| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Cancer Diagnostics |
1 University of Turku, Department of Biotechnology, Turku, Finland.
2 Memorial Sloan-Kettering Cancer Center, New York, NY.
3 Lund University, Department of Laboratory Medicine, Division of Clinical Chemistry, University Hospital UMAS, Malmö, Sweden.
4 University Hospital of Turku, Department of Surgery, Turku, Finland.
aAddress correspondence to this author at: University of Turku, Department of Biotechnology, Tykistökatu 6 A 6th Floor, 20520 Turku, Finland. Fax 358-2-333-8050; e-mail ville.vaisanen{at}pp.nic.fi.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Methods: PSA- and hK2-specific monoclonal antibodies were used in solid-phase, two-site immunofluorometric assays to detect free and total hK2. The total hK2 assay used PSA-specific antibodies to block nonspecific signal. The capture antibody of the free hK2 assay did not cross-react with PSA. To determine the hK2 concentrations in the male bloodstream, total hK2 was measured in a control group consisting of 426 noncharacterized serum samples. Free and total hK2 were measured in plasma from 103 patients with confirmed prostate cancer.
Results: All 426 males in the control group had a total hK2 concentration above the detection limit of 0.0008 µg/L. The median total hK2 concentration was 0.022 µg/L (range, 0.0015–0.37 µg/L). hK2 concentrations were 0.1–58% of total PSA (median, 3.6%). hK2 concentrations were similar in men 41–50 and 51–60 years of age. The ratio of hK2 to PSA steadily decreased from 5–30% at PSA <1 µg/L to 1–2% at higher PSA concentrations. In 103 patients with prostate cancer, the median hK2 concentration in plasma was 0.079 µg/L (range, 0.0015–16.2 µg/L). The median free hK2 concentration was 0.070 (range, 0.005–12.2) µg/L. The proportion of free to total hK2 varied from 17% to 131% (mean, 85%).
Conclusions: The wide variation in the free-to-total hK2 ratio suggests that hK2 in blood plasma is not consistently in the free, noncomplexed form in patients with prostate cancer. The new assay is sufficiently sensitive to be used to study the diagnostic accuracies of free and total hK2 for prostate cancer.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1-antichymotrypsin (ACT). Determining the relative amounts of these forms of PSA can substantially improve the discrimination between prostate cancer (CaP) and benign prostatic hyperplasia (BPH)(5)(6)(7). Human glandular kallikrein 2 (hK2) is a serine protease very similar to PSA. Both genes, KLK2 and KLK3, belong to the human tissue kallikrein gene family, which was recently found to consist of at least 15 members(8). PSA and hK2 share 80% sequence homology(9) as well as the property of being produced mainly in the prostate under hormonal regulation(1)(3)(10)(11)(12)(13). hK2 has also been found in other tissues, but in considerably lower concentrations(14)(15)(16)(17)(18)(19).
Several hK2 immunoassays have been developed in the past few years(20)(21)(22)(23)(24)(25)(26)(27). In the bloodstream, hK2 seems to be present in concentrations of 
1–2% compared with PSA(20)(21)(22)(23). However, the covariance of hK2 and PSA concentrations is generally <60%, suggesting that hK2 might have a function as an independent tumor marker. In recent clinical studies, hK2 has shown promising results for improving the specificity in the early detection of CaP(23)(28)(29). hK2 might also serve to improve prediction of organ-confined cancer(30)(31)(32).
Similar to PSA, hK2 can also be found in different molecular forms. It has been shown in vitro to form a complex with several protease inhibitors, including 2-antiplasmin, ACT, antithrombin III, 2-macroglobulin, C1-inactivator, and plasminogen activator inhibitor-1 (PAI-1)(33)(34)(35). Both free and complexed forms of hK2 have also been found in biological fluids. In seminal fluid, the protein is mainly complexed to protein C inhibitor (PCI)(36). In the bloodstream, gel-filtration studies have suggested that 80–95% of hK2 is in the uncomplexed, free form(20)(23)(24), and up to 20% can be complexed with ACT in sera from patients with CaP that contain high concentrations of hK2(37). Additionally, an inactive zymogen form of hK2 (pro-hK2) has also been found in blood(25)(37).
Because circulating hK2 is present in very low concentrations, the immunoassays used to measure hK2 must have low detection limits and excellent reproducibility. For the PSA "gray zone" of 2–10 µg/L, respective hK2 values will be 0.02–0.5 µg/L. In testing with the first published hK2 assay, up to 57% of samples had hK2 concentrations below the detection limit(20).
In this report we present the production of two novel monoclonal antibodies (MAbs) specific for hK2 and the development of ultrasensitive immunoassays for the measurement of free and total hK2 in human serum. The diagnostic accuracy of the immunoassays will be addressed in a later study.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
production of MABS 11B6 and 7D7
The recombinant hK2 (rhK2) forms fXahK2 and ekhK2 used for immunization have been described elsewhere(39). BALB/c mice were immunized by intraperitoneal injections with 30–40 µg of rhK2. fXahK2 was emulsified in Freund’s complete adjuvant, whereas ekhK2 was mixed with TiterMax Gold (CytRx Corporation). Two to four booster doses (30–50 µg of rhK2) were given at 3- to 4-week intervals. A final booster dose was given to the mice (30–70 µg of rhK2) 3 days before fusion. Approximately 2 months after immunization, blood from each mouse was tested for reactivity against hK2. The mice were killed, and spleen cells were homogenized, fused with SP 2/0 myeloma cells at a 1:1 ratio(44), and applied to microtiter plates.
Serum or cell culture supernatant diluted in DELFIA Assay Buffer (100 µL volume) was added to microtiter plates coated with rabbit anti-mouse IgG antibodies; the plates were subsequently incubated overnight and then washed. Recombinant wild-type PSA (100 µL; 1 mg/L in DELFIA Assay Buffer) was added to one plate, whereas only buffer was added to another. After a 2-h incubation at room temperature, the plates were washed, and europium-labeled hK2 (50 µL; 0.4 mg/L in DELFIA Assay Buffer) was added. After a 60-min incubation, the plates were again washed. DELFIA Enhancement Solution (200 µL) was added, and 60 min later, the fluorescence was measured with a DELFIA 1234 Plate fluorometer.
Supernatants from each cell population were tested for the best signal to hK2 and no cross-reactivity to PSA as described above, and some were selected and cloned by limiting dilution(45). We found that wells 11B6 and 7D7 produced an antibody that reacted well with Eu3+-labeled hK2 and that binding was not disturbed by the presence of PSA. We further subcloned MAbs 11B6 and 7D7 and produced high quantities of the antibodies in a bioreactor [MiniPerm (Heraeus Instruments) and CELLine (Integra Biosciences)]. The MAb was subtyped with a MAb Mouse Isotyping Kit (Gibco), and the immunoglobulin concentration was measured with Eu3+-labeled anti-mouse antibodies. Finally, immunoglobulins were purified on a HiTrap Protein G gel column (Amersham Pharmacia Biotech) and eluted with 0.1 mol/L glycine-HCl (pH 2.5).
antibody specificity to HK2
Three MAbs, 6H10, 7D7, and 11B6, raised against hK2 were tested for their cross-reactivity with hK2 complexes formed in vitro with ACT, PCI, PAI-1, and antitrypsin (AT). The hK2-serpin complexes were formed by incubation of recombinant wild-type hK2 with the protease inhibitor in buffer solution [Tris-HCl (pH 8), 0.1 mol/L NaCl, 0.5 mL/L Tween 20]. A 100-fold excess of inhibitor and overnight incubation at 37 °C were used for ACT and AT. A 5-fold excess of PCI and a 10-fold excess of PAI-1 were used for 3 h at 37 °C. The formed complex was separated from noncomplexed hK2 and inhibitor by gel filtration on a Superdex 75 3.2/30 column (Amersham Pharmacia Biotech). The amount of complexed hK2 was determined with an immunoassay that detects both total hK2 and PSA (Prostatus®; Perkin-Elmer Life Sciences, Wallac Oy). To determine the ability of the different hK2 MAbs to detect complexed hK2, the MAbs were used in four two-site immunoassays as either capture (6H10) or tracer (7D7 and 11B6) in conjunction with MAbs against PSA/hK2 (H117 and H50). The tested combinations were as follows: H117/H50, H117/11B6, H50/7D7, and 6H10/H50. MAbs H117 and H50 recognize both PSA and hK2 with equal affinity(42).
Cross-reactivity to PSA was determined by measuring a sample containing 10 mg/L recombinant PSA but no hK2.
epitope mapping
We determined the locations of the new epitopes in relation to previously determined MAb epitopes(43)(46)(47) by attempting to form two-site (sandwich) immunoassays with antibodies with known epitopes on hK2 and PSA. The hK2-specific antibodies 7D7, 11B6, and 6H10 were used as both capture and tracer in conjunction with the PSA/hK2 MAbs. Affinity constants for the antibodies were determined by the Scatchard method(48).
calibrators
The hK2 calibrator solutions were prepared in Tris-HCl buffer, pH 7.75 (50 mmol/L Tris, 9 g/L NaCl, 0.5 g/L NaN3, 10 g/L bovine serum albumin). The protein concentration of the purified recombinant ekhK2 was determined by absorbance at 280 nm and total protein analysis (Bio-Rad Protein Assay; Bio-Rad Laboratories). The hK2 was diluted to 0.003, 0.006, 0.03, 0.06, 0.3, and 3.0 µg/L, and the final concentrations were confirmed by a PSA immunoassay (PSA F/T Prostatus; Perkin-Elmer Life Sciences, Wallac Oy). The immunoassay uses MAb H117 as capture and MAb H50 as the tracer antibody(49)(50). Both of these antibodies recognize both PSA and hK2 with equal affinity(42).
assay protocols
Total PSA was measured with an in-house assay that uses H117 as capture and H50 as the tracer antibody(49)(50). Both of these antibodies recognize both PSA and hK2 with equal affinity(42). The total PSA result is a combination of PSA and hK2. Because PSA is found at much higher concentrations in biological fluids, this has little effect on the overall result. Biotinylated H117 (100 µL; 3 mg/L) was attached to a streptavidin-coated microtiter plate for 1 h with shaking. The plate was washed, and the calibrator or sample (25 µL) was added together with 100 µL of DELFIA Assay Buffer. After a 1-h incubation with shaking, the plate was washed and the Eu3+-labeled tracer antibody H50 was added (200 µL; 0.5 mg/L). The plate was incubated for 2 h with shaking and then washed. DELFIA Enhancement Solution (200 µL) was added, and 5 min later the time-resolved fluorescence was measured. The reported lower limit of detection for the assay was 0.01 µg/L, and the limit of quantification was <0.1 µg/L.
The free PSA assay was identical to the total PSA measurement except that a free-PSA-specific MAb (5A10) was used as the tracer(49)(50).
The total hK2 assay was a modification of the original assay described by Becker et al.(23). The modifications consisted of preincubation of the capture antibody on streptavidin plates as well as replacing of the blocking and tracer antibodies with MAbs that had been developed at our laboratory and were therefore more readily available. The capture antibody against hK2 had a 5% cross-reactivity to PSA. An excess of anti-PSA MAbs was used to block binding of PSA. Biotinylated 6H10 (100 µL; 3 mg/L) was attached to a streptavidin-coated microtiter plate for 1 h with shaking. The plate was washed, and the blocking antibodies 5H6, 5F7, and 2E9 (previously 2H11, 10, 36, and 2E9) in DELFIA Assay Buffer were added (100 µL; 10 mg/L 5F7 and 2E9; 5 mg/L 5H6). The calibrator or sample (100 µL) was added immediately after the blocking MAbs. After a 2-h incubation with shaking, the plate was washed and the Eu3+-labeled tracer antibody 7G1 (previously H50) was added (200 µL; 0.5 mg/L). After another 1-h incubation with shaking, the plate was washed. DELFIA Enhancement Solution (200µL) was added, and 5 min later, the time-resolved fluorescence was measured.
The free hK2 assay was developed with 11B6 as the capture antibody and 6H10 as tracer. The capture antibody was highly specific for hK2; therefore, the assay required no additional blocking of PSA. The concentration of denatured mouse IgG in DELFIA Assay Buffer was increased to 25 mg/L to assure that there was no interference from nonspecific binding. Biotinylated 11B6 (100 µL; 3 mg/L) was attached to a streptavidin-coated microtiter plates for 1 h with shaking. The plate was washed, and the calibrator or sample (100 µL) was added together with 100 µL of DELFIA Assay Buffer. After a 2-h incubation with shaking, the plate was washed, and the Eu3+-labeled tracer antibody 6H10 was added (200 µL; 0.5 mg/L). The plate was incubated for 1 h with shaking and then washed. DELFIA Enhancement Solution (200 µL) was added, and 5 min later, the time-resolved fluorescence was measured.
determination of detection limits
The minimum detectable concentrations for the assays were determined by nine separate runs with three replicates of the zero diluent and each calibrator. The minimum detectable concentration was calculated by taking 2 SD of the zero calibrator fluorescence counts and comparing the result with the counts obtained from each calibrator. Calibrator concentrations were used to determine the minimum detectable concentration in µg/L.
The limit of quantification in serum matrix was determined by serially diluting male serum with a high hK2 concentration (total hK2, 4.0 µg/L; free hK2, 3.7 µg/L) with a pool of hK2-negative female serum. The dilutions were measured with both hK2 assays in 10 replicates, and the last hK2 concentration that had a CV% <15% between replicates was considered the limit of quantification.
assay precision and linearity on dilution
The within-run and total imprecision of the hK2 assays as well as linearity on dilution were determined with a serum sample containing high hK2 (16.45 µg/L total hK2; 12.82 µg/L free hK2).
For precision studies, the high-hK2 sample was diluted in a pool of hK2-negative female serum to low, normal, and high hK2 concentrations (total hK2, 0.01, 0.09, and 4.28 µg/L, respectively; free hK2, 0.01, 0.06, and 3.34 µg/L, respectively). The free and total hK2 concentrations in these samples were measured in triplicate once or twice a day. Nine separate assays were run within a time period of 6 days.
The within-run imprecision was calculated as the median of the CVs obtained from the three replicas in each assay run. Total imprecision of the hK2 assays was defined as the CV of all measured concentrations in all assay runs.
To determine the linearity on dilution, we diluted the high-hK2 sample 20-, 30-, 100-, and 200-fold in zero diluent and in a pool of hK2-negative female serum. The dilutions were measured in three replicates with the developed hK2 assays. Obtained concentrations were compared with expected values.
serum and heparin-plasma samples
The total hK2 assay was used to estimate the hK2 concentrations in a healthy population. For this purpose we used sera from clinically nondefined men with no indication of prostatic disorders. These consecutive serum samples from male volunteers (n = 426) were collected for other purposes. The men were 42–83 years of age (mean age, 54 years). The samples were stored for 2 years at –20 °C before hK2 measurements.
Free and total hK2 were measured in consecutive heparin-plasma samples (n = 103) that were collected from men who were treated for CaP at the Department of Urology, Turku University Hospital. Thirty patients had organ-confined stage T1/2 cancer; the remaining patients had more advanced cancer or staging data were not available. All samples were taken before treatment was initiated, and patients who had received any kind of hormone therapy for BPH were excluded. The age of the patients varied from 41 to 89 years (median age, 65 years). These samples were stored for less than 1 year at –20 °C before hK2 measurements.
Both serum and heparin-plasma samples were collected from a subset of the CaP patients (n = 6). These were used to confirm that measured hK2 concentrations were comparable in both matrices.
Free and total PSA values were also measured in both population groups. All immunoassay measurements were performed at the Department of Biotechnology, University of Turku, Turku, Finland. The study was conducted between April 2001 and September 2003.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The two novel hK2-specific MAbs, 11B6 and 7D7, recognized complexed hK2 very poorly compared with free hK2. The ratios of hK2-ACT, hK2-PCI, hK2-PAI-1, and hK2-AT to free hK2 were 0.02, 0.31, 0.26, and 0.35, respectively, for 11B6 and 0.07, 0.04, 0.05, and 0.24, respectively, for 7D7.
Although 7D7 was more specific for free hK2, MAb 11B6 was selected as the capture antibody in the free hK2 assay because of its technical qualities. Mainly, it maintains immunoreactivity when biotinylated and thus produces a better signal and provides superior reproducibility. In addition, it was desirable for the two hK2 assays to be as similar as possible, and 7D7 could not bind to hK2 in combination with 6H10 (see below).
When the antibodies were combined in a two-site sandwich assay, the total hK2 assay detected 94% hK2-ACT, whereas the free hK2 assay had a cross-reactivity of <2% (results not shown). The total hK2 assay had a cross-reactivity of <0.001% with PSA. The free hK2 assay did not give a detectable signal when 10 mg/L PSA was measured.
epitope mapping
On the basis of our two-site assays and previously published data, a two-dimensional map of the new epitopes was constructed. The schematic locations of the new epitopes in relation to previously determined MAb epitopes are presented in Fig. 1
. The hK2-specific antibody 7D7 epitope overlaps with 6H10, and these could not be used together in a sandwich assay. Similarly, 11B6 overlapped with 7G1.
|
The affinity constants (L/mol) determined by the Scatchard method were 7.9 x 108 for 11B6 and 7.4 x 108 for 7D7.
analytical characteristics
The detection limits were 0.0008 µg/L for the total hK2 assay and 0.001 µg/L for the free hK2 assay. The limit of quantification (intraassay CV <15%) in serum matrix for the total hK2 assay was 0.003 µg/L, and for the free hK2 assay it was 0.010 µg/L.
The total hK2 assay had a within-assay imprecision (CV) of 2.8%, 1.1%, and 1.6% for the low, normal, and high hK2 controls at concentrations of 0.01, 0.09, and 4.28 µg/L. The corresponding between-assay variations were 11%, 5.8%, and 5.7%. The free hK2 assay had a within-assay imprecision of 3.7%, 1.8%, and 2.0% and between-assay imprecision of 15%, 9%, and 9% for the low, normal, and high hK2 concentrations (0.01, 0.06, and 3.34 µg/L).
linearity
The dose-response of the assay was linear in the calibration range 0.003–3.0 µg/L (Fig. 2
). Linearity was also maintained at higher hK2 concentrations, but because of the low concentration of hK2 in the bloodstream, values >3 µg/L are rare.
|
linearity on dilution and recovery
A high-hK2 male sample containing 16.45 µg/L total hK2 and 12.82 µg/L free hK2 was diluted in hK2-negative female serum or standard diluent. The dilutions were measured with both hK2 assays, and the results were compared with the amount added. The measured total hK2 concentrations were 85–96% (mean, 92%) of expected in serum and 85–95% (mean, 90%) in standard diluent. The measured free hK2 concentrations were 91–100% (mean, 99%) of expected in serum and 92–111% (mean, 100%) in the diluent.
samples from control group
Free and total PSA and total hK2 were measured in 426 serum samples obtained from clinically noncharacterized male volunteers with no reported clinical symptoms of prostatic disorders. These men were between 42 and 83 years of age (mean age, 54 years). All samples had a total hK2 concentration above the detection limit of the immunoassay (0.0008 µg/L). Only 2% (9 of 426) of samples had a total hK2 concentration below the limit of quantification (0.003 µg/L). The total PSA concentrations in these samples ranged from <0.09 to 52.2 µg/L (median, 0.60 µg/L). The median total hK2 concentration was 0.022 µg/L (range, 0.0015–0.37 µg/L).
The hK2 and PSA results were divided into five age groups. Although there was a statistically significant increase in hK2 concentration between different age groups (Kruskal–Wallis, P = 0.003), hK2 remained fairly uniform in the age interval 40–60 years (Fig. 3
). PSA clearly increased in a more age-dependent manner (P <0.001).
|
Because of the imprecision at concentrations below the calibration range, all results could not be used for calculating the ratio of hK2 to PSA. Therefore, samples with a PSA or total hK2 value below the lower limit of the calibration range (0.09 or 0.003 µg/L, respectively) were not included in the hK2/PSA results. Overall, 26 of 426 samples were eliminated. In the remaining 400 samples, hK2 concentrations were 1–5% of total PSA (range, 0.1–58%; median, 3.6%). The hK2/PSA ratio did not vary significantly among the five age groups (P = 0.164; results not shown). Total hK2 increased with increasing total PSA (Fig. 4A
), but the correlation between the protein concentrations was low (r = 0.379; P <0.001). The correlation in the PSA gray area (2–10 µg/L) was even lower, and only barely reached statistical significance (r = 0.325; P = 0.044). At low PSA concentrations, the relative amount of hK2 seemed to be markedly higher than at higher PSA concentrations. The amount of hK2 compared with PSA steadily decreased from 5–30% at PSA concentrations <1 µg/L to 1–2% at higher PSA concentrations (Fig. 4B
).
|
The statistical data for all of the performed assays are shown in Fig. 5
and Table 1
.
|
|
samples from patients with CAP
Total PSA and free and total hK2 were measured in 103 heparin-plasma samples that had been collected from patients with CaP. Samples with PSA or hK2 values above the highest points in the calibration curves (50 and 3 µg/L, respectively) were diluted 1:100 in assay buffer, and the measurement was repeated. The PSA concentrations were between 0.9 and 358 µg/L (median, 10.1 µg/L). Total hK2 ranged from 0.0015 to 16.2 µg/L (median, 0.079 µg/L). Free hK2 concentrations were 0.0048–12.22 (median, 0.070 µg/L).
All samples with a free or total hK2 value below the limits of quantification (0.01 and 0.003 µg/L, respectively) were not included when protein proportions or ratios were calculated. This eliminated 4 samples, and the following analyses were made with a total of 99 samples.
The proportion of free to total hK2 varied from 17% to 131% (mean, 85%). The mean amount of hK2 compared with PSA was 1% (range, 0.1–19%; median, 0.9%).
Total hK2 in relation to PSA behaved similarly in this set of patients and in the control group. Both hK2 and PSA concentrations were higher, but the correlation coefficient was similar (r = 0.29). The hK2/PSA ratio was lower in CaP patients than in control males, but when plotted together the results were in agreement, despite some samples with unusually high hK2 concentrations (>10 µg/L; Fig. 4
).
hK2 concentrations were identical when measured in serum and heparin plasma obtained from the same patient. Measured heparin plasma hK2 concentrations were, on average, 102% of serum hK2 concentrations (range, 97–108%).
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
hK2 is a serine protease that shares extensive homology with hK3 (PSA). Both proteins are androgen-regulated and are produced primarily in the prostate gland(12). Because PSA and hK2 share 80% identity at the amino acid level and have the same molecular mass (30 kDa), early PSA preparations were very likely to also contain hK2(54). Many MAbs raised against PSA also detect hK2, whereas some are specific for PSA and have no cross-reactivity to hK2(55).
By blocking the binding of PSA with an excess of PSA-specific antibodies, we have been able to measure hK2 with use of a capture antibody that has slight (5%) cross-reactivity to PSA(20)(23). The production of rhK2 has also made it possible to raise hK2-specific MAbs that have no cross-reactivity to PSA. Several immunoassays for hK2 have thus been reported in the past few years(21)(22)(23)(24)(25)(26)(27). Most of the developed assays are reported to be assays for total hK2, although they do not necessarily recognize free hK2 and complexed hK2 equally(22) or, alternatively, the detection of free and complexed hK2 was not studied(24). However, Klee et al.(26) have reported an assay specific for the free form of hK2 with <3% cross-reactivity to hK2-ACT.
It might be argued that careful characterization of reaction specificity is not necessary because, based on previous gel-filtration studies, 80–95% of detectable hK2 in serum occurs as the free, noncomplexed form(20)(23)(24). However, these studies were made with only a few samples, which contained unusually high concentrations of PSA and hK2.
To our knowledge, no published study to date has attempted to use immunoassays to measure free and total hK2 from the same group of patients.
Here we report the development of two ultrasensitive immunoassays that detect hK2. Both are sensitive enough to measure hK2 concentrations in sera from a group of men 40 years of age and older. One assay is an improved modification of the original assay described by Becker et al.(23) and detects both free and complexed hK2. The other is a novel assay for the specific measurement of free, uncomplexed hK2.
We first attempted to determine the necessary sensitivity needed for an immunoassay to be able to measure the low concentrations of hK2 in the bloodstream. For this purpose, total hK2 concentrations were measured in samples from 426 male volunteers. We observed a median hK2 concentration of 0.022 µg/L, which is consistent with the value (0.026 µg/L) obtained by Klee et al.(26) but higher than the 0.016 µg/L reported by Finlay et al.(22). The control group in the present study consisted of men with no clinical symptoms of prostatic disorders. However, the presence of prostatic disorders was not ruled out, and in a group of men of this age (mean age, 54 years), it is likely that cancer and hyperplasia are present. This will lead to a slightly increased median hK2. The healthy males in the study of Finlay et al.(22) had a mean age of 35 years.
All samples in the control group had a hK2 concentration above the detection limit of the total hK2 assay. Only 2% of all samples had a hK2 concentration below the limit of quantification. We estimate that 90% of samples would also have a free hK2 value >0.01 µg/L, which is the limit of quantification of the free hK2 assay. Unfortunately, because of a shortage of samples we were not able to measure free hK2 in the same cohort. We conclude that both assays are sensitive enough to measure hK2 concentrations in serum when the purpose is to follow the increase of hK2 in prostatic disorders.
hK2 shows an age dependency that is less evident than that of PSA. Men in their fifties seem to have hK2 concentrations similar to those in men in their forties. Although there is an increase in men above age 60, the increase in PSA values can be more easily appreciated.
In most previously published studies, the amount of hK2 in relation to PSA has been 1–4%(20)(21)(22)(23)(26). For our set of control samples, we observed a median ratio of 3.6%. There was a positive correlation between the concentrations of the two proteins, but the correlation coefficient was low, particularly in the diagnostic gray area of PSA. This suggests that these two proteins follow different secretion and elimination patterns and that hK2 may therefore provide independent clinical information compared with PSA.
Although nearly 80% of all samples had a hK2/PSA ratio <7%, some higher ratios were measured as well. It is noteworthy, however, that any sample with a hK2/PSA ratio >6% in the control cohort of men had a total PSA concentration <1.2 µg/L. Overall, the hK2/PSA ratio was higher with low PSA concentrations (Fig. 4B
). A similar trend can be seen in some previously reported studies(23), in which samples with equally low PSA concentrations were measured. This can be explained to some extent by the fact that PSA exists mainly complexed to ACT, whereas the majority of hK2 seems to be in the free form. The half-life of free PSA, which is eliminated by renal clearance, is considerably shorter than the half-life of complexed PSA(56). Thus, an increase in PSA would lead to formation of a relatively stable PSA-ACT complex, whereas an increase in hK2 concentration would be counteracted by rapid clearance by glomerular filtration. Therefore, although both PSA and hK2 increase in prostatic disorders and a correlation exists between the concentrations, it seems that the increase in hK2 is not as rapid and may well follow a distinctively different course.
The second group of samples was obtained from patients with CaP. The samples were heparin plasma, as opposed to serum in the control group. However, several steps in assay development were repeated with both serum and heparin plasma, and no differences caused by matrix had been observed. Additionally, both serum and heparin plasma were collected from six patients with CaP, and the hK2 concentrations measured were identical.
The PSA and hK2 values in the group with CaP were clearly higher than in the control group. The amount of hK2 compared with PSA was clearly lower in the control group than in the group with CaP. In the vast majority the hK2 concentration was <1% of the total PSA concentration. Although this fact seems very interesting for the purpose of separating patients with CaP from those without, it must be taken into account that the two groups in this study had distinctively different blood concentrations of PSA and hK2. The two groups could be almost entirely separated based on total PSA alone. The difference in hK2/PSA ratio could relate to PSA concentrations or there might be a relationship with prostatic disease. It will be interesting and important to further study this ratio with a more well-defined set of patients with similar PSA values.
The total hK2 values were similar to those reported previously in the literature. However, there was a wide range in the ratio of free to total hK2. Although the mean and median ratios were close to 0.9, the amount of free hK2 varied considerably among individual samples. Previously published data based on gel-filtration studies(20)(23)(24) have indicated that the majority of hK2 is in the free form. This has often been the basis for claims that the equal recognition of free and complexed hK2 is not of critical importance for total hK2 assays. Taking into account that the mean free/total hK2 ratio is close to 90%, it is likely that a serum pool consisting of several blood samples would give a very high free/total ratio. However, based on the presented data, we believe that the amounts of free and complexed hK2 in the bloodstream can vary significantly and that differences in the recognition of these hK2 forms can lead to falsely high or low results.
In some samples, the free hK2 result was higher than the total hK2. This is most likely caused by nonspecific interactions with serum components. We noticed that the free hK2 assay benefits from an increase in the amount of mouse IgG. Despite the optimization of assay components and blocking of nonspecific binding, the concentrations of measured hK2 are so low that even otherwise insignificant reactions can lead to a slight increase in signal. It is unlikely that all of these interferences could be removed; therefore, results at the low end of the measurement range should be interpreted with caution.
It cannot be ruled out that the results could be caused by differences in assay format. Although both assays have a common hK2-binding antibody, the total hK2 assay requires blocking of PSA. This is accomplished with a large excess of PSA-blocking antibodies. The reaction specificities of these antibodies may have a slight effect on the result. The blocking MAbs, although having a very low cross-reactivity to hK2, could conceivably decrease the hK2 signal. Additionally, despite their given names, the total hK2 assay does not detect equally all complexed forms of hK2, whereas the free hK2 assay has substantial cross-reactivity with some of the in vitro-formed hK2 complexes. The highest free/total ratios are in the low hK2 concentration range (<0.05 µg/L), where differences such as the ones mentioned above would have the largest impact.
Despite these cautions our results show that hK2 in the bloodstream is not always primarily present in free form. The relative amounts of free and complexed PSA significantly improve the ability to distinguish between BPH and cancer patients(5)(6)(7). Relative amounts of free and total hK2 might also provide valuable information about patients with increased total PSA. Both free as well as total hK2 might have value in management of CaP. In the future, the free and total hK2 concentrations should be measured in patients with CaP and BPH to determine whether the free/total hK2 ratio has clinical value in separating these two groups.
Should hK2 have a role in CaP diagnostics or prognostics, it will become of great importance to understand exactly what form of the protein is being measured. Measurement of specific molecular forms of PSA has led to improved separation of patients with BPH and CaP(43), and different molecular forms may well have different clinical uses. There seems to be considerable variation in the way antibodies raised against hK2 react with their target molecule. MAb 11B6 binds only minimally to hK2-ACT, whereas 6H10 binds it as strongly as does free hK2. However, both detect the hK2-AT complex similarly. Multicenter studies have shown that different total hK2 assays used to measure the exact same patient sample give significantly different results, leading to low correlation coefficients(57)(58). Even small differences in the molecular structure (nicking, prosequences, or changes in amino acid composition) can affect the way hK2 is recognized by the capture or tracer antibodies(58). Therefore, the reaction patterns of different hK2 antibodies should be well characterized and care should be taken when comparing results obtained with different hK2 assays. The results obtained with one assay format may not reproduce with an assay using different antibodies or a different principle.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
1-antichymotrypsin; CaP, prostate cancer; BPH, benign prostatic hyperplasia; hK2, human glandular kallikrein; PAI-1, plasminogen activator inhibitor-1; PCI, protein C inhibitor; MAb, monoclonal antibody; rhK2, recombinant human kallikrein 2; and AT, antitrypsin. 
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
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;51:222-226.1-antichymotrypsin. Clin Chem 1991;37:1618-1625.1-antichymotrypsin as an indicator of prostate cancer. J Urol 1993;150:100-105.[Web of Science][Medline]
[Order article via Infotrieve]
2-Macroglobulin and C1-inactivator are plasma inhibitors of human glandular kallikrein. Blood Cells Mol Dis 1998;24:412-419.[CrossRef][Medline]
[Order article via Infotrieve]
1-antichymotrypsin complex. Clin Chem 1995;41:1480-1488.1-antichymotrypsin from serum. Urology 1998;51:57-62.[CrossRef][Web of Science][Medline]
[Order article via Infotrieve]
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  T. Steuber, A. J. Vickers, A. M. Serio, V. Vaisanen, A. Haese, K. Pettersson, J. A. Eastham, P. T. Scardino, H. Huland, and H. Lilja Comparison of Free and Total Forms of Serum Human Kallikrein 2 and Prostate-Specific Antigen for Prediction of Locally Advanced and Recurrent Prostate Cancer Clin. Chem., February 1, 2007; 53(2): 233 - 240. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
, mean; error bars, 5th and 95th percentiles; horizontal bars, minimum and maximum; x, 1st and 99th percentiles.
