Expression and Regulation of the Rabbit Uteroglobin Gene in Transgenic Mice
Francesco J. DeMayo, Sami Damak, Thomas N. Hansen, and David W. Bullock Departments of Cell Biology (F.J.D.) and Pediatrics (T.N.H.) Baylor College of Medicine Houston, Texas 77030 Department of Biochemistry and Microbiology Lincoln University (S.D., D.W.B.) Canterbury, New Zealand
INTRODUCTION
The rabbit uteroglobin (UG) gene, with varying lengths of 5' flanking sequence, was introduced into the mouse genome to investigate the DNA sequences required for tissue-specific expression and regulation by steroid hormones. The pattern of expression and steroid hormone regulation of the transgene was compared to the expression and regulation of the endogenous mouse UG-like gene. In the rabbit, UG is induced in the uterus by progesterone and is expressed constitutively in the lungs, where it is weakly regulated by glucocorticoids. Genomic DNA fragments containing the complete UG-coding sequence with 4.0 (UG4.0), 3.0 (UG3.0), 2.3 (UG2.3), or 0.6 (UG0.6) kilobases of 5' flanking sequence were used to establish lines of transgenic mice. Expression of UG mRNA was observed in the lungs of UG4.0 (2/4 lines), UG3.0 (4/4 lines), UG2.3 (1/2 lines), and UG0.6 (4/4 lines) mice. Uterine expression was observed in UG3.0 (3/4 lines), UG2.3 (1/2 lines), and UG0.6 (2/4 lines). In the lungs of UG3.0 and UG2.3 mice, RNA expression was stimulated by treatment with dexamethasone. In the one line of UG3.0 mice examined, UG was regulated by ovarian steroids in the uterus. The endogenous mouse UG-like gene showed the major site of expression to be in the lung. Unlike the transgene, the endogenous gene was more strongly stimulated by glucocorticoids. Thus, we conclude that the cis elements needed for pulmonary expression of UG are contained within the UG2.3 fragment used to generate transgenic mice, but that other elements are required for full glucocorticoid regulation. Also, the transgene did not show the full uterine expression observed in the rabbit, but regulation by the ovarian steroids was observed. (Molecular Endocrinology 5: 311-318,1991)
Uteroglobin (UG) is a secretory protein of the male and female reproductive tract and pulmonary epithelium in rabbits (1, 2). UG was initially named blastokinin, because the timing of expression of this protein implied a potential role in implantation in the rabbit (3). UG binds progesterone specifically (4). Recent comparisons of its amino acid sequence have revealed homology with inhibitors of phospholipase-A2, and an antiinflammatory role has been proposed (5). Initially, UG was thought to be a protein unique to the rabbit; however, UG-like proteins have been reported in rat prostate (6) and lungs (7) and in human uterus (8) and prostate (9). Recently, the existence of homologous UG-like proteins in humans and rats was confirmed by the cloning and sequence analysis of the human 10-kDa Clara cell protein (10) and the rat polychlorobiphenyl-binding protein (pcbb) (11, 12). These proteins showed 61 % and 53% amino acid homology to the rabbit protein, respectively. Although these genes show sequence homology, the tissue distribution of their expression varies between species. Unlike the rabbit, the major site of expression of the rat pcbb gene is in the lung, with minor expression in the female reproductive tract (12). Although the physiological role of UG has not yet been defined, its regulation by steroid hormones has made it a useful model for hormonal control of mammalian gene expression. The effect varies with the steroid and the tissue in which this gene is expressed. In the female reproductive tract, UG is under the predominant influence of progesterone (13, 14), while in the lung its expression is constitutive (15) or weakly stimulated by glucocorticoids (16, 17). In the male reproductive tract, there is evidence that UG is regulated by androgens (18). The rabbit UG protein is encoded in a single copy gene that consists of three exons separated by two introns (19). Analysis of the 5' flanking region by DNAse hypersensitivity studies and steroid
0888-8809/91 /0311 -0318$03.00/0 Molecular Endocrinology Copyright© 1991 by The Endocrine Society
311
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 16:40 For personal use only. No other uses without permission. . All rights reserved.
Vol 5 No. 3
MOL ENDO-1991 312
receptor binding has shown that the putative glucocorticoid- and progesterone-responsive elements are located approximately 2.6 kilobases (kb) from the start of transcription (20, 21). Transfection of a UG-driven bacterial chloramphenicol acteyl transferase gene into Ishikawa cells has also identified estrogen-responsive regulatory elements 252 and 265 nucleotides 5' to the UG gene (22). Determination of the cis elements regulating tissuespecific expression and steroid regulation of the UG gene would shed light on whether the different expression of the UG-like genes in different species is due to a divergence in the cis elements between species or a divergence in the transcription factors required for expression; it would also allow a functional determination of the steroid regulatory elements. Identification of these elements has been impeded by the lack of a suitable cell culture system (23). In this paper we report the use of transgenic mice to overcome this problem. Introduction of foreign genes into the murine genome has allowed evaluation of the enhancer elements needed for the appropriate tissue-specific expression of several genes (24). The rat elastase (25) and the human /3-globin genes (26) have been studied extensively in this way. In the globin gene, the domains required for expression of the transgene independent of the site of integration have been determined (27). This report describes the use of the transgenic mouse to determine the site and regulation of the expression of the rabbit UG gene in murine tissues compared to those of the endogenous UG-like gene.
RESULTS Expression of UG mRNA in Transgenic Mice The fragments of the UG gene used to generate transgenic mice are shown in Fig. 1. Microinjection of these fragments of DNA resulted in four lines of UG4.0, four lines of UG3.0, two lines of UG2.3, and four lines of UG0.6. The number of copies of the transgene ranged from 1 to greater than 10 copies. Figure 2B shows the results of Northern analysis of RNA expression in one
Fig. 1. Organization of the Rabbit UG Gene and Restriction Fragments Used to Generate Transgenic Mice The wavy line indicates vector sequence, the thin line indicates rabbit flanking sequence, and the boxes indicates exons (•) and introns (•). Restriction enzyme sites are C/al (C), AsuW (A), Nae\ (N), Xba\ (X), and Sac\ (S).
28S
18S
•
Fig. 2. Tissue Distribution of the Endogenous Mouse UG-Like Gene in Nontransgenic Mice (A) and the Rabbit UG Gene in Transgenic Mice (B) Northern blots of total tissue RNA (20 ^g) were probed with an oligonucleotide probe derived from the rat pcbb cDNA sequence (A; see text) using the rat lung as a control and a rabbit uteroglobin genomic fragment using rabbit lung and nontransgenic mouse lung as control (B). Sal Gld, Salivary gland.
mouse from line UG 3.0. Rabbit lung RNA served as a positive control, and nontransgenic mouse lung RNA served as a negative control. Rabbit lung shows the 600-nucleotide (nt) band of mature rabbit UG mRNA (19), as well as the unprocessed 3-kb heteronuclear UG mRNA. Under these stringencies no hybridization was observed in the negative mouse lung RNA. This transgenic mouse line expressed UG mRNA of the correct size in lung and uterus, but not in any other tissue examined. All lines of mice containing the UG gene were tested for UG expression in lung and uterus by Northern analysis of 20 ^g total tissue RNA. No UG expression was observed in lungs or uteri of two lines with UG4.0 and one line with UG2.3. In all other lines some UG message could be detected. Table 1 summarizes the tissue distribution of UG expression in transgenic mice. The predominant tissue for expression of the UG gene is the lung. Expression in the female reproductive tract was low and variable. One explanation for this low expression could be that the analysis was conducted without re-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 16:40 For personal use only. No other uses without permission. . All rights reserved.
313
Uteroglobin in Transgenic Mice
Table 1. Tissue Distribution of the Expression of the Rabbit UG Gene in Transgenic Mice FO
Strain
UG4.0
43 7362
B F
UG3.0
5057 5058 5073 5074
F LJ. LJ.
Fragment
Ov
Ut
Te
SV
N
N
Sp
Li
Ki
N N N _
N -
-
_
+ _
Ht
Lu
Ty
SG
N N N _
— N N _
Br
F
159
B
UG0.6
2137 2141 2270 3970
F F LJ.
UG2.3
F
—
—
+
+
N N
N N
N N N _
N N N _
N N N
+ ++ +
_
+
The level of expression is graded, - for no detectable expression, + to + + + + for increasing levels of hybridization signal after Northern analysis, and N for not assayed. Ov, Ovary; Ut, uterus; Te, testis; SV, seminal vesicle; Sp, spleen; Li, liver; Ki, kidney; P, pancreas; M, muscle; Ht, heart; Lu, lung; Ty, thymus; SG, salivary gland; Br, brain. Mice were generated from embryos produced by ICR females bred to either FVB (F) or B6C3F1 (B) male mice. Lines were then established in an ICR background.
gard to the stage of the estrous cycle. To test for stimulation of UG expression, mice were administered the regimen of estrogen and progesterone described by Finn and Publicover (28). UG expression was not detected in uteri from these steroid-stimulated mice (data not shown). Thus, the stage of the estrous cycle is unlikely to influence the ability of the mouse uterus to express UG. The only other sites of expression in transgenic mice were the kidney, testis, salivary gland, seminal vesicles, and thymus, where the frequency and level of expression were low. Figure 2A also shows the distribution of expression of the endogenous mouse homolog to the rabbit UG gene, with rat lung RNA serving as a control. Twenty micrograms of total RNA were isolated from the tissues of nontransgenic mice and probed with a 36-mer oligonucleotide probe chosen from the rat pcbb protein, which showed a high degree of homology to the rabbit UG gene and the human homolog (see Materials and Methods). The predominant site of expression of the mouse homolog, like the transgene, is the lung. Reproductive tract expression was not observed, but the lack of detection of the mouse message could be due to a difference in the sensitivity of probing with an oligonucleotide probe. No expression was observed in the testis or male reproductive tract (data not shown). The only other site of expression of the mouse UG-like gene was the thymus (Fig. 2A). Figure 3 shows the expression of UG mRNA in lungs from representatives of all of the expressing lines of mice. The level of expression was highest in UG3.0 and UG2.3 mice, exceeding that in rabbit lung. The expression of UG in the two lines of UG4.0 mice was lower than that in the rabbit. The expression of UG in the lungs of UG0.6 mice varied with the line of transgenic mouse. A comparison of the expression of UG mRNA in uterine RNA is shown in Fig. 4. These data show that
CM
(o
00 '"J"
io s
co eo o o o t s in io m
28S—
K *- O O o> co ^- r*. f^ IO t- »- N O) »- CM CM CM CO
18S—
4.0
3.0
2.3
0.6
Fig. 3. Analysis of Pulmonary Expression of UG mRNA in Transgenic Mice by Northern Hybridization Expression in the lungs of the mice from line 2270 was at too low a level to be observed in this exposure of this autoradiograph. Fragments UG4.0, UG3.0, UG2.3, and UG0.6 are refered to as 4.0, 3.0, 2.3, and 0.6, respectively.
UG mRNA expression in the mice expressing the transgene is much lower than that in the 3-day pregnant rabbit. Only in line 5074 (UG3.0) was there significant expression in the transgenic mouse uterus. Line 5073 was examined later and found to have a level of UG mRNA similiar to that observed in line 5074 (Table 1). Steroid Regulation Glucocorticoid regulation of UG in the lung was investigated by comparing adrenalectomized mice given glucocorticoids with adrenalectomized mice given PBS. Figure 5 shows the slot blot analysis of the response of UG mRNA expression for mice for each gene fragment and a graphical representation of the mean response of UG message at the three levels of RNA
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 16:40 For personal use only. No other uses without permission. . All rights reserved.
MOL ENDO-1991 314
X) .0 CB
0) (0 3 O
Vol 5 No. 3
CM |s»
CO CO O
CO
Tf
(
D
^
T
-
O
O
O O l f l B r r
CMO)
I O I O I - S M N
M(0
28S
18S
4.0
3.0
2.3
0.6
Fig. 4. Analysis of Uterine Expression of UG mRNA in Transgenic Mice by Northern Hybridization Uterine expression of UG mRNA was at too low a level in the mouse from line 2141 to be observed in this exposure of this autoradiograph. Fragments UG4.0, UG3.0, UG2.3, and UG0.6 are refered to as 4.0, 3.0, 2.3, and 0.6, respectively.
c.
RNA Loading (|ig)
the steady state level of UG expression between gene fragments (P < 0.01). The steady state level of UG expression was significantly increased by treatment with dexamethasone in lines with the UG3.0 and UG2.3 fragments, but not in lines with UG4.0 and UG0.6 fragments (P = 0.017). To test whether the low level of dexamethasone stimulation of the UG transgene may reflect a lack of response of mouse lung to glucocorticoids, nontransgenic adrenalectomized mice were given the same regimen of hormone or placebo, and lung RNA was probed with the rat pcbb oligonucleotide. Figure 6 shows that dexamethasone gave a greater than 2-fold stimulation of the endogenous gene message. Thus, the mouse lung is sensitive to glucocorticoids, but the endogenous mouse gene is influenced more strongly than that of the transgene. Uterine expression of UG was extremely low and variable in the mice bearing the UG transgene. Only two lines, 5073 and 5074 (UG 3.0), expressed the transgene at a level at which regulation could be tested. Mice from line 5074 were administered either oil or the regimen of estrogen and progesterone described by Finn and Publicover (28). Figure 7 shows the results of Northern analysis of the UG transgene by ovarian steroids. The
- 400 E a o ~ 300
E 200 O o 100
1 2 RNA Loading (pg)
4
Fig. 5. Panel 1, Quantitation of UG Expression in Lungs of Transgenic Adrenalectomized Mice Given Dexamethasone (DEX; • — • ) or PBS (O- -O) Carrying UG Fragments UG4.0 (A), UG3.0 (B), UG2.3 (C), and UG0.6 (D) Panel 2 shows the slot blot analysis of 0.1, 0.5, and 1 ^g RNA from pairs of mice (a, b, and c) given PBS or dexamethasone and probed with the rabbit UG gene-specific probe. Panel 3 shows a duplicate slot blot probed for mouse /S-actin. The UG fragments in panels 2 and 3, as in panel 1, are described as UG4.0 (A), UG3.0 (B), UG 2.3 (C), or UG0.6 (D).
loaded onto the slot blot. The data were analyzed using the multiple regression package of the program Statistix (NH Analytical Software, Roseville, MN) running on a personal computer (see Materials and Methods). The analysis revealed a statistically significant difference in
Fig. 6. A, Quantitation of Expression of the Mouse UG-Like Gene in Adrenalectomized Nontransgenic Mice Given Dexamethasone (DEX; • - • ) or PBS (O- -O) B, Slot blot analysis of 1, 2, and 4 ^g total RNA extracted from lung probed with the oligonucleotide probe derived from the rat pcbb cDNA (see text) probe for the mouse UG-like gene. C, A duplicate blot probed for mouse /3-actin. The male mice represent two pairs of siblings from two separate litters.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 16:40 For personal use only. No other uses without permission. . All rights reserved.
315
Uteroglobin in Transgenic Mice
Oi
E+P
28S18S600
NT-
Fig. 7. Northern Analysis of 20 M9 Total Uterine RNA from UG3.0 Mice from Line 5074 Probed with the Rabbit UGSpecific Probe Four littermates were given sesame oil (no. 3503 and 3504) or a regimen of estradiol (E) and progesterone (P) (no. 3494 and 3501). Day 3 pregnant rabbit uterine RNA is shown as a control.
transgene message levels were approximately 2.5-fold greater with the administration of estrogen and progesterone than those in the mice given a placebo. Regulation of the endogenous gene was also examined by treating nontransgenic mice with the same regimen of hormones or placebo. Endogenous message in the reproductive tract could not be detected in either case.
DISCUSSION
The UG gene is expressed maximally in the rabbit uterus during early pregnancy (29). Expression of UG in the rabbit lung is low compared to that in the stimulated uterus and does not vary with pregnancy (30). The pattern of expression of the homologous rat protein is quite different, with the major site of expression being the lungs and only minor expression in the uterus (12). This report shows that the pattern of expression of the mouse homolog to the rabbit UG gene follows the pattern observed in the rat and not that in the rabbit. Also, expression of the rabbit UG gene in the mouse follows the pattern observed for the mouse UG-like gene. This would imply that the species difference in tissue distribution of UG-related genes may be due to a difference in the nature or abundance of transcription factors needed for the appropriate level of expression observed in the different tissues. While these experiments do not distinguish between transcriptional and posttranscriptional effects, evidence in the rabbit indicates that the UG level is regulated at the level of transcription (13-15). All rabbit UG fragments used to generate lines of transgenic mice showed the ability to express the transgene message in the lung. The frequency and level of expression, however, varied with the amount of 5'
sequence flanking the UG gene. The fragments with 3.0 and 2.3 kb of 5' flanking DNA exhibited the highest level of expression, surpassing the level in rabbit lung. The mice with 4.0 kb of 5' flanking DNA (UG4.0) expressed the UG gene at a lower level and a reduced frequency. This observation can be explained by the presence of 3.2 kb of Charon 4A sequences in this construct (Fig. 1), as prokaryotic DNA sequences have been shown to inhibit the expression of /3-globin (26) and a-fetoprotein (31) transgenes in mice. The inhibition of transgene expression by prokaryotic sequences is not universal. For example, expression of the rat elastase gene was not affected when microinjected with 4.3 kb of pBR322 sequences (32). Deletion of UG 5' flanking sequences to -0.6 kb gave increased variability and an overall reduction in the level of expression. Although this demonstrates that the proximal promoter elements needed for pulmonary expression of UG are contained within the 0.6-kb fragment, sequences between -2.3 and -0.6 kb are important for the full pulmonary expression of UG. Pulmonary expression of UG mRNA was increased significantly (P = 0.017) by treatment with dexamethasone. As found in the rabbit (15-17), this effect was small compared to the stimulation caused by progesterone in the rabbit uterus. Stimulation of pulmonary expression of the transgene by dexamethasone was also less than that of the endogenous gene. The effect of dexamethasone depended on the gene construct. No effect was seen with the UG4.0 fragment, probably due to the 3.2 kb of phage sequences, as discussed above. When the 5' flanking sequence was truncated to 0.6 kb, the effect of dexamethasone was lost. Putative sites for the glucorticoid response elements, as found by DNase-1 hypersensitivity and steroid-binding studies, are located in the region of -2.6 kb (20, 21). This putative glucocorticoid response element is present in UG3.0, but not UG2.3, yet both of these fragments were stimulated by dexamethasone. Thus, additional hormone response elements must be present between the A/ael and Xba\ sites in the 5' flanking region of the rabbit UG gene. Although the uterus is the major site of UG expression in the rabbit, expression of this gene in the uteri of transgenic mice was infrequent and could not be correlated with the amount of 5' flanking DNA. While the UG3.0 fragment contained all regions that have been implicated as important in uterine expression in the rabbit (20,21), this fragment was not sufficient to obtain a comparable level of expression in the mouse uterus. The extremely low levels of expression of the transgene in the uterus made it difficult to investigate stimulation of the transgene by ovarian steroids. However, in the line expressing the transgene at an adequate level, regulation of the UG message by ovarian steroids was observed. Additional elements surrounding the UG gene may be necessary for full expression in the uterus. The existence of tissue-specific enhancers at a considerable distance from the promoter is not unprecedented. The albumin gene requires an enhancer 10 kb up-stream
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 16:40 For personal use only. No other uses without permission. . All rights reserved.
Vol 5 No. 3
MOL ENDO-1991 316
from its promoter (33), and in the globin gene enhancers are found 50 kb removed from the gene (27). Alternatively, frans-acting factors in the rabbit uterus specific for the UG promoter (34) may be different or absent in the mouse uterus. This possibility is supported by the known differences in hormonal stimulation required for the establishment of pregnancy in mice and rabbits.
killed to assay for expression in various tissues by Northern hybridization (41). Total tissue RNA was isolated by homogenization in guanidine thiocyanate and separation on a gradient of CsCI2 (42). RNA was denatured and separated by size using formaldehyde-gel electrophoresis. The RNA was transferred to nitrocellulose membranes and probed for the presence of UG mRNA using the 32P-labeled probe described above. Steroid Regulation
MATERIALS AND METHODS
DNA Constructs The DNA used to generate transgenic mice was derived from pUG11.8, a plasmid consisting of an 11.8-kb genomic DNA fragment cloned in the C/al site of pBR322 (19). The 11.8-kb fragment contains the coding region of the rabbit UG gene, with 4.0 kb of 5' flanking sequence and 1.0 kb of 3' flanking sequence. The insert also contains 3.2 kb of Charon 4A DNA at the 5' end and 0.2 kb of Charon 4A sequence at the 3' end of the rabbit DNA sequence. The initial fragment incorporated into the murine genome was the 11.8-kb C/al fragment (UG4.0). Deletions of the 5' flanking sequence were generated by restriction endonuclease digestion of pUG11.8 with AsuW, A/ael, and Xba\ to produce constructs with 3.0 (UG3.0), 2.3 (UG2.3), and 0.6 (UG0.6) kb of 5' flanking sequence, respectively (Fig. 1). All injected fragments contained 0.2 kb of Charon 4A sequence at the 3' end. Generation of Transgenic Mice The DNA to be injected was digested with the appropriate restriction endonuclease and separated from pBR322 sequence by gel electrophoresis using low melting point agarose (Seaplaque, FMC Bioproducts, Rockland, ME). The appropriate fragment was excised, and the DNA was isolated using GeneClean (BIO 101, La Jolla, CA) and diluted to 2 Mg/ml in TE buffer [10 mM Tris (pH 7.5) and 0.25 mM EDTA]. Transgenic mice were produced as described by Hogan et al. (35), with the following modifications. Manipulation of mouse embryos was carried out in medium HT6 (36) supplemented with 5 mg/ ml cytochalasin-B (Sigma, St. Louis, MO) while the embryos were being injected. The culture medium contained 2 Mg/ml of the DNA being injected in order to avoid dilution in case of back-flow into the injection needle during microinjection. The embryos used for microinjection are summarized in Table 1. Integration and Expression Analysis DNA was isolated from tail biopsies taken after weaning of mice born from transferred embryos and was screened by Southern hybridization (37). The isolated DNA (10 ^g) was digested with EcoRI, subjected to agarose gel electrophoresis, and transferred to a Zeta Probe (Bio-Rad, Richmond, CA) membrane under alkaline conditions (38). The 1.4-kb Sacl fragment of pUG11.8, which contains the second and third exons of the UG gene, was labeled with [32P]dCTP (ICN, Irvine, CA) by random priming (39) and used as a hybridization probe under the conditions described by Church et al. (40). Transgenic mice were identified by the presence of the 1.1-kb internal EcoRI fragment. The number of copies of the transgene was estimated by comparison with 1,5, and 10 equivalent copies/haploid genome of pUG11.8 added to normal mouse DNA. Each Southern blot contained nontransgenic mouse DNA as a control. Founder transgenic (Fo) mice were outbred to ICR mice to establish transgenic lines for each integration. Once inheritance was established, one female mouse from each line was
Dexamethasone (Elkins-Sinn, Inc., Cherry Hill, NJ) was administered to intact or adrenalectomized transgenic mice, and UG expression was compared to intact or adrenalectomized mice given PBS. The dexamethasone was given at a dose of 25 /*g, ip, for five injections at 12-h intervals. The mice were killed 6 h after the last injection, and the lungs were taken for RNA analysis. From one transgenic line of each gene construct, three pairs of littermates were allocated to treatment, except for UG0.6, for which only two pairs were used. One member of each pair received dexamethasone, and the other received PBS. Total RNA was applied to nitrocellulose membrane, using a slot blot apparatus (Schleicher and Schuell, Keene, NH). For each mouse, RNA was applied to duplicate slots at three levels (0.1, 0.5, and 1.0 ng). Where necessary, yeast tRNA was added to bring the amount of RNA loaded to 1.0 Mg- One slot was hybridized with 32P-labeled UG RNA transcribed from UG cDNA (43) cloned in pGEM4 (Promega, Madison, Wl) using Sp6 RNA polymerase. The duplicate slot was hybridized with mouse /3actin RNA, prepared similarly from mouse /?-actin cDNA (44) in plasmid cloned pGEM3, and served as a control for uniformity of RNA loading. The blots were hybridized as described by Melton et al. (45). The blots were washed at 68 C in 0.1% sodium dodecyl sulfate and 0.1% SSC (0.015 M NaCI, 0.015 M Na citrate) for 1 h and subsequently treated with RNase. After washing, the membrane was placed against x-ray film and exposed overnight. The slots on the membrane were oriented using the autoradiograph and cut out individually for quantitation of radioactivity by liquid scintillation counting. Expression of the mouse homologous mouse gene was assayed using the oligonucleotide GATCTTCTCCGTGAGCTTCACGATGTTTATTCTGGT, which was kinased with [7-32P]ATP (ICN). This oligonucleotide corresponds to the opposite strand of the rat pcbb cDNA from nucletoides 274310. This area was chosen because it showed a high degree of homology with the rabbit and human cDNA sequences. The Northerns and slot blots were hybridized using the protocol described by Zeff and co-workers (46) with the following modifications. 1) The hybridization temperature was 50 C. 2) The blots were washed for 45 min. 3) Total tissue RNA was analyzed by using 20 ng RNA. Glucocorticoid regulation was conducted using four pairs of male mice produced from a back-cross of ICR x FvB to ICR mice. The mice were adrenalectomized and given the same regimen of hormone or placebo as described above. Slot blot analysis was conducted using 1, 2, and 4 ng total RNA. Regulation of the endogenous mouse gene and the transgene in the uterus by ovarian steroids was investigated using the regimen of estrogen and progesterone described by Finn and Publicover (28). Female litermates were given 100 ng estradiol, sc, for 2 days. Three days after the last estradiol treatment, mice were given 500 ^g progesterone, sc. The mice were then given 500 ng progesterone with 20 ng estradiol, sc, for 2 days. The mice were killed 6 h after the last injection, and total RNA was isolated from the uteri. Control mice were given sesame oil in place of steroid treatment. Northern analysis, using the UG probe for transgenic mice and the pcbb probe for nontransgenic mice, was performed to avoid any problems with background signals due to the low level of uterine UG expression. The loading of RNA was monitored by examining the ethidium bromide staining of the filter after transfer. The fold induction was determined by densitometric
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 16:40 For personal use only. No other uses without permission. . All rights reserved.
317
Uteroglobin in Transgenic Mice
scanning of the autoradiograph using a Kontes Fiber Optic Scanner (Vineland, NJ) and a Hewlett-Packard 3390A Integrator (Houston, TX).
6.
Statistical Analysis The slot blot data were analyzed by multiple regression, using the program Statistix (NH Analytical Software, Roseville, MN) running on a personal computer. Variables were coded as follows: UG expression (Y); construct (X1): UG4.0 = 1, UG3.0 = 2; UG2.3 = 3; UG0.6 = 4; treatment (X2): PBS = 1; dexamethasone = 2; RNA loading (X3): 0.1 ^g = 1- 0.5 ^g = 2,1 ng = 3; replicate (X4); actin expression (X5). Linear models were fitted to the data by stepwise regression, and variables contributing coefficients with t values significant at P < 0.05 were selected. Stepwise regression revealed that actin expression and second or third order interactions between variables contributed little to partitioning of the overall variance. The fitted equation Y = 1962.2X1 - 379.9X12 + 248.1X2 + 552.6X3 3106.9 accounted for 70.5% of the variance. The t value and probabilities for each of the coefficients were 7.71 (P < 0.01), -7.39 (P < 0.01), 2.45 (P < 0.0165), 8.91 (P < 0.01), and -7.89 (P < 0.01), respectively.
7. 8.
9. 10.
11.
12.
Experimental Animals All experiments were conducted using the highest standard for humane care in accordance with the NIH Guide for the Care and Use of Laboratory Animals.
Acknowledgments We thank Ms. Louise Ann Stanley, Mr. Li Li, Ms. Serlina Robinson, and Ms. Celeste Moore for their technical assistance, and Mr. Bruce Robson for statistical advice. We are grateful to Ms. Lisa Gamble and Ms. Beryl Bond for secretarial assistance.
13. 14. 15.
16. 17.
18. Received August 31, 1990. Revision received December 14,1990. Accepted December 14,1990. Address requests for reprints to: Dr. Francesco J. DeMayo, Department of Cell Biology, Baylor College of Medicine, Houston, Texas 77030. This work was supported in part by NIH Grant HD-093778 and the New Zealand Medical Research Council.
REFERENCES 1. Cato ACB, Beato M 1985 The hormonal regulation of uteroglobin gene expression. Anticancer Res 5:65-72 2. Miele L, Cordella-Miele E, Mukherjee AB 1987 Uteroglobin: structure, molecular biology and new perspectives on its functions as a phospholipase A2 inhibitor. Endocr Rev 8:474-490 3. Krishnan RS, Daniel JC 1967 Blastokinin: inducer and regulator of blastocyst development in rabbit uterus. Science 158:490-492 4. Beato M, Baier R 1975 Binding of progesterone to the proteins of the uterine luminal fluid, identification of uteroglobin as the binding protein. Biochim Biophys Acta 392:346-356 5. Miele L, Cordella-Miele E, Facchiano A, Mukherjee AB 1988 Novel anti-inflamatory peptides from the region of
19.
20.
21.
22.
23.
24. 25.
26.
highest similarity between uteroglobin and lipocortin I. Nature 335:726-730 Aunukker G, Seitz J, Heyns W, Kircher C 1985 Ultrastructural localization of uteroglobin immunoreactivity in rabbit lung and endometrium, and rat ventral prostate. Histochemistry 83:413-417 Lopez de Haro MS, Alvarez L, Nieto A 1988 Evidence for the identity of anti-proteinase protein CCSP and uteroglobin. FEBS Lett 232:351-353 Cowan BD, North DH, Whitworth NS, Fujita R, Shumacher EK, Mukherjee AB 1986 Identification of a uteroglobinlike antigen in human uterine like washings. Fertil Steril 45:820-823 Manyak MJ, Kikikawa T, Mukerjee AB 1988 Expression of a uteroglobin-like protein in human prostate. J Urol 140:176-182 Signh G, Katyal SL, Brown WE, Philips S, Kenedy AL, Anthony J, Squelia N 1990 Amino acid and nucleotide sequence of human Clara cell 10 kDa protein. Biochim Biophys Acta 950:329-337 Nordlund-Moller L, Anderson O, Ahlgren R, Schilling J Gillner, M, Gustaffson J-A, Lund J 1990 Cloning, structure and expression of a rat protein for polychlorinated biphenyls. J Biol Chem 21:12690-12693 Hagen G, Wolf M, Katyal SL, Singh G, Beato M, Suske G 1990 Tissue specific expression and 5' flanking gene region of the rat Clara cell 10 kDa protein: comparison to rabbit uteroglobin. Nucleic Acids Res 18:2939-2945 Heins B, Beato M 1981 Hormonal control of uteroglobin secretion and preuteroglobin mRNA content in rabbit endometrium. Mol Cell Endocrinol 21:139-150 Shen X-Z, Tsai M-J, Bullock DW, Woo SLC 1983 Hormonal regulation of rabbit uteroglobin gene transcription. Endocrinology 112:871-875 Savouret J-F, Loosfelt H, Atger M, Milgrom E 1980 Differential hormonal control of a messenger RNA in two tissues: uteroglobin mRNA in the lung and the endometrium. J Biol Chem 255:4131-4136 Lombardero M, Nieto A 1981 Glucocorticoid and developmental regulation of uteroglobin synthesis in rabbit lung. Biochem J 200:487-494 Torkelli T, Krusius T Janne O 1978 Uterine and lung uteroglobins in the rabbit. Biochim Biophys Acta 544:578592 Lopez de Haro MS, Alvarez L, Nieto A1988 Testosterone induces the expression of the uteroglobin gene in rabbit epididymus. Biochem J 250:647-651 Snead R, Day L, Chandra T, Mace M, Bullock DW, Woo SLC 1981 Mosaic structure and mRNA precursor of uteroglobin, a hormone regulated mammalian gene. J Bioi Chem 256:11911-11916 Cato ACB, Geisse S, Wenz M, Westphal HM, Beato M 1984 The nucleotide sequences recognized by the glucocorticoid receptor in the rabbit uteroglobin gene region are located far upstream from the initiation of transcription. EMBO J 3:2771-2778 Jantzen K, Fritton HP, Igo-Kemenes T, Espel E, Janich S, Cato ACB, Mugele K, Beato M 1987 Partial overlapping of binding sequences for steroid hormone receptors and DNase I hypersensitive sites in the rabbit uteroglobin gene region. Nucleic Acids Res 15:4535-4552 Slater EP, Redeuihl G, Theis K, Suske G, Beato M 1990 The uteroglobin promoter contains a noncanonical estrogen responsive element. Mol Endocrinol 4:604-610 Beato M, Chalepakis G, Schnauer M, Slater EP 1989 DNA regulatory elements for steroid hormones. J Steroid Biochem 32:731-748 Brinster RL, Palmiter RD 1986 Introduction of genes into the germ lines of animals. Harvey Lect 80:1-38 Ornitz DM, Palmiter RD, Hammer RM, Brinster RL, Swift GH, MacDonald RJ 1985 Specific expression of an elastase-growth hormone fusion gene in pancreatic acinar cells of transgenic mice. Nature 313:600-602 Townes TM, Lingrel JB, Chen HY, Brinster RL, Palmiter
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 16:40 For personal use only. No other uses without permission. . All rights reserved.
Vol 5 No. 3
MOL ENDO-1991 318
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
RD 1985 Erythroidspecific expression of human B-globin genes in transgenic mice. EMBO J 4:1715-1723 Grosveld F, van Assendelft GB, Greaves DR, Kollias G 1987 Position-independent, high-level expression of the human B-globin gene in transgenic mice. Cell 51:975-985 Finn CA, Publicover M 1981 Cell proliferation and cell death. In: Glasser SR, Bullock DW (eds) Cellular and Molecular Aspects of Implantation. Plenum Press, New York, pp 181-195 Bullock DW, Woo SLC, O'Malley BW 1976 Uteroglobin messenger RNA: translation in vitro. Biol Reprod 15:435443 Kumar NM, Bullock DW 1982 Hybridization analysis of steady state levels of uteroglobin mRNA in rabbit uterus and lung during early pregnancy. J Endocrinol 94:407414 Krumlauf R, Hammer RE, Brinster RL, Chapman VM Tilghman SM 1985 Developmental regulation of a-fetoprotein genes in transgenic mice. Mol Cell Biol 5:16391648 Swift GH, Hammer RE, MacDonald RJ, Brinster RL 1984 Tissue specific expression of rat pancreatic elastase I gene in transgenic mice. Cell 38:639-646 Pinkert CA, Ornitz DM, Brinster RM, Palmiter RD 1987 An albumin enhancer located 10 kb upstream functions along with its promoter to direct efficient, liver-specific expression in transgenic mice. Genes Dev 1:268-276 Rider V, Bullock DW 1988 Progesterone-dependent binding of a trans-acting factor to the uteroglobin promoter. Biochem Biophys Res Commun 156:1368-1375 Hogan B, Constantini F, Lacy E 1986 Manipulating the Mouse Embryo-A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor Quinn P, Barros C, Whittingham DG 1982 Preservation of
37. 38. 39. 40. 41. 42.
43. 44.
45.
46.
hamster oocytes to assess the fertilizing capacity of human spermatozoa. J Reprod Fertil 66:161-168 Southern EM 1979 Detection of specific sequences among DNA fragments separated by gel electrophoresis. J Mol Biol 98:503-517 Reed KC, Mann DA 1985 Rapid transfer of DNA from agarose gels to nylon membranes. Nucleic Acids Res 13:7202-7221 Feinberg AP, Vogelstein B 1983 A technique for radiolabeling DNA restriction endonuclease fragments to a high specific activity. Anal Biochem 132:6-13 Church GM, Gilbert W 1984 Genomic sequencing. Proc Natl Acad Sci USA 81:1991-1995 Thomas P1980 Hybridization of denatured RNA and small DNA fragments transferred to nitrocellulose. Proc Natl Acad Sci USA 77:5201-5205 Chirgwin JM, Prezybyla AE, MacDonald RJ, Rutter WJ 1979 Isolation of biologically active ribonucleic acid from sources enriched in ribonuclease. Biochemistry 18:52945299 Chandra T, Bullock DW, Woo SLC 1981 Hormonally regulated gene expression: steady-state level and nucleotide sequence of rabbit uteroglobin mRNA. DNA 1:19-26 Alonso S, Minty A, Bourlet Y, Buckingham M 1986 Comparison of three actin-coding sequences in the mouse; evolutionary relationships between the actin genes of warm blooded vertebrates. J Mol Evol 23:11-22 Melton DA, Kreig PA, Rebagliati MR, Maniatis T, Zinn K, Green MR 1984 Efficient in vitro synthesis of biologically active RNA and RNA hybridization probes from plasmids containing the SP6 promoter. Nucleic Acids Res 12:70357056 Zeff RA, Gopas J, Steinhauer E, Rajan TV, Nathenson TV 1986 Analysis of somatic cell H-2 variants for class I antigen expression. J Immunol 137:897-903
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 16:40 For personal use only. No other uses without permission. . All rights reserved.
Francesco J. DeMayo, Sami Damak, Thomas N. Hansen, and David W. Bullock Departments of Cell Biology (F.J.D.) and Pediatrics (T.N.H.) Baylor College of Medicine Houston, Texas 77030 Department of Biochemistry and Microbiology Lincoln University (S.D., D.W.B.) Canterbury, New Zealand
INTRODUCTION
The rabbit uteroglobin (UG) gene, with varying lengths of 5' flanking sequence, was introduced into the mouse genome to investigate the DNA sequences required for tissue-specific expression and regulation by steroid hormones. The pattern of expression and steroid hormone regulation of the transgene was compared to the expression and regulation of the endogenous mouse UG-like gene. In the rabbit, UG is induced in the uterus by progesterone and is expressed constitutively in the lungs, where it is weakly regulated by glucocorticoids. Genomic DNA fragments containing the complete UG-coding sequence with 4.0 (UG4.0), 3.0 (UG3.0), 2.3 (UG2.3), or 0.6 (UG0.6) kilobases of 5' flanking sequence were used to establish lines of transgenic mice. Expression of UG mRNA was observed in the lungs of UG4.0 (2/4 lines), UG3.0 (4/4 lines), UG2.3 (1/2 lines), and UG0.6 (4/4 lines) mice. Uterine expression was observed in UG3.0 (3/4 lines), UG2.3 (1/2 lines), and UG0.6 (2/4 lines). In the lungs of UG3.0 and UG2.3 mice, RNA expression was stimulated by treatment with dexamethasone. In the one line of UG3.0 mice examined, UG was regulated by ovarian steroids in the uterus. The endogenous mouse UG-like gene showed the major site of expression to be in the lung. Unlike the transgene, the endogenous gene was more strongly stimulated by glucocorticoids. Thus, we conclude that the cis elements needed for pulmonary expression of UG are contained within the UG2.3 fragment used to generate transgenic mice, but that other elements are required for full glucocorticoid regulation. Also, the transgene did not show the full uterine expression observed in the rabbit, but regulation by the ovarian steroids was observed. (Molecular Endocrinology 5: 311-318,1991)
Uteroglobin (UG) is a secretory protein of the male and female reproductive tract and pulmonary epithelium in rabbits (1, 2). UG was initially named blastokinin, because the timing of expression of this protein implied a potential role in implantation in the rabbit (3). UG binds progesterone specifically (4). Recent comparisons of its amino acid sequence have revealed homology with inhibitors of phospholipase-A2, and an antiinflammatory role has been proposed (5). Initially, UG was thought to be a protein unique to the rabbit; however, UG-like proteins have been reported in rat prostate (6) and lungs (7) and in human uterus (8) and prostate (9). Recently, the existence of homologous UG-like proteins in humans and rats was confirmed by the cloning and sequence analysis of the human 10-kDa Clara cell protein (10) and the rat polychlorobiphenyl-binding protein (pcbb) (11, 12). These proteins showed 61 % and 53% amino acid homology to the rabbit protein, respectively. Although these genes show sequence homology, the tissue distribution of their expression varies between species. Unlike the rabbit, the major site of expression of the rat pcbb gene is in the lung, with minor expression in the female reproductive tract (12). Although the physiological role of UG has not yet been defined, its regulation by steroid hormones has made it a useful model for hormonal control of mammalian gene expression. The effect varies with the steroid and the tissue in which this gene is expressed. In the female reproductive tract, UG is under the predominant influence of progesterone (13, 14), while in the lung its expression is constitutive (15) or weakly stimulated by glucocorticoids (16, 17). In the male reproductive tract, there is evidence that UG is regulated by androgens (18). The rabbit UG protein is encoded in a single copy gene that consists of three exons separated by two introns (19). Analysis of the 5' flanking region by DNAse hypersensitivity studies and steroid
0888-8809/91 /0311 -0318$03.00/0 Molecular Endocrinology Copyright© 1991 by The Endocrine Society
311
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 16:40 For personal use only. No other uses without permission. . All rights reserved.
Vol 5 No. 3
MOL ENDO-1991 312
receptor binding has shown that the putative glucocorticoid- and progesterone-responsive elements are located approximately 2.6 kilobases (kb) from the start of transcription (20, 21). Transfection of a UG-driven bacterial chloramphenicol acteyl transferase gene into Ishikawa cells has also identified estrogen-responsive regulatory elements 252 and 265 nucleotides 5' to the UG gene (22). Determination of the cis elements regulating tissuespecific expression and steroid regulation of the UG gene would shed light on whether the different expression of the UG-like genes in different species is due to a divergence in the cis elements between species or a divergence in the transcription factors required for expression; it would also allow a functional determination of the steroid regulatory elements. Identification of these elements has been impeded by the lack of a suitable cell culture system (23). In this paper we report the use of transgenic mice to overcome this problem. Introduction of foreign genes into the murine genome has allowed evaluation of the enhancer elements needed for the appropriate tissue-specific expression of several genes (24). The rat elastase (25) and the human /3-globin genes (26) have been studied extensively in this way. In the globin gene, the domains required for expression of the transgene independent of the site of integration have been determined (27). This report describes the use of the transgenic mouse to determine the site and regulation of the expression of the rabbit UG gene in murine tissues compared to those of the endogenous UG-like gene.
RESULTS Expression of UG mRNA in Transgenic Mice The fragments of the UG gene used to generate transgenic mice are shown in Fig. 1. Microinjection of these fragments of DNA resulted in four lines of UG4.0, four lines of UG3.0, two lines of UG2.3, and four lines of UG0.6. The number of copies of the transgene ranged from 1 to greater than 10 copies. Figure 2B shows the results of Northern analysis of RNA expression in one
Fig. 1. Organization of the Rabbit UG Gene and Restriction Fragments Used to Generate Transgenic Mice The wavy line indicates vector sequence, the thin line indicates rabbit flanking sequence, and the boxes indicates exons (•) and introns (•). Restriction enzyme sites are C/al (C), AsuW (A), Nae\ (N), Xba\ (X), and Sac\ (S).
28S
18S
•
Fig. 2. Tissue Distribution of the Endogenous Mouse UG-Like Gene in Nontransgenic Mice (A) and the Rabbit UG Gene in Transgenic Mice (B) Northern blots of total tissue RNA (20 ^g) were probed with an oligonucleotide probe derived from the rat pcbb cDNA sequence (A; see text) using the rat lung as a control and a rabbit uteroglobin genomic fragment using rabbit lung and nontransgenic mouse lung as control (B). Sal Gld, Salivary gland.
mouse from line UG 3.0. Rabbit lung RNA served as a positive control, and nontransgenic mouse lung RNA served as a negative control. Rabbit lung shows the 600-nucleotide (nt) band of mature rabbit UG mRNA (19), as well as the unprocessed 3-kb heteronuclear UG mRNA. Under these stringencies no hybridization was observed in the negative mouse lung RNA. This transgenic mouse line expressed UG mRNA of the correct size in lung and uterus, but not in any other tissue examined. All lines of mice containing the UG gene were tested for UG expression in lung and uterus by Northern analysis of 20 ^g total tissue RNA. No UG expression was observed in lungs or uteri of two lines with UG4.0 and one line with UG2.3. In all other lines some UG message could be detected. Table 1 summarizes the tissue distribution of UG expression in transgenic mice. The predominant tissue for expression of the UG gene is the lung. Expression in the female reproductive tract was low and variable. One explanation for this low expression could be that the analysis was conducted without re-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 16:40 For personal use only. No other uses without permission. . All rights reserved.
313
Uteroglobin in Transgenic Mice
Table 1. Tissue Distribution of the Expression of the Rabbit UG Gene in Transgenic Mice FO
Strain
UG4.0
43 7362
B F
UG3.0
5057 5058 5073 5074
F LJ. LJ.
Fragment
Ov
Ut
Te
SV
N
N
Sp
Li
Ki
N N N _
N -
-
_
+ _
Ht
Lu
Ty
SG
N N N _
— N N _
Br
F
159
B
UG0.6
2137 2141 2270 3970
F F LJ.
UG2.3
F
—
—
+
+
N N
N N
N N N _
N N N _
N N N
+ ++ +
_
+
The level of expression is graded, - for no detectable expression, + to + + + + for increasing levels of hybridization signal after Northern analysis, and N for not assayed. Ov, Ovary; Ut, uterus; Te, testis; SV, seminal vesicle; Sp, spleen; Li, liver; Ki, kidney; P, pancreas; M, muscle; Ht, heart; Lu, lung; Ty, thymus; SG, salivary gland; Br, brain. Mice were generated from embryos produced by ICR females bred to either FVB (F) or B6C3F1 (B) male mice. Lines were then established in an ICR background.
gard to the stage of the estrous cycle. To test for stimulation of UG expression, mice were administered the regimen of estrogen and progesterone described by Finn and Publicover (28). UG expression was not detected in uteri from these steroid-stimulated mice (data not shown). Thus, the stage of the estrous cycle is unlikely to influence the ability of the mouse uterus to express UG. The only other sites of expression in transgenic mice were the kidney, testis, salivary gland, seminal vesicles, and thymus, where the frequency and level of expression were low. Figure 2A also shows the distribution of expression of the endogenous mouse homolog to the rabbit UG gene, with rat lung RNA serving as a control. Twenty micrograms of total RNA were isolated from the tissues of nontransgenic mice and probed with a 36-mer oligonucleotide probe chosen from the rat pcbb protein, which showed a high degree of homology to the rabbit UG gene and the human homolog (see Materials and Methods). The predominant site of expression of the mouse homolog, like the transgene, is the lung. Reproductive tract expression was not observed, but the lack of detection of the mouse message could be due to a difference in the sensitivity of probing with an oligonucleotide probe. No expression was observed in the testis or male reproductive tract (data not shown). The only other site of expression of the mouse UG-like gene was the thymus (Fig. 2A). Figure 3 shows the expression of UG mRNA in lungs from representatives of all of the expressing lines of mice. The level of expression was highest in UG3.0 and UG2.3 mice, exceeding that in rabbit lung. The expression of UG in the two lines of UG4.0 mice was lower than that in the rabbit. The expression of UG in the lungs of UG0.6 mice varied with the line of transgenic mouse. A comparison of the expression of UG mRNA in uterine RNA is shown in Fig. 4. These data show that
CM
(o
00 '"J"
io s
co eo o o o t s in io m
28S—
K *- O O o> co ^- r*. f^ IO t- »- N O) »- CM CM CM CO
18S—
4.0
3.0
2.3
0.6
Fig. 3. Analysis of Pulmonary Expression of UG mRNA in Transgenic Mice by Northern Hybridization Expression in the lungs of the mice from line 2270 was at too low a level to be observed in this exposure of this autoradiograph. Fragments UG4.0, UG3.0, UG2.3, and UG0.6 are refered to as 4.0, 3.0, 2.3, and 0.6, respectively.
UG mRNA expression in the mice expressing the transgene is much lower than that in the 3-day pregnant rabbit. Only in line 5074 (UG3.0) was there significant expression in the transgenic mouse uterus. Line 5073 was examined later and found to have a level of UG mRNA similiar to that observed in line 5074 (Table 1). Steroid Regulation Glucocorticoid regulation of UG in the lung was investigated by comparing adrenalectomized mice given glucocorticoids with adrenalectomized mice given PBS. Figure 5 shows the slot blot analysis of the response of UG mRNA expression for mice for each gene fragment and a graphical representation of the mean response of UG message at the three levels of RNA
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 16:40 For personal use only. No other uses without permission. . All rights reserved.
MOL ENDO-1991 314
X) .0 CB
0) (0 3 O
Vol 5 No. 3
CM |s»
CO CO O
CO
Tf
(
D
^
T
-
O
O
O O l f l B r r
CMO)
I O I O I - S M N
M(0
28S
18S
4.0
3.0
2.3
0.6
Fig. 4. Analysis of Uterine Expression of UG mRNA in Transgenic Mice by Northern Hybridization Uterine expression of UG mRNA was at too low a level in the mouse from line 2141 to be observed in this exposure of this autoradiograph. Fragments UG4.0, UG3.0, UG2.3, and UG0.6 are refered to as 4.0, 3.0, 2.3, and 0.6, respectively.
c.
RNA Loading (|ig)
the steady state level of UG expression between gene fragments (P < 0.01). The steady state level of UG expression was significantly increased by treatment with dexamethasone in lines with the UG3.0 and UG2.3 fragments, but not in lines with UG4.0 and UG0.6 fragments (P = 0.017). To test whether the low level of dexamethasone stimulation of the UG transgene may reflect a lack of response of mouse lung to glucocorticoids, nontransgenic adrenalectomized mice were given the same regimen of hormone or placebo, and lung RNA was probed with the rat pcbb oligonucleotide. Figure 6 shows that dexamethasone gave a greater than 2-fold stimulation of the endogenous gene message. Thus, the mouse lung is sensitive to glucocorticoids, but the endogenous mouse gene is influenced more strongly than that of the transgene. Uterine expression of UG was extremely low and variable in the mice bearing the UG transgene. Only two lines, 5073 and 5074 (UG 3.0), expressed the transgene at a level at which regulation could be tested. Mice from line 5074 were administered either oil or the regimen of estrogen and progesterone described by Finn and Publicover (28). Figure 7 shows the results of Northern analysis of the UG transgene by ovarian steroids. The
- 400 E a o ~ 300
E 200 O o 100
1 2 RNA Loading (pg)
4
Fig. 5. Panel 1, Quantitation of UG Expression in Lungs of Transgenic Adrenalectomized Mice Given Dexamethasone (DEX; • — • ) or PBS (O- -O) Carrying UG Fragments UG4.0 (A), UG3.0 (B), UG2.3 (C), and UG0.6 (D) Panel 2 shows the slot blot analysis of 0.1, 0.5, and 1 ^g RNA from pairs of mice (a, b, and c) given PBS or dexamethasone and probed with the rabbit UG gene-specific probe. Panel 3 shows a duplicate slot blot probed for mouse /S-actin. The UG fragments in panels 2 and 3, as in panel 1, are described as UG4.0 (A), UG3.0 (B), UG 2.3 (C), or UG0.6 (D).
loaded onto the slot blot. The data were analyzed using the multiple regression package of the program Statistix (NH Analytical Software, Roseville, MN) running on a personal computer (see Materials and Methods). The analysis revealed a statistically significant difference in
Fig. 6. A, Quantitation of Expression of the Mouse UG-Like Gene in Adrenalectomized Nontransgenic Mice Given Dexamethasone (DEX; • - • ) or PBS (O- -O) B, Slot blot analysis of 1, 2, and 4 ^g total RNA extracted from lung probed with the oligonucleotide probe derived from the rat pcbb cDNA (see text) probe for the mouse UG-like gene. C, A duplicate blot probed for mouse /3-actin. The male mice represent two pairs of siblings from two separate litters.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 16:40 For personal use only. No other uses without permission. . All rights reserved.
315
Uteroglobin in Transgenic Mice
Oi
E+P
28S18S600
NT-
Fig. 7. Northern Analysis of 20 M9 Total Uterine RNA from UG3.0 Mice from Line 5074 Probed with the Rabbit UGSpecific Probe Four littermates were given sesame oil (no. 3503 and 3504) or a regimen of estradiol (E) and progesterone (P) (no. 3494 and 3501). Day 3 pregnant rabbit uterine RNA is shown as a control.
transgene message levels were approximately 2.5-fold greater with the administration of estrogen and progesterone than those in the mice given a placebo. Regulation of the endogenous gene was also examined by treating nontransgenic mice with the same regimen of hormones or placebo. Endogenous message in the reproductive tract could not be detected in either case.
DISCUSSION
The UG gene is expressed maximally in the rabbit uterus during early pregnancy (29). Expression of UG in the rabbit lung is low compared to that in the stimulated uterus and does not vary with pregnancy (30). The pattern of expression of the homologous rat protein is quite different, with the major site of expression being the lungs and only minor expression in the uterus (12). This report shows that the pattern of expression of the mouse homolog to the rabbit UG gene follows the pattern observed in the rat and not that in the rabbit. Also, expression of the rabbit UG gene in the mouse follows the pattern observed for the mouse UG-like gene. This would imply that the species difference in tissue distribution of UG-related genes may be due to a difference in the nature or abundance of transcription factors needed for the appropriate level of expression observed in the different tissues. While these experiments do not distinguish between transcriptional and posttranscriptional effects, evidence in the rabbit indicates that the UG level is regulated at the level of transcription (13-15). All rabbit UG fragments used to generate lines of transgenic mice showed the ability to express the transgene message in the lung. The frequency and level of expression, however, varied with the amount of 5'
sequence flanking the UG gene. The fragments with 3.0 and 2.3 kb of 5' flanking DNA exhibited the highest level of expression, surpassing the level in rabbit lung. The mice with 4.0 kb of 5' flanking DNA (UG4.0) expressed the UG gene at a lower level and a reduced frequency. This observation can be explained by the presence of 3.2 kb of Charon 4A sequences in this construct (Fig. 1), as prokaryotic DNA sequences have been shown to inhibit the expression of /3-globin (26) and a-fetoprotein (31) transgenes in mice. The inhibition of transgene expression by prokaryotic sequences is not universal. For example, expression of the rat elastase gene was not affected when microinjected with 4.3 kb of pBR322 sequences (32). Deletion of UG 5' flanking sequences to -0.6 kb gave increased variability and an overall reduction in the level of expression. Although this demonstrates that the proximal promoter elements needed for pulmonary expression of UG are contained within the 0.6-kb fragment, sequences between -2.3 and -0.6 kb are important for the full pulmonary expression of UG. Pulmonary expression of UG mRNA was increased significantly (P = 0.017) by treatment with dexamethasone. As found in the rabbit (15-17), this effect was small compared to the stimulation caused by progesterone in the rabbit uterus. Stimulation of pulmonary expression of the transgene by dexamethasone was also less than that of the endogenous gene. The effect of dexamethasone depended on the gene construct. No effect was seen with the UG4.0 fragment, probably due to the 3.2 kb of phage sequences, as discussed above. When the 5' flanking sequence was truncated to 0.6 kb, the effect of dexamethasone was lost. Putative sites for the glucorticoid response elements, as found by DNase-1 hypersensitivity and steroid-binding studies, are located in the region of -2.6 kb (20, 21). This putative glucocorticoid response element is present in UG3.0, but not UG2.3, yet both of these fragments were stimulated by dexamethasone. Thus, additional hormone response elements must be present between the A/ael and Xba\ sites in the 5' flanking region of the rabbit UG gene. Although the uterus is the major site of UG expression in the rabbit, expression of this gene in the uteri of transgenic mice was infrequent and could not be correlated with the amount of 5' flanking DNA. While the UG3.0 fragment contained all regions that have been implicated as important in uterine expression in the rabbit (20,21), this fragment was not sufficient to obtain a comparable level of expression in the mouse uterus. The extremely low levels of expression of the transgene in the uterus made it difficult to investigate stimulation of the transgene by ovarian steroids. However, in the line expressing the transgene at an adequate level, regulation of the UG message by ovarian steroids was observed. Additional elements surrounding the UG gene may be necessary for full expression in the uterus. The existence of tissue-specific enhancers at a considerable distance from the promoter is not unprecedented. The albumin gene requires an enhancer 10 kb up-stream
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 16:40 For personal use only. No other uses without permission. . All rights reserved.
Vol 5 No. 3
MOL ENDO-1991 316
from its promoter (33), and in the globin gene enhancers are found 50 kb removed from the gene (27). Alternatively, frans-acting factors in the rabbit uterus specific for the UG promoter (34) may be different or absent in the mouse uterus. This possibility is supported by the known differences in hormonal stimulation required for the establishment of pregnancy in mice and rabbits.
killed to assay for expression in various tissues by Northern hybridization (41). Total tissue RNA was isolated by homogenization in guanidine thiocyanate and separation on a gradient of CsCI2 (42). RNA was denatured and separated by size using formaldehyde-gel electrophoresis. The RNA was transferred to nitrocellulose membranes and probed for the presence of UG mRNA using the 32P-labeled probe described above. Steroid Regulation
MATERIALS AND METHODS
DNA Constructs The DNA used to generate transgenic mice was derived from pUG11.8, a plasmid consisting of an 11.8-kb genomic DNA fragment cloned in the C/al site of pBR322 (19). The 11.8-kb fragment contains the coding region of the rabbit UG gene, with 4.0 kb of 5' flanking sequence and 1.0 kb of 3' flanking sequence. The insert also contains 3.2 kb of Charon 4A DNA at the 5' end and 0.2 kb of Charon 4A sequence at the 3' end of the rabbit DNA sequence. The initial fragment incorporated into the murine genome was the 11.8-kb C/al fragment (UG4.0). Deletions of the 5' flanking sequence were generated by restriction endonuclease digestion of pUG11.8 with AsuW, A/ael, and Xba\ to produce constructs with 3.0 (UG3.0), 2.3 (UG2.3), and 0.6 (UG0.6) kb of 5' flanking sequence, respectively (Fig. 1). All injected fragments contained 0.2 kb of Charon 4A sequence at the 3' end. Generation of Transgenic Mice The DNA to be injected was digested with the appropriate restriction endonuclease and separated from pBR322 sequence by gel electrophoresis using low melting point agarose (Seaplaque, FMC Bioproducts, Rockland, ME). The appropriate fragment was excised, and the DNA was isolated using GeneClean (BIO 101, La Jolla, CA) and diluted to 2 Mg/ml in TE buffer [10 mM Tris (pH 7.5) and 0.25 mM EDTA]. Transgenic mice were produced as described by Hogan et al. (35), with the following modifications. Manipulation of mouse embryos was carried out in medium HT6 (36) supplemented with 5 mg/ ml cytochalasin-B (Sigma, St. Louis, MO) while the embryos were being injected. The culture medium contained 2 Mg/ml of the DNA being injected in order to avoid dilution in case of back-flow into the injection needle during microinjection. The embryos used for microinjection are summarized in Table 1. Integration and Expression Analysis DNA was isolated from tail biopsies taken after weaning of mice born from transferred embryos and was screened by Southern hybridization (37). The isolated DNA (10 ^g) was digested with EcoRI, subjected to agarose gel electrophoresis, and transferred to a Zeta Probe (Bio-Rad, Richmond, CA) membrane under alkaline conditions (38). The 1.4-kb Sacl fragment of pUG11.8, which contains the second and third exons of the UG gene, was labeled with [32P]dCTP (ICN, Irvine, CA) by random priming (39) and used as a hybridization probe under the conditions described by Church et al. (40). Transgenic mice were identified by the presence of the 1.1-kb internal EcoRI fragment. The number of copies of the transgene was estimated by comparison with 1,5, and 10 equivalent copies/haploid genome of pUG11.8 added to normal mouse DNA. Each Southern blot contained nontransgenic mouse DNA as a control. Founder transgenic (Fo) mice were outbred to ICR mice to establish transgenic lines for each integration. Once inheritance was established, one female mouse from each line was
Dexamethasone (Elkins-Sinn, Inc., Cherry Hill, NJ) was administered to intact or adrenalectomized transgenic mice, and UG expression was compared to intact or adrenalectomized mice given PBS. The dexamethasone was given at a dose of 25 /*g, ip, for five injections at 12-h intervals. The mice were killed 6 h after the last injection, and the lungs were taken for RNA analysis. From one transgenic line of each gene construct, three pairs of littermates were allocated to treatment, except for UG0.6, for which only two pairs were used. One member of each pair received dexamethasone, and the other received PBS. Total RNA was applied to nitrocellulose membrane, using a slot blot apparatus (Schleicher and Schuell, Keene, NH). For each mouse, RNA was applied to duplicate slots at three levels (0.1, 0.5, and 1.0 ng). Where necessary, yeast tRNA was added to bring the amount of RNA loaded to 1.0 Mg- One slot was hybridized with 32P-labeled UG RNA transcribed from UG cDNA (43) cloned in pGEM4 (Promega, Madison, Wl) using Sp6 RNA polymerase. The duplicate slot was hybridized with mouse /3actin RNA, prepared similarly from mouse /?-actin cDNA (44) in plasmid cloned pGEM3, and served as a control for uniformity of RNA loading. The blots were hybridized as described by Melton et al. (45). The blots were washed at 68 C in 0.1% sodium dodecyl sulfate and 0.1% SSC (0.015 M NaCI, 0.015 M Na citrate) for 1 h and subsequently treated with RNase. After washing, the membrane was placed against x-ray film and exposed overnight. The slots on the membrane were oriented using the autoradiograph and cut out individually for quantitation of radioactivity by liquid scintillation counting. Expression of the mouse homologous mouse gene was assayed using the oligonucleotide GATCTTCTCCGTGAGCTTCACGATGTTTATTCTGGT, which was kinased with [7-32P]ATP (ICN). This oligonucleotide corresponds to the opposite strand of the rat pcbb cDNA from nucletoides 274310. This area was chosen because it showed a high degree of homology with the rabbit and human cDNA sequences. The Northerns and slot blots were hybridized using the protocol described by Zeff and co-workers (46) with the following modifications. 1) The hybridization temperature was 50 C. 2) The blots were washed for 45 min. 3) Total tissue RNA was analyzed by using 20 ng RNA. Glucocorticoid regulation was conducted using four pairs of male mice produced from a back-cross of ICR x FvB to ICR mice. The mice were adrenalectomized and given the same regimen of hormone or placebo as described above. Slot blot analysis was conducted using 1, 2, and 4 ng total RNA. Regulation of the endogenous mouse gene and the transgene in the uterus by ovarian steroids was investigated using the regimen of estrogen and progesterone described by Finn and Publicover (28). Female litermates were given 100 ng estradiol, sc, for 2 days. Three days after the last estradiol treatment, mice were given 500 ^g progesterone, sc. The mice were then given 500 ng progesterone with 20 ng estradiol, sc, for 2 days. The mice were killed 6 h after the last injection, and total RNA was isolated from the uteri. Control mice were given sesame oil in place of steroid treatment. Northern analysis, using the UG probe for transgenic mice and the pcbb probe for nontransgenic mice, was performed to avoid any problems with background signals due to the low level of uterine UG expression. The loading of RNA was monitored by examining the ethidium bromide staining of the filter after transfer. The fold induction was determined by densitometric
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 16:40 For personal use only. No other uses without permission. . All rights reserved.
317
Uteroglobin in Transgenic Mice
scanning of the autoradiograph using a Kontes Fiber Optic Scanner (Vineland, NJ) and a Hewlett-Packard 3390A Integrator (Houston, TX).
6.
Statistical Analysis The slot blot data were analyzed by multiple regression, using the program Statistix (NH Analytical Software, Roseville, MN) running on a personal computer. Variables were coded as follows: UG expression (Y); construct (X1): UG4.0 = 1, UG3.0 = 2; UG2.3 = 3; UG0.6 = 4; treatment (X2): PBS = 1; dexamethasone = 2; RNA loading (X3): 0.1 ^g = 1- 0.5 ^g = 2,1 ng = 3; replicate (X4); actin expression (X5). Linear models were fitted to the data by stepwise regression, and variables contributing coefficients with t values significant at P < 0.05 were selected. Stepwise regression revealed that actin expression and second or third order interactions between variables contributed little to partitioning of the overall variance. The fitted equation Y = 1962.2X1 - 379.9X12 + 248.1X2 + 552.6X3 3106.9 accounted for 70.5% of the variance. The t value and probabilities for each of the coefficients were 7.71 (P < 0.01), -7.39 (P < 0.01), 2.45 (P < 0.0165), 8.91 (P < 0.01), and -7.89 (P < 0.01), respectively.
7. 8.
9. 10.
11.
12.
Experimental Animals All experiments were conducted using the highest standard for humane care in accordance with the NIH Guide for the Care and Use of Laboratory Animals.
Acknowledgments We thank Ms. Louise Ann Stanley, Mr. Li Li, Ms. Serlina Robinson, and Ms. Celeste Moore for their technical assistance, and Mr. Bruce Robson for statistical advice. We are grateful to Ms. Lisa Gamble and Ms. Beryl Bond for secretarial assistance.
13. 14. 15.
16. 17.
18. Received August 31, 1990. Revision received December 14,1990. Accepted December 14,1990. Address requests for reprints to: Dr. Francesco J. DeMayo, Department of Cell Biology, Baylor College of Medicine, Houston, Texas 77030. This work was supported in part by NIH Grant HD-093778 and the New Zealand Medical Research Council.
REFERENCES 1. Cato ACB, Beato M 1985 The hormonal regulation of uteroglobin gene expression. Anticancer Res 5:65-72 2. Miele L, Cordella-Miele E, Mukherjee AB 1987 Uteroglobin: structure, molecular biology and new perspectives on its functions as a phospholipase A2 inhibitor. Endocr Rev 8:474-490 3. Krishnan RS, Daniel JC 1967 Blastokinin: inducer and regulator of blastocyst development in rabbit uterus. Science 158:490-492 4. Beato M, Baier R 1975 Binding of progesterone to the proteins of the uterine luminal fluid, identification of uteroglobin as the binding protein. Biochim Biophys Acta 392:346-356 5. Miele L, Cordella-Miele E, Facchiano A, Mukherjee AB 1988 Novel anti-inflamatory peptides from the region of
19.
20.
21.
22.
23.
24. 25.
26.
highest similarity between uteroglobin and lipocortin I. Nature 335:726-730 Aunukker G, Seitz J, Heyns W, Kircher C 1985 Ultrastructural localization of uteroglobin immunoreactivity in rabbit lung and endometrium, and rat ventral prostate. Histochemistry 83:413-417 Lopez de Haro MS, Alvarez L, Nieto A 1988 Evidence for the identity of anti-proteinase protein CCSP and uteroglobin. FEBS Lett 232:351-353 Cowan BD, North DH, Whitworth NS, Fujita R, Shumacher EK, Mukherjee AB 1986 Identification of a uteroglobinlike antigen in human uterine like washings. Fertil Steril 45:820-823 Manyak MJ, Kikikawa T, Mukerjee AB 1988 Expression of a uteroglobin-like protein in human prostate. J Urol 140:176-182 Signh G, Katyal SL, Brown WE, Philips S, Kenedy AL, Anthony J, Squelia N 1990 Amino acid and nucleotide sequence of human Clara cell 10 kDa protein. Biochim Biophys Acta 950:329-337 Nordlund-Moller L, Anderson O, Ahlgren R, Schilling J Gillner, M, Gustaffson J-A, Lund J 1990 Cloning, structure and expression of a rat protein for polychlorinated biphenyls. J Biol Chem 21:12690-12693 Hagen G, Wolf M, Katyal SL, Singh G, Beato M, Suske G 1990 Tissue specific expression and 5' flanking gene region of the rat Clara cell 10 kDa protein: comparison to rabbit uteroglobin. Nucleic Acids Res 18:2939-2945 Heins B, Beato M 1981 Hormonal control of uteroglobin secretion and preuteroglobin mRNA content in rabbit endometrium. Mol Cell Endocrinol 21:139-150 Shen X-Z, Tsai M-J, Bullock DW, Woo SLC 1983 Hormonal regulation of rabbit uteroglobin gene transcription. Endocrinology 112:871-875 Savouret J-F, Loosfelt H, Atger M, Milgrom E 1980 Differential hormonal control of a messenger RNA in two tissues: uteroglobin mRNA in the lung and the endometrium. J Biol Chem 255:4131-4136 Lombardero M, Nieto A 1981 Glucocorticoid and developmental regulation of uteroglobin synthesis in rabbit lung. Biochem J 200:487-494 Torkelli T, Krusius T Janne O 1978 Uterine and lung uteroglobins in the rabbit. Biochim Biophys Acta 544:578592 Lopez de Haro MS, Alvarez L, Nieto A1988 Testosterone induces the expression of the uteroglobin gene in rabbit epididymus. Biochem J 250:647-651 Snead R, Day L, Chandra T, Mace M, Bullock DW, Woo SLC 1981 Mosaic structure and mRNA precursor of uteroglobin, a hormone regulated mammalian gene. J Bioi Chem 256:11911-11916 Cato ACB, Geisse S, Wenz M, Westphal HM, Beato M 1984 The nucleotide sequences recognized by the glucocorticoid receptor in the rabbit uteroglobin gene region are located far upstream from the initiation of transcription. EMBO J 3:2771-2778 Jantzen K, Fritton HP, Igo-Kemenes T, Espel E, Janich S, Cato ACB, Mugele K, Beato M 1987 Partial overlapping of binding sequences for steroid hormone receptors and DNase I hypersensitive sites in the rabbit uteroglobin gene region. Nucleic Acids Res 15:4535-4552 Slater EP, Redeuihl G, Theis K, Suske G, Beato M 1990 The uteroglobin promoter contains a noncanonical estrogen responsive element. Mol Endocrinol 4:604-610 Beato M, Chalepakis G, Schnauer M, Slater EP 1989 DNA regulatory elements for steroid hormones. J Steroid Biochem 32:731-748 Brinster RL, Palmiter RD 1986 Introduction of genes into the germ lines of animals. Harvey Lect 80:1-38 Ornitz DM, Palmiter RD, Hammer RM, Brinster RL, Swift GH, MacDonald RJ 1985 Specific expression of an elastase-growth hormone fusion gene in pancreatic acinar cells of transgenic mice. Nature 313:600-602 Townes TM, Lingrel JB, Chen HY, Brinster RL, Palmiter
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 16:40 For personal use only. No other uses without permission. . All rights reserved.
Vol 5 No. 3
MOL ENDO-1991 318
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
RD 1985 Erythroidspecific expression of human B-globin genes in transgenic mice. EMBO J 4:1715-1723 Grosveld F, van Assendelft GB, Greaves DR, Kollias G 1987 Position-independent, high-level expression of the human B-globin gene in transgenic mice. Cell 51:975-985 Finn CA, Publicover M 1981 Cell proliferation and cell death. In: Glasser SR, Bullock DW (eds) Cellular and Molecular Aspects of Implantation. Plenum Press, New York, pp 181-195 Bullock DW, Woo SLC, O'Malley BW 1976 Uteroglobin messenger RNA: translation in vitro. Biol Reprod 15:435443 Kumar NM, Bullock DW 1982 Hybridization analysis of steady state levels of uteroglobin mRNA in rabbit uterus and lung during early pregnancy. J Endocrinol 94:407414 Krumlauf R, Hammer RE, Brinster RL, Chapman VM Tilghman SM 1985 Developmental regulation of a-fetoprotein genes in transgenic mice. Mol Cell Biol 5:16391648 Swift GH, Hammer RE, MacDonald RJ, Brinster RL 1984 Tissue specific expression of rat pancreatic elastase I gene in transgenic mice. Cell 38:639-646 Pinkert CA, Ornitz DM, Brinster RM, Palmiter RD 1987 An albumin enhancer located 10 kb upstream functions along with its promoter to direct efficient, liver-specific expression in transgenic mice. Genes Dev 1:268-276 Rider V, Bullock DW 1988 Progesterone-dependent binding of a trans-acting factor to the uteroglobin promoter. Biochem Biophys Res Commun 156:1368-1375 Hogan B, Constantini F, Lacy E 1986 Manipulating the Mouse Embryo-A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor Quinn P, Barros C, Whittingham DG 1982 Preservation of
37. 38. 39. 40. 41. 42.
43. 44.
45.
46.
hamster oocytes to assess the fertilizing capacity of human spermatozoa. J Reprod Fertil 66:161-168 Southern EM 1979 Detection of specific sequences among DNA fragments separated by gel electrophoresis. J Mol Biol 98:503-517 Reed KC, Mann DA 1985 Rapid transfer of DNA from agarose gels to nylon membranes. Nucleic Acids Res 13:7202-7221 Feinberg AP, Vogelstein B 1983 A technique for radiolabeling DNA restriction endonuclease fragments to a high specific activity. Anal Biochem 132:6-13 Church GM, Gilbert W 1984 Genomic sequencing. Proc Natl Acad Sci USA 81:1991-1995 Thomas P1980 Hybridization of denatured RNA and small DNA fragments transferred to nitrocellulose. Proc Natl Acad Sci USA 77:5201-5205 Chirgwin JM, Prezybyla AE, MacDonald RJ, Rutter WJ 1979 Isolation of biologically active ribonucleic acid from sources enriched in ribonuclease. Biochemistry 18:52945299 Chandra T, Bullock DW, Woo SLC 1981 Hormonally regulated gene expression: steady-state level and nucleotide sequence of rabbit uteroglobin mRNA. DNA 1:19-26 Alonso S, Minty A, Bourlet Y, Buckingham M 1986 Comparison of three actin-coding sequences in the mouse; evolutionary relationships between the actin genes of warm blooded vertebrates. J Mol Evol 23:11-22 Melton DA, Kreig PA, Rebagliati MR, Maniatis T, Zinn K, Green MR 1984 Efficient in vitro synthesis of biologically active RNA and RNA hybridization probes from plasmids containing the SP6 promoter. Nucleic Acids Res 12:70357056 Zeff RA, Gopas J, Steinhauer E, Rajan TV, Nathenson TV 1986 Analysis of somatic cell H-2 variants for class I antigen expression. J Immunol 137:897-903
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 16:40 For personal use only. No other uses without permission. . All rights reserved.
