Vol. 44 No.

6 SCIENCE IN CHINA (Series C) December 2001

Activity identification of chimeric anti-caspase-3 mRNA

hammerhead ribozyme in vitro and in vivo
XU Renhuan (徐人欢)1,2, LIU Jing (刘 静)2, XU Feng (徐 丰)2,
JIANG Ge (江 舸)2, XIE Qing (谢 青)1, ZHOU Xiaqiu (周霞秋)1,
JIN Youxin (金由辛) 2 & WANG Debao (王德宝)2
1. Department of Infectious Diseases，Ruijin Hospital，Shanghai Second Medical University, Shanghai 200025, China;
2. State Key Laboratory of Molecular Biology, Institute of Biochemistry and Cell Biology, Shanghai Institutes of Life
Sciences, Chinese Academy of Sciences, Shanghai 200031, China
Correspondence should be addressed to Zhou Xiaqiu or Jin Youxin (email: yxjin@sunm.shcnc.ac.cn )

Received March 28, 2001

Abstract To study the expression activity of various vectors containing anti-caspase-3 ribozyme
cassettes in vivo, and to further study the role of caspas-3 in the apoptotic pathway, we con-
structed anti-caspase-3 hammerhead ribozyme embedded into the human snRNA U6, and de-
tected the activity of the ribozyme in vitro and in vivo. Meanwhile we compared it with the
self-cleaving hammerhead ribozymes that we previously studied, and with the general ribozyme,
cloned into RNA polymerase II expression systems. The results showed that the three ribozymes,
p1.5RZ107, pRZ107 and pU6RZ107 had the correct structure, and that they could cleave cas-
pase-3 mRNA exactly to produce two fragments: 143nt/553nt. p1.5RZ107 has the highest
cleavage efficiency in vitro, almost 80%. However, the U6 chimeric ribozyme, pU6RZ107, has the
highest cleavage activity in vivo, almost to 65%, though it has lower cleavage activity in vitro. The
cleavage results demonstrated that the pU6RZ107, the U6 chimeric ribozyme, could more
efficiently express and downregulate the level of caspase-3 in vivo, and the ribozyme could
provide an alternative approach to the research into the mechanism of apoptosis and human gene
therapy also.
Keywords: ribozyme, caspase-3, transcript, cleavage reaction, BRL-3A cell.

Apoptosis is a fundamental and complex biological process that enables an organism to kill
and remove unwanted cells. Of all the proteins implicated in the activation and execution of
apoptosis, the ICE-like proteases or caspases (14 so far identified in humans) stand out as being
crucial for this process in diverse metazoan organisms. In mammals, caspases (principally cas-
pase-3) appear to be activated in a protease cascade that leads to inappropriate activation or rapid
disablement of key structural proteins and important signaling, homeostatic and repair enzymes[1].
Recent work has revealed that cell excessive apoptosis is correlated with some diseases, for
example, neurodegenerative disorder and inflammation. So, it is possible to treat these apop-
tosis-relative diseases by inhibition of apoptosis using some inhibitors of caspases[2].
Hammerhead ribozymes are small catalytic RNAs capable of cleaving RNA substrates in an
intermolecular reaction. In the last few years, ribozymes-mediated gene inhibition has been dem-
onstrated many times in cell systems as well as in the whole animals[3, 4]. However, it has become
 No. 6 ACTIVITY IDENTIFICATION OF RIBOZYME in vitro & in vivo 619

quite clear that the design for obtaining effective ribozymes can still be improved and several im-
portant parameters can be optimized: (i) persistent high-level expression in transduced cells; (ii)
targeting of the RNA into the correct subcellular compartment; (iii) appropriate folding of the
RNA molecules for functional activity[5,6].
We previously designed anti-caspase-3 hammerhead ribozymes, which were cloned into the
self-cleaving vector, and have low cleavage efficiency in cells for the ribozyme easy to degrade by
RNase existing in cells, though the ribozymes have perfect cleavage in cell-free condition[7]. In
this study, we constructed anti-caspase-3 hammerhead ribozyme, which was embedded into the
human snRNA U6, and detected the activity of the ribozyme in vitro and in vivo. Meanwhile we
compared it with the self-cleaving hammerhead ribozymes that we previously studied, and with
the general ribozyme, cloned into RNA polymerase II expression systems which have been in de-
velopment for years. The results showed that the U6 chimeric ribozyme has the perfect cleavage
activity in vivo.

1 Materials and methods

1.1 Materials
The self-cleaving vector p1.5 was presented by courtesy of Prof. QI Guorong (Shanghai In-
stitute of Biochemistry, the Chinese Academy of Sciences); T4 DNA ligase, pGEM-T EASY vec-
tor kit, Transcription kit and Trizol kit were purchased from Promega Company. RT-PCR kit was
purchased from Takara Company. α-32P-UTP and α-32P-dATP were purchased from Beijing
Ya-Hui Company. E. coli DH5α was kept in our lab; rat liver cell line, BRL-3A was purchased
from Shanghai Institute of Cell Biology, the Chinese Academy of Sciences; DMEM was pur-
chased from Gibco BRL Company. New born calf serum was purchased from Hyclone Company.
The PCR primer and ribozyme fragments were chemically synthesized in the Beckman
Oligo-1000 DNA synthesizer; human anti-caspase-3 antibody was purchased from Santa Cruz
Company, and human anti-actin antibody was purchased from Gibco Company.

1.2 Methods
1.2.1 Construction of transcription plasmids for target RNAs and expression plasmids for ri-
bozymes
(i) Construction of cell-free transcription plamids for target RNAs—pCaspase-3. Total RNA
was extracted using the Trizol kit from the brain tissue of male Sprague-Dawley rat. The extracted
RNA was reversely transcribed and PCR amplified using a pair of primers in the one step RT-PCR
kit. The products were analyzed and purified on 1 % agarose gel and ligated with pGEM-T EASY
vectors. DNA sequencing results showed that the PCR-amplified fragments were cloned into the
multiple cloning sites of pGEM-T EASY vectors under the control of the T7 promoter. The recon-
structed plasmids were named pCaspase-3. The primers used were, upstream primer: 5′
-CTA
GAA AAA GTG ACC ATG GAC AAC A-3′(20-40nt); downstream primer: 5′
- GGT ACC CCT
 620 SCIENCE IN CHINA (Series C) Vol. 44

TGA ATT TCT CCA GGA AT-3′(641-

660nt) (fig. 1).
(ii) Expression plasmids for ri-
bozymes against rat caspase-3. The
hammerhead ribozyme—Rz107 was de-
signed according to the computer soft-
ware compiled by Professor Zuker (Na-
Fig. 1. Construction of pCaspase-3. The PCR amplified frag- tional Academy of Sciences, Canada).
ments of caspase-3 were cloned into T-vector. The lengths of
The homologous possibility with the gene
RNA transcripts from Spe I-linearized templates should be 696
nt for pCaspase-3. The ribozyme (RZ107) was designed to of rat was excluded by considering the
cleave at a kinetically preferred GUC triplet site that would RNA sequence of rat cell from NCBI
generate fragments of 143 and 553 bases.
Gene bank (GeneBank accession number
U84410). The oligonucleotides of RZ107
were 5′
-CTA GAT CCA TCC TGA TGA GTC CGT GAG GAC GAA ACT TGC TGG TAC-3'
and 5′
-CAGCAA GTT TCG TCC TCA CGG ACTCAT CAG GAT GGA T-3′
.
a) Construction of p1.5RZ107: The two oligonucleotides of RZ107 were ligated into lin-
earized vector p1.5, which has self-cleaving function between Xba I and Kpn I. The reconstruction
was confirmed by DNA sequencing. The reconstruction plasmid then was digested with the EcoR
I and Hind III restriction enzymes, subcloned into the mammal expression vector pcDNA3.1. The
reconstructed plasmid was named p1.5Rz107 (fig. 2(a)).
b) Construction of vector pRZ107: The two oligonucleotides of Rz107 were ligated into lin-
earized vector pcDNA3.1 between Kpn I and Xba I. The reconstruction plasmid was named
pRZ107, and was affirmed by DNA sequencing (fig. 2(b)).
c) Construction of plasmid pU6RZ107: The construction of pBSKneorU6′was the same as
described previously[8]. The cassettes of RZ107 were cloned into pBSKneorU6′between Xba I
and BamH I. The reconstruction plasmid was named pU6RZ107. pU6RZ107 was transcribed by
T7 RNA polymerase in vitro, and transcribed by RNA polymerase III in vivo, driven by human
snRNA U6 promoter (fig. 2(c), (d)).

1.2.2 Transcription in vitro

(i) Cell-free transcription and purification of target RNA. In vitro transcription was carried
out at 37℃ for 90 min using the transcription kit in a 20 μL final volume containing 10 μCi
[α-32P]UTP as described previously[7]. Target RNA samples were purified by cutting off the
autoradiograph bands after running on 6% polyacrylamide 8-mol/L urea gel as described previ-
ously[7].
(ii) Preparation and purification of pCaspase-3. The template of plasmids, p1.5RZ107,
pRZ107 and pU6RZ107 were linearized with Hind III, Apa I and Sca I respectively. In vitro
 No. 6 ACTIVITY IDENTIFICATION OF RIBOZYME in vitro & in vivo 621

Fig. 2. Construction of p1.5RZ107 (a), pRZ107 (b), pU6RZ107 (c) and the expected structure of transcript for pU6RZ107 in vivo
(d). (a) Construction of p1.5RZ107: as the vector has the self-cleaving structures outside the ribozyme, when it was transcribed,
it would present three bands of 5'-cis-RZ (63 nt), 3′
-cis-RZ (50 nt) and RZ (59 nt). The vector of p1.5RZ107 was transcribed by
T7 promoter in vitro, while, it was driven by CMV promoter in vivo. (b) Construction of pRZ107: the lengths of RNA transcripts
from Apa I-linearized templates should be 110 bases in vitro, and it depends on the CMV promoter in vivo. (c) Construction of
pU6RZ107: in this study, the plasmid was transcribed by T7 promoter in vitro, and the lengths of RNA transcripts from Sca
I-linearized templates would be 508 bases, while it will be transcribed by the U6 promoter in vivo instead. (d) The expected
transcripts of pU6RZ107: the stability of the transcripts containing the ribozyme gene should be strongly enhanced by 5′
capping,
which is directed by a 5′
self-complementary hairpin.

transcription and purification were done as described above. The transcription of p1.5RZ107
yielded three bands on the gel: 5′
-cis-Rz(63nt), 3′
-cis-Rz(50nt) and Rz107 (59nt). The bands of
59nt (p1.5RZ107), 110nt (pRZ107) and 508nt (pU6RZ107) were cut off from the gel.

1.2.3 Ribozyme-mediated cleavage of caspase-3 mRNA in vitro. Rz107 and the target RNA
were quantified by measuring their radioactive cpm in 1 μL solution. The cleavage reaction was
carried out in 5μL solution containing 50 mmol/L Tris-Cl (pH 7.5), 20 mmol/L MgCl2 and
20mmol/L NaCl. The substrate in one assay was 10000—20000 cpm. The molar ratio of Rz107 to
substrate RNA was 1:1, which was estimated according to the cpm number combined with the U
number in their RNAs. To finish the reaction, 1 μL stop solution (0.25% bromophenol blue, 0.25%
xylene cyanol FF, 20 mmol/L EDTA and saturated urea) was added. After running a 6% denatur-
ing PAGE and autoradiographing, the cleavage results could be analyzed. The cleavage efficiency
 622 SCIENCE IN CHINA (Series C) Vol. 44

(CE) was calculated from cpm values of the bands of substrates (S) and products (P):
CE=[P/(S+P)]·100%.

1.2.4 Assessment of ribozymes’ functional activity in vivo

(i) Cell culture and stable transfectants. The rat liver cell line BRL-3A was maintained in
DMEM with 10% of new born calf serum (NCS), 100 unit/mL penicillin and 100 μg/mL of strep-
tomycin sulfate at 37℃ in 5% CO2 atmosphere.
To generate stable cell clones, 5×106 cells were collected and electroporated (500 μF, 250 V)
with 10 μg of the plasmid p1.5RZ107, pRZ107, pU6RZ107 and pcDNA3.1 respectively. The
transfected cells were grown in DMEM containing 10% NCS for 48 h before the addition of G418.
After four days of growth in selection medium, the G418-resistant cells were seeded in DMEM
containing 1% methylcellulose and 1 mg/mL of G418 to isolate single cell clones. The clones
were maintained in DMEM medium containing 0.5 mg/mL G418 and analyzed for ribozyme ex-
pression.
(ii) RT-PCR analysis of caspase-3 mRNA. Total RNA from BRL-3A cells containing three
different plasmids was obtained using the Trizol kit. The fragments of caspase-3 and β-actin were
amplified using one-step RT-PCR kit in 2 microgram of total cytoplasmic RNA. The caspase-3
PCR primers are as shown above. The β-actin was used as the internal control. Its primers were:
5′
-ATT TGG CAC CAC ACT TTC TAC A-3′
; 5′
-TCA CGC ACG ATT TCC CTC TCA G-3′
The PCR was carried out for 30 cycles with enough caspase-3 and β-actin primers. The amplified
DNA fragments were 696 bp for caspase-3 mRNA, and 380 bp for the β-actin mRNA.
(iii) Northern blot. Total cellular RNA was extracted from pellets of 106 cells, which were
transfected with p1.5RZ107, pRZ107, pU6RZ107 and pcDNA3.1 using Trizol kit, respectively.
Then we processed the detection of caspase-3 mRNA and actin mRNA as described by Sambrook
et al.[9].
(iv) Western blot. For Western blot analysis, 5×106 cells, transfected with p1.5RZ107,
pRZ107, pU6RZ107 and pcDNA3.1, were harvested and lysed; total proteins were extracted and
the analysis of proteins was processed as described by Sambrook et al.[9].
(v) Measurement of caspase-3 activity. 5×106 cells were washed in phosphate-buffered saline,
lysed, and processed for caspase-3 activity using the ApoAlert caspase-3 colorimetric assay kit
according to the manufacturer instructions.

2 Results

2.1 Identification of transcripts in vitro of pCaspase-3, p1.5RZ107, pRZ107 and pU6RZ107

The lengths of RNA transcribed from Spe I-linearized templates should be 696 nt for pCas-
pase-3 (fig. 1).
Because vector p1.5 is a self-cleaving functional prokaryotic plasmid, its transcription in vi-
tro depends on T7 RNA polymerase in vitro, while the transcription of p1.5RZ107 depends on
 No. 6 ACTIVITY IDENTIFICATION OF RIBOZYME in vitro & in vivo 623

RNA polymerase II in vivo. Because 5′

-cis and 3′
-cis have been self-cleaved, the structure of Rz
was released. Thus, it would decrease the interference of no sense sequences. There were three
bands: 5′
-cis-Rz(63nt), 3′
-cis-Rz(50nt) and Rz107 (59nt)(fig. 3(a)). The results agreed with our
design.
The transcriptions of pRZ107 and pU6RZ107 in vitro were 110 nt and 508 nt respectively
(fig. 3(b), (c)). In theory, the transcribed product U6 mutant RNA was 64-nts long, retaining the
original 5′hairpin, in order to obtain the 5′
-
γ-monomethyl phosphate cap. This initial se-
quence is followed by three restriction sites
(XbaI, SalI and BamHI) and the native U6
uridine-rich 3′terminus. Data from Hela nu-
clear extract in vitro transcription experiments
demonstrated that RNA polymerase III in Hela
nuclear extract can effectively recognize the up- Fig. 3. In vitro transcripts of p1.5RZ107 (a), pRZ107 (b),
stream promoter and enhancer of human U6 and pU6RZ107 (c). (a) In vitro transcripts of p1.5RZ107: as
5′
-cis-RZ and 3′
-cis-RZ have been self-cleaved, the struc-
snRNA gene on pBSKneorU6′
. As pBSKneorU6′
ture of RZ was released. Thus, there were three bands: 5′
-
contains G418 resistant gene neor, it can be inte- cis-RZ (63 nt), 3′
-cis-RZ (50 nt) and RZ (59 nt), and the
grated into eukaryotic genomic DNA and can band of 59 nt was the transcripts of ribozyme. (b) and (c)
stably, continually express U6 mutant RNA in were the transcripts of pRZ107 and pU6RZ107, and the
[8] bands were 110 nt and 508 nt respectively.
vivo . In this study, pU6RZ107 was the tran-
script by T7 RNA polymerease in vitro, while in vivo, it was driven by U6 promoter (fig. 2(c)).
The transcript product is shown in fig. 2(d).

2.2 Detection of ribozyme-mediated cleavage in vitro

The cleavage results demonstrated that the three designed ribozymes, p1.5RZ107, pRZ107
and pU6RZ107 had the correct structure, and that they could cleave caspase-3 mRNA exactly to
produce two fragments: 143nt/553nt. p1.5RZ107 has the highest cleavage efficiency, almost 80%,
while the cleavage efficiencies of pRZ107 and pU6RZ107 were 55% and 50%, respectively.

2.3 Ribozyme-mediated cleavage of caspase-3 mRNA in vivo

To determine the intracellular activities of the ribozymes in BRL-3A cells, caspase-3 mRNA
expression was examined by RT-PCR. The results show that caspase-3 mRNA expression de-
creased markedly in the BRL-3A cells that have been electroporated with pU6RZ107 (fig. 4).
Through analysis of the intensities of caspase-3 and β-actin compared to control, we calculate
the intracellular cleavage efficiency of p1.5RZ107, pRZ107 and pU6RZ107 to be 21%, 42% and
65% respectively (fig. 5).
For analyses of downregulation of caspase-3 mRNA, Northern blot was used to measure the
levels of caspase-3 mRNA. As shown in fig. 6, the level of caspase-3 was markedly decreased
 624 SCIENCE IN CHINA (Series C)
Fig. 4. Cleavage of rat caspase-3 RNA by RZ107 in
vitro. Caspase-3 RNA was incubated with water
(lane 1), pU6RZ107 (lane 2), pRZ107 (lane 3) or
p1.5RZ107 (lane 4) at 37℃ for 30 min. After a 6%
denaturing PAGE and autoradiographing, the cleav-
age results could be analyzed. The results show that
p1.5RZ107 has the highest cleavage efficiency,
almost 80%, while the cleavage efficiencies of
pRZ107 and pU6RZ107 were 55% and 50%,
respectively.
Fig. 6. Northern blot analysis of ribozymes expression in
vivo. Total cellular RNA was extracted, and 20 μg of total
RNA from cells transfected with pU6RZ107 (lane 1),
pRZ107 (lane 2), p1.5RZ107 (lane 3) and pCDNA3.1
(lane 4) was subjected to electrophoresis in a 1% agarose
gel containing 5.4% formaldehyde and transferred to
nylon membranes. The blots were hybridized with probes
specific for caspase-3 and β-actin genes. The result shows
that the transfected-pU6RZ107 cells have the lowest level
of caspase-3.
Vol. 44
Fig. 5. Quantitative RT-PCR of endogenous caspsase-3 in
BRL-3A cells transfected with pCDNA3.1 (lane 1), pU6RZ107
(lane 2), pRZ107 (lane 3) and p1.5RZ107 (lane 4). Total mRNA
was amplified for caspase-3 (696 bp) and β-actin (380 bp).
After electrophoresis and staining with ethidium bromide, the
bands were quantified by densitometry. The cellular cleavage
efficiency of various ribozymes was calculated according to the
ratio of caspase-3 to β-actin band volumes compared to the
control. It showed that the cleavage efficiencies of p1.5RZ107,
pRZ107 and pU6RZ107 were 21%, 42% and 65%, respectively.
with respect to that of p1.5RZ107 and pRZ107
compared to the bands of β-actin correspond-
ingly. The decreased levels were similar to the
results of RT-PCR.
To confirm that the Rz107 could affect
protein production, Western blot was used to
measure levels of caspase-3 protein. The West-
ern blot analysis showed that cells transfected
with pU6RZ107 had significantly lower protein
level of caspases-3 (19%) than that transfected
with pRZ107 (38%) and p1.5RZ107 (19%)
compared to the levels of β-actin (fig. 7).
The effect of Rz107 on caspase-3 expres-
sion was also analyzed functionally. As shown in
fig. 8, protein extracts from cells, which were
 No. 6 ACTIVITY IDENTIFICATION OF RIBOZYME in vitro & in vivo 625

Fig. 7. Analysis by Western blot. For Western blot

analysis, cells were lysed and separated by SDS-
polyacrylamide gel electrophoresis and transferred
onto polyvinylidene difluoride membrane. The result-
ing membrane was blocked with 10% skim milk,
incubated with 2 μg/mL anti-pro-caspase-3 antibody
and actin antibody, and the signals were detected by
use of a Western blotting kit. Lanes 1, 2, 3 and 4 rep-
resent the levels of caspase-3 and actin of cells trans-
fected with pU6RZ107, pRZ107, p1.5RZ107, and
pCDNA3.1, respectively. The results show that lane 1
has the lowest level of caspase-3 protein.
Fig. 8. Analysis of caspase-3 enzyme activity by ELISA. As
shown in fig. 8, protein extracts from cells, which were trans-
fected with three plasmids, were analyzed for caspase-3 en-
zyme activity by ELISA. The values obtained for cells tans-
fected with p1.5RZ107, pRZ107 and pU6RZ107 compare to
the controls, transfected with pCDNA3.1 (lane 1) (the value
represents 100%). Cells transfected with pU6Rz107 showed a
significant decrease in caspase-3 activity (32%) compared
with the cells transfected with other plasmids.

transfected with three plasmids, were analyzed for caspase-3 enzyme activity by ELISA. In con-
trast, cells transfected with pU6Rz107 showed a significant decrease in caspase-3 activity com-
pared with the cells transfected with other plasmids.

3 Discussion

It is believed that apoptosis results from the proteolysis of various cellular components initi-
ated by activated caspases. Because at least 14 caspases have been identified in mammals[10,11], the
precise contribution of individual caspases in this process and how they functionally relate to each
other in vitro and in vivo becomes a key question. Especially, caspase-3 is a frequently activated
death protease. However, the specific requirements of this caspase in apoptosis are until now
largely unknown. Since specific inhibitors for caspase-3 are not available currently, strategies to
assess the exact role of caspase-3 in apoptosis have to depend on target gene manipulating tech-
niques to downregulate its expression, for example, by caspase-3 gene knockout or by using an-
tisense or ribozyme against caspase-3. However, anti-caspase-3 ribozymes that we used in this
study is not only far less expensive than the gene knockout techniques, but also, the ribozymes,
theoretically, will overcome the shortcomings that the gene knockout technique will meet, the
mechanism of compensation. That is to say, because gene knockout is finished during the period
of embryo, other caspases will probably compensate for the function of caspase-3, which has been
abolished. Moreover, though ribozymes share certain characteristics with antisense oligodeoxynu-
cleotides, which are used for the same purposes as ribozyme, because of their catalytic power,
ribozymes are expected to be more efficient than antisense oligodeoxynucleotides[3]. In this study,
 626 SCIENCE IN CHINA (Series C) Vol. 44

we designed three hammerhead ribozymes against rat caspase-3 mRNA, and the results show that
they could cleave target RNA specifically in vitro, and ribozyme cloned into the self-cleaving
vector has the highest cleavage efficiency. The in vitro transcription effect of Rz107 was satisfac-
tory. 5'-cis-Rz and 3'-cis-Rz cut themselves nearly completely and released the intended ribozyme.
The ribozyme’s flanking sequences could be shortened and the effect on ribozyme structure in-
duced by the secondary structure of long flanking sequences would be eliminated. That would
affect the ribozyme’s turnover ratio and/or binding activity as the result of more accurate hybridi-
zation and better cleavage. However, the cleavage efficiency of the ribozyme is limited because
the transcripts, which lack the protective construction in the 5' and 3' ends, is easy to degrade in
vivo.
The transcript of pRZ107 is longer than that of p1.5RZ107. It would form inappropriate con-
formation and affect its hybridization with the target RNA, and then will affect the cleavage effi-
ciency in vitro. While in vivo, because the transcript has some protective construction of 5′cap-
ping and 3′Poly A, it will protect the transcript products from degradation of RNase in vivo. So,
its cleavage efficiency is higher than p1.5RZ107 in vitro. However, the transcript efficiency of
RNA Poly II is limited and the transcript products are at least more than several hundred nucleo-
tides. Therefore the ribozyme is embedded into the whole transcript, and it will affect its appropri-
ate folding and affect its cleavage efficiency in vivo.
Though the cleavage efficiency in vitro of pU6RZ107 is the lowest, because the transcript is
longer than others, in vivo however, it has the perfect cleavage efficiency.
It has become quite clear that the design for obtaining effective ribozymes can still be im-
proved and several important parameters can be optimized. Not only must ribozymes be cloned
under efficient promoters ensuring high levels of expression, but in addition, the RNA context in
which the ribozymes are embedded should provide stability and appropriate conformation for
catalytic activity. Moreover, the cellular compartmentalization of the chimeric ribozyme seems an
essential feature[5, 6]. With the ribozyme sequence embedded into human snRNA, the stability of
the transcript can be further improved by modification of the 3' and 5' terminus of the transcript.
Meanwhile, a disadvantage of Pol II promoters is their requirement for long coding sequences, at
least, a few hundred nucleotides, for efficient transcription. As a result, the ribozyme sequence is
embedded into a long RNA molecule, which can interfere with ribozyme conformation. In contrast,
Pol III promoters, which naturally drive tRNA and snRNA synthesis, are good for very-high-level
non-tissue-specific expression of short RNA. For producing short transcripts polymerase III pro-
moters offer several advantages. The U6 snRNA promoter in our ribozyme constructs gave rise to
105—106 transcripts per transfected cell, in accord with a recent independent report[12]. The addi-
tional U6 snRNA-derived sequences present in the ribozyme RNAs did not apparently interfere
with productive annealing to target, as demonstrated by their efficacy in this study.
The human snRNA U6 promoter is superior to the tRNA promoter. For example, the U6
 No. 6 ACTIVITY IDENTIFICATION OF RIBOZYME in vitro & in vivo 627

promoter does not need the sequence of Box B, except for the 5′
element and 3′end signing. So,
it is possible to replace the other elements with some purpose sequence that we need. tRNA pol III
promoter can insert 100 bp sequences, while U6 RNA can insert at least 400 bp sequence.
In this study, the anti-caspase-3 ribozyme we designed, embedded into U6 RNA, has perfect
cleavage activity in vivo. It would be a desirable alternative approach to the research into the
mechanism of apoptotic pathway in vivo, and also, would be a more attractive tool for human gene
therapy application.
Acknowledgements This work was supported by the High School Science Foundation of Shanghai (Grant No. 98QN60)
and the Chinese Academy of Sciences (Grant No. KSCX2-2-04).

References
1. Andrew, F., Gerard, E., A license to kill, Cell, 1996, 85: 781—784.
2. Thornberry, N. A., Lazebnik, Y., Caspases: Enemies within, Science, 1998, 281: 1312—1316.
3. Kijima, H., Ishida, H., Ohkawa, T. et al., Therapeutic application of ribozymes, Pharmacol. Ther., 1995, 68: 247—264.
4. Phylactou, L. A., Kilpatrick, M. W., Wood, M. J., Ribozymes as therapeutic tools for genetic disease, Hum. Mol. Genet.,
1998, 7(10): 1649—1653.
5. Bettrand, E., Pictet, R ., Grange, T., Can heamerhead ribozymes be efficient tools inactivate gene function? Nucleic Acids
Res., 1994, 22: 293—300.
6. Lieber, A., Strauss, M., Selection of efficient cleavage sites in target RNAs by using a ribozyme expression library, Mol.
Cell Biol., 1995, 15: 540—551.
7. Xu, R. H., Zhou, X. Q., Xie, Q. et al., Preparation and identification of hammerhead ribozyme in vitro against rat cas-
pase-3 mRNA fragment, Chin. J. Hepatol., 2000，8: 361—363.
8. Liu, J., Jin, Y. X., Wang, D. B., A novel vector for abundant expression of antisense RNA, triplex-forming RNA and ri-
bozyme in vivo, High Technology Letters, 2000, 6: 84—88.
9. Sambrook, J., Fritsch, E. F., Maniatis, T., Molecular Cloning: A Laboratory Manual, 2nd ed., New York: Cold Spring
Harbor Laboratory Press, 1989.
10. Porter, A. G., Jänicke, R. U., Emerging roles of caspase-3 in apoptosis, Cell Death Differ, 1999, 6: 99—104.
11. Cryns, V., Yuan, J., Proteases to die for, Genes Dev., 1998, 12: 1551—1570.
12. Narendra, K. V., Anikumar, R. K., Fritz, E., Recent developments in the hammerhead ribozyme field, Nucleic Acids Re-
search, 1998, 26: 5237—5242.