291
Hepatic gene expression patterns in thyroid hormone-treated hypothyroid rats J M Weitzel, S Hamann1, M Jauk, M Lacey1, A Filbry, C Radtke, K A H Iwen, S Kutz, A Harneit, P M Lizardi 1 and H J Seitz Institute of Medical Biochemistry and Molecular Biology, University Hospital Hamburg-Eppendorf, 20246 Hamburg, Germany 1
Department of Pathology, Yale University School of Medicine, New Haven, Connecticut 06520, USA
(Requests for offprints should be addressed to J M Weitzel; Email:
[email protected]) (J M Weitzel and S Hamann contributed equally to this work)
Abstract Thyroid hormone (T3) is essential for normal development, differentiation and metabolic balance. We have performed DNA microarray experiments using hepatic RNA from hypothyroid and T3-treated hypothyroid rats in order to characterize T3-induced gene expression patterns after various time points (6, 24 and 48 h after the administration of the hormone). Sixty-two of 4608 different genes displayed a reproducible T3-response, and cluster analysis divided these differentially regulated genes into six expression patterns. Thirty-six genes were not significantly regulated within the first 24 h. Transient transfection experiments of eight late-induced gene promoters failed to detect a thyroid hormone response element within their regulatory elements, suggesting an indirect activation mechanism(s). In search for an intermediate factor of T3 action, we examined whether various rather ubiquitous transcription factors, peroxisome proliferator-activated receptors (PPARs) and coactivators of the PPAR coactivator 1 family (PGC-1) are regulated by T3. Only PPAR and PERC/PGC-1 exhibit a significant T3-response within the first 6 h after treatment, identifying these factors as candidate components for mediating the late-induced expression pattern. Regulation of early-induced genes within the first 6 h after administration of T3 on transcript levels correlates with altered protein levels after 24 and 48 h in vivo. Journal of Molecular Endocrinology (2003) 31, 291–303
Introduction Thyroid hormone (T3) has a profound influence on normal development, differentiation, and metabolism (Yen 2001). After a single injection of T3 to hypothyroid rats the metabolic rate remains unaffected for the first 20–30 h. After that period, the metabolic rate increases and reaches maximum levels between 72 and 96 h (Tata et al. 1963). This delayed response is surprising since nongenomic effects of T3 are already detectable within minutes (Davis et al. 2002) and genomic effects of T3 are typically detectable within a few hours (Weitzel et al. 2003). For many of the latter genes functional thyroid hormone response elements (TREs) have been reported (Zilz et al. 1990,Weitzel et al. 2001a), indicating a direct mechanism via binding of Journal of Molecular Endocrinology (2003) 31, 291–303 0952–5041/03/031–291 © 2003 Society for Endocrinology
thyroid hormone receptors (TRs), which are ligand-regulated transcription factors. TRs modulate gene activity through alternately silencing or activating transcription by recruitment of either corepressor or coactivator complexes depending on the absence or presence of thyroid hormone (Robyr et al. 2000, Zhang & Lazar 2000). However, several T3-target genes are expressed ‘late’ (24–48 h) after the administration of T3 in vivo, including cytochrome c (Scarpulla et al. 1986) and adenine nucleotide translocator 2 (ANT2) (Dümmler et al. 1996) and, noteworthy, an interaction of TR (via a TRE) with these target genes has not been observed (Li et al. 1997, Weitzel et al. 2003). These observations suggest an induction mechanism via the activation of an intermediate factor(s). Candidate proteins have Online version via http://www.endocrinology.org
Printed in Great Britain
292
J M WEITZEL, S HAMANN
and others ·
Expression patterns in T3-treated rats
been identified which probably serve as intermediate factors, e.g. the nuclear respiratory factors 1 and 2 (NRF-1 and NRF-2, Scarpulla 2002) or the nuclear receptor coactivator PGC-1, which has been shown to interact with NRF-1 (Wu et al. 1999, Knutti & Kralli 2001, Puigserver & Spiegelman 2003). Indeed, binding sites for NRF-1 and/or NRF-2 were detected in several target genes which are known to be regulated by thyroid hormone more than 12 h after the administration of T3 in vivo (Scarpulla et al. 1986, Luciakova & Nelson 1992, Scarpulla 2002, Weitzel et al. 2003). This finding strengthens the argument for an indirect T3-activating mechanism via an intermediate factor(s) since these factor(s) need to be synthesized before they can act. Recently, we detected novel T3-target genes and identified two T3-mediated gene expression patterns after the administration of T3 to hypothyroid rats using cDNA expression arrays. DNA microarrays are powerful tools to analyse differential gene expression and were successfully used to identify novel T3-target genes in mouse, rat, human, cell line and tumour (Feng et al. 2000, Miller et al. 2001, Clement et al. 2002, Flores-Morales et al. 2002, Weitzel et al. 2001b, Wood et al. 2002). However, a large-scale analysis of T3-mediated temporal gene expression patterns has not been performed. Our previous study was hampered by the limited size of the arrays which contained 588 different genes. To receive a more representative set of T3-mediated gene expression patterns we now extended this study to a larger DNA microarray comprising 4608 different probes. In search of an intermediate factor for mediating the late T3-mediated gene expression pattern we extended our study to: (i) transcription factors which are known to be important for the activity of several late-induced T3-target genes, (ii) to peroxisome proliferator-activated receptors (PPARs), since many newly identified T3-target genes contain a functionally characterized PPAR response element, and (iii) to the coactivators PRC and PERC/PGC-1.
Materials and methods Animals and treatments
Adult male Wistar rats (Charles River Wiga, Charles River Deutschland GmbH, Sulzfeld, Journal of Molecular Endocrinology (2003) 31, 291–303
Germany) were housed under controlled conditions (22 C, constant humidity, 12 h:12 h dark–light cycle); food and water were provided ad libitum. Rats were maintained in the animal care facility of the University Hospital Hamburg-Eppendorf according to the UKE animal care guidelines. Hypothyroidism was induced by i.p. injection of Na131I (Amersham Pharmacia) (250 µCi/100 g body weight) 28 days before the experiments. Hyperthyroidism was provoked by an intraperitoneal injection of T3 (50 µg/100 g body weight; Henning) which was repeated after 24 h. Thyroid hormone status was verified by measurements of serum T3, T4 and TSH levels: (hypo: T3,0·4 µg/l; hyper: T3.6·5 µg/l). Zero, 6, 24 and 48 h after thyroid hormone injection, rats were killed by decapitation and the livers were isolated for RNA and protein preparation. RNA isolation and cDNA labelling
Total RNA was prepared from rat livers using RNA-Clean (Hybaid AGS, Heidelberg, Germany) following the manufacturer’s instruction and treated with DNase I (Clontech) (50 U, 30 min, 37 C). One hundred microgram of total RNA (combined from identical amounts of RNA from three different animals per time point to minimize the effect of individual variation of the expression pattern) were labelled by direct incorporation of either Cy3-dUTP or Cy5-dUTP in a reverse transcriptase reaction (Eisen & Brown 1999). Microarray experiments
Probes were hybridized to Mouse 4·6k poly-lysine cDNA microarrays (Yale Keck Microarray Facility) according to the Childs/Kucherlapati protocol (http://sequence.aecom.yu.edu/bioinf/ microarray/protocol4.html). DNA microarrays containing 4608 cDNA fragments (obtained from the IMAGE consortium) were utilized which were not enriched in ‘classical’ T3-target genes and were printed in duplicates. We made an effort to set up a high standard, both experimentally and in the data analysis, in order to maximize the data quality. For example, we tried to avoid the effect of putative artificial signals caused by non-specific dye–cDNA interactions by conducting each microarray hybridization twice, where the second experiment was performed with a reversed Cye-dye www.endocrinology.org
Expression patterns in T3-treated rats ·
assignment (the so called ‘dye flip’). Moreover, we combined the 6 h with the 24 h and the 24 h with the 48 h samples in a way that made it possible to cross-reference the data. Only signals that showed at least a ratio of 2 or 0·5 in all four independent microarray experiments (resulting in a total of eight data points) were considered up- or down-regulated respectively. Because of the reported problems with IMAGE consortium mouse cDNA clones (Halgren et al. 2001), gene identities were first confirmed by sequence analysis. After overnight incubation slides were scanned on a ScanArray 5000 microarray scanner (GSI Lumonics, Farmington Hills, MI, USA) and the generated 16-TIFF images were analyzed by QuantArray microarray analysis software (GSI Lumonics). Cluster analysis and visualization was performed using the Eisen Cluster and Treeview software respectively (Eisen et al. 1998). Northern blotting
For northern blots 10 µg total RNA, combined from identical amounts of three animals per time point, were analyzed and normalized to: (i) the ethidium bromide staining signal and (ii) to the ubiquitin hybridization signal as described previously (Weitzel et al. 2001b). Fold increases were determined from T3-treated hypothyroid rat signal values (6, 24 and 48 h) divided by hypothyroid rat signal values within the same experiment. For hybridization, cDNA probes of particular genes (200–300 bp in length) were generated by RT-PCR amplification, using primers designed to the sequences described in Table 1. The portion of the cDNA which was used as probe for northern hybridization was (whenever possible) not identical to the portion of the cDNA that was used for the production of the microarray. Thus, we generated hybridization probes that served as a sequenceindependent verification of the microarray hybridization data. The amplification products were gel-purified, ligated into pGEM T-easy (Promega) and verified by sequencing.
J M WEITZEL, S HAMANN
and others 293
printing of the arrays. PCR reactions were carried out in duplicates or triplicates with the following PCR primers: CREB for: 5 -ccatcaccactgtaacagtgc3 ; CREB rev: 5 -tgtccatcagtggtctgtgc-3 ; HCF for: 5 -acagaatggtaatgtggtccg-3 ; HCF rev: 5 -agtgatccg cactacagtgg-3 ; HNF-4 for: 5 -caggctcaagaagtgctt cc-3 ; HNF-4 rev: 5 -acacacatccgtgatgttgg-3 ; mtTFA for: 5 -tgaagttgggcgaagtgatc-3 ; mtTFA rev: 5 -atgcattcagtgggcagaag-3 ; Sp1 for: 5 -ccaatgg acaggtcagttgg-3 ; Sp1 rev: 5 -atggaggacagttgagca gc-3 ; YY1 for: 5 -ggttgaagagcagatcattgg-3 ; YY1 rev: 5 -ttcttggcttcattctggc-3 ; PPAR for: 5 -tacatg acatggagaccttg-3 ; PPAR rev: 5 -catagcgaagtcaaac ttgg-3 ; PPAR for: 5 -cagcgtgcatgtgttctacc-3 ; PPAR rev: 5 -tactgcgcaagaactcatgg-3 ; PPAR for: 5 -tattctcagtggagaccgcc-3 ; PPAR rev: 5 -ccg tcttcttgatcacatgc-3 ; PRC for: 5 -tggagagattgaggag tgtacc-3 ; PRC rev: 5 -actcctcttgcagaactgcc-3 ; PERC/PGC-1 for: 5 -ggtattcctcgaaggttaaggc-3 ; PERC/PGC-1 rev: 5 -ccagatacaccgactatgatcc3 . Crossing points of target gene amplification products were normalized to the crossing points of ubiquitin amplification products (Weitzel et al. 2001b). Fold increase of expression levels were determined from T3-treated hypothyroid rat signal values divided by hypothyroid rat signal values within the same experiment of at least three independent experiments. Transient transfection
Transient transfection experiments in human hepatocarcinoma HepG2 cells were conducted as described previously (Weitzel et al. 2001a). Eight promoter fragments of six late T3-induced genes (as indicated in Table 2) were ligated either into pGL2 or pGL3 vectors (Promega) upstream of the luciferase gene. Cytochrome c, COX subunit IV and mtTFA promoter fragments are kind gifts of R C Scarpulla (Northwestern University, Chicago, USA) and R J Wiesner (University of Köln, Germany) and SCP2 and SCPx promoter fragments are kind gifts of J F Strauss III (University of Pennsylvania, USA).
Quantitative PCR
Western blotting
Quantitative real-time RT-PCR was performed as described previously (Weitzel et al. 2001b). Again, PCR primers were (whenever possible) selected from a portion of the cDNA which was not used for
For western blots 50 mg total liver tissue of three different animals per time point (0, 24 or 48 h after injection with 50 µg T3/100 g body weight) were dissolved in SDS-PAGE gel running buffer,
www.endocrinology.org
Journal of Molecular Endocrinology (2003) 31, 291–303
294
J M WEITZEL, S HAMANN
and others ·
Expression patterns in T3-treated rats
Table 1 T3-regulated genes in livers of hypothyroid rats
No. Gene name Cluster 1
1 2 3 4 5 6 7 8 9
2
10 11 12 13 14 15 16 17 18 19 20 21 22
3
0·5
0·2
0·2 S, N
AI430962
metabolism
0·6
0·6
0·4
AI327212
detoxification
0·5
0·5
0·4 S, N
EST (hypothetical protein MMT-7 gene) Ceruloplasmin EST (similar to elongation of very long chain fatty acids-like 2) Ribosomal protein L30 CCAAT-binding transcription factor (Nfix) EST (similar to brain protein 44-like (Brp44l)) Ras-related YPT1 protein
AI414386 AI323769 AI448451
unclassified detoxification metabolism
1·4 1·3 1·5
1·7 1·6 2·4
3·1 2·6 4·7
AI324671 AI528802 AI414240
translation transcription unclassified
1·5 1·7 1·5
1·5 2·9 3·0
2·1 5·3 4·8
AI528700
1·4
2·4
3·0
EST (similar to N-myc downstream regulated 2 (Ndr2)) Multifunctional aminoacyl-tRNA synthetase Alpha-amylase 2B precursor EST (no high similarity to any known gene product) Groucho protein GRG1-L (Tle 1) Apolipoprotein AIV EST (similar to calmodulin binding protein 3 (Strn 3)) EST (similar to karyogamy protein (KAR4)) B-cell Lymphoma 3-encoded protein (BCL3)
AI413204
signal transduction signal transduction translation metabolism unclassified
2·1
3·4
5·1
1·7 1·8 1·8
2·7 2·8 3·9
3·8 4·0 S, N 4·3
1·7 2·2 2·6
5·4 5·4 11·7 14·5 S, N 10·5 7·0
AI429285 AI528691
transcription metabolism signal transduction cell division transcription
2·5 3·5
11·6 5·0
AI385637 AI325068 AI325520
transcription metabolism metabolism
1·0 0·9 1·1
1·6 2·7 1·9
2·0 4·8 S, N 2·6 S, N
AI327085 AI451616
protein folding 1·2 protein targeting 1·3
2·3 4·3
2·5 4·9
AI415331
unclassified
1·2
4·5
6·2
AI430836
cell structure
1·1
2·6
2·5
AI449611
unclassified
1·1
2·5
2·6
AI414038 AI414607 AI415049
metabolism metabolism unclassified
1·2 1·1 1·2
3·4 5·5 4·4
3·5 3·4 3·6
AI325917
detoxification
1·4
6·0
4·2
AI323562 AI324651
metabolism signal transduction cell structure
0·9 0·9
2·6 2·9
2·9 S, N 2·6
0·9
2·1
1·7
Myelin basic protein
27 28 29 30 31
24 h 48 h Verification
detoxification
34
26
6h
AI427530
32 33
25
Fold increase
EST (similar to cystathionine gamma-lyase (Cth)) EST (similar to alcohol dehydrogenase PAN2) Glutathione S-transferase P
EST (similar to frizzled homolog 8 (FZD8)) Adenine nucleotide translocator 2 Mitochondrial electron-transfer-flavoprotein, alpha polypeptide Chaperonin 10 EST (similar to signal recognition particle 54 kDa) EST (no high similarity to any known gene product) EST (similar to myofibrillogenesis regulator MR-2) EST (no high similarity to any known gene product) EST (similar to apolipoprotein AV) EST (similar to steroid 5 alpha-reductase) EST (similar to transport-secretion protein 2.2 (TTS-2.2)) Oxidative stress-inducible protein tyrosine phosphatase (PTPase) 3,2-Trans-enoyl-CoA isomerase c-Src kinase
23 24
Accession no. Function
AI450063 AI325237 AI451318 AI324893 AI326922 AI426660
AI323506
4·7 3·2
(Continued)
Journal of Molecular Endocrinology (2003) 31, 291–303
www.endocrinology.org
Expression patterns in T3-treated rats ·
J M WEITZEL, S HAMANN
and others 295
Table 1 Continued
No. Gene name
Accession no. Function
35
Epidermal growth factor receptor (Egfr)
AI427644
36 37
EST (similar to isocitrate dehydrogenase 1) ATP synthase beta-chain
38
EST (no high similarity to any known gene product) EST (similar to zinc-finger protein (bcl-6)) EST (similar to (predicted) yeast and worm proteins homolog) Aldolase 1, A isoform Farnesyl pyrophosphate synthase (Fpps) EST (similar to neighbor of A-kinase anchoring protein 95) EST (similar to alpha/beta-hydrolases structure containing protein) EST (similar to lipase) EST (no high similarity to any known gene product) EST (no high similarity to any known gene product) EST (similar to MHC region containing the Q region of class I) EST (similar to cytochrome P450) EST (similar to hydroxyacid oxidase 3 (medium-chain)) Guanosine diphosphate (GDP) dissociation inhibitor 3 EST (similar to ubiquitin-conjugating enzyme E2I) EST (similar to neuron specific gene family member 1) EST (similar to ubiquitin specific protease 24) EST (similar to pantothenate kinase 1 (Pank1)) EST (similar to bromodomain-containing protein 3) EST (similar to transcription factor BEN-beta (Gtf2ird1)) Heat-responsive protein 12
Fold increase 6h
24 h 48 h Verification
Cluster 4
39 40 41 42 43 44 45 46 47 5
48 49 50 51 52 53 54 55 56 57 58
6
59 60 61 62
Alpha-tubulin isotype M HMG-CoA synthase EST (similar to CBP/p300-interacting transactivator) NADH-ubiquinone oxidoreductase
0·8
2·1
AI327112 AI327309
signal 0·8 transduction metabolism 0·5 metabolism 0·7
0·8 0·5
2·8 3·6
AI449762
unclassified
0·8
1·8
2·6
AI528676 AI325979
transcription 0·7 unclassified 0·9
1·6 1·7
2·9 3·0
AI323970 AI450071 AI448274
0·9 1·0 1·0
1·7 1·6 1·6
2·7 2·6 3·5
AI449506
metabolism metabolism signal transduction metabolism
0·9
1·5
3·0
AI448275 AI428422
metabolism unclassified
0·9 0·8
1·5 1·0
2·8 2·8
AI428535
unclassified
0·8
1·0
4·1
AI414814
immunology
0·6
1·0
4·9
AI426508 AI325478
metabolism metabolism
0·7 0·7
1·2 1·1
3·4 2·6
AI528719
metabolism
0·8
1·1
2·4
AI666653
0·8
1·1
2·7
0·8
1·3
4·9
0·9
1·2
2·2
AI644488
protein degradation signal transduction protein degradation metabolism
0·7
1·8
8·2
AI428536
transcription 0·9
1·3
2·7
AI428491
transcription 0·9
1·2
2·8
AI893634
signal 1·0 transduction
1·0
3·0
S, N
AI325223 AI324952 AI449991
cell structure 0·5 metabolism 0·4 transcription 0·8
1·3 1·1 1·7
3·1 2·3 1·8
S, P
AI451384
metabolism
2·0
2·4
AI428539 AI414051
0·5
S, N
S, P
The expression ratios represent the mean of four independent experiments, with each gene spotted in duplicates. Gene names, GenBank accession numbers, gene functions and fold increase data are given. Fold increase data were calculated by comparing the 6, 24 and 48 h T3-treated signal intensities to the hypothyroid signal intensities (0 h) (see Result section). Selected genes were verified by sequencing (S), northern hybridisation (N) or quantitative real-time PCR (P).
www.endocrinology.org
Journal of Molecular Endocrinology (2003) 31, 291–303
296
J M WEITZEL, S HAMANN
and others ·
Expression patterns in T3-treated rats
Table 2 Lack of TREs in late T3-induced gene promoters
Gene ANT2 Cytochrome c COX subunit IV L-FABP L-FABP mtTFA SCP2 SCPx
Investigated promoter regions
TRE
rat(−1440/+132) rat(−632/+135) rat(−669/+32) rat(−615/+155) hu(−1623/+132) hu(−424/+42) hu(−809/+73) hu(−3427/+49)
no no no no no no no no
TRE, thyroid hormone response element; ANT2, adenine nucleotide translocator 2; COX, cytochrome c oxidase; L-FABP, liver fatty acid binding protein; SCP2, sterol carrier protein, alternate promoter 2; SCPx, SCP, alternate promoter x.
boiled for 10 min at 95 C and adjusted to an SDS-PAGE. Separated hepatic proteins were transferred to nitrocellulose membranes by electroblotting. Membranes were incubated with a 1:1000 dilution of antibodies to Sp1 (Santa Cruz Biotechnology, Santa Cruz, CA, USA), L-FABP (HyCult, Uden, The Netherlands), mGPDH (Weitzel et al. 2001a) or PGC-1 (Chemicon, Temecula, CA, USA) in 154 mM NaCl, 10 mM NaH2PO4, pH 7·5, 2·5% nonfat dry milk. Antibody binding was detected by carrying out secondary antibody incubations using peroxidaseconjugated anti first IgG antibodies (Dianova, Hamburg, Germany) diluted 1:5000. Secondary antibody was detected using the ECL system according to the manufacturer’s recommendation (Amersham Pharmacia).
livers of three animals per time point (0, 6, 24 and 48 h after administration of T3 in hypothyroid rats) was prepared and used for the labelling reaction and microarray hybridization. Microarray results of authenticated cDNAs were subsequently verified by northern blot, quantitative PCR and western blot experiments. Sixty-two of 4608 genes were shown to be reproducibly T3-responsive. Cluster analysis grouped these 62 differentially regulated genes into six well-defined expression profiles, which are summarized in Fig. 1. Cluster analysis
The six observed clusters can be further subdivided into three functional groups. The first group exhibits a quick and steadily intensifying response (cluster 1 representing continuous down-regulation, while cluster 2 represents the reverse pattern, Figs 1 and 2). A second group displays a weak or even slightly negative response within the first 6 h followed by an increased expression after 48 h (cluster 3 and 5) and a third group that shows a distinct down-regulation after 6 h and is subsequently up-regulated (cluster 4 and 6) (Figs 1 and 2). In general, after 6 h approximately half of the genes were down- and half of the genes were up-regulated, whereas after 48 h only 3 of 62 are down-regulated in response to T3 (cluster 1, Fig. 1). Among the T3-responsive genes, which we have identified in our study, several genes have already been described to be T3 regulated, e.g. ANT2, myelin basic protein, ATP synthase subunit or apolipoprotein AIV (Dümmler et al. 1996, Miller et al. 2001, Clement et al. 2002, Flores-Morales et al. 2002). Thus, our screening protocol was successful in identifying differentially T3-regulated genes.
Results Microarray analyses
Verification of microarray hybridization data
Utilizing DNA microarray technology we investigated 4608 different genes to gain detailed information about T3-mediated gene expression patterns in rat liver in vivo. Total RNA from the
Representative examples of differentially regulated genes from each of the six different gene profile clusters were selected for further characterization by either northern blot experiments or quantitative
Figure 1 Expression dynamics of T3-responsive genes in rat liver in vivo. Temporal expression profiles of 62 arrayed features exhibiting change over time in response to T3 were hierarchically clustered by Euclidean distance measurement and visualized in a clustergram using Treeview software (see Materials and methods). Microarray experiments for each time point are in columns; individual genes are in rows. Red indicates increasing mRNA levels and green denotes decreasing mRNA levels at distinct time points after the administration of T3. The degree of colour saturation reflects the magnitude of the ratio (see colour key at the bottom). Journal of Molecular Endocrinology (2003) 31, 291–303
www.endocrinology.org
Expression patterns in T3-treated rats ·
www.endocrinology.org
J M WEITZEL, S HAMANN
and others 297
Journal of Molecular Endocrinology (2003) 31, 291–303
298
J M WEITZEL, S HAMANN
and others ·
Expression patterns in T3-treated rats
Journal of Molecular Endocrinology (2003) 31, 291–303
www.endocrinology.org
Expression patterns in T3-treated rats ·
real-time PCR. For northern blot analysis, identical amounts of total RNA (combined from three different animals per time point) were electrophoresed and visualized by staining with ethidium bromide (data not shown). Additional quality control was performed by hybridization with the ‘housekeeping’ gene ubiquitin (data not shown). In summary, we were able to reproduce the cluster profiles which have been detected by microarray hybridization using northern blot hybridization and quantitative real-time PCR (Fig. 2). Transient transfection analysis of late T3-induced genes
Since at least 36 out of the 62 differently regulated genes (58%) were not significantly regulated within the first 24 h (clusters 3 and 5), we tested the hypothesis that eight gene promoters of six late-induced genes (e.g. ANT2 (Table 1, No. 21)) (Scarpulla et al. 1986, Luciakova & Nelson 1992, Weitzel et al. 2001b) are directly regulated via a TRE. As indicated in Table 2 (and data not shown) we were unable to detect a TRE in these gene promoters by transient transfection experiment in human hepatocarcinoma HepG2 cells. As positive control we used the well-characterized TREcontaining promoter B of the rat mitochondrial glycerol-3-phosphate dehydrogenase gene (an early-induced gene (Dümmler et al. 1996)) under identical experimental conditions (Weitzel et al. 2001a). Candidates for an intermediate factor of T3 action
Numerous distinguishable T3-mediated expression patterns, together with a lack of TREs in late-induced gene promoters, suggested multiple pathways of T3-mediated modulation of gene expression which probably includes the action of an intermediate factor(s). In this study we analyzed: (i) rather ubiquitously expressed transcription factors, (ii) ligand-dependent transcription factors and (iii)
J M WEITZEL, S HAMANN
and others 299
coactivators of the PGC-1 family as candidates for an intermediate factor of T3 action, which were either undetectable by microarray hybridisation or not present on the array (data not shown). In order to analyze the gene expression levels 0, 6, 24 and 48 h after the administration of T3 we performed quantitative real-time PCR, as described above. We observed no significant T3-regulation of the transcription factors CREB, HCF, HNF4, Sp1 and YY1, but discerned a regulation of the mitochondrial transcription factor A (mtTFA) (Table 3). Analyses of the liganddependent transcription factors PPARs (peroxisome proliferator-activated receptors) revealed a different T3-mediated regulation of the three known isoforms. Whereas PPAR was late-induced by T3, PPAR was not induced and PPAR was down-regulated within the first 6 h and remains reduced for at least 48 h after the administration of T3. The coactivators PRC (Andersson & Scarpulla 2001) and PERC/PGC-1 (PGC-1-related estrogen receptor coactivator (Kressler et al. 2002, Lin et al. 2002)) are regulated within the first 6 h after T3 treatment and the abundance remains high for at least 48 h. The T3-mediated expression patterns of transcription factors and coactivators are summarized in Table 3. Alteration of protein concentrations in response to T3
In order to verify that an increased concentration of mRNA molecules (which we measured by microarray experiments, northern analyses and quantitative PCR) also leads to increased protein abundance, we performed an analysis by western blot. Western blot experiments were conducted using identical amounts of total protein extract from livers of untreated and T3-treated rats (0, 24 and 48 h after the administration of T3) that were loaded onto an SDS-PAGE and blotted according to standard protocols. Not surprisingly, protein quantities were not altered when the amount of corresponding mRNA
Figure 2 Cluster profiles. Gene expression profiles were algorithmically subdivided into six clusters. The average expression profile of each gene cluster is shown in a line graph; n, number of array features within each cluster. Selected genes from the different gene clusters were verified by northern blot hybridization or quantitative real-time PCR. Liver total RNA, combined from three hypothyroid (0 h) and three T3-treated hypothyroid rats (6, 24 and 48 h after the administration of T3) per time point, were analyzed. Cth, cystathionine -lyase; Apo AIV, apolipoprotein AIV; ATP synth., ATP synthase subunit. www.endocrinology.org
Journal of Molecular Endocrinology (2003) 31, 291–303
300
J M WEITZEL, S HAMANN
and others ·
Expression patterns in T3-treated rats
Table 3 Differentially regulated candidate intermediate factors in T3-treated rats mRNA levels (fold increase)
Protein levels (fold increase)
6h
24 h
48 h
24 h
48 h
0·8 0·7 0·8 1·6 0·8 1·4
1·1 0·9 1·6 3·6 0·7 1·7
n.d. n.d. n.d. n.d. 0·8 n.d.
n.d. n.d. n.d. n.d. 1·0 n.d.
2) peroxisome proliferator-activated receptors (PPARs) PPAR 1·5 2·8 2·2 PPAR 1·2 1·3 1·1 PPAR 0·5 0·5 0·3
1·1 1·0 0·7
1·2 1·0 0·6
3) PPAR coactivator 1 (PGC-1) PGC-1 13·0 PRC 1·7 PERC/PGC-1 5·0
2·3 n.d. n.d.
2·8 n.d. n.d.
Gene 1) transcription factors CREB 1·4 HCF 0·8 HNF-4 1·2 mtTFA 1·8 Sp1 1·4 YY1 1·4
family of coactivators 9·4 8·7 1·7 2·5 6·9 6·5
Fold increase data from at least three quantitative real-time PCR experiments or western blot experiments are given. The values were calculated by comparing the 6, 24 and 48 h hyperthyroid signal intensities to the hypothyroid signal intensities (0 h) within the same experiment. n.d., not done.
did not change (e.g. for Sp1, Fig. 3A). When mRNA abundance did not shift within the first 24 h after T3-treatment (e.g. liver fatty acid
binding protein (L-FABP), apolipoprotein AI and apolipoprotein AIV), we did not observe alteration in protein concentrations within 48 h either (Fig. 3A and data not shown). In contrast, those genes which were induced within the first 6 h also display increased protein abundance after 24 h which is more pronounced after 48 h (e.g. mGPDH and PGC-1, Fig. 3B).
Discussion
Figure 3 Western blot analysis of differentially T3-regulated proteins. After animals were treated with T3 for the time indicated, 50 mg total liver lysates were analyzed by SDS-PAGE. Western blot analysis was carried out using anti-Sp1, anti-L-FABP, anti-mGPDH or anti-PGC-1 antibodies as described in Material and methods. Journal of Molecular Endocrinology (2003) 31, 291–303
T3 has a profound influence in mitochondrial biogenesis and metabolic balance. Consistent with this long known observation, DNA microarray analysis identified numerous genes implicated in metabolic pathways to be regulated by T3 (Feng et al. 2000, Miller et al. 2001, Weitzel et al. 2001b, Clement et al. 2002, Flores-Morales et al. 2002, Wood et al. 2002). Our microarray analysis detected changes in the expression of ANT2, apolipoprotein AIV, HMG-CoA synthase and ATP synthase subunit, all four cases representing examples of T3-regulated genes that are associated with metabolic pathways. However, the majority of the differentially regulated genes are not primarily involved in metabolic functions; they are associated www.endocrinology.org
Expression patterns in T3-treated rats ·
with a wide variety of cellular pathways, encompassing translation, protein turnover, cell structure and apoptosis-associated proteins (Table 1). Cluster analysis divided the 62 differentially regulated genes into six expression profiles (Figs 1 and 2). Remarkably, the vast majority of T3responsive genes were found to be represented by clusters 2, 3 and 5 (52 of 62 genes) (Fig. 1). In contrast, cluster 1 (3 of 62 genes), cluster 4 (3/62) and cluster 6 (4/62) apparently depict rarely occurring patterns. It is interesting to compare the very similar T3-dependent gene expression patterns in rat liver and Xenopus laevis metamorphosis. Metamorphosis in Xenopus laevis is characterized by a non-functional thyroid gland in embryos, which is converted into a functional (thyroid hormonesecreting) organ during larval development – a classical T3-dependent differentiation model (Shi et al. 1996, Berry et al. 1998). Among the differentially T3-regulated genes during metamorphosis, Shi and co-workers described a biphasic expression pattern for the neural-specific tubulin subunit in Xenopus (Denver et al. 1997) and, noteworthy, we identified a highly similar biphasic expression pattern for an -tubulin isoform in rat liver (cluster 6, Table 1, No. 59). This suggests an evolutionarily conserved principle of T3 activation of homologous genes in amphibian and mammals. Numerous distinguishable T3-mediated expression patterns suggest multiple pathways of T3-mediated modulation of gene expression. This is supported by recent findings of gene-specific effects of TR, TR and SRC-1 on T3-regulated gene expression (Flores-Morales et al. 2002, Takeuchi et al. 2002, Sadow et al. 2003), although temporal expression patterns have not been analyzed in these studies. Furthermore, some genes are known to be endogenously regulated by T3 in vivo; however, a thyroid hormone response element (TRE) has not yet been identified within their regulatory elements (Table 2). Since the great majority of TREs are located several hundred base pairs upstream of the transcriptional start site (Yen 2001, Weitzel et al. 2003) it is unlikely that a TRE is located upstream of the investigated promoter sequences; however we cannot completely exclude this possibility. Moreover, all these genes are regulated late ( 12–24 h) in response to T3. This leads to the assumption that there might be an www.endocrinology.org
J M WEITZEL, S HAMANN
and others 301
intermediate factor of T3 action, modulating the expression of late-induced, no TRE-containing genes. Since the synthesis of this intermediate factor requires time, the T3 response of the late-induced target genes exhibits a latency period. In our previous study, we suggested the transcription factor NRF-1 and the coactivator PGC-1 as candidate factors for such an intermediate factor. Unfortunately, a common sequence motif in late-induced genes (e.g. a common NRF-1 site) has not yet been described in late-induced gene promoters (Weitzel et al. 2003). We therefore tested the hypothesis whether other transcription factors might serve as intermediate factors. However, neither Sp1 nor YY1, CREB, HCF or HNF-4 (either transcriptional regulators of late-induced genes or PGC-1 interaction partners (Evans & Scarpulla 1989, Basu et al. 1993, Seelan & Grossman 1997, Tagami et al. 1999, Zaid et al. 1999, Yoon et al. 2001, Lin et al. 2002)) have been shown to be regulated by T3 (Table 3). In contrast, mtTFA, the major transcription factor of the mitochondrial genome, is up-regulated within 48 h after administration of T3, in agreement with previous studies (Garstka et al. 1994). Many of the detected target genes are involved in lipid metabolism (e.g. apolipoprotein AIV, enoylCoA isomerase, HMG-CoA synthase, L-FABP, sterol carrier protein 2 (SCP2)) and are known to be regulated by PPARs (Hsu et al. 2001, Poirier et al. 2001). Furthermore, the concentration of activated fatty acids is increased in hyperthyroidism (Müller & Seitz 1981). We therefore evaluated the hypothesis that PPAR transcription factors might serve as intermediate factors. Surprisingly, the observed expression pattern of PPAR isoforms , and exhibits a completely different profile in response to T3 (Table 3). Solely PPAR was (down-)regulated within 6 h and remains downregulated for at least 48 h, a very unusual expression profile (see cluster 1, Fig. 1). Since PPAR has also been described as a differentiation factor for adipogenesis rather than a regulator of fatty acid oxidation (Lowell 1999), down-regulation may primarily be involved in differentiation processes in response to T3. Finally, we examined whether expression of the coactivators of the PGC-1 family is regulated by T3. We previously described PGC-1 to be dramatically regulated by T3 in rat liver within 6 h (Weitzel et al. 2001b). Very similar results have now Journal of Molecular Endocrinology (2003) 31, 291–303
302
J M WEITZEL, S HAMANN
and others ·
Expression patterns in T3-treated rats
been reported for PGC-1 related coactivator (PRC) and PERC/PGC-1. Thus, the PGC-1 family members of coactivators are highly attractive candidates for mediating the late induction pattern (Table 3 and Fig. 3). Investigations of the steady-state level of mRNA concentrations are hampered by the question of whether modulation of the mRNA concentration also reflects a modulation of the protein concentration. Hence we tested various candidate proteins, where the corresponding mRNA is: (i) not regulated by T3, (ii) late regulated by T3 and (iii) early regulated by T3. We observed that a change in protein concentration did not occur within 48 h, when there is no significant alteration in the mRNA within the first 24 h (Fig. 3A and Table 3). However, in those cases where we observed an increase in the mRNA abundance within 6 h after T3-treatment (early induction), the protein concentration was also altered after 24 h and the change was even more pronounced 48 h after administration of T3 (Fig. 3B and Table 3). In summary, our data reveal a complex expression pattern in response to T3 in rat liver in vivo, suggesting various signal transduction pathways of T3 action. Beside the ‘classical’ pathway of T3-mediated gene regulation by thyroid hormone receptor binding to TREs, an additional pathway appears to be mediated by transcription factors like NRF-1 and PPAR (but not CREB, HCF, HNF-4, NRF-2, Sp1, YY1, PPAR, PPAR) and coactivators like the PGC-1 family of coactivators. It has been shown that ectopic expression of PGC-1 mimics T3-mediated effects (Wu et al. 1999, Knutti & Kralli 2001, Yoon et al. 2001, Puigserver & Spiegelman 2003). It will be interesting to address the question of whether signal pathways are lost after depletion of these factors, as soon as suitable knockout models become available.
Acknowledgements We are indebted to Britta Schwarzloh for help with the LightCycler, and to Richard C Scarpulla (Northwestern University, Chicago, USA), Jerome F Strauss III (University of Pennsylvania, USA) and Rudolf J Wiesner (University of Köln, Germany) for kind gifts of plasmid DNA. This work was supported by grants from the Hamburgische Journal of Molecular Endocrinology (2003) 31, 291–303
Wissenschaftliche Stiftung and the Deutsche Forschungsgemeinschaft (WE 2458/3–1) to J M W.
References Andersson U & Scarpulla RC 2001 PGC-1-related coactivator, a novel, serum-inducible coactivator of nuclear respiratory factor 1-dependent transcription in mammalian cells. Molecular and Cellular Biology 21 3738–3749. Basu A, Park K, Atchison ML, Carter RS & Avadhani NG 1993 Identification of a transcriptional initiator element in the cytochrome c oxidase subunit Vb promoter which binds to transcription factors NF-E1 (YY-1, delta) and Sp1. Journal of Biological Chemistry 268 4188–4196. Berry DL, Schwartzman RA & Brown DD 1998 The expression pattern of thyroid hormone response genes in the tadpole tail identifies multiple resorption programs. Developmental Biology 203 12–23. Clement K, Viguerie N, Diehn M, Alizadeh A, Barbe P, Thalamas C, Storey JD, Brown PO, Barsh GS & Langin D 2002 In vivo regulation of human skeletal muscle gene expression by thyroid hormone. Genome Research 12 281–291. Davis PJ, Tillmann HC, Davis FB & Wehling M 2002 Comparison of the mechanisms of nongenomic actions of thyroid hormone and steroid hormones. Journal of Endocrinological Investigation 25 377–388. Denver RJ, Pavgi S & Shi YB 1997 Thyroid hormone-dependent gene expression program for Xenopus neural development. Journal of Biological Chemistry 272 8179–8188. Dümmler K, Müller S & Seitz HJ 1996 Regulation of adenine nucleotide translocase and glycerol 3-phosphate dehydrogenase expression by thyroid hormones in different rat tissues. Biochemical Journal 317 913–918. Eisen MB & Brown PO 1999 DNA arrays for analysis of gene expression. Methods in Enzymology 303 179–205. Eisen MB, Spellman PT, Brown PO & Botstein D 1998 Cluster analysis and display of genome-wide expression patterns. PNAS 95 14863–14868. Evans MJ & Scarpulla RC 1989 Interaction of nuclear factors with multiple sites in the somatic cytochrome c promoter. Characterization of upstream NRF-1, ATF, and intron Sp1 recognition sequences. Journal of Biological Chemistry 264 14361–14368. Feng X, Jiang Y, Meltzer P & Yen PM 2000 Thyroid hormone regulation of hepatic genes in vivo detected by complementary DNA microarray. Molecular Endocrinology 14 947–955. Flores-Morales A, Gullberg H, Fernandez L, Stahlberg N, Lee NH, Vennstrom B & Norstedt G 2002 Patterns of Liver Gene Expression Governed by TRbeta. Molecular Endocrinology 16 1257–1268. Garstka HL, Facke M, Escribano JR & Wiesner RJ 1994 Stoichiometry of mitochondrial transcripts and regulation of gene expression by mitochondrial transcription factor A. Biochemical and Biophysical Research Communications 200 619–626. Halgren RG, Fielden MR, Fong CJ & Zacharewski TR 2001 Assessment of clone identity and sequence fidelity for 1189 IMAGE cDNA clones. Nucleic Acids Research 29 582–588. Hsu MH, Savas U, Griffin KJ & Johnson EF 2001 Identification of peroxisome proliferator-responsive human genes by elevated expression of the peroxisome proliferator-activated receptor alpha in HepG2 cells. Journal of Biological Chemistry 276 27950–27958. Knutti D & Kralli A 2001 PGC-1, a versatile coactivator. Trends in Endocrinology and Metabolism 12 360–365. Kressler D, Schreiber SN, Knutti D & Kralli A 2002 The PGC-1related protein PERC is a selective coactivator of estrogen receptor alpha. Journal of Biological Chemistry 277 13918–13925. www.endocrinology.org
Expression patterns in T3-treated rats · Li R, Luciakova K, Zaid A, Betina S, Fridell E & Nelson BD 1997 Thyroid hormone activates transcription from the promoter regions of some human nuclear-encoded genes of the oxidative phosphorylation system. Molecular and Cellular Endocrinology 128 69–75. Lin J, Puigserver P, Donovan J, Tarr P & Spiegelman BM 2002 Peroxisome proliferator-activated receptor gamma coactivator 1 beta (PGC-1 beta), a novel PGC-1-related transcription coactivator associated with host cell factor. Journal of Biological Chemistry 277 1645–1648. Lowell BB 1999 PPARgamma: an essential regulator of adipogenesis and modulator of fat cell function. Cell 99 239–242. Luciakova K & Nelson BD 1992 Transcript levels for nuclearencoded mammalian mitochondrial respiratory-chain components are regulated by thyroid hormone in an uncoordinated fashion. European Journal of Biochemistry 207 247–251. Miller LD, Park KS, Guo QM, Alkharouf NW, Malek RL, Lee NH, Liu ET & Cheng SY 2001 Silencing of Wnt signaling and activation of multiple metabolic pathways in response to thyroid hormone-stimulated cell proliferation. Molecular and Cellular Biology 21 6626–6639. Müller MJ & Seitz HJ 1981 Starvation-induced changes of hepatic glucose metabolism in hypo- and hyperthyroid rats in vivo. Journal of Nutrition 111 1370–1379. Poirier H, Niot I, Monnot MC, Braissant O, Meunier-Durmort C, Costet P, Pineau T, Wahli W, Willson TM & Besnard P 2001 Differential involvement of peroxisome-proliferator-activated receptors alpha and delta in fibrate and fatty-acid-mediated inductions of the gene encoding liver fatty-acid-binding protein in the liver and the small intestine. Biochemical Journal 355 481–488. Puigserver P & Spiegelman BM 2003 Peroxisome proliferator-activated receptor-gamma coactivator 1 alpha (PGC-1 alpha): transcriptional coactivator and metabolic regulator. Endocrine Reviews 24 78–90. Robyr D, Wolffe AP & Wahli W 2000 Nuclear hormone receptor coregulators in action: diversity for shared tasks. Molecular Endocrinology 14 329–347. Sadow PM, Chassande O, Gauthier K, Samarut J, Xu J, O’Malley BW & Weiss RE 2003 Specificity of thyroid hormone receptor subtype and steroid receptor coactivator-1 on thyroid hormone action. American Journal of Physiology Endocrinology and Metabolism 284 E36–E46. Scarpulla RC 2002 Nuclear activators and coactivators in mammalian mitochondrial biogenesis. Biochimica et Biophysica Acta 1576 1–14. Scarpulla RC, Kilar MC & Scarpulla KM 1986 Coordinate induction of multiple cytochrome c mRNAs in response to thyroid hormone. Journal of Biological Chemistry 261 4660–4662. Seelan RS & Grossman LI 1997 Structural organization and promoter analysis of the bovine cytochrome c oxidase subunit VIIc gene. A functional role for YY1. Journal of Biological Chemistry 272 10175–10181. Shi YB, Wong J, Puzianowska-Kuznicka M & Stolow MA 1996 Tadpole competence and tissue-specific temporal regulation of
www.endocrinology.org
J M WEITZEL, S HAMANN
and others 303
amphibian metamorphosis: roles of thyroid hormone and its receptors. Bioessays 18 391–399. Tagami T, Park Y & Jameson JL 1999 Mechanisms that mediate negative regulation of the thyroid-stimulating hormone alpha gene by the thyroid hormone receptor. Journal of Biological Chemistry 274 22345–22353. Takeuchi Y, Murata Y, Sadow P, Hayashi Y, Seo H, Xu J, O’Malley BW, Weiss RE & Refetoff S 2002 Steroid receptor coactivator-1 deficiency causes variable alterations in the modulation of T(3)-regulated transcription of genes in vivo. Endocrinology 143 1346–1352. Tata JR, Ernster L, Lindberg O, Arrhenius, E Pedersen S & Hedman R 1963 The action of thyroid hormone at the cell level. Biochemical Journal 86 408–428. Weitzel JM, Kutz S, Radtke C, Grott S & Seitz HJ 2001a Hormonal regulation of multiple promoters of the rat mitochondrial glycerol-3-phosphate dehydrogenase gene: identification of a complex hormone-response element in the ubiquitous promoter B. European Journal of Biochemistry 268 4095–4103. Weitzel JM, Radtke C & Seitz HJ 2001b Two thyroid hormonemediated gene expression patterns in vivo identified by cDNA expression arrays in rat. Nucleic Acids Research 29 5148–5155. Weitzel JM, Iwen KAH & Seitz HJ 2003 Regulation of mitochondrial biogenesis by thyroid hormone. Experimental Physiology 88 121–128. Wood WM, Sarapura VD, Dowding JM, Woodmansee WW, Haakinson DJ, Gordon DF & Ridgway EC 2002 Early gene expression changes preceding thyroid hormone-induced involution of a thyrotrope tumor. Endocrinology 143 347–359. Wu Z, Puigserver P, Andersson U, Zhang C, Adelmant G, Mootha V, Troy A, Cinti S, Lowell B, Scarpulla RC & Spiegelman BM 1999 Mechanisms controlling mitochondrial biogenesis and respiration through the thermogenic coactivator PGC-1. Cell 98 115–124. Yen PM 2001 Physiological and molecular basis of thyroid hormone action. Physiological Reviews 81 1097–1142. Yoon JC, Puigserver P, Chen G, Donovan J, Wu Z, Rhee J, Adelmant G, Stafford J, Kahn CR, Granner DK, Newgard CB & Spiegelman BM 2001 Control of hepatic gluconeogenesis through the transcriptional coactivator PGC-1. Nature 413 131–138. Zaid A, Li R, Luciakova K, Barath P, Nery S & Nelson BD 1999 On the role of the general transcription factor Sp1 in the activation and repression of diverse mammalian oxidative phosphorylation genes. Journal of Bioenergetics and Biomembranes 31 129–135. Zhang J & Lazar MA 2000 The mechanism of action of thyroid hormones. Annual Review of Physiology 62 439–466. Zilz ND, Murray MB & Towle HC 1990 Identification of multiple thyroid hormone response elements located far upstream from the rat S14 promoter. Journal of Biological Chemistry 265 8136–8143.
Received 7 April 2003 Accepted 17 June 2003
Journal of Molecular Endocrinology (2003) 31, 291–303
