Estrogen Regulation of Creatine Kinase-B in the Rat Uterus
Brian T. Pentecost, Lori Mattheiss, Herbert W. Dickerman, and S. Anand Kumar* Wadsworth Center for Laboratories and Research New York State Department of Health Albany, New York 12201-0509
Creatine kinase-B (CKB) synthesis is rapidly and specifically induced by estrogen in the uterus of the immature rat. This study indicates that this elevation is due at least in part to increases in the levels of mRNA for CKB. The stimulation of CKB mRNA levels is rapid (a 7- to 10-fold increase is detected 1-3 h after estrogen administration), but transient, as levels return to near control values by 6 h. Analysis of cDNAs to both uterine and brain CKB mRNA indicate that the same sequence is expressed in both tissues despite earlier observations of heterogeneity of the protein isolated from the two tissues. A 1.7-kilobasepair DNA fragment containing the CKB promoter and 5' flanking sequences confers estrogen sensitivity on expression of the bacterial chloramphenicol acetyl transferase gene in HeLa cells on cotransfection with an estrogen-receptor expression vector. However, the CKB promoter sequences lack any motif with convincing similarity to the currently accepted consensus estrogen response element GGTCAnnnTGACC. (Molecular Endocrinology 4: 1000-1010, 1990)
elude: mol wt, as determined by sodium dodecyl sulfatepolyacrylamide gel electrophoresis (brain, 46 kDa; uterus, 49 kDa) (5); isoelectric point, as determined by gel electrofocusing (brain, 6-6.5; uterus, 5.3-5.6) (5); and differential reactivity with several monoclonal antibodies (6). This suggested that different CKB genes might be expressed in estrogen-responsive and nonresponsive tissues or, as reported for chicken CKB (7), that alternative splicing of a common transcript might occur. Another aspect of estrogen regulation of rat uterine CKB gene expression is whether the hormonal effect is direct or requires additional protein synthesis. Extensive studies by Katzenellenbogen and Gorski (8) demonstrated, using uterine organ culture, that the action of estrogen could not be duplicated by other steroids or cAMP, and that the degree of induction of CKB synthesis (measured as IP) correlated with nuclear receptor occupancy (8). Importantly, indirect experiments indicated that synthesis of the IP mRNA was not blocked by cycloheximide, but was abolished by actinomycin-D (9, 8), fulfilling the best criteria then available for establishing direct transcriptional regulation. This suggests that CKB provides a useful model for studies of estrogen regulation of gene expression. We now report on analysis of uterine CKB using molecular probes. We demonstrate that the levels of the CKB mRNA are rapidly but transiently stimulated by estrogen and show that receptor-dependent estrogen-modulated expression of chloramphenicol acetyl transferase (CAT) activity in cell culture can be conferred by the CKB promoter and 5' flanking sequences. Our results also indicate that the differences between brain and uterine CKB protein are not due to variations in the sequences expressed.
INTRODUCTION
Estrogen stimulation of a specific induced protein (IP) before general increases in protein synthesis in the immature rat uterus was originally detected by Notides and Gorski (1). Its identity was unknown until cytosol mixing studies by Reiss and Kaye (2) demonstrated that IP copurified with the B isozyme of creatine kinase (CK; ATP; creatine A/-phosphotransferase, EC 2.7.3.2) from rat brain. Subsequently, an estrogen-inducible increase in the amount of immunoprecipitable creatine kinase-B (CKB), upon in vitro translation of uterine RNA, provided more direct evidence that CKB mRNA levels were stimulated by the hormone (3, 4). Studies from our laboratory on IP purified solely from rat uteri demonstrated that the homogenous uterine protein has CKB activity (5), but that it is not identical to CKB isolated from rat brain. Variant properties in-
RESULTS Similar CKB Sequences Are Expressed in Brain and Uterus Our initial studies addressed the relationship of CKB sequences expressed in rat brain and uteri, to both resolve the discrepancies seen on analysis of the enzyme from the two tissues and ensure that the correct
0888-8809/90/1000-1010$02.00/0 Molecular Endocrinology Copyright © 1990 by The Endocrine Society
1000
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 September 2016. at 00:39 For personal use only. No other uses without permission. . All rights reserved.
Creatine Kinase-B Regulation
gene was later selected for analysis of regulation at the molecular level. A 1300-basepair (bp) uterine CKB cDNA was isolated from a Agt11 library constructed using uterine mRNA of estrogen-treated immature rats and subcloned to give pRUTCKB (Fig. 1). Seven small discrepancies between the uterine cDNA sequence and the published rat brain cDNA sequence (10) were detected; three of these differences were also found on comparison with the published rat CKB gene sequence (11) (Fig. 1). However, we found that our independently isolated rat brain sequence (pCKBBR5) was identical to the uterine sequence at those points, suggesting that the conflicts were due to synthesis artifacts or sequencing errors in the published data (10, 11). Our data eliminate a Hpall site at position 810. Restriction analysis of the cDNA inserts of both pRUTCKB and pCKBBR5 could not detect a Hpall site at 810, confirming our sequence assignment. In addition, the single difference in the coding region affecting the translated product causes rat CKB to become homologous at that point to CK from all other species that have been determined (12). The 5' end of both of our CKB clones was within the coding region determined by Benfield et al. (10), owing to cleavage at an internal EcoRI site during construction. The cDNA insert of pRUTCKB contained no sequences corresponding to a poly-A tract, despite preparation of polyadenylated RNA and priming with polydT primers, presumably owing to incomplete second strand synthesis and trimming by S1 nuclease. The last base of the uterine sequence corresponded to the final transcript-derived base of the published CKB1 brain cDNA (11,10), which is followed by a poly-A tract. Our rat brain cDNA was also polyadenylated, but contained three additional bases (dCdTdA) before the poly-A tract. The extra bases were compatible with the sequence of the published rat CKB genomic clone (11). Heterogeneity of 3' untranslated sequences in dog CKB cDNAs has been proposed as possible evidence of multiple genes; however, we consider that the variation in the rat sequences is probably due to imprecision in processing of the mRNA precursors, as also proposed by other groups (13, 14). The labeled cDNA insert of pRUTCKB hybridized to similarly sized [1.6 kilobases (kb)] RNA species in Northern blots of brain and uterine total RNA (Fig. 2B), with no evidence of multiple forms in either tissue. Uterine CKB Is Specifically Stimulated by 17/3Estradiol (E2) Preliminary Northern blot analysis of RNA from immature female rats (17-19 days old) indicated that significant elevation of CKB mRNA levels occurred after estrogen administration (Fig. 2B). Additional analysis was, therefore, carried out, primarily using slot blots for improved quantification. Significant background levels of the CKB mRNA were detected in RNA from control rats (Figs. 2B and 3A). E2 administration (12 Mg/kg) caused a rapid but tran-
1001
sient increase in CKB mRNA levels detectable 15-30 min after treatment (data not shown). A 7- to 10-fold increase was found after 1 h, and levels peaked in the 1 - to 3-h period before returning to approximately control values by 6 h (Fig. 3A). There was some interexperimental variation in both the timing and magnitude of the response, with levels at 1 h occasionally exceeding those 3 h posttreatment. Levels of actin and /?-tubulin mRNAs were also monitored as controls. Tubulin demonstrated only a limited (~3-fold) and gradual response over the 6-h period (Fig. 2A and our unpublished slot blot analysis), as found by Loose-Mitchell and co-workers (15), while actin mRNA levels showed a definite rise but with a lag period distinctly greater than that for CKB mRNA (Fig. 3B). The results for actin may be complicated by the presence of several isoforms (16,17) detectable by the actin probe used (18). Nevertheless, little increase was found until 3 h after estrogen treatment. Analysis with a ribsomal probe (our unpublished data) indicates that samples from the various groups contain similar amounts of RNA. There was an approximately linear relationship (up to 12 Mg/kg E2) between E2 administered and stimulation of CKB mRNA levels 3 h later (Fig. 4). This pattern of response is similar to that seen for uterine wet weight at 3 h and correlates with nuclear estrogen receptor (ER) occupancy (19). This contrasts with the response of actin mRNA levels, which were significantly elevated 3 h after E2 administration by the lowest dose (0.75 Mg/kg) evaluated (Fig. 4). The apparent requirement for only fractional receptor occupancy to significantly elevate actin mRNA levels is similar to results for glucose oxidation (20) and c-fos mRNA levels (15) in the uterus. Hormonal specificity of the uterine CKB mRNA response was evaluated in groups of 20-day-old female rats. Neither progesterone, dihydrotestosterone, nor dexamethasone (all at 120 Mg/kg) caused noticeable increases in uterine CKB mRNA levels 1 h after administration (Fig. 5). E2 (30 Mg/kg), in contrast, caused a greater than 10fold increase in CKB mRNA levels over control levels in the same experimental series (Fig. 5). Subsequent analysis of the CKB genomic sequence (Fig. 6) failed to identify sequences resembling the hormone response element for nonestrogenic steroids. Tamoxifen, an antiestrogen that can sometimes act as an agonist, neither increased nor decreased uterine CKB mRNA levels at 1 h (Fig. 5). Genomic Clone Isolation Rat genomic sequences encoding the CKB gene were isolated from Charon 4a rat genomic libraries (21) by screening with the labeled insert of pRUTCKB and rescreened with the nonconserved 3' untranslated region of the CKB cDNA to confirm that they were related to CKB rather than CKM. Subcloned fragments predicted to encode the promoter and adjacent sequences
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 September 2016. at 00:39 For personal use only. No other uses without permission. . All rights reserved.
MOL ENDO-1990 1002
Vol 4 No. 7
151
GAATTCCCTGATCTGAGCAGCCACAACAACCATATGGCCAAGGTGCTGACCCCCGAGCTG GluPheProAspLeuSerSerHisAsnAsnHisMetAlaLysValLeuThrProGluLeu
210
211
TACGCGGAGCTCCGTGCCAAGTGCACGCCGAGCGGCTTTACGTTGGACGACGCCATCCAG TyrAlaGluLeuArgAlaLysCysThrProSerGlyPheThrLeuAspAspAlalleGln
270
2 71
ACTGGCGTAGACAATCCGGGCCACCCGTACATCATGACAGTGGGTGCAGTGGCGGGCGAC ThrGlyValAspAsnProGlyHisProTyrlleMetThrValGlyAlaValAlaGlyAsp
330
3 31
GAGGAGAGTTACGACGTATTCAAGGACCTTTTCGACCCCATCATCGAGGACCGGCACGGC GluGluSerTyrAspValPheLysAspLeuPheAspProIlelleGluAspArgHisGly
39 0
391
GGCTACCAGCCCAGTGATGAGCACAAGACTGACCTCAACCCAGACAACCTGCAGGGCGGC GlyTyrGlnProSerAspGluHisLysThrAspLeuAsnProAspAsnLeuGlnGlyGly
4 50
4 51
GATGACCTGGACCCCAACTACGTGCTGAGCTCGCGGGTGCGCACAGGCCGAAGCATCCGC AspAspLeuAspProAsnTyrValLeuSerSerArgValArgThrGlyArgSerlleArg
510
511
GGCTTCTGCCTCCCCCCTCACTGCAGCCGTGGGGAGCGCCGCGCCATCGAGAAGCTGGCA GlyPheCysLeuProProHisCysSerArgGlyGluArgArgAlalleGluLysLeuAla
570
571
GTAGAAGCCCTGTCCAGCCTAGATGGCGACCTGTCTGGCAGGTACTATGCGCTCAAGAGC ValGluAlaLeuSerSerLeuAspGlyAspLeuSerGlyArgTyrTyrAlaLeuLysSer
630
631
ATGACCGAGGCGGAGCAGCAGCAGCTCATTGACGACCACTTCCTCTTCGACAAGCCTGTG GACGAG MetThrGluAlaGluGlnGlnGlnLeuIleAspAspHisPheLeuPheAspLysProVal AspGlu
690
691
TCGCCTCTGCTGCTGGCCTCCGGCATGGCCCGCGACTGGCCGGATGCTCGCGGCATTTGG SerProLeuLeuLeuAlaSerGlyMetAlaArgAspTrpProAspAlaArgGlylleTrp
7 50
751
CACAATGACAATAAGACGTTCCTGGTGTGGATCAACGAGGAGGACCACCTGCGGGTTATC [C] HisAsnAspAsnLysThrPheLeuValTrpIleAsnGluGluAspHisLeuArgVallle
810
811
TCCATGCAGAAAGGGGGCAACATGAAGGAAGTTTTCACGCGATTCTGCACTGGCCTCACT SerMetGlnLysGlyGlyAsnMetLysGluValPheThrArgPheCysThrGlyLeuThr
870
871
CAGATTGAAACTCTCTTCAAGTCTAAGAACTATGAGTTCATGTGGAACCCTCACCTGGGC GlnlleGluThrLeuPheLysSerLysAsnTyrGluPheMetTrpAsnProHisLeuGly
930
931
TACATCCTCACGTGCCCATCCAACCTGGGCACTGGGCTTCGGGCAGGCGTGCACATCAAG [A] TyrlleLeuThrCysProSerAsnLeuGlyThrGlyLeuArgAlaGlyValHisIleLys
990
991
CTGCCCCACCTGGGAAAGCACGAGAAGTTCTCGGAGGTGCTCAAGCGACTGCGGCTTCAG LeuProHisLeuGlyLysHisGluLysPheSerGluValLeuLysArgLeuArgLeuGln
1050
1051
AAGCGAGGCACAGGTGGTGTGGACACCGCTGCTGTGGGTGGAGTTTTTGATGTCTCCAAC LysArgGlyThrGlyGlyValAspThrAlaAlaValGlyGlyValPheAspValSerAsn
1110
1111
GCTGACCGCCTGGGCTTCTCGGAGGTGGAGCTGGTGCAGATGGTGGTGGACGGAGTGAAG
117 0
AlaAspArgLeuGlyPheSerGluValGluLeuValGlnMetValValAspGlyValLys 1171
CTACTCATTGAGATGGAGCAGCGGCTTGAGCAGGGTCAGCCCATTGACGACCTCATGCCT LeuLeuIleGluMetGluGlnArgLeuGluGlnGlyGlnProIleAspAspLeuMetPro
12 30
12 31
GCCCAGAAGTGAAGCC.TGGCCCTAGCCACCACCAGGCTGCCGCTTCCTAACTTATTACC AlaGlnLysEnd [G]
1290
1291
CGGGCAGTGCCCGCCATGCATCCTTGATGTTTGCCGCCGCCTGGCGGCTGAGCCCTTAGC TGCC...GGG
13 50
13 51
CTCGCTGTAGAGACTTCTGTCGCCCTGGGTAGAG..TTTATTTTT.CTGATGGCTAAGCT [CC] T
1410
1411
GTTGCAGACACTGAAATAAATTAGGGTTTGGCCTGCC
Fig. 1. Sequence of a Uterine cDNA to CKB The cDNA nucleotide sequence numbering is based on the proposed transcriptional start site of the gene (11), and the cDNA begins at an internal EcoRI site. Conflicts with published rat brain cDNA (10) and genomic (11) sequences are indicated below the uterine sequence. Conflicts between the uterine cDNA and published brain sequence (10), but not the gene, are in brackets.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 September 2016. at 00:39 For personal use only. No other uses without permission. . All rights reserved.
Creatine Kinase-B Regulation
1003
B
1
1
Fig. 2. Northern Blot Analysis of Uterine and Brain RNA Replicate blots were probed with /?-tubulin (A) and CKB (B) cDNAs. Track 1, Estrogenized brain; track 2, control uteri; track 3, estrogenized uteri. E2 was administered at 30 ^g/kg; controls received vehicle alone. Animals were killed 3 h after injection. Each track contained 5 Mg total RNA resolved in agarose-formaldehyde gels.
were subjected to sequencing, and results indicated that the clone was essentially identical to that analyzed by Benfield and co-workers (11). The sequence (Fig. 6) had the unusual dual promoter structure noted for both the human (22, 23) and rat CKB genes (11). Analysis was, therefore, concentrated on sequences up-stream from the promoter, the most likely region to contain sequences conferring estrogen sensitivity on the CKB gene and for which no data were available beyond - 1 kilobasepair (kbp). A final continuous sequence extending to -2900 bp was established (Fig. 6), as virtually all hormonal response elements in other genes have been found less than 2.5 kbp up-stream of the relevant promoters. Similarity to other CKB genomic sequences was assessed using the GCG Gap program. Similarity to the published rat CKB gene (11) for the region +400 to -900 was in excess of 95% and was 99% on exclusion of the region -540 to -490, which is GC rich and had previously been sequenced in only one orientation, apparently leading to compression artifacts. These values confirm that the CKB gene described is equivalent to that previously reported. Similarity to the human gene was high (90%) in the region from the proposed transcriptional start site out to -120 bp, but then poor (60%) for the remaining available data out to -400 bp (22), with no contiguous blocks of homologous bases greater than 10 bp. The limited human CKB gene sequence was, therefore, uninformative in the search for regulatory sequences outside the proximal promoter region. Studies on the transfected human CKB gene do, however, indicate that this family of sequences is capable of supporting expression of active CKB enzyme (23). The up-stream sequences of the CKB gene (Fig. 6) were screened for c/s-elements resembling the known examples of estrogen response elements (EREs) from
other genes. This was performed in the GCG analysis package using a consensus matrix derived from the EREs from chicken and Xenopus vitellogenins (24, 25) and human pS2 (26) and PRL (27, 28) genes. The analysis was inconclusive, as no potential motifs with less than three mismatches to the perfect palindromic element GGTCAnnnTGACC were detected, whereas the functional examples on which the survey was based have a maximum of two deviations. The best examples from the immediate flank are located at -1150 to -1162, -544 to -556, and -338 to -350. Additional surveys of the body of the rat CKB gene did not give any better candidate response elements, and delineation of active regulatory elements will have to rely on functional assays and receptor binding studies. CKB Promoter Sequences Can Confer Estrogen Sensitivity We have undertaken preliminary studies to demonstrate that DNA sequences associated with the CKB promoter can confer the capacity for regulation by estrogen. Transfections were carried out using the calcium phosphate (29) technique on HeLa cells. This cell line is readily transfected and has high levels of a number of nuclear transcription factors. These include SP1 (30), for which potential binding sites are found adjacent to a possible ERE motif up-stream of the CKB promoter (see Discussion). Human CKB has previously been succesfully expressed from a CKB fragment transfected into HeLa cells (23), but estrogen responsiveness was not evaluated. Cotransfection with pHEO, an ER expression clone (31), was necessary, as HeLa lacks functional ERs. Transfection and cell culture conditions were established using a positive control construct (pEREKAT4) containing a synthetic ERE linked to the herpes simples virus (HSV)-TK promoter regulating CAT expression. CAT expression from pEREKAT4 was enhanced by estrogen (Fig. 7, tracks 6-9) and dependent on cotransfection with the receptor clone (Fig. 7, tracks 9 and 10). Estrogen modulation of CAT expression from pCKBICAT, containing CKB sequences -1700 to +5 linked to CAT sequences of the promoterless pOCAT2 vector (32), depended on the presence of the receptor clone, indicating that estrogen action was mediated via its receptor (Fig. 7, tracks 4 and 5). The maximum degree of enhancement was only moderate (7-fold) compared with that which could be achieved with pEREKAT4 (30-fold; Fig. 7, tracks 6 and 9). We examined the sensitivity to estrogen of CAT expression from pCKBICAT in light of the high quantities of estrogen required to elevate CKB mRNA levels in the rat uterus. Low sensitivity to the hormone was again found, with little stimulation of CAT activity seen below 10~8 M (Fig. 7, tracks 1-4). This was in contrast to results with pEREKAT4, which gave a near-maximal response with concentrations 2 orders of magnitude lower (Fig. 7, tracks 6-9). In an alternative approach to the same
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 September 2016. at 00:39 For personal use only. No other uses without permission. . All rights reserved.
Vol 4 No. 7
MOL ENDO-1990 1004
4000
p3 5 2000 _l
z CO
2
4
6
TIME (hours after estradiol administration) 40 0 0 B
03
|
gp2 gpi *A B C D" IT B C
2000
03
CON
1 hr 3 hr
Hi JMBlilWlHiffiwII M l
6hr
gp3 DWA
B
C
DA
HHRHH
WmWm mHmKmmmwm
TIME (hours after estradiol adminstration) Fig. 3. Temporal Response of Estrogen-Modulated Gene Expression in the Uterus Replicate filters of total RNA from uteri of estrogenized rats (12 ^g/kg) were probed with labeled cDNAs to determine the responses of CKB (A) and actin (B). Each point represents the mean ± SE for three groups of five rats. A series of dilutions (A, 800 ng; B, 400 ng; C, 200 ng; D, 100 ng) was made for each sample; autoradiograms of the data used to construct the graphs are shown in the inset. CON, Control.
question we found that reduction of the amount of cotransfected receptor clone below 1 ^g also reduces CAT expression from pCKBICAT (Fig. 8), again indicating a low sensitivity. In contrast, pEREKAT4 demonstrated the highest activity, with aproximately 0.3 j*g pHEO, and activity declined with further increases in the amount of cotransfected receptor clone.
DISCUSSION Our data support the observation of Kaye and coworkers that a rapidly estrogen induced protein of the rat uterus, first detected by Notides and Gorski (1), is CKB. Previous results from our group suggested that the CKB from the uterus is not identical to that from
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 September 2016. at 00:39 For personal use only. No other uses without permission. . All rights reserved.
Creatine Kinase-B
Regulation
1005
A B C D A B C D A B C D
10
20
30
ESTRADIOL (*jg/kg body weight) Fig. 4. Effect of Varying Estrogen Treatment on Rat Uterine CKB mRNA Expression Three groups of five rats received each dose of E2 (ip in saline). Animals were killed 3 h after treatment, and levels of CKB and actin mRNAs (mean ± SE) were determined by slot blot analysis. Autoradiographic results for CKB are shown in the inset (A, 800 ng total RNA; B, 400 ng; C, 200 ng; D, 100 ng). The absolute levels of CKB and actin mRNAs cannot be determined from the figure due to differences in probe specific activities and imperfect homology of the human actin probe to the various rat actin RNAs.
TAM
CON
» i t I DEX
I PROG
I
I DHT
Fig. 5. Hormonal Specificity in the Regulation of Uterine CKB mRNA Levels Three groups of five rats were subjected to each treatment (ip in saline). Animals were killed at 1 h, and 1 /*g uterine RNA from each group (A, B, and C) was probed for CKB mRNA. E2, 30 /ig/kg. CON, Saline control; TAM, tamoxifen (120 ng/ kg); DEX, dexamethasone (120 /xg/kg); PROG, progesterone (120 Mg/kg); DHT, 5a-hydrotestosterone (120 /xg/kg).
the brain. The current study finds no differences between the coding regions of CKB cDNAs derived from the two tissues. It remains possible that CKB is subject to tissue-specific posttranslational modifications, as is seen, for example, in glycosylation of liver and testicular transferrin (33). Human CKB contains sulfonated gly-
cosidic chains (34), and a small variation in sulfonation could account for the difference in pi between uterine and brain CKB, while differences in glycosylation could significantly affect peptide mol wt estimated by gel electrophoresis. Several genomic sequences related to, but distinct from, CKB, CKM, and mitochondrial CK have been isolated, but it is likely that these represent processed pseudogenes, rather than genes for expressed isoforms (11). The time course for accumulation of uterine CKB mRNA approximately parallels the increase in nuclear receptor occupancy (19) and together with earlier data on the sensitivity to actinomycin-D suggests that the effect is due to a selective increase in CKB transcription. Yet this remains to be formally proven by more direct measurements of CKB transcription. The mechanism for the rapid decrease in CKB mRNA in the 3- to 6-h period is not clear and requires studies of both transcription and mRNA turnover. Further work is required to determine whether enhanced transcription continues during this period and whether turnover of the CKB mRNA in the declining phase is enhanced or similar to that in control uteri. Our data indicate that the rapid response of uterine CKB mRNA levels to E2 occurs despite a low sensitivity to the hormone. It is possible
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 September 2016. at 00:39 For personal use only. No other uses without permission. . All rights reserved.
Vol 4 No. 7
MOL ENDO-1990 1006
-917
ctgctggcagctggaggcagggtgtgaacgcctgtgttcacacattccaattcacagccg -858
-677
ggtccctcaggaaagaacctggggatttgaagattcaaaacagtctctaggagctcagtg -618
-617
tcttgaattttttagggtccgggtccaagggtcttggctaggttccttagggcccgccca -558
-378 -318
-1877
caagctgagcaagagacaggagctctgcaggagctctggggacagaacagtggtggcttt
-2 57
tgaagcaggtctagggtatctacggtcctgggatttcgtcttcgagatcctgagcgaqcg -198
-137
ctgggggtgtcgagggttgttgcctcggacaaagcggcggcaccaccccaaagcgcgggr -78
-77 -17
caatqaaatqaatqqqctataaataoccaccaatqqqagqccqacgacqcqcccctt.viq -18 .1 agctcagggagcagcgaGCGGCCGTCGTTCTTCTGCGCTGCGCCCGGAGCTCCAAGCACA 4 2
43
GGCAGTCTGCGTTCCTGCTCCGTCCGGAATCCCGgtgagcgggtcggagggtgaggggct 102
223
cttggtgacgtccgaggaaagcttggggggtccggattggtctgcaggtctctggagacc 282
28 3
ggtgtgtaaagctcctctgatcccgctcttccccgcagCCTGCCGCCGCCGCCGCCGCCG 34 2
34 3
CCATGCCCTTCTCCAACAGCCACAACACGCAGAAGCTGCGCTTTCCGGCCGACGCTGAATTC 404 MetProPheSerAsnSerHisAsnThrGlnLysLeuArgPheProAlaGluAlaGluPhe
-1818
Fig. 6. DNA Sequence of the Promoter and 5' Flanking Sequences of a Rat CKB Gene Numbering is based on the transcriptional start site proposed by Benfield et al. (11). Potential promoter elements are underlined.
pCKBlCAT
pEREKAT
AcC TOTAL DNA 25pg
CAT PLASMIDS 2pg
ERPLASMID 2pg
E R + E 2 (M)o
I0"10 I0' 9 I0"8 2
3
4
+
+
+
I0'10 I0" 9 I0" 8
7
8
9
Fig. 7. Estrogen-Dependent Expression of CAT Sequences Driven by the CKB Promoter or a Chimeric ERE-HSV-TK Promoter HeLa cells were transfected with 2 M g CKB-CAT (pCKBlCAT; tracks 1-5) or 2 M g ERE-TK-CAT (pERETKAT4; tracks 6-10) constructs together with 2 nQ of the ER clone pHEO (except for tracks 5 and 10) and sufficent pBSk- DNA to adjust DNA to 25 ng. After transfection (see Materials and Methods) cultures received E2 as indicated. Protein (15 ng) from pCKBI CAT extracts was assayed for CAT activity in a 2-h 37 C incubation (5 ^g for 1 h with pERETKAT4 extracts). C, Control.
that this relative insensitivity reflects the interaction of the receptor complex with the CKB gene, as a similar low sensitivity is seen on transfection with CAT sequences linked to the CKB promoter and 5' flanking sequences. We propose that the effect is physiologically relevant, despite the low sensitivity to hormone, since previous studies (35, 36) of the rat estrous cycle could detect enhanced synthesis of CKB (as IP) only in proestrus, when circulating estrogen levels are at their highest. The rapidity of the estrogen-induced increase
in uterine CKB mRNA levels compared with the more estrogen-sensitive actin and c-fos (15) mRNAs suggests that the temporal response of specific genes does not directly correlate with sensitivity. Indeed, our results suggest that restriction of the classical IP effect to the uterus may reflect the limited number of tissues with sufficient numbers of ERs to drive transcription on the rat CKB gene. The basal levels of CKB mRNA in control rats probably represent constitutive transcription of this housekeeping gene. Circulating levels of estrogen in
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 September 2016. at 00:39 For personal use only. No other uses without permission. . All rights reserved.
Creatine Kinase-B Regulation
|
1007
60
Brian T. Pentecost, Lori Mattheiss, Herbert W. Dickerman, and S. Anand Kumar* Wadsworth Center for Laboratories and Research New York State Department of Health Albany, New York 12201-0509
Creatine kinase-B (CKB) synthesis is rapidly and specifically induced by estrogen in the uterus of the immature rat. This study indicates that this elevation is due at least in part to increases in the levels of mRNA for CKB. The stimulation of CKB mRNA levels is rapid (a 7- to 10-fold increase is detected 1-3 h after estrogen administration), but transient, as levels return to near control values by 6 h. Analysis of cDNAs to both uterine and brain CKB mRNA indicate that the same sequence is expressed in both tissues despite earlier observations of heterogeneity of the protein isolated from the two tissues. A 1.7-kilobasepair DNA fragment containing the CKB promoter and 5' flanking sequences confers estrogen sensitivity on expression of the bacterial chloramphenicol acetyl transferase gene in HeLa cells on cotransfection with an estrogen-receptor expression vector. However, the CKB promoter sequences lack any motif with convincing similarity to the currently accepted consensus estrogen response element GGTCAnnnTGACC. (Molecular Endocrinology 4: 1000-1010, 1990)
elude: mol wt, as determined by sodium dodecyl sulfatepolyacrylamide gel electrophoresis (brain, 46 kDa; uterus, 49 kDa) (5); isoelectric point, as determined by gel electrofocusing (brain, 6-6.5; uterus, 5.3-5.6) (5); and differential reactivity with several monoclonal antibodies (6). This suggested that different CKB genes might be expressed in estrogen-responsive and nonresponsive tissues or, as reported for chicken CKB (7), that alternative splicing of a common transcript might occur. Another aspect of estrogen regulation of rat uterine CKB gene expression is whether the hormonal effect is direct or requires additional protein synthesis. Extensive studies by Katzenellenbogen and Gorski (8) demonstrated, using uterine organ culture, that the action of estrogen could not be duplicated by other steroids or cAMP, and that the degree of induction of CKB synthesis (measured as IP) correlated with nuclear receptor occupancy (8). Importantly, indirect experiments indicated that synthesis of the IP mRNA was not blocked by cycloheximide, but was abolished by actinomycin-D (9, 8), fulfilling the best criteria then available for establishing direct transcriptional regulation. This suggests that CKB provides a useful model for studies of estrogen regulation of gene expression. We now report on analysis of uterine CKB using molecular probes. We demonstrate that the levels of the CKB mRNA are rapidly but transiently stimulated by estrogen and show that receptor-dependent estrogen-modulated expression of chloramphenicol acetyl transferase (CAT) activity in cell culture can be conferred by the CKB promoter and 5' flanking sequences. Our results also indicate that the differences between brain and uterine CKB protein are not due to variations in the sequences expressed.
INTRODUCTION
Estrogen stimulation of a specific induced protein (IP) before general increases in protein synthesis in the immature rat uterus was originally detected by Notides and Gorski (1). Its identity was unknown until cytosol mixing studies by Reiss and Kaye (2) demonstrated that IP copurified with the B isozyme of creatine kinase (CK; ATP; creatine A/-phosphotransferase, EC 2.7.3.2) from rat brain. Subsequently, an estrogen-inducible increase in the amount of immunoprecipitable creatine kinase-B (CKB), upon in vitro translation of uterine RNA, provided more direct evidence that CKB mRNA levels were stimulated by the hormone (3, 4). Studies from our laboratory on IP purified solely from rat uteri demonstrated that the homogenous uterine protein has CKB activity (5), but that it is not identical to CKB isolated from rat brain. Variant properties in-
RESULTS Similar CKB Sequences Are Expressed in Brain and Uterus Our initial studies addressed the relationship of CKB sequences expressed in rat brain and uteri, to both resolve the discrepancies seen on analysis of the enzyme from the two tissues and ensure that the correct
0888-8809/90/1000-1010$02.00/0 Molecular Endocrinology Copyright © 1990 by The Endocrine Society
1000
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 September 2016. at 00:39 For personal use only. No other uses without permission. . All rights reserved.
Creatine Kinase-B Regulation
gene was later selected for analysis of regulation at the molecular level. A 1300-basepair (bp) uterine CKB cDNA was isolated from a Agt11 library constructed using uterine mRNA of estrogen-treated immature rats and subcloned to give pRUTCKB (Fig. 1). Seven small discrepancies between the uterine cDNA sequence and the published rat brain cDNA sequence (10) were detected; three of these differences were also found on comparison with the published rat CKB gene sequence (11) (Fig. 1). However, we found that our independently isolated rat brain sequence (pCKBBR5) was identical to the uterine sequence at those points, suggesting that the conflicts were due to synthesis artifacts or sequencing errors in the published data (10, 11). Our data eliminate a Hpall site at position 810. Restriction analysis of the cDNA inserts of both pRUTCKB and pCKBBR5 could not detect a Hpall site at 810, confirming our sequence assignment. In addition, the single difference in the coding region affecting the translated product causes rat CKB to become homologous at that point to CK from all other species that have been determined (12). The 5' end of both of our CKB clones was within the coding region determined by Benfield et al. (10), owing to cleavage at an internal EcoRI site during construction. The cDNA insert of pRUTCKB contained no sequences corresponding to a poly-A tract, despite preparation of polyadenylated RNA and priming with polydT primers, presumably owing to incomplete second strand synthesis and trimming by S1 nuclease. The last base of the uterine sequence corresponded to the final transcript-derived base of the published CKB1 brain cDNA (11,10), which is followed by a poly-A tract. Our rat brain cDNA was also polyadenylated, but contained three additional bases (dCdTdA) before the poly-A tract. The extra bases were compatible with the sequence of the published rat CKB genomic clone (11). Heterogeneity of 3' untranslated sequences in dog CKB cDNAs has been proposed as possible evidence of multiple genes; however, we consider that the variation in the rat sequences is probably due to imprecision in processing of the mRNA precursors, as also proposed by other groups (13, 14). The labeled cDNA insert of pRUTCKB hybridized to similarly sized [1.6 kilobases (kb)] RNA species in Northern blots of brain and uterine total RNA (Fig. 2B), with no evidence of multiple forms in either tissue. Uterine CKB Is Specifically Stimulated by 17/3Estradiol (E2) Preliminary Northern blot analysis of RNA from immature female rats (17-19 days old) indicated that significant elevation of CKB mRNA levels occurred after estrogen administration (Fig. 2B). Additional analysis was, therefore, carried out, primarily using slot blots for improved quantification. Significant background levels of the CKB mRNA were detected in RNA from control rats (Figs. 2B and 3A). E2 administration (12 Mg/kg) caused a rapid but tran-
1001
sient increase in CKB mRNA levels detectable 15-30 min after treatment (data not shown). A 7- to 10-fold increase was found after 1 h, and levels peaked in the 1 - to 3-h period before returning to approximately control values by 6 h (Fig. 3A). There was some interexperimental variation in both the timing and magnitude of the response, with levels at 1 h occasionally exceeding those 3 h posttreatment. Levels of actin and /?-tubulin mRNAs were also monitored as controls. Tubulin demonstrated only a limited (~3-fold) and gradual response over the 6-h period (Fig. 2A and our unpublished slot blot analysis), as found by Loose-Mitchell and co-workers (15), while actin mRNA levels showed a definite rise but with a lag period distinctly greater than that for CKB mRNA (Fig. 3B). The results for actin may be complicated by the presence of several isoforms (16,17) detectable by the actin probe used (18). Nevertheless, little increase was found until 3 h after estrogen treatment. Analysis with a ribsomal probe (our unpublished data) indicates that samples from the various groups contain similar amounts of RNA. There was an approximately linear relationship (up to 12 Mg/kg E2) between E2 administered and stimulation of CKB mRNA levels 3 h later (Fig. 4). This pattern of response is similar to that seen for uterine wet weight at 3 h and correlates with nuclear estrogen receptor (ER) occupancy (19). This contrasts with the response of actin mRNA levels, which were significantly elevated 3 h after E2 administration by the lowest dose (0.75 Mg/kg) evaluated (Fig. 4). The apparent requirement for only fractional receptor occupancy to significantly elevate actin mRNA levels is similar to results for glucose oxidation (20) and c-fos mRNA levels (15) in the uterus. Hormonal specificity of the uterine CKB mRNA response was evaluated in groups of 20-day-old female rats. Neither progesterone, dihydrotestosterone, nor dexamethasone (all at 120 Mg/kg) caused noticeable increases in uterine CKB mRNA levels 1 h after administration (Fig. 5). E2 (30 Mg/kg), in contrast, caused a greater than 10fold increase in CKB mRNA levels over control levels in the same experimental series (Fig. 5). Subsequent analysis of the CKB genomic sequence (Fig. 6) failed to identify sequences resembling the hormone response element for nonestrogenic steroids. Tamoxifen, an antiestrogen that can sometimes act as an agonist, neither increased nor decreased uterine CKB mRNA levels at 1 h (Fig. 5). Genomic Clone Isolation Rat genomic sequences encoding the CKB gene were isolated from Charon 4a rat genomic libraries (21) by screening with the labeled insert of pRUTCKB and rescreened with the nonconserved 3' untranslated region of the CKB cDNA to confirm that they were related to CKB rather than CKM. Subcloned fragments predicted to encode the promoter and adjacent sequences
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 September 2016. at 00:39 For personal use only. No other uses without permission. . All rights reserved.
MOL ENDO-1990 1002
Vol 4 No. 7
151
GAATTCCCTGATCTGAGCAGCCACAACAACCATATGGCCAAGGTGCTGACCCCCGAGCTG GluPheProAspLeuSerSerHisAsnAsnHisMetAlaLysValLeuThrProGluLeu
210
211
TACGCGGAGCTCCGTGCCAAGTGCACGCCGAGCGGCTTTACGTTGGACGACGCCATCCAG TyrAlaGluLeuArgAlaLysCysThrProSerGlyPheThrLeuAspAspAlalleGln
270
2 71
ACTGGCGTAGACAATCCGGGCCACCCGTACATCATGACAGTGGGTGCAGTGGCGGGCGAC ThrGlyValAspAsnProGlyHisProTyrlleMetThrValGlyAlaValAlaGlyAsp
330
3 31
GAGGAGAGTTACGACGTATTCAAGGACCTTTTCGACCCCATCATCGAGGACCGGCACGGC GluGluSerTyrAspValPheLysAspLeuPheAspProIlelleGluAspArgHisGly
39 0
391
GGCTACCAGCCCAGTGATGAGCACAAGACTGACCTCAACCCAGACAACCTGCAGGGCGGC GlyTyrGlnProSerAspGluHisLysThrAspLeuAsnProAspAsnLeuGlnGlyGly
4 50
4 51
GATGACCTGGACCCCAACTACGTGCTGAGCTCGCGGGTGCGCACAGGCCGAAGCATCCGC AspAspLeuAspProAsnTyrValLeuSerSerArgValArgThrGlyArgSerlleArg
510
511
GGCTTCTGCCTCCCCCCTCACTGCAGCCGTGGGGAGCGCCGCGCCATCGAGAAGCTGGCA GlyPheCysLeuProProHisCysSerArgGlyGluArgArgAlalleGluLysLeuAla
570
571
GTAGAAGCCCTGTCCAGCCTAGATGGCGACCTGTCTGGCAGGTACTATGCGCTCAAGAGC ValGluAlaLeuSerSerLeuAspGlyAspLeuSerGlyArgTyrTyrAlaLeuLysSer
630
631
ATGACCGAGGCGGAGCAGCAGCAGCTCATTGACGACCACTTCCTCTTCGACAAGCCTGTG GACGAG MetThrGluAlaGluGlnGlnGlnLeuIleAspAspHisPheLeuPheAspLysProVal AspGlu
690
691
TCGCCTCTGCTGCTGGCCTCCGGCATGGCCCGCGACTGGCCGGATGCTCGCGGCATTTGG SerProLeuLeuLeuAlaSerGlyMetAlaArgAspTrpProAspAlaArgGlylleTrp
7 50
751
CACAATGACAATAAGACGTTCCTGGTGTGGATCAACGAGGAGGACCACCTGCGGGTTATC [C] HisAsnAspAsnLysThrPheLeuValTrpIleAsnGluGluAspHisLeuArgVallle
810
811
TCCATGCAGAAAGGGGGCAACATGAAGGAAGTTTTCACGCGATTCTGCACTGGCCTCACT SerMetGlnLysGlyGlyAsnMetLysGluValPheThrArgPheCysThrGlyLeuThr
870
871
CAGATTGAAACTCTCTTCAAGTCTAAGAACTATGAGTTCATGTGGAACCCTCACCTGGGC GlnlleGluThrLeuPheLysSerLysAsnTyrGluPheMetTrpAsnProHisLeuGly
930
931
TACATCCTCACGTGCCCATCCAACCTGGGCACTGGGCTTCGGGCAGGCGTGCACATCAAG [A] TyrlleLeuThrCysProSerAsnLeuGlyThrGlyLeuArgAlaGlyValHisIleLys
990
991
CTGCCCCACCTGGGAAAGCACGAGAAGTTCTCGGAGGTGCTCAAGCGACTGCGGCTTCAG LeuProHisLeuGlyLysHisGluLysPheSerGluValLeuLysArgLeuArgLeuGln
1050
1051
AAGCGAGGCACAGGTGGTGTGGACACCGCTGCTGTGGGTGGAGTTTTTGATGTCTCCAAC LysArgGlyThrGlyGlyValAspThrAlaAlaValGlyGlyValPheAspValSerAsn
1110
1111
GCTGACCGCCTGGGCTTCTCGGAGGTGGAGCTGGTGCAGATGGTGGTGGACGGAGTGAAG
117 0
AlaAspArgLeuGlyPheSerGluValGluLeuValGlnMetValValAspGlyValLys 1171
CTACTCATTGAGATGGAGCAGCGGCTTGAGCAGGGTCAGCCCATTGACGACCTCATGCCT LeuLeuIleGluMetGluGlnArgLeuGluGlnGlyGlnProIleAspAspLeuMetPro
12 30
12 31
GCCCAGAAGTGAAGCC.TGGCCCTAGCCACCACCAGGCTGCCGCTTCCTAACTTATTACC AlaGlnLysEnd [G]
1290
1291
CGGGCAGTGCCCGCCATGCATCCTTGATGTTTGCCGCCGCCTGGCGGCTGAGCCCTTAGC TGCC...GGG
13 50
13 51
CTCGCTGTAGAGACTTCTGTCGCCCTGGGTAGAG..TTTATTTTT.CTGATGGCTAAGCT [CC] T
1410
1411
GTTGCAGACACTGAAATAAATTAGGGTTTGGCCTGCC
Fig. 1. Sequence of a Uterine cDNA to CKB The cDNA nucleotide sequence numbering is based on the proposed transcriptional start site of the gene (11), and the cDNA begins at an internal EcoRI site. Conflicts with published rat brain cDNA (10) and genomic (11) sequences are indicated below the uterine sequence. Conflicts between the uterine cDNA and published brain sequence (10), but not the gene, are in brackets.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 September 2016. at 00:39 For personal use only. No other uses without permission. . All rights reserved.
Creatine Kinase-B Regulation
1003
B
1
1
Fig. 2. Northern Blot Analysis of Uterine and Brain RNA Replicate blots were probed with /?-tubulin (A) and CKB (B) cDNAs. Track 1, Estrogenized brain; track 2, control uteri; track 3, estrogenized uteri. E2 was administered at 30 ^g/kg; controls received vehicle alone. Animals were killed 3 h after injection. Each track contained 5 Mg total RNA resolved in agarose-formaldehyde gels.
were subjected to sequencing, and results indicated that the clone was essentially identical to that analyzed by Benfield and co-workers (11). The sequence (Fig. 6) had the unusual dual promoter structure noted for both the human (22, 23) and rat CKB genes (11). Analysis was, therefore, concentrated on sequences up-stream from the promoter, the most likely region to contain sequences conferring estrogen sensitivity on the CKB gene and for which no data were available beyond - 1 kilobasepair (kbp). A final continuous sequence extending to -2900 bp was established (Fig. 6), as virtually all hormonal response elements in other genes have been found less than 2.5 kbp up-stream of the relevant promoters. Similarity to other CKB genomic sequences was assessed using the GCG Gap program. Similarity to the published rat CKB gene (11) for the region +400 to -900 was in excess of 95% and was 99% on exclusion of the region -540 to -490, which is GC rich and had previously been sequenced in only one orientation, apparently leading to compression artifacts. These values confirm that the CKB gene described is equivalent to that previously reported. Similarity to the human gene was high (90%) in the region from the proposed transcriptional start site out to -120 bp, but then poor (60%) for the remaining available data out to -400 bp (22), with no contiguous blocks of homologous bases greater than 10 bp. The limited human CKB gene sequence was, therefore, uninformative in the search for regulatory sequences outside the proximal promoter region. Studies on the transfected human CKB gene do, however, indicate that this family of sequences is capable of supporting expression of active CKB enzyme (23). The up-stream sequences of the CKB gene (Fig. 6) were screened for c/s-elements resembling the known examples of estrogen response elements (EREs) from
other genes. This was performed in the GCG analysis package using a consensus matrix derived from the EREs from chicken and Xenopus vitellogenins (24, 25) and human pS2 (26) and PRL (27, 28) genes. The analysis was inconclusive, as no potential motifs with less than three mismatches to the perfect palindromic element GGTCAnnnTGACC were detected, whereas the functional examples on which the survey was based have a maximum of two deviations. The best examples from the immediate flank are located at -1150 to -1162, -544 to -556, and -338 to -350. Additional surveys of the body of the rat CKB gene did not give any better candidate response elements, and delineation of active regulatory elements will have to rely on functional assays and receptor binding studies. CKB Promoter Sequences Can Confer Estrogen Sensitivity We have undertaken preliminary studies to demonstrate that DNA sequences associated with the CKB promoter can confer the capacity for regulation by estrogen. Transfections were carried out using the calcium phosphate (29) technique on HeLa cells. This cell line is readily transfected and has high levels of a number of nuclear transcription factors. These include SP1 (30), for which potential binding sites are found adjacent to a possible ERE motif up-stream of the CKB promoter (see Discussion). Human CKB has previously been succesfully expressed from a CKB fragment transfected into HeLa cells (23), but estrogen responsiveness was not evaluated. Cotransfection with pHEO, an ER expression clone (31), was necessary, as HeLa lacks functional ERs. Transfection and cell culture conditions were established using a positive control construct (pEREKAT4) containing a synthetic ERE linked to the herpes simples virus (HSV)-TK promoter regulating CAT expression. CAT expression from pEREKAT4 was enhanced by estrogen (Fig. 7, tracks 6-9) and dependent on cotransfection with the receptor clone (Fig. 7, tracks 9 and 10). Estrogen modulation of CAT expression from pCKBICAT, containing CKB sequences -1700 to +5 linked to CAT sequences of the promoterless pOCAT2 vector (32), depended on the presence of the receptor clone, indicating that estrogen action was mediated via its receptor (Fig. 7, tracks 4 and 5). The maximum degree of enhancement was only moderate (7-fold) compared with that which could be achieved with pEREKAT4 (30-fold; Fig. 7, tracks 6 and 9). We examined the sensitivity to estrogen of CAT expression from pCKBICAT in light of the high quantities of estrogen required to elevate CKB mRNA levels in the rat uterus. Low sensitivity to the hormone was again found, with little stimulation of CAT activity seen below 10~8 M (Fig. 7, tracks 1-4). This was in contrast to results with pEREKAT4, which gave a near-maximal response with concentrations 2 orders of magnitude lower (Fig. 7, tracks 6-9). In an alternative approach to the same
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 September 2016. at 00:39 For personal use only. No other uses without permission. . All rights reserved.
Vol 4 No. 7
MOL ENDO-1990 1004
4000
p3 5 2000 _l
z CO
2
4
6
TIME (hours after estradiol administration) 40 0 0 B
03
|
gp2 gpi *A B C D" IT B C
2000
03
CON
1 hr 3 hr
Hi JMBlilWlHiffiwII M l
6hr
gp3 DWA
B
C
DA
HHRHH
WmWm mHmKmmmwm
TIME (hours after estradiol adminstration) Fig. 3. Temporal Response of Estrogen-Modulated Gene Expression in the Uterus Replicate filters of total RNA from uteri of estrogenized rats (12 ^g/kg) were probed with labeled cDNAs to determine the responses of CKB (A) and actin (B). Each point represents the mean ± SE for three groups of five rats. A series of dilutions (A, 800 ng; B, 400 ng; C, 200 ng; D, 100 ng) was made for each sample; autoradiograms of the data used to construct the graphs are shown in the inset. CON, Control.
question we found that reduction of the amount of cotransfected receptor clone below 1 ^g also reduces CAT expression from pCKBICAT (Fig. 8), again indicating a low sensitivity. In contrast, pEREKAT4 demonstrated the highest activity, with aproximately 0.3 j*g pHEO, and activity declined with further increases in the amount of cotransfected receptor clone.
DISCUSSION Our data support the observation of Kaye and coworkers that a rapidly estrogen induced protein of the rat uterus, first detected by Notides and Gorski (1), is CKB. Previous results from our group suggested that the CKB from the uterus is not identical to that from
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 September 2016. at 00:39 For personal use only. No other uses without permission. . All rights reserved.
Creatine Kinase-B
Regulation
1005
A B C D A B C D A B C D
10
20
30
ESTRADIOL (*jg/kg body weight) Fig. 4. Effect of Varying Estrogen Treatment on Rat Uterine CKB mRNA Expression Three groups of five rats received each dose of E2 (ip in saline). Animals were killed 3 h after treatment, and levels of CKB and actin mRNAs (mean ± SE) were determined by slot blot analysis. Autoradiographic results for CKB are shown in the inset (A, 800 ng total RNA; B, 400 ng; C, 200 ng; D, 100 ng). The absolute levels of CKB and actin mRNAs cannot be determined from the figure due to differences in probe specific activities and imperfect homology of the human actin probe to the various rat actin RNAs.
TAM
CON
» i t I DEX
I PROG
I
I DHT
Fig. 5. Hormonal Specificity in the Regulation of Uterine CKB mRNA Levels Three groups of five rats were subjected to each treatment (ip in saline). Animals were killed at 1 h, and 1 /*g uterine RNA from each group (A, B, and C) was probed for CKB mRNA. E2, 30 /ig/kg. CON, Saline control; TAM, tamoxifen (120 ng/ kg); DEX, dexamethasone (120 /xg/kg); PROG, progesterone (120 Mg/kg); DHT, 5a-hydrotestosterone (120 /xg/kg).
the brain. The current study finds no differences between the coding regions of CKB cDNAs derived from the two tissues. It remains possible that CKB is subject to tissue-specific posttranslational modifications, as is seen, for example, in glycosylation of liver and testicular transferrin (33). Human CKB contains sulfonated gly-
cosidic chains (34), and a small variation in sulfonation could account for the difference in pi between uterine and brain CKB, while differences in glycosylation could significantly affect peptide mol wt estimated by gel electrophoresis. Several genomic sequences related to, but distinct from, CKB, CKM, and mitochondrial CK have been isolated, but it is likely that these represent processed pseudogenes, rather than genes for expressed isoforms (11). The time course for accumulation of uterine CKB mRNA approximately parallels the increase in nuclear receptor occupancy (19) and together with earlier data on the sensitivity to actinomycin-D suggests that the effect is due to a selective increase in CKB transcription. Yet this remains to be formally proven by more direct measurements of CKB transcription. The mechanism for the rapid decrease in CKB mRNA in the 3- to 6-h period is not clear and requires studies of both transcription and mRNA turnover. Further work is required to determine whether enhanced transcription continues during this period and whether turnover of the CKB mRNA in the declining phase is enhanced or similar to that in control uteri. Our data indicate that the rapid response of uterine CKB mRNA levels to E2 occurs despite a low sensitivity to the hormone. It is possible
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 September 2016. at 00:39 For personal use only. No other uses without permission. . All rights reserved.
Vol 4 No. 7
MOL ENDO-1990 1006
-917
ctgctggcagctggaggcagggtgtgaacgcctgtgttcacacattccaattcacagccg -858
-677
ggtccctcaggaaagaacctggggatttgaagattcaaaacagtctctaggagctcagtg -618
-617
tcttgaattttttagggtccgggtccaagggtcttggctaggttccttagggcccgccca -558
-378 -318
-1877
caagctgagcaagagacaggagctctgcaggagctctggggacagaacagtggtggcttt
-2 57
tgaagcaggtctagggtatctacggtcctgggatttcgtcttcgagatcctgagcgaqcg -198
-137
ctgggggtgtcgagggttgttgcctcggacaaagcggcggcaccaccccaaagcgcgggr -78
-77 -17
caatqaaatqaatqqqctataaataoccaccaatqqqagqccqacgacqcqcccctt.viq -18 .1 agctcagggagcagcgaGCGGCCGTCGTTCTTCTGCGCTGCGCCCGGAGCTCCAAGCACA 4 2
43
GGCAGTCTGCGTTCCTGCTCCGTCCGGAATCCCGgtgagcgggtcggagggtgaggggct 102
223
cttggtgacgtccgaggaaagcttggggggtccggattggtctgcaggtctctggagacc 282
28 3
ggtgtgtaaagctcctctgatcccgctcttccccgcagCCTGCCGCCGCCGCCGCCGCCG 34 2
34 3
CCATGCCCTTCTCCAACAGCCACAACACGCAGAAGCTGCGCTTTCCGGCCGACGCTGAATTC 404 MetProPheSerAsnSerHisAsnThrGlnLysLeuArgPheProAlaGluAlaGluPhe
-1818
Fig. 6. DNA Sequence of the Promoter and 5' Flanking Sequences of a Rat CKB Gene Numbering is based on the transcriptional start site proposed by Benfield et al. (11). Potential promoter elements are underlined.
pCKBlCAT
pEREKAT
AcC TOTAL DNA 25pg
CAT PLASMIDS 2pg
ERPLASMID 2pg
E R + E 2 (M)o
I0"10 I0' 9 I0"8 2
3
4
+
+
+
I0'10 I0" 9 I0" 8
7
8
9
Fig. 7. Estrogen-Dependent Expression of CAT Sequences Driven by the CKB Promoter or a Chimeric ERE-HSV-TK Promoter HeLa cells were transfected with 2 M g CKB-CAT (pCKBlCAT; tracks 1-5) or 2 M g ERE-TK-CAT (pERETKAT4; tracks 6-10) constructs together with 2 nQ of the ER clone pHEO (except for tracks 5 and 10) and sufficent pBSk- DNA to adjust DNA to 25 ng. After transfection (see Materials and Methods) cultures received E2 as indicated. Protein (15 ng) from pCKBI CAT extracts was assayed for CAT activity in a 2-h 37 C incubation (5 ^g for 1 h with pERETKAT4 extracts). C, Control.
that this relative insensitivity reflects the interaction of the receptor complex with the CKB gene, as a similar low sensitivity is seen on transfection with CAT sequences linked to the CKB promoter and 5' flanking sequences. We propose that the effect is physiologically relevant, despite the low sensitivity to hormone, since previous studies (35, 36) of the rat estrous cycle could detect enhanced synthesis of CKB (as IP) only in proestrus, when circulating estrogen levels are at their highest. The rapidity of the estrogen-induced increase
in uterine CKB mRNA levels compared with the more estrogen-sensitive actin and c-fos (15) mRNAs suggests that the temporal response of specific genes does not directly correlate with sensitivity. Indeed, our results suggest that restriction of the classical IP effect to the uterus may reflect the limited number of tissues with sufficient numbers of ERs to drive transcription on the rat CKB gene. The basal levels of CKB mRNA in control rats probably represent constitutive transcription of this housekeeping gene. Circulating levels of estrogen in
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 September 2016. at 00:39 For personal use only. No other uses without permission. . All rights reserved.
Creatine Kinase-B Regulation
|
1007
60
