Clinical Chemistry 43: 2244-2250, 1997;
(Clinical Chemistry. 1997;43:2244-2250.)
© 1997 American Association for Clinical Chemistry, Inc.
Sensitive and specific cytokeratin 18 reverse transcription-polymerase chain reaction that excludes amplification of processed pseudogenes from contaminating genomic DNA
Peter Tschentscher,
Christoph Wagenera and
Michael Neumaier
Department of Clinical Chemistry, Medical Clinic, University Hospital Eppendorf, Martinistr. 52, D-20251 Hamburg, Germany.
a Author for correspondence. Fax 0049-40-4717-4621; e-mail wagener{at}uke
 |
Abstract
|
|---|
Processed pseudogenes of residual contaminating genomic DNA
interfere with a sensitive detection of cytokeratin 18 (CK18) mRNA by
reverse transcription and polymerase chain reaction (RT-PCR). This may
cause false-positive results when CK18 mRNA is used as a marker for
ectopic tumor cells in specimens from cancer patients. To establish a
sensitive CK18 RT-PCR by excluding the amplification of processed
pseudogenes, the following strategy was chosen: (a) CK18
pseudogene sequences were cloned from genomic DNA by PCR;
(b) cDNA-specific primers were designed on the basis of
mismatches between pseudogenes and cDNA; (c) PCR conditions
were adjusted to reach maximum sensitivity and specificity. Epithelial
cells (1–10) could be detected in 1 mL of blood. Among the numerous
CK18 genes homologous to the transcribed gene, at least two different
processed pseudogenes exist that are highly homologous to each other
and to the exons of the transcribed CK18 gene.
|
Introduction
|
|---|
The sensitive detection of tissue-specific mRNA is used
increasingly for the detection of micrometastatic tumor cells in
peripheral blood, bone marrow, lymph nodes, and other tissues. For some
of the target genes, processed pseudogenes are present in the genome.
Because the amplification products of cDNA and processed pseudogenes
are of the same size, residual contaminating genomic DNA in the RNA
preparation may cause false-positive results. Cytokeratin 18
(CK18) is an intermediate-sized keratin-like filament
characteristic for epithelial cells. Expression of the human CK18
polypeptide was detected at the protein level in a large number of
tumors, cultured carcinoma cell lines, and normal epithelia
(1). To make use of the widespread occurrence of CK18 in
carcinoma cells, antibodies against this filament have been applied to
the detection of ectopic tumor cells in bone marrow and other specimens
(2)(3). The detection was sufficient in
specificity but limited in sensitivity. To lower the detection limit,
different groups tried to transfer this diagnostic approach to the
nucleotide level detecting CK18 mRNA by polymerase chain reaction after
a reverse transcription into cDNA (RT-PCR). They failed for the simple
reason that their negative controls, samples without carcinoma cells,
turned out to be positive (4)(5). In contrast
to the findings of the protein studies, this outcome was interpreted as
a nonepithelial expression of CK18, e.g., in hematopoetic cells. In
this study, we show that contaminating genomic DNA with processed
pseudogene sequences of CK18 interferes with the RNA detection by
RT-PCR. To solve this problem, an RNA-specific nested RT-PCR assay was
established with the use of sequence information obtained from cloned
pseudogene PCR products amplified from genomic DNA.
|
Materials and Methods
|
|---|
pcr
All PCRs were performed in 50 mmol/L KCl, 10 mmol/L Tris-HCl (pH
8.3), 1.5 mmol/L MgCl2, 0.1 g/L gelatin, 0.25 mmol/L
of each dNTP (Pharmacia), 0.4 or 0.8 µmol/L of each primer, and 5
units of Taq polymerase (Life Technologies) in a 100-µL
reaction with the use of a Biozym MiniCycler (Hameln) and PCR tubes
from Sarstedt. Primers were synthesized with the use of phosphoramidite
chemistry on an oligonucleotide synthesizer Gene Assembler Plus
(Pharmacia).
cloning of pseudogene sequences
Genomic DNA was prepared from peripheral blood leukocytes of a
healthy human volunteer by following established procedures
(6) and was treated with 30 ng/L RNase (Life Technologies)
for 1 h at 37 °C. For amplification of genomic pseudogene
sequences, primers were synthesized according to the published sequence
of the CK18 cDNA (7): L,
5'-ATGAGATTCACCACTCGCTCCACCT-3'; R,
5'-ATGCCTCAGAACTTTGGTGTCATTGG-3'.
RNase-treated DNA (1 µg) served as a template in a PCR with 0.4
µmol/L of primers R and L for 32 cycles at: 97 °C (1 min); an
annealing temperature of 62 °C (2 min), which is optimal according
to the Oligo 4.01 program (Primer Analysis Software, National
Biosciences, Plymouth, MN); and an extension step at 72 °C (2 min),
followed by a final extension at 72 °C (10 min). The PCR products of
five reactions in separate tubes were cloned into the multicloning site
of the pCRII vector (Invitrogen). Plasmid preparations of transformed
One-Shot-cells (Invitrogen) were checked by restriction fragment
analysis to identify clones containing a complete 1.3-kb insert for
each of the five initial reactions. The inserts were sequenced with the
use of the Taq DyeDeoxy Terminator cycle sequencing system
of Applied Biosystems.
rna extraction and cdna synthesis
RNA was extracted by the acid guanidinium
thiocyanate–phenol–chloroform method (8). To reduce
contamination of genomic DNA, an additional step of DNase digestion
(Boehringer Mannheim; 40 units in a 400-µL reaction with 10 units of
RNase inhibitor in 5 mmol/L MgSO4 at 25 °C) was
introduced before the RNA was precipitated. The enzyme was inactivated
by incubating the RNA preparation for 5 min at 90 °C. For
measurement of sensitivity of the nested PCR assay, cell culture
monolayers of the human colon carcinoma cell line HT29 were
trypsinated, pelleted, washed, and resuspended in phosphate-buffered
saline, counted in the basophil/lobularity channel of a cell counter
(H1, Technicon), and serially diluted (105 to
100 cells/mL) in peripheral blood from a healthy human
volunteer with 6 x 109/L leukocytes and 153 g/L
hemoglobin. After the blood-collection needle was introduced into the
volunteer's vein, the first milliliters of blood were discarded, and
the sample was collected in a fresh tube to minimize the risk of
contamination with epithelial cells from skin. The dried RNA of 0.5 mL
of blood or HT29 cell dilution in blood was resuspended in 80 µL of
10 mmol/L Tris-HCl and 1 mmol/L EDTA (pH 8.0). Random hexamer
(Pharmacia)-primed cDNA synthesis was performed in a 20-µL reaction
with 10 µL of the RNA preparation and 1 µL of SUPERSCRIPT II RNase
H- Reverse Transcriptase (Life Technologies) with the use
of buffer and conditions recommended by the manufacturer.
nested primers and pcr protocol
With the use of substitutions found by comparison of the highly
homologous sequences of the PCR products and the CK18 cDNA, primers
were constructed that should preferentially amplify the cDNA: X,
5'-TGCTCACCACACAGTCTGAT-3'; Y, 5'-CACTTTGCCATCCACTAGCC-3'; X',
5'-TGGAGGACCGCTACGCCCTA-3'; Y', 5'-CCAAGGCATCACCAAGACTA-3'.
Nested PCR was performed with 10 µL of the cDNA with 0.8 µmol/L of
primers X and Y in the first PCR of 40 cycles at 95 °C (50 s) and
54 °C (30 s) and 7 µL of the first PCR product with 0.8 µmol/L
of primers X' and Y' in the second PCR of different numbers of cycles
with the same steps of temperature. Each PCR was started with an
initial denaturation of 1 min 20 s at 95 °C and terminated with
a final extension of 1 min at 72 °C. Following established
procedures (6) aliquots of the PCR products were
electrophoresed in TAE buffer on a 1.3% agarose gel containing
ethidium bromide and photographed under ultraviolet light to estimate
size and amount.
patients and bone marrow collection
Bone marrow samples were collected from five patients undergoing
surgery for malignant disease (cancer of esophagus, cardia, and lung)
and from one patient with a benign disease (chronic pancreatitis). An
incision was made in the skin with a scalpel before the needle was
introduced, to minimize the risk of contamination with epithelial
cells. Quality of RNA preparation was checked by PCR amplification of
ß2-microglobulin. Bone marrow samples were obtained
by informed consent. The procedures were in accordance with the
Helsinki Declaration of 1975, as revised in 1983.
|
Results
|
|---|
pseudogene interference with rna detection
The analytical problem of processed pseudogene interference with a
sensitive detection of CK18 RNA is illustrated in Fig. 1
. Conventional primers were designed according to the cDNA
sequence without consideration of pseudogene sequences. RNA was
prepared from peripheral blood from a healthy human volunteer and
transcribed into cDNA. In a control experiment with the same RNA
preparation the reverse transcriptase was omitted in the cDNA synthesis
step. A nested PCR was performed on both samples at optimal annealing
temperatures with 40 cycles in the first and 30 cycles in the second
PCR step. The transcribed (lane T) as well as the untranscribed sample
(lane T-) generated a PCR fragment corresponding to
the size of 392 bp predicted from the cDNA sequence. The result
suggests the following. (a) Residual genomic DNA in the RNA
preparation is sufficient to generate a cDNA-independent amplification
product. (b) Although the primers span introns, the
amplification product is the same size as a cDNA-derived product.
Consequently, detection of CK18 RNA cannot be distinguished from
amplification of genomic DNA with a conventional nested PCR protocol.
View larger version (91K):
[in this window]
[in a new window]
|
Figure 1. CK18-nested PCR with conventional primers designed
according to the cDNA sequence without consideration of pseudogene
sequences (40 cycles in the first step, 30 cycles in the second step,
expected size of cDNA product: 392 bp).
Ethidium bromide-stained agarose gel. Lane T, peripheral
blood of a healthy human volunteer; lane T-,
same preparation, reverse transcriptase omitted.
|
|
strategy for cloning processed pseudogenes
The strategy to clone processed pseudogenes by PCR is depicted in
Fig. 2
: RNA was isolated from a CK18-expressing cell line. DNA was
prepared from peripheral blood leukocytes from a healthy human
volunteer. After RT of RNA into cDNA, a PCR was performed at an optimal
annealing temperature for the primers L and R, which span the whole
cDNA as indicated in Fig. 2
. The PCR fragment corresponded to the size
of 1.3 kb predicted by the cDNA sequence (Fig. 2
, lane C). When
identical PCR conditions were applied to RNase-treated genomic DNA
instead of cDNA, a fragment of identical size was obtained, indicating
the amplification of processed pseudogene(s) (lane G). Such products of
five independent PCRs were cloned and sequenced. Each of the cloned
fragments contained base substitutions when compared with the cDNA.
Some substitutions were found in only one clone but not in the
remaining clones. These substitutions most probably reflect
misincorporations due to Taq polymerase errors. However, a
number of identical base substitutions were found at identical
positions in more than one clone. Because it is highly improbable that
these identical substitutions reflect Taq polymerase errors,
they indicate specific differences between the sequences of cDNA and
processed pseudogenes. The sequences were aligned and screened for
identical base substitutions at identical positions. As shown in Fig. 3
, three clones (1, 2, and 5) contained one segment corresponding
to clone 3 and a second segment corresponding to clone 4. The finding
that different segments of two main sequences are present in a given
clone suggests that the five PCR clones may not represent five
different pseudogenes. During PCR, an incompletely synthesized DNA
strand of one pseudogene may hybridize with another pseudogene in the
annealing step of the next cycle. Because each fragment corresponds to
either one of the two different clones 3 or 4, the five PCR clones may
originate from only two different processed pseudogenes.
View larger version (26K):
[in this window]
[in a new window]
|
Figure 2. Strategy for cloning processed pseudogenes.
PCR amplification with primers L and R (annealing temperature
62 °C). Primers are positioned at the 5' and 3' ends of the CK18
cDNA sequence to amplify a full-length cDNA product of 1.3 kb.
Lane C, HT29 cDNA; lane N, negative control (no
DNA); lane G, RNase-treated genomic DNA.
|
|
View larger version (26K):
[in this window]
[in a new window]
|
Figure 3. Schematic presentation of CK18 cDNA, intron/exon
boundaries, and pseudogene sequences.
Arrows indicate the position and size of the introns in the
genomic DNA. Light and dark bars of the
pseudogenes reflect two sequence motifs with identical base
substitutions as compared with the cDNA. Positions of primers X, Y, X',
and Y' relative to the cDNA and CK18 gene are indicated.
|
|
design of ck18 cdna-specific primers
The aligned sequences were screened for regions in which all
pseudogene sequences deviated from the cDNA sequence. These conditions
were fulfilled for primers designated X, Y, X', and Y'. Additional
substitutions were introduced to decrease the stability of primer
annealing with the pseudogene sequences. In Fig. 4
, the primers X and Y of the first PCR step and the
corresponding regions of cDNA and pseudogenes are shown. Primer
positions relative to the cDNA and the gene are depicted in Fig. 3
. The
primers span two introns. From the published sequences, a fragment size
of 1054 bp is predicted for the genomic fragment and a size of 413 bp
for the cDNA fragment. Because the size of processed pseudogenes
corresponds to the size of the cDNA, their amplification would also
produce a 413-bp fragment. To prove that the amplification of processed
pseudogenes can be prevented with cDNA-specific primers at appropriate
annealing temperatures, the RNase-treated genomic DNA of the initial
experiment (Fig. 2
) was amplified with the use of primers X and Y. As
shown in the left part of Fig. 5
, fragments of 413 bp are amplified at 63 and 65 °C. The
fragment size corresponds to the size predicted from the cDNA. Because
the amplification started from genomic DNA, the fragments are derived
from processed pseudogenes. With increasing annealing temperature, the
only fragment amplified exhibits a size of 1054 bp corresponding to the
fragment derived from the CK18 gene. Because the exon sequences of the
gene and the cDNA are identical, the experiment indicates that the
primers should be suited for a specific amplification of CK18 cDNA even
in the presence of processed pseudogenes. To verify this assumption,
CK18 cDNA from HT29 cells was amplified at an annealing temperature at
which an amplification of processed pseudogenes is excluded (Fig. 5
, lanes P1 and P2). Isolated RNA from HT29
cells as well as the RNA prepared from peripheral blood supplemented
with these cells produced a fragment of the expected size of 413 bp.
Contaminating genomic DNA in the blood sample (lane P2)
generated a second fragment of 1054 bp. Because the amplification of
processed pseudogenes is excluded at this high annealing temperature,
the smaller fragment originates from cDNA and the larger fragment from
genomic DNA.
View larger version (29K):
[in this window]
[in a new window]
|
Figure 4. Sequence comparison of the pseudogene PCR clones with the
CK18 cDNA and construction of primers X and Y.
Nucleotide substitutions are boxed.
|
|
View larger version (39K):
[in this window]
[in a new window]
|
Figure 5. Amplification products of a PCR with primers X and Y
(expected size of cDNA product: 450 bp).
Ethidium bromide-stained agarose gel. Annealing temperatures are
indicated below the lanes. Lanes 1–8, 1 µg of
RNase-treated genomic DNA; lane N, negative control (no
DNA); lane P1, cDNA of HT29 cells; lane
P2, cDNA of carcinoma cells in peripheral blood.
|
|
establishment and optimization of a nested pcr protocol
The position of primers X, Y, X', and Y' are depicted in Fig. 3
.
The primers X' and Y' of the second PCR step produce a cDNA fragment of
210 bp and a fragment of 589 bp when the gene is amplified, because the
primers span an intron of 379 bp. Taking into account the need of
stringent annealing conditions effected by high annealing temperatures,
the temperature/time profile of the nested PCR was optimized
empirically by detecting RNA of human carcinoma cells HT29, diluted in
blood of a healthy human volunteer. A two-step PCR with 50 s
denaturation at 95 °C and 30 s annealing at 54 °C without
any extra time for extension was most effective. Because of the short
annealing time, these conditions favor the generation of short
fragments from cDNA vs larger fragments from genomic DNA. To cool the
effective temperature in the PCR tube to the desired 70–72 °C
during the short annealing step, it was necessary to adjust a lower
temperature of 54 °C in the thermal cycler. Omitting reverse
transcriptase and RNase treatment of the samples reconfirmed that the
produced PCR fragments were strictly RNA/cDNA-dependent and did not
originate from processed pseudogenes (not shown).
Applying a constant 40 cycles of this profile in the first PCR meant
that the sensitivity of the nested PCR assay was determined by the
number of cycles during the second PCR (Fig. 6
). As shown on Fig. 6
, top, 102 (lane 2) to
103 (lane 3) epithelial cells per milliliter of peripheral
blood could be detected with 25 cycles in the second PCR. With 30
cycles in the second PCR (Fig. 6
, middle), 100 (lane 0) to
101 (lane 1) tumor cells were detected. The peripheral
blood sample without tumor cells remained negative for the
cDNA-dependent 210-bp fragment (lane B). The appearance of the genomic
fragment of 589 bp reflects variable amounts of contaminating genomic
DNA in the RNA preparations. Because of a competition between cDNA and
genomic DNA, the genomic product, if present in the RNA preparation,
becomes apparent only at low amounts of cDNA in the sample (lane 0).
However, the genomic amplification product does not disturb a specific
detection of the cDNA because the smaller fragment is preferably
amplified. When the number of cycles in the second PCR is further
increased, a 210-bp product becomes visible even in normal peripheral
blood without tumor cells (Fig. 6
, bottom, lane B). Corresponding
samples with RNase treatment (lane BR) or without reverse
transcriptase (lane B-) remain negative for this fragment
and show the larger fragment of 589 bp, which in this case is not
inhibited by competitive amplification of the 210-bp cDNA product.
Thus, in contrast to the amplification product of the PCR with
conventional primers (Fig. 1
), the 210-bp fragment of lane B is
strictly dependent on the presence of cDNA, indicating a low
concentration of CK18 RNA in the samples (low level transcription).
View larger version (81K):
[in this window]
[in a new window]
|
Figure 6. CK18-nested PCR products (expected size of cDNA product:
210 bp).
Ethidium bromide-stained agarose gels. Lanes 0–5,
100 to 105 HT29 carcinoma cells per milliliter
of peripheral blood; lane N, negative control
(H2O instead of blood in guanidinium thiocyanate solution);
lane H, negative control (H2O instead of RNA in
the cDNA synthesis); lane B, peripheral blood without
carcinoma cells; lane B-, preparation B,
reverse transcriptase omitted; lane BR,
preparation B, cDNA synthesis after RNase treatment. Top, 25
cycles in second PCR; middle, 30 cycles in second PCR;
bottom, 40 cycles in second PCR.
|
|
examination of bone marrow aspirates
To prove the diagnostic utility of the CK18 RT-PCR protocol
developed in this study, bone marrow aspirates from patients with
esophageal, gastric, and lung cancer were analyzed. As negative
control, the bone marrow aspirate of a patient with pancreatitis was
used (Fig. 7
). Seven of 11 samples from patients with malignant disease
scored positive, indicating the presence of CK18 mRNA-expressing cells
in their bone marrow aspirates. Six results were clearly positive
(lanes 5, 6, 7, 9, 11, and 12); in one case only a faint band of
expected size could be detected (lane 2). The bone marrow sample from
the patient with chronic pancreatitis was negative (lane 1).
View larger version (25K):
[in this window]
[in a new window]
|
Figure 7. Examination of bone marrow samples by CK18-nested PCR.
Ethidium bromide-stained agarose gel. Lane H, negative
control (H2O instead of RNA in the cDNA synthesis);
lane P, positive control (RNA of carcinoma cells in
peripheral blood); lane 1, bone marrow sample from a patient
with benign disease (chronic pancreatitis); lanes 2–12,
bone marrow samples from cancer patients.
|
|
|
Discussion
|
|---|
Previous studies have shown that 15–20 CK18 homologous genes
exist in the human genome, including the single active gene and a large
number of pseudogenes (9). Beside those pseudogenes, which
retain the exon-intron arrangement of their productive counterpart, a
second category has been defined as processed pseudogenes
(10) —intronless, nontranscribed genes that are highly
homologous to the exons of the transcribed genes. In RT-PCR assays,
mRNA is transcribed into cDNA, which is amplified by PCR. Both cDNA and
processed pseudogenes lack introns. Furthermore, the DNAs are highly
homologous. For this reason, processed pseudogenes may generate PCR
fragments of the same size as fragments originating from the cDNA.
Thus, processed pseudogenes may present an important analytical problem
in RT-PCR assays because residual genomic DNA in the RNA preparation
will generate false-positive results.
Cytokeratins have been used widely as protein targets for the detection
of micrometastatic cells by immunocytochemical methods
(2)(3)(11). Among the different
cytokeratins, CK18 has been the preferred target. In principle, the
CK18 mRNA should also be a suitable target for highly sensitive
detection of tumor cells by RT-PCR. However, initial attempts to set up
a CK18 RT-PCR failed because positive signals were generated in normal
bone marrow and peripheral blood also in the absence of epithelial
cells (4)(5). For the following reasons, we
assumed that these false-positive results are caused by processed
pseudogenes. In Southern transfers, fragments of identical size were
generated both from genomic and cDNA (12). As shown
here, a PCR product of the expected cDNA size was readily obtained from
genomic DNA without RT and after treatment with RNase (Fig. 2
, lane G).
The present study was undertaken to establish a CK18 RT-PCR assay for
the detection of micrometastatic carcinoma cells. Assuming the
existence of processed pseudogenes, primers had to be chosen that would
hybridize to the cDNA only. To achieve this goal, it was necessary to
provide sequence information on the putative pseudogenes. In principle,
this could be done by genomic cloning. To avoid the time-consuming
cloning and screening procedures, we took a more direct approach by
cloning the genomic PCR products that would disturb a cDNA assay.
Experimental procedures were such that only processed pseudogenes were
cloned. The sequences of five independent clones were highly
homologous, but differed by single base substitutions. Single
nucleotide substitutions in one sequence may be caused by
misincorporation caused by Taq polymerase errors. However,
identical substitutions at the same position in more than one clone
most probably indicate a real difference between the amplified genomic
sequences. To estimate the number of processed pseudogenes correctly,
another possible PCR artifact has to be considered. Incompletely
synthesized DNA strands of one pseudogene could hybridize with another
pseudogene in the annealing step of the next cycle. For this reason the
PCR clones may be composed of parts of different genomic sequences and
thus present a false number of processed pseudogenes. However, by
focusing on different segments of an alignment, each cloned sequence
could be assigned to one of two sequence motifs (Fig. 3
). Therefore,
the existence of at least two different processed CK18 pseudogenes in
the human genome seems reasonable.
On the basis of the substitutions in the pseudogenes, primers
were designed that should allow the specific amplification of the CK18
cDNA. To decrease the stability of annealing with the pseudogenes, a
few additional mismatches were introduced. At the appropriate annealing
temperature, a good amplification of intron-including CK18 gene
sequences was obtained from genomic DNA without any amplification of
processed pseudogene sequences (Fig. 5
, lanes 5–7). Because the CK18
cDNA is identical with the corresponding exons of the active gene, this
is a good basis for optimizing a cDNA-specific RT-PCR.
The successful design of cDNA-specific primers on the basis of
mismatches to the pseudogene sequences proves the feasibility of the
cloning procedure taken in this study. To our knowledge, this is the
first report in which processed pseudogenes were cloned by PCR. This
direct approach should facilitate the design of cDNA-specific primers
for other RT-PCRs in which processed pseudogenes may cause false
results.
Time and temperature of the annealing and extension steps of the nested
PCR was optimized for the best yield of the desired 210-bp product. A
two-step PCR without an extension step and with a short annealing
period was most effective because this preferentially amplified small
fragments, avoiding competitive synthesis of the intron-including
products derived from the active CK18 gene and putative nonprocessed
pseudogenes of residual genomic DNA. As a control for RNA specificity,
reverse transcriptase was omitted. To reach the required effective
annealing temperature within an annealing time of only 30 s, it
was necessary to lower the temperature in the thermal cycler to
54 °C.
At a constant number of 40 cycles in the first PCR, the number of
cycles in the second (nested) PCR was adopted to reach maximum
sensitivity. In reconstitution experiments, 100 carcinoma cells per
milliliter of normal peripheral blood (that is, ~17
cells/106 leukocytes) could easily be detected by
performing ~25 cycles in the second PCR. This is comparable with the
described detection limit of an RT-PCR assay established for
cytokeratin 19 (13). Increasing the number of cycles to 30
made possible the detection of 10 to 1 cells/mL of blood (1.7 to 0.17
calculated cells per 106 leukocytes), although in this
range specificity became more precarious. With a further increase of
cycles, a background RNA signal was detectable also in the sample
without carcinoma cells (Fig. 6
, bottom, lane B). RNA specificity of
the assay was reconfirmed by negative results in the corresponding
samples without reverse transcriptase (lane B-) and after
RNase treatment (lane BR). These mRNA-based signals may be
traced back to a background transcription in normal blood cells.
However, although the first milliliter of blood had been discarded and
a new collection tube was used when blood was drawn from the cubital
vein, contamination of the sample by CK18-positive epithelial cells
from skin cannot be formally excluded.
In conclusion, we describe the cloning of CK18-processed pseudogene
sequences from human genomic DNA. On the basis of differences in the
sequence of pseudogenes and cDNA, a highly sensitive and specific CK18
RT-PCR was developed. In reconstitution experiments, between 1 and 10
epithelial cells could be detected in 1 mL of peripheral blood. In bone
marrow aspirates from patients with different carcinomas, positive
results were obtained in 7 of 11 samples, and the bone marrow of a
patient with nonmalignant disease scored negative. In an ongoing study
of bone marrow aspirates from tumor patients, the results obtained by
the CK18 RT-PCR compare well with the results obtained by the
carcinoembryonic antigen RT-PCR described previously (14).
|
Acknowledgments
|
|---|
We thank Cornelia Lübcke for her technical assistance in RNA
preparation and Jakob Izbicki (Surgical Clinic, University Hospital
Eppendorf, Hamburg, Germany) and his coworkers for provision of bone
marrow samples from cancer patients.
|
References
|
|---|
-
Moll R, Franke WW, Schiller DL, Geiger B, Krepler R. The catalog of human cytokeratins: patterns of expression in normal epithelia, tumors and cultured cells. Cell 1982;31:11-24.
[Web of Science][Medline]
[Order article via Infotrieve]
-
Schlimok G, Funke I, Bock B, Schweiberer B, Witte J, Riethmüller G. Epithelial tumor cells in bone marrow of patients with colorectal cancer: immunocytochemical detection, phenotypic characterization, and prognostic significance. J Clin Oncol 1990;8:831-837.
[Abstract]
-
Lindemann F, Schlimok G, Dirschedl P, Witte J, Riethmüller G. Prognostic significance of micrometastatic tumor cells in bone marrow of colorectal cancer patients. Lancet 1992;340:685-689.
[Web of Science][Medline]
[Order article via Infotrieve]
-
Traweek ST, Liu J, Battifora H. Keratin gene expression in non-epithelial tissues. Detection with polymerase chain reaction. Am J Pathol 1993;142:1111-1118.
[Abstract]
-
Brown DC, Purushotham AD, Birnie GD, George WD. Detection of intraoperative tumor cell dissemination in patients with breast cancer by use of reverse transcription and polymerase chain reaction. Surgery 1995;117:96-101.
-
Sambrook J, Fritsch EF, Maniatis T. Molecular cloning: a
laboratory manual. Cold Spring Harbor, NY: Cold Spring Harbor
Laboratory Press, 1989;1553pp..
-
Kulesh DA, Oshima RG. Complete structure of the gene for human keratin 18. Genomics 1989;4:339-347.
[Web of Science][Medline]
[Order article via Infotrieve]
-
Chomczynski P, Sacchi N. Single-step method of RNA isolation by acid guanidinium thiocyanate–phenol–chloroform extraction. Anal Biochem 1987;162:156-159.
[Web of Science][Medline]
[Order article via Infotrieve]
-
Kulesh DA, Oshima RG. Cloning of the human keratin 18 gene and its expression in nonepithelial mouse cells. Mol Cell Biol 1988;8:1540-1550.
[Abstract/Free Full Text]
-
Vanin EF. Processed pseudogenes. Characteristics and evolution. Biochim Biophys Acta 1984;782:231-241.
[Medline]
[Order article via Infotrieve]
-
Pantel K, Izbicki J, Passlick B, Angstwurm M, Häussinger K, Thetter O, Riethmüller G. Frequency and prognostic significance of isolated tumor cells in bone marrow of patients with non-small-cell lung cancer without overt metastases. Lancet 1996;347:649-653.
[Web of Science][Medline]
[Order article via Infotrieve]
-
Neumaier M, Gerhard M, Wagener C. Diagnosis of micrometastases by the amplification of tissue-specific genes. Gene 1995;159:43-47.
[Web of Science][Medline]
[Order article via Infotrieve]
-
Datta YH, Adams PT, Drobyski WR, Ethier SP, Terry VH, Roth MS. Sensitive detection of occult breast cancer by the reverse-transcriptase polymerase chain reaction. J Clin Oncol 1994;12:475-482.
[Abstract]
-
Gerhard M, Juhl H, Kalthoff H, Schreiber HW, Wagener C, Neumaier M. Specific detection of carcinoembryonic antigen-expressing tumor cells in bone marrow aspirates by polymerase chain reaction. J Clin Oncol 1994;12:725-729.
[Abstract]
Top
Abstract
Introduction
