Biochemical Characterization and Immunolocalization of Prostate-specific Antigen in Human Term Placenta, Ferdinando Mannello, 1* Manuela Malatesta, 1 Eugenio Fusco, 2 Giuseppe Bianchi, 3 Antonella Cardinali, 1 and Giancarlo Gazza- nelli 1 ( 1 Instituto di Istologia ed Analisi di Laboratorio, Fa- colta ` di Scienze Matematiche, Fisiche e Naturali. Universita ` Studi, Via Zeppi, 61029 Urbino (PS), Italy; 2 Divisione di Ostetricia e Ginecologia, AUSL 2, Ospedale Civile, Urbino, Italy; 3 Laboratorio Analisi, AUSL 2, Ospedale Civile, Urbino, Italy; * author for correspondence: fax 39-722- 322370, e-mail mannello@bio.uniurb.it) Prostate-specific antigen (PSA) has been biochemically and molecularly characterized as a 33-kDa androgen- dependent glycoprotein related to the kallikrein family of serine proteases, with chymotrypsin- and kallikrein-like enzymatic activity (1-4). PSA has long been thought to be produced exclusively by the prostate cells and has been used as a tumor marker for diagnosis and monitoring of prostate cancer (5, 6). Recently, it was found in several nonprostatic tissues and body fluids (7), although no physiologic role of PSA is known in these tissues (8). Immunoreactivity and gene expression studies have char- acterized human PSA as a steroid hormone-regulated serine protease (9-12). Considering the studies carried out in amniotic fluids and in healthy endometrium (13-15), we undertook the present study on PSA characterization and immunolocalization in the term placentas collected from five women (ages, 25–37 years) undergoing routine deliv- eries (40  2 weeks). After the membranes were stripped, each placenta was weighed and placed on ice in a sterile solution of 9 g/L NaCl and 5 mmol/L glucose, trans- ported to the laboratory, and processed within 30 min. Samples of entire placentas were minced and homoge- nized as described (16). After sonication, the lysates were centrifuged at 9000g at 4 °C for 30 min, and the superna- tants were stored at -80 °C until analysis. n-Butanol- extracted fractions were prepared from cytosolic extracts, according to a previously described method (17). Blood was also drawn from healthy pregnant women (n = 15; ages, 23–38 years); after the blood clotted, the sample was centrifuged at 500g for 10 min and the serum stored at -30 °C until assay. Free and total PSA concentrations were determined in serum and cytosolic extracts of pla- centas, using an automated enzyme immunoassay with a detection limit of 0.01 g/L (AxSYM ® PSA, Abbott Lab- oratories) (9, 16, 18). Placental extracts were serially di- luted in PSA-negative female serum and reanalyzed to exclude the possibility of matrix artifacts. The analytical recovery of purified PSA added to placental extracts was also tested. Decidual extract components were separated on a 600  9 mm column of Sephacryl S-200 (Pharmacia Biotech) and eluted (16). Reagents and equipment for Western blotting were purchased from Bio-Rad. Our protocols were followed throughout, using anti-human PSA monoclonal and polyclonal mouse antibody (Dako) (16, 18). For thermal inactivation studies, placental extract was incubated in a thermostable water bath for different lengths of time at 55 °C or at different temperatures (45, 55, 65, and 75 °C) for 30 min. Five repeated freeze-thaw cycles were also performed using placental extract ali- quots stored in cryotubes at -30 °C. Moreover, we tested PSA stability at 25 °C both in different buffer systems (phosphate, HEPES, 1,4-piperazinediethanesulfonic acid, Tris-HCl, and borate-citrate-phosphate-HCl) as well as at different pH values. The results obtained from five differ- ent placental specimens (mean  SE) were statistically analyzed using the StatView, Ver.4. package (Abacus Concepts) on a Macintosh Power PC (Apple). For the electron microscopic analysis, term placenta was fixed in a mixture of 40 g/L paraformaldehyde and 5 g/L glutaraldehyde in 0.1 mol/L So ¨ rensen phosphate buffer, pH 7.4, at 4 °C for 2 h. After the placenta was washed in So ¨ rensen phosphate buffer, the free aldehydes were blocked in 0.5 mol/L ammonium chloride in phos- phate-buffered saline at 4 °C for 45 min; the placenta was then dehydrated through graded concentrations of etha- nol and finally embedded in LRWhite resin (MultiLab). Polymerization was carried out under ultraviolet light at room temperature. For immunocytochemical analyses, ultrathin sections were placed on nickel grids, floated for 3 min on normal goat serum, and incubated for 17 h at 4 °C with rabbit anti-human PSA antiserum (Biomeda). After the sections were rinsed, they were reacted for 30 min at room temperature with the secondary gold-conju- gated antibody (Jackson ImmunoRes Laboratories). Fi- nally, sections were rinsed, air-dried, and stained with uranyl acetate. As controls, some grids were treated with the incubation buffer without the primary anti-PSA anti- body. The present work was carried out in accordance with the ethical standards of Helsinki Declaration of 1975, as revised in 1983. The average serum PSA content of the women exam- ined (n = 15) was 0.15  0.03 g/L. The linearity and interference studies revealed a good correlation between PSA concentration and dilution (serum of pregnant women, n = 15, r 2 = 0.98; placental extracts, n = 5, r 2 = 0.99), demonstrating that “placenta matrix” (constituted of lipids, hemoglobin, hormones, and proteins) did not affect the performance of PSA assays. Analytical recovery of purified PSA added to cytosolic placental extract was 97%  4%. Assay reproducibility (CV) was determined by assaying placenta samples in replicates of two or three in at least four independent analyses; the within-run CV was 2.5% and the between-run CV was 4.4%. The mean concentration of PSA in term placenta tissues (n = 5) was 56  8 g/L, with 30% in the free, noncomplexed form (16.8  2.3 g/L). n-Butanol extraction of term placental homogenates revealed that the major proportion (91%) of PSA immu- noreactivity was in the aqueous phase, but the altered recovery of PSA subfractions made it an unsuitable method for isolation of PSA from biological samples. The PSA immunogram of a Sephacryl S-200 column of placental extracts revealed that the major immunoreactiv- ities were in fractions 35– 80, where the antichymotrypsin (ACT)-PSA complex and free PSA were expected (19); in Technical Briefs Clinical Chemistry 44, No. 8, 1998 1735