Functional precision oncology using patient-derived assays: bridging genotype and phenotype.

Genomics-based precision medicine has revolutionized oncology but also has inherent limitations. Functional precision oncology is emerging as a complementary approach that aims to bridge the gap between genotype and phenotype by modelling individual tumours in vitro. These patient-derived ex vivo models largely preserve several tumour characteristics that are not captured by genomics approaches and enable the functional dissection of tumour vulnerabilities in a personalized manner. In this Review, we discuss several examples of personalized functional assays involving tumour organoids, spheroids and explants and their potential to predict treatment responses and drug-induced toxicities in individual patients. These developments have opened exciting new avenues for precision oncology, with the potential for successful clinical applications in contexts in which genomic data alone are not informative. To implement these assays into clinical practice, we outline four key barriers that need to be overcome: assay success rates, turnaround times, the need for standardized conditions and the definition of in vitro responders. Furthermore, we discuss novel technological advances such as microfluidics that might reduce sample requirements, assay times and labour intensity and thereby enable functional precision oncology to be implemented in routine clinical practice.

A cloned cell line mediating natural killer cell function inhibits immunoglobulin secretion.

We previously described a cloned cell line that combines information for a unique display of cell surface antigens and specialized function similar to activated natural killer (NK) cells. In addition to conventional cellular targets such as the YAC-1 and MBL-2 lymphomas, this cloned line also lysed lipopolysaccharide-activated B lymphocytes. To determine whether some NK cells can inhibit B cell function, we tested the ability of NK-like clones to suppress Ig secretion in vitro and in vivo. These cloned cells suppressed Ig secretion when they constituted as few as 0.2% of the total cell population and inhibition did not require identity at the H-2 locus. We suggest that some NK cells might recognize non-major histocompatibility complex gene products on activated B lymphocytes and lyse these cells, and this might represent a fundamental cell-cell interaction that regulates antibody secretion by activated B cells.

Sequence and structure of a human glucose transporter.

The amino acid sequence of the glucose transport protein from human HepG2 hepatoma cells was deduced from analysis of a complementary DNA clone. Structural analysis of the purified human erythrocyte glucose transporter by fast atom bombardment mapping and gas phase Edman degradation confirmed the identity of the clone and demonstrated that the HepG2 and erythrocyte transporters are highly homologous and may be identical. The protein lacks a cleavable amino-terminal signal sequence. Analysis of the primary structure suggests the presence of 12 membrane-spanning domains. Several of these may form amphipathic alpha helices and contain abundant hydroxyl and amide side chains that could participate in glucose binding or line a transmembrane pore through which the sugar moves. The amino terminus, carboxyl terminus, and a highly hydrophilic domain in the center of the protein are all predicted to lie on the cytoplasmic face. Messenger RNA species homologous to HepG2 glucose transporter messenger RNA were detected in K562 leukemic cells, HT29 colon adenocarcinoma cells, and human kidney tissue.

Nonspecific cross-reacting antigen 50/90 is elevated in patients with breast, lung, and colon cancer.

A total of 22 genes have been identified in the carcinoembryonic antigen (CEA) gene family. The protein products of this family are highly homologous and include CEA, biliary glycoprotein, nonspecific cross-reacting antigen 50/90 (NCA 50/90), NCA 95, and pregnancy-specific beta-glycoprotein. We used a monoclonal antibody with high affinity to develop a specific enzyme-linked immunosorbent assay (ELISA) method for NCA 50/90 in serum and plasma. Our calibrators were based on affinity-purified recombinant protein from a baculovirus expression system. No significant reactivity with purified CEA, recombinant NCA 95, or recombinant biliary glycoprotein was found by Western blot analysis or in the ELISA method. Only 1 of 15 sera from pregnant women (chorionic gonadotropin > 1000 ng/ml) was positive in the NCA 50/90 ELISA, suggesting that this method does not detect pregnancy-specific glycoprotein. A cutoff value of 18 ng/ml was established based on the 95% value of serum and plasma from 147 healthy volunteers. Only 3 of 31 serum and plasma samples from patients with clinically inactive breast cancer were elevated above the cutoff value, but 44% of 136 samples from patients with clinically active breast cancer were positive. NCA 50/90 measurements were elevated in 7 of 25 patients with active breast cancer whose CEA and CA 15-3 values were below cutoff, and NCA 50/90 values do not correlate with CEA in breast cancer. In addition, we found sensitivities of 70, 39, and 42% for lung cancer, colon cancer, and leukemia, respectively. The sensitivity for non-small cell lung cancer was 85%, however, compared to 50% for small cell lung cancer. Serum from leukemia patients showed an overall sensitivity of 43%, but 71% (10 of 14) sera from patients with chronic myelogenous leukemia were positive compared to, for example, chronic lymphocytic leukemia where 0 of 7 sera had NCA 50/90 values above the cutoff. These studies suggest that NCA 50/90 may have clinical utility in the management of patients with a variety of cancers.

The glucose transporter in human fibroblasts is phosphorylated in response to phorbol ester but not in response to growth factors.

The possibility that the stimulation of hexose transport in human fibroblasts by phorbol myristate acetate (PMA), insulin, platelet-derived growth factor (PDGF) or epidermal growth factor (EGF) is associated with phosphorylation of the glucose transporter has been investigated. The time and concentration dependencies of the stimulation of transport by these agents under conditions identical to those used for phosphorylation were determined. Each agent, when used at the concentration that resulted in the maximal increase in transport rate, elicited this effect within 30 min of exposure. The extent of stimulation ranged from 15 to 70%. For determination of phosphorylation of the glucose transporter, fibroblasts were incubated for 16 h with [32P]Pi and exposed to the agonist for 30 min; the transporter was then isolated from a detergent lysate of the cells by immunoprecipitation with a monoclonal antibody. Under these conditions, there was no phosphorylation of transporter in basal cells and only PMA caused detectable incorporation of phosphate into the transporter. Thus, it is unlikely that the stimulation of glucose transport by insulin, PDGF and EGF involve transporter phosphorylation.

Antigen binding properties of highly purified bispecific antibodies.

A panel of three bispecific monoclonal antibodies (bsMAbs) binding to follitropin (FSH) and to beta-galactosidase have been prepared by fusion of hybridoma cell lines resistant to oubain and neomycin. One of these bispecific antibodies contains heavy chains of the same IgG subclass, and two are composed of heavy chains of different IgG subclasses. We have investigated methods for the purification of bispecific antibodies from hybrid hybridoma supernatant fluids grown in serum-free medium. Following ammonium sulfate precipitation, bispecific antibodies can be purified in a single step by mixed mode ion-exchange HPLC on Bakerbond Abx columns. In one case, three species were resolved by ion-exchange HPLC and functional analysis showed that two peaks contained parental antibodies, and the third contained the bispecific. Ion-exchange HPLC purification of serum-free preparations from two other hybrid hybridomas resolved seven protein-containing peaks, only one of which was active in a bispecific ELISA. The equilibrium affinity constants for each of the parental antibodies for both FSH and beta-galactosidase were determined and found to be similar to those of the purified bsMAbs. Further, the association of FSH to one binding site on a bispecific antibody was shown to have no effect on the equilibrium binding constant for beta-galactosidase binding to the other site. Our results suggest that bsMAbs can be readily purified from hybrid hybridomas by a simple and rapid method, and the binding of antigen to one binding site on a bsMAb is independent of antigen binding to the second site.

The function but not the expression of rat liver inhibitory guanine nucleotide binding protein is altered in streptozotocin-induced diabetes.

Adenylate cyclase activity was examined as a measure of inhibitory guanine nucleotide binding protein (Gi) function in liver plasma membranes from rats made chemically diabetic by streptozotocin (STZ) treatment. Clonidine activation of the alpha 2 adrenergic receptor, which activates Gi, inhibited forskolin--stimulated adenylate cyclase activity in control membranes. However, there was no effect on adenylate cyclase activity in membranes from STZ diabetic animals. Also, a polyclonal antipeptide antibody was raised to a highly conserved segment of the Gi alpha 2 subunit. This antibody specifically recognizes a 41 kilodalton protein, is blocked by an excess of peptide, does not recognize the alpha-subunit of transducin, and immunoprecipitates a 41 kilodalton protein which was ADP-ribosylated by pertussis toxin. Immunoblots using this antibody detect no difference between normal and STZ diabetic animals in the level of liver plasma membrane Gi expression. Therefore, STZ-induced diabetes altered the function of Gi but had no effect on Gi expression.

Measurement of complexed PSA improves specificity for early detection of prostate cancer.

OBJECTIVES: Prostate-specific antigen (PSA) is the most useful of all tumor markers. Although the sensitivity is impressive, low specificity results in a lack of cancer detection in a significant proportion of patients undergoing prostate biopsy. Several recent studies have addressed the need for improved specificity. Of all these approaches, the free/total PSA ratio appears to be the most promising. Given that most circulating PSA is complexed to alpha1-antichymotrypsin, and that this moiety represents a greater proportion of the total PSA in those men with carcinoma, we set out to determine whether complexed PSA would improve specificity in the detection of men with prostate cancer. METHODS: Archival sera were obtained from 300 men, 75 of whom had biopsy-proved prostate cancer. All sera had been previously stored at -70 degrees C for variable periods. An investigative assay for complexed PSA (Bayer) was used. The Tandem-R free and total PSA assays (Hybritech) were used according to the manufacturer's recommendations. RESULTS: Among all patients, specificities for the total PSA, free/total PSA, and complexed PSA alone were 21.8%, 15.6%, and 26.7%, respectively, at cutoffs yielding 95% sensitivity. Similar equivalence or superior performance, in terms of specificity relative to the free/total PSA ratio, was seen at other sensitivity thresholds and other total PSA ranges. CONCLUSIONS: Complexed PSA alone performs better than total PSA or the free/total PSA ratio and obviates the need for a second analyte determination. We believe this marker may offer significant enhancement in PSA testing with significant economic advantages.

Clinical utility of serum HER-2/neu testing on the Bayer Immuno 1 automated system in breast cancer.

UNLABELLED: The clinical utility of automated serum HER-2/neu measurements in breast cancer run on the Bayer random analyzer Immuno 1 was analyzed in several steps: [a] The reference interval was determined for 242 normal healthy pre- and postmenopausal females. [b] The clinical specificity of serum HER-2/neu to separate healthy controls from 210 patients with non-malignant breast--and non-breast diseases was calculated. [c] The clinical sensitivity of cross-sectional serum HER-2/neu values for 204 patients (pts) with stage I-IV breast cancer was established. [d] Specimens from 103 stage IV breast cancer pts were tested for their parallel between serial serum HER-2/neu results and disease course. RESULTS: [a] The value of 13.03 ng/ml exceeded 95% of the results from the healthy female population. Based on the mean +2 standard deviations value of 14.7 ng/dl, the upper limit of normal was established at 15 ng/ml. [b] The specificity for benign breast diseases and other benign non-breast diseases was 98.0% and 94.6%, respectively. [c] The correlation of increased serum HER-2/neu levels and stage of breast cancer revealed the best sensitivity of 40% for stage IV disease. [4] Thirty-eight (36.9%) of 103 stage IV patients had initial HER-2/neu values > 15 ng/ml, 33 of whom showed longitudinal HER-2/neu concentrations which paralleled the clinical course of the disease giving a sensitivity of 86.8%.

Multicenter evaluation of the performance and clinical utility in longitudinal monitoring of the Bayer Immuno 1 complexed PSA assay.

We conducted a multicenter evaluation of the analytical and clinical performance of the automated Bayer Immuno 1 complexed PSA (cPSA) assay, and compared assay performance to the Bayer Immuno 1 PSA assay. We sought to determine whether measurements of cPSA could be of clinical utility in the management of patients with prostate cancer. Results of the 10-day imprecision across three evaluation sites produced total CV < 2.50% and an analytical sensitivity of 0.02 microgram/L. There was an increased trend in clinical sensitivity for prostate cancer with increasing stage of disease (71-86%). Clinical specificity for patients with benign urogenital disease was 74.8%, and for other nonprostate diseases ranged from 91.1-100%. Retrospective serial monitoring of 155 patients with prostate cancer demonstrated concordance of cPSA measurements to clinical status for 97% of the patients analyzed. Results from the clinical studies using the Bayer Immuno 1 cPSA assay were comparable to results obtained with the Bayer Immuno 1 PSA assay. The Bayer Immuno 1 cPSA assay demonstrates analytical performance and clinical effectiveness in the management of prostate cancer patients during the course of disease and therapy.

Monoclonal antibodies to the glucose transporter from human erythrocytes. Identification of the transporter as a Mr = 55,000 protein.

Using the preparation of purified glucose transporter from human erythrocytes as antigen, we have prepared and characterized six monoclonal antibodies. Three of these antibodies have been shown to be to the glucose transporter by several criteria: they immunoprecipitate the transport activity, the cytochalasin B binding activity, and 75% of the protein from the solubilized purified preparation. The remaining three antibodies were shown to recognize the same polypeptide by a Western blot procedure. All of the antibodies reacted with the deglycosylated transporter and are thus against peptide determinants; most bound to the cytoplasmic domain of the transporter. The antibodies exhibited a range of effects on cytochalasin B binding, from slight enhancement to modest inhibition to strong inhibition; for this reason they must bind to at least three different epitopes. Western blot analysis of erythrocyte membranes prepared in the presence of protease inhibitors showed that all six antibodies bound to a polypeptide of average Mr = 55,000. Moreover, by immunological assay this polypeptide accounted for 5.3% of the membranes protein, a value similar to that given by cytochalasin B binding. Thus, the proposal that the native transporter is a Mr = 100,000 polypeptide is highly unlikely. The antibodies also react with the glucose transporter in other human cell types, but not with that in rodent or avian cells.

Novel immunoassay for the measurement of complexed prostate-specific antigen in serum.

Serum prostate-specific antigen (PSA) is an effective diagnostic tool for detection of prostate cancer (CaP) at an early and potentially curable stage, but specificity is low. Studies have shown that the proportion of serum PSA complexed with alpha-1-antichymotrypsin (ACT) is higher in men with CaP than in men with benign prostate disease. We developed a novel immunoassay for complexed PSA based on the unique binding properties of a monoclonal antibody that fails to bind free PSA in the presence of antibodies specific for free PSA. The assay measured mixtures of free and complexed PSA accurately, and the measured values of free + complexed PSA in artificial mixtures and in patient sera were equivalent to the measured value of total PSA. Both the serum concentration and the proportion of complexed PSA was substantially higher in patients with CaP compared with patients with benign prostate disease. The cPSA assay may have utility in improving specificity in screening for prostate cancer.

The glucose transporter in 3T3-L1 adipocytes is phosphorylated in response to phorbol ester but not in response to insulin.

At maximally active concentrations with 20-min exposure, insulin and phorbol myristate acetate (PMA) stimulated hexose transport in 3T3-L1 adipocytes by 11- and 2-fold, respectively. The potential role of phosphorylation of the glucose transporter (GT) in these stimulations was investigated by the isolation of GT through immunoprecipitation from ortho[32P]phosphate-labeled 3T3-L1 adipocytes. It was found that there was no significant 32P incorporation into GT from basal adipocytes after 2- or 18 h-labeling in the presence of 0.5 mCi of 32Pi/ml. Furthermore, under these labeling conditions, insulin treatment for 1, 4, or 30 min failed to stimulate the phosphorylation of GT. Also, there was no detectable phosphate incorporation into GT upon reversal of insulin-stimulated hexose transport by the removal of insulin (half-time for reversal approximately 8 min). In contrast to these results, exposure of adipocytes to PMA (1 microM) for 20 min elicited a phosphorylation of GT to the extent of about 0.1 phosphate/GT molecule. Exposure of cells to both insulin and PMA resulted in a 3-fold increase in the level of phosphate in GT compared to that seen with PMA alone. Possibly this increase is due to the translocation of GT to the plasma membrane where it is a better substrate for activated protein kinase C. Stimulation of hexose transport was the same with the combined treatment of insulin and PMA compared to that seen with insulin alone. These results indicate that neither a change in the phosphorylation state of the GT nor activation of protein kinase C is involved in the mechanism by which the insulin receptor stimulates glucose transport.

Technicon Immuno 1 PSA assay measures both free and alpha-1-antichymotrypsin-complexed prostate-specific antigen on an equimolar basis.

In this study, we investigated the immunoreactivity of the Technicon Immuno 1 PSA assay, a monoclonal-polyclonal sandwich immunoassay, with free and ACT-complexed PSA. Assay calibrators prepared with free PSA (standard Immuno 1 calibrators) and calibrators consisting of 90% ACT-complexed and 10% free PSA (90:10 calibrators) yielded virtually identical PSA recoveries at all concentrations tested. Concentrations of total PSA at approximately 4 and 10 ng/mL, prepared in varying ratios of free PSA to PSA-ACT complex, recovered from 92-104% at 4 ng/mL and from 98-102% at 10 ng/mL. Additionally, an excellent correlation of serum total PSA values from a panel of 40 prostatic cancer patient samples was obtained using calibration curves generated with the standard Immuno 1 calibrators or the 90:10 calibrators. These results demonstrate that the Technicon Immuno 1 PSA assay measures free and ACT-complexed PSA on an equal molar basis.

Effect of desialylation on binding, affinity, and specificity of 56 monoclonal antibodies against MUC1 mucin.

We evaluated 56 monoclonal antibodies (MAbs), submitted to the ISOBM TD-4 Workshop, for changes in binding following desialylation of the MUC1 molecule and for epitope specificity. Antibody binding of MAbs was assayed by an ELISA method using microtiter plates coated with the MUC1 mucin obtained from supernatants of the ZR75-1 cell line. The MUC1 mucin was desialylated directly on the plate by treatment with neuraminidase. For each MAb, binding to untreated mucin was compared over a range of antibody concentrations. The concentration at which binding was half-maximal (K50) was determined for all antibodies whose binding reached saturation in the assay. Results showed that K50 values for MAb binding to untreated MUC1 mucin varied from 10(-10) to 10(-6) M. These data suggest that MAbs to MUC1 mucin bind with a broad range of intrinsic affinities. Desialylation was found to have variable effects on antibody binding, in that binding was either increased, decreased, or unchanged. No relationship was found between the apparent affinities for untreated mucin and changes in binding following desialylation. Among the 56 Workshop MAbs, 33 were found reactive with synthetic peptides which mimic the MUC1 tandem repeat. We determined the epitope specificity of the 33 MAbs by competitive binding using 10 amino acid peptides corresponding to various regions of the 20-amino acid tandem repeat domain of MUC1. All antibodies which recognized epitopes in the 1-10 amino acid region of the tandem repeat showed increased binding to desialylated mucin. Antibodies to other peptide epitopes showed no consistent pattern of change in binding following desialylation. Our results suggest that sialic acid residues on the MUC1 mucin may contribute either positively or negatively to antibody binding. In addition, our results suggest that improved antibody selection methods could provide MAbs with improved selectivity for cancer-derived mucin compared with mucin from normal tissues. This could form the basis of improved biomarker assays for breast cancer.

Alloreactive T-cell clones. Ly phenotypes predict both function and specificity for major histocompatibility complex products.

We have studied the association of Ly phenotype with function and specificity for major histocompatibility complex (MHC) products by examining the properties of 21 T-cell clones derived from B10 anti-B10.D2 and B10.A anti-B10.D2 mixed lymphocyte cultures (MLC). T cells were selected after MLC solely on the basis of Ly phenotype, cloned by limiting dilution, and tested for stability of Ly phenotype, function and specificity for class I or class II MHC products. Sixteen Ly-1+2- and five Ly-1-2+ T-cell clones were tested. The clones selected for the Ly-1+2- phenotype maintained this phenotype, expressed helper but not lytic function, and recognized class II MHC products (I-Ad or I-Ed). All Ly-1-2+ clones maintained this phenotype, possessed cytolytic but not helper activity, and recognized class I MHC products (Dd and Ld). Our data therefore confirm at the clonal level the original observations of a remarkably consistent correlation between Ly markers, MHC specificity, and function. They suggest that the expression of Ly antigens on T-cell clones forms part of a genetic program for each of these specialized cells that also determines their function and MHC specificity.

Purification of ras GTPase activating protein from bovine brain.

In cytosolic extracts of bovine brain, we detected ras GTPase activating protein (GAP) activity that stimulated the GTP hydrolytic activity of normal c-Ha-ras p21 but not that of the oncogenic [Val12]p21 variant. GAP was purified 19,500-fold by a five-column procedure involving DEAE-Sephacel, Sepharose 6B, orange dye and green dye matrices, and Mono Q resins. A single major protein band of 125 kDa was observed on NaDodSO4/polyacrylamide gels that correlated with the elution of GAP activity on Mono Q. Purified GAP was devoid of inherent GTP hydrolytic activity, suggesting that it was a regulator of ras intrinsic GTPase activity. Under submaximal velocity conditions, the second-order rate constant of GTP hydrolysis at 24 degrees C for p21-GTP + GAP (4.5 X 10(6) M-1.sec-1) was at least 1000-fold greater than that for [Val12]p21-GTP + GAP (less than 3 X 10(3) M-1.sec-1).

Can complexed prostate specific antigen and prostatic volume enhance prostate cancer detection in men with total prostate specific antigen between 2.5 and 4.0 ng./ml.

PURPOSE: We assessed whether complexed prostate specific antigen (PSA) and complexed PSA referenced variables would enhance prostate cancer detection in men with serum total PSA between 2.5 and 4.0 ng./ml. MATERIALS AND METHODS: Transition zone and total prostate gland volumes were determined in 151 men who underwent prostate biopsy using an 11 core biopsy strategy. In addition to measuring the Bayer section sign complexed PSA assay, we also calculated 2 computed complexed PSA values (Hybritech parallel total PSA--Hybritech free PSA and Bayer total PSA--Hybritech free PSA). We calculated 8 volume referenced variables using total and complexed PSA, and 2 computed complexed PSA values by dividing each value by the total prostate and transition zone volumes. RESULTS: Of the 151 patients 37 (24.5%) had cancer. In 10 of the 37 men with cancer (27%) a positive core was present in only 1 or more of the 5 alternate regions not sampled by conventional sextant biopsies. At 92% sensitivity a cutoff value of 2.3 ng./ml. for complexed and 31% for free-to-total PSA provided 42% and 11% specificity, respectively (p <0.001). In the 116 men with a total prostate volume of 30 cc or greater at 92% sensitivity the specificity of complexed PSA density (55%) and complexed PSA adjusted for transition zone volume (52%) were better than that of complexed (40%) and free-to-total (11%) PSA. In the 35 men with a total prostate volume of less than 30 cc at 92% sensitivity the specificity of complexed PSA (50%), complexed PSA density (55%) and complexed PSA adjusted for transition zone volume (55%) were significantly better than that of free-to-total PSA (8%, p <0.001). The area under the curve of complexed PSA was almost identical to that of the 2 computed complexed PSA calculations. CONCLUSIONS: A substantial proportion of men with total PSA values between 2.5 and 4.0 ng./ml. had prostate cancer. Complexed and computed complexed PSA were more specific than the free-to-total PSA ratio when total PSA was between 2.5 and 4.0 ng./ml. A 2.3 ng./ml. threshold for complexed and computed complexed PSA appears to stratify prostate biopsy results in men with total PSA between 2.5 and 4.0 ng./ml. The computed complexed PSA calculation appears to be equivalent to the complexed PSA serum assay for detecting cancer. Volume referenced complexed PSA performed better than complexed PSA in men with a total prostate volume of 30 cc or greater compared to men with a total prostate volume of less than 30 cc.

Measurement of four tumor marker antigens in the sera of pregnant women.

We sought to determine the maternal serum levels of four tumor-associated antigens during the three trimesters of pregnancy in healthy women. CEA, CA 228, CA 15-3, and Her2/neu oncogene product p105 assay values were determined for 90 healthy pregnant women during the three trimesters of pregnancy at five participating evaluation sites. Results were compared to means and cut-off values determined for healthy nonpregnant women. Differences in assay values in the 1st and 3rd trimester were analyzed for statistical significance (Student's t-test). CEA, CA 228 and CA 15-3 assay values in general were found to be within the normal range. CA 15-3 and Her2/neu p105 serum assay values were above the cut-off (3.3% and 8.2%, respectively) and were significantly elevated in the 3rd trimester as compared to the 1st trimester of pregnancy (P < 0.05 and P < 0.001, respectively). CEA and CA 228 may be of potential value in monitoring pregnant women with malignant disease. Normal elevations in 3rd trimester serum Her2/neu p105 and CA 15-3 assay values should be considered when monitoring a pregnant patient with malignant disease.

Analysis of monoclonal antibodies to prostate-specific antigen: reactivity with native and recombinant prostate-specific antigen.

Epitope mapping analysis was performed on 53 antibodies submitted to the ISOBM TD-3 PSA Workshop. Western blotting and N-terminal amino acid sequencing, using both native and recombinant human prostate-specific antigen (rPSA), identified four different epitope groups for native PSA. Under reducing conditions native PSA was not recognized by 18/53 antibodies suggesting they reacted with conformation-dependent epitopes. Nine other antibodies reacted with a rPSA polypeptide doublet of 34-35 kD corresponding to different rPSA glycoforms. From sequence mapping studies 22/53 antibodies bound epitopes within amino acid residues 25-85, 3/53 antibodies bound to epitopes within residues 86-220, while 10/53 antibodies bound epitopes within residues 221-261. These results indicate that there are multiple immunogenic epitopes localized on the PSA molecule.