| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Molecular Pathology and Genetics |
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, Université Catholique de Louvain, 1200 Bruxelles, Belgique.
2
Queen Astrid Military Hospital, Rue Bruyn, 2, 1120
Bruxelles, Belgique.
3
Hôpital Necker, Laboratory of
Clinical Chemistry, 149, rue de Sèvres, 75015 Paris, France.
a Address correspondence to this author at: Laboratoire de Biologie Moléculaire Clinique, Clos Chapelle-aux-Champs, 30 - UCL/30.46, B-1200 Brussels, Belgium. Fax 00/32.2.764.39.59; e-mail gala{at}sang.ucl.ac.be.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
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 |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
nested rna-pcr or nested rt-pcr
(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 TrizolTM reagent (Life
Technologies, Gibco-BRL). They were processed immediately or frozen at
-80 °C. The procedure for total RNA isolation was carried out
according to the manufacturer's instructions; 2.5 µg of total RNA
were reverse transcribed with pdN6 (random hexamers: 1.6 µmol/L;
Expedite Nucleic Acid Synthesis System, Millipore)
(4)(9) and the SuperscriptTM RNase
H- 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 °C for 2 min)
was followed by repeated cycles (94 °C for 50 s, 61 °C for
50 s, and 72 °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 µL of PCR mix included the Dynazyme or Taq DNA polymerase as mentioned above, 1.5 mmol/L MgCl2, 200 µmol/L deoxynucleotide triphosphates, and 200 nmol/L of each primer. In the second PCR round, 2.5 µL 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.
psa and psm primer sequences
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 bp and a nested product of 193 bp. PSM primers were identical to previously described primers (13), except for their 5' ends, which were shortened.
controls at the mrna and cdna level
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 RT-PCR. 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, MgCl2, 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 °C for 2 min) was followed by 32 PCR cycles (94 °C for 50 s, 60 °C for 50 s, 72 °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).
detection of illegitimate transcripts according to the
number of rt-pcr cycles
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.
|
limit of detection
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 LNCaP in 10 mL of RPMI 1640. Each suspension was assessed in triplicate on a Kova® Glasstic® 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/L white blood cells (Technicon H3 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 10/L) with 2.10 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, Malmö, Sweden). PSM cDNA (1.2 kb) was amplified with the following primers, and cloned into BamHI and SalIsites 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 101 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 PCRs to determine the CV. The limit of dilution was defined as the last dilution giving a positive PCR result with a CV <20%.
sequencing of psa and psm pcr products or hybridization
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).
cell lysates
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 °C and thawed at 37 °C
three times. Debris were discarded after centrifugation, and the
supernatant was frozen at -80 °C until assayed.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A, 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 PCRs, we found
8 of 127 (6.3%) positive results, only 3% (2 of 66) in men and 9.8%
(6 of 61) in women.
psm rt-pcr results
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 bood donors' cDNA. As with PSA, positive signals obtained were comparable (data not shown).
frequency of detection of psa and psm transcripts according to
number of cycles
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.
detection limit of pcr
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. 2
A 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.
|
|
controls
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 results
Sequencing of the 355-bp PSA PCR products obtained with various
cell lines (LNCaP, MCF-7, IM9, 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'-ACACAGGCCAGGTATTTCAG-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 bp (from
nucleotide 396 to 519) in MCF-7 and 71 bp (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).
protein assay results
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).
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
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 10 PSA-negative cells (7) and about 2 LNCaP in 10 nucleated blood cells (8)(9)(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 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 K562 cells, about 2 LNCaP were detected in 10 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.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  P. Mhawech-Fauceglia, D. J. Smiraglia, W. Bshara, C. Andrews, J. Schwaller, S. South, D. Higgs, S. Lele, F. Herrmann, and K. Odunsi Prostate-Specific Membrane Antigen Expression Is a Potential Prognostic Marker in Endometrial Adenocarcinoma Cancer Epidemiol. Biomarkers Prev., March 1, 2008; 17(3): 571 - 577. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Patel, P. J. Whelan, S. Prescott, S. C. Brownhill, C. F. Johnston, P. J. Selby, and S. A. Burchill The Use of Real-Time Reverse Transcription-PCR for Prostate-Specific Antigen mRNA to Discriminate between Blood Samples from Healthy Volunteers and from Patients with Metastatic Prostate Cancer Clin. Cancer Res., November 15, 2004; 10(22): 7511 - 7519. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. H.N. Wong The Sporadic Nature of Shedding Cells in Blood: Multiple RNA Diagnostic Testing and Prognostication of Cancer Progression Clin. Chem., September 1, 2003; 49(9): 1429 - 1431. [Full Text] [PDF]
|
|
|
|
|
|
D. Schamhart, J. Swinnen, K.-H. Kurth, A. Westerhof, R. Kusters, H. Borchers, and C. Sternberg Numeric Definition of the Clinical Performance of the Nested Reverse Transcription-PCR for Detection of Hematogenous Epithelial Cells and Correction for Specific mRNA of Non-Target Cell Origin as Evaluated for Prostate Cancer Cells Clin. Chem., September 1, 2003; 49(9): 1458 - 1466. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Ylikoski, K. Pettersson, J. Nurmi, K. Irjala, M. Karp, H. Lilja, T. Lovgren, and M. Nurmi Simultaneous Quantification of Prostate-specific Antigen and Human Glandular Kallikrein 2 mRNA in Blood Samples from Patients with Prostate Cancer and Benign Disease Clin. Chem., August 1, 2002; 48(8): 1265 - 1271. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. A. Fava, R. Desnoyers, S. Schulz, J. Park, D. Weinberg, E. Mitchell, and S. A. Waldman Ectopic Expression of Guanylyl Cyclase C in CD34+ Progenitor Cells in Peripheral Blood J. Clin. Oncol., October 1, 2001; 19(19): 3951 - 3959. [Abstract] [Full Text]
|
|
|
|
 |
|
 A. Ylikoski, M. Karp, K. Pettersson, H. Lilja, and T. Lovgren Simultaneous Quantification of Human Glandular Kallikrein 2 and Prostate-Specific Antigen mRNAs in Peripheral Blood from Prostate Cancer Patients J. Mol. Diagn., August 1, 2001; 3(3): 111 - 122. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. de Cremoux, J. M. Extra, M. G. Denis, J. Y. Pierga, E. Bourstyn, C. Nos, K. B. Clough, E. Boudou, E. C. Martin, A. Müller, et al. Detection of MUC1-expressing Mammary Carcinoma Cells in the Peripheral Blood of Breast Cancer Patients by Real-Time Polymerase Chain Reaction Clin. Cancer Res., August 1, 2000; 6(8): 3117 - 3122. [Abstract] [Full Text]
|
|
|
|
 |
|
 T. Notomi, H. Okayama, H. Masubuchi, T. Yonekawa, K. Watanabe, N. Amino, and T. Hase Loop-mediated isothermal amplification of DNA Nucleic Acids Res., June 15, 2000; 28(12): e63 - e63. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Ylikoski, M. Sjoroos, A. Lundwall, M. Karp, T. Lovgren, H. Lilja, and A. Iitia Quantitative Reverse Transcription-PCR Assay with an Internal Standard for the Detection of Prostate-specific Antigen mRNA Clin. Chem., September 1, 1999; 45(9): 1397 - 1407. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Mannello, S. Barulli, M. Malatesta, S. Mancini, P. Leoni, and G. Gazzanelli Prostate-specific Antigen Expression in Normal Human Bone Marrow Cells Clin. Chem., July 1, 1999; 45(7): 1102 - 1103. [Full Text] [PDF]
|
|
|
|
|
|
F. Mannello, M. Malatesta, F. Luchetti, S. Papa, S. Battistelli, and G. Gazzanelli Immunoreactivity, Ultrastructural Localization, and Transcript Expression of Prostate-specific Antigen in Human Neuroblastoma Cell Lines Clin. Chem., January 1, 1999; 45(1): 78 - 84. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Ogura, Y. Agata, K. Watanabe, R. M. McCormick, Y. Hamaguchi, Y. Aso, and M. Mitsuhashi RNA chip: quality assessment of RNA by microchannel linear gel electrophoresis in injection-molded plastic chips Clin. Chem., November 1, 1998; 44(11): 2249 - 2255. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Urata, M. Wakiyama, M. Iwase, M. Yoneda, S. Kinoshita, N. Hamasaki, and D. Kang New sensitive method for the detection of the A3243G mutation of human mitochondrial deoxyribonucleic acid in diabetes mellitus patients by ligation-mediated polymerase chain reaction Clin. Chem., October 1, 1998; 44(10): 2088 - 2093. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Mannello, F. Luchetti, D. Lancioli, S. Battistelli, S. Papa, and G. Gazzanelli Prostate-specific Antigen Expression in Neoplastic Human Myeloid Cell Lines Clin. Chem., September 1, 1998; 44(9): 1991 - 1993. [Full Text] [PDF]
|
|
| ||||||||||
