Printer Friendly

Expression of prostate-specific antigen and prostate-specific membrane antigen transcripts in blood cells: implications for the detection of hematogenous prostate cells and standardization.

The serum biomarkers prostatic acid phosphatase and prostate-specific antigen (PSA) have both been widely used for diagnosis and clinical monitoring of patients with prostate cancer [1].(5) However, they are not accurate enough to predict occult invasion or metastatic disease, which are diagnosed at the time of radical surgery in almost half of the patients believed to have localized disease [2]. Detection of circulating PSA-positive malignant cells by amplification of PSA mRNA has opened new expectations [3] and proved to be useful to identify cancer cells in lymph nodes [4]. To better identify early extra-prostatic spread of cancer, blood samples of patients with clinically localized or metastatic disease have also been assessed [5-7]. The nested PCR method was introduced to improve the limit of detection of PSA mRNA detection [8-10]. Even with reported sensitivities of one positive cell/ [10.sup.6] negative cells, 80-85% of the patients with early prostate cancer were found negative for PSA mRNA by PCR. This was also true in 50-70% of patients with advanced disease, therefore questioning the clinical value of molecular staging [11]. Attempts have been made to improve further the limit of detection of the technique by using an increased number of amplification cycles in each round (up to 40 for each) of the nested reverse transcriptase (RT)-PCR reaction [12]. However, this procedure detected PSA transcripts in normal peripheral blood, constituting a major limitation for cancer clinical staging.

Prostate-specific membrane antigen (PSM), an integral transmembrane glycoprotein, is another promising prostate-specific marker [9]. The sensitivity of PSM mRNA as a prostatic marker has already been investigated, with a nested PCR procedure, in a couple of studies [9, 13]. Both concluded that, stage for stage, the RT-PCR detection was more often positive for PSM than for PSA, regardless of androgen deprivation therapy, which is known to increase the expression of PSM. Using a conventional nested RT-PCR (25 cycles-restricted PCR round) and the same standard PSA- and PSM-positive control as in previously validated assays [9, 13], we assessed the expression of PSA and PSM mRNA in a wide range of negative controls, including many hematological cell lines and a large cohort of healthy blood donors. The intraassay variation was explored by repeating several times (from 4 to 39 times) the PCR. We confirm that PSA and PSM mRNA are found in cell lines of nonprostate origin as well as in normal blood cells. It is noteworthy that PSA has been induced by steroids in the MCF-7 breast cell line [14] but remained undetected at a basal concentration in other studies [9, 12] even after 2 x 40 PCR rounds [12]. In the current study, PSA mRNA was found at a very low concentration in the unstimulated MCF-7 cell line and readily detected after exposure to cycloheximide, a protein inhibitor and enhancer of illegitimate transcription. Current data highlight the need for extended quality-control study for preventing amplification of illegitimate transcripts.

Materials and Methods


(a) Nonhematological cell lines (CL) investigated in this study comprised: LNCaP, a prostate carcinoma CL expressing PSA and PSM, obtained from S. Loric, Hopital Necker, Paris, France; MCF-7, a breast adenocarcinoma CL and Chinese hamster (Cricetulus griseus) ovary transformed fibroblast CL (CHO) were obtained from ATCC. Other MCF-7 clones were also obtained from A. Belliard, Laboratoire de Biologie Generale, Universite de Liege, Sart-Tilman, Liege, and from J.-P. Delville, Universite Libre de Bruxelles, Bruxelles; (b) hematological CL tested were the following: U937, a histiocytic CL; Jurkat and CEM, two T-cell acute lymphoblastic leukemia CL; KG1, K562, and HL-60, three nonlymphoid CL; IM9, a B-lymphoblastoid/multiple myeloma CL; 8226, a multiple myeloma CL; DG75 and Raji, two Burkitt lymphoma CL; and DOHH2, a follicular lymphoma CL. Except for 8226 (ECAAC) and DG75 (the German Collection of Microorganisms and Cell Cultures--DSMZ), all the other CL were obtained from ATCC. They were grown with RPMI 1640 and 100 mL/L fetal calf serum, supplemented with 105 U/L penicillin and 0.1 g/L streptomycin; confluent cultures were dispersed with trypsin-EDTA once a week as described [13]; (c) blood specimens used as negative controls were obtained from 29 healthy blood donors: men (n = 15; ages 23 to 47, mean age 32, median age 30) and women (n = 14; ages 23 to 70, mean age 39.5, median age 38.5).


(a) Venous blood (2 x 10 mL) was collected in EDTA-treated tubes (Sarstedt) and processed within 1 h. The buffy coats obtained after centrifugation on Ficoll Hypaque were washed and placed directly in 1.5 mL of Trizol[TM] reagent (Life Technologies, Gibco-BRL). They were processed immediately or frozen at -80[degrees]C. The procedure for total RNA isolation was carried out according to the manufacturer's instructions; 2.5 [micro]g of total RNA were reverse transcribed with pdN6 (random hexamers: 1.6 [micro]mol/L; Expedite Nucleic Acid Synthesis System, Millipore) [4, 9] and the Superscript[TM] RNase [H.sup.-] reverse transcriptase (Gibco-BRL). Amplification of the cDNAs was based on previously described procedures [9, 13] and performed in a DNA thermal cycler 480 (Perkin-Elmer). An initial denaturation step (95[degrees]C for 2 min) was followed by repeated cycles (94[degrees]C for 50 s, 61[degrees]C for 50 s, and 72[degrees]C for 90 s) and a final extension for 10 min.

The procedure for the amplification of the PSA and PSM cDNA sequences was identical except that 0.5 U of Dynazyme DNA polymerase (Finnzyme Life Science) was used for PSM and 1 U of Ampli Taq DNA polymerase (Perkin-Elmer) for PSA: 50 [micro]L of PCR mix included the Dynazyme or Taq DNA polymerase as mentioned above, 1.5 mmol/L Mg[Cl.sub.2,] 200 [micro]mol/L deoxynucleotide triphosphates, and 200 nmol/L of each primer. In the second PCR round, 2.5 [micro]L of the first PCR product was further amplified with the nested primers for 25 cycles. Ten microliters of the product was electrophoresed on a 2% agarose gel and visualized by ethidium bromide staining. Assays were repeated several times, with the same RNA, to confirm the results.

(b) In a separate set of experiments, we compared the amplification of PSA and PSM transcripts using 0.5 U of Dynazyme or 1 U of AmpliTaq DNA polymerase. This analysis was performed, for PSA and PSM, on cell lines (MCF-7, IM9, Jurkat, HL-60, K562) and 10 healthy blood donor samples.
The sequences are as follows:


New primers designed on the basis of sequence data obtained from the European Molecular Biology gene bank were used to avoid amplification of the highly homologous human glandular kallikrein gene. The predicted PSA outer primers amplified product is 485 bp, whereas the nested PSA product is 355 bp. PSM amplification yielded an outer amplified product of 416 by and a nested product of 193 bp. PSM primers were identical to previously described primers [13], except for their 5' ends, which were shortened.


A strict procedure was followed to avoid cross-contamination with PCR products. RNA is extracted in a pre-PCR room totally remote (at another floor) from the laboratory where cDNA amplification is performed. The first and second PCR rounds are not performed in the same room. PCR products are manipulated in a third distant post-PCR room. Each step (i.e., RNA extraction, first and second amplifications) is programmed on different days.

For cell lines as well as for the control blood samples, mRNA extraction was carried out one by one to avoid cross-contamination between samples. In any series of reactions, contamination at the DNA level was excluded by performing PCR analysis without reverse transcriptase. A water control, containing no cDNA template, was also used to detect carryover; cDNA isolated from a positive control cell line (LNCaP) and negative control (CHO) was amplified in each set of PCR. For LNCaP, cDNA diluted 1/10 was used for both first- and second-round PCR to decrease the risk of contamination by PCR products.

Amplification of the protooncogene Abelson was originally set up to detect the bcr-abl rearrangement characterizing Philadelphia-positive hematological neoplasms [15]. It is commonly used in our department as a reverse transcription and amplification internal control for RTPCR. Primers derived from the Abelson gene sequence were therefore used accordingly for all the human cell lines and normal blood specimens.

RT-PCR was performed with random hexamers, primers, dNTPs, Mg[Cl.sub.2], and Dynazyme DNA polymerase as described above. It is noteworthy that Dynazyme and AmpliTaq can be indistinctly used for Abelson amplification (personal data). An initial denaturation step (95[degrees]C for 2 min) was followed by 32 PCR cycles (94[degrees]C for 50 s, 60[degrees]C for 50 s, 72[degrees]C for 1 min) and a final extension for 10 min.

A pair of primers specific for the housekeeping gene coding ribosomal protein S14 (accession number: Genbank M35008) was designed to control both RNA transcription and cDNA amplification steps in the CHO cell line. Sense and antisense primers were located in exon 2 and exon 3, respectively:

The PCR profile was as described above for PSA and PSM cDNA amplification.

For MCF-7, an additional mRNA extraction was performed on cells cultured for 7 h in a medium containing 1 g/L of the protein synthesis inhibitor cycloheximide (Sigma) as described [16].


We examined the effect of gradually decreasing the number of cycles (25, 23, 21, and 19) for PSA and PSM, using a few cell lines (IM9, Jurkat, MCF7, CEM, and Raji) and normal blood donors (for details, see Table 3). We used healthy blood donors' cDNA samples, which had been already assessed for PSA and PSM amplification. For each particular sample, this comparison was made during the same PCR procedure: tubes having completed the requested number of cycles were sequentially removed from the PCR thermal cycler.


The detection limit of our protocol was tested in three distinct ways:

(a) LNCaP mixed with fresh venous blood: We prepared three suspensions containing [10.sup.6] LNCaP in 10 mL of RPMI 1640. Each suspension was assessed in triplicate on a Kova[R] Glasstic[R] Slide 10 with grids (Hycor Biomedical), according to the manufacturer's instruction, and pooled. Tenfold serial dilutions were then performed. The detection limit was assessed by mixing 2 mL of each dilution of LNCaP cells to 20 mL of fresh venous blood containing 5.2 x [10.sup.9]/L white blood cells (Technicon H*3 RTC, Bayer Diagnostics Division). Total RNA was extracted from the mixed specimens and was assayed for PSA and PSM in individual reactions by RT-PCR, as described above. To assess the reproducibility, we repeated LNCaP mixing in healthy blood donors three times over 6 months, assessing each aliquot in triplicate.

(b) LNCaP mixed in K562, a PSA- and PSM-negative cell line: Limit of detection was performed, as described above, by mixing 2 mL of serial tenfold dilutions of LNCaP (1 to 108/L) with [2.10.sup.7] K562, a cell line showing no transcription of PSA or PSM (in the experimental conditions of the study). As for fresh venous blood, the supplementation experiment with K562 was repeated three times over the last 6 months. In both assays, PCR reactions were performed in triplicate on cDNA from the same extraction.

(c) Serial dilution of plasmids containing PSA and PSM cDNA: Limit of detection was determined by performing PCR with serial dilutions of PSA cDNA cloned into pGEMPSA-7 (gift from A. Lundwall, Lund University, Malmo, Sweden). PSM cDNA (1.2 kb) was amplified with the following primers, and cloned into BamHI and SalI-sites of pJRD184 [17]:

Before serial dilutions, PSA and PSM cDNA containing plasmids were first linearized with the digestion enzymes EcoRI and XhoI, respectively. Total digestion was controlled on agarose gel. Linearized plasmids were diluted in 10 mmol/L Tris buffer, pH 7.80, containing 1 g/L salmon sperm DNA, as described [18]. Aliquots containing 1 to [10.sup.5] molecules were assessed in triplicate.

For each of the three methods, we reassessed the last dilution giving a PCR-positive result and the following first PCR-negative dilution by performing 10 additional PCBs to determine the CV. The limit of dilution was defined as the last dilution giving a positive PCR result with a CV <20%.


Sequence analysis was performed on PSA cDNA from several cell lines (LNCaP, MCF-7, IM9, CEM, and Jurkat) as well as every PSA RT-PCR-positive blood donor. PSM cDNA sequencing was performed on KG1, MCF-7, and four normal donors. The targeted products were sequenced in both orientations. The sequencing reactions were carried out on an automated ABI 373 A apparatus (Applied Biosystems) by using the Taq Dye Deoxy Terminator Cycle Sequencing kit from the same manufacturer according to its instructions.

Specificity of the remaining PSA and PSM cDNA PCR products was assessed by transfer onto Hybond N+ nylon membranes (Zeta-Probe, Bio-Rad) and hybridization with the internal PSA or PSM oligoprobes, respectively. Southern transfer, probe radiolabeling, hybridization, and autoradiography were performed by standard techniques as described [13].


An IRMA (Hybritech) was used to quantify PSA protein both in culture medium and in cell lysates according to the manufacturer's instructions. Lysates were prepared as described [12]. Briefly, cells pellets were frozen at -80[degrees]C and thawed at 37[degrees]C three times. Debris were discarded after centrifugation, and the supernatant was frozen at -80[degrees]C until assayed.



Results, which are illustrated in Fig. 1A, are summarized in Table 1 for CL and in Table 2 for normal blood samples. As expected, the PSA signal was consistently found in the PSA-expressing LNCaP after the first PCR run. While PSA mRNA remained undetectable in HL-60, 8226, K562, and CHO cells, it was occasionally found after the second PCR round in MCF-7 previously reported as PSA and PSM negative [9, 12]. Positivity was confirmed with MCF-7 from distinct origins. PSA cDNA was also found in several hematological CL. The frequency of positivity varied from line to line, some giving frequent positive results (IM9, Jurkat, CEM, Raji) while others being only occasionally positive (KG1, DG75, U937, DOHH2). The performance of the Dynazyme and AmpliTaq DNA polymerase compared on MCF-7, IM9, Jurkat, and KG-1 cDNA gave a positive signal with the same intensity and at the same frequency (data not shown).


Interestingly, healthy blood donors were also found positive for PSA mRNA (Table 2). A total of 6 of 31 (19.4%) healthy blood donors showed at least one blood sample positive for the PSA mRNA (Table 2). The frequency reached 11.8% (2 of 17) in men and 28.4% (4 of 14) in women. The nested PCR was repeated several times for each donor from the same cDNA. Considering the total number of positive PCBs, we found 8 of 127 (6.3%) positive results, only 3% (2 of 66) in men and 9.8% (6 of 61) in women.


Results, which are illustrated in Fig. 1B, are also summarized in Table 1 for cell lines, and in Table 2 for normal blood samples. The frequency of PSM mRNA detection (30-100% of the RT-PCR reactions) was usually higher than what was observed with PSA on the same cell line; 5 of 11 assessed cell lines remained negative, including U937, DOHH2, HL-60, K562, and CHO. Cell lines expressing a PSA-positive and a PSM-negative signal were also found (U937, DOHH2). Also, a positive PSM signal was detected in each MCF-7 sample, though the line is considered to lack PSM expression. Interestingly, the same consistent positivity was observed with IM9. PSM mRNA was frequently detected in healthy blood donors: A positive signal was identified in most of the PCR reactions (24 of 25; 96%).

The performance of the Dynazyme and AmpliTaq DNA polymerase was compared on a few cell lines (MCF-7, KG-1, HL-60, and CHO) and healthy blood donors' cDNA. As with PSA, positive signals obtained were comparable (data not shown).


The progressive decrease in the number of cycles prevented the amplification of illegitimate transcripts (Table 3). It was evidenced for PSA at 2 x 19 cycles in cell lines, but not in blood donor samples, no PSA-positive signal being observed in this set of experiments. The same observation was made on cell lines and healthy blood donors with PSM primers. Except for IM9, which remained consistently positive throughout the experiment, other cell lines and blood donor samples, including those still positive at 21 cycles, were negative at 19 cycles.


The limit of detection for PSA and PSM, after two rounds of 19 or 25 cycles, in either venous blood or cell line K562, is given in Table 4. For each method, results were reproducible in the three successive sets of experiments (identical lowest limit of detection as determined by a CV <20%). Because of the amplification of illegitimate transcripts, no evaluation could be done with PSM primers in whole blood at 2 x 25 cycles. The performance of the PCR assay on serial dilutions of plasmid containing the PSA or PSM cDNA insert was equivalent (Table 4, Fig. 2A and B). At 2 x 19 cycles, 100 copies were consistently and strongly detected, whereas dilutions containing 10 copies gave only a very faint band in 58% (11 of 19) of the PCR reactions. At 2 x 25 cycles, dilution corresponding to 1 cDNA molecule gave a strong PCR signal with both cDNA PSA and PSM plasmid in 65% (25 of 39) and 70% (27 of 39) of the reactions, respectively. In comparison, the following dilution gave a percentage of positivity for PSA and PSM cDNA of 12% (6 of 39) and 14% (5 of 39) respectively. These results are in accordance with the Poisson distribution.


Negative controls (control without reverse transcription and water control) remained consistently negative all through the experiments. Unexpected results prompted us to include additional controls: (a) positive results in cell lines were reassessed after extraction of those cells together with the CHO cell line, which remained consistently negative for both markers throughout the experiments; (b) original PSA primers and new PSA primers, the latter especially selected to differ from kallikrein sequences, were compared and hybridization with the internal PSA oligoprobe confirmed the origin of the bands detected (data not shown); (c) PSA amplification of mRNA extracted from MCF-7 with and without exposure to cycloheximide was positive in 10 of 10 and 3 of 10 PCR reactions, respectively (Fig. 3).


Sequencing of the 355-bp PSA PCR products obtained with various cell lines (LNCaP, MCF-7, M49, CEM, and Jurkat) demonstrated a perfect homology between amplified cDNA and the PSA-specific sequence (data not shown). In the Jurkat cell line, a point mutation (G to A; Asp to Asn) mapped in exon 3 at position 345. To verify the allelic status of this mutation, we also sequenced the genomic DNA corresponding to exon 3 by using the following sense and antisense primers: 5'-ACACA000CAGGTATTTCAG-3' from Israeli et al. [9] and 5'-CAGGCGTACACTCCTCTGG-3', respectively. The presence of a heterozygous base pair substitution was demonstrated.


Although sequence analysis of the 193-bp PSM PCR product obtained from KG1 was homologous to the reported PSM sequence, amplification product from MCF-7 revealed the presence of a silent point mutation (polymorphism at position 993: T to C). The same analysis was repeated with the PSM-PCR product from four healthy donors and indicated an identical polymorphism in three of them.


As illustrated in Fig. 1A, smaller-size PSA-PCR products (~230 or ~280 bp) were occasionally found in some positive samples, either as a single band (MCF-7, 1 of 10; IM9, 1 of 20; Jurkatt, 6 of 22; DG75, 1 of 2; U937, 1 of 1; normal blood control, 1 of 2) or in addition to the expected 355-bp fragment (MCF-7, 1 of 10; IM9, 6 of 20; CEM, 4 of 17; Raji, 2 of 6). Sequencing of these fragments derived from either MCF-7 or Jurkat revealed, for both products, a perfect homology with the PSA sequence except for the presence of a deletion that spans 123 by (from nucleotide 396 to 519) in MCF-7 and 71 by (from nucleotide 307 to 378) in Jurkat. It is noteworthy that homologous sequences mapped at both extremities of each deletion (CAGC for the 123-bp deletion, CAGGCCAGGT for the 71-bp deletion).


For each cell line, an IRMA was performed on the culture supernatants and cytosolic fractions. Although consistently positive for the LNCaP culture medium and cytosolic fraction, both assays failed to detect the PSA protein in any of the other cell lines, even those frequently found to express PSA mRNA (data not shown).


Cross-contamination is a major drawback potentially affecting specificity and reproductibility of nested PCR reaction. Special efforts were made to control the results as emphasized in Materials and Methods and Results. Sequencing of the PSA or PSM amplified signals, obtained from cell lines or normal blood, confirmed that PSA and PSM mRNAs were entirely identical to mRNAs found in prostate cells. Moreover, the procedure identified a PSM polymorphism in some but not all of the PSM-positive blood samples, and a heterozygous point mutation in the PSA sequence of the Jurkat cell line, hence ruling out a cross-contamination.

Our results extend previous observations on the detection of PSA in normal peripheral blood and nonprostatic malignant or normal tissues. PSA mRNA and (or) protein has indeed been detected in cell lines from either myeloid (HL-60), lung (SK-MES-1), or ovarian (BG-1) origin [12] and in clinical samples such as normal or malignant breast tissues, breast secretions, and ovarian carcinoma [18, 19]. Among the breast tumor cell lines, the androgen-stimulated T47D, but not BT-20, was reported to express PSA mRNA [18, 20]. Although induction of PSA can be achieved by steroid stimulation in the MCF-7 breast cell line [14], lack of PSA mRNA and protein expression has been claimed for unstimulated MCF-7 after nested PCR procedure on the basis of two rounds of 25 cycles (25 x 2) [9] or even two rounds of 40 cycles (40 x 2) [12].

We did not obtain similar results with our assay. Surprisingly, we detected a specific PSA mRNA in about 25% of the PCR reactions with MCF-7. PSA mRNA was detected randomly in the PCR tubes, attesting for its very low basal expression. These results were confirmed with three MCF-7 clones from different origins, excluding a potential subclonal evolution characterizing one particular cell line. Our data are therefore in agreement with reports on the positivity of breast carcinomas and some breast tumor cell lines. Discrepant results could be explained by a variable level of transcription of the PSA gene in malignant cells originating from the same original clone. This phenomenon has indeed already been stated for PSA mRNA in LNCaP [21]. We also found a promiscuous expression of PSA mRNA in a wide variety of hematological cell lines and normal blood controls, but compared with Smith et al. [12], it occurred already after 2 x 25 cycles of amplification. Interestingly, several PCR reactions performed in parallel on cDNA from the same extraction showed a random detection of the specific message in blood samples and most cell lines. The frequency of positive results varied from line to line. Most importantly, Table 2 clearly points out that the frequency of positivity can be as high as 12% in men's samples if only tested once, but drops to 3% upon multiple testing of the same cDNA. This highlights the stochastic effects in RT-PCR results at high sensitivity, hence the importance of repetitive testing in clinical samples.

In Smith et al.'s study [12], detection of PSA mRNA was clearly related to a very high limit of detection of the assay, since the frequency of PSA-positive normal blood samples reached 100% when a 40 x 2 cycles procedure was used and dropped to 77% with a 25 x 2 cycles amplification. Our 25 x 2 cycles assay for PSA mRNA appears, however, less sensitive, with only 20% of the normal blood controls being found, at least once, positive for PSA, HL-60 being repeatedly negative. It is interesting to note that HL-60, positive in Smith et al.'s study, was also found negative by Galvan et al., who used a highly sensitive procedure combining RT-PCR and time-resolved fluorometry [22]. Whereas the limit of detection of Smith et al.'s assay was not assessed, our level for PSA, at 2 x 25 cycles, was identical to previous data reporting a level of 1 LNCaP in 106 PSA-negative cells [7] and about 2 LNCaP in [10.sup.6] nucleated blood cells [8-10]. In terms of PSA cDNA copies detection, our limit of detection was <100 times that for time-resolved fluorometry (160 cDNA copies for a signal-to-background ratio of 10) [22] but in the range of two other groups who detected <10 molecules [18], or 5 copies of PSA cDNA plasmid by hot-start RT-PCR protocol [23]. It is noteworthy that the latter group lowered the limit of detection to one LNCaP in [10.sup.8] peripheral blood mononuclear cells and samples of blood and bone marrow from healthy donors, a level much higher than what was reached by any other protocol so far. Despite this unusual performance, they did not report any false-positive result. The reasons for this major discrepancy with our data remain so far only hypothetical. As discussed above for MCF-7, this could be due to the varying amount of PSA mRNA in LNCaP in response to cell growth and number of passages [21]. In our study, a number of cycles decreased to 19 x 2 clearly avoided amplification of PSA transcripts in the cell lines assessed, except in LNCaP, and affected greatly the limit of detection. Similarly to Smith et al.'s data, PSA protein was detected neither in the supernatants nor in the cellular lysates of any cell line except in the positive control LNCaP.

The detection of PSM mRNA was based on a previously described assay [13]. The primers used were identical although shortened at their 5' end. In some lines (MCF-7, IM9, Jurkat, KG-1), a positive PSM signal was detected in nearly all the PCR reactions or 50% of them (8226), which was more frequent than what was observed with PSA mRNA. Like PSA, PSM mRNA was not detected in a few cell lines (HL-60, K562, CHO). It is noteworthy that the great majority of the blood donors were found positive, often repeatedly. Altogether, these findings are consistent with a frequency of PSM mRNA detection usually higher than PSA, not only in cell lines but also in normal peripheral blood. This is in accordance with a higher limit of detection of the PSM signal, evidenced previously [9, 13], and with the unexplained positivity of some negative controls restricted to PSM but not PSA mRNA amplification [9]. Accordingly, the detection limit was assessed with PSM primers at two rounds of 25 cycles. The result obtained by mixing LNCaP in PSM-negative K562 was identical to Israeli et al.'s data with MCF-7 [9]. At that limit of detection, illegitimate transcripts were, however, detected in whole blood, making impossible the evaluation of limit of detection with nucleated blood cells. A similar observation, based on the use of Israeli et al.'s primers [9], has just been reported by another group during the review processing of this manuscript [24]. Although data on limiting dilution of PSM cDNA are lacking in the literature, making impossible any comparison of results, the similarity of PCR results after amplification of PSA and PSM cDNA was striking. This demonstrates an equal performance of the assay for amplification of both markers, hence its suitability for their simultaneous quantification on the same samples.

A decrease in the number of cycles reduced the amplification of PSM transcripts in all blood donor samples and cell lines assessed, except IM9 and LNCaP. Detection limit was assessed at 2 x 19 cycles because no positive signal was found in blood donor samples with this procedure. Whereas the limit of detection dropped to 1 LNCaP in [10.sup.4] K562 cells, about 2 LNCaP were detected in [10.sup.6] nucleated cells, which is very close to Loric et al.'s data [13]. Experiments performed on both sites on the same clinical samples and on the same aliquots of LNCaP mixed with fresh venous blood confirmed that our respective assays have a similar although not yet identical limit of detection (data not shown).

The current and previous data illustrate the need for multicentric standardization of all currently used RT-PCR assays designed to detect PSA- or PSM-positive prostatic cells in extraprostatic tissues and fluids. The potential limitations of the method due to detectable PSA mRNA in nonprostate cells have already been emphasized [12, 25]. Our data confirm this point and demonstrate that such limitations apply to the PSM marker, which is also expressed in a nonspecific manner. Absence of protein expression associated with the presence of prostate-specific transcripts in human nonprostate cells corresponds to a well-known phenomenon called illegitimate transcription [26]. It is characterized by the presence of an extremely low number of transcripts, even <1 mRNA per 100 or 1000 cells, of a tissue-specific gene in any cell types. Illegitimate transcripts are related to a low-level activity of the normal gene promoter [16] and allow, by their presence in normal blood lymphocytes, the study of inherited diseases [27]. A comparable phenomenon helped us to identify polymorphisms or heterozygous mutations in some PSA and PSM cDNA nonprostate cell lines and healthy blood donors. Increase of mRNA transcripts in MCF-7 cells treated with cycloheximide is part of the same phenomenon and differs from the membrane receptor-mediated stimulation by steroids. Such an increase can theoretically be sustained by an enhancement of the mRNA stability, by the loss of labile transcriptional repressors, or the inhibition of protein involved in a mechanism of autotranscriptional repression [16, 28, 29]. We can speculate that factors regulating directly or indirectly the transcription or the stability of the mRNA in normal nonprostate cells might affect the specificity of the assay used for the detection of prostate metastases (e.g., exogenous factors such as drugs, hormones), but confirmation will require further studies. Whether every cell is capable of generating illegitimate transcripts or, alternatively, if occasional cells in an otherwise nonexpressing tissue are able to produce comparatively high amounts of the transcript of interest has to be determined [27] but remains an important issue for the limit of detection of the assays.

In conclusion, an illegitimate transcription of PSA and PSM mRNA may be detected by apparently conventional assays. Current data point out that the ability to detect these transcripts depends directly on technical conditions and the cell type investigated. Although this required further studies, it could also be affected by factors modulating directly or indirectly the transcription and the mRNA stability. As PSA and PSM transcripts are sometimes detected at a very low frequency in nonprostate cells, quality-control procedures therefore require a repetitive testing and adjustment of the limit of detection on the basis of a large number of normal blood samples to avoid any "random" nonprostate-specific signal [30]. Multicentric standardization could adequately address this problem and determine a clinical cutoff for positivity.

We thank Etienne De Graeve for the statistical analysis and John Isaacs, Johns Hopkins Oncology Center, Prostate and Breast Center Laboratories, Baltimore, MD, for critical reading and useful discussion.


[1.] Oesterling J. Prostate specific antigen: a critical assessment of the most useful tumor marker for adenocarcinoma of the prostate. J Urol 1991;145:907-23.

[2.] Olsson CA. Reverse transcriptase polymerase chain reaction in prostate cancer. In: Murphy G, Griffiths K, Denis L, Khoury S, Chatelain C, Cockett AT, eds. Proceedings of the First International Consultation in Prostate Cancer. Jersey, UK: Science Communication International, 1997:121-32.

[3.] Moreno JG, Croce CM, Fischer R, Monne M, Vihko P, Mulholland SG, Gomella LG. Detection of hematogenous micrometastasis in patients with prostate cancer. Cancer Res 1992;52:6110-2.

[4.] Deguchi T, Doi T, Ehara H, Ito S, Takahashi Y, Nishino Y, et al. Detection of micrometastatic prostate cancer cells in lymph nodes by reverse transcriptase--polymerase chain reaction. Cancer Res 1993;53:5350-4.

[5.] Katz AE, Olsson CA, Raffo AJ, Cama C, Perlman H, Seaman E, et al. Molecular staging of prostate cancer with the use of an enhanced reverse transcriptase-PCR assay. Urology 1993;43: 765-75.

[6.] Olsson CA, De Vries GM, Benson MC, Raffo A, Buttyan R, Cama C, et al. The use of RT-PCR for prostate-specific antigen assay to predict potential surgical failures before radical prostatectomy: molecular staging of prostate cancer. Br J Urology 1996;77: 411-7.

[7.] Ghossein RA, Scher HI, Gerald WL, Kelly WK, Curley T, Amsterdam A, et al. Detection of circulating tumor cells in patients with localized and metastatic prostatic carcinoma: clinical implications. J Clin Oncol 1995;13:1195-200.

[8.] Seiden MV, Kantoff PW, Krithivas K, Propert K, Bryant M, Haltom E, et al. Detection of circulating tumor cells in men with localized prostate cancer. J Clin Oncol 1994;12:2634-9.

[9.] Israeli RS, Miller WH Jr, Su SL, Powell CT, Fair WR, Samadi DS, et al. Sensitive nested reverse transcription polymerase chain reaction detection of circulating prostatic tumor cells: comparison of prostate-specific membrane antigen and prostate-specific antigen-based assays. Cancer Res 1994;54:6306-10.

[10.] Jaakkola S, Vornanen T, Leinonen J, Rannikko S, Stenman UH. Detection of prostatic cells in peripheral blood: correlation with serum concentrations of prostate-specific antigen. Clin Chem 1995;2:182-6.

[11.] Diamandis EP. Prostate cancer, prostate-specific antigen, and the polymerase chain reaction [Editorial]. Clin Chem 1995;2:177-9.

[12.] Smith MR, Biggar S, Hussain M. Prostate-specific messenger RNA is expressed in non-prostate cells: implications for detection of micrometastases. Cancer Res 1995;55:2640-4.

[13.] Loric S, Dumas F, Eschwege P, Blanchet P, Benoit G, Jardin A, Lacour B. Enhanced detection of hematogenous circulating prostatic cells in patients with prostate adenocarcinoma by using nested reverse transcription polymerase chain reaction assay based on prostate-specific membrane antigen. Clin Chem 1995; 12:1698-704.

[14.] Yu H, Diamandis EP, Zarghami N, Grass L. Induction of prostate specific antigen by steroids and tamoxifen in breast cancer cell lines. Breast Cancer Res Treat 1994;32:291-300.

[15.] Cross NC, Hughes TP, Feng L, O'Shea P, Bungey J, Marks DI, et al. Minimal residual disease after allogeneic bone marrow transplantation for chronic myeloid leukaemia in first chronic phase: correlations with acute graft-versus-host-disease and relapse. Br J Haematol 1993;84:67-74.

[16.] Chelly J, Hugnot JP, Concordet JP, Kaplan JC, Khan A. Illegitimate (or ectopic) transcription proceeds through the usual promoters. Biochem Biophys Res Commun 1991;179:553-7.

[17.] Heusterspreute M, Kennedy N, Emery S, Ha Thi V, Tournis-Gamble S, Davison J. Restriction site bank vector IV. pJRD184, a 3793 plasmid vector having 43 unique cloning sites Gene 1985;39: 299-304.

[18.] Zarghami N, Diamandis EP. Detection of prostate-specific antigen mRNA and protein in breast tumors. Clin Chem 1996;3:361-6.

[19.] Diamandis EP, Yu H. New biological functions of prostate-specific antigen? [Editorial]. J Clin Endocrinol Metab 1995;80:1515-7.

[20.] Monne M, Croce C, Yu H, Dimandis EP. Molecular characterization of prostate-specific antigen messenger RNA expressed in breast tumors. Cancer Res 1994;54:6344-7.

[21.] Heston WDW. Detection of hematogenous dissemination of prostatic cancer by RT-PCR with primers specific for prostate-specific membrane antigen [Editorial]. Clin Chem 1995;12:1687-8.

[22.] Galvan B, Christopoulos TK, Diamandis EP. Detection of prostate-specific antigen mRNA by reverse transcription polymerase chain reaction and time-resolved fluorometry. Clin Chem 1995;12: 1705-9.

[23.] Corey E, Arfman EW, Liu AY, Vessella RL. Improved reverse transcriptase-polymerase chain reaction protocol with exogenous internal competitive control for prostate-specific antigen mRNA in blood and bone marrow. Clin Chem 1997;3:443-52.

[24.] Lintula S, Stenman UH. The expression of prostate-specific membrane antigen in peripheral blood leukocytes. J Urol 1997;157: 1969-72.

[25.] Henke W, Jung M, Luin M, Schelte H, Berndt C, Rudolph B, et al. Increased analytical sensitivity of RT-PCR of PSA mRNA decreases diagnostic specificity of detection of prostatic cells in blood. Int J Cancer 1997;70:52-6.

[26.] Sarkar G, Sommer S. Access to a messenger sequence or its protein product is not limited by tissue or species specificity. Science 1989;244:331-4.

[27.] Cooper DN, Berg LP, Kakkar W, Reiss J. Ectopic (illegitimate) transcription: new possibilities for the analysis and diagnosis of human genetic disease. Ann Med 1994;26:9-14.

[28.] Mahadevan LC, Edwards DR. Signalling and superinduction. Nature 1991;349:747-8.

[29.] Herrick D, Parker R, Jacobson A. Identification and comparison of stable and unstable mRNAs in Saccharomyces cerevisiae. Mol Cell Biol 1990;10:2269-84.

[30.] Jung RJ, Ahmad-Nejad P, Wimmer M, Gerhard M, Wagener C, Neumaier M. Quality management and influential factors for the detection of single metastatic cancer cells by reverse transcriptase polymerase chain reaction.


(1) Laboratory of Clinical Molecular Biology, Department of Biochemistry, Clos-Chapelle-aux-Champs, UCL 30.46, and Department of Urology, Ave. Hippocrate, 30, Cliniques Universitaires Saint-Luc, Universite Catholique de Louvain, 1200 Bruxelles, Belgique. (3) Queen Astrid Military Hospital, Rue Bruyn, 2, 1120 Bruxelles, Belgique. (4) Hopital Necker, Laboratory of Clinical Chemistry, 149, rue de Sevres, 75015 Paris, France.

(5) Nonstandard abbreviations: PSA, prostate-specific antigen; RT-PCR, reverse-transcribed PCR; PSM, prostate-specific membrane antigen; and CL, cell line.

* Address correspondence to this author at: Laboratoire de Biologie Moleculaire Clinique, Clos Chapelle-aux-Champs, 30 - UCL/30.46, B-1200 Brussels, Belgium. Fax 00/32.2.764.39.59; e-mail

Received July 24, 1997; revision accepted December 3, 1997.
Table 1. PSA and PSM transcripts: frequency of detection
 in tumor and hematological cell lines.

Cell lines PSA PSM

MCF-7 10/39 (25.6%) 19/19 (100%)
IM9 21/26 (80.8%) 15/15 (100%)
Jurkat 22/30 (73%) 10/12 (83.3%)
CEM 17/27 (63%) ND
KG-1 2/26 (7.7%) 13/15 (86.7%)
DG75 2/13 (15.4%) ND
Raji 11/16 (68.8%) 2/7 (29%)
U937 1/14 (7%) 0/9 (0%)
DOHH2 1/12 (8%) 0/9 (0%)
HL-60 0/16 (0%) 0/9 (0%)
8226 0/21 (0%) 4/9 (44.4%)
K562 0/22 (0%) 0/15 (0%)
CHO 0/30 (0%) 0/20 (0%)

The total number of PCR reactions and the frequency of
positive results (%) is given for each cell line.

ND: not done.

All cell lines are from hematological origin except
MCF-7 and CHO.

LNCaP was positive for PSA and PSM in
100% of the PCR reactions (data not shown).

Table 2. PSA and PSM transcripts: frequency of detection
in normal blood donors.

 Men (n = 17) Women (n= 14)

21 0/2 4/6 23 0/4 2/3
23 0/4 1/3 24 0/4 3/3
23 0/4 3/3 24 0/4 9/9
24 0/4 0/1 25 0/4 8/9
24 0/4 3/3 32 0/4 3/3
26 0/2 1/2 37 0/4 3/3
27 1/6 2/2 38 0/4 ND
28 0/4 2/3 39 0/4 ND
29 0/4 3/3 41 1/4 1/3
30 0/4 2/2 41 1/4 ND
30 0/4 2/3 43 0/4 2/3
36 0/4 5/6 52 0/4 ND
39 0/4 1/2 70 2/7 ND
39 0/4 6/6 70 2/6 ND
41 1/4 2/2
45 0/4 3/3
47 0/4 2/2

Donors with at least one
positive sample
Men (n = 17) 2/17 (11.8%) 16/17 (94.1%)
Women (n = 14) 4/14 (28.6%) 8/8 (100%)
Total 6/31 (19.4%) 24/25 (96%)
Positive nested RT-PCR
In men 2/66 (3.0%) 42/51 (82.3%)
In women 6/61 (9.8%) 31/36 (86.1%)
Total 8/127 (6.3%) 73/87 (83.9%)

The number of blood donors and the number of samples found
positive for PSA or PSM transcripts are given for men and
women. No significant statistical difference was observed
between men and women (Fisher test and [[chi].sup.2] with
Yates correction).

Table 3. Frequency of detection of PSA and PSM
transcripts and number of PCR cycles.

 2 x 19 2 x 21 2 x 23 2 x 25
IM9 0/8 1/8 5/8 6/8
Jurkat 0/4 0/4 0/4 0/4
MCF7 0/4 0/4 0/4 1/4
CEM 0/3 3/3 3/3 3/3
Raji 0/8 0/8 2/8 3/8
M 29 0/2 0/2 0/2 0/2
M 45 0/2 0/2 0/2 0/2
M 30 0/2 0/2 0/2 0/2
M 29 0/2 0/2 0/2 0/2
IM9 6/6 6/6 6/6 6/6
Jurkat 0/6 0/6 3/6 2/6
MCF7 0/4 0/4 3/4 4/4
KG1 0/4 0/4 3/4 2/4
M 23 0/2 0/2 1/2 2/2
M 45 0/6 0/6 5/6 4/6
M 30 0/2 0/2 1/2 1/2
M 23 0/3 2/2 ND 2/2
M 24 0/3 2/3 ND 3/3
M 29 0/3 1/2 ND 2/2
F 37 0/3 3/3 ND 3/3

F, female; M, male; ND, not done.

Table 4. Limit of detection of PSA and PSM: comparison
 between two rounds of 19 and 25 cycles.

 2 x 19 cycles 2 x 25 cycles
LNCaP diluted in K562
PSA 1/[10.sup.4] 1/[10.sup.6]
PSM 1/[10.sup.4] 1/[10.sup.6]
LNCaP in whole blood (number of
LNCaP/mL-number of LNCaP/
PSA [10.sup.3]-192 10-1.9
PSM 10-1.9 Not assessable
Plasmids (number of copies
PSA 100 (faint: 10) 1
PSM 100 (faint: 10) 1

PBMC, peripheral blood mononuclear cell.
COPYRIGHT 1998 American Association for Clinical Chemistry, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1998 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Molecular Pathology and Genetics
Author:Gala, Jean-Luc; Heusterspreute, Michel; Loric, Sylvain; Hanon, France; Tombal, Bertrand; Van Cangh,
Publication:Clinical Chemistry
Date:Mar 1, 1998
Previous Article:Simultaneous analysis of plasma free fatty acids and their 3-hydroxy analogs in fatty acid [beta]-oxidation disorders.
Next Article:Molecular beacons: a new approach for semiautomated mutation analysis.

Related Articles
Simultaneous quantification of prostate-specific antigen and human glandular kallikrein 2 mRNA in blood samples from patients with prostate cancer...
Prostate-specific antigen expression in normal human bone marrow cells.
Prostate-specific antigen: a cancer fighter and a valuable messenger?
Numeric definition of the clinical performance of the nested reverse transcription-PCR for detection of hematogenous epithelial cells and correction...
Immunoreactivity, ultrastructural localization, and transcript expression of prostate-specific antigen in human neuroblastoma cell lines.
Improved reverse transcriptase--polymerase chain reaction protocol with exogenous internal competitive control for prostate-specific antigen mRNA in...
Biochemical characterization and immunolocalization of prostate-specific antigen in human term placenta.
Molecular detection of circulating prostate cells in cancer II: comparison of prostate epithelial cells isolation procedures. (Technical Briefs).
Quantitative reverse transcription-PCR assay with an internal standard for the detection of prostate-specific antigen mRNA.
Prostate-specific antigen in serum during the menstrual cycle.

Terms of use | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters