microarray analyses of circadian timing: the genomic basis of biological time. J Neuroendocrinol 15: 991–1002, 2003.[CrossRef][ISI][Medline]
  18. Duffield GE, Best JD, Meurers BH, Bittner A, Loros JJ, and Dunlap JC. Circadian programs of transcriptional activation, signaling and protein turnover revealed by microarray analysis of mammalian cells. Curr Biol 12: 551–557, 2002.[CrossRef][ISI][Medline]
  19. Earnest DJ and Sladek CD. Circadian vasopressin release from perifused rat suprachiasmatic explants in vitro: effects of acute stimulation. Brain Res 422: 398–402, 1987.[CrossRef][ISI][Medline]
  20. Earnest DJ, Liang FQ, DiGiorgio S, Gallagher M, Harvey B, Earnest B, and Seigel G. Establishment and characterization of adenoviral E1A immortalized cell lines derived from the rat suprachiasmatic nucleus. J Neurobiol 39: 1–13, 1999.[CrossRef][ISI][Medline]
  21. Earnest DJ, Liang FQ, Ratcliff M, and Cassone VM. Immortal time: circadian clock properties of rat suprachiasmatic cell lines. Science 283: 693–695, 1999.[Abstract/Free Full Text]
  22. Gannon RL and Rea MA. Glutamate receptor immunoreactivity in the rat suprachiasmatic nucleus. Brain Res 622: 337–342, 1993.[CrossRef][ISI][Medline]
  23. Gillette MU and Reppert SM. The hypothalamic suprachiasmatic nuclei: circadian patterns of vasopressin secretion and neuronal activity in vitro. Brain Res Bull 19: 135–139, 1987.[CrossRef][ISI][Medline]
  24. Gladstone Institutes. Gene archives and GenMAPPs. In: GenMAPP: Gene Map Annotator and Pathway Profiler. http://www.GenMAPP.org (10 Sept. 2003).
  25. Golden SS and Canales SR. Cyanobacterial circadian clocks—timing is everything. Nat Rev Microbiol 1: 191–199, 2003.[CrossRef][ISI][Medline]
  26. Grima B, Lamouroux A, Chelot E, Papin C, Limbourg-Bouchon B, and Rouyer F. The F-box protein slimb controls the levels of clock proteins period and timeless. Nature 420: 178–182, 2002.[CrossRef][ISI][Medline]
  27. Harley CW, Farrell RC, and Rusak B. Daily variation in the distribution of glycogen phosphorylase in the suprachiasmatic nucleus of syrian hamsters. J Comp Neurol 435: 249–258, 2001.[CrossRef][ISI][Medline]
  28. Heid CA, Stevens J, Livak KJ, and Williams PM. Real time quantitative PCR. Genome Res 6: 986–994, 1996.[Abstract]
  29. Hille B. Potassium channels and chloride channels. In: Ionic Channels of Excitable Membranes, edited by Sinauer HB. Sunderland, MA: Sinauer, 1992, p. 130–135.
  30. Hollnagel A, Oehlmann V, Heymer J, Rüther U, and Nordheim A. Id genes are direct targets of bone morphogenetic protein induction in embryonic stem cells. J Biol Chem 274: 19838–19845, 1999.[Abstract/Free Full Text]
  31. Holmes RP and Assimos DG. Glyoxylate synthesis, and its modulation and influence on oxalate synthesis. J Urol 160: 1617–1624, 1998.[CrossRef][ISI][Medline]
  32. Hurst WJ, Mitchell JW, and Gillette MU. Synchronization and phase-resetting by glutamate of an immortalized SCN cell line. Biochem Biophys Res Commun 298: 133–143, 2002.[CrossRef][ISI][Medline]
  33. Kleiber J, Kalousek F, Swaroop M, and Rosenberg LE. The general mitochondrial matrix processing protease from rat liver: structural characterization of the catalytic subunit. Proc Natl Acad Sci USA 87: 7978–7982, 1990.[Abstract/Free Full Text]
  34. Klein DC, Moore RY, and Reppert SM (Editors). Suprachiasmatic Nucleus: The Mind’s Clock. New York: Oxford Univ. Press, 1991.
  35. Ko HW, Jiang J, and Edery I. Role of Slimb in the degradation of Drosophila protein phosphorylated by doubletime. Nature 240: 673–678, 2002.
  36. Kriegsfeld LJ, Demas GE, Lee SE, Dawson TM, Dawson VL, and Nelson RJ. Circadian locomotor analysis of male mice lacking the gene for neuronal nitric oxide synthase (nNOS–/–). J Biol Rhythms 14: 20–27, 1999.[Abstract/Free Full Text]
  37. Kriegsfeld LJ, Drazen DL, and Nelson RJ. Circadian organization in male mice lacking the gene for endothelial nitric oxide synthase (eNOS–/–). J Biol Rhythms 16: 142–148, 2001.[Abstract/Free Full Text]
  38. Lavialle M and Servière J. Circadian fluctuations in GFAP distribution in the Syrian hamster suprachiasmatic nucleus. Neuroreport 4: 1243–1246, 1993.[ISI][Medline]
  39. Lee CK, Klopp RG, Weidruch R, and Prolla TA. Gene expression profile of aging and its retardation by caloric restriction. Science 285: 1390–1393, 1999.[Abstract/Free Full Text]
  40. Lovenberg TW, Baron BM, de Lecea L, Miller JD, Prosser RA, Rea MA, Foye PE, Racke M, Slone AL, Siegel BW, Danielson PE, Sutcliffe JG, and Erlander MG. A novel adenyl cyclase-activating serotonin receptor (5-HT7) implicated in the regulation of mammalian circadian rhythms. Neuron 11: 449–458, 1993.[CrossRef][ISI][Medline]
  41. Matsuo T, Yamaguchi S, Mitsui S, Emi A, Shimoda F, and Okamura H. Control mechanism of the circadian clock for timing of cell division in vivo. Science 302: 255–259, 2003.[Abstract/Free Full Text]
  42. McCallum CD, Do H, Johnson AE, and Frydman J. The interaction of the chaperonin tailless complex polypeptide 1 (TCP1) ring complex (TRiC) with ribosome-bound nascent chains examined using photo-cross-linking. J Cell Biol 149: 591–602, 2000.[Abstract/Free Full Text]
  43. McDonald MJ and Rosbash M. Microarray analysis and organization of circadian gene expression in Drosophila. Cell 107: 567–578, 2001.[CrossRef][ISI][Medline]
  44. McGowan KM, Long SD, and Pekala PH. Glucose transporter gene expression: regulation of transcription and mRNA stability. Pharmacol Ther 66: 465–505, 1995.[CrossRef][ISI][Medline]
  45. Meijer JH and Schwartz WJ. In search of the pathways for light-induced pacemaker resetting in the suprachiasmatic nucleus. J Biol Rhythms 18: 235–249, 2003.[Abstract/Free Full Text]
  46. Meijer JH, van der Zee EA, and Dietz M. Glutamate phase shifts circadian activity rhythms in hamsters. Neurosci Lett 86: 177–183, 1988.[CrossRef][ISI][Medline]
  47. Mirande M and Waller JP. Molecular cloning and primary structure of cDNA encoding the catalytic domain of rat liver aspartyl-tRNA synthetase. J Biol Chem 264: 842–847, 1989.[Abstract/Free Full Text]
  48. Mizzen LA, Chang C, Garrels JI, and Welch WJ. Identification, characterization, and purification of two mammalian stress proteins present in mitochondria, grp 75, a member of the hsp 70 family and hsp 58, a homolog of the bacterial groEL protein. J Biol Chem 264: 20664–20675, 1989.[Abstract/Free Full Text]
  49. Moore RY. Organization and function of a central nervous system circadian oscillator: the suprachiasmatic hypothalamic nucleus. Fed Proc 42: 2783–2789, 1983.[ISI][Medline]
  50. Mori T, Binder B, and Johnson CH. Circadian gating of cell division in cyanobacteria growing with average doubling times of less than 24 hours. Proc Natl Acad Sci USA 93: 10183–10188, 1996.[Abstract/Free Full Text]
  51. National Center for Biotechnology Information. Tools for data mining. In: National Center for Biotechnology Information Homepage. http://www.ncbi.nlm.nih.gov/ (10 Sept. 2003).
  52. Newman GC, Hospod FE, Patlak CS, and Moore RY. Analysis of in vitro glucose utilization in a circadian pacemaker model. J Neurosci 12: 2015–2021, 1992.[Abstract]
  53. Niethammer M, Kim E, and Sheng M. Interaction between the C terminus of NMDA receptor subunits and multiple members of the PSD-95 family of membrane-associated guanylate kinases. J Neurosci 16: 2157–2163, 1996.[Abstract]
  54. Panda S, Antoch MP, Miller BH, Su AI, Schook AB, Straume M, Schultz G, Kay SA, Takahashi JS, and Hogenesch JB. Coordinated transcription of key pathways in the mouse by the circadian clock. Cell 109: 307–320, 2002.[CrossRef][ISI][Medline]
  55. Pellerin L and Magistretti PJ. Neuroenergetics: calling upon astrocytes to satisfy hungry neurons. Neuroscientist 10: 53–62, 2004.[Abstract/Free Full Text]
  56. Pittendrigh CS. Temporal organization: reflections of a Darwinian clock-watcher. Annu Rev Physiol 55: 16–54, 1993.[ISI][Medline]
  57. Prosser RA, Dean RR, Edgar DM, Heller HC, and Miller JD. Serotonin and the mammalian circadian system. I. In vitro phase shifts by serotonergic agonists and antagonists. J Biol Rhythms 8: 1–16, 1993.[ISI][Medline]
  58. Pruitt KD and Maglott DR. RefSeq and LocusLink: NCBI gene-centered resources. Nucleic Acids Res 29: 137–140, 2001.[Abstract/Free Full Text]
  59. Ralph MR, Foster RG, Davis FC, and Menaker M. Transplanted suprachiasmatic nucleus determines circadian period. Science 247: 975–978, 1990.[ISI][Medline]
  60. Rea MA, Glass JD, and Colwell CS. Serotonin modulates photic responses in the hamster suprachiasmatic nuclei. J Neurosci 14: 3635–3642, 1994.[Abstract]
  61. Reppert SM and Weaver DR. Molecular analysis of mammalian circadian rhythms. Annu Rev Physiol 63: 647–676, 2001.[CrossRef][ISI][Medline]
  62. Reppert SM and Weaver DR. Coordination of circadian timing in mammals. Nature 418: 935–941, 2002.[CrossRef][ISI][Medline]
  63. Resendez E Jr, Wooden SK, and Lee AS. Identification of highly conserved regulatory domains and protein-binding sites in the promoters of the rat and human genes encoding the stress-inducible 78-kilodalton glucose-regulated protein. Mol Cell Biol 8: 4579–4584, 1988.[ISI][Medline]
  64. Rutter J, Reick M, and McKnight SL. Metabolism and the control of circadian rhythms. Annu Rev Biochem 71: 307–331, 2002.[CrossRef][ISI][Medline]
  65. Schaffer R, Landgraf J, Accerbi M, Simon V, Larson M, and Wisman E. Microarray analysis of diurnal and circadian-regulated genes in Arabidopsis. Plant Cell 13: 113–123, 2001.[Abstract/Free Full Text]
  66. Schwartz WJ. SCN metabolic activity in vitro. In: Suprachiasmatic Nucleus: The Mind’s Clock, edited by Klein DC, Moore RY, and Reppert SM. New York: Oxford Univ. Press, 1991, p. 144–156.
  67. Shen H, Watanabe M, Tomasiewicz H, and Glass JD. Genetic deletions of NCAM and PSA impair circadian function in the mouse. Physiol Behav 73: 185–193, 2001.[CrossRef][ISI][Medline]
  68. Shirakawa S and Moore RY. Glutamate shifts the phase of the circadian neuronal firing rhythm in the rat suprachiasmatic nucleus in vitro. Neurosci Lett 178: 47–50, 1994.[CrossRef][ISI][Medline]
  69. Storch KF, Lipan O, Leykin I, Viswanathan N, Davis FC, Wong WH, and Weitz CJ. Extensive and divergent circadian gene expression in liver and heart. Nature 417: 78–83, 2002.[CrossRef][ISI][Medline]
  70. Ueda HR, Chen W, Adachi A, Wakamatsu H, Hayashi S, Takasugi T, Nagano M, Nakahama K, Suzuki Y, Sugano S, Iino M, Shigeyoshi Y, and Hashimoto S. A transcription factor response element for gene expression during circadian night. Nature 418: 534–539, 2002.[CrossRef][ISI][Medline]
  71. Ueda HR, Matsumoto A, Kawamura M, Iino M, Tanimura T, and Hashimoto S. Genome-wide transcriptional orchestration of circadian rhythms in Drosophila. J Biol Chem 277: 14048–14052, 2002.[Abstract/Free Full Text]
  72. van den Pol AN, Hermans-Borgmeyer I, Hofer M, Ghosh P, and Heinemann S. Ionotropic glutamate-receptor gene expression in hypothalamus: localization of AMPA, kainate, and NMDA receptor RNA with in situ hybridization. J Comp Neurol 343: 428–444, 1994.[CrossRef][ISI][Medline]
  73. Vannucci SJ and Simpson IA. Developmental switch in brain nutrient transporter expression in the rat. Am J Physiol Endocrinol Metab 285: E1127–E1134, 2003.[Abstract/Free Full Text]
  74. Wang D, Kranz-Eble P, and De Vivo DC. Mutational analysis of GLUT1 (SLC2A1) in Glut-1 deficiency syndrome. Hum Mutat 16: 224–231, 2000.[CrossRef][ISI][Medline]
  75. Wang J, Yang X, Zhou P, and Han H. Cloning of mouse genomic ribosomal protein L6 and analysis of its promoter. Biochim Biophys Acta 1576: 219–224, 2002.[ISI][Medline]
  76. Watanabe A, Hamada T, Shibata S, and Watanabe S. Effects of nitric oxide synthase inhibitors on N-methyl-D-aspartate-induced phase delay of circadian rhythm of neuronal activity in the rat suprachiasmatic nucleus in vitro. Brain Res 646: 161–164, 1994.[CrossRef][ISI][Medline]
  77. Young MW. Life’s 24-hour clock: molecular control of circadian rhythms in animal cells. Trends Biochem Sci 25: 601–606, 2000.[CrossRef][ISI][Medline]




This Article
Abstract
Full Text (PDF)
Supplemental Tables and Figures
All Versions of this Article:
21/3/370    most recent
00224.2004v1
Citation Map
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Download to citation manager
Google Scholar
Articles by Menger, G. J.
Articles by Earnest, D. J.
Articles citing this Article
PubMed
PubMed Citation
Articles by Menger, G. J.
Articles by Earnest, D. J.


HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Visit Other APS Journals Online
Copyright © 2005 by the American Physiological Society.