Retinol-binding proteins (RBP) are secreted by the porcine uterus under the influence of progesterone and consist of multiple charge forms. Evidence has been ...
BIOLOGY OF REPRODUCTION 48, 998-1005 (1993)
Porcine Uterine Retinol-Binding Proteins Are Identical Gene Products to the Serum Retinol-Binding Protein' MELODY L. STALLINGS-MANN, WILLIAM E. TROUT, and R. MICHAEL ROBERTS 2 Departments of Animal Sciences and Biochemistry, University of Missouri, Columbia, Missouri 65211 ABSTRACT Retinol-binding proteins (RBP) are secreted by the porcine uterus under the influence of progesterone and consist of multiple charge forms. Evidence has been previously presented by this laboratory that these uterine RBP are distinct from serum RBP. We have followed the secretion of the uterine RBP during two stages of pseudopregnancy, examined their properties and amino acid sequences, and attempted to clone their cDNA. Analysis of the charge forms present in uterine flushes by anion-exchange chromatography showed that forms 1 (p < 0.01) and 3 (P < 0.05) predominated at Day 13, whereas forms 2 (p < 0.05) and 4 (p < 0.01 ) were most abundant at Day 45. All four charge forms appeared to form stable complexes with transthyretin (TfR) and were recognized by antiserum to human serum RBP on Western blots. Several cDNA clones isolated from an endometrial cDNA library all appeared to code for a protein identical to classical RBP. Off-blot amino acid sequencing of the first ten residues of two of the more divergent charge forms of uterine RBP indicated complete sequence identity with pig serum RBP. These data suggest that the uterine RBP charge forms may be slightly modified forms of a single protein product corresponding to the classical form of RBP. The change in appearance of the charge forms during pseudopregnancy is probably due to chemical modifications. These modifications do not appear to influence the binding of each charge form to TTR.
INTRODUCTION The uterus of the pig synthesizes and secretes large amounts of protein in response to progesterone. The major proteins synthesized by the uterus include uteroferrin [1, 2], serine protease inhibitors of the Kunitz class [3], glycoproteins belonging to the serpin superfamily of proteins [4, 5], and retinol-binding proteins (RBP) [6-8]. These proteins are believed to be important for conceptus growth and development. RBP is a small protein (21 kDa), classically described as a product of the liver. It is secreted into the bloodstream, where it acts as a transport protein for the alcohol form of vitamin A (retinol). RBP circulates as a complex with another larger protein, transthyretin (TR). Association with TTR is thought to stabilize the retinol-RBP complex as well as prevent loss of RBP through glomerular filtration [9]. Although RBP is a product of the liver, it has been shown to be secreted by a variety of extrahepatic sites as well [10]. These include the visceral yolk sac in the rat [11, 12]; conceptus and placental tissues in the pig [13, 14], sheep [15, 16], and cow [17]; and the uterus of the pig [6-8], sheep [18], and cow [19]. In the pig, secretion of RBP by the uterus is thought to transport retinol from the maternal circulation to the conceptuses during pregnancy. In contrast to the liver, the secretion of uterine RBP appears to be under steroid control. Intrauterine concentrations increased in ovariectomized gilts in response to both short-term [6] and long-term progesterone treatment [7].
More recently, the progesterone-dependent secretion of uterine RBP was shown to be markedly increased by estrogen in ovariectomized gilts [8]. Furthermore, Northern analysis of RNA isolated from uterine endometrium indicated increased levels of RBP mRNA in pregnant gilts at the time of conceptus elongation, suggesting that conceptus estrogen may trigger uterine RBP synthesis. Ion-exchange chromatography of RBP purified from porcine uterine flushings revealed three to four distinct charge forms [7]. Microheterogeneity has been previously demonstrated to occur with human [20-23], rat [24], and chicken [23] serum RBP as well as porcine conceptus RBP [13] and may be the result of loss of amide groups from a single gene product [20]. In contrast, direct amino acid sequencing of three of the charge forms of porcine uterine RBP indicated that the sequences, while similar to serum RBP, were unique [7]. The aim of our study was to obtain evidence for structural and functional differences among the different forms of uterine RBP and to clone cDNAs corresponding to the unique forms. MATERIALS AND METHODS Animals and Surgical Procedures Sexually mature cross-bred gilts were observed daily for estrus in the presence of intact boars. The onset of estrus was designated as Day 0. Pseudopregnancy was induced by daily injection of 2.5 mg of estradiol valerate in sesame seed oil on Day 11 and Day 12 in gilts utilized on Day 13 of pseudopregnancy and on Day 11 to Day 14 in gilts maintained through Day 45 of pseudopregnancy [25]. Gilts on Day 13 and Day 45 of pseudopregnancy were subjected to surgery in which each uterine horn was flushed with 30 ml 0.9% (w/v) NaCl to collect the accumulated proteins [26].
Accepted December 14, 1992. Received October 26, 1992. 'This work was supported by USDA grant #89-37240-4586. This work is journal series number 11,791 of the Missouri Agricultural Experiment Station. 2 Correspondence: 158 Animal Sciences Research Center, University of Missouri, Columbia, MO 65211. FAX: (314) 882-6827.
998
PORCINE UTERINE RBP
Uterine flushes were centrifuged at 1000 x g for 15 min and stored at -20°C. Flushings were not dialyzed in order to avoid loss of retinol from RBP. Purification of RBP The uterine RBP were partially purified according to the method of Clawitter et al. [7] from uterine flushings obtained from Day 45 pseudopregnant pigs. This process involved passage through a CM-cellulose column (Pharmacia LKB, Piscataway, NJ) at pH 8.2 to remove strongly basic proteins, gel filtration on Sephadex G-100 (Pharmacia LKB) to collect the Mr = 20 000-30 000 peak that contained the RBP, and high-performance anion-exchange chromatography on an Ultropac TSK-DEAE-5PW LKB HPLC column (Pharmacia LKB; see below and Fig. 2). The apo- or vitamin-free form of each isoelectric variant (see [7]) was pooled with its corresponding holoform. Pig serum RBP was purified as described by Clawitter et al. [7]. Fast Separation of Uterine RBP Charge Forms Approximately 50-100 1L of nondialyzed uterine flushings were loaded onto an Ultropac TSK-DEAE-5PW LKB HPLC column and eluted with a linear gradient (0.075-0.25 M NaCI in 0.01 M Tris-HC1, pH 8.2, 30 ml; 30 min). Protein in the eluant was monitored by absorbance at 280 nm. Retinol was monitored by relative fluorescence (excitation at 325 nm, emission at 460 nm) with a Shimadzu Model RF-551 fluorescence HPLC monitor (Kyoto, Japan). The area under each fluorescent peak was expressed as a percentage of the total fluorescence area as quantified by a Shimadzu Model C-R3A Chromatopac integrator. Association of RBP with TR Approximately 10 lRg of pig serum RBP and 30 jg of partially purified pig uterine RBP were each separately incubated with excess (125 jIg) TTR (Calbiochem, La Jolla, CA) and excess (10 ILg) retinol (Sigma Chemical Co., St. Louis, MO) in a 50-jl reaction (volume adjusted with 0.03 M Na2HPO 4, 0.15 M NaCl, pH 7.0) for 4 h at room temperature. The unbound RBP was separated from TTR-bound RBP by gel filtration on a Superose 12 FPLC column (Pharmacia LKB) and eluted with 0.03 M Na 2HPO4 , 0.15 M NaCl (pH 7.0) at a flow rate of 0.5 ml/min. Elution of RBP was monitored by fluorescence as described above. Oligonucleotide Primers A family of degenerate oligonucleotide probes was designed to correspond to amino acids 11-18 of one of the unique uterine RBP, peak 2,3 (Fig. 1), as previously reported [7]. An antisense oligonucleotide, RBP-240 [14], was also synthesized to represent the retinol binding pocket, a highly conserved region in human, rat, and rabbit RBP [27]. Additional antisense RBP primers (RBP-157 and RBP-293) were designed that corresponded to regions within the
999
coding region of the porcine RBP conceptus cDNA [14]. The remaining primers were designed for use in sequencing of the cDNAs in either X-gtll (GTF and GTR) or in plasmid (M13F and M13R) [28]. The sequences of these oligonucleotides are as follows: uRBP, 5'GCNAAA(G)GAA(G)GTNGAT(C)ATGAAA(G)GCN3'; RBP-240, 5'GGCAAACACGAAGGAGTAGCTGTCAGCACAGGTGCCATC3'; RBP- 157, 5' CTTGGCCATGGCGTACCAGGTGCCGGAGAA3'; RBP-293, 5'TGCGCACACGTCCCAGTrAT3'; GTF, 5' GGTGGCGACGACTCCTGGAGCCCG3'; GTR, 5'TTGACACCAGACCAACTGGTAATG3'; M13F, 5'TGACCGGCAGCAAAATG3'; M13R, 5'AACAGCTATGACCATG3' Preparationof Nucleic Acid Probes The oligonucleotides (4 Lg) were end-labeled by the forward reaction of polynucleotide kinase to a specific activity of approximately 1 x 108 cpm/ig with 200 Ci [y32 P]dATP (6000 Ci/mmol; New England Nuclear-DuPont, Boston, MA). Radiolabeled oligonucleotides were separated from free [y-32P]dATP by centrifugation at 1000 x g through 0.8 ml Sephadex G-50 [29]. A uterine cDNA (BT13), which was identical to the 3'618 bp of a cDNA for porcine conceptus RBP [14], was used for additional screening of the cDNA library. BT13 (400 ng) was radiolabeled with [a- 32P]dATP (3000 Ci/mmol; ICN Biomedicals, Inc., Irvine, CA) by priming with random hexamers [30]. Labeled probe was separated from free [32P]dATP by centrifugation at 1000 x g through 0.8 ml Sephadex G50, yielding a probe with specific activity of approximately 1 x 109 cpm/ikg. Screening of cDNA Library The oligonucleotides representing the unique uterine RBP were used to screen greater than 300 000 plaque-forming units (PFU) of a X-gtll cDNA library constructed from polyadenylated RNA isolated from uterine endometrium of pigs at Day 60 of pregnancy [31] by standard procedures [29] under appropriate low-stringency conditions (see [32]). The library was subsequently rescreened with the RBP-240 probe (150 000 PFU) and with the BT13 RBP cDNA probe (60 000 PFU). Positive phage plaques were identified by exposure to x-ray film (XAR-5; Eastman Kodak, Rochester, NY) at -80C for 16-24 h and purified by secondary and tertiary screening. Subcloning and Sequence Analysis Inserts corresponding to four uterine clones (U2, U4, U8, and BT13) were excised from X-gtll by digestion with the restriction enzyme, EcoRI, and subcloned into pUC13 or pUC19 plasmids. The 5'-end of five apparently full-length X clones U11, U18, U22, U24, and U25 were amplified by an asymmetric polymerase chain reaction (PCR) [33] by em-
1000
STALLINGS-MANN ET AL.
Sample sequenced
Amino acid sequence obtained or inferred 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Peak 2,3
ala
pro
ala
glu
val
val
pro
val
val
pro
ala
lys
glu
val
asp
met
lys
ala
Peak 2 [DEAE]
thr
gly
ala
glu
lys
ser
gly
val
val
pro
ala
val
glu
val
asp
met
arg
arg
Peak 2 [PVDF]
glu
arg
asp
-
arg
val
ser
ser
phe
arg
Peak 4 [PVDF]
glu
arg
asp
-
arg
val
ser
ser
phe
arg
-
arg
asp
-
arg
val
ser
-
phe
arg
val
lys
glu
asn
phe
asp
-
ala
glu
arg
asp
cys
arg
val
ser
ser
phe
arg
val
lys
glu
asn
phe
asp
lye
ala
glu
arg
asp
cys
arg
val
ser
ser
phe
arg
val
lys
glu
asn
phe
asp
lye
ala
glu
arg
asp
cys
arg
val
ser
ser
phe
arg
val
lys
glu
asn
phe
asp
lye
ala
Porcine Serum RBP
Human Serum RBP (cDNA) Porcine Uterine RBP (cDNA) Porcine Conceptus REP (cDNA)
FIG. 1. Comparison of amino acid sequences of the uterine RBP with human and porcine serum and conceptus RBP. Hyphens (-) indicate a lack of a major signal in a given cycle. The sequence reported for porcine uterine RBP was obtained from the cDNA reported in this paper. Peak 2,3 and porcine serum RBP [7]; human serum RBP 124, 37]; porcine conceptus RBP 14].
ploying primers flanking the X-gtl 1 EcoRI insertion site and a limiting quantity of an internal antisense RBP primer (RBP293). RBP-293 corresponds to amino acids 65-71 of porcine conceptus RBP [14]. Plasmid inserts and asymmetric PCR products were sequenced by the dideoxy method [34, 35] with M13 forward and reverse primers [28] and internal RBP primers. Western Blot Analysis Approximately 0.4-1zg samples of each peak of protein were loaded onto each lane and electroblotted onto polyvinylidene difluoride membranes (PVDF, Immobilon, Millipore, Bedford, MA) following one-dimensional PAGE. The blotted proteins were detected as follows: nonspecific binding sites were saturated by 30-min incubation at room temperature in 5% (w/v) nonfat dry milk (NFDM) in Trisbuffered saline (TBST; 10 mM Tris-HC1, 150 mM NaC1, 0.05% Tween 20, pH 8.0). The blots were washed three times for 5 min each in TBST and then incubated overnight at 4°C
on a rocking platform in either rabbit anti-human serum RBP (Dako Corp., Santa Barbara, CA) or normal rabbit serum (NRS) diluted 1:400 in TBST that contained 1% (w/v) NFDM. Blots were washed three times (5 min each) with TBST before incubation with the second antibody, an alkaline phosphatase-conjugated goat anti-rabbit immunoglobulin (Ig) G (Promega, Madison, WI) diluted 1:2000 in TBST. After a 2-h incubation at room temperature and three washes in TBST, the bands were visualized by addition of nitro-blue tetrazolium and 5-bromo-4-chloro-3-indolyl phosphate (Promega). Porcine serum RBP (0.4 jig) and prestained lowmolecular weight proteins (Bio-Rad Laboratories, Richmond, CA) were loaded as standards. Amino Acid Sequencing Approximately 4 g of each uterine RBP charge form was subjected to one-dimensional PAGE and then electroblotted onto PVDF (Immobilon, Millipore) in a 0.025 M Tris, 0.192 glycine buffer (pH 8.2) at 4°C for 2.5 h at 25 V. Pre-
PORCINE UTERINE RBP - A280 012-
*
......
r
Relative Fluorescence
008-
- 600
-400
Y
L
3
2
LL
*
u
004-
-2002
. , ,
........................................- ...........- .......... I
:
,, ,
L.
,,
..:,
, , ,, .,
.,
.
'...I
I
I
I
I
10
20
30
40
...._-.._
'.
I
50
Volume (mil
FIG. 2. High-performance anion-exchange chromatography of the partially purified RBP following gel filtration. Chromatography was performed on an LKB Ultropac TSK DEAE-5PW HPLC column. The gradient was linear (0.01-0.025 M NaCI; 60 ml; 60 m i d . Polypeptides were detected by continuous monitoring of the column effluent at 280 nm. Retinol-bound polypeptides were detected by continuous monitoring of the column effluent for fluorescence at an excitation of 325 nm and an emission of 460 nm. The fluorescent peaks were numbered 1 through 4.
stained low-molecular weight proteins (Integrated Separation Systems, Hyde Park, MA, or Bio-Rad) were loaded as standards. The blots were immediately stained with Coomassie Brilliant blue dye. The protein bands of apparent M, 21 000 from the four selected peaks that had been identified by Coomassie Brilliant blue staining on the PVDF blots were excised and subjected to NH2-terminalamino acid sequence analysis by the Edman degradation method (Protein Core Facility, University of Missouri, Columbia, MO) on an Applied Biosystems, Inc. (ABI, Foster City, CA) model 470 protein sequencer with on-line analysis for phenylthiohydantoin derivatives.
Densitornetry The intensity of the Coomassie Brilliant blue-stained bands associated with uterine RBP peaks 2 and 3 on polyacryamide gels and PVDF blots was determined by densitometry with a Bio-Rad Model 620 video densitometer and the accompanying one-dimensional analysis and integration software. Statistical Analysis Fluorescence associated with each RBP peak was expressed as a percentage of the total retinol-associated fluorescence detected in uterine flushings obtained from each pig. The percentages were analyzed by ANOVA, with peak number and day of pseudopregnancy as main effects. Treatment means were separated with the Student-Neumann-Keuls Multiple Range test [36]. RESULTS
Characterization of the I s o f o m at Diffent Stages of Pseudopregnancy Anion-exchange chromatography of an acidic protein fraction of relatively low molecular weight (25 000) ob-
Minuter
FIG. 3. Representative fluorescence profile obtained during DEAE-HPLC of uterine flushings obtained from a Day 13 and a Day 45 pseudopregnant gilt. Chromatography was performed on an LKB Ultropac TSK DEAE-5PW HPLC column. The gradient was linear (0.075-0.25 M NaCI; 30 ml; 30 m i d . Retinol-bound polypeptides were detected by continuous monitoring of the column effluent for fluorescence at an excitation of 325 nm and an emission of 460 nm. Day 13 is represented by the solid line and closed circles, Day 45 by the solid line and open circles. Data shown here are from single gilts.
tained from uterine flushings of pseudopregnant gilts has previously revealed the presence of about eight peaks of protein eluting relatively late in the gradient (0.12-0.19 M NaCl; Fig. 2) [7]. Four of these contained bound retinol whereas the remaining four appeared to represent apoforms (lachng the vitamin). Thus, if excess retinol was added to the preparations prior to ion-exchange chromatography, only four peaks were observed [7]. Two-dimensional electrophoresis of this low molecular weight fraction prior to ion-exchange chromatography revealed two major spots, and additional minor polypeptide spots of lower pI but of identical size (M, 22 000) were frequently seen [7]. Since these RBP variants obtained by ion-exchange chromatography have been reported to differ in their NH2-terminal amino acid sequences [7], it was of interest to determine whether the proportion of these isoforms changed under different physiological conditions. Therefore, ion-exchange chromatography was used to measure the quantity of the different charged forms of RBP present on Day 13 and Day 45 of pseudopregnancy. Day 13 was chosen to simulate the time of maternal recognition of pregnancy, when the RBP can first be detected in uterine flushings in large amounts [B], while Day 45 was chosen because it represents the stage at which Clawitter et al. [7] first purified the RBP from uterine flushings and demonstrated the presence of at least four apparently distinct forms. Anion-exchange chromatography of uterine flushings from five different gilts at Day 13 of pseudopregnancy revealed two major peaks of fluorescence (Fig. 3), which eluted at 9.6 min (0.13 M NaCl) and 15.3 min (0.16 M NaCl). Upon comparison with the protein profiles from partially purified RBP from uterine flushings (Fig. 2), the first peak from the
1002
STALLINGS-MANN ET AL.
TABLE 1. Distribution of charge variants of RBP in uterine flushings of pseudopregnant pigs.
Peak
Day 13* pseudopregnancy (12)
-- o--
Day 45* pseudopregnancy (10)
URBP
TTR URBP
6.ld
1
59.9c
2
7.0"
3 4
26.8" ¢ 1.3
13.8 34 .1d
3.8
3.4
30.3
b
I
b
I cc =
7i
SE (ANOVA)
0r
*Percentage of total retinol-associated fluorescence. Number in parentheses indicates number of animals. a-dMeans within a row differ: b(p < 0.05); cd(p < 0.01).
Day 13 flushings corresponded with RBP peak 1 and the second with RBP peak 3. Peaks 1 and 3 accounted for 59.9% and 26.8%, respectively, of the total fluorescent material present (Table 1). Peak 4, in particular, was barely detectable. By contrast, anion-exchange chromatography of Day 45 material (n = 5) revealed a significant enhancement of peak 2 (0.14 M NaCI; 11.0 min; 30.3% of total) and peak 4 (0.19 M NaCI; 19.3 min; 34.1% of total; Fig. 3 and Table 1). At Day 45, peaks 2 and 4 were significantly higher (p < 0.05 andp < 0.01, respectively) and peaks 1 and 3 significantly lower (p < 0.01 andp < 0.05, respectively) than at Day 13. Binding of Porcine Uterine Forms of RBP to 7TR Partially purified uterine RBP from Day 45 of pseudopregnancy (i.e., material fractionated by cation-exchange chromatography and gel filtration, but not subjected to anion-exchange chromatography) and purified pig serum RBP were allowed to bind to an excess of human TTR, and the bound and unbound RBP were separated by gel filtration. Pig uterine RBP (Fig. 4) and pig serum RBP (data not shown) eluted at 30 min. When pig serum RBP was allowed to incubate with TTR prior to chromatography, most of the retinol-associated fluorescence eluted at approximately 24 min (data not shown). This shift in elution position corresponded with the expected elution position of a TTR-RBP complex (Mr = 76 000) on a Superose 12 column. Likewise, when partially purified uterine RBP was incubated with TTR, most of the fluorescence eluted at 24 min (Fig. 3). These results indicated that the binding of the uterine RBP to TTR was similar to that of serum RBP. Cloning and Sequencing of the cDNA for Uterine RBP A degenerate series of oligonucleotides was synthesized to represent the full range of codon variation for amino acids 11-18 of the REP in peak 2,3 of Clawitter et al. [7] (Fig. 1). This probe was used to screen more than 3 x 105 plaques of a porcine endometrial cDNA library from Day 60 of pregnancy. No positive plaques were identified. A second oligonucleotide probe (RBP-240) was constructed on
0
10
20
30
40
Minutes
FIG. 4. High-performance gel filtration chromatography of the partially purified uterine RBP (URBP) charge forms bound to TTR. Chromatography was performed on an LKB Superose-12 FPLC column by using a 0.15 M NaCI buffer (0.01 M Tris, pH 8.2). TTR was detected by continuous monitoring of the column effluent at 280 nm. RBP-bound TTR was detected through measurement of retinol by continuous monitoring of the column effluent for fluorescence at an excitation of 325 nm and an emission of 460 nm.
the basis of a very highly conserved segment of polypeptide present in all known mammalian RBP. A total of 21 strongly positive plaques were obtained from 15 x 104 plaques screened. An additional 6 x 10 4 plaques were hybridized to one of the uterine clones, BT13 (the terminal 618 bp of the RBP cDNA), obtained from the first screen with the oligonucleotide. This procedure provided eight additional positive recombinants. Clearly the RBP was relatively abundant in the library. Although nine of these clones were sequenced, only five contained sequences that corresponded to the NH2-terminus portion of the protein. However, the inserts from all nine clones were identical where they overlapped and shared 100% sequence identity with the RBP cDNA from porcine conceptuses whose sequence we have previously reported [14]. Western Blot Analysis Western blot analysis of all four uterine RBP charge forms, as well as purified pig serum RBP, with a rabbit antiserum directed against human serum REP revealed a single major immunoreactive band, M, = 21 000, in each lane (Fig. 5). These bands were not recognized by serum from a rabbit that had not been immunized with human serum RBP. Purification and Sequencing of the Uterine RBP The RBP were purified by gel filtration and anion-exchange procedures (Fig. 2) as previously reported [7] and identified by their characteristic fluorescence. The most basic form was designated peak 1 and the most acidic form
PORCINE UTERINE RBP
1003
FIG. 5. Western blot of the uterine RBP. Approximately 0.4 izg of each peak and serum RBP were analyzed by SDS-PAGE on 12.5% (w/v) gels. Proteins were blotted on PVDF membranes. The RBP were detected by using a 1:400 dilution of either NRS (lanes 1-5) or rabbit anti-human serum RBP (lanes 610). Molecular weight markers are on the right. Peak 1 (lanes 1, 6); peak 2 (lanes 2, 7); peak 3 (lanes 3, 8); peak 4 (lanes 4, 9); serum RBP (lanes 5, 10).
peak 4 (Fig. 2). Since each fluorescent peak represented only vitamin-loaded forms, the vitamin-free forms, previously identified as the protein peaks eluting immediately after the fluorescent peaks [7], were pooled with their respective holoforms. Samples from each combined peak were then analyzed by SDS-PAGE. When such gels were stained by Coomassie Brilliant blue dye under standard conditions, a major band of protein (Mr = 21 000) was detectable in each lane (Fig. 6), which by densitometric scanning ac-
counted for > 80% of the stained material on the gel. The polypeptides on an unfixed and unstained but otherwise identical gel were transferred electrophoretically to a PVDF membrane and then stained (Fig. 6). Numerous protein bands not clearly evident on the polyacrylamide gel were present, and the putative RBP band in this case constituted only about 30% of the total stained protein. The 21 000-Mr bands corresponding to peaks 2 and 4 were then excised and subjected directly to 10 cycles of microsequencing. Such
FIG. 6. Comparison of the uterine RBP analyzed by one-dimensional SDS-PAGE after DEAE chromatography on similarly stained gels and PVDF blots. SDS-PAGE was performed in 12.5% (w/v) gels. Proteins were blotted on PVDF membranes in a 0.025 M Tris, 0.192 M glycine, pH 8.2 buffer. Both gels and blots were stained with Coomassie Brilliant blue dye. The positions of the molecular weight markers are shown on both sides. Approximately 4 ipg of protein were loaded onto each lane. Peak 1 (lanes 1, 5); peak 2 (lanes 2, 6); peak 3 (lanes 3, 7); peak 4 (lanes 4, 8).
1004
STALLINGS-MANN ET AL.
an off-blot procedure was not used by Clawitter et al. [7], who had sequenced the pooled material from each peak, assuming on the basis of electrophoretic analysis that it largely represented pure RBP. The off-blot sequences (Fig. 1) were identical for peak 2 and for the most acidic form, peak 4, of the RBP variants and corresponded exactly with that of serum RBP. The amino acid at position 4, which on the basis of cDNA data was probably cysteine, could not be identified. It became clear, therefore, that the amino acid sequence data previously published [7] were incorrect and that the anomalous sequences noted most probably had arisen from the presence of contaminating proteins in the samples. To confirm this supposition, a sample of protein in peak 2 from the anion-exchange column was subjected directly to amino acid sequencing (Fig. 1). Multiple signals were clearly evident, and the major sequence inferred was distinct from that of RBP. DISCUSSION During the original purification of porcine uterine RBP, several forms of similar molecular weight but distinct charge were identified by anion-exchange chromatography. These charge variants were assumed to correspond to the two to three low-molecular weight, acidic polypeptides that could be routinely visualized on stained gels after two-dimensional electrophoresis of uterine flushings obtained during early pregnancy (Days 12 and 13) [38-40] and pseudopregnancy (Day 45) [7] in the pig. As many as five such spots, differing only slightly in charge, have been noted for porcine conceptus RBP [13]. However, microheterogeneity has also been reported for serum RBP in humans [20-23], rats [24], and chickens [23] and has been attributed to loss of amide groups from the protein, resulting in lowered pI [20, 41]. By contrast, Clawitter et al. [7] suggested that the porcine uterine RBP were a family of unique proteins since the individual members had unique NH2-terminal amino acid sequences. Furthermore, these uterine RBP were unusual in at least two other respects. They were synthesized in large quantities at an extra-hepatic site in response to progesterone, and they were not complexed with TTR. Indeed, TTR cDNA have not been detected in the uterine cDNA library, but they were abundant in a porcine liver library (W. Trout, unpublished results). It was clearly of interest, therefore, to determine whether any functional differences existed between these unusual uterine isoforms. In a first series of experiments, we examined whether the pattern of secretion of uterine RBP charge forms changed over time, reasoning that similar changes in expression might occur during pregnancy and could reflect differing requirements for retinol during fetal development. These experiments clearly showed that there was a shift in the complement of isoforms that were recovered in uterine flushings. At Day 13 of pseudopregnancy, peaks 1 and 3 predominated, whereas peaks 2 and 4 were clearly highest at Day 45.
However, despite their apparent unique sequences, it became clear that the uterine RBP were immunologically related to serum RBP and were also able to bind to the serum carrier protein TTR, although they did not exist in such complexes in the uterine flushings. In serum, TTR probably helps stabilize the retinol-RBP complex and reduces the renal clearance rate [9]. It is considered to bind to highly conserved regions of RBP, since the two proteins can form complexes even when they have been isolated from evolutionarily distant species [42, 43]. In fact, one region of RBP that has been proposed to bind TTR is a highly conserved surface-located segment of 19 amino acids at the NH2-terminus of the molecule [44]. Because the sequences of the uterine RBP appeared to differ from serum RBP in this conserved region, their ability to bind TTR was unexpected. An attempt to clone the cDNA by employing oligonucleotide probes that had been constructed to reflect the unique NH2-terminal sequences of the uterine RBP failed. By contrast, an oligonucleotide designed to represent an internal amino acid sequence within the retinol-binding pocket of all known serum RBP successfully identified numerous cDNA in the uterine library. It was evident that these coded for a single classical RBP and not for the elusive uterine variants. All the above data strongly suggested that the sequences reported by Clawitter et al. [7] were incorrect. It is now clear that although the RBP in each peak was considered to be greater than 80% pure, the content of contaminating proteins had been severely underestimated, probably as a result of inadequate fixation on the electrophoretic gels used to assess purity. Direct off-blot sequencing of the putative RBP bands demonstrated that the most acidic (peak 4) and a basic (peak 2) variant had identical NH 2-terminal sequences and were probably closely related to the liver form of RBP and to the RBP cloned from porcine conceptuses. What, then, is the cause of the multiple RBP isoforms? The most likely explanation is that they arise by loss of amide groups or some other modification that leads to a reduction in their pI while the proteins are within the uterine tract. These changes probably occur progressively while the proteins are in contact with tissue and were, therefore, more extreme in the Day 45 pseudopregnant animals. It seems unlikely that these changes have any physiological significance. ACKNOWLEDGMENTS We thank August Rieke for arrangement of animals, Harriet Francis for help with surgeries, Dr. David Chin for protein sequencing, and Dr. Joe Forrester for synthesis of oligonucleotides.
REFERENCES 1. Roberts RM, Raub TJ, Bazer FW. The role of uteroferrin in transplacental iron transport in the pig. Fed Proc 1986; 45:2513-2518.
PORCINE UTERINE RBP 2. Roberts RM, Bazer FW. The functions of uterine secretions. J Reprod Fertil 1988; 82:875-892. 3. Fazleabas AT, Bazer FW, Roberts RM. Purification and properties of progesteroneinduced plasmin/trypsin inhibitor from uterine secretions of pigs and its immunocytochemical localization in the pregnant uterus. J Biol Chem 1982; 257:68866897. 4. Murray MK, Malathy PV, Bazer FW, Roberts RM. Structural relationship, biosynthesis and immunocytochemical localization of uteroferrin-associated basic glycoproteins. J Biol Chem 1989; 264:4143-4150. 5. Malathy P-V, Imakawa K, Simmen RCM, Roberts RM. Molecular cloning of the uteroferrin-associated protein, a major progesterone-induced serpin secreted by the porcine uterus, and the expression of its mRNA during pregnancy. Mol Endocrinol 1990; 4:428-440. 6. Adams KL, Bazer FW, Roberts RM. Progesterone-induced secretion of a retinolbinding protein in the pig uterus. J Reprod Fertil 1981; 62:39-47. 7. Clawitter J, Trout WE, Burke MG, Araghi S, Roberts RM. A novel family of progesterone-induced, retinol-binding proteins from uterine secretions of the pig. J Biol Chem 1990; 265:3248-3255. 8. Trout WE, HallJA, Stallings-Mann ML, Anthony RV, Galvin JM, Roberts RM. Steroid regulation of the synthesis and secretion of retinol-binding protein by the uterus of the pig. Endocrinology 1992; 130:2557-2564. 9. Blaner WS. Retinol-binding proteins: the serum transport protein for vitamin A. Endocr Rev 1989; 10:308-316. 10. Soprano DR, Soprano KJ,Goodman DS. Retinol-binding protein messenger RNA levels in the liver and in extrahepatic tissue of the rat. J Lipid Res 1986; 27:166171. 11. Makover A, Soprano DR, Wyatt ML, Goodman DS. An in situ-hybridization study of the localization of retinol-binding protein and transthyretin messenger RNAs during fetal development. Differentiation 1989; 40:17-25. 12. Soprano DR, Soprano KJ, Goodman DS. Retinol-binding protein and transthyretin mRNA levels in visceral yolk sac and liver during fetal development in the rat. Proc Natl Acad Sci USA 1986; 83:7330-7334. 13. Harney JP, Mirando MA, Smith LC, Bazer FW. Retinol-binding protein: a major secretory product of the pig conceptus. Biol Reprod 1990; 42:523-532. 14. Trout WE, McDonnell , Kramer KK, Baumbach GA, Roberts RM. The retinolbinding protein of the expanding pig blastocyst: molecular cloning and expression in trophectoderm and embryonic disc. Mol Endocrinol 1991; 5:1533-1540. 15. Liu KH, Brewton RG, Baumbach GA, Godkin JD. Characterization of protein production by ovine placental membranes: identification of a placental retinol-binding protein. Endocrinology 1991; 129:126-132. 16. Liu KH, Gao K, Baumbach GA, Godkin JD. Purification and immunolocalization of ovine placental retinol-binding protein. Biol Reprod 1992; 46:23-29. 17. Liu KH, Baumbach GA, Gillevet PM, GodkinJD. Purification and characterization of bovine placental retinol-binding protein. Endocrinology 1990; 127:2696-2704. 18. Dore JE Jr, Liu KH, Roberts MP. Endometrial and embryonic expression of ovine retinol-binding protein. Biol Reprod 1992; 46(suppl 1):142 (abstract). 19. Thomas PGA, Leslie MV, Hansen PJ. Retinol binding protein is produced by the bovine endometrium and accumulates in uterine secretions in a progesteronedependent manner. Anim Reprod Sci 1992; 27:55-66. 20. Raz A, Shiratori T, Goodman DS. Studies on the protein-protein and proteinligand interactions involved in retinol transport in plasma. J Biol Chem 1970; 245:1903-1912. 21. Peterson PA, Berggard I. Isolation and properties of a human retinol-transporting protein. J Biol Chem 1971; 246:25-33. 22. Peterson PA. Characteristics of a vitamin A-transporting protein complex occurring in human serum. J Biol Chem 1971; 246:34-43. 23. Kopelman M, Mokady S, Cogan U. Comparative studies of human and chicken retinol-binding proteins and prealbumins. Biochim Biophys Acta 1976; 439:442448.
1005
24. Colantuoni V, Romano V, Bensi G, Santoro C, Costanzo F, Raugei G, Cortese R. Cloning and sequencing of a full length cDNA coding for human retinol-binding protein. Nucleic Acids Res 1983; 11:7769-7776. 25. Frank M, Bazer FW, Thatcher WW, Wilcox CJ. A study of prostaglandin F2 as the luteolysin in swine: III. Effects of estradiol valerate on prostaglandin F, progestins, estrone and estradiol concentrations in the utero-ovarian vein of non-pregnant gilts. Prostaglandins 1977; 14:1183-1196. 26. Bazer FW, Sharp DC, Roberts RM. Collection and analysis of female genital tract secretions. In: Daniel JC (ed.), Methods in Mammalian Reproduction. New York: Academic Press; 1978: 503-527. 27. Sundelin J, Laurent BC, Anundi H, Tragardh L, Larhammar D, Bjorck L, Eriksson U, Akerstrom B, Jones A, Newcomer M, Peterson PA, Rask L.Amino acid sequence homologies between rabbit, rat and human serum retinol-binding proteins. J Biol Chem 1985; 260:6472-6480. 28. Messing J. New M13 vectors for cloning. Methods Enzymol 1983; 101:20-78. 29. Sambrook J, Fritsch EF, Maniatis T. Molecular Cloning-A Laboratory Manual, 2nd ed. Cold Spring Harbor: Cold Spring Harbor Laboratory; 1989. 30. Fernberg AP, Vogelstein B. A technique for radiolabeling DNA restriction endonuclease fragments to high specific activity. Anal Biochem 1983; 132:6-13. 31. Simmen RCM, Baumbach GA, Roberts RM. Molecular cloning and temporal expression during pregnancy of the messenger ribonucleic acid encoding uteroferrin, a progesterone-induced uterine secretory protein. Mol Endocrinol 1989; 2:253-262. 32 32. Davis LG, Dibner MD, Battey JF. Hybridization with synthetic p end-labeled probe. In: Basic Methods in Molecular Biology. New York: Elsevier Science Publ. Corp.; 1986: 75-78. 33. McCabe PC. Production of single-stranded DNA by asymmetric PCR In: Innis MA, Gelfard DH, Sninsky , White TJ (eds.), PCR Protocols: A Guide to Methods and Applications. San Diego: Academic Press; 1990: 76-83. 34. Sanger F, Nicklen S, Coulson AR. DNA sequencing with chain-terminating inhibitors. Proc Natl Acad Sci 1977; 74:5463-5467. 35. Chen E, Seeburg PH. Supercoil sequencing: a fast and simple method for sequencing plasmid. DNA 1985; 4:165-170. 36. Snedecor GW, Cochran WG. Statistical Methods, 6th ed. Ames, IA: Iowa State University Press; 1967. 37. Rask L, Anundi H, Peterson PA. The primary structure of the human retinol-binding protein. FEBS Lett 1979; 104:55-58. 38. Geisert RD, Renegar RH, Thatcher WW, Roberts RM, Bazer FW. Establishment of pregnancy in the pig. I. Interrelationships between preimplantation development of the pig blastocyst and uterine endometrial secretions. Biol Reprod 1982; 27:925939. 39. Geisert RD, Zavy MT, Wettemann RB, Biggers BG. Length of pseudopregnancy and pattern of uterine protein release as influenced by time and duration of oestrogen administration in the pig. J Reprod Fertil 1987; 79:163-172. 40. Morgan GL, Geisert RD, Zavy MT, Shawley RV, Fazleabas AT. Development of pig blastocysts in a uterine environment advanced by exogenous oestrogen. J Reprod Fertil 1987; 80:125-131. 41. Vahlquist A, Nilsson SF, Peterson PA. Isolation of the human retinol binding protein by affinity chromatography. Eur J Biochem 1971; 20:160-168. 42. Berni R, Stoppini M, Zapponi MC, Meloni ML, Monaco HL, Zanotti G. The bovine plasma retinol-binding proteins and pre-albumins studied by fluorescence polarization. Eur J Biochem 1990; 192:507-513. 43. Kopelman M, Cogan U, Mokady S, Shinitzky M. The interaction between retinolbinding proteins and pre-albumins studied by fluorescence polarization. Biochim Biophys Acta 1976; 439:449-460. 44. Cowan SW, Newcomer ME, Jones TA Crystallographic refinement of human serum retinol binding protein at 2A resolution. Proteins 1990; 8:44-61.
