| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
a Department of Physiology,
b Division of Comparative Medicine, Department of Medical Biochemistry and Microbiology,
c Section of Medical Biochemistry, and Department of Medical Cell Biology, Uppsala University, 751 23 Uppsala, Sweden
ABSTRACT
This study has established the rat as an animal model for the analysis of the biological role of fetal antigen 1 (FA1), a protein previously described in humans and mice. FA1 was purified from rat amniotic fluid by immunospecific affinity chromatography. Immunochemical identity between mouse and rat FA1 was established by crossed tandem immunoelectrophoresis. Molecular size was analyzed by mass spectrometry (33 kDa). The amino acid composition was determined, and the amino acid sequence was analyzed. The overall amino acid composition and sequence of the 28 first N-terminal amino acids were identical to the corresponding parts of rat preadipocyte factor 1 and rat adrenal zona glomerulosa protein. Extensive sequence similarity was found between rat and mouse FA1 (86%) and between rat and human FA1 (82%). The concentration of FA1 in fetal serum, maternal serum, urine, and amniotic fluid in rats was determined using an ELISA. The highest concentrations were found in fetal serum and amniotic fluid around Day 18 of pregnancy. This is the first report on the physicochemical characteristics and compartmentalization of rat FA1.
conceptus, developmental biology, pregnancy
INTRODUCTION
Human fetal antigen 1 (hFA1), isolated from second trimester aminiotic fluid, was first described by Fay et al. [1]. Human FA1 is an epidermal-growth-factor-like glycoprotein present in fetal hepatocytes, fetal adrenal cortex, and glandular cells of the early pancreas primordium and around the blood islands of the yolk sac [2–6]. During pregnancy, high levels of hFA1 have been identified in maternal and fetal circulation and in urine and amniotic fluid [1, 2]. In adult tissues, hFA1 is much less abundant but is present in insulin-producing beta cells of the pancreas, the pituitary gland, and the adrenal cortex [2, 6]. Human FA1 is expressed in neuroendocrine tumors, and its identity to the gene products of delta-like (dlk) and human adrenal-specific (pG2) mRNA has been demonstrated by Jensen and et al. [2].
The biological role of hFA1 is unknown, and a murine model for the study of FA1 has recently been developed [7]. Mouse FA1 (mFA1) is a single chain glycoprotein with a molecular mass of 42–50 kDa, and the N-terminal amino acid sequence has been reported to exhibit 74% identity to hFA1 and 100% identity to the translated cDNAs referred to as mouse delta-like protein (dlk), preadipocyte factor 1 (pref-1), and stromal-cell-derived protein 1 (SCP-1) [8]. Mouse FA1 is thus the secreted, processed molecule encoded by the mRNA defined by these identical mouse cDNAs [7]. At midpregnancy, the maternal serum level of mFA1 increases and the concentration in circulation is positively correlated with the number of fetuses [7].
The aim of the present study was to expand the model studies and to analyze the biological mechanisms involved in the metabolism of FA1. For practical reasons, we established a rat model for FA1. This is thus the first report on the isolation, purification, and characterization of rat FA1 (rFA1). We also quantified rFA1 in amniotic fluid, serum, and urine samples from pregnant and nonpregnant female rats and in serum from rat fetuses and neonates.
MATERIALS AND METHODS
Animals
Virgin female outbred Sprague-Dawley rats (B&K Universal, Sollentuna, Sweden; body weight, 200–280 g) were housed in groups of four in Macrolon cages (Techniplast, Buguggia, Italy) size 4 with free access to tap water. The rats were fed a standard diet (R36; Lactamin, Stockholm, Sweden) ad lib. The temperature was maintained at 21°C ± 1°C, and the relative humidity was maintained at 50% ± 15%. The lighting regime was 12L:12D.
The animals were housed with males overnight, and mating was confirmed by analysis of vaginal smears the following morning. The day following the overnight mating was designated Day 0 of pregnancy.
The mated animals were transferred to individual metabolism cages (Scanbur, Køge, Denmark) and kept for 16–20 h for collection of urine. Before the rats were returned to conventional cages, blood samples were obtained from the tip of the tail into heparinized tubes.
For antigen preparation, rats (n = 41) were killed on Days 12–21 of pregnancy by an overdose of CO2, and blood was collected by decapitation. The amniotic membranes were washed, punctured, and emptied in a petri dish, and the fluid was transferred to plastic tubes. The amniotic fluid was centrifuged (3000 rpm, 10 min, 4°C) to remove debris and red blood cells. The supernatant was kept at -20°C until used. Blood was collected from fetuses and neonates by decapitation. Serum was separated and stored at -20°C until used.
Antigen
Rat FA1 was purified from amniotic fluid obtained at Days 12–21 of pregnancy. The antigen was purified by affinity chromatography as described by Bachmann et al. [7] using affinity-purified rabbit anti-mFA1 (kindly donated by Prof. B. Teisner, University of Odense, Odense, Denmark) coupled to cyanogen bromide (CNBr)-activated Sepharose (Pharmacia, Uppsala, Sweden) using the manufacturer's standard protocol.
Antisera
Antisera against rat amniotic fluid (rAF) and purified rFA1, respectively, were prepared in New Zealand White rabbits (2.5 kg, SVA, Uppsala, Sweden). Pooled rat amniotic fluid (0.2 ml) or purified rFA1 (~25 µg) was mixed with 0.5 ml Freund's complete adjuvant (Sigma, St. Louis, MO) and injected i.m. into rabbits. Two weeks later, the rabbits were injected with the same amount of the respective antigen emulsified with Freund's incomplete adjuvant (Sigma). This procedure was repeated twice at 2-wk intervals. Blood was collected from the rabbits before they received the booster injections, and rabbits were exsanguinated during anesthesia 10 days after the final injection.
The antiserum against rFA1 was absorbed liquid phase to monospecificity [9] with adult rat serum and nonadsorbed material from affinity chromatography of amniotic fluid.
Immunoelectrophoresis
Crossed immunoelectrophoresis [10] and rocket immunoelectrophoresis [11] were performed essentially as described by Jensen et al. [6] and Fay et al. [1]. Rabbit anti-mFA1 antiserum [7] or rabbit anti-rFA1 or rabbit anti-rAF at a concentration of 1:25 (v:v); 1:50 (v:v); or 1:125 (v:v), respectively, were used in the gels. Mouse FA1 and rAF or purified rFA1 were used as antigens. Rocket immunoelectrophoresis was used to monitor rFA1 during affinity chromatography.
Crossed tandem immunoelectrophoresis [12] was performed as described by Krogh and Hau [13] using rabbit anti-rFA1 or rabbit anti-mFA1 [7] at a concentration of 1:25 antibody:gel (v:v) in the two-dimensional gels. Undiluted amniotic fluid from pregnant mice and rats (10 µl) was used as antigen.
Enzyme-Linked Immunosorbent Assay
The ELISA [14] was based on mouse or rat reagents and performed in microtiter plates as a capture assay as previously described [7]. The plates were incubated overnight with test samples and calibrator (rAF in twofold dilution series). Biotinylated affinity-purified rabbit anti-mFA1 was used as detector antibody, and peroxidase-conjugated streptavidin (Sigma), H2O2, and O-phenylendiamine (Sigma) were used to develop the reaction.
Mass Spectrometry
Purified rFA1 was analyzed by MALDI mass spectrometry (Kratos III; Kratos, Manchester, UK) according to the manufacturer's manual.
Amino Acid Analysis
To determine the amino acid composition of rFA1, 170 µl of the sample solution that had been purified by immunospecific affinity chromatography was lyophilized in a hydrolysis tube. To the residue was added 2 ml 6 M HCl containing 20 nmol norleucine as an internal standard and 1 mg/ml reagent grade phenol. Hydrolysis was carried out for 24 h at 110°C in a thoroughly evacuated and sealed Pyrex hydrolysis tube.
The hydrolysate was evaporated on a rotary evaporator at 45°C, and the residue was dissolved in 200 µl pH 2.2 application buffer. An aliquot of 100 µl was analyzed with an amino acid analyzer (Alpha-Plus; LKB, Stockholm, Sweden) using the standard protein hydrolysate program with sodium citrate buffers and ninhydrin detection.
Data were collected with a CR2-AX integrator (Shimadzu, Kyoto, Japan). The values for threonine, serine, glucosamine, and galactosamine were corrected for hydrolytic loss using the standard recovery values of 0.96, 0.90, 0.50, and 0.50, respectively. The results were normalized on the basis of the recovery of the internal standard, norleucine. The values for half-cystine and methionine were only estimates because a performic acid oxidation procedure was not performed.
Amino acid sequencing of the 28 first N-terminal amino acids of purified rFA1 was performed using a protein sequencer (Applied Biosystems 477A Protein Sequencer; Pharmacia) as described in the manufacturer's manual.
Statistics
An analysis of variance was used for statistical analyses, and P values <0.05 were considered significant.
Ethical Permit
The immunization and other experimental protocols involving animals were approved by the regional laboratory animal ethics committee.
RESULTS
Immunochemical Analysis of mFA1 and rFA1 by Crossed Immunoelectrophoresis and Tandem Crossed Immunoelectrophoresis
Rat FA1 (10 µl, ~200 µg/ml) purified by immunospecific affinity chromatography showed four precipitates when tested against unabsorbed rabbit anti-rFA1 antiserum or rabbit anti-rAF in crossed immunoelectrophoresis. However, when rFA1 was diluted, only one precipitate remained visible. When undiluted rFA1 was tested against anti-rFA1 absorbed with normal rat serum and amniotic fluid depleted of FA1 by immunoadsorption or against monospecific anti-mFA1, only one precipitate was seen. The tandem crossed immunoelectrophoresis analysis of mFA1 and rFA1 showed complete fusion of precipitates without spur formation, demonstrating immunochemical identity between mouse and rat FA1 (Fig. 1).
|
Amino Acid Analysis of rFA1
The sequence of the 28 first N-terminal amino acids of the purified rFA1 showed 100% identity to the corresponding section of pref-1 of the rat (amino acids 24–60) (Fig. 2), 86% identity to mouse dlk, and 82% identity to human dlk. The identity between the first 28 amino acids of mouse and human dlk was 71%. The amino acid composition of the purified rFA1 showed 90–100% identity to the corresponding part of rat pref-1 and rat adrenal zona glomerulosa protein (ZOG) (amino acids 24–304).
|
Levels of rFA1 in Serum of Nonpregnant and Pregnant Adult Female Rats, Neonates, and Fetuses
The concentration of rFA1 in serum of nonpregnant adult rats was approximately 0.01 µg/ml. In pregnant rats, the serum level of rFA1 increased dramatically during the second half of gestation (Fig. 3), the highest levels occurring between Days 15 and 20 (~0.2 µg/ml). In amniotic fluid and fetal serum, the concentration of rFA1 was about two orders of magnitude higher (~10 and 18 µg/ml, respectively). The level of rFA1 peaked 1 day earlier (Day 17) in fetal serum than in amniotic fluid and maternal serum (Table 1).
|
|
The relationship between maternal serum concentration of rFA1 and the number of fetuses (corrected for day of pregnancy) demonstrated that the concentration of FA1 in maternal rat serum was positively correlated with the number of fetuses (n = 41; r = 0.434; P < 0.002) (Fig. 4).
|
Levels of rFA1 in Urine Compared with Corresponding Serum of Pregnant and Nonpregnant Females
Rat FA1 could be quantified in all samples by the ELISA. There was no significant difference between serum and urine levels in nonpregnant and pregnant rats until late pregnancy (Table 2). However, on Day 19 of pregnancy, significantly increased levels were recorded and the concentration of rFA1 was significantly higher in urine than in serum (P < 0.01). The average concentration of rFA1 in urine was approximately twice that in serum, and in individual animals it was always higher in urine than in serum.
|
DISCUSSION
Using monospecific rabbit anti-mFA1 antibodies, FA1 was isolated from rAF. Rat FA1 showed immunochemical identity with mFA1 in crossed tandem immunoelectrophoresis using rabbit anti-mFA1 and rabbit anti-rFA1 antibodies. Purified rFA1 was used to produce anti-rFA1 antisera in rabbits.
Mass spectrometry of the intact rFA1 glycoprotein showed a population of molecules with molecular masses ranging from 28 to 36 kDa, with the bulk of the material (95%) appearing around 33 kDa. The molecular mass calculated from of the amino acid composition analysis was 34 kDa. These results indicate that the number of amino acids of the circulating rFA1 varies slightly and that the protein is glycosylated to some degree. Results from sugar analyses, SDS-PAGE, and gel filtration (data not shown) also support the conclusion that rFA1 is a glycoprotein of slightly variable protein chain length and glycosylation. These findings are in agreement with previous reports on mFA1 and hFA1 [15].
The amino acid composition of rFA1 was found to be identical (100%) to the corresponding part of rat pref-1. The amino acid sequence of the N-terminus of rFA1 (28 amino acids) showed 86% identity (24/28) between rat and mouse FA1, 82% identity (23/28) between rat and human FA1, but only 71% identity (20/28) between mouse and human FA1. With the assumption that the unidentified amino acids were cysteine or tryptophan, it is reasonable to conclude that rFA1 is identical to the circulating gene product of rat pref-1 [8]. An identical genetic code has also been reported for the ZOG protein of rat adrenal tissue [16]. However, because the present analysis was restricted to 28 of a total of approximately 260 amino acids, the results may be different from an analysis comprising the entire molecule.
Using the ELISA technique, FA1 was detected in all rat body fluids tested. In serum from nonpregnant adult rats, the average concentration of FA1 was 0.01 ± 0.003 µg/ml. In pregnant rats, the serum concentration of FA1 increased dramatically during the second half of gestation, with a maximum around Day 18. The concentration of rFA1 in amniotic fluid on Day 18 was approximately 50 times higher (11 ± 4.55 µg/ml) than in the maternal serum (0.19 ± 0.11 µg/ml). On Day 17, the concentration in fetal serum (18 ± 2.93 µg/ml) was approximately twice that in amniotic fluid (9 ± 2.51 µg/ml). The concentration of rFA1 seems to peak 1 day earlier in fetal serum than in amniotic fluid and maternal serum, indicating that rFA1 in amniotic fluid originates from the fetus and not from the mother. The positive correlation between the concentration of rFA1 in maternal serum and in amniotic fluid suggests that the elevated levels of rFA1 in maternal serum are caused by transplacental transfer of the molecule from amniotic fluid to maternal circulation.
In the urine of pregnant rats, the variation of the FA1 level followed the same pattern as that observed in the corresponding serum samples at approximately the same concentrations until Day 15 of pregnancy. The ratio of rFA1 in urine versus blood from Day 15 and onwards was appropximately 2:1, indicating that rFA1 is actively concentrated in the kidneys.
This is the first report on FA1 in the rat. The physicochemical characteristics and the compartmentalization are very similar to what has been reported for human and mouse FA1. The amino acid sequence analyses indicated that rat and human FA1 show a higher degree of homology than do mouse and human FA1. Considering the importance of rats in physiological and metabolic studies, we believe that the rat will be a useful model in studies of the biological role of FA1. Studies on the possible hormonal regulation of FA1 synthesis are in progress.
ACKNOWLEDGMENTS
Our sincere thanks go to Dr. B. Teisner, who was the first to identify FA1 in human amniotic fluid, and his chief technician, Mrs. J. Brandt, for setting up the assays at Uppsala University and generously sharing reagents and expertise. The meticulous technical assistance of Ms. Else-Marie Andersson is gratefully acknowledged, as well as the expert statistical advice of Dr. J. Hagelin.
FOOTNOTES
First decision: 17 September 1999.
1 This work was supported by the Swedish Medical Research Council, projects 12567 and 7475. 
2 Correspondence: Hans-Erik Carlsson, Department of Physiology, Division of Comparative Medicine, Box 572, Uppsala University, SE-751 23 Uppsala, Sweden. FAX: 46 18 501740; hans-erik.carlsson{at}bmc.uu.se
Accepted: February 2, 2000.
Received: July 29, 1999.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
, Individual values; 
, expected trendy values