Báo cáo khoa học: Binase cleaves cellular noncoding RNAs and affects coding mRNAs potx

Binase lowers the amount of RNA in the cellsBinase lowered cell viability and increased the numbers of apoptotic cells in lines FDC-P1iR1171 and HE-KhSK4, but did not affect cells from l

coding mRNAs

Vladimir A Mitkevich1, Nickolai A Tchurikov1, Pavel V Zelenikhin2, Irina Yu Petrushanko1,

Alexander A Makarov1and Olga N Ilinskaya1,2

1 Engelhardt Institute of Molecular Biology, Russian Academy of Sciences, Moscow, Russia

2 Department of Microbiology, Kazan State University, Kazan, Russia

Introduction

RNA is an active player in oncogenesis [1–3] One of

the signs of neoplastic transformation is the enhanced

accumulation of rRNA and tRNA and broadly altered

expression of microRNAs (miRNAs), all of which are

categorized as noncoding RNA RNases therefore

possess therapeutic possibilities for cancer treatment,

as RNA damage caused by RNases could be an

alter-native to standard DNA-damaging chemotherapeutics

[4–8] The ribonucleolytic activity of exogenously applied RNases is essential for their cytotoxicity [4,5,9] Degradation of tRNA by onconase (Rana pipi-ens RNase) [10,11], 28S rRNA by Aspergillus RNase a-sarcin [12] and 16S rRNA by colicin E3 (Escherichia coli RNase) [13] have been shown Bacillus amylolique-faciens RNase (barnase) cleaves a wide range of noncoding tRNAs and rRNAs, thus significantly

Keywords

apoptosis; cellular RNA degradation;

cytotoxicity; gene expression; RNase

Correspondence

A A Makarov or O N Ilinskaya, Engelhardt

Institute of Molecular Biology, Russian

Academy of Sciences, Moscow, Russia

Fax: +7 499 1351405

Tel: +7 499 1354095

E-mail: aamakarov@eimb.ru;

olga.ilinskaya@ksu.ru

(Received 28 July 2009, revised 8 October

2009, accepted 2 November 2009)

doi:10.1111/j.1742-4658.2009.07471.x

Bacterial RNases are promising tools for the development of anticancer drugs Neoplastic transformation leads to enhanced accumulation of rRNA and tRNA, and altered expression of regulatory noncoding RNAs Cleav-age of RNA in cancer cells is the main reason for the cytotoxic effects of exogenic RNases We have shown that binase, a cytotoxic ribonuclease from Bacillus intermedius, affects the total amount of intracellular RNA and the expression of proapoptotic and antiapoptotic mRNAs For four cell lines, we visualized cellular RNA by fluorescence microscopy, and determined RNA levels, viability and apoptosis by flow cytometry We found that the level of cellular RNA was decreased in cells that were sensi-tive to the cytotoxic effects of binase The RNA level was lowered by 44%

in HEK cells transfected with the hSK4 gene of the Ca2+-activated potassium channels (HEKhSK4) and by 20% in kit-transformed myeloid progenitor FDC-P1iR1171 cells The most significant decrease in RNA levels was registered in the subpopulations of apoptotic cells However, the binase-induced RNA decrease did not correlate with apoptosis Kit-transformed cells with binase-induced RNA decrease retained viability if the interleukin-dependent proliferation pathway was activated Using quan-titative RT-PCR with RNA samples isolated from the binase-treated HE-KhSK4 cells, we found that the amount of mRNA of the antiapoptotic bcl-2 gene in vivo was reduced about two-fold In contrast, expression of the proapoptotic genes p53 and hSK4 was increased 1.5-fold and 4.3-fold, respectively These results show that binase is a regulator of RNA-depen-dent processes of cell proliferation and apoptosis

Abbreviations

IL, interleukin; miRNA, microRNA; PIC, Protease Inhibitor Cocktail; RT, reverse transcriptase.

decreasing the total amount of intracellular RNA in

ovarian carcinoma SKOV-3 cells [14] Zhao et al [15]

have shown that one of the targets of onconase is

short interfering RNA, probably within the

RNA-induced silencing complex

Recent data on the cytotoxic effects of microbial

RNases, such as Bacillus intermedius RNase (binase)

[16–18], barnase [14], Streptomyces aureofaciens RNase

Sa3 [19] and the 5K cationic mutant of RNase Sa

[18,20,21], suggest that RNases of the T1 superfamily

have promise as a basis for developing new antitumor

drugs

Binase is a highly cationic guanyl-specific RNase

that catalyzes RNA cleavage without the need for

metal ions and cofactors Cloning and sequencing of

the binase gene have been reported [22], and a

three-dimensional structure at 1.65 A˚ resolution was

deter-mined by Polyakov et al [23] We have shown that

binase selectively inhibits the growth of cells,

express-ing ras [16] and kit [17] oncogenes

In this work, we investigated what effects binase has

on the total amount of intracellular RNA and on the

expression of proapoptotic and antiapoptotic mRNAs

We provide evidence that the level of noncoding

cellu-lar RNA decreases in cells that are sensitive to the

cytotoxic effects of binase This effect is most easily

observed in a subpopulation of apoptotic cells

However, the binase-induced RNA decrease does not

correlate with apoptosis The viability of the

kit-trans-formed FDC-P1iR1171 cells, whose RNA level is

decreased by binase, does not change if the interleukin

(IL)-dependent proliferation pathway is activated We have found in vivo that binase affects the quantity of proapoptotic and antiapoptotic mRNAs, as expression

of the p53 and hSK4 genes was increased and expres-sion of the bcl-2 gene was reduced

Results

Visualization of the cellular RNA The most intensive fluorescence in IHKE cells, after RNA dye treatment, was found in the prenuclear part

of the endoplasmic reticulum (Fig 1Aa–c) Among IHKE cells treated with binase and those not treated with binase, there were no individual cells that signifi-cantly differed from the others by their fluorescence However, cells treated with binase tended to show a reduction in the radius of the fluorescing zone around the nucleus (Fig 1B), thus demonstrating a decrease in the amount of RNA in this zone

Apoptotic cells from FDC-P1iR1171 culture, visu-ally selected on the basis of characteristic morphology and high granulation had a higher fluorescence inten-sity than viable cells in samples both treated and untreated with binase (Fig 1C) At the same time, among the apoptotic cells, individuals with practically

no RNA were detected (Fig 1Ca) The amount of such cells in the total population did not exceed 5%

In order to obtain quantitative data on the RNA level

in the cells before and after binase treatment, flow cytometry was used

C a

a

b

Fig 1 Visualization of cellular RNA by

stain-ing with SYTO RNASelect (A) RNA in IHKE

cells: (a) without staining; (b) merged image;

(c) with staining (B) IHKE cells untreated

(a) and treated (b) with binase, (C) RNA in

binase-treated FDC-P1iR1171 cells: (a)

without staining; (b) merged image; (c) with

staining White arrows (Ca) show individual

apoptotic cells without RNA Cells were

treated with 40 l M binase for 48 h.

Binase lowers the amount of RNA in the cells

Binase lowered cell viability and increased the numbers

of apoptotic cells in lines FDC-P1iR1171 and

HE-KhSK4, but did not affect cells from line FDC-P1

(Table 1) The apoptosis-inducing effect of binase was

accompanied by a lowered level of RNA (Table 1)

The decline in RNA levels was not proportional to the

intensity of the cytotoxic effect: in the cells most

sensi-tive to binase, from line FDC-P1iR1171, the total drop

in the amount of RNA after binase treatment was up

to 20% in comparison with untreated cells, whereas in

line HEKhSK4 it was up to 44% (Table 1) The most

significant decrease in RNA levels in both cell lines

was seen in the subpopulation of apoptotic cells

(Fig 2) Also, binase reduced the proportions of

HEKhSK4 and FDC-P1iR1171 cells with high

fluo-rescence intensities (Fig 3), whereas for the FDC-P1

cells this effect was not shown

Binase increases the poly(I) hydrolysis rate in

cells

The ribonucleolytic activity of the nuclear and

cyto-solic fractions of FDC-P1 and FDC-P1iR1171 cells

was measured with a poly(I) substrate stable to pyrimi-dine-specific eukaryotic RNases [24], in order to deter-mine the activity of the guanyl-specific binase after its penetration into the cells Without binase treatment, the rate of hydrolysis of poly(I) in these fractions was less than 0.2–0.4 relative units per cell, whereas treat-ment with binase led to similar effects for both cell lines: within 24 h after addition of binase, the hydro-lysis rate in both the nuclear and the cytosolic fractions increased two-fold to six-fold (Fig 4A,D) The differ-ence in the rate of hydrolysis of poly(I) between the treated and native cells remained the same for 48 h (Fig 4B,E), and then disappeared after 72 h of enzyme action (Fig 4C,F)

IL removes the apoptotic, but not the RNA-decreasing, effect of binase When added to the cultivation medium of FDC-P1iR1171 cells, IL led to an increase in the mean RNA level in the population by 20% (Fig 5, +IL col-umns) In cells grown on medium without IL, in 72 h binase increased the number of apoptotic cells by 30% and lowered the viability of the culture by 70%, while decreasing the total amount of RNA by only 13% as

Table 1 Viability, apoptosis and RNA content of binase-treated cells (48 h, 40 l M ) The viability of control cells grown without RNases was set to 100% The total amount of binase-treated cells was set as 100% The amounts of apoptotic cells in untreated cultures were 5 ± 3% for HEKhSK4 cells and 18 ± 7% for FDC-P1 and FDCP1iR1171 cells The RNA levels of cells grown without RNases were set as equal to 100%.

Cell line

Viability of binase-treated cells (%)

Apoptotic cells (%)

RNA content of binase-treated cells (%) Early apoptotic cells Late apoptotic cells

a According to [17].

Fig 2 Binase lowers the amount of RNA in the cells Intracellular RNA contents in sub-populations of viable (A) and apoptotic (B) HEKhSK4 and FDC-P1iR1171 cells untreated and treated with binase (+Bi) (48 h, 40 l M ), detected by staining with SYTO RNASelect.

compared with cells untreated with binase (Fig 5,

)IL+Bi and )IL columns) It was found that IL did

not decrease the catalytic activity of binase in vitro

(data not shown) In the presence of IL, binase did not

affect the viability of the FDC-P1iR1171 cells and

their transition into apoptosis, despite the fact that the

amount of RNA in them decreased by 23% as

com-pared with those grown on medium without binase

(Fig 5, +IL+Bi and +IL columns)

Hydrolysis of the hSK4 mRNA by RNases in vitro

Figure 6 shows patterns of the RNA fragments

observed after the digestion of a single-stranded and a

double-stranded hSK4 mRNA sample with binase, the

nontoxic RNase Sa, and its highly cytotoxic cationic

mutant 5K RNase Sa [20], used for comparison All of

the RNases actively digested the ssRNA, and the

full-length RNA product was not observed: RNase Sa and

5K RNase Sa generated RNA fragments in the range

from 20 to 200 nucleotides, whereas binase digested

the same sample more deeply into very short

frag-ments, with only a small amount being found in the

19–31 nucleotide range (Fig 6A,B), and the 32

P-labeled dsRNA was only partially cleaved by excessive

amounts of RNase Sa, 5K RNase Sa, and binase The

distribution of the dsRNA hydrolysis products was

shifted towards longer fragments (Fig 6A,B) Binase

was clearly much less active on the dsRNA substrate

than RNase Sa and 5K RNase Sa The patterns of

RNA fragments generated from dsRNA by RNase Sa

and 5K RNase Sa are practically the same About

15% of the labeled products from ssRNA and 6%

from dsRNA were found in the 19–35 nucleotide

region for both 5K RNase Sa and RNase Sa As for binase, less than 1% of the label was revealed in this region for both ssRNA and dsRNA samples

Binase affects the quantity of proapoptotic and antiapoptotic mRNAs in vivo

To test whether the expression of some genes that are involved in the control of apoptosis is affected by bin-ase, we selected three genes – the proapoptotic p53 tumor suppressor [25], the antiapoptotic bcl-2 [26], and the ion channel gene hSK4, with a dual function in apoptosis [27] Using quantitative RT-PCR on RNA samples isolated from the HEKhSK4 cells treated by binase, we observed that, after 24 h of binase treat-ment, the amount of bcl-2 mRNA was slightly reduced, and that after 48 h it had dropped about two-fold (Fig 7) In contrast, the p53 steady-state

Fig 3 Binase reduces the proportions of HEKhSK4 and

FDC-P1iR1171 cells with high RNA contents The ratio of the

subpopula-tions with low (blank sectors) and high (filled sectors) RNA content

in HEKhSK4, FDC-P1 and FDC-P1iR1171 cells untreated and treated

with binase (+Bi) (72 h, 40 l M ) Subpopulations were selected

according to the cytometric distribution of fluorescence intensity.

A

B

C

D

E

F

Fig 4 Binase increases the poly(I) hydrolysis rate in cells Cleav-age of poly(I) by nuclear and cytosolic fractions of FDC-P1 cells (A–C) and FDC-P1iR1171 cells (D–F) untreated (blank columns) and treated (filled columns) with 40 l M binase for 24 h (A, D), 48 h (B, E), and 72 h (C, F).

expression was maintained during the first 24 h of

treatment, and increased 1.5-fold after 48 h (Fig 7)

Expression of the hSK4 gene increased about two-fold

after 24 h of treatment with binase, and then rose

more than four-fold as compared with the steady-state

expression These data showed that binase does not

simply lead to the degradation of some RNAs, but

affects the regulation of gene transcription

Discussion

The binase-induced RNA decrease does not

correlate with apoptosis

There are no doubts that cytotoxic RNases affect

cel-lular RNA They cleave a wide range of substrates:

tRNA by onconase [10,11], 28S rRNA by a-sarcin

[12], 16S rRNA by colicin E3 [13], and tRNA and all

types of rRNA by barnase [11] It is assumed that the

cytotoxicity of the onconase is associated with its

abil-ity to cleave miRNAs [15] and to degrade RNA to

form products similar to short interfering RNAs [28]

Binase induces apoptosis in HEK and HEKhSK4

embryonic kidney cells [18], K562 myelogenous

leuke-mia cells, FDC-P1iR1171 transgenic myeloid progeni-tor cells expressing activated kit oncogene [17], and NIH3T3 ras-expressing fibroblasts [16], but does not affect normal fibroblasts [16] and normal FDC-P1 myeloid progenitor cells [17] The lack of the specific target (kit) for binase in the latter cells does not induce their death even after internalization of the enzyme, as shown by the increase in the rate of hydrolysis of poly(I) after 24 h of binase treatment (Fig 4A,D) This increase precedes the binase-induced apoptosis, which develops within 48 h after treatment only in kit-transformed cells (Table 1) and disappears in 72 h, probably due to the action of intracellular proteases (Fig 4C,F)

Here, we have demonstrated that the apoptosis-inducing action of binase on FDC-P1iR1171 and HE-KhSK4 cells is accompanied by a decrease in the total amount of RNA (Table 1) and in the number of cells with a high RNA content (Fig 3), whereas in the case

of FDC-P1 cells, which are insensitive to binase, these effects are not observed However, even though the decrease in RNA in HEKhSK4 cells is more than two times greater than in FDC-P1iR1171 cells, the latter are more susceptible to apoptosis (Table 1) Also, onconase induces apoptosis in mitogen-stimulated lymphocytes, but does not affect the total RNA content [29] Thus, the reduction in the RNA level is not directly connected with the induction of the apop-tosis process by the RNase Additional evidence for this is the retention of viability of binase-treated FDC-P1iR1171 cells in the presence of IL, despite the fact that the RNA content in these cells is 23% lower than without the binase (Fig 5) Addition of IL leads to the activation of its own antiapoptosis and proliferation pathways, regardless of the Kit-dependent signaling pathways [30], thus canceling the binase-induced growth depression of the kit-transformed cells There-fore, RNA breakdown by itself does not cause cell death

The effect of binase-induced reduction of the RNA content is greatest in the subpopulation of apoptotic cells

Under binase action, the total RNA content in the apoptotic subpopulation of HEKhSK4 cells decreases much more severely than in the viable subpopulation (Fig 2) Among FDC-P1iR1171 cells, binase lowers the RNA levels only in apoptotic cells Apoptotic FDC-P1iR1171 cells demonstrated higher fluorescence levels of the RNA-bound dye than viable cells, even though, physiologically, they could not have had greater RNA amounts (Fig 1Cb) It is possible that

Fig 5 IL abolishes the apoptotic, but not the RNA-decreasing,

effect of binase Viability (blank columns), the amount of apoptotic

cells (black columns) and cellular RNA content (shaded columns)

for FDC-P1iR1171 cells untreated and treated with 40 l M binase

(+Bi) for 72 h in the presence (+IL) or absence ( )IL) of IL The

via-bility and cellular RNA contents of cells grown without binase and

IL were taken as 100% The amount of apoptotic cells is expressed

as a proportion of the total number of cells of each type.

the high fluorescence levels are caused by the dye

bind-ing to the rRNA, which is freed from the proteins, as

their expression in apoptotic cells stops The possibility

that the increase in fluorescence intensity was caused

by the emergence of additional dye-binding sites on

the RNA during hydrolysis was ruled out, because

RNA fluorescence during experiments in vitro with the

SYTO RNASelect dye steadily decreased with time

under binase action (data not shown) Even though

the number of apoptotic cells increases under binase

action (Table 1), their fluorescence decreases (Fig 2B),

and some of them are not even stained by SYTO

RNASelect (Fig 1Ca, arrows) These cells are

proba-bly in the late stages of apoptosis, and do not contain

macromolecular RNA Such cells are also found in

insignificant numbers in the populations without

binase treatment The death of cell lines that are

sensi-tive to binase treatment is more significant for the

FDC-P1iR1171 line, in which only 3% of late

apopto-tic cells are left, than for the HEKhSK4 line, where about 10% of such cells remain (Table 1) Even though, under binase action, the level of RNA in apoptotic cells of the HEKhSK4 line dropped more severely than in cells from line FDC-P1iR1171 (Fig 2B), they remained in the culture for much longer periods of time This once more confirms that there is

no straight correlation between lowered RNA levels and cell death

Noncoding RNAs are obligatory substrates for binase

Considering the significant drop in the RNA level of transformed cells (20–44%; Table 1), it can be con-cluded that binase uses noncoding RNA as a substrate,

as only approximately 5% of the genome output con-sists of protein-coding mRNAs [31] This is in agreement with information about the massive cleavage of tRNA

A

B

Fig 6 Digestions of the hSK432P-labeled

transcripts by RNases (A) Cleavage

prod-ucts of the hSK4 mRNA fragment digested

in vitro with 5K RNase Sa, RNase Sa, and

binase The original ssRNA and the same

RNA after annealing with the excess of

unlabeled antisense RNA were digested

with the RNases The samples were

sepa-rated in a sequencing gel, and

autoradio-graphed using X-ray film Lane 1: untreated

RNA Lanes 2–4: products of ssRNA

diges-tion with 5K RNase Sa, RNase Sa, and

bin-ase, respectively Lanes 5–7: products of

dsRNA digestion with 5K RNase Sa, RNase

Sa, and binase, respectively (B) Enlarged

fragment of the same gel M: RNA marker,

with lengths of fragments in nucleotides.

The brackets show the RNA fragments

corresponding in size (19–31 nucleotides) to

small RNAs involved in RNA

inteference-related regulation mechanisms.

and rRNA in SKOV-3 cells by a close analog of binase,

i.e barnase, which was visualized using gel

electrophore-sis in polyacrylamide gel [14] In this work, it was shown

that the relative abundance of tRNA and 5.8S, 5S, 18S

and 28S rRNA after barnase treatment was significantly

decreased As the two RNases are very highly

homolo-gous and their biochemical properties are almost

identi-cal [32], one can assume that binase should have

practically the same effect on RNA as barnase and

should cleave both tRNA and rRNA in cells

The decrease of the RNA content is caused by the

cat-alytic effect of the enzyme in the cells Binase treatment

of FDC-P1iR1171 and FDC-P1 myeloid progenitor cells

leads, within 24 h, to a significant increase of the RNase

hydrolytic activity on poly(I), in both the cytosolic and

the nuclear fractions (Fig 4A) These results agree with

the data on the decrease in the radius of the fluorescent

zone formed by RNA staining in IHKE cells (Fig 1B)

After 48 h, when the poly(I) hydrolysis rate returns to

the background level (Fig 4B,C), the apoptotic action

of binase becomes apparent (Table 1)

Features of the RNA hydrolysis process in vitro

do not allow toxic and nontoxic RNases to be

distinguished

The effects on the dsRNA are especially significant in

the case of antitumor activities of RNases [6]

Recently, the effect of onconase on the dsRNA was demonstrated [33] We have shown that the single-strand-preferring RNase binase [23] cleaves the dsRNA (Fig 6) Degradation of the dsRNA by binase occurs through the same mechanism as described for RNase

Sa and a number of other single-strand-preferring RNases [34,35]

The patterns of the RNA hydrolysis products obtained with microbial RNases under investigation (Fig 6) do not allow us to elucidate any special fea-tures relating to the cytotoxicity of the RNases Thus, the nontoxic RNase Sa and its cytotoxic 5K mutant [20] have the same effect on the dsRNA (Fig 6), whereas binase, which is less active towards the dsRNA, is cytotoxic [20] Anionic RNase Sa is non-toxic [20], whereas the cationic RNases – binase and 5K RNase Sa, which have essentially different patterns

of RNA hydrolysis products – are both cytotoxic [20]

Binase affects gene expression

We observed upregulation of the KCachannels and the apoptotic p53 genes and downregulation of the antia-poptotic bcl-2 gene under binase action (Fig 7) These changes are typical for apoptosis [36], and indicate that the mRNA of antiapoptotic genes can, to some extent, contribute to the overall binase-induced decrease in the cellular RNA levels Activation of p53 mRNA synthe-sis is not in agreement with the overall decrease in the RNA levels, and thus indicated a regulatory role for binase Besides this, activation of expression of the

KCachannel gene, which is important for shrinkage of the apoptotic cell, develops only at the mRNA level

We have determined that binase, like 5K RNase Sa, depresses the functional activity of the KCa channels [18] This indicates that an exogenous binase, after penetrating into the cells, can affect the regulatory mechanisms of RNA-dependent processes of transcrip-tion and translatranscrip-tion

Degradation of the available RNA by exogenous RNases puts the cellular RNA levels out of balance and disrupts the regulatory processes, thus leading to the induction of apoptosis It is possible to assume that the effect on gene expression will take place if the length of the duplex fragments generated by bacterial RNases corresponds to that of small, noncoding RNAs involved in RNA regulation mechanisms The possibility of such duplex formation was shown in vitro

by experiments on digestion of the hSK4 gene mRNA

by the RNases used in this study (Fig 6) Similarly, it has been demonstrated that onconase cleaves tRNA

in vitro, yielding fragments 5–40 nucleotides in length [11] The possibility also cannot be excluded that

Fig 7 Binase affects the quantity of proapoptotic and antiapoptotic

mRNAs The mRNA expression levels of genes hSK4, bcl-2 and

p53 in HEKhSK4 cells untreated and treated with 15 l M binase for

24 h and 48 h.

RNases may degrade small cellular RNAs, which are

repressors of gene translation This deserved special

attention, as it is known that some miRNAs from the

miR17–92 cluster are amplified and overexpressed in

human cancers, thus blocking the translation of

proa-poptotic genes [37] Selective degradation of these

miRNAs could potentially prevent the repression of

apoptosis, and therefore be exploited for anticancer

chemotherapy [3] A similar mechanism was suggested

for onconase [15] Recently, it has been reported that

small RNA duplexes generated by hydrolysis with

E coli RNase III mediate effective RNA interference

in mammalian cells [38] Thus, by targeting different

regulatory RNAs with cytotoxic RNases, it might be

possible to regulate the expression of certain genes

Conclusions

The choice between life and death of a cell subjected

to exogenous RNases depends on the characteristic

pattern of hydrolysis products of the cellular RNA,

which reflects the results of complicated interactions

between molecular determinants of RNases on one

side [5,7] and the cellular targets on the other

[16,17,20] The results of this investigation show that

binase reduces the amount of RNA in sensitive cells,

but this decrease by itself is not fatal, as it is the

dis-ruption of RNA-dependent regulatory processes that

causes cell death The development of an approach

based on information about the RNA regulation

network and the creation of more selective cationic

microbial RNases targeting specific RNAs could be

promising

Experimental procedures

Enzymes

Binase (12.3 kDa) was isolated from the culture fluid of

E coli BL21 carrying plasmid pGEMGX1⁄ ent ⁄ Bi as a

homogeneous protein The enzyme purification was carried

out by a procedure described in [32] Binase was assayed

for catalytic activity towards synthetic substrates [32] and

yeast RNA [39] The cells were treated with binase at a

concentration of 40 lm, except for those used for RT-PCR

analysis (see below), which were treated with 15 lm binase,

because when they were treated with 40 lm binase it was

impossible to extract enough of the cellular RNA for

analy-sis RNase Sa and 5K RNase Sa (D1K, D17K, E41K,

D25K, and E74K) [40,41] were gifts from J M Scholtz

and C N Pace (Texas A&M University System Health

Science Center, College Station, TX, USA)

Cell cultures

Four different cell lines were used Human embryonic kid-ney cells transfected with the DNA encoding the human small conductance Ca2+-activated K+ channel type hSK4 (HEKhSK4) and immortalized human kidney epithelial embryo cells (IHKE) were obtained from the Rudolf-Buchheim Institute of Pharmacology (Giessen, Germany) Normal myeloid progenitor cells (FDC-P1) and transgenic myeloid progenitor cells expressing the activated kit onco-gene (FDC-P1iR1171) were obtained from the Heinrich-Pette Institute of Experimental Virology and Immunology (Hamburg, Germany) The IHKE cells were cultured according to [42], and the HEKhSK4, P1 and FDC-P1iR1171 cells were cultured as described previously [17,18]

RNA visualization by fluorescence microscopy

RNA in the IHKE and FDC-P1iR1171 cells was stained with 2.5 lm SYTO RNASelect (Molecular Probes, Eugene,

OR, USA) for 30 min at 37C The fluorescence intensi-ties of intracellular RNA after staining and washing off

of the extra dye were analyzed using saved images obtained with a Leica DM 6000B fluorescence microscope

fw4000 software

Determination of RNA level, viability and apoptosis by flow cytometry

HEKhSK4, FDC-P1 and FDC-P1iR1171 cells were pre-loaded with 5 lm SYTO RNASelect (excitation at 490 nm; emission at 530 nm) for 30 min at 37C for RNA detec-tion Cell viability was assessed by adding propidium iodide (Molecular Probes) at a final concentration of 10 lgÆmL)1

to the cell suspension for 1–2 min before measurements Apoptosis was verified with annexin V–fluorescein isothio-cyanate (Molecular Probes) [43] and propidium iodide [44] double staining All measurements were performed on a Beckman Coulter Epix XL4 flow cytometer (Fullerton, CA, USA)

In order to show that degradation of the RNA does not lead to an increase in the fluorescence intensity values, the fluorescence of a solution of RNA, stained with SYTO RNASelect, in the presence of binase was measured SYTO RNASelect (0.5 lm) was added to 11 lm yeast RNA (Sigma-Aldrich, St Louis, MO, USA) in Tris buffer (10 mm Tris⁄ HCl, 140 mm NaCl, pH 7.0) After 10 min of RNA staining, 1.8· 10)8m binase was added, and at 530 nm a decrease in the fluorescence intensity by 80% in 15 min was observed on a Cary Eclips fluorimeter (Varian, Palo Alto,

CA, USA)

Extraction of the cytosolic and nuclear fractions

of the cell, and determination of the RNase

activity in these fractions

FDC-P1iR1171 and FDC-P1 cells were washed in ice-cold

NaCl⁄ Pi(Sigma-Aldrich) Then, lysis buffer (10 mm Hepes,

10 mm KCl, 2 mm MgCl2, 0.5% NP-40 in double-distilled

H2O, pH 7.9) supplemented with Protease Inhibitor

Cock-tail (PIC; Roche Diagnostics, Mannheim, Germany) was

added, and the suspension was centrifuged at 1500 g for

5 min at 4C The supernatant was used as the cytosolic

fraction, and the pellet was resuspended in a low-salt

solu-tion (20 mm Hepes, 2.5 lL of 25% glycerol, 2 mm MgCl2,

0.1 mm KCl, 1 mm EDTA, pH 7.9) supplemented with

PIC, and again centrifuged at 1500 g for 5 min at 4C

The pellet was resuspended in Tris buffer (10 mm Tris⁄ HCl,

3 mm MgCl2, 2 mm 2-mercaptoethanol, 5 mm CH3COONa,

0.5% SDS, 0.5 mm CaCl2, PIC, pH 7.6), and treated as

described in [45] in order to obtain the nuclear fraction

Cleavage of poly(I) (Sigma-Aldrich) by the nuclear and

cytosolic fractions was measured as described by Yakovlev

et al [46] RNase activity was expressed in relative units,

which show the increase in absorption at 248 nm in 1 min

divided by the number of cells used to obtain the cytosolic

and nuclear fractions The number of cells was calculated

using a Leica DMI 4000B light microscope

RNA synthesis

For synthesis of the hSK4 32P-labeled sense transcript, a

cDNA clone (pcDNA3; Invitrogen, Carlsbad, CA, USA)

was used The32P-labeled sense and the unlabeled antisense

transcripts were synthesized in vitro using T7 RNA

poly-merase on DNA templates digested with NotI or BamHI,

respectively The reactions were performed in 20 lL of a

solution containing 1 lg of DNA, 40 mm Tris⁄ HCl (pH

7.5), 6 mm MgCl2, 2 mm spermidine, 10 mm NaCl, 10 mm

dithiothreitol, 1 unitÆlL)1 RNasin (where one unit is

defined as the amount of RNasin Ribonuclease inhibitor

required to inhibit the activity of 5 ng of ribonuclease A by

50%; as defined by the manufacturer, Promega, Madison,

WI, USA), ATP, GTP, CTP (500 lm each), 10 lm UTP,

and 1 lm [32P]UTP[aP] (6000 CiÆmmol)1, ‘Phosphor-center’

of the Russian Academy of Sciences, Moscow, Russia), and

20 units of the T7 RNA polymerase The conditions used

favor the synthesis of full-length RNA products

Annealing

To ensure the annealing of full-length molecules, RNA

preparations were gel-purified in 4% acrylamide gel, using a

mirVana miRNA detection kit for elution (Ambion, Austin,

TX, USA) About 0.6 ng of antisense RNA was mixed with

0.03 ng of32P-labeled sense RNA in 10 lL of solution

con-taining 100 mm NaCl, 10 mm Tris⁄ HCl (pH 7.5), 2 mm MgCl2, and 0.5 units of RNasin Annealing was performed for 5 h under mineral oil, using a temperature shift from 50

to 35C Annealing was tested by gel-shift assay, which demonstrated the complete annealing of 32P-labeled sense RNA with a 20-fold excess of unlabeled antisense RNA (data not shown) Self-annealing of32P-labeled ssense RNA under the same conditions was performed as a control

RNase treatment

The RNase digestions were performed in 20 lL-well plates containing 2 lL of annealed samples and 3 lL of annealing mixture Up to 0.1 lg of RNase (5K RNase Sa, RNase Sa, and binase) were added to the wells, and the incubation was performed for 30 min at 37C Then, 7 lL of dyes containing formamide and 2 mm EDTA were added The mixture was maintained at 90C for 3 min, and used for electrophoresis

Fractionation

For fractionation, 0.2-mm-thick denaturing 12% polyacryl-amide gel containing 8.3 m urea was used Two-microliter samples were loaded onto each lane The 32P-labeled RNA markers were synthesized on pGEM1 DNA digested with EcoRI or SmaI After electrophoresis, the gel was washed with 10% acetic acid, dried, scanned using a Cyclone phos-phoimager (Packard Instruments, Meriden, CT, USA), and exposed to X-ray film

RT-PCR analysis

RNA was isolated from about 150· 103

HEKhSK4 cells using the Trizol reagent (Invitrogen) Samples were treated with DNase using a DNA-free kit (Ambion), and approxi-mately 2 lg of RNA Specific primers and Moloney murine leukemia virus reverse transcriptase (RT) (Promega) were used for synthesis of cDNAs corresponding to the hSK4 mRNA, p53 mRNA and bcl-2 mRNA, according to the manufacturer’s instructions Each PCR was performed with the cDNA template (RT+) and the same RNA probe with-out addition of RT The number of PCR cycles varied from

28 to 37 Primers for RT-PCR were selected according to the primer selection tool program (http://biotools umassmed.edu/)

The primers for cDNA synthesis were as follows:

mRNA), 5¢-GGCCCTTCTGTCTTGAACATGAG-3¢ (for

AAA-3¢ (for bcl-2 mRNA) For quantitative PCR with cDNA corresponding to sense transcripts of hSK4, bcl-2, and p53, the following primers were used: 5¢-GAAGCCTG GATGTTCTACAAACATA-3¢ and 5¢-AAGCAGCTCAGT

CAGGGCATCC-3¢, 5¢-CTAATGGTGGCCAACTGGAG

ACT-3¢ and 5¢-GTTTTGTTTATTATACCTTCTTAAGTT

TT-3¢, and 5¢-AGACCGGCGCACAGAGGAAGAGAA-3¢

and 5¢-CTTTTTGGACTTCAGGTGGCTGG-3¢, respectively

Conditions for linear PCR for each set of primers were

selected in the preliminary experiments using Mastercycler

personal (Eppendorf, Hamburg, Germany) The final PCR

products were separated in mixed 1% agarose⁄ 2%

Nu-Sieve agarose gels, and the separation data were evaluated

using quantity one quantitation software (Bio-Rad,

Her-cules, CA, USA) Statistical analysis of the fractionated

DNA fragments obtained in five independent experiments

was performed with origin software The identity of the

amplified DNA fragments was confirmed by sequencing

The RT-PCR data were normalized to ribosomal human

5.8S RNA as follows Two-microliter aliquots from the

final 40 lL cDNA probes synthesized using different

specific primers on total RNA preparations (see above)

were used for new cDNA synthesis with the 5.8S RNA

spe-cific primer 5¢-GCTCAGACAGGCGTAGCCCCGGGA-3¢

One microliter of the sample was then taken for PCR using

5.8S gene-specific primers 5¢-CGGTGGATCACTCGGC

The data from RNA preparations corresponding to

differ-ent constructs were normalized using quantity one

Acknowledgements

We thank M Scholtz and N Pace for presenting us with

RNase Sa and its 5K mutant We thank C Stocking for

donating the FDC-P1 and FDC-P1iR1171 cells We

thank A Koschinski for preparing the IHKE cells This

work was supported by the Molecular and Cellular

Biol-ogy Program of the Russian Academy of Sciences, by

the RFBR (grant 07-04-01051), by the Russian Federal

Programs (contracts 02.512.11.2198, 02.512.12.2014,

2.1.1⁄ 920, and 02.740.11.0391), and by a Grant of the

President of the Russian Federation for young scientists

(MK-162.2009.4)

References

1 Scholzowa E, Malik R, Sevcik J & Kleibl Z (2007)

RNA regulation and cancer development Cancer Lett

246, 12–23

2 Johnson DL & Johnson SA (2008) Cell biology RNA

metabolism and oncogenesis Science 320, 461–462

3 Zhang W, Dahlberg JE & Tam W (2007) MicroRNAs

in tumorigenesis: a primer Am J Pathol 171, 728–738

4 Leland P & Raines R (2001) Cancer chemotherapy:

ribonucleases to the rescue Chem Biol 8, 405–413

5 Makarov AA & Ilinskaya ON (2003) Cytotoxic

ribonucleases: molecular weapons and their targets

FEBS Lett 540, 15–20

6 Libonati M, Gotte G & Vottariello F (2008) A novel biological action acquired by ribonuclease

through oligomerization Curr Pharm Biotechnol 9, 200– 209

7 Makarov AA, Kolchinsky A & Ilinskaya ON (2008) Binase and other microbial RNases as potential anticancer agents BioEssays 30, 781–790

8 Chang C, Newton DL, Rybak SM & Wlodawer A (2002) Crystallographic and functional studies of a modified form of eosinophil-derived neurotoxin (EDN) with novel biological activities J Mol Biol 317, 119–130

9 Kim DH, Behlke MA, Rose SD, Chang MS, Choi S & Rossi JJ (2005) Synthetic dsRNA Dicer substrates enhance RNAi potency and efficacy Nat Biotechnol 23, 222–226

10 Saxena SK, Sirdeshmukh R, Ardelt W, Mikulski SM, Shogen K & Youle RJ (2002) Entry into cells and selec-tive degradation of tRNAs by a cytotoxic member of the RNase A family J Biol Chem 277, 15142–15146

11 Suhasini AN & Sirdeshmukh R (2006) Transfer RNA cleavages by onconase reveal unusual cleavage sites

J Biol Chem 281, 12201–12209

12 Olmo N, Turnay J, Gonzalez de Butitrago G, Lopez de Silanes I, Gavilanes JG & Lizarbe MA (2001) Cytotoxic mechanism of the ribotoxin alpha-sarcin Induction of cell death via apoptosis Eur J Biochem 268, 2113–2123

13 Lancaster LE, Wintermeyer W & Rodnina MV (2007) Colicins and their potential in cancer treatment Blood Cells Mol Dis 38, 15–18

14 Edelweiss E, Balandin TG, Ivanova JL, Lutsenko GV, Leonova OG, Popenko VI, Sapozhnikov AM & Deyev

SM (2008) Barnase as a new therapeutic agent trigger-ing apoptosis in human cancer cells PLoS ONE 3, e2434

15 Zhao H, Ardelt B, Ardelt W, Shogen K & Dar-zynkiewicz Z (2008) The cytotoxic ribonuclease oncon-ase targets RNA interference (siRNA) Cell Cycle 7, 3258–3261

16 Ilinskaya O, Decker K, Koschinski A, Dreyer F & Repp H (2001) Bacillus intermedius ribonuclease as inhibitor of cell proliferation and membrane current Toxicology 156, 101–107

17 Ilinskaya ON, Zelenikhin PV, Petrushanko IY, Mitke-vich VA, Prassolov VS & Makarov AA (2007) Binase induces apoptosis of transformed myeloid cells and does not induce T-cell immune response Biochem Biophys Res Commun 361, 1000–1005

18 Ilinskaya ON, Koschinski A, Repp H, Mitkevich VA, Dreyer F, Scholtz JM, Pace CN & Makarov AA (2008) RNase-induced apoptosis: fate of calcium-activated potassium channels Biochimie 90, 717–725

19 Sevcik J, Urbanikova L, Leland PA & Raines RT (2002) X-Ray structure of two crystalline forms of a