| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Hybritech Incorporated, a subsidiary of Beckman Coulter, Inc., San Diego, CA 92121.
2 Department of Urology/Surgery, UCLA School of Medicine, Los Angeles, CA 90095.
aAddress correspondence to this author at: Beckman Coulter, Inc, 7330 Carroll Rd., San Diego, CA 92121-2302. Fax 858-621-4610; e-mail sdmikolajczyk{at}beckman.com.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Methods: The BPSA antigen was purified by HPLC, and murine monoclonal antibodies were prepared by standard methods. A fluorogenic ELISA was developed with high specificity for BPSA and no cross-reactivity with other forms of PSA.
Results: The BPSA immunoassay had a lower limit of detection of 6 ng/L and a cross-reactivity of <1% with all other clipped and nonclipped forms of PSA. The BPSA antibody was specific for the internal Lys182 cleavage site that characterizes BPSA. Biopsy-negative men with a median total PSA of 4.8 µg/L had a median of 0.22 µg/L BPSA, representing 25% of the free PSA in serum. BPSA ranged from 0% to 60% of the free PSA in serum. BPSA in a cohort of cancer serum also comprised 25% of the free PSA. Control serum from women or men without increased PSA had nondetectable BPSA.
Conclusions: BPSA is a significant percentage of the free PSA in BPH serum but not in control serum. The presence of prostate cancer does not alter the relative proportions of BPSA in sera with <10 µg/L PSA. BPSA has a wide distribution of concentrations in the serum and may provide clinical information for the study of men with BPH.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1-antichymotrypsin, which is commonly referred to as the "complexed" form of PSA, and noncomplexed PSA, called "free" PSA. Subsequent studies have shown that the relative concentrations of free PSA are higher in benign disease (6)(7). The biological mechanism for the increased concentrations of free PSA in benign disease is not fully understood, but it is generally accepted that free PSA is composed of enzymatically inactive PSA. This hypothesis is supported by studies on BPH tissue nodules that showed that nodular PSA was more enzymatically inactive than seminal plasma PSA (8). PSA has been most widely characterized in seminal plasma, for which several internally clipped and inactive forms have been described (9). Although each of these inactive PSA forms may also be present in the serum at some concentration, the major percentage of free (noncomplexed) PSA in serum is now known to comprise at least three distinct forms of inactive PSA, which have been recently reviewed (10). This includes one form of PSA that has been identified as the proenzyme, or precursor form of PSA (pPSA), which is associated with cancer (11)(12)(13)(14). A second PSA form appears to be composed largely of intact PSA that is similar to native, active PSA except that it is enzymatically inactive (15)(16)(17). A third form of free PSA, called "benign" PSA (BPSA), is an internally cleaved or degraded form of PSA that is more highly associated with BPH tissues (18). BPSA is simultaneously clipped at Lys145 and Lys182. Additional studies have shown that BPSA is also present in seminal plasma (19). The BPSA in these studies was purified and analyzed by analytical HPLC methods because PSA is present at gram per liter concentrations in tissues and seminal plasma. The critical questions remained: was BPSA also found in the serum of men with BPH and, if so, at what concentrations? The measurement of BPSA in serum is far more difficult than in tissue or seminal plasma because PSA is present at only microgram per liter concentrations. This report describes the development and characterization of a research immunoassay to measure BPSA in serum and the first quantification of BPSA concentrations in men with BPH and cancer.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The BPSA assay protocol is as follows. Biotinylated capture anti-PSA mAb PSM773 (100 µL; 0.25 mg/L), diluted in 10 mmol/L Tris–0.15 mol/L NaCl, pH 8.5, containing 10 g/L bovine serum albumin (mAb buffer), was incubated for 1 h with shaking in a streptavidin-coated microtiter plate. The plate was washed with 10 mmol/L Tris–0.15 mol/L NaCl, pH 8.5, containing 0.5 mL/L Tween 20 and 50 µL of blocking buffer (10 mmol/L Tris–0.68 mol/L NaCl, pH 8.5, containing 10 mL/L Tween 20, 10 mL/L Triton X-100, 200 mL/L horse serum, and 47 g/L bovine serum), followed by 50 µL of either the calibrator in mAb buffer or sample and allowed to incubate with shaking for 2 h. The plate was washed, and the alkaline phosphatase-labeled anti-BPSA mAb, PS2E290, was added (100 µL; 2 mg/L) and allowed to incubate with shaking for 1 h. The plate was washed, and 100 µL of 4-methylumbelliferyl phosphate was added and allowed to incubate with shaking for 5 min. The plates were then read at 5, 60, and 120 min. Relative fluorescence was measured in a Victor 1420 multilabel counter (Wallac, EG&G).
The serum cohorts tested included the following: BPH [biopsy-negative men with increased PSA (PSA, 1.5–17 µg/L)]; cancer [biopsy-positive (PSA, 2.6–9.7 µg/L)]; urologic outpatients [men who received outpatient examination but who were not suspected of having cancer or benign disease (PSA, 0.17–2.3 µg/L)]; young men [men <30 years of age in apparent good health (PSA, 0.21–0.9 µg/L)]; females [young women <30 years of age (PSA nondetectable)]; and prostatectomy [post-radical prostatectomy serum with minimal PSA (PSA, 0–0.34 µg/L)]. PSA and free PSA were measured with Hybritech brand Tandem-MP PSA and FPSA assays (Beckman Coulter, Inc.). The analytical minimum detectable concentration (MDC) of these assays is 0.05 µg/L. All patients providing serum signed an informed consent approved in advance by Western Institutional Review Board (Seattle, WA).
The biological MDC for BPSA was calculated as the mean concentration of BPSA in 20 female sera plus 2 SD of the mean. The cross-reactivity of the BPSA assay was calculated as the slope of the BPSA calibration curve divided into the slope of the calibration curve for PSA that did not contain the internal clip at Lys182 [PSA182(-)]. The concentrations of all free PSA forms were determined with the Hybritech brand Tandem FPSA (Beckman Coulter).
Transurethral resection of the prostate (TURP) tissue was frozen in liquid nitrogen and homogenized in phosphate-buffered saline (PBS) as described previously (18). PSA was purified from the filtered supernatant solution by passage over an immunoaffinity column containing the bound anti-PSA mAb, PSM773 (Beckman Coulter), at 5 g/L of resin. PSM773 has been shown previously to be a PSA-specific mAb and to have specificity for mature, cleaved, and precursor forms of PSA (21)(22)(23). The column was washed with 40 volumes of PBS containing 1 mL/L reduced Triton X-100 (Sigma), and the PSA was eluted with 100 mmol/L glycine, pH 2.5, containing 200 mmol/L NaCl. The eluate was immediately neutralized with 1 mol/L Tris, pH 8.0 (100 mL/L). For the purification of PSA from seminal plasma, seminal plasma was diluted 1:20 in PBS (1 mL of seminal plasma plus 20 mL of PBS), and the PSA was purified as described above.
High-performance hydrophobic interaction chromatography (HIC-HPLC) was performed with a polypropylaspartamide column [PolyLC; 250 x 4.6 mm (i.d.); pore size, 1000 Å; distributed by Western Analytical]. Samples were applied in 1.5 mol/L ammonium sulfate and eluted with a gradient. Buffers were as follows: buffer A, 1.2 mol/L sodium sulfate–25 mmol/L sodium phosphate, pH 6.3; buffer B, 50 mmol/L sodium phosphate–50 mL/L 2-propanol. The gradient was 0–30% B over 1 min and 30–60% B over 10 min, followed by equilibration in buffer A. High-sensitivity peak detection was obtained with a Varian Model 9070 (Varian Medical Systems) scanning fluorescence detector with excitation at 232 nm and emission at 334 nm to detect the tryptophan residues in protein.
N-Terminal sequence analysis of the samples was performed on an Applied Biosystems Model 492 protein sequencer. Purified PSA and peaks collected by HIC-HPLC were applied directly to Prosorb cartridges (Applied Biosystems), which were then washed with three 0.1-mL volumes of 0.1 mL/L trifluoroacetic acid, and the membrane was applied to the sequencer.
PSA samples were incubated with 1 g/L lysyl endopeptidase C (LysC; Wako Chemicals USA) in PBS, pH 7, for 30 min to obtain comparable proportions of BPSA and PSA entirely cleaved at Lys182 only [PSA182(+)]. The LysC reaction was quenched by the addition of a 10-fold molar excess of aprotinin before analysis by HIC-HPLC. The internal cleavage sites of the different HIC-HPLC peaks were determined by N-terminal sequencing.
Sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) was performed using 4–20% gradient minigels (Invitrogen) under reducing or nonreducing conditions, as indicated. Samples for sequencing were resolved by SDS-PAGE and electroblotted onto polyvinyl difluoride membranes (Applied Biosystems). The bands were visualized with Coomassie Brilliant Blue G250, excised, and applied directly to the Applied Biosystems Model 492 protein sequencer.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, peak A], as has been described previously (19). The PSA in peak A contained 
35% PSA cleaved at Lys145 and 4% cleaved at Lys182, which are typical values for seminal plasma PSABPSA(-). Peak A contained all forms of seminal plasma PSA except BPSA. After digestion with LysC, the PSA of peak A was converted into two new peaks (Fig. 1
, peaks B and C). N-Terminal sequencing confirmed that peak B represented PSA entirely cleaved at Lys182 only [PSA182(+)] and that peak C represented PSA cleaved at both Lys145 and Lys182 (the BPSA form of PSA).
|
SDS-PAGE of the peaks in Fig. 1
was performed under standard reducing conditions with 2-mercaptoethanol, and the identities of the bands were further confirmed by acid sequencing (Fig. 2
). Lane 1 in Fig. 2
shows the purified PSA starting material, of which the majority is intact PSA but which also contains internal peptide bond cleavages. The bands at 22 and 10 kDa are the fragments containing residues 1–145 and 146–237, respectively, resulting from cleavage at Lys145. Lane 2 in Fig. 2
shows that peak B contains no intact PSA and is almost entirely cleaved at Lys182 [PSA182(+)]. The fragments at 29 and 6 kDa represent fragments containing residues 1–182 and 183–237, respectively. Lane 3 in Fig. 2
shows the fragment pattern of peak C (BPSA). The main bands are found at 22, 6, and 4 kDa and represent fragments comprising residues 1–145, 146–182, and 183–237, respectively. These results demonstrate that very pure and homogeneous BPSA and PSA182(+) can be isolated after treatment of PSA with LysC.
|
In addition to the natural BPSA present in seminal plasma, PSA also naturally contains minor amounts of PSA182(+) as demonstrated by the SDS-PAGE blot and N-terminal sequencing and Western blot analysis of peak A in Fig. 1
. BPSA contains clips at Lys145 and Lys182, whereas PSA182(+) contains only a single clip at Lys182. Seminal plasma PSA typically contains 7–10% BPSA compared with 3–4% PSA182(+), or approximately two to three times more BPSA than PSA182(+). Prostate transition zone tissues obtained from TURP samples were analyzed to determine whether this reflected similar proportions of BPSA and PSA182(+) in BPH tissue.
The HIC-HPLC profile of the PSA forms purified from TURP tissue is shown in Fig. 3A
. The clearly resolved PSA and BPSA are equivalent to peaks A and C, respectively, in Fig. 1
. To determine the concentration of any unresolved PSA182(+) that might be present in the PSA eluting at 10 min in Fig. 3A
, an immunoaffinity column was constructed to purify all PSA forms containing the clip at Lys182. The mAb PS2E290 was used for the immunoaffinity column because this mAb was demonstrated to have high selectivity for PSA containing the Lys182 clip (Fig. 4
). This includes PSA with only the Lys182 clip and PSA containing clips at Lys182 and Lys145.
|
|
The PSA from Fig. 3A
was passed over the PS2E290 immunoaffinity column, and the elution profile of the PSA that bound to the column is shown in Fig. 3B
. Because only PSA containing the clip at Lys182 is retained on the PS2E290 immunoaffinity column, Fig. 3B
shows the relative proportions of BPSA and PSA182(+), both of which were confirmed by N-terminal sequencing. In the sample shown in Fig. 3
, the PSA182(+) represented 21% of the eluted Lys182-clipped PSA forms, whereas BPSA represented the remaining 79%. The same analysis of six TURP tissues from different patients is shown in Table 1
, which lists the relative percentages of BPSA and PSA182(+) in each of the individual TURP tissues. Despite the wide range of naturally occurring BPSA present in these tissues (3–24% of the total PSA), the PSA182(+) remained a relatively constant minority percentage of the BPSA, suggesting that both resulted from the same mechanism. Therefore, the PSA182(+) represented a mean (SE) of 31% ± 7% of the Lys182-clipped PSA in TURP tissue, whereas BPSA represented 69% ± 7%.
|
Mice were immunized with various PSA forms, including native BPSA purified from seminal plasma, in vitro prepared BPSA purified after treatment with LysC or trypsin, and PSA182(+) obtained after LysC treatment. The pure and precisely characterized forms of BPSA and PSA182(+) shown in Figs. 1
and 2
were used to evaluate the binding and specificities of the resulting mAbs. The negative-control PSA was PSABPSA(-) (Fig. 1
, peak A).
The antibodies with the highest specificity for BPSA and low cross-reactivity with PSABPSA(-) were determined to have specificity for the presence of the Lys182 clip in PSA. Some mAbs had a preference for BPSA compared with PSA182(+) or vice versa, but none was sufficiently specific between these two forms to allow for the development of an immunoassay specific for the individual BPSA or PSA182(+) forms. The mAb PS2E290 was selected as suitable for assay development because it is strongly and equally reactive to both BPSA and PSA182(+) but has negligible reactivity to PSA that does not contain the clip at Lys182 (Fig. 4
). In Fig. 4
, the assay shows a minor reactivity to the PSABPSA(-) (peak A in Fig. 1
) because it contains a minor amount of PSA182(+). When the PSABPSA(-) was subsequently passed over the PS2E290 immunoaffinity column to remove the minor PSA182(+) contamination, the assay reactivity dropped to nearly zero (Fig. 4
). The reactivity of this PSA containing no clips at Lys182, termed PSA182(-), is the true negative PSA control for the BPSA assay. PSA182(-) is devoid of the 10% of PSA containing the clip at Lys182 but otherwise contains all of the PSA forms typically found in purified seminal plasma PSA.
The BPSA assay is therefore defined as specific for PSA that contains the clip at Lys182, which includes both BPSA and PSA182(+). The lower limit of detection of the assay in buffer was 6 ng/L. The biological MDC of the assay as determined with female serum was 0.023 µg/L. Serum BPSA was linear on dilution (slope, 1.02; r = 0.998). BPSA was found to be relatively stable in serum. The analysis of BPSA assay control calibrators added to female serum and stored at -70 °C showed no significant variation from the mean after more than 6 months, with an intraassay CV of 5.1% and an interassay CV of 8.7% (n = 159). The BPSA concentrations in freshly thawed serum specimens showed no differences after 48 h of storage at 4 °C (data not shown).
The assay of BPSA forms in the serum of men with BPH is shown in Fig. 5
. The term BPH in this case indicates men with increased PSA and negative biopsy but does not necessarily imply men with symptomatic BPH. These men had a mean (SE) International Prostate symptom score of 12.0 ± 8.5 and a Qmax (maximum urine flow) of 11.1 ± 4.9 mL/s. The median percentage of free PSA in these men was 19% compared with 16% in the cancer cohort (Wilcoxon P = 0.007; Table 2
). The BPSA ranged from 0% to 60% of the free PSA in both cohorts, and both had median BPSA/free PSA values of 25% (Wilcoxon P = 0.59). There was, however, no significant correlation between the percentage of free PSA and the percentage of the free PSA present as BPSA in either the benign (r = -0.14) or cancer (r = -0.20) groups. BPSA represented 4.9% of the total PSA in the benign group and 4.0% in the cancer group. The median BPSA/total PSA values were significantly different between the cancer and benign cohorts (Wilcoxon P = 0.014). The most representative control set for interfering substances for the cancer and benign cohorts was the urologic outpatient controls because these men comprised a random patient profile of men without significantly increased PSA or other evidence of prostate disease. Measurable BPSA was found only in those outpatients with mildly increased PSA, but not in men without measurable PSA. The other control groups of young men, women, and post-radical prostatectomy patients served to demonstrate that the assay did not recognize significant artifacts in divergent sera. None of the points in these latter control groups were above the MDC. Table 2
provides the general values for these serum cohorts.
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In this study we have shown that BPSA formation in the tissues is closely paralleled by the cleavage at Lys182 only, which suggests a common proteolytic pathway. As seen in Table 1
, the proportion of PSA182(+) correlates with BPSA. For example, TURP samples containing either 3% or 24% of their PSA as BPSA both contained the same relative proportion (21%) of PSA182(+), suggesting that PSA182(+) formation is not independent of BPSA. It is possible that PSA182(+) is initially formed in BPH tissues as a precursor to BPSA and that the subsequent clip at Lys145 occurs by the same proteolytic mechanism that usually produces the large percentage of PSA clipped only at Lys145 in the seminal plasma.
The understanding of these tissue forms of PSA is important because any PSA in the blood is thought to derive from the tissues. It is interesting that Western blot studies of free PSA forms in BPH serum suggest that BPH serum can contain significant quantities of a 29-kDa form of PSA consistent with PSA182(+) (24). The BPSA assay described in the present study measures both BPSA and PSA182(+) equally. It is possible that the development of individual assays that measure each form individually could offer some additional insight, but currently, such mAbs have not been identified. It is not anticipated that such mAbs will be pursued because the prevailing evidence suggests that that the clip at Lys182 is the important feature of these BPH-associated forms of PSA and that both are formed in parallel. Thus, the BPSA assay measures the BPH-associated forms of PSA, which is functionally defined as PSA containing the clip at Lys182.
There are some common features in the current work with a reported assay that measures PSA-I, defined as the free PSA that is not clipped at Lys145 (17), but there are also some important differences. The PSA-I assay is reported to measure inactive forms of free PSA not clipped at Lys145, which would most likely include most proPSA forms. With the PSA-I assay, one can calculate, by difference, the remainder of the free PSA not detected by the PSA-I assay, i.e., free PSA that is clipped at Lys145. It is assumed that this latter calculation would substantially include BPSA but could also include some percentage of PSA clipped only at Lys145. However, from the present study it can be seen that approximately one-third of the PSA that is measured in the BPSA assay may be PSA182(+), which does not contain a clip at Lys145. The PSA-I assay for cancer detection may be confounded to some extent by the presumed measurement of the BPH-associated PSA182(+). Thus, although there may be substantial overlap in the forms of free PSA measured in the BPSA assay compared with (the inverse of) the PSA-I assay, these assays clearly measure different populations of free PSA and these differences may be important. In the future, careful analysis of well-defined patient populations will be necessary to determine which forms of free PSA offer the greatest diagnostic value for different disease states.
BPSA represents a subpopulation of free PSA that is present in the serum of men with increased PSA but with no diagnosis of prostate cancer (i.e., biopsy-negative). BPSA is increased in the prostate tissues of men with enlarged prostates and pathologic BPH, as indicated by the presence of BPH nodules (18). However, because men with prostate cancer may also have enlarged prostates and coexisting BPH, this marker alone would not be expected to significantly distinguish prostate cancer from BPH. Stamey et al. (25) have proposed that most of the increased serum PSA in men with <9 µg/L PSA is attributable to BPH. This limited comparison of biopsy-positive and -negative sera indicates that BPSA is not increased because of cancer and is, in fact, significantly higher in benign serum as a percentage of total PSA (Table 2
). However, these limited cohorts are not sufficiently powered to determine a clear role for BPSA discrimination of cancer or a potentially more specialized role in other aspects of cancer detection, such as men with very high or low free PSA concentrations. For cancer detection, we have previously shown that proPSA forms, and particularly the truncated proPSA forms of free PSA, are highly associated with prostate cancer (26).
The application of BPSA toward the study of BPH is complex, and many questions remain. It is not clear why some men have almost no BPSA whereas in others the BPSA comprises more than one-half of the free PSA in serum. Because BPSA is associated with pathologic BPH at the tissue level, it is intriguing to speculate about a serum correlation with prostate volume, nodular development, symptomatic BPH properties such as bladder obstruction, prostate hormonal balance, or any number of other potential indicators or consequences of BPH. Another important area for investigation is the serum BPSA response to BPH drug treatment.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1-antichymotrypsin. Clin Chem 1991;37:1618-1625.1-antichymotrypsin is the major form of prostate-specific antigen in serum of patients with prostatic cancer: assay of the complex improves clinical sensitivity for cancer. Cancer Res 1991;51:222-226.The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  C. STEPHAN, H. RITTENHOUSE, H. CAMMANN, M. LEIN, M. SCHRADER, S. DEGER, K. MILLER, and K. JUNG New Markers and Multivariate Models for Prostate Cancer Detection Anticancer Res, July 1, 2009; 29(7): 2589 - 2600. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Saerens, J. Kinne, E. Bosmans, U. Wernery, S. Muyldermans, and K. Conrath Single Domain Antibodies Derived from Dromedary Lymph Node and Peripheral Blood Lymphocytes Sensing Conformational Variants of Prostate-specific Antigen J. Biol. Chem., December 10, 2004; 279(50): 51965 - 51972. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Jung, J. Reiche, A. Boehme, C. Stephan, S. A. Loening, D. Schnorr, W. Hoesel, and P. Sinha Analysis of Subforms of Free Prostate-Specific Antigen in Serum by Two-Dimensional Gel Electrophoresis: Potential to Improve Diagnosis of Prostate Cancer Clin. Chem., December 1, 2004; 50(12): 2292 - 2301. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Wu, L. Zhu, U.-H. Stenman, and J. Leinonen Immunopeptidometric Assay for Enzymatically Active Prostate-Specific Antigen Clin. Chem., January 1, 2004; 50(1): 125 - 129. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
