Alterations in the expression of muscle-specific genes mediated by troponin C antisense oligodeoxynucleotide

EXPERIMENTAL

CELL

RESEARCH

192,

227-235

(1991)

Alterations in the Expression of Muscle-Specific Genes Mediated by Troponin C Antisense Oligodeoxynucleotide GOPAL THINAKARAN Department

of Molecular

Biology

and Genetics,

AND JNANANKUR

University

reprint,

requests

should

Guelph,

Ontario,

Canada

NIG

2 WI

oligomer to TnC mRNA may have triggered a complex array of compensatory processes. 0 1991 Academic press, hc.

The effectiveness of an antisense oligodeoxynucleotide to troponin C (TnC) mRNA in blocking expression of TnC in differentiated chicken myotubes was examined. An 18-nucleotide-long sequence common to both fast and slow isoforms of TnC mRNAs was chosen as the target sequence. The oligomer was found to be efficiently taken up by myotubes. However, the intracellular half-life of the oligomer was found to be only 3 h. Results of studies using different concentrations of oligomer for 3 h in the culture medium showed that compared to the untreated control culture, myotubes incubated with 20 pm antisense oligomer showed a 30% reduction in the steady-state level of TnC mRNAs. An increase of incubation period to 12 h with additions of fresh culture medium containing 20 Mm antisense oligomer every 3 h failed to produce any further reduction of TnC mRNA level. Concomitant to the decrease of TnC mRNAs in antisense oligomer-treated cells, the steady-state levels of cr-actin and a-tropomyosin mRNAs were also reduced by approximately 20 to 40%. Analysis of the homology of the sense sequence of this oligomer with that of a-actin and cr-tropomyosin mRNAs suggested that reduction in the level of cu-actin and a-tropomyosin mRNAs was not due to direct hybridization of the antisense oligomer to these mRNAs. Comparison of TnC polypeptide synthesis in untreated and oligomer-treated myotubes showed approximately 70% reduction of fast TnC polypeptide synthesis in antisense oligomer-treated cells. In contrast, slow TnC polypeptide synthesis was not significantly reduced in treated cells. Similarly, cr-actin and Lu-tropomyosin polypeptide synthesis remained close to the level of untreated cells. Furthermore, analysis of transcription of various muscle-specific mRNAs showed increased synthesis of both TnC and cY-tropomyosin mRNAs in antisense oligomer-treated myotubes. On the other hand, synthesis of actin mRNAs was not altered by this treatment. These results showed that antisense oligomer was effective in significantly reducing TnC polypeptide synthesis in chicken myotubes. Furthermore, these results suggest that treatment of myotubes with antisense

’ To whom

of Guelph,

BAG’

INTRODUCTION During terminal differentiation of muscle cells, various muscle-specific proteins are synthesized and assembledinto the functional contractile apparatus. Regulation at the level of transcription plays the central role in muscle differentiation by directing the accumulation of contractile protein mRNAs. Furthermore, transcriptions of these genes are coordinately regulated during differentiation [ 1, 21. The contractile proteins are assembled in a defined stoichiometry to form the myofibrils. Mechanisms regulating their quantitative synthesis and accumulation are not fully understood. The outcome of the absence or of the increased rate of degradation of one of the polypeptide complements of the contractile unit is also unknown. One approach to identify and study these mechanisms is to selectively block the expression of one of the proteins forming the myofilament and examine its effect on the other contractile proteins. Intracellular expression of an antisense RNA has been used to accomplish specific inhibition of gene expression (reviewed in [3]). A simpler alternative is the use of antisense oligodeoxynucleotides which are complementary to unique mRNA sequences (for reviews see 14-61). The mechanism of oligodeoxynucleotide transport through a specific receptor into living cells has been characterised [7, 81. Once inside the cell, the oligodeoxynucleotide can form an RNA:DNA duplex with the target mRNA and may interfere with the process of translation. In addition to this mechanism the targeted mRNA present in the form of an RNA:DNA duplex may also be cleaved by RNase H [9]. Introduction of antisense oligodeoxynucleotides into a variety of cells has been shown to specifically inhibit viral gene expression [lo-121, cellular proliferation [ 13-161, neurite outgrowth [ 17, 181, T-cell receptor expression [ 191, and lymphokine biosynthesis [20]. Recently, antisense oligomer targeted against myogenin was shown to block morphological differen-

be addressed. 227

0014.4827/91

Copyright 0 1991 All rights of reproduction

$3.00

by Academic Press, Inc. in any form reserved.

228

THINAKARAN

tiation and expression of nicotinic acetylcholine receptor in BC3H-1 cells [21]. In addition, antisense RNA has been shown to inhibit the synthesis of myosin heavy chain and a-actinin proteins in Dictyostelium discoideum [22, 231. For this investigation we chose to use antisense oligodeoxynucleotide to selectively block troponin C (TnC) gene expression in differentiated muscle cells. TnC is one of the three polypeptides constituting the troponin complex. The activation of muscle contraction is mediated by the TnC-calcium interaction. The genes for the contractile proteins troponin C, T, and I encoding the calcium regulatory complex are transcriptionally coactivated during differentiation of muscle cells [24]. TnC has two known isoforms, the slow and the fast TnC. The slow or cardiac isoform of TnC is present in both cardiac and skeletal muscle, whereas the fast TnC is present only in the skeletal muscle [25]. Furthermore, it was found that primary cultures of chicken skeletal muscle cells express predominantly the fast TnC mRNA and a small amount of slow TnC mRNA [26; C. Berezowsky and J. Bag, unpublished results]. Therefore, in this study, we have attempted to reduce the level of expression of both isoforms of TnC in primary chicken skeletal muscle cultures by using an antisense oligodeoxynucleotide targeted against a sequence common in both mRNAs. The effect of this inhibition on the gene expression of a few other contractile proteins was also analyzed. Our results suggest that reduction of TnC mRNA level in myotubes could influence the level of mRNAs for other contractile proteins. In addition a compensatory mechanism may operate at the level of transcription to minimize the effect of the antisense oligodeoxynucleotide.

MATERIALS

AND

METHODS

synthesis. The nucleotidesequence of the and fast TnC cDNAs [2’7,28] was analyzed for sequence homology. An 18nucleotide-long sequence common to both fast and slow TnC mRNAs was chosen for use as the target sequence in our studies. A single antisense oligodeoxynucleotide was therefore used to inhibit expression of both slow and fast TnC mRNAs in cultured chicken skeletal myotubes. The oligodeoxynucleotides were custom synthesized by the Midland Certified Reagent Co., Texas, using phosphoramidite chemistry and further purified by anion-exchange HPLC. The sequences of the oligomers were verified using chemical sequencing technique [29]. Cell culture. Primary myoblast cultures were prepared from Day 11 chicken embryos as described previously [30]. Briefly, pectoral and leg muscles, dissected free of skin and connective tissue, were minced and trypsinized (0.25% trypsin) for 15 min. The suspension was then passed through nylon monofilament screen (Tetko Inc., NY) to remove aggregates. The collected single cells were pelleted by mild centrifugation, resuspended in Dulbecco’s modified Eagle’s medium (DMEM) containing 5% horse serum and 5% chick embryo extract (GIBCO), and plated in gelatin-coated tissue culture dishes. Cultures were maintained at 37°C in a humidified atmosphere containing 5% CO,. Twenty-four hours after plating, the medium was replaced by Oligodeoxynucleotide

chickenslow

AND

BAG

DMEM containing 5% horse serum and 2% chick embryo extract. Cytosine arabinoside (Sigma) to 10 nM final concentration was added to the medium after 48 h to eliminate the contaminating fibroblasts. The cells fused routinely after 2 days in culture and more than 70% of nuclei were found in multinucleated cells. All experiments were done 5 days after the culture was initiated. Determination of oligodeorynucleotide uptake and stability. Oligodeoxynucleotides were 5’ end-labeled with [Y-~‘P]ATP (7000 Ci/ mmol; 1 Ci = 37 GBq, ICN) using T, polynucleotide kinase (BRL) and were purified by passage through a Bio-Gel P-4 column (Bio-Rad). The purity of end-labeled oligomer was checked by electrophoresis on a denaturing 20% polyacrylamide gel. Chicken skeletal myotubes derived from plating 1 X lo5 cells/cm’ in 24.well plates (Corning) were used for uptake studies. To each well, 5 X lo6 cpm of ,5’-“‘P-labeled oligodeoxynucleotide mixed with unlabeled oligomer to a final concentration of 5 PM was added in 0.2 ml of DMEM containing 2% heat-inactivated horse serum. A single lot of heat-inactivated horse serum which had minimal nuclease activity, determined by the extent of stability of 5’ end-labeled oligodeoxynucleotide, was used throughout the study. After incubating for 0,3,6,12, or 18 h at 37”C, medium was removed and cells were washed five times with 0.5 ml of phosphate-buffered saline. Cells were then lysed in 100 ~1 of lysis buffer (10 mMTris, pH 7.5,10 mMNaCl,3 mMMgCl,, 0.05% Nonidet-P40, 0.5% sodium dodecyl sulfate (SDS), 0.01% proteinase K) and the lysate was extracted with an equal volume of phenol. The aqueous phase was removed and reextracted with a mixture of phenol:chloroform:isoamyl alcohol (25:24:1). Radioactivity present in the culture medium, cell washes, lysate, and aqueous phase was determined by liquid scintillation counting. The percentage of oligodeoxynucleotide taken up by the cells was calculated as described previously [15]. Briefly, the amount of radioactivity in the aqueous phase was divided by the sum of radioactivity present in the culture medium, washes, and the cell lysate. To determine the oligodeoxynucleotide stability, aliquots of the culture medium and aqueous phases derived from the cell lysate were lyophilized, dissolved in 10 ~1 of 80% deionized formamide, 0.01% xylene cyanole, and 0.01% bromophenol blue, and subjected to electrophoresis on a denaturing 20% polyacrylamide gel. After electrophoresis, gels were fixed, dried, and exposed to X-AR films (Kodak) at -80°C for autoradiography. RNA isolation and Northern blot analysis. Skeletal myotubes, derived from plating 7 X lo5 tells/35-mm dish, were treated with 0, 10, and 20 FM anti-TnC or 20 FM sense-TnC oligomer for 3 h in DMEM containing 2% heat-inactivated horse serum. For 12-h treatments with antisense oligomer, the medium was removed and fresh oligomer containing medium was added every 3 h. Total RNA was isolated by the acid guanidinium thiocyanate method 1311. Polysomal and postpolysomal RNA fractions were prepared as previously described [32]. RNA samples were size fractionated on a 1.5% agarose gel containing 1.1 M formaldehyde and 10 mM sodium phosphate 1331 and transferred to Zetaprobe membrane (Bio-Rad) in 20X SSC (1X SSC = 0.15 M NaCl, 0.015 M sodium citrate, pH 7.4). Following transfer, the membranes were baked at 80°C for 2 h and washed at 60°C for 1 h in 0.1X SSC and 0.1% SDS. Prehybridization and hybridization were carried out at 65°C for 16 h in 0.5 M NaH,PO,, pH 7.2.1 mM EDTA, 7% SDS, and 1% bovine serum albumin. The probes were labeled to l-5 X 10’ cpmlwg DNA using a nick-translation kit (IBI) according to manufacturer’s instructions and added at 2 X lo6 cpm/ml of the hybridization solution. Blots were washed at room temperature in 40 mM NaH,PO,, pH 7.2, 1 mM EDTA, 5% SDS, and 0.5% bovine serum albumin for 30 min followed by two washes in 40 mM NaH,PO,, pH 7.2, 1 mM EDTA, and 1% SDS, first at room temperature for 30 min and then at 65°C for 30 min. Autoradiograms were scanned using a densitometer and the signal intensity was normalized to that obtained with /3- and y-actin. The probes used for hybridization were as follows: TnC probe was a mixture of a 753.bp EcoRI fragment of a MlS-derived vector and a 655.bp HindIII, Sal1 fragment of pMC3B corresponding to the full-length cDNAs of chicken

INHIBITION Slow Troponm 5

OF

TnC

WITH

C

IAUG -+,+

~~

999 9aa GGU GAU GAG GAU GCU GGG gca gas

20

148

Fast Troponin

3

-I--

165

655

C

5’ AK

-i

~

~

999 CaC GGU GAU GAG GAU GCU GGG cm 9aa -

43

177 ANTISENSE

OL.IGOMER

SENSE OLIGOMER

194

3’

pi753

3’ CCA CTA CTC CTA CGA CCC 5‘

5’ GGT GAT GAG GAT GCT GGG 3’

FIG. 1. Complementary region of the slow and fast TnC mRNA corresponding to the antisense oligomer. Initiation codon and the target sequence are shown in capital letters. The nucleotide sequence of the antisense oligomer is also shown in capital letters.

fast and slow TnC mRNAs, respectively [27, 281; probes for cu-actin, tu-tropomyosin, and fl-tubulin were linearized pAC269, pTm5ru, and pT2 plasmids, respectively [34-361. Measurement of transcription. For transcription assay, skeletal myotubes derived from plating 1 X lo5 cells/cm* in 12-well plat,es (Flow) were used. Cells were treated with the indicated oligomers for 12 h as described. RNA transcribed during the final 3 h of incubation was labeled by adding 4 mCi/ml [32P0,3] (ICN) to the culture medium. RNA was isolated as described above. For hybridization, denatured, linearized cDNA (5 fig) was slot blotted onto Zetaprobe membranes using the slot-blot apparatus (BRL) and the membrane was baked at 80°C for 2 h. Prehybridization was performed in 1.5X SSPE (lx SSPE = 0.15 M NaCl, 0.01 M NaH,PO,* H,O, 0.02 M EDTA), 1% SDS, 0.5% bovine serum albumin, 0.2 mg/ml tRNA, and 0.5 mg/ml sheared, denatured salmon sperm DNA at 50°C for 24 h. Hybridization was performed in the same buffer containing, in addition, 0.2 mg/ml tRNA, 0.5 mg/ml salmon sperm DNA, and 1 X 106cpm labeled RNA at 50°C for 40 h. Blots were washed successively for 15 min at room temperature in 2~ SSC, 0.1% SDS, then in 0.5X SSC, 0.1% SDS, and finally in 0.1X SSC, 0.1% SDS. Scanning densitometry was performed as described. Skeletal [““S]Methionine labeling and electrophoresis of proteins. myotubes derived from plating 1 X lo5 cells/cm’ in 24-well plates were treated with the indicated oligomers for 12 h as described. L[?S]Methionine (Amersham) was added at a concentration of 250 &i/ml to label the proteins synthesized during the final 3 h of incubation. Cells were washed with cold phosphate-buffered saline and lysed in lysis buffer [37]. Samples containing equal amounts of radioactivity (2 X lo5 cpm) were subjected to IEF/SDS-PAGE [38] and autofluorography as previously described [39].

ANTISENSE

229

OLIGOMER

labeled oligomer was added to the culture medium. After 0, 3, 6, 12 and 18 h of incubation, the cells were washed and lysed. The percentage of uptake was calculated from the radioactivity present in the aqueous phase as described under Materials and Methods (Fig. 2). The amount of oligomer that had entered the cell was linear with time for up to 6 h. There was a 2.5-fold increase in the uptake between 6 and 12 h and it remained at about 2% for up to 18 h. Denaturing gel electrophoresis of the samples from culture medium revealed that a 50% loss of 32P-labeled oligomer added in the culture medium had occurred within 3 h (Fig. 3A). Incubation of 32P-labeled oligomer with the medium alone, however, did not show a similar loss of 32P-label from the oligomer (Fig. 3C). Incubation in the presence of myotubes was therefore necessary for degradation of the oligomer. Recovery of undegraded oligomer from the cell lysate was maximal after 3 h of incubation with the oligomer and had decreased to onethird of that level by 6 h (Fig. 3B). Very little labeled oligomer was detectable at 12 h in both the culture medium and the cell extract. Oligodeoxynucleotides are readily degraded by 5’ and 3’ exonucleases and also by endonucleases [ 141. The lower band present in the samples from culture medium (Fig. 3A) represents partially degraded products. The stability of oligomers has been found to vary between cell lines. In HL-60 cells oligomers have been found to be stable for up to 3 days [14], whereas in PC12 cells oligomers were stable for only 6-12 h [ 171. In our studies we have not detected significant levels of undergraded labeled oligomer after a 6-h incubation. At this point we are, however, not certain if loss of radioactivity from the oligomer was the result of removal of the 5’ [3”PO;3] from the 5’ end-labeled oligomer by phosphomonoesterase activity [lo]. Presence

RESULTS

Oligodeoxynucleotide Skeletal Myotu,bes

Uptake

and Stability

in Chicken

The sequences of the oligomers used in this study and the target sequence in the TnC mRNAs are shown in Fig. 1. The strategy for choosing the TnC antisense oligomer (anti-TnC) is outlined under Materials and Methods. The efficiency of oligomer internalization and the intracellular stability vary depending on the cell type and culture conditions employed. To study the oligomer uptake by chicken skeletal myotubes, 5’ end-

HOURS

OF INCUBATION

FIG. 2. Uptake of anti-TnC oligodeoxynucleotide by chicken skeletal myotubes. Chicken skeletal myotubes were incubated with 5’-32P-labeled anti-TnC oligomer (5 PM) for 0, 3,6, 12, and 18 h. Cells were lysed and lysates were phenol:chloroform extracted. Percentage of uptake was calculated from the counts recovered in the aqueous phase as described under Materials and Methods. Bars represent &SEM from four different experiments.

230

THINAKARAN 0

3

6

12

18

0.5

3

6

12

18h

a b

mnt-

A

El

C

FIG. 3. Extracellular and intracellular stability of anti-TnC oligomer. Myotubes were incubated with 5’ end-labeled anti-TnC oligomers and samples were processed to extract aqueous phase as described in the legend to Fig. 2. Aliquots of culture medium collected at the end of incubation period (A) and aqueous phases obtained from the cell lysates (B) were analyzed by polyacrylamide gel electrophoresis under denaturing conditions. (C) Anti-TnC was incubated for 3 h at 37°C with Hanks’ balanced salt solution (a) or DMEM containing 2% heat-inactivated horse serum (b) and analyzed by electrophoresis.

of various low-molecular-weight bands in samples during the early stages of the incubation suggests that 3’-5’ exonuclease activity was also involved in degradation of the oligomer in myotubes. Anti-TnC

Treatment Decreases TnC mRNA Level

In various studies, the sequences for antisense oligodeoxynucleotides used were complementary to the first 15-28 nucleotides of the 5’ coding region, in some cases including the initiation signal [ 14, 16, 191. In this study we chose an 18-nucleotide-long sequence; the first nucleotide of this oligomer is 134 and 127 nucleotides away from the A of the AUG initiation codon of fast and slow TnC mRNAs, respectively. This sequence was common to both fast and slow TnC mRNA and therefore was used to allow hybrid formation with both fast and slow TnC mRNAs. The effect of various concentrations of the anti-TnC oligomer in the culture medium was analyzed. The steady-state levels of TnC mRNAs in myotubes following treatment with 0, lo,20 PM anti-TnC or 20 PM sense-TnC oligomer for 3 h were therefore determined. For this analysis total RNA was isolated at the end of the incubation period and analyzed on RNA gel blots (Fig. 4) using probes to detect both fast and slow TnC mRNAs. Cells treated with 20 PM anti-TnC showed 34% reduction in total TnC mRNA level compared to untrtttp). However, there was no reduction in the TnC mRNA level in 20 PM sense-TnC treated cells (Fig. 4B). This value was obtained after normalizing the signals using 28 S rRNA

AND

BAG

levels. Similar results were obtained when either p- and y-actin or P-tubulin mRNA levels determined by reprobing the same RNA blot with respective probes were used for normalization. Furthermore, the reduction of TnC mRNA levels reported here were reproducible in several experiments within a +2% range. In this analysis, it was not possible to resolve the small difference in size between the slow and the fast TnC mRNAs by agarose gel electrophoresis. The steady-state levels, therefore, represent the combined level of both iso-mRNAs. Since both TnC isoforms are functionally similar, at this point, no further attempt was made to separate the two iso-mRNAs. To examine whether the level of TnC mRNA could be further reduced by increasing the length of incubation, cells were treated with 0, 10, or 20 PM anti-TnC for 12 h as described under Materials and Methods. Northern blot analysis of RNA samples revealed that even after 12 h of treatment with 20 PM anti-TnC, approximately 30% (+5) inhibition of TnC mRNA level was achieved in four independent experiments (Fig. 5, top). This reduction of TnC mRNA level was similar to that found after only 3 h of incubation with the anti-TnC oligomer. Therefore, although the anti-TnC treatment decreased the TnC mRNA level in the skeletal myotubes, inhibition was not proportional to the length of treatment (Table 1). Inhibition

of a-Actin and a-Tropomyosin mRNA Levels

To examine if the reduction of the steady-state level of TnC mRNA had any effect on the expression of other

0

10 20

0

20pM

x I

p+ractin-2 (Y actin-

-TnC

TnC-’

-18s

TmL

A

B

FIG. 4. Steady-state levels of several muscle-specific mRNAs in myotuhes. (A) Chicken myotuhes were treated with 0, 10, or 20 ).LLM anti-TnC oligomer for 3 h. Total RNA was isolated from the cells at the end of treatment and electrophoresed (15 rg/lane) in 1.5% agarose gel. The blot was hybridized with a mixture of 32P-labeled slow TnC, fast TnC, and actin cDNAs (top) and reprobed with Lu-tropomyosin cDNA (bottom). (B) Myotuhes were treated with 0 or 20 PM sense-TnC for 3 h, and RNA was isolated and processed as described. The blot was hybridized with a mixture of the slow and fast TnC probes (top). Ethidium bromide-stained RNA gel is also shown (bottom).

INHIBITION

0

OF

TnC

WITH

ANTISENSE

gomer failed to induce differentiation of HL-60 cells in their study. The anti-TnC target sequence is conserved in the chicken slow and fast, mouse slow, and human slow and fast TnC mRNAs [27,28,40,41]. But a computer homology search of the GenBank database (63.0) using the FASTA program [42] failed to show significant homology to any of the sequence present in chicken skeletal actin or a-tropomyosin cDNAs. We have, therefore, concluded that the levels of a-actin and cr-tropomyosin mRNAs in the anti-TnC-treated cells were not decreased due to the oligomer forming a hybrid with either of these mRNAs. It is possible that the reduction in the amount of TnC mRNA triggered a compensatory mechanism(s) leading to reduced levels of mRNAs for the functionally related thin-filament proteins, cY-actin and cu-tropomyosin. It should be noted that the amount of RNA recovered from 20 PM anti-TnC-treated cells was somewhat lower than that from other samples. The levels of TnC, a-tropomyosin, and ol-actin mRNAs were normalized using the amount of 28 S rRNA as internal standards. Therefore these normalized values for the steady-state levels of the mRNAs are slightly greater than they appear from the Northern blot (Figs. 4A and 5). These values (Table 1) also show that the P- and y-actin mRNA levels were not affected due to anti-TnC treatment. In all the samples from myotubes, significant levels of /3-actin mRNA were found. Presence of high levels of P-actin may be due to a delay in switching of actin isotype under the conditions of these experiments. Judging from the degree of fusion of these cells it was assessed that at least 70% of cells were in a multinucleated state. Biochemical differentiation was evidenced by the presence of high levels of mRNAs for muscle-specific proteins like TnC and cu-tropomyosin. To enrich the myotube population in our cultures, treatment with cytosine arabinoside [30, 431 was carried out. Nevertheless, it is difficult to rule out the possibility of the presence of

10 20 /AM

-fl+Y actin --ad actin -TnC

;:

-Tm

FIG. 5. tubes treated under p- and shows shows

Steady-state levels of muscle-specific mRNAs in myofollowing 12 h anti-TnC oligomer treatment. Myotubes were with the indicated amounts of anti-TnC for 12 h as described Materials and Methods. Levels of muscle-specific mRNAs and y-actin mRNAs were analyzed by RNA blotting. The top panel the levels of TnC and actin mRNAs and the bottom panel the levels of ru-tropomyosin mRNA.

muscle-specific genes, the blots from three experiments were reprobed with chicken a-actin and cu-tropomyosin probes. Hybridization analysis showed that the cY-actin mRNA level was reduced by 21 and 27%, respectively, in the 3- and 12-h anti-TnC-treated cells (Figs. 4A and 5, top panels), whereas the cu-tropomyosin level was reduced by 22 and 41% after 3 and 12 h of anti-TnC treatment, respectively (Figs. 4A and 5, bottom panels, and Table 1). The level of p- and y-actin mRNA was not affected in these anti-TnC-treated cells. The antisense-mediated inhibition of gene expression is highly sequence specific. Holt et al. [ 141 reported that even a 2-bp mismatch in the 15-nt-long anti-myc oliTABLE Antisense

Inhibition

231

OLIGOMER

1

of Muscle-Specific

Gene

Expression” Steady-state levels (X of untreated

Time Oligomer Anti-TnC Anti-TnC Anti-TnC Anti-TnC Sense-TnC

in culture (h)

Concentration (rcM)

3 12 3 12

10 10 20 20

3

20

Note. The numbers in parentheses a Quantitation of Northern blots * Not determined.

represent the values normalized shown in Figs. 4 and 5. All values

of mRNAs cells)

/3- and y-actin

TnC

a-actin

100 100 100 100 (92.21) NDb

78.51 79.0 65.19 70.04 (64.58) 118.63

93.8 81.24 79.37 73.43 (67.71) ND

using the P-tubulin mRNA level. were normalized using 28 S rRNA

levels.

cu-tropomyosin 89.82 89.16 77.72 59.26 (54.65) ND

THINAKARAN

232 a

b

c

d

r

0

formation of di- or trisomes which will sediment in the polysomal fraction under the conditions used for subcellular fractionation. This may also explain why RNase H may have failed to degrade all of the TnC mRNA in anti-TnC-treated cells.

20 PM

FIG. 6. Subcellular distribution of TnC mRNA. Polysomal and postpolysomal fractions from chicken myotubes treated with 0 or 20 +M anti-TnC were isolated. RNA from the fractions were hybridized to a mixture of 32P-labeled slow and fast TnC cDNAs. Lanes a and b are polysomal and postpolysomal fractions from untreated myotubes; lanes c and d are polysomal and postpolysomal fractions from 20 pm anti-TnC-treated myotubes, respectively.

2530% fibroblasts in these cultures. Therefore, the presence of different amounts of fibroblast contamination in oligomer-treated cultures may also be responsible for the reduction in the steady-state levels of mRNAs for contractile proteins. However, this possibility is unlikely for several reasons. First of all, experiments were performed using the same batch of culture which was enriched for differentiating myoblasts by selectively killing dividing cells by cytosine arabinoside treatment. Second, when the p-actin level was normalized using rRNA as control, the P-actin level showed no change in oligomer-treated cells. The presence of higher levels of fibroblasts in oligomer-treated cultures should have shown an increased level for the nonmuscle P-actin mRNA. Subcellular Distribution

AND BAG

of TnC mRNA

One of the mechanisms of blocking gene expression by antisense DNA or RNA acts by arresting translation of target mRNA by hybrid formation with the 5’ end of the mRNA [4]. Therefore, we examined if the TnC mRNA present in the anti-TnC-treated cells was being translated. For this analysis distribution of cytoplasmic TnC mRNA between polysomal and free populations was measured. Polysomal and free fractions were prepared from cells treated with anti-TnC oligomer. RNA isolated from these fractions was analyzed on an RNA gel blot. Hybridization analysis revealed that following treatment of myotubes with anti-TnC oligomer, the available undegraded population of TnC mRNA remained bound to polysomes (Fig. 6). Therefore, it appears that the anti-TnC treatment was unable to arrest translation of the TnC mRNA remaining in these cells. To block expression of both slow and fast TnC mRNAs we have deliberately chosen a common nucleotide sequence which is 134 and 127 nucleotides downstream from the initiation codons in the fast and slow TnC mRNAs, respectively. It is therefore possible that in antisense-treated cells, a small number of ribosomes could bind to TnC mRNA and stall near the target sequence which was present as RNA:DNA hybrid. This may allow

Decreased TnC mRNA Level Increases Transcription Rate The possibility of increased transcription of the slow and the fast TnC genes following treatment with antiTnC oligomer was also examined. For this analysis the RNA synthesized in the cells during anti-TnC treatment was labeled with 32[PO;3] and used to hybridize with cDNAs bound to membranes. Quantitation of the resulting autoradiograms revealed that the transcription of both the slow and the fast TnC genes was reproducibly higher in the anti-TnC-treated cells in two independent experiments. Results from one experiment are presented in Fig. 7. The results show that the synthesis of fast TnC mRNA was increased by approximately 28%. The increase of slow TnC mRNA synthesis was however less than that of fast TnC mRNA synthesis and was only 13% above the control level. Similarly the transcription of the cu-tropomyosin gene was also increased by approximately 34% in anti-TnC oligomertreated cells. In contrast to the antisense oligomer, the transcription of both TnC and Tm mRNAs remained unchanged in cells treated with the sense oligomer. The combined synthesis of cy-and /3-actin mRNA was measured in this analysis by using the chicken skeletal muscle cDNA clone that hybridizes to both mRNAs. The results of our analysis show a small decrease in the

160

/I

1

ACTIN

SLOW

TNC

FAST TNC

TROPOMYOSIN

FIG. 7. Effect of anti-TnC treatment on the relative rates of muscle-specific gene transcription. Chicken myotubes were treated with 0 (Cl) and 20 pM anti-TnC (B) or 20 FM sense-TnC oligomer (Cl) for 12 h and RNA transcribed were labeled as described under Materials and Methods. The 32P-labeled RNA was hybridized to appropriate cDNA inserts bound to nylon membrane. Signals from the resulting autoradiogram were quantitated. Values from untreated cells were taken as 100%.

INHIBITION

OF

TnC

WITH

synthesis of these mRNAs in anti-TnC-treated cells. From these results we cannot conclude if cu-actin mRNA synthesis was affected by the antisense oligomer. These results however suggest that the observed increase of TnC and cu-tropomyosin mRNA synthesis was not a nonspecific effect on the transcription of all genes. Anti-T& Treatment TnC Protein

Resulted

in Decreased

Synthesis

233

OLIGOMER

a’

N STnC S‘

of

To examine the effect of the decreased steady-state level of TnC mRNAs on the synthesis of TnC polypeptide following antisense oligomer treatment, we labeled the proteins synthesized during the anti-TnC treatment with [35S]Methionine. The labeled proteins were then analyzed by two-dimensional polyacrylamide gel electrophoresis (IEFSDS-PAGE) followed by autofluorography. The result of this analysis shows that cells treated with 20 pM anti-TnC synthesized approximately 70% less fast TnC polypeptide than sense-TnCtreated or untreated cells (Fig. 8). Decrease in the synthesis of fast TnC polypeptide was somewhat higher than the decreased steady-state level of TnC mRNAs observed in these cells. When the intensities of slow TnC polypeptide spots in Figs. 8A and 8B were compared visually, it appeared that anti-TnC-treated cells synthesized slightly more of this polypeptide. Quantitation of the autoradiogram by using a densitometer, however, did not show any difference in the intensities. Nevertheless it is clear from these studies that a significant reduction in the synthesis of TnC polypeptide can be achieved by this antisense oligomer. Furthermore, synthesis of cu-actin polypeptide was not affected by this treatment. This is in contrast to the decrease in the level of cu-actin mRNA in anti-TnCtreated cells. It is possible that the synthesis of cu-actin polypeptide was compensated by modulating the translation of this mRNA. In addition it should be noted that although the steady-state level of P-actin mRNA appeared higher than the level of P-actin mRNA in the culture (Fig. 4A), the synthesis of cr-actin polypeptide was much higher than that of the /3-actin. This may have resulted from better translation of a-actin mRNA in differentiated cells. Furthermore, the results show that compared to untreated control cells (Fig. 8A) the synthesis of cr-tropomyosin polypeptide was reduced only 25% in cells treated with anti-TnC (Fig. 8B). No reduction in w-tropomyosin synthesis was observed in sense-TnCtreated cells (Fig. SC). DISCUSSION

Use shown genes. grown

ANTISENSE

of antisense oligodeoxynucleotides has been in several systems to block expression of specific In most of such studies, nondifferentiated cells in suspension culture were used [lo, 13-16, 19,

b’ STn c FTnC

c’

FTnC

FIG. 8. Protein synthesis in anti-TnC-treated cells. Chicken myotubes were treated with 0 and 20 pM sense-TnC or 20 ~i?f antiTnC oligomer for 12 h and proteins synthesized were labeled using [?5l]Methionine. Equivalent counts were subjected to IEF/SDSPAGE (12.5% polyacrylamide gels) followed by autofluorography. Panels A-C: Autoradiograms of proteins synthesized by untreated cells (A), 20 PM antisense-TnC-treated cells (B), and 20 pM senseTnC-treated cells (C). Panels a’-c’: the respective enlargements of the acidic region of the autoradiograms showing the TnC isoforms and myosin light chain proteins (a’, b’, c’) are shown. 1, fl- and y-actin; 2, u-actin; 3, fi-tropomyosin; 4, a-tropomyosin; FLCl, FLCB, fast myosin light chains; N, nonmuscle myosin light chain; STnC, slow troponin C; FTnC, fast troponin C. Radioactive spots were identified according to previously published two-dimensional gel patterns [47,48].

20, 441. In our investigation we have used terminally differentiated cells to examine if synthesis of a differentiation specific marker can be influenced by the antisense oligodeoxynucleotide. Our results show that significant inhibition of expression of TnC polypeptide in differentiated chicken myotubes can be achieved by this method. Anti-TnC oligomer at 20 ~Mconcentration was more effective than that at 10 pM in inhibiting the TnC gene expression. But increasing the incubation period did not result in further inhibition. There appears to be an optimum level of inhibition that can be achieved by this approach. The results of our studies show that 20 pM anti-TnC

234

THINAKARAN

oligomer reduced the steady-state level of TnC mRNA by 30-35% which resulted in approximately 70% inhibition of TnC polypeptide synthesis. This level of inhibition was reproducible in several experiments. There was no further reduction of TnC mRNA level either at a higher concentration of the oligomer or after a longer incubation period (results not shown). It is not known why some oligomers exhibit a greater inhibitory effect than the others. The proximity of the antisense target sequence to translation initiation is an important factor in blocking the translation of target mRNA. But oligomers targeted against internal coding sequences and 3’ noncoding regions have also been shown to be effective in inhibiting specific gene expression [15, 171. The failure to observe a more significant reduction in the steady-state level of TnC in our study may be due to the presence of target sequences as a double-stranded structure inside the cells (Fig. 9) and, therefore, TnC mRNA may not be readily available to form a hybrid with the anti-TnC oligomer. Alternatively, compensation by increased transcription of the TnC genes may also be responsible for this observation. Our results show a 70% reduction in the synthesis of TnC polypeptide as a result of approximately 30% reduction in the steady-state level of TnC mRNA. This observation suggests that in addition to reduction in the steady-state level of TnC mRNA the translation of this mRNA was also inhibited in antisense oligomer-treated

Slow

Ttlc

Fast U-U

c-u G

G A G G A

G G G G A

G G A

A

A

A

@

AUGGCU:i&JGC 117

G G G C A C

A

c:* c:o ;i& G AC: u:a A:U

TnC

167

c:s c:a A:@ c:ma G &# AC:' IJ:* A:@ c:m C A G : CAAGAC G:C G:C G:C

U:A G:C G:C GGACGGU:AAAGAGG 214 138

AND

148

GGT GAT GAG GAT GCT GGG

165

111

.. .

. ..

. ..

-.-

. ..

. ..

194

t.UC!-NM

194

_..

.._

.G.

-.-

. ..

211

MLC

If

395

.A.

_._

C..

_w.

412

MLC! 3f

208

.A.

C.A C.A

..A C.A C-A

__.

C..

.._

225

Slow

TnC TnC

Past

FIG. 10. Sequence homology and chicken alkali myosin light bases are shown for the nonmuscle and two fast myosin light chain [45,46]. These mRNAs code for the skeletal muscle myosin alkali light indicates sequence homology.

between anti-TnC target sequence chain mRNAs. The mismatched myosin light chain (MLC-NM) mRNAs (MLC 1’ and MLC 3’) nonmuscle myosin light chain and chains 1 and 3, respectively. (*1

cells. It is therefore likely that the TnC mRNA recovered in the polysomal fraction of antisense-treated cells was associated with fewer ribosomes than in the samples from untreated control cells. The antisense oligomer is a sequence-specific inhibitor of gene expression. But we observed a reduction in the steady-state level of cu-actin and cr-tropomyosin mRNAs along with TnC mRNA. It is unlikely that antiTnC oligomer formed a hybrid with either of the mRNAs and was cleaved by RNase H because a homology search using the FASTA program [42] failed to show any homology between these sequences. The TnC mRNA target sequence however showed 89% homology (2 bp mismatch) with a nonmuscle chicken myosin alkali light chain mRNA sequence (Fig. 10) [45]. In addition, chicken myosin light chain 1 and 3 mRNA sequences [46] had 6 bp mismatch when compared to the target sequence. Two-dimensional electrophoresis of labeled proteins showed that there is no decrease in nonmuscle myosin light chain and myosin light chain polypeptides 1 and 3 in the anti-TnC-treated cells (Fig. 8). So the observed decrease in the steady-state level of Lu-actin and cu-tropomyosin mRNAs and an increase in the transcription of cr-tropomyosin gene in the antiTnC treated cells may be the result of a yet unknown compensatory mechanism(s). This work was supported by a grant-in-aid from the Natural Sciences and Engineering Research Council of Canada. We thank Drs. D. W. Cleveland, A. R. MacLeod, A. R. Means, F. C. Reinach, and R. J. Schwartz for providing the cDNA probes. REFERENCES 1. 2. 3.

FIG. 9. Predicted secondary structure of the anti-TnC target region of TnC mRNA. Both the slow and the fast mRNAs show a similar secondary structure. The anti-TnC target sequence is shaded. The method of Martinez [49] and the energy parameters of Salser [50] (RNAFOLD program (C) 1987 Scientific and Educational Software) were used to generate the secondary structures.

BAG

4. 5. 6.

7.

Hastings, K. E. M., and Emerson, C. P., Jr. (1982) Proc. Ncztl. Acad. Sci. USA 79,1553-1557. Lawrence, J. B., Taneja, K., and Singer, R. H. (19891 Deu. Biol. 133, 235-246. Green, P. J., Pines, O., and Inouye, M. (1986) Annu. Reu. Biothem. 55, 569-597. Walder, J. (1988) Genes Deu. 2, 502-504. Stein, C. A., and Cohen, J. S. (1988) Cancer Res. 48,2659-2668. Toulme, J.-J., and Helene, C. (1988) Gene 72, 51-58. Loke, S. L., Stein, C. A., Zhang, X. H., Nakanishi,

M.,

Suba-

INHIBITION singhe, C., Cohen, J. S., and Neckers, Acad. Sci. USA 86,3474-3478.

OF

L. M.

(1989)

TnC Proc.

WITH N&l.

8.

Yakubov, L. A., Deeva, E. A., Zarytova, V. F., Ivanova, E. M., Ryte, A. S., Yurichenko, L. V., and Vlassov, V. V. (1989) Proc. Natl. Acad. Sci. USA 86, 6454-6458.

9.

Walder, R. Y., and USA 85,5011-5015.

10. 11.

Walder,

J. A. (1988)

Proc.

Natl.

Acad.

ANTISENSE 27.

Reinach, 2371-2376.

28.

Putkey, J. A., Carroll, Biol. 7, 1549-1553.

29.

Maxam, A. M., and Gilbert, W. (1980) in Methods in Enzymology (Grossman, L., and Moldare, K., Eds.), Vol. 65, pp. 499-560, Academic Press, San Diego. Fischbach, G. D. (1972) Deu. Biol. 26407-429.

Sci.

Zamecnik, P. C., Goodchild, J., Taguchi, Y., and Sarin, P. S. (1986) Proc. Natl. Acad. Sci. USA 83, 4143-4146. Westermann, P., Gross, B., and Hoinkis, G. (1989) Biomed. &ochim. Acta 48, 85-93. Matsukura, M., Zon, G., Shinozuka, K., Robert-Guroff, M., Shimada, T., Stein, C. A., Mitsuya, H., Wong-Staal, F., Cohen, J. S., and Broder, S. (1989) Proc. Natl. Acad. Sci. USA 86,424444248. Heikkila, R., Schwab, G., Wickstrom, E., Loke, S. L., Pluznik, D. H., Watt, R., and Neckers, L. M. (1987) Nature (Londonj 328, 445-449.

235

OLIGOMER

30. 31.

F. C., and

Chomczynski, 156-159.

Karlsson,

R. (1988)

J. Biol.

S. I,., and Means,

P., and

Sacchi,

Chem.

A. R. (1987)

N. (1987)

Anal.

263,

Mol.

Cell.

Biochem.

162,

32.

Pramanik, S. K., Meadus, them. 172,355-362.

33.

Meinkoth,

34.

Schwartz, A. (1980)

35.

MacLeod,

36.

Valenzuela, P., Quiroga, M., Kirschner, M. W., and Cleveland, 289, 650-655.

37. 38.

O’Farrell, P. H. (1975) Endo, T., and Masaki,

A. M., and Calabretta, B. (1989) Proc. Nutl. 337993383. A., Littauer, U. Z., and Ginzburg, I. (1988)

39.

Bonner, W. M., and Laskey, R. A. (1974) Eur. J. Biochem. 46, 83-88. Parmacek. M. S., and Leiden, J. M. (1989) J. Biol. Chem. 264, 13,217-13,225.

Caceres, A., and Kosik, K. S. (1990) Nature (London) 343,461463. Zheng, H., Sahai, B. M., Kilgannon, I?., Fotedar, A., and Green, D. R. (1989) Proc. Natl. Acad. Sci. USA 86, 3758-3762.

41.

20.

Harel-Bellan, A., Durum, S., Muegge, Farrar, W. L. (1988) J. Exp. Med. 168,

A. K., and

43.

21.

Brunetti, 5963. Knecht, 1086.

265,5960-

44.

12.

13.

14. 15.

16. 17. 18. 19.

22. 23. 24. 25.

26.

Holt, J. F., Redner, R. L., and Nienhuis, A. W. (1988) Mol. Cell. Biol. 8, 963-973. Wickstorm, E. L., Bacon, T. A., Gonzalez, A., Freeman, D. L., Lyman, G. H., and Wickstorm, E. (1988) Proc. Natl. Acad. Sci. USA 85, 1028-1032. Anfossi, G., Gewirtz, Acad. Sci. 1JSA 86, Teichman-Weinberg, Gene 72, 297-307.

A., and Goldfine,

I. R. (1990)

K., Abbas, 2309-2318. J. Biol.

Chem.

40.

42.

Gahlmann, Mol. Biol.

W. J., and Bag, J. (1988)

J., and Wahl,

G. (1984)

Anal.

A. R. (1982)

R., Wade,

Eur.

K. N., and Dugaiczyk,

J. Biochem.

126,

R., Gunning,

P., and Kedes,

Marcus-Sekura, and Quinnan,

C. J., Woerner, A. M., Shinozuka, G. V., Jr. (1987) Nucleic Acids Rex

46.

Nabeshima, Y., Fujii-Kuriyama, Y., Muramatsu, K. (1984) Nature (London) 308, 333-338.

47.

Mikawa, Biochem. Bandman, 508-518. Martinez,

Received Revised

June 12,199O version received

September

3, 1990

236,

48. 49. 50.

J.

Pearson, W. R., and Lipman, D. J. (1988) Proc. Natl. Acad. Sci. USA 85, 2444-2448. Graham, F. L., and Whitmore, 6. F. (1970) Cancer Res. 30, 2627-2634.

Schulze, H., Huckriede, A., Noegel, A. A., Schleicher, M., and Jockusch, B. M. (1989) EMBO J. 8, 3587-3593. Bucher, E. A., Maisonpierre, P. C., Konieczny, S. F., and Emerson, C. P., Jr. (1988) Mol. Cell. Biol. 8, 4134-4142. Buckingham, M. E., and Minty, A. J. (1983) in Eukaryotic Genes: Their Structure, Activity and Replication (MacLean, W., Gregory, S. P., and Flavell, R. A., Eds.), pp. 365-395, Butterworths, London. Toyota, N., and Shimada, Y. (1983) Cell 33, 297-304.

Science

L. (1988)

201, 379-391.

Nabeshima, Kuriyama,

W. F. (1987)

W. J., fLondon)

J. Biol. Chem. 250, 4007-4021. T. (1982) J. Biochem. 92, 1457-1468.

45.

Loomis,

293-297.

Zaldivar, J., Rutter, D. W. (1981) Nature

1081L

D. A., and

J. Bio-

138,267-284.

Biochem.

R. J., Haron, J. A., Rothblum, Biochem. 19, 5883-5890.

Eur.

Y., Nabeshima, Y. (1987) J. Biol.

Y., Chem.

Nonomura, Y., and 262, 10,60810,612.

T., Takeda, S., Shimizu, T., and 89, 1951-1962. E., Matsuda, R., and Strohman. H. M. (1984)

Salser, W. (1977) 985-1002.

K., Zon,

Nucleic

Acids

Cold Spring

Harbor

Rex

G.,

15,5749-5763. Fujii-

M., and Ogata,

Kitaura, (1982)

T. (1981) Dell. Biol.

J. 93,

12, 323-334.

Symp.

Quant.

Biol.

42,