Division of Hematology/Oncology, Department of Medicine,
University of Florida, Gainesville, Florida
Previous studies in this laboratory showed that the overexpression of
human aldehyde dehydrogenase class-1 (ALDH-1) with a retroviral vector
resulted in increased resistance to 4-hydroperoxycyclophosphamide (4-HC), an active metabolite of cyclophosphamide. The present study
examined the effect of ALDH-1 antisense RNA expression on ALDH-1
activity and sensitivity to 4-HC toxicity. Three different ALDH-1 cDNAs
were synthesized that are either missing the N terminus (N), C terminus
(C), or both (NC) and subcloned into the BamHI cloning
site of pLXSN retroviral vector in the antisense (AS) orientation
(AS-N, AS-C, and AS-NC, respectively). It was demonstrated that the
overexpression of each of the AS constructs in K562 leukemic cells and
A549 lung cancer cells results in suppression of ALDH-1 mRNA and
enzymatic activity. Furthermore, the AS-N and AS-NC were generally more
effective than AS-C in reducing the ALDH-1 activity. Both K562 and A549
cells expressing the ALDH-1 AS became significantly more sensitive to
4-HC toxicity as demonstrated by clonogenic and liquid culture assays.
The increase in 4-HC sensitivity was in correlation with the degree of
suppression of ALDH-1 activity. Moreover, such increase in 4-HC
sensitivity, especially with AS-N and AS-NC, was to a similar degree
seen with the use of diethylaminobenzaldehyde, a specific inhibitor of
ALDH-1. These results indicate that ALDH-1 expression and activity can
be specifically and effectively suppressed by AS RNA and lead to
increased sensitivity to 4-HC.
 |
Introduction |
Aldehyde
dehydrogenase class-1 (ALDH-1) and class-3 (ALDH-3) are established
molecular determinants of sensitivity to oxazaphosphorines, a group of
important anticancer drugs that include cyclophosphamide, 4-hydroperoxycyclophamide (4-HC), mafosfamide, ifosfamide, and 4-hydroperoxyifosfamide. Both enzymes have been shown to catalyze the
detoxification of these drugs (Sreerama and Sladek, 1997
). The
overexpression of either ALDH-1 (Magni et al., 1996; Moreb et al.,
1996, 1998) or ALDH-3 (Bunting and Townsend, 1996) has been shown to
induce resistance to 4-HC or mafosfamide in vitro. These studies have
implications for the use of these enzymes in gene therapy strategies in
vivo mainly by increasing resistance of normal hematopoietic
progenitors and thus increasing the therapeutic index of the
oxazaphosphorine drugs.
With the same technology of gene transfer, a different strategy can be
pursued in which tumor cells are manipulated to down-regulate the
expression of ALDH-1 and thus become more sensitive to treatment with
oxazaphosphorines. One such strategy is to use antisense (AS) therapy
to suppress ALDH-1 protein production. Current strategies of AS therapy
target oncogenes or drug-resistance genes (Mercola and Cohen, 1995;
Zhang, 1996), with either AS oligonucleutides or in vivo expression of
AS molecule delivered by a vector. Advantages and disadvantages of each
approach have been discussed in Mercola and Cohen (1995), Tonkinson and
Stein (1996), Zhang (1996), and Crucio et al. (1997). One of the
advantages for gene transfer approach is the theoretical ability to
selectively target the AS expression in tumor cells.
Targeting the expression of ALDH-1 can be of great clinical
significance due to its presence in many tumor types (Sreerama and
Sladek, 1997) and due to the fact that oxazaphosphorines are widely
used effective drugs. This study focuses on the effects of ALDH-1 AS
RNA on ALDH-1 expression and activity and subsequently on the
resistance to 4-HC with K562 leukemic cells and A549 lung cancer cells
in vitro. Our results show that ALDH-1 AS successfully decreases ALDH-1
expression and increases sensitivity to 4-HC.
| |
Materials and Methods |
Cell Cultures.
K562 leukemia and A549 nonsmall cell lung
cancer (NSCLC) cell lines were obtained from American Type Culture
Collection (Rockville, MD). The cells were maintained in liquid culture
with RPMI 1640 culture medium with 10% fetal bovine serum (FBS) and
used when in the exponential log phase of their proliferation.
TF-1, a leukemic cell line known for its high constitutive
expression of ALDH-1, was used as a positive control.
ALDH-1 AS cDNA Synthesis and Construction of Expression
Vector.
Generation of the ALDH-1 cDNA with truncated 5' and 3'
termini was performed by polymerase chain reaction with primers
containing BamHI cloning sites and the 1570-base pair (bp)
cloned ALDH-1 cDNA template (Moreb et al., 1996). The actual ALDH-1
cDNA is 1506 bp in length, including translational start and stop
codons. The difference between 1570 and 1506 is the result of a 64-bp 5' extension that is part of the nontranslated leading sequence for
ALDH-1. Three different AS cDNAs were synthesized as follows. The
1422-bp 3' truncated ALDH-1 cDNA was synthesized by using the 5' primer
(5'-GGATCCCGATCAGAACCAAATTGCTGAC-3') and the 3' primer
(5'-GGATCCACACTGTTCCTGCCT-3'), which removed 148 bp 5' of the
translational stop codon or the C terminus (AS-C); the 1351-bp 5'
truncated ALDH-1 cDNA was synthesized with the 3' primer (5'-GGCCATCCCGTTATGAGTTCTTCTGAGAGA-3') and the 5' primer
(5'-GGATCCAGAAGGAGATAAGGA-3'), which removed 219 bp from the 5' end,
including the 64-bp extension and 155 bp 3' of the translational start
codon or the N terminus (AS-N); and the 1203-bp 5'/3' truncated ALDH-1
cDNA was synthesized with the 5' primer (5'-GGATCCAGAAGGAGATAAGGA-3'),
which removed 219 bp from the 5' end, including the 64-bp extension and
155 bp 3' of the translational start codon as well as the 3' primer (5'-GGATCCACACTGTTCCTGCCT-3'), which removed 148 bp 5' of the translational stop codon (AS-NC).
In each case, the polymerase chain reaction fragments were cloned into
the Novagen pT7 Blue T-vector and mapped by restriction digestion
before being subcloned into the BamHI cloning site of the
pLXSN retroviral vector (Miller and Rosman, 1989) whose restriction map
has been previously published (Moreb et al., 1998). The pLXSN/truncated ALDH-1 cDNA clones oriented in the AS direction relative to the 5' long
terminal repeat promoter as determined by digestion with BglII, EcoRI, and HindIII in the case
of the 1422-bp 3' truncated ALDH-1 cDNA and with SspI,
EcoRI, and HindIII in the case of the 1351-bp 5'
truncated ALDH-1 cDNA, as well as the 1203-bp 5'/3' truncated ALDH-1
cDNA, were used in the AS experiments.
Transfection of Tumor Cells.
K562 and A549 cells were
transfected separately with 10 µg of each of the pLXSN vector or the
vector constructs with the different AS cDNAs (AS-N, AS-NC, and AS-C)
with electroporation (Bio-Rad gene pulser; Bio-Rad Laboratories,
Richmond, CA) as described in Moreb et al. (1996). The cells were
selected in 1 mg/ml G418 (Geniticin; Life Technologies, Gaithersburg,
MD) for 2 to 3 weeks, and the resulting cell population was used for
all the experiments.
ALDH-1 Enzyme Activity.
ALDH-1 enzyme activity assay was
performed every 1 to 2 weeks on 0.5 to 2.0 × 107 cells collected by centrifugation from both
wild-type (WT) and transfected cells. The activity was measured as
described in Moreb et al. (1996). Briefly, cells were washed with 1×
PBS, lysed with 6 ml of buffer solution [50 mM Tris (pH 8), 25 mM
EDTA, 5 mM 
-mercaptoethanol, 1 mM phenylmethylsulfonyl fluoride, and
0.1% Sarcosyl], and centrifuged at 4°C and 40,000 rpm for 1 h.
The supernatant was filtered through a 0.22-µm filter and used to
determine ALDH-1 activity. Aliquots of 1 ml of the supernatant from
each cell sample (WT, AS-N, AS-NC, and AS-C) were used, to which 5 mM
NAD+ and 5 mM propionaldehyde (substrate) were
added. ALDH-1 enzyme activity was monitored at 340 nm in a Beckman
DU-70 spectrophotometer at 37°C. A total of five readings of ALDH-1
activity was obtained each time for each experimental group. A control
reaction in which propionaldehyde was not added monitored the
endogenous rate of NAD+ reduction.
Western Blot Analysis.
Western blot analysis was performed
to demonstrate changes in the protein levels of ALDH-1 and ALDH-3.
However, due to low levels of expression of both enzymes in K562 cells,
adequate Western analysis was possible only for A549 cells. Protein was
extracted from cell lysates as described above, and equal amounts for
each experimental group were size separated in a 10% denaturing
SDS-polyacrylamide gel. Three identical gels were performed. One was
stained with Coommassie blue to verify equal protein loading and the
other two were electrotransferred onto nitrocellulose membranes that were blocked for 1 h in 5% nonfat dry milk dissolved in
Tris-buffered sodium containing 0.05% Tween 20. The specific proteins
were then visualized with either chicken anti-human ALDH-1 or ALDH-3
primary antibodies (supplied generously by Dr. Norman Sladek,
University of Minnesota, Minneapolis) at 1:400 dilution, secondary
antibody (horseradish peroxidase-labeled rabbit anti-chicken antibody; Sigma Chemical Co., St. Louis, MO) at 1:8000 dilution, and the enhanced
chemiluminescence (ECL) method (ECL Western blotting kit; Amersham,
Arlington Heights, IL).
Southern Analysis.
DNA was extracted and quantitated from
K562 cells of the four experimental groups (WT, AS-N, AS-NC, and AS-C)
as described in Moreb et al. (1998). To determine genomic integration
of the AS cDNAs, 25 µg of DNA from each group was digested with
BamHI, BstXI, EcoRI, and
HindIII, then separated on a 0.8% agarose gel. After
capillary blotting onto a nylon membrane (MSI, Westborough, MA) in 20×
standard saline citrate overnight, the membrane was exposed to UV
(Stratalinker UV Crosslinker 1800; Stratagene, La Jolla, CA) and then
baked at 80°C for 2 h. The blot was hybridized overnight at
65°C with random-primer [32P]deoxy cytidine
5'-triphosphate (dCTP)-labeled ALDH-1 cDNA, and washed three times at
room temperature before being autoradiographed. Blots were exposed to
X-ray film with two intensifying screens at 
80°C.
Northern Analysis.
Northern analysis was used to detect
expression of the AS RNA and its effect on ALDH-1 mRNA. Total RNA was
extracted from K562 cells of each experimental group with the
Chomczynski and Sacchi method (Chomczynski and Sacchi, 1987). Total RNA
(20 µg/lane) was separated on 1.2% agarose formaldehyde gel and
blotted onto a nylon membrane (MSI) by capillary transfer for 2 days.
The blots were hybridized overnight at 42°C with random-primer
[32P]dCTP-labeled ALDH-1 cDNA, and washed with
0.2× standard saline citrate and 0.1% SDS first at room temperature
and afterward at 56°C before being autoradiographed. Blots were
exposed to X-ray film with two intensifying screens at 80°C. These
same blots were stripped and reprobed twice, once with Neo cDNA for
correlation of Neo and ALDH-1 expression, and another with
glyceraldehyde-3-phosphate dehydrogenase (GAPDH), a housekeeping gene,
to assess equal loading of RNA or any nonspecific effects related to
the AS expression.
In Vitro 4-HC Treatment Assay.
Diethylaminobenzaldehyde
(DEAB), 4-HC, and phenylketophosphamide, a synthetic 4-HC analog not
metabolized by ALDH-1, were generously supplied by Dr. Michael Colvin
(Duke University, Durham, NC) and kept at 20°C until used. A
solution of 1 µg/µl 4-HC was prepared in culture medium, filter
sterilized (0.2 µm), and kept on ice immediately before use. DEAB (an
ALDH-1 inhibitor) also was prepared 10 to 15 min before use. To
investigate the effect of AS RNA on sensitivity to 4-HC, K562 cells
(5 × 104 cells/ml) or A549 cells (2.5 × 104 cells/ml) from each experimental group
(WT, AS-N, AS-NC, and AS-C) were treated with 4-HC and incubated for 30 min at 37°C. In some experiments, a separate WT cell group was
treated with 25 µM DEAB for 15 min before the addition of 4-HC to
compare the effect of the inhibitor versus ALDH-1 AS on sensitivity to
4-HC. After 4-HC treatment, cells were washed twice with chilled
culture medium (RPMI with 10% FBS), then plated. K562 cells were
plated in methylcellulose containing 25% FBS or in liquid culture
in RPMI containing 10% FBS, as described in Moreb et al. (1998). Colonies were counted on day 7 of methylcellulose cultures with an
inverted microscope. The total number of viable cells in liquid cultures was determined twice within a 7-day period. Viability was
determined by the trypan blue exclusion criteria. A549 cells were
plated in liquid colony assay with four 35-mm petri dishes per group.
Colonies (>10 cells) adhered to the bottom of the plate were counted
on day 4 with an inverted microscope. Untreated A549 cells were plated
similarly at 250 cells/ml/dish.
Statistical Analysis.
Statistical significance of the
difference between experimental groups was calculated with Student's
t test for two means. A P value of <.05 was
considered statistically significant
| |
Results |
ALDH-1 Activity Measurements in K562 Cells.
To determine the
effect of AS RNA on ALDH-1 expression, ALDH-1 enzyme activity was
measured in all experimental groups. Because the results with WT cells
and those transfected with pLXSN vector only were similar in multiple
experiments for both K562 and A549 (Table
1), only the results with WT cells are
reported throughout the manuscript. Furthermore, transfected cells had
similar morphology and proliferation characteristics as WT cells.
ALDH-1 activity was measured over a 3-month period for each AS K562
cell type. Initially, all of the AS cells showed a significant decrease
in ALDH-1 activity, however, ALDH-1 activity was restored by day 40 of
culture and while the cells were maintained without G418 (Table
2). ALDH-1 AS expression was regained and
a decrease in the ALDH-1 activity was achieved again in the same cells
on reexposure and maintenance in 1 mg/ml G418. Figure
1 represents the mean ± S.E. of the
activity measured over several weeks of culture for each experimental
group. The ALDH-1 activity in AS-N and AS-NC cells was significantly
lower than that for WT cells (P < .001). Although
reduction in ALDH-1 activity was seen in the AS-C cells as well (Table
2), it did not reach significance when the overall mean was compared
with that of the WT cells. This is most likely explained by lack of
consistent effect of AS-C on blocking the translation of ALDH-1 protein
and fluctuation in its activity over time in the AS-C K562 cells.
View this table:
[in this window]
[in a new window]
|
TABLE 1
Effect of expression of pLXSN retroviral vector only on ALDH-1 enzyme
activity and 4-HC sensitivity compared with WT cells
|
|
View larger version (18K):
[in this window]
[in a new window]
|
Fig. 1.
ALDH-1 activity of the four experimental groups: K562
WT, AS-N, AS-C, and AS-NC. Columns represent the mean ± S.E. of
at least 35 measurements taken on different occasions over 3 months of
continuous cell culture. The P values for AS-N and AS-NC
were <.001 compared with WT.
|
|
ALDH-1 AS Integration and Expression in K562 Cells.
In view of
the reduction in ALDH-1 activity, and to verify that this reduction is
indeed the result of the ALDH-1 AS cDNA genomic integration and mRNA
expression, several studies were performed. Figure
2 shows a Southern analysis in which DNA
from the different K562 experimental groups was digested with four different restriction enzymes, blotted, and labeled with
32P-labeled ALDH-1 cDNA. Digestion with
BamHI, the cloning site for ALDH-1 cDNAs in the pLXSN
vector, reveals two DNA bands that were common among all groups and
most likely represent the genomic ALDH-1 DNA. Other bands of different
sizes (1351, 1422, and 1203 bp) represent the AS ALDH-1 cDNAs: AS-N,
AS-C, and AS-NC, respectively. Digestion with EcoRI, which
cuts once in the vector upstream to the ALDH-1 cDNA, and
HindIII, which cuts once in the vector downstream to the
ALDH-1 cDNA, both showed two extra bands with all AS groups compared
with WT DNA, indicating two different integration sites. Digestion with
a noncut restriction enzyme, BstXI, resulted in at least two
DNA fragment larger than the vector that are not seen in the WT DNA and
are different in size among the different AS clones. Overall, these
results demonstrate the genomic integration of the AS vector constructs
in two different sites.
View larger version (87K):
[in this window]
[in a new window]
|
Fig. 2.
Southern analysis of genomic DNA obtained from K562
WT, K562 transformed with ALDH-1 cDNA minus 200 bp of the N terminus
(N), K562 transformed with ALDH-1 cDNA minus 140 bp of the C terminus
(C), and K562 transformed with ALDH-1 cDNA minus both termini (NC)
digested with several restriction enzymes and probed with a randomly
primed [32P]dCTP-labeled 1570-bp ALDH-1 cDNA detected by
autoradiography over a 2-week period incubated at 80°C and with two
intensifying screens.
|
|
Northern analysis was performed to detect the presence of retrovirally
transcribed ALDH-1 AS mRNA in the different K562 experimental groups
(Fig. 3). The results indicate the
presence of the expected AS mRNA in all three AS K562 cells (Fig. 3,
top, top band), whereas the WT ALDH-1 mRNA [top, lower band (~3
kb)] can be seen in the TF-1 cells (first lane) and WT K562 (second
lane) but not detected in the cells expressing the ALDH-1 AS mRNA.
Total RNA from TF-1 leukemic cells was included as a positive control
for WT ALDH-1 because these cells are known to have high levels of WT
ALDH-1 mRNA (Moreb et al., 1995, 1998). Neomycin resistance gene was detected only in the AS transfected K562 cells (Fig. 3, middle).
View larger version (59K):
[in this window]
[in a new window]
|
Fig. 3.
Northern analysis of total RNA obtained from WT TF-1,
WT K562 (WT), K562 transformed with ALDH-1 cDNA minus 140 bp of the C
terminus (C), K562 transformed with ALDH-1 cDNA minus 140 bp of the N
terminus (N), and K562 transformed with ALDH-1 cDNA minus both
termini (NC) probed with a randomly primed
[32P]dCTP-labeled 1570-bp ALDH-1 cDNA, 790-bp neomycin
phosphotransferase cDNA, or GAPDH cDNA detected by autoradiography for
either 14 days (ALDH-1), 7 days (Neo), or 24 h (GAPDH) incubated
at 80°C and with two intensifying screens. Ethidium bromide
staining of the original Northern agarose gel is shown as a second
loading control.
|
|
Figure 4 shows the results of similar
Northern analysis as in Fig. 3, however, the RNA was obtained from the
same K562 experimental groups at a later time of cell culture that
corresponded to the increase in the ALDH-1 activity in the three AS
experimental groups while cultured in the absence of G418 (Table 2).
Figure 4 shows the loss of AS mRNA in all cell lines (AS-C, AS-N, and
AS-NC) with rebound increase in the WT ALDH-1 mRNA. These findings
correlate very well with the increase in ALDH-1 activity shown in all
these cells in Table 2, and confirm the loss of the retrovirally
transfected ALDH-1 AS expression. The neomycin resistance gene (Fig. 4,
middle) also was undetectable. However, GAPDH expression was not
affected by the expression of ALDH-1 AS or lack of it (Figs. 3 and 4).
View larger version (75K):
[in this window]
[in a new window]
|
Fig. 4.
Northern analysis of total RNA obtained from WT TF-1,
WT K562 (WT); K562 transformed with ALDH-1 cDNA minus 148 bp of the C
terminus (C); K562 transformed with ALDH-1 cDNA minus 219 bp of the N
terminus (N), including 64 bp of uncoded sequence; K562 transformed
with ALDH-1 cDNA minus both termini (NC); and the AM12 packaging cell
line. As opposed to the Northern analysis in Fig. 3, the K562 cells
expressing ALDH-1 AS were grown for >1 month in the absence of G418
before RNA was extracted. Northern blot was probed with a randomly
primed [32P]dCTP-labeled 1570-bp ALDH-1 cDNA, 790-bp
neomycin phosphotransferase cDNA, or GAPDH cDNA detected by
autoradiography for 24 h (for all three different probes),
incubated at 80°C and with two intensifying screens. Longer
exposure for 7 days still did not show any other bands. Ethidium
bromide staining of the original Northern agarose gel is shown as a
second loading control.
|
|
ALDH-1 AS Overexpression Sensitizes K562 Cells to 4-HC.
To
determine whether ALDH-1 AS RNA expression led to increased sensitivity
to 4-HC, K562 cells, either WT or expressing ALDH-1 AS (AS-C, AS-N and
AS-NC), were treated with various doses of 4-HC (10 and 15 µg/ml) and
cultured in methylcellulose colony assay or liquid culture. Table 2
shows the results obtained from one of three representative experiments
with similar results. Data represent the mean ± S.D. of the total
colonies counted as well as the percentage recovery after 4-HC
treatment compared with untreated cells. It is shown that cells
expressing ALDH-1 AS were more sensitive to 4-HC than WT cells with
AS-N and AS-NC exhibiting significantly increased sensitivity
(P < .005) compared with AS-C. These results
correspond well to the ALDH-1 activity results shown in Fig. 1. Table
3 also shows that the addition of 25 µM
DEAB 10 min before the treatment of WT cells with 10 µg/ml 4-HC
affected 4-HC sensitivity to a similar proportion seen with AS-N and
AS-NC. Furthermore, similar treatment with 15 µg/ml phenylketophosphamide showed no significant difference between the WT
and AS cells, indicating the specificity of the ALDH-1 AS effects.
Table 4 shows the summary results of
three liquid culture experiments in which K562 cells were treated with
10 µg/ml 4-HC. The results are expressed as the mean ± S.D. of
total viable cells in three flasks for each experimental group. Again,
the results demonstrate significantly lower cell survival and recovery
in all ALDH-1 AS groups compared with WT cells (P < .05).
Table 5 shows the results of one of two
representative experiments conducted similarly to those described in
Table 3. These results again demonstrate that with the loss of ALDH-1
AS effect and the rebound increase in ALDH-1 activity and mRNA (Fig.
4), the relative resistance to 4-HC is restored in the AS cells (AS-C, AS-N, and AS-NC) compared with WT K562 cells. The increase in relative
resistance is reflected in significantly (P < .025)
higher recovery of total colonies in all three AS groups compared with WT cells.
ALDH-1 AS Overexpression Reduces ALDH-1 Activity and Sensitizes
A549 Cells to 4-HC.
To study the effectiveness of the AS approach
when large amounts of the target mRNA is present, we overexpressed the
same AS constructs in A549 NSCLC cell line, which is known to have very
high levels of both ALDH-1 and ALDH-3 (Sreerama and Sladek, 1997).
The same retroviral constructs with AS-N, AS-C, and AS-NC were
electroporated into A549 cells. After selection in G418, cells were
analyzed for ALDH-1 enzyme activity and protein levels and treated with
4-HC in vitro. Table 6 shows significant
decrease in enzyme activity (with propionaldehyde as a substrate) in
all AS groups compared with WT cells. Similar decrease in activity was
demonstrated when other substrates, acetaldehyde and benzaldehyde, were
used in the reaction (data not shown). In addition, Table 6 shows
proportional increase in sensitivity to 4-HC in all AS cells.
The effect of the different AS constructs expression on ALDH-1 and
ALDH-3 proteins was studied with Western blot analysis. The same
protein samples used for the enzyme activity were used for the Western
analysis shown in Fig. 5. The results
confirm those obtained by enzyme activity assay and demonstrate
significant decrease in ALDH-1 protein levels in all AS groups, with
some variability. However, ALDH-3 protein levels were relatively
unaffected.
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 5.
Western blot analysis of protein extracted from A549
cells transfected with the different ALDH-1 AS constructs (AS-NC, AS-C,
and AS-N) and WT cells. Two identical blots were labeled with either
chicken anti-human ALDH-1 or ALDH-3 and horseradish peroxidase-labeled
secondary antibody. Detection is by the ECL detection system.
|
|
| |
Discussion |
The relationship between ALDH-1 and resistance to the
oxazaphosphorines has been established for many years mainly through correlative studies, use of ALDH-1 inhibitors, and most recently by
gene transfer studies (Kohn and Sladek, 1985; Sladek and Landkamer, 1985; Russo and Hilton, 1988; Bunting and Townsend, 1996; Magni et al.,
1996; Moreb et al., 1996, 1998; Sreerama and Sladek, 1997). Because of
several other potential mechanisms of resistance against oxazaphosphorines (McGowan and Fox, 1986; Yuan et al., 1990; Friedman et al., 1992; O'Conner et al., 1992), the exact contribution of ALDH-1
to that resistance in any cell type is still being debated and studied.
In this report, we show that suppression of ALDH-1 expression (Figs. 3
and 5) and activity (Tables 2 and 6) with gene transfer to express
ALDH-1 AS RNA in two different cell lines, results in a significant
increase in the sensitivity to 4-HC, an active metabolite of
cyclophosphamide. The increase in sensitivity to 4-HC in some of the
cells with ALDH-1 AS was similar to that seen with DEAB, a specific
inhibitor of ALDH-1.
AS strategies for gene silencing have attracted much attention in
recent years and despite potential obstacles, AS technology (oligonucleutides or AS genes) has been widely used in the laboratory for studying gene function and in recent years introduced into the
treatment of cancer patients. Because cyclophosphamide (CP) is one of
the most commonly used alkylating agents and because one of the
resistance mechanisms to CP is high levels of ALDH-1, a strategy to
reduce ALDH-1 protein with gene transfer to express AS RNA in tumor
cells may increase the therapeutic index of CP and lead to a higher
rate of complete responses. Our studies demonstrate that such a
strategy can be effective, both in K562 cells with low constitutive
expression of ALDH-1 and in A549 cells with very high constitutive
expression of ALDH-1, and in the presence of high levels of ALDH-3 as
well. Given the promising results with A549 NSCLC cell line, an AS
approach can potentially be used in the treatment of locally advanced
NSCLC cells. Similar to phase I gene therapy studies already reported
(Roth et al., 1996; Tursz et al., 1996), an adenoviral vector
containing the ALDH-1 AS can be directly injected into locally advanced
NSCLC cells, then treated with i.v. CP. Such strategy to reduce the
levels of ALDH-1 may increase the therapeutic index of CP and lead to
higher rates of response.
Another approach to increasing sensitivity of tumors to CP has been
reported by Chen et al. (1996) and is based on the use of cytochrome
P450 (CYP) gene CYP2B1, which activates the prodrug CP into its
4-hydroxy-CP active metabolite. Such activation occurs normally in the
liver in vivo, however, overexpression of the CYP2B1 in tumor cells was
shown to increase susceptibility to CP and ifosfamide. Such approach
might have limited use in tumors expressing high levels of ALDH-1 or
ALDH-3. Therefore, our approach could have greater clinical
implications and could possibly be used in conjunction with other
approaches such as the one mentioned above.
The significance of aldehyde dehydrogenases in the oxidation of
retinaldehyde to retinoic acid has been known for many years (Lee et al., 1991; Labrecque et al., 1995; Bhat et al., 1996; Zhao et al., 1996). Because of the recently demonstrated critical role
of ALDH-1 in embryonic differentiation by converting retinaldehyde to
retinoic acid (Berggren et al., 1999; McCaffery et al., 1999), our AS
approach to studying the function of ALDH-1 may have another potentially significant use in studying embryonic differentiation.
Of interest is the difference in effectiveness of the different AS
constructs we used. Our results suggest that AS sequences missing the N
terminus (AS-N and AS-NC) are more effective in blocking ALDH-1
activity than the sequence missing the C terminus (AS-C). Such effect
has been described in Ch'ng et al. (1989) and the most likely
explanation is that the sense-AS duplex is protected against the
ribosome's unwinding activity during RNA translation when the AS
molecule is missing the N terminus (Shaki-Eshleman and Liebhaber,
1988). Thus, in spite of good levels of AS-C mRNA (Fig. 3), AS-C was
less effective in reducing the ALDH-1 activity and increasing 4-HC
sensitivity in K562 cells (Fig. 1 and Table 3). Similar differences,
although to a lesser degree, were seen in A549 cells (Fig. 5 and Table
6). Several other factors that determine the effectiveness of an AS
molecule have been reported and include the following: 1) rate of
transcription of the AS versus sense gene; 2) stability of the AS gene;
3) target site of the AS RNA; 4) rate of duplex formations; and 5)
location of the AS gene versus that of the targeted gene (Mercola and
Cohen, 1995; Zhang, 1996; Arndt and Hank, 1997). Any one or combination of these factors could have contributed to the different effectiveness of AS-C.
Finally, our results suggest that growing the ALDH-1 AS cells in the
absence of G418 selective pressure, results in the loss of AS mRNA
expression and rebound increase in the WT ALDH-1 mRNA (Fig. 4). We are
not aware of any previous reports on such rebound phenomenon with the
AS strategy. Although it is reversible after re-exposure to G418 in
vitro, it is clearly of concern because according to our results (Table
5), it can result in the opposite effect, i.e., an increase in
resistance to 4-HC or CP of the affected tumor cells. However,
sensitizing the cells to the toxicity of 4-HC or CP may result in the
elimination of all cells expressing the AS and, therefore, such rebound
phenomenon may not be clinically significant. In vivo studies with an
animal model are essential to address the significance of such
phenomenon. Last, the mechanism for such loss of AS expression was most
likely due to the absence of selection pressure because re-exposure to
G418 resulted in the restoration of AS effect and continuous exposure
to 0.4 to 1 mg/ml G418 maintains the expression of the different AS
mRNAs. Furthermore, repeat Southern analysis, with DNA obtained from all the experimental groups during the loss of AS expression and digested with BamHI, continued to show the presence of the
different AS cDNAs (data not shown).
In summary, we have defined ALDH-1 AS constructs that can effectively
reduce the ALDH-1 expression and activity and increase sensitivity to
4-HC in vitro. Our studies again demonstrate the significant
contribution of ALDH-1 to the resistance against oxazaphosphorines and,
therefore, the potential clinical application of ALDH-1 AS RNA in
cancer treatment. Any such clinical application will have to
specifically target the tumor cells to avoid introducing the ALDH-1 AS
into normal cells.
ALDH-1, aldehyde dehydrogenase class-1;
4-HC, 4-hydroperoxycyclophosphamide;
AS, antisense;
NSCLC, nonsmall cell lung
cancer;
FBS, fetal bovine serum;
bp, base pair;
WT, wild type;
ECL, enhanced chemiluminescence;
dCTP, deoxy cytidine 5'-triphosphate;
GAPDH, glyceraldehyde-3-phosphate dehydrogenase;
DEAB, diethylaminobenzaldehyde;
CP, cyclophosphamide;
CYP, cytochrome P450.
![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
Top
Abstract
Introduction
