Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The proto-oncogene c-myc plays a critical role in the control of cell proliferation, differentiation, and apoptosis. It is expressed in nearly all replicating cells and expression is reduced after terminal differentiation. Abnormal elevation of this basic-loop-helix-loop transcriptional activator is associated with several proliferation-related pathologies such as neoplasms and vascular restenosis after angioplasty (Hoffman et al., 1996; Potter and Marcu, 1997). The role of c-myc is not well understood in the process of tissue regeneration. The regeneration process in the liver presents an elegant model for studying the function of this gene during tissue proliferation in an in vivo setting because: 1) c-myc expression is highly up-regulated during the process (Fausto and Webber, 1994; Kren et al., 1996), 2) most genetic and phenotypic events during liver regeneration are tightly regulated and documented in the literature (Kay and Fausto, 1997; Michalopoulos and DeFrances, 1997), and 3) the signaling event for initiation of the regenerative process [70% partial hepatectomy (PH)] is free of toxic side-effects and can be delivered rapidly and precisely (Steer, 1995).

The present study uses antisense oligonucleotides for transient inhibition of c-myc expression during liver regeneration to characterize the function of this protein in the regenerative process. Previous studies by our group have demonstrated that the PH model is well-suited for studies involving inhibition of genes in the regenerating liver by use of several different chemical analogs of antisense oligonucleotides (Arora and Iversen, 2000). The present study uses phosphorodiamidate morpholino oligomers (PMOs) that represent a novel DNA chemistry with a six-membered morpholine ring instead of a deoxyribose sugar and the charged phosphodiester internucleoside linkage replaced by an uncharged phosphorodiamidate linkage (Summerton and Weller, 1997a). The lack of internucleoside charge allows PMOs to avoid nonspecific effects observed with the more commonly used phosphorothioate analogs that bind to cellular and extracellular proteins. Furthermore, PMOs are highly resistant to various nucleases and proteases (Hudziak et al., 1996) and extremely efficient inhibitors of translation via a nonRNase H, sequence-specific steric blockade process (Giles et al., 1998, 1999).

Several antisense oligonucleotides designed to block expression of c-myc are being evaluated for clinical applications in a variety of proliferation-related disorders. A recent report (Tinel et al., 1999) indicates concern that antisense inhibition of c-myc also may alter the expression of cytochrome P-450 (CYP) 3A enzymes. The CYP3A subfamily (Shimada et al., 1994) is a prominent component for xenobiotic biotransformation, ultimately determining the systemic disposition and pharmacokinetics of a variety of therapeutic drugs. Hence, combination of c-myc antisense strategies with existing drugs used in treatment of patients with neoplasms or cardiovascular disease may result in untoward drug interactions.

This study was designed to test the hypothesis that antisense PMOs targeted to the c-myc gene can inhibit the expression of c-myc protein in the regenerating rat liver after 70% PH. The potential for altering the expression of CYP3A enzymes by use of c-myc antisense PMOs also was examined.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animals. Sprague-Dawley rats (Simonsen, Gilroy, CA) weighing between 200 and 225 g were housed in plastic cages in the Laboratory Animal Resources facility at Oregon State University in Corvallis. The animals were maintained in a climate-controlled room with 12-h light/dark cycle and allowed access to a commercial rat chow and tap water ad libitum. All animal protocols conformed to the ethical guidelines of the 1975 Declaration of Helsinki and were approved by the Institutional Animal Care and Use Committee of Oregon State University.

PH. The PH procedure was performed under sterile conditions by a board-certified veterinary surgeon (B.L.S.) by the method described previously (Higgins and Anderson, 1931) with aseptic technique. Rats were anesthetized with isoflurane (Abbott Laboratories, North Chicago, IL) and their ventral surface was shaved along the median-line and swabbed with Betadine. A midline incision was made to expose the liver and the medial and left lateral lobes were securely ligated and then excised. This resulted in removal of ~65 to 70% of the total liver. The abdominal incisions were closed in two layers.

PMO Administration. All PMOs were synthesized at AVI BioPharma (Corvallis, OR) as previously described (Summerton and Weller, 1997b). Purity was >90% full-length as determined by reversed phase HPLC and matrix-assisted laser desorption/ionization time-of-flight mass spectroscopy. Lyophilized PMO compounds were dissolved in sterile saline for injection (Sigma Chemical Co., St. Louis, MO) and filtered through 0.2-µm Acrodisc filters (Gelman Sciences, Inc., Ann Arbor, MI). All injections were made i.p. in a typical volume of 0.5 ml.

Immunoblot Analysis. Levels of all proteins were determined by Western blots with lysates from remnant livers with standard techniques. Lysates were prepared immediately after animal euthanization by homogenizing tissue in lysis buffer (1× PBS, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.2% SDS). Complete Mini EDTA-free (Boehringer-Mannheim, Indianapolis, IN) protease inhibitor cocktail tablets were added to fresh lysates at 8 ml/tablet. Resultant samples were centrifuged at 15,000g for 20 min at 4°C and supernatant was stored at 80°C until use. Note that c-myc was detectable only in fresh, never freeze-thawed lysates. A 10% v/v SDS/acrylamide gel with a 6% SDS/acrylamide stacking gel on top was prepared. Each sample was prepared by mixing 0.05 mg of lysate protein in 0.01 ml of SDS and 5% 2-mercaptoethanol, and loaded onto the gel. All reagents for Western blot were from Sigma Chemical Co. The gel had 20-mA constant current passed through it with a model 3000Xi electrophoresis power supply (Bio-Rad, Richmond, CA) until the tracking dye migrated to the running gel. The current was then increased to 30 mA until the tracking dye migrated off the gel. The gels were then soaked in transfer buffer (192 mM glycine, 25 mM Tris base, pH = 8.3) for 20 min. The protein was then transferred from the gel to methanol-soaked Immobilon-P transfer membranes (Millipore, Bedford, MA) at 480 mV for 60 min. The membranes were allowed to dry, resoaked in methanol, and then incubated with blocking buffer (20 mM Tris base, 150 mM NaCl, 3% nonfat milk, and 0.3% Tween 20) for 1 h at room temperature. The membranes were incubated for 2 h with primary antibodies diluted 1:1000 in blocking buffer. Monoclonal primary antibodies for c-myc (clone C33), proliferating cell nuclear antigen (PCNA, clone PC10), and p21^waf-1 (clone F5) were all obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Monoclonal antibodies to p53 (Ab3) were from Oncogene Research Products, Cambridge, MA, and -actin (clone AC-40) was from Sigma Chemical Co. Immunoblot analysis for CYP3A2 was performed on liver S-9 fractions. Polyclonal primary antibodies for rat CYP3A2 were purchased from Gentest (Woburn, MA). The membranes were washed repeatedly with wash buffer (1× PBS, 0.3% Tween 20), and then incubated for 30 min with appropriate secondary antibody (1:2000) conjugated with horseradish peroxidase (p53 primary antibody is directly biotinylated and was incubated with streptavidin-horseradish peroxidase conjugate). The membranes were again washed repeatedly with wash buffer and then incubated for 1 min with enhanced chemiluminescence Western Blot Reagents (Amersham, Arlington Heights, IL). The membranes were exposed to Kodak Biomax film for 15 to 30 s and the film developed. -actin immunodetection was performed to confirm that all lanes were loaded with similar amounts of protein by stripping the same blot in 100 mM 2-mercaptoethanol, 2% SDS, 62.5 mM Tris-HCl, pH 6.7, for 1 h at 50°C followed by the washing and blocking steps as described.

Isolation of Hepatocytes and Flow Cytometric Analysis of DNA Content. Individual hepatocytes were isolated from the whole liver by incubating ~0.5 g of fresh minced liver sample in 10 ml of 125 mg/100 ml type IV collagenase (Sigma Chemical Co.) solution at 37°C for 10 min. The sample was gently shaken four to five times during the incubation. The sample was then passed through an 18-gauge needle three times and filtered through sterile 100 and 40 µm meshes successively. The filtrate was spun down and resuspended in 10 ml of cold 70% ethanol. The cell cycle distribution of the isolated liver cells was done by the method of Telford (Fraker et al., 1995). The following day, 1 × 10⁶ cells were removed from the ethanol and resuspended in 500 µl of Telford staining reagent [33 µg/ml disodium EDTA (Sigma Chemical Co.), 124 U RNase A (93 U/mg; Sigma Chemical Co.), 50 µg/ml propidium iodide, and 1 µl/ml Triton X-100 (Sigma Chemical Co.) in 1× PBS] for 2 h at 4°C. Cells were passed through a 40-µm filter a second time immediately before flow cytometer analysis on a Coulter Epic XL-MCL machine (excitation wavelength of 488 nm). Cell cycle analysis was done on Phoenix Systems Multicycle software package.

HPLC Detection of PMOs in Liver Tissue. Liver tissue was homogenized with lysis buffer (1× PBS, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.2% SDS) in a dounce homogenizer. The lysates were digested with pronase E (10 µl; 20 mg/ml) at 37°C for 1 to 3 h and samples centrifuged at 15,000g for 20 min. The supernatant was analyzed for presence of PMO AVI-4126 by reversed phase HPLC. Then a 50-µl aliquot of the supernatant was injected on to a PLRP-S column (300 Å pore size; 15 cm × 4.6 mm; 8 µm particle size; Polymer Instruments, Amherst, MA) with a Varian HPLC pump (model 9010 inert) equipped with a variable wavelength UV detector (model 9050) and an AI-200 autosampler (100-µl injector loop volume). The mobile phases (A, 10 mM NaOH and B, 10 mM NaOH/80% acetonitrile) were prepared with HPLC grade solvents and filtered through a 0.2-µm filter before use. The pump gradient program was 5%B (1 min), 15%B (10 min), 15%B (20 min), and 50%B (30 min) at a flow rate of 1 ml/min. The UV detection was done at 254 nm.

Measurement of CYP3A2. The activity of CYP3A2 was measured with erythromycin demethylation assay (Gonzalez, 1989). The samples were prepared by mixing 1.0 mg of S-9 fraction protein, 0.4 mM erythromycin (Sigma Chemical Co.), and 1.0 mM NADPH (Sigma Chemical Co.) in a final volume of 1 ml in 0.1 M potassium phosphate buffer (pH = 7.4). The samples were incubated for 15 min at 37°C. The product was then assayed by the colorimetric method developed by Nash (1953). The samples were mixed with 0.5 ml of 17% perchloric acid (Sigma Chemical Co.) and centrifuged on a Joaun (Winchester, VA) centrifuge at 15,000 rcf for 5 min. The samples were placed in a new tube and mixed with 0.4 ml of Nash reagent (0.02 M 2,4-pentanedione, 0.6% v/v glacial acetic acid, and 3.9 M ammonium acetate) and incubated at 70°C for 20 min. The samples were allowed to cool and read on the spectrophotometer at 412 nm. Absorbencies were compared with a standard curve generated from known concentrations of formaldehyde. Activities were recorded as micromoles of formaldehyde per milligram protein per minute.

Statistical Analysis. All data are reported as means ± S.E. as determined by the computer program InStat2 (GraphPad, San Diego, CA). The P values also were calculated by InStat2 with the Tukey multiple comparison test. P values of <.05 were designated with an asterisk and those of <.005 were designated with a double asterisk. Standard curves and graphs were generated with Prism v2.0 (GraphPad). Prism software also was used to calculate linear regression, slope, and correlation coefficients.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Various Antisense and Control PMO Sequences Tested in Rat PH Model. Several antisense PMO sequences with differing target regions in the c-myc mRNA and lengths ranging from 20-mer to 28-mer were used to suppress the activity of c-myc in the rat liver regeneration model (Table 1). In addition, the effect of substituting uracil for thymine (PMO 1-22-21) in the base composition of the PMOs was studied. Note that two of the antisense PMOs, AVI-4126 and 1-22-115, have one mismatch in their sequences. This was done to accommodate the use of these compounds as potential therapeutic drugs because the sequences are completely complementary to the human c-myc mRNA (Genbank accession nos. D10493 and D90467). For puposes of comparison, all PMOs were injected i.p. at a dose of 0.5 mg/kg immediately following 70% PH. The ratio of the wet weight of the regenerating liver and the body weight of the animals 24 h after PH was used as one of the functional indices to determine the activity of the PMOs. AVI-4126 was determined to be the optimum human antisense sequence in a separate study (Hudziak et al., 2000). All data described henceforth in these experiments were generated with AVI-4126 as the antisense PMO and its scrambled sequence as control PMO.

c-Myc Antisense PMO Causes a Dose-Dependent and Sequence-Dependent Reduction in c-myc Protein in Regenerating Liver. Immunoblot analyses were performed to determine c-myc protein levels in fresh rat liver lysates. The c-myc protein band was detectable in lysates from regenerating livers 4 h after PH (Fig. 1). The c-myc protein was undetectable to barely detectable in rats that did not undergo PH, irrespective of treatment with AVI-4126. There was an up-regulation in c-myc protein levels after PH that was observed in rats treated either with saline or 0.5 mg/kg scrambled AVI-4126. There was a dose-dependent decrease in the intensity of the c-myc band with increasing doses of the antisense PMO AVI-4126, starting at 0.1 mg/kg and the signal completely disappearing at 12.5 mg/kg. -Actin immunodetection on the same blots resulted in similar band intensities in all lanes. This inhibition of the c-myc protein was confirmed in three separate sets of identically treated rats.

View larger version (32K): [in this window] [in a new window]   Fig. 1.   Representative immunoblot from triplicate analysis of c-myc in rat liver lysates. Rats were either nonhepatectomized or were allowed to recover for 4 h after PH, as indicated. The following were the lane-by-lane treatments administered to rats: 1) saline, 2) 0.5 mg/kg AVI-4126, 3) saline, 4) 0.1 mg/kg AVI-4126, 5) 0.5 mg/kg AVI-4126, 6) 2.5 mg/kg AVI-4126, 7) 12.5 mg/kg AVI-4126, and 8) 0.5 mg/kg AVI-4126 scrambled. The immunoblot was stripped and reprobed to determine -actin levels (bottom).

PMO is Detectable in Regenerating Liver 24 h after PH. Reversed phase HPLC detection of PMO AVI-4126 was performed in lysates of regenerating rat livers 24 h following PH (Fig. 2, A-D). This time point was selected because all parameters in the study described henceforth were measured 24 h after PH. The UV absorbance peak representing the PMO was readily detectable in the pronase-E-treated liver lysates only in rats treated with the PMO. Furthermore, the peak size was a function of the PMO dose administered immediately after PH.

View larger version (14K): [in this window] [in a new window]   Fig. 2.   Reversed phase HPLC chromatograms of rat liver tissue to detect presence of full-length PMO 24 h after administration. A, PMO standard; B, liver from rat treated with 12.5 mg/kg AVI-4126; C, liver from rat treated with 2.5 mg/kg AVI-4126; and D, liver from rat treated with saline.

c-Myc Antisense PMO Causes Reduction in Liver PCNA Expression. PCNA is a component of the DNA polymerase- and an accurate indicator of DNA synthesis activity (Foley et al., 1993; Assy et al., 1998). PCNA levels were determined in liver lysates by immunoblot analysis of three sets of animals, and a representative blot is presented in Fig. 3. PCNA levels were detectable but low in nonhepatectomized rat livers and were not affected by treatment with AVI-4126. As expected, an induction was observed 24 h after PH in control rats that were treated with saline or 0.5 mg/kg scrambled AVI-4126. Increasing doses of antisense PMO AVI-4126 caused a dose-dependent reduction in PCNA band intensities with the highest dose (12.5 mg/kg), reducing PCNA expression to an undetectable level.

View larger version (37K): [in this window] [in a new window]   Fig. 3.   Representative immunoblot analysis of PCNA in rat liver lysates. Rats were either nonhepatectomized or were allowed to recover for 24 h after PH, as indicated. The following were the lane by lane treatments administered to rats: 1) saline, 2) 0.5 mg/kg AVI-4126, 3) saline, 4) 0.1 mg/kg AVI-4126, 5) 0.5 mg/kg AVI-4126, 6) 2.5 mg/kg AVI-4126, 7) 12.5 mg/kg AVI-4126, and 8) 0.5 mg/kg AVI-4126 scrambled. The immunoblot was stripped and reprobed to determine -actin levels (bottom).

c-Myc Antisense PMO Treatment Reduces Population of Liver Cells in G₂ Phase of Cell Cycle. Cell cycle distribution of hepatocytes in the regenerating liver was determined by analysis of their total DNA content. Hepatocytes were isolated from the liver tissue and incubated with Telford reagent to label DNA with propidium iodide. The label was quantitated with flow cytometric techniques followed by determination of cell cycle distribution based on calculation of area under the curves. The cells with 2n DNA were classified as G₀/G₁ populations, those with 4n DNA were classified as G₂/M populations, and those with DNA content >2n but <4n were assigned to be in the S phase of cell cycle. Representative cell cycle distribution curves for 24-h PH rats treated with 0.5 mg/kg AVI-4126 scrambled and AVI-4126 are shown in Fig. 4, A and B, respectively. The ratio of G₂:G₀ cell populations in individual livers was used to determine the effect of treatment of rats with the antisense PMO (Fig. 4C). The mean G₂:G₀ ratio (as a percentage) dropped in a dose-dependent manner from 29.1 to 18.0 for rats treated with vehicle alone (saline) and 2.5 mg/kg AVI-4126, respectively. One group of rats was pretreated with 0.5 mg/kg AVI-4126 two hours before PH surgery. This group had an average G₂:G₀ ratio (percentage) of 17.5. 

View larger version (12K): [in this window] [in a new window]   Fig. 4.   Effect of treatment with AVI-4126 on the cell cycle distribution of hepatocytes in the regenerating liver. Typical histograms for 24-h PH rats treated with 0.5 mg/kg of AVI-4126 scrambled (A) and AVI-4126 (B). Absolute area of individual peaks depends on total number cells counted. The mean of G2/G0 ratio (percentage) with increasing doses of AVI-4126 is plotted in C. (2hr) indicates PMO was administered 2 h before PH surgery. See Materials and Methods for details. Number of animals for each treatment regimen varied from 3 to 13.

c-Myc Antisense PMO Reduces Expression of Cell Cycle Checkpoint Protein p53 but not cdk Inhibitor p21^waf-1. Levels of the cell cycle checkpoint protein p53 and its downstream partner p21^waf-1, an inhibitor of cyclin-dependent kinase (cdk)-2, were determined in the liver samples by immunoblot analysis of tissue from three sets of rats. Representative blots are presented in Fig. 5. As reported previously (Kren et al. 1996), an increase was observed in the expression of both proteins after PH. The levels of p53 decreased hand in hand with increasing doses of antisense PMO AVI-4126 in regenerating livers 24 h after PH. However, the changes in p53 expression had no effect on the levels of p21^waf-1 in the PH rat livers, which remained steady with increasing doses of the antisense PMO.

View larger version (38K): [in this window] [in a new window]   Fig. 5.   Representative immunoblot analysis of p53 and p21waf-1 in rat liver lysates. Rats were either nonhepatectomized or were allowed to recover for 24 h after PH, as indicated. The following were the lane by lane treatments administered to rats: 1) saline, 2) 0.5 mg/kg AVI-4126, 3) saline, 4) 0.1 mg/kg AVI-4126, 5) 0.5 mg/kg AVI-4126, 6) 2.5 mg/kg AVI-4126, 7) 12.5 mg/kg AVI-4126, and 8) 0.5 mg/kg AVI-4126 scrambled. The immunoblots were stripped and reprobed to determine -actin levels (see corresponding gels, bottom).

c-Myc Antisense Causes Dose-Dependent Reduction in Liver CYP3A2 Levels and Activity. Levels of CYP3A2 were measured by immunoblot analysis of liver S-9 fractions in addition to the functional determination of its activity by erythromycin N-demethylation assay. No change in CYP3A activity was detected in nonhepatectamized AVI-4126-treated rats. A decrease in liver erythromycin N-demethylation activity from 0.87 ± 0.05 to 0.32 ± 0.03 µmole formaldehyde/mg was observed 24 h after PH in saline-treated rats (Fig. 6B). This is consistent with previous reports in literature about reduction in CYP activity in regenerating rat liver 24 h after PH (Arora et al., 1998). A dose-dependent decrease in erythromycin N-demethylation activity was observed from 0.32 ± 0.03 to 0.02 ± 0.005 µmol formaldehyde/mg protein/min in vehicle (saline) only to 12.5 mg/kg in AVI-4126-treated PH rats, respectively. The pattern of decrease in functional CYP3A activity in PH rats treated with AVI-4126 also was observed in immunoblot band intensity of the 3A2 protein in liver S-9 fractions (Fig. 6A). Stripping and reprobing of the same immunoblot with -actin antibodies resulted in similar band intensities in all lanes.

View larger version (50K): [in this window] [in a new window]   Fig. 6.   A, representative immunoblot analysis of CYP3A2 in rat liver S-9 fractions. Rats were either nonhepatectomized or were allowed to recover for 24 h after PH, as indicated. The following were the lane by lane treatments administered to rats: 1) saline, 2) 0.5 mg/kg AVI-4126, 3) saline, 4) 0.1 mg/kg AVI-4126, 5) 0.5 mg/kg AVI-4126, 6) 2.5 mg/kg AVI-4126, 7) 12.5 mg/kg AVI-4126, and 8) 0.5 mg/kg AVI-4126 scramled. The immunoblot was stripped and reprobed to determine -actin levels (bottom). B, erythromycin N-demethylase activity in the same matrix of liver S-9 fractions. Data are expressed as micromoles formaldehyde/milligram/minute. For each treatment regimen, n = 3. **P < .005 from saline-treated PH rats.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The present study presents an in vivo approach to examining regulation of gene expression in the liver. Antisense PMOs were able to suppress the expression of c-myc in the regenerating rat liver in a sequence-specific and dose-dependent manner after systemic administration (Fig. 1). The sequence of AVI-4126 contains a single cytosine-to-adenine mispair at the 3' end of the oligomer. This sequence was selected because of its complementarity to the human c-myc mRNA and these studies were intended to provide insights into the activity of the human antisense sequence. The data presented indicate that a single mispair at the 3' end of the oligomer does not prevent antisense activity. The targets for the majority of the tested sequences were clustered at the translation initiation region of the c-myc mRNA (Genbank accession no. Y00396). This approach was selected because antisense PMOs have been reported to act as steric inhibitors of translation with no RNase H activity (Giles et al., 1999). Confirmation of delivery of PMO to its site of action (regenerating liver) after systemic administration was demonstrated by reversed phase HPLC and found to be dose-dependent.

The functional effect of c-myc suppression on the rate of proliferation in the regenerating liver was determined by immunoblot analysis of PCNA, which is a component of the DNA polymerase-. The level of this nuclear protein has been determined to be an accurate indicator of proliferative activity (Assy et al., 1998) and found to be comparable with more traditional approaches such as use of [³H]thymidine in the rat liver regeneration model (Foley et al., 1993). We observed a dose-dependent reduction in PCNA levels in the regenerating liver in rats treated with c-myc antisense PMO (Fig. 3). At the highest dose of 12.5 mg/kg AVI-4126, we found no detectable immunoblot signal for PCNA in the liver. We also observed a concomitant reduction in the percentage of G₂:G₀ ratio of cell cycle distribution in the regenerating liver at the same time point (24 h after PH) from 29.1 to 18.0 for saline-treated and 2.5 mg/kg AVI-4126-treated rats, respectively (Fig. 4C). The antiproliferative nature of c-myc antisense PMO is clearly apparent when both molecular and physiological endpoints are considered.

A reduction in levels of the classic G₁/S cell cycle checkpoint protein p53 was observed with increasing doses of AVI-4126 (Fig. 5). This was contrary to our original expectation because a reduction in p53 is typically associated with loss of cell cycle checkpoint activity and increased proliferation (Ko and Prives, 1996). However, our data agree with a recent mechanistic study in the literature (Kirch et al., 1999) that suggests that the expression of c-myc is reduced 5-fold after a mutation in the myc/max binding site in the promoter region of the human p53 gene. We studied the expression of the protein product of the p21^waf-1 gene, a classic downstream partner of p53 that mediates its cell cycle checkpoint activity by inhibition of cdks (El-Deiry, 1998), to resolve this apparent dilemma. We found that the reduction of the p53 protein with increasing doses of AVI-4126 had no effect on the p21^waf-1 protein, which remained steady irrespective of antisense treatment in the regenerating liver. The dissociation between the protein levels of p53 and p21^waf-1 suggests that the antiproliferative activity of c-myc antisense PMO is not mediated at the cdk level by p21^waf-1.

The possibility of modulation of CYP enzymes is an important concern for any agent that could potentially be used as a drug. Our findings with CYP3A2 immunoblot analysis (Fig. 6) suggest that CYP3A2 is down-regulated in a dose-dependent manner in the regenerating liver after treatment with c-myc antisense PMO. Note that this immunoblot, detected with a polyclonal antibody raised against rat CYP3A2, displays two distinct bands that migrate closely and fluctuate similarly after various treatments. Huss and Kasper (1998) speculate that these represent CYP3A2 and CYP3A23 as the major and minor forms of rat CYP3A, respectively.

Tinel et al. (1999) have recently reported that a phosphorothioate backbone c-myc antisense compound, a 15-mer with the target region overlapping AVI-4126 on the c-myc mRNA, can cause up-regulation of CYP3A in interleukin-2-treated rat hepatocytes. These appear to be contradictory findings compared with the immunoblot analyses and erythromycin N-demethylation data presented in Fig. 6. These contradictions could possibly be explained by the differences in the models used. Tinel et al. (1999) used primary cultures of rat hepatocytes treated with interleukin-2 to induce c-myc compared with our in vivo rat model in which c-myc induction occurs physiologically after PH. However, note that Burgess et al. (1995) have previously reported, with a 15-mer phosphorothioate oligonucleotide similar in sequence to the one used by Tinel et al. (1999), that antiproliferative activity of that oligonucleotide in smooth muscle cells is caused by a nonantisense mechanism resulting from four contiguous guanosine residues in a oligonucleotide with a phosphorothioate backbone (the G-quartet). The PMO chemistry used in our study does not support the formation of G-quartets. Despite the differences in data generated from the two models, the alteration in CYP3A2 activity after inhibition of c-myc remains an important concern that warrants further investigation.

The PMO chemistry offers distinct advantages over second-generation antisense oligonucleotides such as phosphorothioates, methylphosphonates, and phosphotriesters by offering improved efficacy, stability, and delivery (Summerton and Weller, 1997a). The first advantage is the neutrality of the PMO chemistry, which avoids the potential for the formation of biologically active G-quartets and interactions with cations found on the surface of proteins, lipids, and carbohydrates. The second advantage for the PMO chemistry is its inability to recruit RNase H. The RNase H competent oligonucleotide chemistries such as phosphorothioates lack specificity, in part, because duplexes as short as five base pairs in length can be cleaved by this enzyme (Crouch and Dirksen, 1982). Cleavage of the RNA in such short duplexes might occur once in every 1000 bases or approximately once in every transcript from every gene. Therefore, the advantages of the PMO chemistry represent fundamental improvements in the selective inhibition of genes with the antisense approach.

These studies have detailed an overall slow down of the regenerative process in rat liver after PH by use of c-myc antisense PMO. c-Myc expression is inhibited in a dose-dependent and sequence-dependent manner. The decline of c-myc protein in the regenerating liver is associated with decrease in weight gain and PCNA in the regenerating liver. There is also a concomitant reduction in the levels of the cell cycle checkpoint protein p53 and the population of cells in the G₂ phase of the cell cycle. Our data in the liver regeneration model are consistent with observations in other systems that c-myc is an important positive regulator of cell proliferation. We conclude that selective inhibition of c-myc with antisense PMO is feasible in vivo and presents a potential therapeutic strategy for several proliferation-related disorders.