Physiol. Genomics AJP: Endocrinology and Metabolism
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Physiol. Genomics 29: 280-289, 2007. First published February 6, 2007; doi:10.1152/physiolgenomics.00199.2006 Free Article
1094-8341/07 $8.00
This Article
Free upon publication Free Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Figure and Tables
Right arrow All Versions of this Article:
29/3/280    most recent
00199.2006v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Menger, G. J.
Right arrow Articles by Earnest, D. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Menger, G. J.
Right arrow Articles by Earnest, D. J.
Received 11 September 2006; accepted in final form 1 February 2007.
Physiological Genomics 29:280-289 (2007)
1094-8341/07 $8.00 © 2007 American Physiological Society

Circadian profiling of the transcriptome in NIH/3T3 fibroblasts: comparison with rhythmic gene expression in SCN2.2 cells and the rat SCN

Gus J. Menger2,3, Gregg C. Allen1,2, Nichole Neuendorff1, Sang-Soep Nahm1,2, Terry L. Thomas2,3, Vincent M. Cassone2,3 and David J. Earnest1,2,3

1 Department of Neuroscience and Experimental Therapeutics, Texas A&M University Health Science Center, College of Medicine
2 Center for Research on Biological Clocks, 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
 
To screen for output signals that may distinguish the pacemaker in the mammalian suprachiasmatic nucleus (SCN) from peripheral-type oscillators in which the canonical clockworks are similarly regulated in a circadian manner, the rhythmic behavior of the transcriptome in forskolin-stimulated NIH/3T3 fibroblasts was analyzed and compared relative to SCN2.2 cells in vitro and the rat SCN. Similar to the circadian profiling of the SCN2.2 and rat SCN transcriptomes, NIH/3T3 fibroblasts exhibited circadian fluctuations in the expression of the core clock genes, Per2, Cry1, and Bmal1, and 323 functionally diverse transcripts, many of which regulate cellular communication. Overlap in rhythmic transcripts among NIH/3T3 fibroblasts, SCN2.2 cells, and the rat SCN was limited to these clock genes and four other genes that mediate fatty acid and lipid metabolism or function as nuclear factors. Compared with NIH/3T3 cells, circadian gene expression in SCN oscillators was more prevalent among genes mediating glucose metabolism and neurotransmission. Coupled with evidence for the rhythmic regulation of the inducible isoform of nitric oxide synthase (iNos) in SCN2.2 cells and the rat SCN but not in fibroblasts, studies examining the effects of a NOS inhibitor on metabolic rhythms in cocultures containing SCN2.2 cells and untreated NIH/3T3 cells suggest that the gaseous neurotransmitter nitric oxide may play a key role in SCN pacemaker function. This comparative analysis of circadian gene expression in SCN and non-SCN cells may have important implications in the selective analysis of circadian signals involved in the coupling of SCN oscillators and regulation of rhythmicity in downstream cells.

suprachiasmatic nucleus; pacemaker; oscillator; rhythm; clock genes; Per1; Per2; Bmal1 (Mop3); Cry1; Cry2


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 GRANTS
 REFERENCES
 
CIRCADIAN CLOCKS HAVE EVOLVED as endogenous biological timekeeping mechanisms that coordinate a wide range of biological processes with the 24-h period of the daily solar cycle. In mammals, the suprachiasmatic nuclei (SCN) of the anterior hypothalamus function as a master clock that regulates overt circadian rhythms throughout the organism (18). The SCN clock is also capable of producing self-sustained circadian oscillations in many of its intrinsic molecular and cellular activities. SCN cells are marked by circadian regulation of neuropeptide secretion, cellular metabolism, and electrical activity in vivo and in vitro (13, 32). Moreover, recent gene profiling studies indicate that up to 10% of the transcriptome is rhythmically expressed in the SCN in vivo and in immortalized rat SCN cells (26, 29). These endogenous molecular oscillations in the SCN are a critical property of the clock mechanism and presumably its output signals that regulate circadian rhythmicity in other cells or tissues. The SCN timekeeping mechanism consists of interlocking feedback loops in which the core components, Bmal1 (Mop3), Period1 (Per1), Per2, Cryptochrome1 (Cry1), Cry2, and Rev-erb{alpha}, are rhythmically regulated through mutual interactions with their protein products. In turn, the molecular feedback loops involving these genes coordinate downstream rhythmicity in SCN-specific output genes that mediate clock-controlled oscillations in other neural or endocrine tissues.

Although the SCN is vital to the function of the mammalian circadian system, the expression and rhythmic regulation of these core clock genes are not confined to this neural locus. Instead, oscillations in the same genes forming the molecular core of the SCN clockworks occur widely in many peripheral cells and tissues (33, 38, 39). Studies using in vitro models have yielded compatible evidence indicating that peripheral or non-SCN cells are also capable of expressing clock gene oscillations. Rat-1 and NIH/3T3 fibroblasts exposed to serum shock or forskolin treatment are similar to immortalized cells derived from the rat SCN (SCN2.2) with regard to rhythmic fluctuations in clock gene mRNA abundance (1, 3, 4, 6). However, this similarity in the oscillatory nature of core clock elements does not provide for equivalent functional properties, because SCN cells have the capacity to act as pacemakers whereas peripheral tissues and cultured fibroblasts do not. For example, SCN2.2 cells, but not fibroblasts, restore circadian behavior when transplanted into arrhythmic, SCN-lesioned hosts in vivo (11) and also confer metabolic and molecular oscillations to cocultured cells via the secretion of an unknown diffusible signal (3). Collectively, these findings raise the important question of why the molecular machinery found in the SCN and SCN2.2 cells does not propagate similar pacemaker function in fibroblasts.

A potential explanation is that SCN output signals necessary for its function in coordinating rhythmicity among individual clock cells and in downstream oscillators may not be expressed or circadian-regulated in stimulated fibroblasts. In the present study, we explored this possibility by analyzing global and circadian facets of gene expression in NIH/3T3 fibroblasts. The transcriptome in forskolin-stimulated fibroblasts was profiled over two circadian cycles and then compared with the patterns of gene expression observed in SCN2.2 cells and the rat SCN (26). This comparison of the NIH/3T3 and SCN transcriptomes was used to identify the extent of overlap and fundamental differences in their basic expression and circadian regulation of specific genes. Using bioinformatic tools, we next examined the functional distribution of both nonrhythmic and circadian-regulated genes that were found only in forskolin-stimulated NIH/3T3 fibroblasts or in SCN2.2 cells and the rat SCN. Because SCN-specific rhythms of gene expression were especially evident within pathways associated with intercellular communication, we probed implications of this information in differentiating candidate signals for the synchronization of SCN clock cells or pacemaker regulation of circadian rhythmicity in other cells. Based on evidence for the role of nitric oxide (NO) in resetting the SCN clock (8, 37) and the present observation that the inducible isoform of nitric oxide synthase (iNos or Nos2), an isozyme involved in NO production, is rhythmically expressed in the rat SCN and SCN2.2 cells but not in NIH/3T3 fibroblasts, this diffusible messenger was targeted for functional analysis. These experiments used our coculture model to determine whether treatment with the NOS inhibitor, NG-nitro-L-arginine methyl ester (L-NAME) alters the endogenous oscillatory and circadian pacemaker functions of SCN2.2 cells.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 GRANTS
 REFERENCES
 
NIH/3T3 cultures and RNA extraction.
For each of three biological replicates, NIH/3T3 cells (passages 4 and 5) were seeded on culture dishes (60 mm; Corning, Corning, NY) and maintained at 37°C and 5% CO2 in Dulbecco's minimum essential medium (DMEM; Invitrogen, Carlsbad, CA) supplemented with 20% fetal bovine serum (FBS; HyClone, Logan, UT), 3,000 µg/ml glucose and 292 µg/ml L-glutamine. At 48-h intervals, the medium was changed and cultures were expanded 1:3 or 1:4. Prior to experimental analysis, cultures were plated in multiple T-75 flasks. At ~24 and 44 h after plating, the culture medium was changed so as to respectively lower the FBS concentration to 10% and then to 5%. To facilitate cell cycle and circadian oscillation synchronization across cultures, cells were subjected to medium replacement and exposed to serum-free medium (DMEM) containing 15 µM forskolin (Calbiochem, La Jolla, CA) for 2 h. Cultures were rinsed and thereafter maintained in serum-free DMEM. Immediately after forskolin treatment, cells were harvested from individual flasks (approximate density: 1.3 x 106 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 ethanol precipitations.

Affymetrix U74v2 GeneChip analysis.
Prior to analysis, the quality of all NIH/3T3 RNA samples was assessed by electrophoresis on 1% agarose gels containing 0.1 µg/ml ethidium bromide. Labeled cRNAs were produced from purified RNA collected at each time point and hybridized on mouse U74v2 GeneChips. Experimental procedures including double-stranded cDNA synthesis and biotinylated cRNA preparation were conducted according to recommended protocols described in Affymetrix GeneChip Expression Analysis Technical Manual (23) and processed as previously described for microarray analysis of SCN2.2 cells and the rat SCN (26). To assure the quality of labeling and fragmentation efficiency, unfragmented and fragmented cRNA products were analyzed on an Agilent 2100 bioanalyzer prior to hybridization on arrays. Fragmented biotinylated cRNA (15 µg) was hybridized on Affymetrix GeneChip mouse U74 version 2 arrays at 45°C and 60 rpm in a GeneChip Hybridization Oven 640 (Affymetrix, Santa Clara, CA) for 16 h.

Following 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 antistreptavidin biotinylated antibody (Vector Laboratories, Burlingame, CA), 1x morpholine ethane sulfonic 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 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 prior to uploading into GeneSpring 7.3 software (Agilent Technologies, Palo Alto, CA).

Data processing and global analyses.
GeneChip signal intensity data from three biological replicates of NIH/3T3 cells were uploaded into GeneSpring 7.2 and filtered in a similar fashion to that used for SCN2.2 cells (26). 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 NIH/3T3 cells. 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. 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 7.3 "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 by methods similar to those described previously (7, 25, 26). 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). In addition, we imposed the requirement that rhythmic transcripts show at least one pair of nonoverlapping standard error bars between time points with the highest and lowest values. The analytical and amplitude criteria used to identify cycling transcripts in NIH/3T3 cells are consistent with those applied in experimental analyses using cDNA or oligonucleotide arrays to profile circadian gene expression (2, 7, 25, 26, 31). Microarray data have been deposited in the National Center for Biotechnology Information's (NCBI) Gene Expression Omnibus database (accession numbers GSE 1654, 1673, 5810).

To assess the false positive rate for profiling rhythmically expressed genes in our analysis, data files were randomized by 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 written to the output file sequentially, 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 analytical and amplitude criteria used to identify cycling transcripts in the original data sets for NIH/3T3 cells. Based on this randomization analysis, we estimate false positive rates of ~21% for NIH/3T3 cells.

Bioinformatics and validation.
Because some genes surpassing circadian expression filters were represented on the mouse U74v2 GeneChip multiple times, we used GeneSpring 7.3 and bioinformatic tools available as links from the NCBI (27, 30), including basic local alignment search tool (5), to identify unique genes with circadian profiles in NIH/3T3 fibroblasts. To compare the expression profiles of specific genes in these fibroblasts with those found in SCN2.2 cells or the rat SCN, we identified genes that were commonly represented on the mouse (U74v2) and rat (U34a) GeneChips using the "Genome Homology" tool in GeneSpring 7.3. For a small subset of clock and clock-controlled genes, quantitative PCR (qt-PCR) was used to validate rhythmic expression in NIH/3T3 fibroblasts. Circadian expression of cAMP responsive element modulator (Crem), fatty acid synthase (Fasn), aryl hydrocarbon receptor nuclear translocator (Arnt), fibroblastic growth factor 7 (Fgf7), stearoyl CoA desaturase (Scd1), the transcription factor Sef/If2 (Tcf4) and the core clock genes Per2, Bmal1 (Mop3), Cry1, was validated by qt-PCR. Quantification of relative mRNA abundance was performed using 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. Using the cDNA equivalent of 30 ng of total RNA, we then PCR amplified duplicate aliquots of each sample in an ABI PRISM 7700 sequence detection system (16). The following probes and primers were designed using PrimerExpress software (ABI): mPeriod 2 forward: 5'-ATGCTCGCCATCCACAAGA-3'; mPeriod 2 reverse: 5'-GCGGAATCGAATGGGAGAAT-3'; mCry1 forward: 5'-CTGGCGTGGAAGTCATCGT-3'; mCry1 reverse: 5'-CTGTCCGCCATTGAGTTCTATG-3'; mBmal1 (Mop3) forward: 5'-CCAAGAAAGTATGGACACAGACAAA-3'; mBmal1 (Mop3) reverse: 5'-GCATTCTTGATCCTTCCTTGGT-3'; Aryl hydrocarbon forward: 5'-GCATGGGCTCACGAAGGT-3'; Aryl hydrocarbon reverse: 5'-AACAGGGTCCACGGAGCTAGT-3'; cAMP responsive element modulator forward: 5'-GGCTGCTGCCACAGGTG-3'; cAMP responsive element modulator reverse: 5'-CACCTTGTGGCAAAGCAGTAGT-3'; Fgf7 forward: 5'-AAGGGACCCAGGAGATGAAGA-3'; Fgf7 reverse: 5'-TGCCACAATTCCAACTGCC-3'; Scd1 forward: 5'-AACACCATGGCGTTCCAAA-3'; Scd1 reverse: 5'-GGTGGGCGCGGTGAT-3'; CypA forward: 5'-TGTGCCAGGGTGGTGACTT-3'; CypA reverse: 5'-TCAAATTTCTCTCCGTAGATGGACTT-3'.

To control for differences in sample RNA content, either cyclophilin A (CypA) mRNA or ribosomal RNA (rRNA) was amplified with the cDNA equivalent of 1 ng of total RNA from the same samples. For all reactions, the comparative CT method described in the ABI Prism 7700 Sequence Detection System User Bulletin #2 was utilized to calculate the relative mRNA abundance for a given target gene. Using this method, we normalized the amount of target gene mRNA in each sample first to corresponding CypA mRNA or rRNA levels and then relative to a calibrator consisting of pooled cDNA from multiple samples that was analyzed on each reaction plate. Relative abundance of target mRNA was represented as a percentage of the maximal value obtained within an individual experiment.

Analysis of circadian phase and gene tree clustering.
Consistent with our transcriptional profiling analysis of SCN2.2 cells (26), the phase of rhythmic genes in NIH/3T3 cells was determined by comparing the first peak of cyclic mRNA abundance (maxima of normalized signal intensity) in relation to the first four sampling intervals and to the circadian expression profiles of core clock genes. Based on this comparative relationship, rhythmically expressed genes showing peak mRNA abundance during the first (hour 0), second (hour 6), third (hour 12), or fourth (hour 18) sampling interval with recurrent 24-h peaks thereafter were respectively assigned to phase group (PG) I, II, III, and IV. For genes within each phase group, gene tree clustering was used to identify clusters of transcripts with similar expression patterns. Genes displaying common expression patterns were positioned on nearby branches separated by a distance of 0.02 or less using the Gene Tree clustering tool in GeneSpring 7.3. Clustered branches fulfilling this separation threshold were comparatively labeled in relation to the circadian expression profiles of clock or other clock-controlled genes within the cluster.

NO and functional properties of SCN2.2 cells.
Using our coculture model (3), this experiment was conducted to examine the effects of L-NAME, a highly competitive, slowly reversible inhibitor of all three isoforms of NOS in the central nervous system (CNS) (22), on the generation of endogenous metabolic rhythms and/or the communication this rhythmicity to cocultured NIH/3T3 fibroblasts. SCN2.2 cells derived from a single passage were plated on multiple six-well companion plates (Falcon, Oxnard, CA), and these cultures were treated at confluence with 100 µM L-NAME (Calbiochem, La Jolla, CA) or its inactive enantiomer, NG-nitro-D-arginine methyl ester (D-NAME, 100 µM), for 48 h (0–48 h). In parallel to these L-NAME- and D-NAME-treated SCN2.2 cells, cultures of NIH/3T3 fibroblasts were established and maintained separately on cell-impermeable inserts (23 mm, pore size = 1 µm). Following experimental treatment, SCN2.2 cells were re-exposed to normal growth medium and then cocultured for 60 h with inserts containing untreated NIH/3T3 cells (48–108 h). Samples were collected at 4-h intervals to determine 2-deoxyglucose (2-DG) uptake in SCN2.2 cells during the period of L-NAME treatment (0–48 h) and in both coculture compartments for the remainder of the experiment (48–108 h). This analysis was performed on three companion wells or inserts at each time point. Individual samples were first analyzed with a liquid scintillation counter to determine 2-DG uptake and then processed for extraction of soluble protein.

Measurement of 2-DG uptake.
SCN2.2 and NIH/3T3 cells were assayed for uptake of 2-DG using methods described previously (3, 11, 28). Confluent cultures were incubated for 1 h with 14C-labeled 2-DG (0.1 mCi/ml, American Radiological) diluted into culture medium. Following this incubation, cells were rinsed twice with Dulbecco's PBS (dPBS) without calcium or magnesium and removed from companion wells and inserts by trypsinization at 37°C for 5 min. Cells were then harvested by lysis with dPBS/1% SDS buffer. Aliquots of cell lysates (100–200 µl) were placed in scintillation vials in triplicate and then counted on a Beckman scintillation counter (Fullerton, CA). Determinations of 2-DG uptake were normalized for sample protein content as measured by the bicinchoninic acid method (Pierce, Rockford, IL). Peak values for rhythmic 2-DG uptake were identified when relative 2-DG levels were at least 1.5-fold higher than those observed during the preceding or succeeding minimum. Time-dependent fluctuations were identified by one-way analysis of variance. Paired comparisons between determinations of 2-DG uptake at specific time points were analyzed post hoc for statistical differences by the Newman-Keuls sequential range test.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
GRANTS
 REFERENCES
 
Global and circadian properties of gene expression in NIH/3T3 cells.
Global expression and circadian regulation of the NIH/3T3 transcriptome was examined over two cycles. In forskolin-stimulated NIH/3T3 cells, 5,830 (47%) of the 12,500 probe sets on the mouse U74v2 GeneChip fulfilled expression criteria. Of these detected probe sets, most (94%) displayed either stable (n = 2,391) or noncircadian (n = 3,111) patterns of expression (Table 1, top). Relative to the total number of probe sets, a small proportion (2.6%) consisting of 157 unique genes and 166 expressed sequence tags (ESTs) or genes of unknown function was distinguished by circadian profiles in fibroblasts (Fig. 1). Oscillations in NIH/3T3 expression profiles for some of these transcripts and the core clock genes Per2, Cry1, and Bmal1 (Mop3) were validated by qt-PCR (Supplemental Fig. S1). (The online version of this article contains supplemental material.) For NIH/3T3 rhythms in mRNA abundance, the amplitude of differences between peak and minimum levels ranged from 1.5- to 3.6-fold. The clock genes Per2 and Cry1 were among those oscillations with the highest rhythm amplitude in NIH/3T3 cells. The extent of circadian expression in the NIH/3T3 transcriptome is consistent with recent studies of serum-treated rat-1 fibroblasts (9) and in rat 3Y1 embryonic fibroblasts (14) in which the relative percentage (1–2%) and number of rhythmic genes ({approx}41–85) were similarly low. Based on direct comparisons of homologous transcripts, NIH/3T3 cells showed only a limited set of rhythmically expressed genes in common with serum-treated rat-1 fibroblasts (10%) and embryonic fibroblasts (15%) (9, 14). Annexin, Rora, Cdk4, Ubiquitin-like protein 4, histone H2a, Ig{kappa} chain V-region, and RAN GTPase activating protein 1 represent genes with circadian profiles that were common among NIH/3T3 cells and these fibroblast lines (Supplemental Table S1). The overlap in circadian gene expression between NIH/3T3 and rat-1 fibroblasts was especially evident within cell signaling genes associated with the Ras/MAP kinase pathway (9). The limited degree to which the circadian regulation of specific genes was common among NIH/3T3, rat-1, and rat 3Y1 embryonic fibroblasts may reflect differences in experimental analysis across these studies, including the developmental and/or species-specific origin (mouse vs. rat) of these lines, the type of stimulus used to induce rhythmicity (forskolin vs. serum shock), specific criteria for identifying rhythmicity, sampling interval and duration of analysis and the array type (Affymetrix or cDNA). The narrow scope and overlap of fibroblast oscillations in gene expression reported in this and previous studies are similar to that found among peripheral tissues in vivo (10). Collectively, these observations support the view that circadian regulation of the transcriptome is largely cell or tissue specific so as to provide for the local coordination of important processes associated with their distinct functions.


View this table:
[in this window]
[in a new window]

  		Table 1. Comparison of gene expression and function in NIH/3T3 fibroblasts, SCN2.2 cells, and the rat SCN

 

Figure 1
View larger version (22K):
[in this window]
[in a new window]

  		Fig. 1. Venn diagram comparing circadian expression of homologous or functionally related genes in NIH/3T3 cells with that reported in microarray analyses of rat-1 (9) and rat 3Y1 fibroblasts (14). Specific genes with rhythmic profiles in NIH/3T3 cells and rat 3Y1 or rat-1 fibroblasts are listed in Supplemental Table S1. For each fibroblast line the percentage of rhythmic gene expression in proportion to the total number of genes or probe sets examined is denoted in parentheses.

 
Circadian phase and gene tree analyses.
Based on comparisons of first peak of cyclic mRNA abundance in relation to the sampling intervals during the initial cycle (see MATERIALS AND METHODS), rhythmic genes in NIH/3T3 cells were distributed across all phases of the circadian cycle. The circadian patterns and phase distributions of the clock genes Per2 and Bmal1 (Mop3) and several other rhythmic transcripts were corroborated by qt-PCR (Supplemental Fig. S1). In NIH/3T3 fibroblasts, peak mRNA levels were observed at hour 0 and 24 for Per2 and at hour 12 and 36 for Bmal1 (Mop3) so the profiles for these clock genes were respectively categorized as PG I and PG III oscillations. Of the 323 rhythmically expressed unique genes or ESTs in NIH/3T3 fibroblasts, 238 exhibited PG I oscillations in which peak mRNA abundance coincided with the zenith of Per2 expression, whereas 23 exhibited PG III oscillations in which peak levels were contemporaneous with the crest of Bmal1 expression. The remaining groups of 25 and 37 transcripts with circadian profiles were distributed in PG II and IV, respectively.

Rhythmically expressed transcripts were analyzed further with gene tree clustering to determine whether the full temporal profiles were similar among a number of genes within each phase group. Within each phase group, clusters of rhythmic genes with similar expression profiles were assigned to a common branch. Gene lists of transcripts contained in these clusters are available at http://genenet.tamu.edu/servlet/GSWG. An interesting ramification of this analysis is that the specific temporal patterns of expression differ even among core clock components and other rhythmic genes with coincident peaks of mRNA abundance and thus common phase designation.

Functional analyses of nonrhythmic and circadian gene expression in NIH/3T3 cells.
To assess how fundamental cellular processes are differentially expressed in forskolin-stimulated NIH/3T3 fibroblasts, the Gene Ontology lists in GeneSpring 7.3 were used to separately organize noncircadian and clock-controlled genes according to seven broad functional categories: 1) energetics, 2) cell communication, 3) protein dynamics, 4) cell development, 5) defense and detoxification, 6) cytoskeleton and adhesion, and 7) unknown or ESTs. Of the 5,502 detected probe sets with nonrhythmic expression profiles in forskolin-stimulated fibroblasts, {approx}2,100 functionally annotated genes were analyzed in this fashion. Functional categorization of these non-rhythmic genes revealed that 2.6% control energetic processes (n = 55), 1.3% regulate aspects of neurotransmission (n = 27), 20.4% are involved in cellular communication (n = 428), 11.5% mediate the dynamics of gene expression by controlling protein translation, degradation, or trafficking (n = 241), 51.6% regulate aspects of cell development including growth and maintenance (n = 1,082), 2.6%, are associated with defense and detoxification (n = 55), and 9.9% affect cytoskeletal elements, cellular adhesion, or components of the extracellular matrix (n = 208). Approximately 3,400 genes could not be assigned to a specific category due to a lack of sufficient functional annotation (Table 1, bottom).

The 157 unique genes with circadian expression profiles in NIH/3T3 cells were similarly segregated into these broad categories and then subdivided further into specific functional clusters. In NIH/3T3 fibroblasts, 11% of the annotated genes with rhythmic expression profiles (n = 17) were involved in the regulation of energetic processes. these genes with energetic functions were subdivided into four functional clusters: 1) glucose metabolism and mitochondrial energy transduction (n = 2), 2) lipid and fatty acid metabolism (n = 3), 3) transporters of energy metabolites (transporters) (n = 4), and 4) miscellaneous metabolism (n = 8) (Supplemental Table S2). Compared with the SCN where there is ample evidence for the circadian regulation of genes associated with glucose metabolism and mitochondrial energy transduction (26), it is noteworthy that only two genes in this functional cluster were rhythmically expressed in NIH/3T3 fibroblasts. The greatest proportion/number of genes with circadian profiles (35%, n = 55) were involved in other forms of cellular communication. Rhythmic genes in this category were further subdivided into functional clusters: 1) neurotransmission (n = 2), 2) cytosolic signaling factors and transducers (n = 25), 3) nuclear factors (n = 22), 4) G protein-coupled receptors and associated proteins (n = 5), and 5) extracellular factors (n = 1). Regulation of the cAMP pathway was common for rhythmically expressed genes that function as cytosolic signaling factors and transducers (n = 2) or as nuclear factors (n = 5). Twenty-six genes (17%) with rhythmic patterns of expression in NIH/3T3 cells were involved in different aspects of protein dynamics including protein degradation and synthesis (n = 15), protein sorting and trafficking (n = 7), and protein modification and folding (n = 4). Twenty-four rhythmic genes (15%) exhibited common functions in the regulation of cellular development. Rhythmic genes within the cell development category were involved in the regulation of cell cycle (n = 5), maintenance of DNA and chromatin (n = 10), and growth and differentiation (n = 9). The defense and detoxification category encompassed 13% of the rhythmic genes in NIH/3T3 fibroblasts (n = 21). Circadian regulation was observed in 14 annotated genes with cytoskeletal or cellular adhesion functions (9%) (Supplemental Table S2). Differential representation on the mouse U74v2 GeneChip may be an influential factor in the observed prevalence of nonrhythmic genes mediating cell development and of rhythmic genes involved in cell communication because analysis using the Simplified Gene Ontology lists in GeneSpring 7.3 indicates that the probe sets on the array were biased for genes in these functional categories ({approx}21.3 and 11.1%, respectively).

Cross-model comparison of global and circadian properties.
Because our analyses of circadian gene expression in the rat SCN and SCN2.2 cells used the same array technology and applied the same expression and rhythm amplitude criteria as well as the cross-correlation with cosine waves (26), we next compared the NIH/3T3 and SCN transcriptomes with regard to global properties and temporal regulation of gene expression. The breadth of gene expression in NIH/3T3 and SCN2.2 cells was similar (47 and 45%, respectively) and greater than that observed in the rat SCN (33%). In NIH/3T3 fibroblasts, 5,830 out of the 12,500 probe sets (47%) on the mouse U74v2 GeneChip surpassed expression criteria. Relative to the 8,800 probe sets on the rat U34a GeneChip, 3,993 in SCN2.2 cells and 2,929 in the rat SCN fulfilled expression criteria (Table 1, top). Further comparisons revealed that the proportion of genes with stable expression profiles was greater in NIH/3T3 cells (19%, n = 2,391) than in SCN2.2 cells (12%, n = 1,035) or the rat SCN (7%, n = 584), whereas the extent of circadian gene expression was similar across these oscillator models (2.6, 1.8, and 3.4%, respectively; Table 1, top).

We also compared NIH/3T3 fibroblasts, SCN2.2 cells, and the rat SCN with regard to the functional distribution of genes with circadian profiles (Fig. 2). In all of these oscillator models, circadian fluctuations were observed in genes with a wide range of functions although rhythmic gene expression was typically more prevalent within certain functional categories or clusters. NIH/3T3 fibroblasts, SCN2.2 cells, and the rat SCN exhibited some similarities in the functional distribution of rhythmically expressed genes with cytosolic signaling factors, transducers, and nuclear factors in the cellular communication category commonly representing a large proportion of the annotated genes with rhythmic profiles. Corresponding to this functional distribution, genes mediating cellular communication are also prevalent among the cadre of rhythmic transcripts observed in mammalian peripheral tissues such as the liver, heart, and kidney (10, 17, 29, 35, 36). However, there were notable differences in the functional configurations of circadian gene expression in NIH/3T3 fibroblasts relative to those found in SCN2.2 cells and the rat SCN. Compared with the functional distribution of SCN oscillations, circadian rhythmicity in NIH/3T3 cells was differentially evident among genes associated with maintenance of DNA and chromatin (N), defense and detoxification (P), and cytoskeletal elements and adhesion (Q). Relative to NIH/3T3 fibroblasts, circadian profiles in both SCN2.2 cells and the rat SCN were more prevalent among genes involved in glucose metabolism and mitochondrial energy transduction (A), neurotransmission (E), the regulation of G protein-coupled receptors and associated proteins (G), and protein modification and folding (L). The extent of circadian regulation was similar among all three oscillator models only for genes affiliated with cell growth and differentiation (O) (Fig. 2).


Figure 2
View larger version (32K):
[in this window]
[in a new window]

  		Fig. 2. Comparative functional classification of circadian gene expression in NIH/3T3 fibroblasts, suprachiasmatic nucleus (SCN) 2.2 cells, and the rat SCN. In accord with our previous microarray analyses (26), rhythmically expressed transcripts in these oscillator models were comparatively subdivided into 15 functional clusters, Glucose metabolism and mitochondrial energy transduction (A), Lipid and fatty acid metabolism (B), Transporters (C), Miscellaneous metabolism (D), Neurotransmission (E), Extracellular factors (F), G protein-coupled receptors and associated proteins (G), Cytosolic signaling factors and transducers (H), Nuclear factors (I), Protein degradation and synthesis (J), Protein sorting and trafficking (K), Protein modification and folding (L), Cell Cycle (M), Maintenance of DNA and chromatin (N), Growth and differentiation (O), and two functional categories, Defense and Detoxification (P), and Cytoskeletal Elements and Adhesion (Q). Bars depict the percentage of rhythmic transcripts in each functional cluster or category relative to the total number of annotated genes with circadian profiles in NIH/3T3 fibroblasts (n = 157), SCN2.2 cells (n = 116), and the rat SCN (n = 163). Specific functional clusters and categories in which the proportion of rhythmic gene expression was differentially greater in NIH/3T3 fibroblasts or in SCN2.2 cells and the rat SCN are respectively labeled in red and blue.

 
We then assessed NIH/3T3 fibroblasts, SCN2.2 cells, and the rat SCN for overlap as well as fundamental differences in the temporal regulation of 2,255 probe sets with common representation on both the mouse U74v2 and rat U34a GeneChips (Fig. 3). Of these 2,255 probe sets, 413 (18%) fulfilled expression criteria used in this study. Many of these genes (n = 344) were distinguished by stable (n = 60) or noncircadian (n = 284) expression profiles in NIH/3T3 fibroblasts, SCN2.2 cells, and the rat SCN (Table 1, top), and functional annotations were identified for 162 of these nonrhythmic probe sets using our filters. Functional analysis of the commonality in the noncircadian expression of these annotated genes (n = 162) across all three oscillator models revealed that these genes control diverse biological processes but were most prevalent in the cell communication (18.5%, n = 30) and cellular development categories (66.7%, n = 108) (Table 1, bottom). It is interesting that the NIH/3T3 fibroblasts, SCN2.2 cells, and the rat SCN showed similar proportions of noncircadian gene expression in the cellular development and defense and detoxification categories (Table 1, bottom).


Figure 3
View larger version (15K):
[in this window]
[in a new window]

  		Fig. 3. Venn diagram illustrating overlap and differences in the circadian expression of commonly represented genes among NIH/3T3 fibroblasts, SCN2.2 cells, and the rat SCN. Probe sets for homologous or functionally related genes with mutual representation on both the mouse U74v2 and rat U34a GeneChips (n = 2,255) were analyzed and rhythmically expressed genes in NIH/3T3 cells were compared with those identified previously in SCN2.2 cells and the rat SCN (26).

 
Overlap in circadian gene expression across NIH/3T3 fibroblasts, SCN2.2 cells, and the rat SCN was limited to four of the commonly represented genes on the mouse and rat GeneChips (0.2%, Fig. 3): Scd1, long chain fatty acyl CoA synthase (Facl), Tcf4, and Ngfi-A/early growth response 1 (Egr-1) (Supplemental Table S2). These rhythmic genes are involved in fatty acid or lipid metabolism (Scd1 and Facl), or function as nuclear factors in cell communication pathways (Tcf4 and Egr-1). Circadian expression of the core clock genes Per2, Cry1, and Bmal1 was also common among all three models, although this observation is based on qt-PCR analysis because the mouse and rat GeneChips are largely different with regard to the representation of specific clock genes. The limited degree of overlap in circadian gene expression between NIH/3T3 fibroblasts, SCN2.2 cells, and the rat SCN is consistent with reported comparisons of microarray data indicating that only 10% of the rhythmically expressed transcripts were common to the SCN, peripheral tissues, and cultured fibroblasts (10) and that the canonical clock genes constituted the principal portion of this overlap in circadian gene expression. In both this review and the present comparative analyses, it is noteworthy that nuclear factors were identified as a prevalent functional category among the remaining common genes with circadian profiles. This observation suggests that the circadian regulation of nuclear factors may play a critical role in the common oscillatory function of different tissues throughout the body.

To complement this analysis of overlap between experimental models, the temporal expression profiles of the 2,255 commonly represented probe sets were subsequently assessed for differences between NIH/3T3 fibroblasts and SCN cells. For genes and ESTs with noncircadian profiles, 194 (8.6%) were expressed in NIH/3T3 fibroblasts but not in SCN2.2 cells and the rat SCN. Of the annotated genes with noncircadian profiles that were exclusively expressed in NIH/3T3 cells, many were associated with cell communication (n = 44, 23%) or cellular development (n = 15, 8%) (Supplemental Table S2). It is interesting that the differential expression of noncircadian genes in NIH/3T3 cells includes three in the energetics category that function either in the transport of neurotransmitters [5-hydroxytryptamine (serotonin) transporter (Slc6a4) and glycine transporter 1 (Glyt1)] or in the regulation of glucose metabolism [glycerol-3-phosphate dehydrogenase (Gdm1)]. Only 10 noncircadian genes (0.4%) were expressed in SCN2.2 cells and the rat SCN but not in NIH/3T3 fibroblasts. These genes with SCN-specific noncircadian profiles included defender against cell death (Dad-1), prostaglandin F2 receptor negative regulator (Ptgfrn), chemokine (C-X3-C motif) ligand 1 (Cx3cl1), ADP-ribosylation factor 4 (Arf4), farensyl diphosphate synthase (Fdps), isopentenyl-diphosphate delta isomerase (Idi1), par-3 (partitioning defective 3) homolog (Pard3), actin beta (Actb), serum and glucocorticoid-regulated kinase (Sgk), and gephyrin (Gphn).

For rhythmic genes, 26 (1.2%) were marked by circadian expression in NIH/3T3 fibroblasts but not in SCN2.2 cells and the rat SCN (Fig. 3). Annotated genes with fibroblast-specific circadian profiles were primarily involved in the regulation of cellular communication (n = 8, 0.4%) and defense and detoxification pathways (n = 6, 0.3%). On the basis of our previous microarray study (26), 28 annotated genes, 8 of which were identified by supplementary screening analyses, exhibited circadian expression patterns in SCN2.2 cells and the rat SCN but not in NIH/3T3 fibroblasts (Table 2). It is noteworthy that functionally annotated genes with SCN-specific rhythms of expression were predominantly involved in the regulation of cellular communication (n = 16) and energetic processes related to metabolism and transport (n = 7). The distribution of rhythmically expressed genes in the former functional category may have some significance with regard to SCN-specific signals necessary for the endogenous generation of self-sustained rhythmicity, temporal synchrony among individual clock cells, or pacemaker coordination of circadian rhythmicity in downstream oscillators. iNos or Nos2, syntaxin 4 (Stx4a), glycine receptor, alpha 1 subunit (Glra1), cytosolic phospholipase A2 (Pla2g4a), SAP kinase/mitogen-activated protein kinase 12 (Mapk12/Erk-6), serine-threonine kinase receptor/activin a receptor type II-like 1 (Acvrl1), calcitonin/calcitonin-related polypeptide, alpha (Calca), and pregnancy upregulated non-ubiquitously expressed CaM kinase (Pnck) are notable examples of genes in intercellular communication pathways with circadian expression profiles that were unique to SCN2.2 cells and the rat SCN. The identification of Nos2 as a distinctive SCN output signal using this comparative microarray and bioinformatics approach is particularly intriguing in relation to previous functional studies targeting the other isoforms of this enzyme, neuronal Nos (nNos or Nos1) and endothelial Nos (eNos or Nos3). These studies suggest that NO signaling is not essential for SCN circadian function because mutant mice lacking nNos or eNos show normal activity rhythms with regard to their free-running patterns, entrainment to light-dark cycles, and phase-shifting responses to light (19, 20). However, the function of Nos2 in the SCN and the regulation of circadian rhythms has not been examined, presumably based on the implications of these studies and the prevailing bias that this isozyme is expressed by macrophages (24), but not SCN cells.


View this table:
[in this window]
[in a new window]

  		Table 2. Functional categorization of annotated genes with circadian profiles in SCN cells but not NIH/3T3 fibroblasts

 
NO and functional properties of SCN2.2 cells.
To probe the utility of these comparative observations on SCN-specific rhythms of gene expression in differentiating key circadian signals that mediate the distinctive functional properties of the SCN, we targeted the diffusible messenger, NO, for supplemental analysis using our SCN2.2-NIH/3T3 coculture model. The NO signaling pathway provided a focus for this functional analysis because: 1) diffusible signals mediate the pacemaker function of SCN2.2 cells in conferring rhythmicity to cocultured fibroblasts (3) and the restoration of behavioral rhythmicity by SCN transplants in vivo (34); and 2) our gene profiling comparisons and qt-PCR validation indicate that iNos or Nos2, an isozyme involved in NO production, is rhythmically expressed in SCN2.2 cells and the rat SCN (26), but not in NIH/3T3 fibroblasts. Experiments were conducted to determine whether treatment of SCN2.2 cells with L-NAME, reversible inhibitor of all three isoforms of NOS in the CNS (22), alters their endogenous oscillator and/or pacemaker properties. Control cultures of SCN2.2 cells treated with the inactive enantiomer, D-NAME, exhibited circadian rhythms of 2-DG uptake during (0–48 h) and after (48–108 h) the period of treatment (Fig. 4). When cocultured with D-NAME-treated SCN2.2 cells, untreated NIH/3T3 fibroblasts were also characterized by the circadian regulation of 2-DG uptake, with a 4-h phase delay in their rhythmicity relative to that observed in SCN2.2 cells (Fig. 4). This phase relationship between SCN2.2 and NIH/3T3 rhythms in 2-DG uptake is similar to that observed in previous studies (3). For both cell types in cocultures containing D-NAME-treated SCN2.2 cells and untreated NIH/3T3 fibroblasts, peak levels of 2-DG uptake were approximately twofold greater than the corresponding minimum and the peak-to-peak interval, which is indicative of the period for the rhythm in glucose utilization, consistently 24 h. Rhythmic 2-DG uptake reached peak values at 0, 24, 48, 72, and 96 h in D-NAME-treated SCN2.2 cells and at 52, 76, and 100 h in untreated NIH/3T3 fibroblasts. During treatment with the NOS inhibitor (0–48 h) L-NAME (0–48 h), 2-DG uptake in SCN2.2 cells showed no evidence of circadian rhythmicity (Fig. 4) and remained at intermediate levels with respect to the circadian maxima and minima observed in control cultures of SCN2.2 cells. Following L-NAME treatment (48–108 h), the temporal profiles of 2-DG uptake in both SCN2.2 cells and cocultured NIH/3T3 fibroblasts were marked by circadian rhythmicity. However, posttreatment rhythms in 2-DG uptake were altered in both coculture compartments. Peak levels of 2-DG uptake were observed at 64 and 96 h posttreatment in SCN2.2 cells and at 68 and 100 h in cocultures of untreated NIH/3T3 cells. Although the 4-h phase difference between SCN2.2 and NIH/3T3 metabolic rhythms was maintained during all posttreatment cycles, it is noteworthy that the peak-to-peak interval for rhythmic 2-DG uptake in both cell types was increased to 32 h.


Figure 4
View larger version (35K):
[in this window]
[in a new window]

  		Fig. 4. Temporal profiles of 2-deoxyglucose (2-DG) uptake in cocultures containing L-NAME- or D-NAME-treated (100 µM) SCN2.2 cells (top) and untreated NIH/3T3 fibroblasts (bottom). SCN2.2 cells were treated for 48 h with the NOS inhibitor L-NAME (dashed line, {triangleup}) and then cocultured for 60 h with untreated NIH/3T3 cells (dashed line, {circ}). Control cultures of SCN2.2 cells were treated with its inactive enantiomer D-NAME (solid line, {blacktriangleup}) and then cocultured with untreated NIH/3T3 fibroblasts (solid line, •). The shaded region in the top panel demarcates the interval of SCN2.2 treatment with L-NAME or D-NAME (hour 0–48). Symbols denote determinations of 2-DG uptake (n = 3, means ± SE) at 4-h intervals. Values for 2-DG uptake are plotted as a function of time such that time 48 denotes when treated SCN2.2 cells were first cocultured with untreated NIH/3T3 fibroblasts. *Sampling intervals during which peak values for 2-DG uptake were significantly greater (P < 0.05) than those observed during preceding or succeeding minima.

 
One possible explanation for the observed loss of rhythmic 2-DG uptake in SCN2.2 cultures during the treatment interval is that L-NAME-mediated inhibition of NOS and NO signaling may disrupt the clock mechanism by altering the biological effects of NO through the intracellular second messenger cGMP and/or on posttranslational modification of cellular proteins (15, 21). It is also possible that the loss of metabolic rhythmicity in SCN2.2 cultures may reflect L-NAME inhibition of NO function as a diffusible messenger in the intercellular coupling or synchronization of clock cell populations. Decreased coupling strength among multiple oscillators is thought to produce changes in rhythm amplitude or period length from the population mean and thus result in irregular ensemble rhythmicity (12). Thus, the disruption of rhythmic 2-DG uptake may be derived from the effects of L-NAME on NO function in the coordination of metabolic oscillations among populations of SCN2.2 clock cells. However, the implications of these results are subject to the caveat that global inhibition of NOS activity with L-NAME does not allow distinctions between Nos2 and other isoforms, neuronal Nos (nNos or Nos1) and endothelial Nos (eNos or Nos3), with regard to their relative importance in SCN circadian function. In addition, the effects of L-NAME-induced NOS inhibition on SCN2.2 pacemaker function were not studied during the period of treatment because the coculture model is not well suited for differentiating the effects of pharmacological agents on a specific cell type due to the ready diffusion of these agents across coculture compartments. Consequently, further analysis using siRNA or antisense approaches will be necessary to directly assess Nos2 function in the endogenous rhythm-generating and circadian pacemaker properties of SCN2.2 cells. Nonetheless, the identification of Nos2 as an SCN-specific circadian signal in our comparative microarray analyses and the L-NAME-induced rhythm disturbances in SCN2.2-NIH/3T3 cocultures suggest that NO may play an important role in the distinctive oscillatory and pacemaker functions of SCN2.2 cells.

In summary, previous studies have suggested comparisons of array data provide negligible information beyond the indication that circadian gene expression is highly disparate among different tissues or cell types. Our comparative analyses of gene expression profiles in NIH/3T3 fibroblasts, SCN2.2 cells, and the rat SCN are certainly consistent with this general observation but also underscore some implications of the limited overlap and substantial differences in rhythmic transcripts across these oscillator models. In conjunction with the canonical clock genes, Per2, Cry1 and Bmal1, genes that mediate fatty acid or lipid metabolism (Scd1 and Facl) or function as nuclear factors in cell communication pathways were the only rhythmically expressed transcripts in common among NIH/3T3 fibroblasts, SCN2.2 cells, and the rat SCN. Further analysis is warranted to determine whether genes in these functional clusters are responsible for the common oscillatory properties of the SCN and other tissues or cells throughout the body. The present identification of SCN-specific differences provides further insight into why the SCN in vivo and SCN2.2 cells, but not fibroblasts, are capable of generating self-sustained rhythmicity and functioning as circadian pacemakers. Nos2 and other rhythmic genes involved in the regulation of intercellular communication were differentially identified as critical circadian signals expressed only in SCN cells. On the basis of our analysis of NOS function in this study, these genes will provide a suitable focus for identifying signaling pathways responsible for the coupling of SCN oscillators or the coordination of rhythmicity in other cells and tissues.


   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 PO1 NS-39546 (D. J. Earnest, V. M. Cassone, and T. L. Thomas).


   ACKNOWLEDGMENTS
 
The authors thank Kim Lu for technical assistance and Phil Beremand 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 Neuroscience and Experimental Therapeutics, 238 Reynolds Medical Bldg., College Station, TX 77843-1114 (e-mail: dearnest{at}tamu.edu).


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

  1. Akashi M, 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, 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, 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, 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, Lipman DJ. Basic local alignment search tool. J Mol Biol 215: 403–410, 1990.[CrossRef][ISI][Medline]
  6. Balsalobre A, Damiola F, Schibler U. A serum shock induces circadian gene expression in mammalian tissue culture cells. Cell 93: 929–937, 1998.[CrossRef][ISI][Medline]
  7. Correa A, Lewis ZA, Greene AV, March IJ, Gomer RH, Bell-Pedersen D. Multiple oscillators regulate circadian gene expression in Neurospora. Proc Natl Acad Sci USA 100: 13597–13602, 2003.[Abstract/Free Full Text]
  8. Ding JM, Chen D, Weber ET, Faiman LE, Rea MA, Gillette MU. Resetting the biological clock: mediation of nocturnal circadian shifts by glutamate and NO. Science 266: 1713–1717, 1994.[Abstract/Free Full Text]
  9. Duffield GE, Best JD, Meurers BH, Bittner A, Loros JJ, 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]
  10. Duffield GE. DNA microarray analyses of circadian timing: the genomic basis of biological time. J Neuroendocrinol 15: 991–1002, 2003.[CrossRef][ISI][Medline]
  11. Earnest DJ, Liang FQ, Ratcliff M, Cassone VM. Immortal time: circadian clock properties of rat suprachiasmatic cell lines. Science 283: 693–695, 1999b.[Abstract/Free Full Text]
  12. Enright JT. Temporal precision in circadian systems: a reliable neuronal clock from unreliable components? Science 209: 1542–1545, 1980.[Abstract/Free Full Text]
  13. Gillette MU, 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]
  14. Grundschober C, Delaunay F, Puhlhofer A, Triqueneaux G, Laudet V, Bartfai T, Nef P. Circadian regulation of diverse gene products revealed by mRNA expression profiling of synchronized fibroblasts. J Biol Chem 276: 46751–4678, 2001.[Abstract/Free Full Text]
  15. Hanafy KA, Krumenacker JS, Murad F. NO, nitrotyrosine, and cyclic GMP in signal transduction. Med Sci Monit 7: 801–819, 2001.[Medline]
  16. Heid CA, Stevens J, Livak KJ, Williams PM. Real time quantitative PCR. Genome Res 6: 986–994, 1996.[Abstract/Free Full Text]
  17. Kita Y, Shiozawa M, Jin W, Majewski RR, Besharse JC, Greene AS, Jacob HJ. Implications of circadian gene expression in kidney, liver and the effects of fasting on pharmacogenomic studies. Pharmacogenetics 12: 55–65, 2002.[CrossRef][ISI][Medline]
  18. Klein DC, Moore RY, Reppert SM. Suprachiasmatic Nucleus: The Mind's Clock. New York: Oxford Univ. Press, 1991.
  19. Kriegsfeld LJ, Demas GE, Lee SE, Dawson TM, Dawson VL, 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]
  20. Kriegsfeld LJ, Drazen DL, Nelson RJ. Circadian organization in male mice lacking the gene for endothelial nitric oxide synthase (eNOS-/-). J Biol Rhythms 16: 142–148, 2001.[Abstract]
  21. Krumenacker JS, Murad F. NO-cGMP signaling in development and stem cells. Mol Genet Metab 87: 311–314, 2006.[CrossRef][ISI][Medline]
  22. Lambert LE, Whitten JP, Baron BM, Cheng HC, Doherty NS, McDonald IA. Nitric oxide synthesis in the CNS endothelium and macrophages differs in its sensitivity to inhibition by arginine analogues. Life Sci 48: 69–75, 1991.[CrossRef][ISI][Medline]
  23. Lee CK, Klopp RG, Weidruch R, Prolla TA. Gene expression profile of aging and its retardation by caloric restriction. Science 285: 1390–1393, 1991.
  24. Lowenstein CJ, Snyder SH. Nitric oxide, a novel biologic messenger. Cell 70: 705–707, 1992.[CrossRef][ISI][Medline]
  25. McDonald MJ, Rosbash M. Microarray analysis and organization of circadian gene expression in Drosophila. Cell 107: 567–578, 2001.[CrossRef][ISI][Medline]
  26. Menger GJ, Kim L, Thomas T, Cassone VM, Earnest DJ. Circadian profiling of the transcriptome in immortalized rat SCN cells. Physiol Genomics 21: 370–381, 2005.[Abstract/Free Full Text]
  27. National Library of Medicine, and National Institutes of Health. National Center for Biotechnology Information (NCBI). http.//www.ncbi.nlm.nih.gov/, 2004–2005.
  28. Newman GC, Hospod FE, Patlak CS. Kinetic model of 2-deoxyglucose metabolism using brain slices. J Cereb Blood Flow Metab 10: 510–526, 1990.[ISI][Medline]
  29. Panda S, Antoch MP, Miller BH, Su AI, Schook AB, Straume M, Schultz G, Kay SA, Takahashi JS, Hogenesch JB. Coordinated transcription of key pathways in the mouse by the circadian clock. Cell 109: 307–320, 2002.[CrossRef][ISI][Medline]
  30. Pruitt KD, Maglott DR. RefSeq and LocusLink: NCBI gene-centered resources. Nucleic Acids Res 29: 137–140, 2001.[Abstract/Free Full Text]
  31. Schaffer R, Landgraf J, Accerbi M, Simon V, Larson M, Wisman E. Microarray analysis of diurnal and circadian-regulated genes in Arabidopsis. Plant Cell 13: 113–123, 2001.[Abstract/Free Full Text]
  32. Schwartz WJ. SCN metabolic activity in vitro. In: Suprachiasmatic Nucleus: The Mind's Clock, edited by Klein DC, Moore RY, Reppert SM. New York: Oxford Univ. Press, 1991, p. 144–156.
  33. Shearman LP, Zylka MJ, Weaver DR, Kolakowski LF Jr, Reppert SM. Two period homologs: circadian expression and photic regulation in the suprachiasmatic nuclei. Neuron 19: 1261–1269, 1997.[CrossRef][ISI][Medline]
  34. Silver R, LeSauter J, Tresco PA, Lehman MN. A diffusible coupling signal from the transplanted suprachiasmatic nucleus controlling circadian locomotor rhythms. Nature 382: 810–813, 1996.[CrossRef][Medline]
  35. Storch KF, Lipan O, Leykin I, Viswanathan N, Davis FC, Wong WH, Weitz CJ. Extensive and divergent circadian gene expression in liver and heart. Nature 417: 78–83, 2002.[CrossRef][Medline]
  36. Ueda HR, Chen W, Adachi A, Wakamatsu H, Hayashi S, Takasugi T, Nagano M, Nakahama K, Suzuki Y, Sugano S, Iino M, Shigeyoshi Y, Hashimoto SA. A transcription factor response element for gene expression during circadian night. Nature 418: 534–539, 2002.[CrossRef][Medline]
  37. Watanabe A, Hamada T, Shibata S, 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]
  38. Yamazaki S, Numano R, Abe M, Hida A, Takahashi R, Ueda M, Block GD, Sakaki Y, Menaker M, Tei H. Resetting central and peripheral circadian oscillators in transgenic rats. Science 288: 682–685, 2000.[Abstract/Free Full Text]
  39. Zylka MJ, Shearman LP, Weaver DR, Reppert SM. Three period homologs in mammals: differential light responses in the suprachiasmatic circadian clock and oscillating transcripts outside of brain. Neuron 20: 1103–1110, 1998.[CrossRef][ISI][Medline]



This article has been cited by other articles:


Home page
 Mol. Pharmacol.Home page
X. Qu, R. P. Metz, W. W. Porter, V. M. Cassone, and D. J. Earnest
Disruption of Clock Gene Expression Alters Responses of the Aryl Hydrocarbon Receptor Signaling Pathway in the Mouse Mammary Gland
Mol. Pharmacol., November 1, 2007; 72(5): 1349 - 1358.
[Abstract] [Full Text] [PDF]

This Article
Free upon publication Free Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Figure and Tables
Right arrow All Versions of this Article:
29/3/280    most recent
00199.2006v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (3)