Keywords: uterine fluid, porcine, uterine protein, platelet aggregation, lymphocyte blastogenesis, protein purification

ABSTRACT

Microheterogeneous retinol-binding proteins (RBPs) are secreted by the porcine uterus under the influence of progesterone. Reported functional capacities of these proteins have included their ability to bind retinol and immunosuppressive effects on mitogen-induced T-lymphocytes. In this study, porcine uterine RBPs were isolated, purified, and tested for effects on mitogen-induced lymphocyte proliferation and thrombin-induced platelet aggregation. The proteins were purified and two charge forms were tested for activity. Uterine flushings were collected on Day 15 after estrus and fractionated by Sephadex G-75. Subsequent fractionation by Mono P HR 5/20 fast protein liquid chromatography (FPLC) chromatofocusing yielded two peaks containing two bands each as determined by native-polyacrylamide gel electrophoresis (PAGE). Fractionation by Mono Q HR 5/5 anion exchange column resulted in an identical separation pattern. Slab isoelectric focusing revealed four bands with a pI range of 4.7 to 5.2. Fractionation of all four proteins with a preparative isoelectric focusing cell resulted in isolation of two of the bands to homogeneity as determined by sodium dodecyl sulfate (SDS)- and native-PAGE. Individually each of the two bands, as well as the purified set of uterine RBPs, completely inhibited thrombin-induced human platelet aggregation at physiological levels (13.6 µM).

INTRODUCTION

The uterus of the pig secretes a variety of proteins. Under the influence of progesterone, a large percentage of the protein profile consists of a family of low molecular weight (19,000 - 22,000) acidic proteins which have been shown to bind retinol (1,2). Porcine uterine retinol-binding proteins (RBPs) are heterogeneous as revealed by native-PAGE (3,4,5) and two-dimensional-(2D)-SDS-PAGE (6,7). Several distinct isoelectric variants have been reported ranging in number from four (3,4,7) to six (5,6). Microheterogeneity has also been demonstrated to occur with serum RBPs of the human (8,9), rat (10) and chicken (11), placental RBPs of the cow (12), sheep (13) and human (14), as well as RBPs from the porcine conceptus (15). Recently it has been reported that all charge forms of uterine RBPs from the pig were recognized by antiserum to human serum RBPs on western blots (16). Two of the more divergent charge forms of uterine RBPs showed complete amino acid sequence identity with pig serum RBP. It has been suggested that the uterine RBPs charge forms may be slightly modified forms of a single protein product corresponding to a classical form of RBPs (16).

During pregnancy, porcine uterine RBPs are believed to function to transport retinol from the maternal circulation to the conceptus (17). The pig has a non-invasive epitheliochorial placentation. Because of the superficial attachment of the trophoblast, the conceptus is unable to obtain nurients directly from the maternal blood supply. Instead, the conceptus must depend on carrier proteins from the uterine epithelium to supply nutrient requirements. The proper concentration of retinol is vital to both cell differentiation and embryonic development (reviewed in (18)).

In addition to the passive carrier role of RBP, studies have indicated that this protein may also play an active role in cell regulation. For example, although there is conflicting evidence (19,20), this protein family has been reported to inhibit mitogen-induced lymphocyte proliferation (21,22). Uteroglobin from the rabbit is a protein similar to porcine uterine RBP. Like uterine RBP, it has been shown to be secreted under the influence of progesterone and found to be a major protein component of uterine fluid during early pregnancy at the time of implantation (23,24). It is a low molecular weight acidic protein which has been shown to both bind progesterone (25), as well as effect cell function (26,27,28). Uteroglobin has been demonstrated to contain anti-inflammatory properties based on its ability to inhibit both monocyte and neutrophil chemotaxis and phagocytosis in vitro (26,27). It has also been shown to inhibit thrombin-induced platelet aggregation (28). These properties are attributed to its potent ability to inhibit phospholipase A₂ (PLA₂) (29). The aim of this study was to purify porcine uterine RBPs to homogeniety and to test their effects on thrombin-induced platelet aggregation and mitogen-induced lymphocyte proliferation.

MATERIALS AND METHODS

Collection of Porcine Uterine Fluid

Uterine flushings were collected from mature cycling pigs of Yorkshire, Duroc and Landrace breeds on Day 15 after estrus. The flushings were collected according to the method as described by Segerson et al. (19). Reproductive tracts were surgically removed postmortem and each horn was subsequently flushed with 50 ml sterile 0.01 M PBS (pH 7.4, 4°C) to collect the accumulated proteins (19). Flushings were stored at -70°C.

Gel Filtration Chromatography

Each uterine flushing was membrane filtered (0.2 µm) and concentrated overnight via vacuum concentration using cellulose dialysis tubing. Concentrated samples were applied to a 72 x 2.6 cm column of Sephadex G-75 (Pharmacia LKB, Piscataway, NJ), equilibrated in 0.01 M phosphate buffered saline (PBS), pH 7.4 and run with a mean flow rate of 5 ml/h into 5 ml fractions. Column eluents were monitored and all protein concentrations were determined by method of Lowry et al. (30) using bovine serum albumin (BSA) as a standard. Fractions eluting between 250-300 ml were pooled, membrane sterilized, concentrated by YC05 Amicon ultrafiltration (Amicon, Danvers, MA) and stored at -135°C. Deionized water was used to prepare all solutions. All purification procedures were carried out at 4°C.

Electrophoresis

The presence of RBPs was verified by native-PAGE on a PhastSystem (Pharmacia LKB) using a PhastGel (8-25%) and separated for 220 Vh. Fractions were analyzed by SDS-PAGE using a PhastGel (8-25%) and separated for 110 Vh. The SDS-PAGE program on a PhastSystem was modified using a constant 220 V in replacement of 250 V. Both native and SDS gels were developed with Coomassie brilliant blue (PhastGel Blue R) using the staining method for the native-PAGE PhastSystem program. Molecular weight markers (Sigma, St. Louis, MO) included BSA (66,000), ovalbumin (45,000), glyceraldehyde-3-phosphate dehydrogenase (36,000), carbonic anhydrase (29,000), trypsinogen (24,000), trypsin inhibitor (20,100) and -lactalbumin (14,200). The isoelectric points of the RBPs were determined on a PhastSystem using PhastGel isoelectric focusing (IEF) media. The PhastGel IEF 4-6.5 was prefocused for 75 Vh at 2000 V, 2 mA. Gels were run at 2000 V limiting, 3.5 W and 5 mA for 654 Vh (~32 min). The pH gradient was determined using calibrated pI markers pH 2.5-6.5 (Pharmacia LKB). Standards included human carbonic anhydrase B (6.55), bovine carbonic anhydrase B (5.85), ß-lactoglobulin A (5.20), soybean trypsin inhibitor (4.55), glucose oxidase (4.15) and amyloglucosidase (3.50).

Fast Protein Liquid Chromatography (FPLC)

Chromatofocusing was performed on a FPLC Mono P HR 5/20 column (Pharmacia LKB) equilibrated in 0.025 M bis Tris, pH 6.7 at 20°C. Equilibrated RBPs Sephadex G-75 aliquots were separated using a 0.5 ml/min flow rate, a 0.1 AUFS, a 0.5 cm/min chart speed, and collected in 0.5 ml fractions. The column was eluted with 10% polybuffer 74, pH 5.0 creating a 7-5.0 pH gradient. The absorbance of column effluent was monitored continuously at 280 nm.

Anion exchange was performed on a FPLC Mono Q HR 5/5 column (Pharmacia LKB) at 20°C. Pooled Sephadex G-75 fractions containing RBPs were equilibrated over Superose 12 hr and applied in 500 µl aliquots. The Mono Q HR 5/5 column was equilibrated in 50 mM Tris-HCl, 1 mM dithiothreitol, pH 8.0 and eluted with a linear gradient of 0-0.5 M NaCl in 30 min. The elution buffer consisted of the starting buffer containing 1 M NaCl. The column was run using a 0.5 ml/min flow rate, a 0.1 AUFS, a 0.5 cm/min chart speed and collected in 0.5 ml fractions. All FPLC column fractions containing protein were concentrated in a SpeedVac (Savant, Farmingdale, N.Y.), and extensively dialyzed against 0.01 M PBS.

Preparative Isoelectric Focusing

Final purification of uterine RBP was achieved within a pH 4.5-5.4 ampholyte gradient using a Rotofor Cell (Bio-Rad Laboratories, Richmond, CA). Samples of up to 15 mg of the G-75 fractionated RBPs fractions were desalted into deionized water over a PD-10 column (Pharmacia LKB). The RBP, in a 1.1% ampholyte solution (55 ml) was loaded into the electrofocusing Rotofor Cell and focused for 4 h at 4°C at a constant 12 W. The final parameters of 1400 V and 8 mA were held for at least 2 h. After focusing, 20 fractions were harvested and immediately concentrated by a SpeedVac concentrator. Fractions were incubated at room temperature for 30 min in 1 M NaCl and dialyzed against 2 liters 0.01 M PBS (4 changes in18 h). Each sample was then fractionated over a 52 x 2.5 cm Sephadex G-50 (Pharmacia LKB) column equilibrated in 0.01 M PBS at 4°C. Experimental conditions were identical to those described for G-75 gel filtration.

Lymphocyte Proliferation Assay

Phytohemagglutinin-induced blastogenesis of porcine lymphocytes in the presence of a dose range of porcine uterine flushings (PUF) was conducted with minor variations of methods previously described by Murray et al. (21). Methodological differences include the following. Lymphocytes were isolated with Sepracell-MN. Optimal concentrations of lymphocytes (for maximal blastogenesis) were determined to be 1.0 to 2.5 x 10⁶ cells/ml with phytohemagglutinin at 5 µg/ml. At 48 hours of culture, 0.5 µCi [³H]-thymidine was added to each culture well (microtiter plate). Counts were done in Ready Safe (Beckman Instruments) in a Beckman LS8000 liquid scintillation counter. Results were expressed in comparison to maximally stimulated control cultures less background, as a percent of the control.

Platelet Aggregation Assay

Human platelets were used in these studies because of their ready availability and because of the reliability of our aggregation assays which were developed for human platelets. Human blood was collected by venipuncture from healthy volunteers who reported to be free of medication for at least 10 days prior to donation. Coagulation was prevented with acid citrate dextrose (85 mM trisodium citrate, 71 mM citric acid, 111 mM dextrose) in a ratio of 1:6 (v/v). Whole blood was centrifuged at 120 x g for 15 min at room temperature and the plasma drawn off and recentrifuged at 1100 x g for 15 min. The resulting platelet pellet was resuspended in 9.5 ml calcium-free modified Tyrode's buffer (pH 6.5) containing 2 mM MgCl₂ (31). Accumulation of ADP was prevented by adding 0.1 ml (60 µg/ml) of apyrase per wash. To remove calcium, EGTA (5.26 mM final concentration) was added prior to centrifugation of the first wash only. Following resuspension, the platelets were centrifuged at 1100 x g for 10 min at room temperature. The wash procedure was repeated three times and the final platelet pellet was resuspended in Tyrode's buffer (pH 7.4) with 2 mM CaCl₂ and 1 mM MgCl₂ (31). Platelet aggregation studies were performed using a dual channel aggregometer (Chrong-log Corporation) interfaced to an Apple II Microcomputer (32). An Epson FX-80 printer was used to produce continuous recordings of platelet aggregation curves. All aggregation experiments were performed in siliconized glass cuvettes containing 0.35 ml platelet suspension (3 x 10⁸ platelets/ml). Uterine RBPs were preincubated with platelets for 1 min at 37°C under constant stirring at 1000 rpm. Parallel control aggregation experiments were conducted in the absence of RBPs but with 0.35 ml of 0.01 M PBS buffer to maintain volume equivalence in control and treated assays. An empirically determined amount (0.014 U/ml) of thrombin (Sigma, St. Louis, MO) was added to induce an approximately 60% aggregation response in the control. The reaction was measured over four min. Aggregation was quantitated by light transmittance and reported as percentage of maximum transmittance. Two controls were used in each assay: PBS and RBP dialysate. Dialysate was obtained by removing RBPs from the purification buffer by amicon filtration and was used as a control in every experiment. Sample protein was also heat inactivated (56°C for 30 min) and tested for its effect on thrombin-induced platelet aggregation.

RESULTS

Gel Filtration Chromatography

Porcine uterine secretions were chromatographed by Sephadex G-75 into three protein fractions (Figure 1). Uterine RBPs eluted in fractions 50-59 (250 and 300 ml) as determined by native-PAGE. A total of up to 30 mg of protein was recovered in this set of fractions (designated peak 3) obtained from flushings of the uterus of a single Day 15 cycling gilt. The recovery of RBPs from this first step represented a yield of 22-30% of total protein from crude flushings. Peak 3 was concentrated for further fractionation.

Electrophoresis

A typical native-PAGE pattern of the uterine starting material and RBPs following Sephadex G-75 fractionation are shown in (Figure 2), lane 1 and 2, respectively. The RBPs consistently appeared as 4 bands, although up to 6 bands have been observed in individual samples. SDS-PAGE analysis of these samples revealed a molecular weight of 22,000 for RBP.

Isoelectric Focusing

The isoelectric points (pIs) of RBPs were determined using a PhastSystem and PhastGel IEF 6-4.5. The isoelectric focusing pattern of RBPs under non-reducing conditions revealed four bands focusing between pH 5.2 and pH 4.6 as shown in (Figure 3). The IEF pattern of RBPs following fractionation by Sephadex G-75 is shown in Fig. 3, lane 2. RBPs which was subsequently fractionated by the Rotofor cell is shown in Fig. 3, lane 3. The pH gradient was determined using calibrated low pI standards (lanes 1 and 4), and the pI of the RBPs were determined to be 5.20, 4.96, 4.90 and 4.75 for the four isoforms present in porcine uterine secretions.

FPLC Chromatofocusing

Proteins were eluted from a Mono P HR 5/20 FPLC column as two peaks by employing a pH gradient 7-5.0 as shown in (Figure 4). Native-PAGE analysis showed peak 1 contained the two least acidic proteins (Figure 4, lane 1) and peak 2 contained the two most acidic proteins (Figure 4, lane 2). The least acidic proteins in peak 1 eluted at pH 5.6 and the most acidic proteins at pH 5.1.

FPLC Anion Exchange Chromatography

Fractionation of pooled Superose-12 FPLC RBPs over a Mono Q HR 5/5 column revealed similar results. (Figure 5) shows bound proteins were eluted with two peaks with a linear gradient of 0-0.5 M NaCl. Peak 1 was eluted at 0.13-0.15 M NaCl and peak 2 was eluted at 0.23-0.24 M NaCl. Analysis by native-PAGE are shown in (Figure 6A) which revealed peaks 1 and 2 contained the two least acidic proteins (lane 1) and the two most acidic proteins (lane 2), respectively. SDS-PAGE analysis revealed that peaks 1 and 2 each migrated as a single band as shown in (Figure 6B) lanes 1 and 2, respectively.

Preparative Isoelectric Focusing

Purification of RBPs by a Rotofor cell resulted in the highest resolution of the four protein bands. (Figure 7) shows the pH profile and a native-PAGE analysis of Sephadex G-75 fractionated RBPs (lanes 1 and 8) along with Rotofor fractions 14-9 (lanes 2-7, respectively). Rotofor fractions 16 and 4 consisted of single bands as determined by native-PAGE as shown in (Figure 8, lane 1 and 3), respectively. The homogeneous band in fraction 16 was eluted at pH 5.2 and fraction 4 was eluted at pH 4.8. Refractionation of fractions 5-15 (Figure 8, lane 2) at a narrow pH range of 4.8-5.2 did not result in resolution of the central bands. Single band homogeneity of fractions 16 and 4 had been achieved as shown in lane 1 and 3, respectively.

In one run, a total of 13.6 mg of G-75 fractionated RBPs was loaded on the Rotofor. A recovery of 2.38 mg in fractions 5-15 was obtained consisting of only four RBPs bands. This represents a yield of 17.5% of the RBPs from the Rotofor and 3.9% from the original flushing from one porcine uterus. The yield of RBPs varied between Rotofor preparations, but final overall yields of up to 9.5% have been obtained. The fractions exhibiting one homogeneous band each, contained 0.52 mg in fraction 4 and 0.26 mg in fraction 16 representing a 3.8% and 1.9% yield from a Rotofor fractionation, respectively and an overall yield of 0.84% and 0.42%, respectively.

Effect of PUF and RBPs on Lymphocyte Proliferation

A total of nine blastogenesis experiments with PUF and/or RBPs derived from as many animals were conducted. In no case did an inhibition of blastogenesis occur, at any concentration, due to either PUF or RBP. In spite of intensive effort to duplicate conditions of the previous experiments (21,22), we did not find suppression of PHA-induced blastogenesis. In fact, both PUF and RBPs were consistently stimulatory to blastogenesis compared with controls. However, the stimulatory activity was not related to concentration of either PUF or RBPs. Results of these experiments are summarized in (Table 1).

Platelet Aggregation Assay

A Rotofor preparation of four pooled RBPs and two individual isoforms of RBP, were tested for their effects on thrombin-induced human platelet aggregation. The mixture of four RBPs inhibited thrombin-induced aggregation in washed human platelets in a concentration dependent manner (Figure 9). Complete inhibition was achieved with 13.6 µM RBP. The most acidic (fraction 4) and the least acidic (fraction 16) RBPs both caused complete inhibition of thrombin-induced platelet aggregation at 13.6 µM as shown in (Figure 10 A and 10B.). Limited quantities of the RBPs isoforms precluded running dose response curves for these proteins. Platelet aggregation was not inhibited in any experiment when platelets were incubated with 70µl PBS alone, 70µl dialysate, or 13.6 µM heat inactivated sample protein (data not shown).

DISCUSSION

Porcine uterine RBPs consists of a family of microheterogeneous proteins of similar molecular weight with four distinct isoelectric variants. The heterogeneous RBPs family as well as two individual isoelectric variants were isolated by a three-step-procedure. Fractionation of uterine secretions by Sephadex G-75 gel filtration resulted in the isolation of the small molecular weight proteins. Preparative isoelectric focusing using a Rotofor produced purified heterogeneous RBPs and allowed isolation of homogeneous preparations of the most basic and most acidic variants. Yields of 3.9% of total uterine protein were acheived for the heterogeneous RBPs and 0.42% to 0.84% for individual isoforms.

Analytical isolation of uterine RBPs from the pig has been reported using gel filtration, batch treatment with CM-cellulose and anion exhange chromotography (2). Purification methods of RBPs involving repetitive chromotographic steps typically have yielded a low recovery of protein (2,33). In addition to the high yield of uterine RBPs achieved with the procedure described here, this is the first report of the purification of a single, homogeneous porcine uterine RBPs isoelectric variant.

The isoelectric measurements of the four RBPs ranged from 5.2-4.8 as determined by slab gel isoelectric focusing. These pI measurements were an order of magnitude lower than previously reported measurements of 6.1 to 6.3 as determined by 2D-SDS-PAGE (3,7). They are also lower than the isoelectric points reported for porcine conceptus RBPs of 5.6 to 6.5 also determined by 2D-SDS-PAGE (15). Subsequent analysis of the proteins by FPLC anion exchange chromatography revealed two major peaks which eluted at 0.14 M NaCl and 0.23 M NaCl. Each peak contained two charge variants as revealed by native-PAGE. These charge variants correspond to the ionic charge determination reported by Stallings-Mann et al. (16).

In previous studies by the authors (21,22), PUF and the fraction we now call RBPs were suppressive to porcine lymphocyte blastogenesis. Other, more recent experiments, have not supported these observations. For example, Segerson et al. (19,20) reported suppression of lymphocyte blastogenesis by a 230 kD protein from PUF but no inhibition by proteins ranging below 50 kD. Our results in the current study are in agreement with the more recent reports (19,20), but differ from them in that we find no suppression by unfractionated PUF. The reasons for the disparity regarding effect on lymphocyte blastogenesis between the earlier sets of experiments, those by Segerson et al. (19,20), and the current report are not apparent.

Secretion of RBPs by the endometrium during the peri-implantation period may be important for local transport of retinoids to the developing conceptus (1). Various effects of retinoids on tissue and cell function have been demonstrated (18,34,35). But the effects of RBPs on cell regulation remain to be elucidated. Several uterine proteins have demonstrated cell regulatory action. For example, uteroferrin, secreted from the porcine endometrium, has been shown to be a hematopoietic growth factor (36) and uteroglobin has been shown to contain anti-chemotactic and anti-inflammatory activities (26,27). Uteroglobin has also been shown to inhibit thrombin-induced platelet aggregation (28). These properties are attributed to their ability to inhibit PLA₂ activity (29).

In this study, porcine uterine RBPs inhibited thrombin-induced platelet aggregation in a dose dependent mannner. It is unlikely that these inhibitory effects may be attributed to the action of retinol or retinol bound to RBP. Retinol has been demonstrated to stimulate thrombin-induced human platelet aggregation through the activation of PLA₂ (37). Inhibition of platelet aggregation by RBPs was eliminated by a heat treatment of 56°C for 30 min, and there was no significant difference between platelet aggregation incubated without sample protein or with the negative control (RBP dialysate). These results indicate that the RBP proteins were responsible for the effects demonstrated and were not caused by an artifact of the buffer.

Furthermore, two RBP variants, purified to homogeneity, individually caused complete inhibition of thrombin-induced platelet aggregation at 13.6 µM (Fig. 10). The activity of these individual proteins appear to be similar to each other and to RBPs as a group. These results are consistent with the report that the uterine RBP variants are slightly modified forms of a single protein product corresponding to serum RBP (16) and that the isoforms are likely to be functionally similar as well.

The effect of serum RBP on platelet activation in vivo has not been reported. Normal levels of RBP in human serum range between approximately 40-60 µg/ml (38) as opposed to 200 µg/ml (6) in the uterus. This represents 13-20% of the amount of uterine RBPs reported here to completely inhibit thrombin-induced platelet aggregation. In addition, RBP is typically bound to the plasma protein, transthyretin, in the blood (38) and thus may not be available to express anti-platelet properties in the circulation. In contrast, RBPs were not found to be complexed with transthyretin in the porcine uterus (16). The results of this study indicate that porcine uterine RBPs may potentially function as an inhibitor of cell action and specifically block platelet aggregation locally in the uterus. Inhibition of platelet aggregation may be important during implantation and later in pregnancy to maintain blood flow in the placental capillary beds.

CONCLUSIONS

Porcine uterine secretions contain four isoforms of retinol binding protein (RBPs). Gel permeation chromatography with Sephadex G-75 isolated the four RBP isoforms from the other major proteins of the uterine secretions. Isoelectric points were determined by isoelectric focusing to be 5.20, 4.96, 4.90, and 4.75, for the different isoforms. FPLC chromatofocusing resolved the RBPs into two fractions, one of which was composed primarily of the less acidic isoforms (pI 5.20 and 4.96) and another which contained the more acidic isoforms (pI 4.90 and 4.75). Similarly, FPLC anion exchange chromatography resolved RBPs into two fractions with similar composition to the fractions resulting from chromatofocusing. Preparative isoelectric focusing with a narrow pH gradient (pH 4.5 to 5.4) in a Rotofer cell resolved the RBPs into a set of 16 fractions, of which fractions 4 and 16 were nearly homogenous for isoforms with pIs of 4.75 and 5.20, respectively. Fractions 5 through 15 were mixtures of the four isoforms.

This study did little to clarify the issue of immunosuppressive activity in porcine uterine secretions. In contrast to studies of porcine uterine RBPs reporting inhibition of lymphocyte blastogenesis, e.g., (21,22), and in agreement with those which have reported no suppression or even stimulation of blastogenesis, e.g., (19,20), this study found no suppression and slight stimulation of blastogenesis by RBPs. However, in contrast to other reports (19,20), we found that unfractionated PUF was not suppressive. In this study the effects of unfractioned uterine secretions and Sephadex G-75 fractionated RBPs on lymphocyte blastogenesis were stimulatory compared with controls but not in a dose-related manner.

When assayed for effect on human platelet aggregation, RBPs were inhibitory in a dose-related manner. 13.6 µM RBP completely inhibited thrombin-induced platelet aggregation. Isolated isoforms of RBP (13.6 µM) were completely inhibitory to thrombin-induced platelet aggregation, suggesting that all four isoforms are inhibitory to platelet aggregation. Inhibition of platelet aggregation may be an important function of uterine RBPs during implantation and placentation.

ACKNOWLEDGEMENTS

This research was supported by the College of Arts and Sciences, the Interdisciplinary Doctoral Program in Molecular and Cellular Biology, and the College of Osteopathic Medicine, as well as a research grant from the Ohio University Research Committee.

REFERENCES

© 1996 Epress Inc.