Characterization of a Potent and Specific Class of Antisense Oligonucleotide Inhibitor of Human Protein Kinase C-α Expression*

The use of antisense oligonucleotides to inhibit the expression of targeted mRNA sequences is becoming increasingly commonplace. Although effective, the most widely used oligonucleotide modification (phosphorothioate) has some limitations. In previous studies we have described a 20-mer phosphorothioate oligodeoxynucleotide inhibitor of human protein kinase C-α expression. In an effort to identify improved antisense inhibitors of protein kinase C expression, a series of 2′ modifications have been incorporated into the protein kinase C-α targeting oligonucleotide, and the effects on oligonucleotide biophysical characteristics and pharmacology evaluated. The incorporation of 2′-O-(2-methoxy)ethyl chemistry resulted in a number of significant improvements in oligonucleotide characteristics. These include an increase in hybridization affinity toward a complementary RNA (1.5° C per modification) and an increase in resistance toward both 3′-exonuclease and intracellular nucleases. These improvements result in a substantial increase in oligonucleotide potency (>20-fold after 72 h). The most active compound identified was used to examine the role played by protein kinase C-α in mediating the phorbol ester-induced changes in c-fos, c-jun, and junB expression in A549 lung epithelial cells. Depletion of protein kinase C-α protein expression by this oligonucleotide lead to a reduction in c-jun expression but not c-fos orjunB. These results demonstrate that 2′-O-(2-methoxy)ethyl-modified antisense oligonucleotides are 1) effective inhibitors of protein kinase C-α expression, and 2) represent a class of antisense oligonucleotide which are much more effective inhibitors of gene expression than the widely used phosphorothioate antisense oligodeoxynucleotides.

The use of antisense oligonucleotides to inhibit the expression of targeted mRNA sequences is becoming increasingly commonplace. Although effective, the most widely used oligonucleotide modification (phosphorothioate) has some limitations. In previous studies we have described a 20-mer phosphorothioate oligodeoxynucleotide inhibitor of human protein kinase C-␣ expression. In an effort to identify improved antisense inhibitors of protein kinase C expression, a series of 2 modifications have been incorporated into the protein kinase C-␣ targeting oligonucleotide, and the effects on oligonucleotide biophysical characteristics and pharmacology evaluated. The incorporation of 2-O-(2-methoxy)ethyl chemistry resulted in a number of significant improvements in oligonucleotide characteristics. These include an increase in hybridization affinity toward a complementary RNA (1.5°C per modification) and an increase in resistance toward both 3-exonuclease and intracellular nucleases. These improvements result in a substantial increase in oligonucleotide potency (>20-fold after 72 h). The most active compound identified was used to examine the role played by protein kinase C-␣ in mediating the phorbol ester-induced changes in c-fos, c-jun, and junB expression in A549 lung epithelial cells. Depletion of protein kinase C-␣ protein expression by this oligonucleotide lead to a reduction in c-jun expression but not c-fos or junB. These results demonstrate that 2-O-(2-methoxy)ethyl-modified antisense oligonucleotides are 1) effective inhibitors of protein kinase C-␣ expression, and 2) represent a class of antisense oligonucleotide which are much more effective inhibitors of gene expression than the widely used phosphorothioate antisense oligodeoxynucleotides.
The protein kinase C (PKC) 1 family of isozymes is composed of at least 11 different, but structurally related serine/threonine kinases. These can be subdivided on the basis of structural and biochemical similarities into three groups, the conventional (␣,␤I, ␤II, and ␥), the novel (␦, ⑀, , and ), and the atypical ( and ) (1-3). The classic PKCs and the novel PKCs are activated by 1,2-diacylglycerol, which is generated by phospholipase cleavage of membrane phospholipids. These phospholipases are regulated by many growth factors and hormones, and it is therefore widely thought that PKC isozymes play an important role in regulating cell proliferation and differentiation, as well as short-term cellular responses, such as secretion and ion flux (1). The identification of multiple members of the PKC family has lead to speculation that individual isozymes play different roles in regulating different cell functions (4). Much evidence is available to support this hypothesis. For example, expression profiles of the individual family members is extremely heterogeneous, both at the tissue and the subcellular levels (4,5). In addition, the substrate specificities of purified proteins are very different, and the responses of isozymes to stimuli differ not just between isotypes, but also between the same isotype stimulated in different cell types.
Considerable effort has been made over the last 10 years to develop isozyme-specific inhibitors of PKC to allow the dissection of the PKC-dependent signaling pathways (6). These efforts are hampered by the similarities in protein structure of the many isozymes of PKC, which make the identification of specific, small molecule enzyme inhibitors difficult. To overcome this difficulty, we have used antisense oligonucleotides to inhibit the expression of individual isozymes of PKC (7,8). Antisense oligonucleotides can be targeted to mRNA sequences which are unique to a given PKC isozyme, leading to the selective inhibition in expression of that isozyme (9,10). The long half-lives of some PKC proteins (and other proteins which have been targeted with antisense oligonucleotides) have proven problematic, as we have found that the most widely used oligonucleotide modification available, the phosphorothioate (PϭS) oligodeoxynucleotide, is metabolized in cells over time. This leads to a loss of activity over a 48 -72-h period, which can make inhibition of some PKC isozymes difficult (7,8,11).
The factors which govern oligonucleotide activity are complex. Two important parameters are the affinity with which an oligonucleotide hybridizes to a target mRNA, and the ability of the oligonucleotide to withstand degradation by intracellular nucleases (12)(13)(14)(15)(16)(17)(18)(19). We have therefore sought to improve these characteristics, with the anticipation that this would lead to the identification of antisense oligonucleotides with improved pharmacological activity compared with those presently available. This should allow for the development not just of improved antisense inhibitors of PKC, but of a more generalized class of antisense effective against any mRNA target which encodes a protein with a long half-life.
2-O-methyl (2Ј-M) and 2Ј-deoxy containing oligonucleotides are contrasted. Antisense oligonucleotides can inhibit the expression of proteins by a number of potential mechanisms (21)(22)(23)(24). One of the most effective oligonucleotide-dependent mechanisms for reducing protein expression is to cause an RNase H-mediated cleavage in the hybridized target mRNA (7,25,26). Unfortunately, however, 2Ј-alkyl modifications (such as 2Ј-MOE) do not support RNase H-mediated mRNA cleavage (12,18,27,28). This can be overcome by the inclusion of 2Ј-deoxy residues into an antisense oligonucleotide, in combination with 2Ј-alkyl modifications, in a motif that will support RNase H cleavage (chimeric oligonucleotides) (11, 15, 29 -31). In our studies reported here, an oligonucleotide containing 2-MOE modification in such a configuration was found to be at least 20-fold more active than conventional PϭS oligodeoxynucleotides at reducing the expression of PKC-␣ mRNA in A549 lung carcinoma cells. This inhibition resulted in a time dependent and oligonucleotide-specific reduction in expression of PKC-␣ protein.
This has allowed us to examine the role played by PKC-␣ in regulating the expression of members of the AP-1 family of transcription factors in A549 lung carcinoma cells. Activation of PKC by phorbol esters leads to an increase in expression of the fos and jun family members, by both increased transcription and increases in mRNA stability. The PKC isoform responsible for this increase is not clear, as A549 cells express multiple phorbol ester binding PKC isozymes, including PKC-␣, ␦, ⑀, and . Depletion of PKC-␣ from A549 cells had a dramatic effect on the increase in c-jun mRNA expression, reducing the up-regulation to levels seen in control cells. In contrast, the phorbol ester dependent increase in c-fos and junB mRNA expression was not inhibited.

EXPERIMENTAL PROCEDURES
Cell Culture-Human A549 lung carcinoma cells were obtained from the American Type Tissue Collection (ATCC). Cells were grown in Dulbecco's modified Eagle's medium containing 1 g of glucose/liter and 10% fetal calf serum and routinely passaged when 90 -95% confluent.
Measurement of Oligonucleotide T m -Absorbance versus temperature profiles were performed as described previously (33). Briefly, antisense oligonucleotides were hybridized to complementary RNA strands and T m values and free energies of duplex formation were obtained. Values are the averages of three experiments.
Oligonucleotide Treatments of Cells-A549 cells were grown to 60 -70% confluence in T-75 flasks. The cells were then washed twice with Dulbecco's modified Eagle's medium and then 5 ml of Dulbecco's modified Eagle's medium containing 20 g/ml N-[1-(2, 3-dioleyloxy)propyl]-n, n,n-trimethylammonium chloride/dioleoyl phosphatidylethanolamine (DOTMA/DOPE) (Lipofectin®)(Life Technologies, Inc.) solution was added to the flasks. Oligonucleotides were added to the required concentration from a 10 M stock solution and the flask swirled to mix the solutions. The cells were then incubated at 37°C for 4 h and then the DOTMA/DOPE/oligonucleotide solution was aspirated off and replaced with medium for the indicated time.
Oligonucleotide Metabolism in A549 Cells-A549 cells were treated with 500 nM oligonucleotides as described above and allowed to recover for 72 h. At this time, metabolites were recovered and analyzed by capillary gel electrophoresis as described previously (11). After digestion with proteinase K, oligonucleotide metabolites were recovered by sequential passages through an anion exchange column and a reverse phase column. Analysis of the samples by capillary gel electrophoresis was performed on a Beckman 5010 P/ACE capillary electrophoresis unit.
Measurement of PKC, jun, and fos mRNA Expression-Total mRNA was extracted from cells and resolved on agarose gels as described previously (7). These were transferred to nylon membrane (Bio-Rad) and probed with 32 P-radiolabeled cDNA probes for different PKC isozymes (7). Additionally, gels were probed with [ 32 P]cDNA probes for c-fos, c-jun, junB (ATCC). Gels were routinely stripped and reprobed with radiolabeled human glycerol-3-phosphate dehydrogenase (GAPDH) probe to confirm equal loading. Radioactive bands were quantitated using a PhosphorImager, and typically we measure only the upper of the two PKC-␣ transcripts, although both are reduced with identical kinetics upon oligonucleotide treatment of cells. 2 Measurement of PKC Protein Expression-PKC isozyme protein expression was determined by Western blotting (7). The antibodies used were obtained as indicated. PKC-␣, UBI; PKC-␦, Santa Cruz Biotechnology; PKC-⑀, a gift from Dr. Doriano Fabbro, Novartis Pharmaceuticals; PKC-, BioMol; PKC-, Santa Cruz Biotechnology; and PKC-(UBI).
Measurement of PKC Enzyme Activity-A549 cells were treated with oligonucleotides for 3 days. Cells were then washed in cold phosphate-buffered saline, scraped, and pelleted into a sample preparation buffer (50 mM Tris-HCl, pH 7.5, 5 mM EDTA, 10 mM EGTA, 50 mM 2-mercaptoethanol, 1 mM phenylmethylsulfonyl fluoride, 10 mM benzamidine). A cytosolic fraction was prepared by centrifugation at 100,000 ϫ g for 1 h at 4°C. PKC enzyme activity was determined by measuring the ability of the cytosolic protein extract to phosphorylate a synthetic peptide substrate in the absence or presence of phosphatidylserine in an enzyme-linked immunosorbent-based assay according to the manufacturers instructions (MBL Co. Ltd., Nagoya, Japan). The final concentrations of the reaction mixture used were 25 mM Tris-HCl, pH 7.0, 3 mM MgCl 2 , 0.1 mM ATP, 2 mM CaCl 2 , 0.5 mM EDTA, 1 mM EGTA, 5 mM 2-mercaptoethanol, Ϯ50 g/ml phosphatidylserine. PKC activity is defined as phosphatidylserine-dependent kinase activity.

Determination of Minimum Oligodeoxynucleotide Residues
Required for Oligonucleotide Activity-The inhibition of PKC-␣ mRNA expression by the uniform phosphorothioate oligodeoxynucleotide sequence used here is believed to be mediated by RNase H (7). The 2Ј-modifications examined in the present study do not form substrates for RNase H, and therefore need to be incorporated into the oligonucleotide in combination with oligodeoxynucleotide residues to effect this mechanism of mRNA degradation (18) (20,34). The number of contiguous oligodeoxynucleotide residues required in an oligonucleotide to support RNase H cleavage of a hybridized RNA have been proposed to range from 3 to 8 (15,31,(35)(36)(37)(38)(39). To determine the requirements for this oligonucleotide sequence, we have initially incorporated 4, 6, or 8 contiguous oligodeoxynucleotide  I Sequence, chemistry, structure, and hybridizing affinity of the oligonucleotides used in the present study The 20-mer antisense oligonucleotide targeted to 3Ј-untranslated sequences in the human PKC-␣ mRNA were synthesized with the indicated 2Ј-sugar modifications (underlined sequences). The center of the antisense oligonucleotides (not underlined) was oligodeoxynucleotide and phosphorothioate. The 2Ј-modified regions were synthesized with either phosphorothioate (lowercase s between bases) or phosphodiester (lower case o between bases) backbone linkages. In addition, the parent phosphorothioate oligodeoxynucleotide (ISIS 3521) and a phosphodiester oligodeoxynucleotide (ISIS 11485) were synthesized. Oligonucleotide sequence is shown 5Ј-3Ј. For indicated T m values, antisense oligonucleotides were hybridized to complementary RNA strands and T m values and free energies of duplex formation were obtained as described under "Experimental Procedures." Values are the averages of three experiments. residues (deoxy gap) into the center of a full phosphorothioated 2Ј-O-methyl oligonucleotide, and determined the ability of these oligonucleotides to reduce PKC-␣ mRNA expression in human lung A549 cells at 500 nM concentration. As shown previously, a fully 2Ј-O-methyl modified compound was unable to reduce PKC-␣ mRNA expression (Fig. 1, A and B). However, increasing the number of oligodeoxynucleotide residues (deoxy gap) present in the oligonucleotide resulted in a progressive increase in the ability of the sequence to reduce PKC-␣ mRNA expression. A contiguous stretch of 8 oligodeoxy residues gave a greater than 90% reduction in expression of PKC-␣ mRNA (Fig. 1, A and B).
Design and Hybridization Thermodynamics of 2Ј-Modified Oligonucleotides-A series of oligonucleotides were subsequently synthesized with 8 contiguous central oligodeoxynucleotide residues flanked with either 2Ј-O-methyl or 2Ј-MOEmodified sugar residues. The 3Ј end base of each oligonucleotide was left oligodeoxynucleotide for synthetic reasons. PϭS backbone linkages were always retained in the central oligodeoxynucleotide sequence of the oligonucleotide to maintain resistance to endonucleases in this part of the molecule (15). The flanking sequences were also prepared with either phosphorothioate and phosphodiester backbones ( Table  I) Oligonucleotide Nuclease Resistance-The nuclease resistance of the chimeric oligonucleotides shown in Table I was determined using a number of different strategies. An in vitro nuclease assay was used to examine the ability to withstand digestion by a snake venom phosphodiesterase (a 3Ј-exonuclease). Under conditions which resulted in 50% digestion of the full-length PϭS oligodeoxynucleotide (ISIS 3521) (60 min incubation time with the nuclease), approximately 75% of a chi- Procedures." At these times metabolites were extracted and resolved by capillary gel electrophoresis after spiking of the sample with a T27 oligonucleotide standard. Migration times of full-length (20-mer) starting material was determined by comparison with this standard and these peaks are indicated.

FIG. 4. Dose response for the reduction in PKC-␣ mRNA expression in A549 cells by modified oligonucleotides 24 h after oligonucleotide addition to cells.
A, A549 cells were treated with the indicated concentration of oligonucleotide (0, 25, 50, or 100 nM) and DOTMA/DOPE as described under "Experimental Procedures" for a period of 4 h. Cells were then washed and allowed to recover for a further 20 h. At this time PKC-␣ mRNA expression was determined by Northern blotting. The gels were stripped and reprobed for expression of a housekeeping gene (G3PDH) to confirm equal loading. The oligonucleotides used were ISIS 9605, 9606, 8329, 5357, and 3521 (see Table  I for oligonucleotide chemistries). B, quantitation of the above gels. ƒ, ISIS 9606; E, ISIS 9605; छ, ISIS 8329; Ⅺ, ISIS 5357; ‚, ISIS 3521. meric 2Ј-O-methyl/PϭO compound (ISIS 8329) was degraded (Fig. 2). In contrast, only approximately 40% of the 2Ј-MOE/ PϭO oligonucleotide (ISIS 9605) was degraded, demonstrating that even combined with a PϭO backbone, this latter modification provides greater nuclease resistance than that obtained by a PϭS oligodeoxynucleotide substituent. When the two 2Ј-O-modified sugar residues were evaluated in the context of a PϭS backbone they provided considerable enhancement of stability (Fig. 2). The 2Ј-MOE was superior, demonstrating no digestion for the duration of the experiment.
Experiments were also performed to evaluate the effects of incorporating the 2Ј-O-methyl and 2Ј-MOE modifications as PϭS on oligonucleotide stability in tissue culture cells. A549 cells were treated with oligonucleotides (500 nM in the presence of cationic liposomes) and the oligonucleotide metabolites extracted from cells 72 h later and resolved by capillary gel electrophoresis. At this time, extensive metabolism of ISIS 3521 had occurred consistent with the successive removal of 3Ј-bases by 3Ј-exonucleases resulting in the appearance of n-1, -2 etc. metabolites (Fig. 3). Some metabolism of the 2Ј-O-methyl containing oligonucleotide (ISIS 5357) was also apparent (results not shown). In contrast, no metabolism of ISIS 9606 (the 2Ј-MOE modified oligonucleotide) was found (Fig. 3).

Effect of 2Ј-Modifications on Ability of Oligonucleotides to Reduce Expression of PKC-␣ mRNA Expression-Transfection
of the PϭS oligodeoxynucleotide (ISIS 3521) into A549 cells results in a concentration-dependent reduction in PKC-␣ mRNA expression after 24 h (7). The IC 50 for this reduction was approximately 100 nM. In the context of a PϭS backbone, the 2Ј-O-methyl (ISIS 5357) and 2Ј-MOE (ISIS 9606) containing oligonucleotides demonstrate approximately a 2-and 5-fold increase in potency, respectively (Fig. 4,A and B) as a consequence of the enhanced hybridizing affinity of these two compounds. In the context of a PϭO backbone, the 2Ј-O-methyl oligonucleotide (ISIS 8329) is inactive, even though this molecule has a substantially higher T m than the parent PϭS oligodeoxynucleotide (ISIS 3521). In contrast, the phosphodiester containing 2Ј-MOE compound (ISIS 9605) is about 4-fold more active than ISIS 3521 (Fig. 4, A and B).
PKC-␣ protein has a very long half-life (approximately 24 h) (40) and therefore to substantially reduce expression of this protein (by Ͼ80%) should require oligonucleotide activity for at least three half-lives of the protein. The ability of the 2Ј-MOE modification to withstand nuclease digestion prompted us to determine whether oligonucleotides containing these modifications could reduce expression of PKC-␣ mRNA for extended periods of time. Oligonucleotides ISIS 3521, ISIS 9605, and ISIS 9606 (as well as two scrambled control oligonucleotides) were transfected into A549 cells and PKC-␣ mRNA expression determined 72 h later. At this time ISIS 3521 is inactive at concentrations up to 100 nM, whereas ISIS 9605 and ISIS 9606 are able to maintain reduced levels of PKC-␣ expression with an oligonucleotide IC 50 of approximately 100 nM (Fig. 5). The specificity in oligonucleotide-dependent reduction in PKC-␣ mRNA expression by 2Ј-MOE containing oligonucleotides was also examined. First, two scrambled control oligonucleotides (ISIS 12963 and ISIS 13009) with the same base composition and chemistry as ISIS 9605 and ISIS 9606, respectively, were found to be without effect on PKC-␣ mRNA expression at concentrations as high as 200 nM (Fig. 5). Second, when the expression of multiple PKC-isozyme mRNA transcripts from cells treated with ISIS 9606 was examined, only the expression of PKC-␣ was found to be reduced (Fig. 6). Levels of PKC-␦, ⑀, , and mRNA were unaffected.

Effect of 2Ј-MOE Modified Oligonucleotide on PKC-␣ Protein Expression and PKC Enzyme Activity in A549 Cells-
The ability of ISIS 9606 to maintain reduced levels of PKC-␣ mRNA expression for up to 72 h suggested to us that this oligonucleotide would effectively reduce PKC-␣ protein expression. A549 cells were treated with 100 nM ISIS 9606 and PKC-␣ protein expression was determined by Western blotting. PKC-␣ protein expression was reduced in a time-dependent manner, consistent with a half-life of the protein of approximately 24 h. The oligonucleotide-mediated reduction in PKC-␣ protein expression was again shown to be specific for this isozyme, as levels of PKC-␦ and -were unaffected by ISIS 9606 (Fig. 7).
We next determined whether ISIS 9606 was able to deplete cells of PKC enzyme activity. Using an enzyme-linked immunosorbent-based assay which measures phosphatidylserine-dependent kinase activity we were able to demonstrate an overall reduction in PKC enzyme activity of approximately 70% (Fig.  8A). Samples of protein extract used in the assay were analyzed by Western blotting to confirm a specific reduction in PKC-␣ protein expression by ISIS 9606 (Fig. 8B). A scrambled control oligonucleotide (ISIS 13009) was without effect on PKC-␣ pro-tein expression or kinase activity. The remaining kinase activity present after oligonucleotide is likely to be due to a combination of residual PKC-␣ protein as well as other phospholipiddependent kinases present in A549 cells (PKC-␦, ⑀, and ).
Effect of PKC-␣ Reduction on Phorbol Ester-mediated AP-1 Gene Expression in A549 Cells-AP-1 is a sequence-specific transcriptional complex composed of members of the fos and jun families (41). Activation of the AP-1 family of transcription factors is complex, involving both transcriptional and posttranslational regulation. The phorbol ester class of tumor promoters are known to be strong inducers of AP-1 via activation of PKC. In A549 cells, 12-O-tetradecanoylphorbol-13-acetate (TPA) causes a time-dependent accumulation of mRNA transcripts of multiple members of the AP-1, including c-jun, junB, and c-fos. A maximum increase in expression of these transcripts occurs at between 30 and 60 min after TPA addition (data not shown). In order to determine whether PKC-␣ is involved in regulating this response, A549 cells were treated with ISIS 9606 prior to treatment with TPA. The induction of both c-fos and junB mRNA were unaffected by PKC-␣ depletion, however, the up-regulation of c-jun was almost completely inhibited (Fig. 9). A scrambled control oligonucleotide was without effect on the up-regulation of any of these AP-1 family members (Fig. 9).

DISCUSSION
Antisense oligonucleotides must possess certain characteristics in order for them to demonstrate full biological activity. Oligonucleotides must be sufficiently resistant to intracellular nucleases, have sufficient affinity for targeted mRNA to bind with a high degree of specificity and fidelity, and possess a mechanism for inhibiting the expression of the protein encoded in the target mRNA. The instability of conventional PϭO DNA to nucleases largely precludes their use as antisense oligonucleotides. This has lead to the replacement by a sulfur atom for an equatorial oxygen atom in the phosphate backbone of oligodeoxynucleotides, resulting in the widely used PϭS modification (17,23,42). The PϭS modification provides considerable stability to both exo-and endonucleases, and are widely used as specific inhibitors of gene expression. The advantages in nuclease stability obtained with this modification come with a price, as each incorporation of a PϭS generates a chiral center and reduces the binding affinity for target mRNA by 1-0.5°C. Furthermore, although PϭS oligonucleotides are less sensitive to nucleases, they will degrade in cells over time (11).
To overcome these drawbacks, considerable research has been undertaken to identify oligonucleotide modifications which provide a more attractive pharmacological profile than PϭS oligodeoxynucleotides (27,43,44). These include modifications to the oligonucleotide backbone (44 -46), the oligonucleotide base (47,48), and the C2Ј position of the ribose (15, 49 -52). To date, however, very few modifications have been fully characterized with respect to their pharmacological activity in tissue culture and animals. In the present study, we have characterized the biophysical and pharmacological activities of oligonucleotides containing 2Ј-MOE modifications and compared them to both PϭS oligodeoxynucleotides and 2Ј-Omethyl containing oligonucleotides. As 2Ј-MOE and 2Ј-Omethyl modifications do not support RNase H-mediated cleavage of hybridized mRNA, we have generated "chimeric" oligonucleotides, composed of both 2Ј-modifications and 2Ј-deoxy residues.
The incorporation of 2Ј-MOE modifications into a 20-mer oligonucleotide has a dramatic effect on the ability of the sequence to hybridize to a target mRNA, approximately 1.5°C per base compared with approximately 1°C per base for the 2Ј-O-methyl modification. The increase in hybridizing affinity obtained by the 2Ј-MOE modification is thought to result from the conformation of the sugar and the backbone. The sugar pucker is believed to change from C2Ј-endo (associated with B-form DNA) to C2Ј-exo, which more closely resembles RNA, and RNA/RNA duplexes are more stable than DNA/DNA duplexes (20). In addition, this limited rotational freedom may produce an enhanced steric effect that limits nucleases from digesting the 3Ј-phosphodiester, resulting in the increased nuclease resistance seen with this modification. Irrespective of the mechanism involved, the 2Ј-MOE modification enhances the nuclease resistance of a PϭO oligonucleotide to at least that of a PϭS oligonucleotide. Combined with a PϭS modification, the 2Ј-MOE containing oligonucleotide exhibits further resistance to intracellular nucleases.
Increases in oligonucleotide affinity for a targeted mRNA have been reported to result in increased oligonucleotide potency, provided a mechanism for preventing protein synthesis is maintained (15,19,28,32,48). The 2Ј-MOE modified oligonucleotides were about 5-fold more potent than the parent PϭS oligodeoxynucleotide sequence (at 24 h) and Ͼ20-fold more potent after 72 h, both in the context of either a PϭS or a PϭO backbone. The increase in potency for the 2Ј-MOE modification versus the PϭS oligodeoxynucleotide is the result of a combination of enhanced nuclease resistance an increase in hybridizing affinity. This modification is clearly superior to the 2Ј-Omethyl modification. The increased potency of the oligonucleotide did not lead to any decrease in specificity. Control oligonucleotides containing the same chemistry as the active sequence, but with a randomized base composition were without effect on PKC-␣ mRNA and protein expression. In addition, only expression of PKC-␣ was reduced by oligonucleotide treatment of the cells.
Relatively little is known about the specific signaling roles played by individual members of the PKC family of isozymes, although recent observations are beginning to shed light on this field. For example, PKC-appears to associate with the B-cell antigen receptor complex and is involved in regulating lymphocyte signaling (53); PKC-is thought to mediate transcriptional activation of the human transglutaminase 1 gene (54); and PKC-has been shown to selectively stimulate the transcription factor complex AP-1 in T-lymphocytes (55). Our strategy has been to use oligonucleotides to help define functions for PKC isozymes, this has proven successful for PKC-␣, -⑀, and -(7) (56,57).
Our previous studies indicated that PKC-␣ expression was required for the phorbol ester-mediated up-regulation of the cell adhesion molecule ICAM-1 expression in A549 cells (7). The 5Ј-regulatory domain of the ICAM-1 gene has been shown to contain multiple promoter sites (58 -60). Of these, the most important regulators of ICAM-1 gene expression appear to be NFB and TPA-responsive elements (61). In A549 cells the key regulators of AP-1 that are expressed are c-jun, junB and c-fos (Fig. 9). 2 The effectiveness of the 2Ј-MOE modified oligonucleotides at reducing PKC-␣ protein expression allowed us to examine whether the phorbol ester-mediated increase in expression of these three transcription factors could be regulated by this PKC isozyme. Of the three, only the expression of c-jun was inhibited, suggesting that PKC-␣ increases ICAM-1 expression through a mechanism which requires c-jun expression.
In summary, we have described the pharmacological characterization of a potent and specific class of antisense oligonucleotide inhibitors of gene expression. These 2Ј-MOE containing oligonucleotides are at least 20 times more active than the widely used PϭS oligodeoxynucleotides. This increase in activity is derived from increases in both hybridizing affinity toward the targeted mRNA and in a substantial increase in resistance toward intracellular nucleases. The oligonucleotide has been used to demonstrate a role for PKC-␣ in regulating c-jun expression in A549 epithelial cells, suggesting that this TPAresponsive element binding transcription factor may be required for complete up-regulation of ICAM-1 mRNA expression in response to phorbol esters.