Wallenberglaboratory, 4th floor, University Hospital MAS, SE-205 02 Malmö, Sweden (e-mail: email@example.com).
ABSTRACT: Beta-microseminoprotein (MSP) is one of the most abundant proteins secreted by the prostate gland. Because MSP is also synthesized in nonreproductive organs, the establishment of a solid relationship between the levels of MSP in serum and semen is crucial for future studies connecting MSP with aging or diseases of the prostate gland. We developed a specific, competitive, europium-based immunoassay to measure MSP in serum and seminal plasma. We also produced recombinant MSP in insect cells using baculo virus and purified it to homogeneity by a novel approach with ethanol extraction and gel filtration. The median values of MSP in 205 young men were 12 μg/L (2.5–97.5 percentile, 4.9–26 μg/L) in serum and 0.53 g/L (2.5–97.5 percentile, 0.13–2.0 g/L) or 1.8 mg (2.5–97.5 percentile, 0.32–6.6 mg) in seminal plasma. MSP in serum showed significant correlation to MSP in seminal plasma (r = .50, P < .001). Significant correlations were also found in seminal plasma between MSP and prostate-specific antigen (PSA) (r = .65, P < .001) and between MSP and Zn2+ (r = .54, P < .001). The yield of recombinant MSP in culture medium was 35 mg/L or higher, and recovery following ethanol extraction was 80%–90%. MSP in serum reflects the prostate secretion of MSP, and correlations were also found in seminal plasma between MSP and PSA and Zn2+. This suggests that MSP in serum can be used as a marker of prostate secretion, despite the contribution from extra prostatic tissues.
Biomarkers for the detection and monitoring of prostate cancer have been the topic of research for many years. Prostate-specific antigen (PSA), a secretion product from the prostate that also passes into the blood, has been the marker that has attracted the most interest, and today, the measurement of PSA in serum is a well-established test for detection of prostate cancer. However, the efficiency of PSA is far from ideal, and the search for additional markers is pursued widely, mainly to improve the specificity for early detection and monitoring of the disease. About 10%–40% of PSA in blood occurs in free, uncomplexed forms (fPSA), whereas the main part is bound to α1-antichymotrypsin (Lilja et al, 1991; Stenman et al, 1991). There is evidence to suggest that measurement of the proportion of fPSA to total PSA (tPSA) enhances discrimination of men with prostate malignancy from men with no evidence of cancer in the prostate gland (Christensson et al, 1993; Catalona et al, 1998). This enhancement is incomplete, and it is still relevant to state that no serum marker or combination of markers have the ability to both detect and distinguish aggressive and less aggressive forms of prostate cancer with sufficient certainty.
Beta-microseminoprotein (MSP), which like PSA is a predominant protein secreted by the prostate gland, has attracted much less interest as a biomarker for prostatic disease compared with PSA (Dube et al, 1987; Abrahamsson et al, 1988, 1989). MSP is synthesized as a preprotein of 114 amino acid residues, from which a 20-residue signal peptide is cleaved off to form the mature protein; another name of MSP is prostatic secretory protein of 94 amino acids (Mbikay et al, 1987). The molecular mass is 11 kd, and the polypeptide chain is not glycosylated, despite an apparent molecular size of 16 kd on sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). Studies of the three-dimensional structure by nuclear magnetic resonance (NMR) have shown the molecule to consist of 2 distinct domains that form a rather extended structure (Ghasriani et al, 2006).
Early functional studies reported that MSP display inhibin activity and could suppress secretion of follicle-stimulating hormone from the pituitary gland, which resulted in the alternative name beta-inhibin. These claims were later retracted and there has been no consensus regarding the function of MSP ever since. MSP has also been suggested to be one of the immunoglobulin binding factors in seminal plasma (Liang et al, 1991; Kamada et al, 1998). More recently, 2 studies reported evidence to show that MSP forms high-affinity complexes with 2 related Cys-rich proteins: PSP94-binding protein in blood plasma and cysteine-rich secretory protein 3 (CRISP-3) in semen (Reeves et al, 2005; Udby et al, 2005). Possibly, these observations will give a clue to the function of MSP.
MSP is not solely synthesized by the prostate epithelium, in that the protein also can be detected in nonreproductive organs such as in the respiratory and gastrointestinal tracts, where the gastric mucosa especially shows high expression (Ulvsback et al, 1989; Weiber et al, 1990, 1999). Accordingly, MSP can be measured in serum of both men and women, but the levels in serum from women were found to be around two-thirds of those measured in men (Abrahamsson et al, 1989). Elevated serum concentrations of MSP are reported in hyperplasia and neoplasia of the prostate gland, and increased levels have also been seen in gastric carcinoid disease (Dube et al, 1987; Abrahamsson et al, 1989; Weiber et al, 1999). The nonprostate-specific features of MSP give rise to concerns regarding the value of MSP as a tumor marker for prostatic disease. However, of particular interest for the consideration of MSP as a prostate tumor marker are some recent observations: transcriptional down-regulation of MSP in prostate cancer (Liu et al, 1993), an inhibitory effect of MSP on growing prostate cancer cells (Lokeshwar et al, 1993; Shukeir et al, 2003), and the finding that patients with low serum concentrations of MSP have a high probability of having prostate cancer detected at biopsy (Nam et al, 2006).
Because MSP also has extraprostatic origin, the concentrations measured in serum cannot immediately be associated only with the contribution from the prostate gland. To the best of our knowledge, no previous data are reported on whether there is an association between the secretory release of MSP into semen and the levels at which MSP occurs in serum. We have therefore determined the MSP levels in healthy young men and characterized the correlation between MSP in seminal plasma and serum and other prostate components, such as Zn2+ and PSA. To pursue this work, we developed a new MSP immunoassay and worked out procedures for the expression of recombinant MSP and its purification, which we also report in this paper.
Materials and Methods
Written consent was given by the voluntary donors who provided blood and semen samples for this study. The investigation was approved by the Research Ethics Committee at Lund University (application LU 385–99; date of approval September 22, 1999).
Ejaculates and blood samples were collected from 305 young, healthy Swedish military conscripts as previously described (Richthoff et al, 2002; Sävblom et al, 2005). The sampling was not done on the same day as the examination of physical fitness. Semen samples were taken prior to the serum samples. Semen volumes were calculated from the weights of the samples, assuming a density of 1.0 g/mL. The different PSA forms in serum and seminal plasma and the Zn2+ in seminal plasma were determined in earlier studies (Elzanaty et al, 2002; Sävblom et al, 2005). The Zn2+ concentration was determined with a colorimetric method (Makino et al, 1982). A total of 100 subjects were excluded in this study because of lack of semen samples, leaving 205 cases for analysis of MSP. To investigate the risk of selection bias, the study group of 205 subjects and the excluded subjects were compared with respect to prostate parameters by the Mann-Whitney test. No significant differences were found between the 2 subgroups regarding the previously measured amount and concentration of PSA and Zn2+ in semen or the concentrations of fPSA and tPSA in serum.
Protein and Antiserum
MSP was purified from human seminal plasma essentially as previously described (Fernlund et al, 1994). The pure protein was dialyzed against water and lyophilized. The dry protein was weighed with an accuracy of 1% on a Cahn 28 Automatic electro balance scale (AB Ninolab, Upplands Väsby, Sweden) and then dissolved in a buffer containing 20 mM phosphate and 50 mM NaCl, pH 7.0, to a final concentration of 1.00 mg/mL. Aliquots of the solution were stored at −20°C until use. The rabbit anti-human MSP serum has previously been described (Abrahamsson et al, 1989). Biotinylated goat-anti-rabbit antibody (GAR) was purchased from DACO Norden AB (Solna, Sweden).
Labeling with Eu3+
The labeling of MSP with Eu3+ (Eu-MSP) was made with the DELFIA Eu labeling kit 1244–302 (Wallac Oy, Turku, Finland) according to the manufacturer's instructions. Briefly, 0.45 mg of MSP was dissolved in 600 μL of labeling buffer (50 mM NaHCO3 and 0.15 M NaCl, pH 8.5) containing 0.2 mg of labeling reagent and incubated for 15 hours at room temperature. After the separation of the labeled product from excess reagent on a Sepharose 6B column (Amersham Biosciences, Uppsala, Sweden), the yield was estimated by measuring the delayed fluorescence on an Autodelphia 1235 automatic immunoassay system (Wallac Oy) after the addition of enhancement solution. The MSP concentration was determined by absorbance measurements at 280 nm with an absorbance coefficient of 1.47 for an MSP concentration of 1.00 mg/mL. The mean recovery of Eu-MSP from the Sepharose 6B column was 78%.
All dilutions were made with a buffer containing 50 mM Tris-HCl, pH 7.75, 0.9% NaCl, 0.05% NaN3, 0.01% Tween 20, 0.05% bovine-γ-globulin, 20 μM diethylaminepentaacetic acid (DPTA), 0.5% bovine serum albumin, and 20 μg/mL amaranth (assay buffer). Washings were made with a buffer containing 5 mM Tris-HCl, pH 7.75, 0.9% NaCl, 0.1% germal II, and 0.005% Tween 20. The incubations were made by shaking at room temperature.
The amount of GAR was titrated by incubation of Streptavidin Microtitration strips (Innotrac Diagnostics Oy, Turku, Finland) with different concentrations of biotinylated GAR in 200 μL for 1 hour. After washing, 200 μL of a solution containing anti-MSP diluted 1:4000 and 50 ng of the tracer were added. After additional washes, delayed fluorescence was measured. The anti-MSP antiserum was titrated by incubating 1 ng of tracer in 200 μL of serially diluted antiserum for 4 hours. Different incubation times were then evaluated, showing the most reproducible results after 4 hours of incubation, which was then used for the assay.
The final assay was done with an Autodelphia 1235 automatic immunoassay system (Wallac). As above, all incubations were made at room temperature by shaking. Each well of the streptavidin plates was coated with 300 ng of the biotinylated GAR diluted in 200 μL of assay buffer for 1 hour. After 2 wash steps, 50 μL of sample, 1 ng of Eu-labeled MSP in 50 μL assay buffer, and 100 μL of antiserum diluted 1:16 000, were added to the wells, and the plates were incubated for another 4 hours. After 4 wash steps, 200 μL of enhancement solution was added, and the plates were incubated for 5 minutes prior to measurement of the delayed fluorescence. The standard curve comprised 7 points, with concentrations from 0 to 130 μg/L assay buffer, and calculations were made by the Multicalc program (Wallac). Serum samples were analyzed without prior dilution, but the seminal plasma samples were routinely diluted 1:50 000, in some cases 1:10 000, in assay buffer. The following control samples were included in each assay run: 3 levels of MSP in assay buffer, 2 levels of MSP added to serum, and pooled seminal plasma from 7 donors. All study and control samples were run in duplicate. The standards and controls were prepared from stock solutions, diluted in assay buffer, and frozen in aliquots that were thawed just before use and then handled in the same way as the serum and seminal plasma samples.
The sensitivity, or lower limit of detection of the assay, was defined as the concentration obtained by the mean count minus 2 standard deviations of 32 zero-standard readings in the same assay. The intra-assay coefficient of variation (CV) was determined at 3 different levels. Each sample was run 20 times in 1 assay, and the mean value with CV was calculated. The intra-assay CV was also estimated for every sample because of duplicates. The interassay variation was estimated by analysis of variance from the values for the duplicates in 10 consecutive assays.
The linearity and recovery were evaluated by adding MSP to a female sample, with an endogenous MSP value of 11 μg/L, to a final concentration of 200 μg/L. The sample was then serially diluted with either serum or assay buffer, and the recovery was reported as the ratio between the expected concentration and the measured concentration.
The stability of Eu-MSP was tested by comparing the values of assay standards in runs with aliquots that were newly thawed or stored in the refrigerator overnight.
The statistical analyses were done with the SPSS 13.0 software (SPSS Inc, Chicago, Illinois). The correlations were calculated using Spearman ρ method.
Recombinant Protein Expression
MSP was produced by recombinant expression in insect cells according to the Bac-to-Bac system (Invitrogen AB, Stockholm, Sweden). The generation of a cDNA encoding human MSP has been described elsewhere (Ghasriani et al, 2006). The cDNA was cloned into the EcoRI site of the transfer vector pFastBac, and the insertion was verified by DNA sequencing. The recombinant plasmid was introduced in Escherichia coli H10Bac, carrying a bacmid with the baculo virus genome. Colonies containing recombinant baculo virus were identified by blue/white selection and the insertion was confirmed by polymerase chain reaction using the 17-mer and 16-mer M13/pUC universal and reverse sequencing primers (New England Biolabs, In Vitro Sweden AB, Stockholm, Sweden). Mini preparations of recombinant bacmid DNA were transfected into Sf 9 insect cells using Cellfectin (Invitrogen AB, Stockholm, Sweden). Recombinant baculo virus was harvested from the supernatant and titrated by viral plaque assay. The infection of Trichoplusia ni H5 cells was done essentially according to the manual provided by the supplier, with several pilot expressions to optimize the time and multiplicity of infection (MOI) for expression. The expression of recombinant MSP was monitored with the immunoassay and by Western blot as previously described (Valtonen-André et al, 2005).
Isolation of Recombinant MSP
Frozen culture medium (200 mL) from cells infected with the recombinant baculo virus was used for purification of MSP. Immediately after thawing, 94.4 g of (NH4)2SO4 was added to yield 70% saturation (ie, a concentration of 2.8 M). The solution was stirred gently with a magnetic stirrer at room temperature for 1 hour and then centrifuged at 5000 × g for 20 minutes. The precipitate was discarded, and 1 volume of 99.5% ethanol was added to 4 volumes of the supernatant. The mixture was stirred as above for 30 minutes and again centrifuged as above. The upper ethanol phase was saved, and 1 volume of 99.5% ethanol was added to 16 volumes of the remaining water/salt phase, followed by incubation and centrifugation as above. The ethanol phases from the first and second extraction were combined, and 1 volume of n-butanol was added to 4 volumes of the pooled ethanol phase. After 30 minutes of stirring, the phases were separated by centrifugation at 5000 × g for 10 minutes. The lower water phase was saved, whereas the ethanol phase was subjected to a new round of extraction with the same amount of n-butanol. Again, a water phase was formed, which was isolated as above and then combined with the water phase from the first extraction. The water phase was dialyzed against 50 mM Tris pH 7.4 and 0.1 M NaCl concentrated by ultrafiltration with an Omega 3K filter (PALL Life Sciences, Ann Arbor, Michigan) to a final volume of approximately 1–2 mL and subjected to gel filtration at 4°C using a 1.6 × 100 cm Sephadex G-75 column (Amersham Biosciences). The column was eluted at a rate of 4.9 mL/h with a buffer containing 25 mM Tris-HCl pH 8.0, 0.15 M NaCl, and 0.02% NaN3, and 2.4-mL fractions were collected. The recovery was estimated by the immunoassay. The purification was also monitored by silver-stained SDS-PAGE with 12% polyacrylamide gels using the Mini PROTEAN II system (Bio-Rad Laboratories AB, Sundbyberg, Sweden) (Laemmli, 1970).
The high-affinity binding between streptavidin-coated microtitration strips and the biotinylated GAR catcher was used in the first incubation step of the assay. Different GAR concentrations were tested, ranging from 100 to 1000 ng/200 μL. There was an increase in the signal that paralleled the increase of GAR concentration, but at above 300 ng/well, the signal leveled out, and this amount was selected for the assay. The performance of the MSP antiserum was evaluated by monitoring the decrease in fluorescence when serial dilutions of the antiserum were incubated with 1 ng of tracer. An antiserum dilution of 1:32 000, which yielded a signal that was approximately 80% of the maximal signal, was selected. When no antiserum was included, the nonspecific binding represented only about 1% of the maximum signal.
Seven standard points were chosen in the range of 0–130 μg/L, and a standard curve was generated with the logarithmic concentration on the x-axis plotted against the ratio between the detected signal and the maximal signal on the y-axis (Figure 1). The sensitivity, or lower limit of detection of the assay, was 4.9 μg/L. The intraassay CV for samples on 3 levels were 13% (x̄ = 5.5 μg/L), 3% (x̄ = 47 μg/L), and 4% (x̄ = 97 μg/L). The interassay CV for each of the 3 levels was 6% (x̄ = 5.0 μg/L), 3% (x̄ = 55 μg/L), and 3% (x̄ = 88 μg/L). The analytical recovery of added purified MSP ranged from 102% to 109% and was independent of whether serum or buffer was used as dilution matrix, and dilution series of the seminal plasmas down to 1:50 000, had good reproducibility. Also, the stability of Eu-MSP was high, yielding comparable values in the assay independent of whether the tracer was freshly thawed or stored in the refrigerator overnight.
MSP in Serum and Seminal Plasma
The MSP concentration was measured in serum and seminal plasma from 205 young males, and the median values were calculated. Five serum values were below the analytical detection limit of the assay and were reported as 0 μg/L. The median MSP concentration in serum was 12 μg/L (2.5–97.5 percentile, 4.9–26 μg/L) and in seminal plasma 0.53 g/L (2.5–97.5 percentile, 0.13–2.0 g/L). The total amount of MSP in the seminal plasmas showed a median of 1.8 mg (2.5–97.5 percentile, 0.32–6.6 mg). The MSP concentration in serum correlated significantly to both the concentration (r = .47, P < .001) and the amount (r = .50, P < .001) of MSP in seminal plasma (Figure 2). Hence, the levels of MSP in serum normally corresponds to less than 10−4 of the levels in semen.
Correlation Between MSP and Other Prostate-Secreted Components
Associations between MSP and PSA in seminal plasma and serum were assessed (Table). MSP in serum was not significantly associated with either tPSA or fPSA in serum or the PSA levels in seminal plasma. However, similar to the previously reported findings of a significant association between PSA in seminal plasma and fPSA in serum (Sävblom et al, 2005), there was also a low but statistically significant correlation between MSP in seminal plasma and fPSA in serum. The highest correlation was seen between MSP and PSA in seminal plasma (Figure 3). These correlations were not importantly affected (but were slightly higher) when the amounts of the substances in seminal plasma, instead of the concentrations, were used in the calculations. As expected, the correlation between the concentration of Zn2+ in seminal plasma (median Zn2+ of 1.5 mM; 2.5–97.5 percentile, 0.29–4.1 mM) and MSP in seminal plasma was also high (r = .54, P < .001) (Figure 4), whereas there was no association between the concentration or total amount of Zn2+ in seminal plasma and the concentration of MSP in serum.
Expression and Purification of Recombinant MSP
Recombinant MSP was produced in insect cells using the baculo virus expression system. Infection at a MOI of 100 yielded an MSP concentration of 35 mg/L in the culture medium after 96 hours of growth. The recombinant protein displayed the same size as native MSP on SDS-PAGE and gave a strong and specific reaction with the polyclonal antiserum in Western blotting. A 2-step method based on ethanol extraction and gel filtration was developed for the purification of recombinant MSP from the culture medium. The recovery of MSP from the ethanol extraction step was 80%–90%. The extraction also reduced the volume considerably, to around 5%–10% of the initial cell culture volume. The purity of the preparation, assessed with silver staining after SDSPAGE, showed enrichment of MSP and some other low- and high-molecular mass proteins after the ethanol extraction. To further purify MSP, the material was subjected to gel filtration on a Sephadex G-75. This final step yielded a homogeneous MSP preparation with very few low abundant contaminants, as shown on the silver-stained SDS-PAGE gel (Figure 5).
In this study, we demonstrate a correlation between MSP in serum and seminal plasma, which supports the potential use of MSP as a serum marker of prostate secretory function. Interestingly, the correlation is even higher than those previously reported between the secretory release of PSA into seminal plasma and the levels of fPSA in serum (Sävblom et al, 2005). One reason might be because of differences in the ability of the molecules to pass from the prostate into the blood stream. Another explanation could be differences in plasma clearance rates. In contrast, neither PSA nor MSP in seminal fluid correlates to the levels of tPSA in serum, which further illustrates that accumulation of tPSA in the blood is not importantly associated with the secretory capacity of the prostate.
In an earlier study, it was shown that MSPs in serum of women have a median concentration of around 65% of the values in men (Abrahamsson et al, 1989). The source of this nonprostatic MSP could be the trachea or the stomach because the expression in these organs is high. Such extraprostatic MSP most likely affects the serum values measured in this study, but the contribution of prostate-secreted MSP is presumably high, given the correlation between the serum and seminal plasma values. The median value of 12 μg/L reported for the serum MSP concentration in our study of military conscripts is double that of the median value reported in a previous study (Abrahamsson et al, 1989). This could be at least in part because of differences in the composition of each study cohort, but most likely it is also due to differences in assay design, and comparison between values obtained with dissimilar methods which are not uniformly standardized should be interpreted with great caution.
As expected, the prostate parameters measured in seminal plasma all closely correlate with MSP in seminal plasma. The strongest correlation was that between MSP and PSA, but their covariance in seminal plasma is not currently known to reflect any functional relationship. It is well known that the Zn2+ found in seminal plasma is secreted mainly by the prostate, and the concentration in the ejaculate is around 100-fold higher than that in blood plasma (Mann and Lutwak-Mann, 1981; Shivaji et al, 1990). In semen, Zn2+ also strongly correlates with MSP in this study. Whether this cation has any biological role associated with MSP is not known, and because MSP has no Zn2+ binding sites, the correlation could merely reflect their common glandular secretory origin.
MSP is a stable protein in various laboratory settings, and an investigation of the proteolytic degradation of human seminal plasma proteins under acidic conditions has shown that MSP was one of the proteins most resistant to proteolysis by the aspartic protease progastricsin (Szecsi and Lilja, 1993). Stability in the face of proteolytic degradation, combined with covariation with other prostate-secreted components and to the serum levels of MSP, suggests that MSP could be a suitable semen marker for the secretory function of the prostate. A major advantage with the newly developed MSP assay is the microtiter plate format, which has enabled automation of the assay. Its usefulness was clearly demonstrated with the samples from military conscripts because the sensitivity of the assay made it possible to measure MSP in 98% of the serum samples and, after appropriate dilution, in all of the seminal plasma samples.
An MSP assay for clinical use, requires reliable access to protein for preparation of reagents, and the generation and purification of recombinant MSP described in this paper enables us to produce considerable amounts of the protein without having to rely on sperm donors. The baculo virus system was selected because of the high yield and the use of eukaryotic cells, with their higher probability of correct folding of the peptide chain that contains 5 disulfides. The expression in insect cells could affect posttranslational modifications, but because MSP is nonglycosylated and also is lacking other modifications, this was not a problem. It has previously been shown that ethanol added to seminal plasma containing 2.8 M (NH4)2SO4 forms an organic phase that is enriched with MSP (Fernlund et al, 1994). We used the same strategy to enrich MSP from the culture medium. By adding n-butanol to the organic phase, the polarity is decreased in such a way that a water phase is recreated and simultaneously MSP is quantitatively transferred to the water phase. A similar procedure has been described for the purification of Green Fluorescent Protein from the jellyfish Aequorea victoria (Yakhnin et al, 1998). The modified method used in this study presents an excellent yield of MSP after ethanol extraction, and the subsequent gel filtration chromatography gives an MSP product of high purity.
In this study, we developed an immunoassay that we used to measure MSP in serum and seminal plasma of 205 young males. The levels of MSP in serum were found to reflect the prostate release of MSP, and strong correlations were also found in seminal plasma between MSP and the prostate parameters PSA and Zn2+. Furthermore, the methods described for the generation and purification of recombinant MSP will facilitate future MSP studies and assay designs.
Margareta Persson, Gun-Britt Eriksson, Kerstin Håkansson, and Ingrid Wigheden are acknowledged for excellent technical assistance.
Supported by the MAS Cancer Foundation and The Swedish Cancer Society, projects 3555 and 4423.