Isis Pharmaceuticals, Inc., Departments of Pharmacokinetics
(R.Z.Y., R.S.G., L.M., A.A.L.), Antisense Drug Discovery (M.G., K.L.,
R.C.), and Pharmacology (H.Z., N.M.D.), Carlsbad, California
ISIS 22023 is a modified phosphorothioate antisense oligonucleotide
targeting murine Fas mRNA. Treatment of mice with ISIS 22023 reduced
Fas expression in liver in a concentration-dependent and
sequence-specific manner, which completely protected mice from
fulminant death induced by agonistic Fas antibody. In this study, we
characterized the relationships in mice between total dose
administered, dose to the target organ, and ultimately, the intracellular concentration within target cell types to the
pharmacologic activity of ISIS 22023. After subcutaneous injection,
ISIS 22023 distributed to the liver rapidly and remained in the liver
with the t1/2 ranging from 11 to 19 days,
depending on dose. There were apparent differences in patterns of
uptake and elimination in different types of liver cells.
Oligonucleotide appeared within hepatocytes rapidly, whereas the peak
concentrations in Kupffer cells were delayed until 2 days after dose
administration. Hepatocytes cleared oligonucleotide the most rapidly,
whereas Kupffer cells appeared to retain oligonucleotide longer. The
reduction of Fas mRNA levels (pharmacodynamic response)
paralleled the increase of oligonucleotide concentration in mouse liver
with maximum mRNA reduction of 90% at 2 days after a single 50 mg/kg
subcutaneous administration. Moreover, the pharmacodynamics of ISIS
22023 correlated better with the pharmacokinetics in hepatocytes,
supporting the concept that the presence of oligonucleotide in target
cells results in reductions in mRNA and, ultimately, pharmacologic
activity. These results provide a comprehensive understanding of the
kinetics of an antisense drug at the site of action and demonstrate
that the reductions in mRNA induced by this antisense oligonucleotide correlate with its concentrations in cell targets.
 |
Introduction |
A
guiding principle in pharmacology is that there should be a correlation
between dose and pharmacologic activity of a therapeutic agent. At
times it is not always possible to define such relationships because of
poor absorption, first-pass effects, elimination, or any one of a
multitude of factors. In practice, plasma concentrations are often used
as a surrogate for dose, thereby correcting for absorption,
elimination, or metabolic factors. Clearly, dose administered and
plasma levels are both surrogates for understanding the
pharmacokinetics at the target site. Although these principles are well
accepted and thoroughly validated for low molecular weight drugs, with antisense therapeutic agents there have always been questions about
pharmacokinetics and tissue uptake. Little is known about the effective
concentrations at the target sites for antisense oligonucleotides.
Phosphorothioate oligodeoxynucleotides have been studied extensively in
recent years as potential antisense therapeutic agents and have been
shown to be effective in the treatment of a broad range of diseases,
including viral infections, inflammatory diseases, and various types of
cancer (Monia et al., 1992
; Bennett et al., 1996; Cowsert, 1997). The
first antisense phosphorothioate oligodeoxynucleotide drug, Vitravene,
was approved by the FDA in 1998 for treatment of cytomegalovirus retinitis.
ISIS 22023 is a 20-mer phosphorothioate oligonucleotide that
incorporates 2'-O-(2-methoxy) ethyl (MOE) modification on
the five nucleotides on both 3'- and 5'-termini (Zhang et al., 2000). MOE modifications provide enhanced resistance to nucleases, provide a
longer target organ half-life, and have reduced toxicity (McKay et al.,
1999; Henry et al., 2000). In addition, these modifications also
increase the affinity of an antisense oligonucleotide for complementary
target mRNA, resulting in enhanced potency and specificity (Altmann et
al., 1996; Baker et al., 1997; McKay et al., 1999). Although the
pharmacokinetics, tissue, suborgan, and subcellular distribution of
phosphorothioate oligonucleotides have been well characterized in
rodents (Cossum et al., 1993; Agrawal et al., 1995; Zhang et al., 1995;
Graham et al., 1998), primates (Grindel et al., 1998), and humans
(Agrawal et al., 1991; Iversen et al., 1994; Glover et al., 1997;
Stevenson et al., 1999), this is the first complete correlation of the
pharmacokinetics and pharmacodynamics of an antisense oligonucleotide.
ISIS 22023 was designed to target a sequence within the translated
region of the murine Fas mRNA. Previous studies have shown that ISIS
22023 inhibited Fas expression in a dose-dependent and sequence-specific manner both in vitro and in vivo (Zhang et al., 2000). Treatment with ISIS 22023 and its subsequent reduction in Fas
mRNA levels have been shown to protect mice from liver injury in both
agonistic Fas antibody and acetaminophen-induced fulminant hepatitis
models and to completely protect mice from death induced by
antagonistic Fas antibody.
For an antisense oligonucleotide to exhibit a pharmacologic response,
the administered agent must distribute to its target organ and be taken
up by the target cells, where it can bind to its cognate mRNA. In this
study, we characterized the pharmacokinetics and pharmacodynamics of
ISIS 22023 in mouse liver, the target organ, and in the different cell
types within that organ. The results demonstrated that the time course
of mRNA reduction (pharmacologic activity) correlated closely with the
time course of ISIS 22023 concentrations in hepatocytes and that Fas
mRNA reduction in mouse liver protected mice from Fas antibody-induced
fulminant hepatitis.
| |
Materials and Methods |
Oligonucleotides.
ISIS 22023 is a 20-nucleotide
phosphorothioate antisense oligonucleotide with MOE modifications at
positions 1 through 5 and 15 through 20 (underlined below). In
addition, the cytosines in the 5'- and 3'-ends of the compound were
modified to contain a 5-methyl group (5-methyl cytosine,
CM). The sequence of ISIS 22023 is
TCMCMAGCMACMTTTCMTTTTCMCMGG.
ISIS 13866, a 2'-MOE modified oligonucleotide at positions 15 through
21 with a sequence of 5'-GCG TTT GCT CTT CTT
CMTT GCMG TTT TTT-3', was used as the internal standard for quantitation of ISIS 22023 in
plasma and tissues. ISIS 22023 and ISIS 13866 were synthesized using an
automated DNA synthesizer model 380B (Applied Biosystems, Inc., Foster
City, CA) and purified as previously described (Baker et al., 1997).
The purity of the compounds used in this study was found to be 89.4%
and 88.8% for ISIS 22023 and ISIS 13866, respectively.
Animals and Treatments.
Female BALB/c mice, 8 weeks
old (16-25 g), were purchased from The Jackson Laboratory (Bar Harbor,
ME) and maintained in compliance with Isis Institutional Animal Care
and Use Committee Guidelines in an Association for Assessment and
Accreditation of Laboratory Animal Care International
(AAALAC)-accredited facility. The animals were housed in
polycarbonate cages (three mice per cage) and allowed food and water ad
libitum. ISIS 22023 solutions were formulated in sterile saline at
concentrations of 1, 2, 3.5, and 5 mg/ml. The mice were randomly
assigned to three groups such that three mice were sacrificed at each
time point per group. Group 1 received a single subcutaneous (s.c.)
injection of ISIS 22023 at a doses of 10, 20, 35, or 50 mg/kg (dose
volume, 10 ml/kg). Group 2 was treated with a single s.c. injection of
50 mg/kg ISIS 22023 and was used to study suborgan kinetics. Group 3 was treated with one, two, and three doses of 50 mg/kg of ISIS 22023 administered subcutaneously every other week.
At each scheduled sampling time point (see Table
1) for groups 1 and 3, three mice were
anesthetized using Avertin (containing 1.25% w/v tribromoethanol,
1.25% v/v tert-amyl alcohol in 0.9% saline), and blood
samples were collected from each mouse by cardiac puncture using EDTA
as the anticoagulant. Approximately 100-mg aliquots of liver from each
mouse were homogenized in 1 ml of RLT buffer (Qiagen, Santa Clarita,
CA) containing 10 µl of 
-mercaptoethanol (Sigma Chemical Co., St.
Louis, MO) immediately after sample collection, and frozen on dry ice
for RNA analysis. The rest of the liver was placed in labeled vials and
frozen for capillary gel electrophoretic (CGE) analysis. All samples
were stored at 
70°C until analysis.
For group 2, the liver was perfused in situ at each sampling time point
to prepare purified liver cell fractions.
Preparation of Suborgan Fractions.
Hepatocyte, Kupffer
cells, and endothelial cells were isolated from the mouse liver as
described previously (Graham et al., 1998). Briefly, mouse liver was
perfused in situ with a collagenase buffer. After perfusion, the
prepared cell suspension was used to isolate parenchymal and
nonparenchymal cells by differential centrifugation and plating. The
nonparenchymal cells were further separated into Kupffer and
endothelial cells. Flow cytometry was used to characterize the enriched
parenchymal and nonparenchymal cells by observing the forward and side
scatter display. The result showed that the hepatocytes were 95% free
of other contaminating cells, and the nonparenchymal cells were 92%
free of parenchymal cells (Graham et al., 1998). Viability, as
determined by Trypan Blue exclusion, was generally greater then 85 to
95%.
Sample Extraction and Analysis of ISIS 22023 in Plasma and
Liver.
An aliquot of plasma (100 µl) for each sample was spiked
with a known concentration of internal standard and extracted using solid-phase extraction and analyzed by CGE analysis (Leeds et al.,
1996).
For liver samples, the method was similar except that tissues were
weighed, homogenized in a Fastprep homogenizer (BIO 101, Inc.,
Vista, CA), and then the material was extracted as described (Leeds et
al., 1996) with the exception that a phenyl-bonded SPE column (Supelco
Inc., Bellefonte, PA) was used. Extraction of oligonucleotide from
suborgan fractions was the same as previously described (Graham et al.,
1998). Extracted samples were analyzed by CGE using a Beckman P/ACE
model 5010 capillary electrophoresis instrument (Beckman Instruments,
Irvine, CA) with UV detection at 260 nm.
The concentrations of ISIS 22023 and metabolites in the plasma and
liver samples were calculated from the ratio of the absorbances, based
on the starting concentration of internal standard (ISIS 13866) as
previously described (Leeds et al., 1996). The limit of quantitation
for this assay has been estimated to be 0.07 µg/ml and 0.35 µg/g
oligonucleotide in plasma and liver, respectively.
RNA Isolation and RNase Protection Assay.
Total RNAs were
extracted from mouse liver using the RNeasy kit (Qiagen). An RNase
protection assay was performed as suggested in the RiboQuant manual
(PharMingen, San Diego, CA). RNase protection assay template mApo-3 and
the custom templates (PharMingen) were used as probes. Total RNA (20 µg) of each treatment, either from cell culture or mouse liver, were
used and analyzed on 6% denaturing polyacrylamide gel. Gels were then
scanned by a PhosphorImager (model 300 Series, Molecular Dynamics,
Sunnyvale, CA).
Pharmacokinetic Analysis.
The plasma concentration-time
profile after a single subcutaneous administration of 50 mg/kg ISIS
22023 was fitted to a one-compartment open model using WinNonlin 1.5 (Scientific Consulting, Inc., Cary, NC). ISIS 22023 was rapidly cleared
from plasma and distributed into tissues. Whole body clearance cannot
be monitored using the terminal elimination phase in plasma because the
existing methods lack sufficient sensitivity to assay such low levels
of oligonucleotides. However, the elimination of ISIS 22023 from liver
could be measured directly, and the liver concentration data after
single subcutaneous injection of 10 to 50 mg/kg were analyzed using a
noncompartmental model (WinNonlin 1.5).
Cell-specific kinetics showed that the hepatocyte-concentration
(parenchymal cells) data declined in a bi-exponential fashion over
time. Moreover, noncompartmental analysis indicated that both liver
uptake and elimination of ISIS 22023 might be saturated at high dose.
Because of the limited number of observations collected, the model was
simplified by choosing a two-compartment linear model (WinNonlin 1.5).
Pharmacodynamic Analysis.
The relation between Fas mRNA
reductions and ISIS 22023 concentrations in hepatocytes was
characterized using a pharmacodynamic model. Fas mRNA levels following
ISIS 22023 treatment were normalized with the Fas mRNA levels from
control animals (treated with saline):
|
(1)
|
The relationship between inhibitory activity of Fas mRNA
(E) and concentration of ISIS 22023 in liver was described
by adopting a inhibitory sigmoidal
Emax model:
|
(2)
|
where Emax is the maximum level
of Fas mRNA, E0 is the baseline level,
EC50 is the concentration of ISIS 22023 required
for half-maximal reduction of Fas mRNA, C is the
concentration of ISIS 22023 in hepatocyte, and n is the
sigmoidicity factor.
Integrated Pharmacokinetic/Pharmacodynamic Model.
ISIS 22023 specifically binds to its target (Fas mRNA), and it is this hybrid
target mRNA that will then be degraded by RNase H enzymatic activity.
Indirect response models assume that the drug produces its action by
stimulating or inhibiting the production or dissipation of a
pharmacologic response (Dayneka et al., 1993). Therefore, ISIS 22023 binding to Fas mRNA was assumed to stimulate RNase H activity, which
could induce the degradation of Fas mRNA (Fig.
1). The pharmacodynamic model is based on
a scheme in which Fas mRNA is synthesized with a zero-order rate
constant (Rin) and enzymatically
cleaved by a first-order degradation rate
(Kdeg). ISIS 22023 activates RNase H
to degrade Fas mRNA, thus increasing Kdeg. The relationship between ISIS
22023 concentration in hepatocytes and Fas mRNA levels is described in
the following equation:
|
(3)
|
where Emax is the maximum
effect on Kdeg,
EC50 is the concentration of ISIS 22023 required
for half-maximal stimulation of RNase H, C is the
concentration of ISIS 22023 in hepatocyte,
Kdeg is the Fas mRNA degradation rate
constant, Rin is the synthesis rate of
Fas mRNA, n is the sigmoidicity factor, and mRNA is the normalized Fas mRNA level at time t.
| |
Results |
Pharmacokinetics and Suborgan Kinetics.
After subcutaneous
injection, ISIS 22023 appeared rapidly in plasma and declined
monoexponentially with time (Fig. 2).
Maximum plasma concentration (Cmax) of
intact ISIS 22023 was approximately 118.6 µg/ml and was observed at
0.5 h after subcutaneous administration of 50 mg/kg drug.
Concentrations of ISIS 22023 in plasma decreased rapidly with time and
were below the limit of detection 7 h after dose administration
(<0.07 µg/ml). Similar to other phosphorothioate oligonucleotides,
the disappearance of ISIS 22023 from plasma was the result of
distribution to tissues. The calculated plasma half-life actually
represents a distribution half-life, not the true elimination
half-life. Because phosphorothioate oligonucleotides (MOE
oligonucleotides in particular, vide infra) have long tissue half-lives
and are rapidly cleared from blood, their tissue levels are not
accurately predicted from plasma concentrations. In addition, the assay
method employed was not sensitive enough to quantitate plasma
concentrations in the elimination phase. Thus, the plasma kinetics of
oligonucleotides cannot be used to predict drug tissue levels nor to
understand tissue clearance.
View larger version (11K):
[in this window]
[in a new window]
|
Fig. 2.
Plasma concentration-time profile of ISIS 22023 following subcutaneous administration of a single 50 mg/kg dose of ISIS
22023 to mice. Symbols represent observed concentrations in plasma
(error bars represent standard deviation, n = 3).
The solid line represents predicted concentrations in plasma using
nonlinear regression.
|
|
To characterize tissue kinetics, concentrations of ISIS 22023 in liver
were measured directly. The concentration of ISIS 22023 in liver
reached a maximum 2 days following subcutaneous administration and
decreased slowly with time (Fig. 3).
Using noncompartmental analysis, the half-lives calculated from the
log-linear portion of the concentration versus time curves ranged from
8.8 to 18.8 days over the dose range of 10 to 50 mg/kg. The area under
the concentration-time curve (AUC) for liver increased approximately 4.5-fold over a 5-fold range of doses suggesting linear kinetics, but
Cmax increased less than that
predicted on the basis of dose alone (Table
2). Elimination half-lives from liver
became longer with higher doses, increasing 2-fold from 10 to 50 mg/kg
dose. Taken together, these data suggest that both the uptake and the removal of ISIS 22023 from mouse liver were diminished as dose escalated.
View larger version (17K):
[in this window]
[in a new window]
|
Fig. 3.
ISIS 22023 concentration in whole liver following
subcutaneous administration of a single 10 to 50 mg/kg dose of ISIS
22023 to mice. Symbols represent observed concentrations in liver
(error bars represent standard deviation, n = 3).
The solid lines represent predicted concentrations in liver using
nonlinear regression.
|
|
Assessment of the kinetics of ISIS 22023 in whole liver homogenates
assumes that distribution within the various cells and compartments
within the liver is homogeneous. It is known that Fas expression is
predominantly localized in hepatocytes (Ogasawara et al., 1993), where
the mRNA would be targeted by oligonucleotides. To better understand
the suborgan distribution of MOE oligonucleotides and further identify
the correlation to pharmacodynamics, the pharmacokinetics within
hepatocytes, Kupffer cells, and endothelial cells were studied
following a single 50 mg/kg subcutaneous dose of ISIS 22023.
The time course of ISIS 22023 in different cell types was significantly
different. Several trends are noticeable in the analysis of the
kinetics in the individual cell types (Fig.
4). First, the highest concentration of
ISIS 22023 was observed in endothelial cells, followed by the Kupffer
cells. The concentration of ISIS 22023 observed in hepatocytes was the
lowest. Second, the uptake by Kupffer cells appears to be more delayed
than the uptake by hepatocytes or endothelial cells. As hepatocyte
concentrations diminish the Kupffer cell concentrations appear to
increase. These data may reflect a redistribution of oligonucleotide
between the different cell types in liver. From immunohistochemistry
studies with other oligonucleotides, it is clear that Kupffer cells are a repository for oligonucleotides long after dosing (M. Butler, unpublished data). The data obtained in these studies using
direct quantitative analysis are consistent with that observation.
View larger version (18K):
[in this window]
[in a new window]
|
Fig. 4.
Concentrations of ISIS 22023 in hepatocyte, Kupffer
cells and endothelial cells following single subcutaneous
administration to mice in terms of number of ISIS 22023 molecules per
cell. Error bars represent standard deviation (n = 3).
|
|
To make comparisons with whole liver concentrations of ISIS 22023, concentrations of ISIS 22023 in terms of per cell basis [ISIS 22023 molecules (×106) per cell] was converted to
micrograms of ISIS 22023 per gram of liver basis by eq. 4:
![<UP>ISIS 22023 concentration</UP>(&mgr;<UP>g/g liver</UP>)=[<UP>ISIS 22023 molecules </UP>(<UP>×10<SUP>6</SUP></UP>)/<UP>cell</UP>]×[<UP>number of cells/g liver</UP>]×[<UP>mol. wt.</UP>]×[1,000,000 &mgr;<UP>mol/mol</UP>]/6.02×10<SUP>23</SUP>](/content/vol296/issue2/fulltext/388/img004.gif)
|
(4)
|
where 1 g of mouse liver contains 128 × 106 hepatocytes (Kedderis and Held, 1996; Kawada,
1997) and molecular weight of ISIS 22023 is 7570 g/mol.
Hepatocytes contained the greatest amount of oligonucleotide based on
concentrations per gram liver because of the greater cell population.
The 
-phase half-life of ISIS 22023 in hepatocytes was 4.37 days
(Table 3), which is somewhat more rapid
than the -phase half-life obtained for whole liver, 9.85 days. (The
terminal phase for liver and hepatocytes should be equivalent at
equilibrium.) Thus, the pharmacokinetics of ISIS 22023 in hepatocytes
is not well represented by the kinetics of the whole liver.
View this table:
[in this window]
[in a new window]
|
TABLE 3
ISIS 22023 pharmacokinetics in hepatocyte following a single 50 mg/kg
ISIS 22023 administered subcutaneously
|
|
A two-compartment suborgan pharmacokinetic model (described in Fig. 1)
was attempted with hepatocytes being the central compartment. All
other cell types, connective tissue and interstitium (nonhepatocyte), were assumed to comprise the peripheral compartment. Based on the experimental results, elimination of oligonucleotide from liver may
be a saturable process. Michaelis-Menten elimination was attempted.
However, because of the limited number of observations, the model was
simplified by assuming a first-order elimination from the central
compartment (Table 3).
In Vivo Pharmacodynamics.
The pharmacologic effect of ISIS
22023 occurs in Fas-expressing cells when ISIS 22023 binds to Fas mRNA
and the RNA strand of the heteroduplex is digested by RNase H (Zhang et
al., 2000). Degradation of mRNA is very rapid, occurring within minutes
(Wu et al., 1998). In contrast, the processes of absorption,
distribution, and elimination of oligonucleotides, occur from hours to
days. This difference in rates of action versus rates of delivery
implies that the pharmacologic response of oligonucleotides will be
limited by delivery to the targeted RNA, not enzymatic action, and
predicts a direct correlation between drug concentration at the site
and pharmacologic activity.
After treatment of mice with different doses of ISIS 22023, the
liver Fas mRNA levels fit a typical sigmoidal dose response to the
hepatocyte concentration (Fig. 5A) and to
the whole liver concentration (Fig. 5B). The pharmacodynamic parameters
were estimated according to the inhibitory
Emax model (eq. 2), and the obtained EC50 was 16.1 ± 9.3 µg of ISIS 22023 in
hepatocytes per gram of liver (Table 4).
View larger version (13K):
[in this window]
[in a new window]
|
Fig. 5.
Relationship of Fas mRNA reduction with hepatocyte
ISIS 22023 concentrations (A) and with liver ISIS 22023 concentrations
(B). Symbols represent observed Fas mRNA levels in liver (error bars
represent standard error, n = 3). Solid lines
represent predicted Fas mRNA levels in liver using an inhibitory
sigmoidal Emax model.
|
|
Activation of Fas by agonist Fas monoclonal antibodies in hepatocytes
triggers a proapoptotic signal transduction pathway leading to
mortality from acute liver injuries. Decrease in Fas expression would
be predicted to protect mice from fulminant hepatitis initiated from
agonist Fas monoclonal antibodies. The relationship of Fas mRNA
reduction with mortality induced by agonistic Fas antibody was
characterized using a sigmoidal Emax
model. The more marked the reduction in mRNA expression the lower the
mortality rate. To achieve a 50% reduction of the mortality rate, mRNA
levels have to be maintained at 44% of normal level.
Pharmacokinetic/Pharmacodynamic Relationship following Single and
Multiple Administration.
The integrated
pharmacokinetic/pharmacodynamic model was based on the data from a
single-dose study (eq. 3). The estimated parameters are shown in Table
5. Fas mRNA levels decreased rapidly as
ISIS 22023 concentration increased following subcutaneous
administration. The maximum mRNA reduction was approximately 90% and
occurred 2 days after a single dose of 50 mg/kg (Fig.
6), concurrent with the maximum
concentration of ISIS 22023 in liver and hepatocytes. As the
concentration of ISIS 22023 in hepatocytes declined, appreciable inhibition of Fas mRNA also diminished and Fas mRNA levels began to
rise, presumably, because the expression of new mRNA exceeded the
degradation of old Fas mRNA. By 14 days after dose administration, when
concentrations of ISIS 22023 in hepatocytes were relatively low, on the
order of 10 µg/g of liver, Fas mRNA levels returned to baseline
values. However, total liver concentration was still relatively high,
in the range of 50 µg/g. This observation suggests that total liver
concentration was biased by the relatively high concentrations in the
nonhepatocyte cell fractions at later time points and that hepatocyte
concentrations fell below an active level well before the total burden
of drug in the liver was very much reduced from peak levels (Fig. 6).
When the pharmacokinetic models obtained from fitting data from a
single dose of 50 mg/kg were used in a simulation of multiple dose
administration (Fig. 7A), the simulated
whole liver cleared oligonucleotide more slowly than the simulated
hepatocytes as was predicted from the single dose data. The
concentration minima between doses were more marked for the simulated
hepatocyte data compared with whole liver. When the model was
superimposed on data collected from the multiple dose study, the
typical saw-tooth appearance of multiple dose regimens was observed in
the concentration data from liver and correspondingly, the inverse
pattern was observed for mRNA levels, such that maxima for liver levels
occurred concurrently for minima of mRNA levels (Fig. 7, A and B).
Simulation of the hepatocyte data with its deeper valleys appears to be
in even better concordance with the mRNA data, as was observed in
single dose studies.
View larger version (21K):
[in this window]
[in a new window]
|
Fig. 6.
Pharmacokinetic/pharmacodynamic relationship of ISIS
22023 in mouse liver following subcutaneous administration of a single
50 mg/kg dose of ISIS 22023 to mice. Symbols represent observed
concentrations or Fas mRNA levels (error bars represent standard
deviation for ISIS 22023 concentrations, or standard error for Fas mRNA
levels, n = 3). Solid lines represent predicted
ISIS 22023 concentrations or Fas mRNA levels using nonlinear
regression.
|
|
View larger version (64K):
[in this window]
[in a new window]
|
Fig. 7.
A, pharmacokinetic/pharmacodynamic relationship of
ISIS 22023 in mouse liver following subcutaneous administration of
three 50 mg/kg doses of ISIS 22023 administered every other week to
mice. Symbols represent observed concentrations or Fas mRNA levels
(error bars represent standard deviation for ISIS 22023 concentrations,
or standard error for Fas mRNA levels, n = 3).
Solid lines represent predicted ISIS 22023 concentrations or Fas mRNA
levels using nonlinear regression. B, expression of Fas mRNA and the
reduction in mouse liver by antisense oligonucleotide treatment. Mice
were treated with 50 mg/kg ISIS 22023 for 1, 2, or 3 doses following
subcutaneous administration every other week. Animals were sacrificed
and total RNA extracted from liver at 1, 4, 7, 14, and 28 days after
the last dose. Fas mRNA expression was determined as described under
Materials and Methods.  , ISIS 22023, observed;  ,
mRNA, observed.
|
|
| |
Discussion |
Phosphorothioate oligonucleotides, including those with
MOE-modified termini, are rapidly cleared from plasma and distribute to
tissues where they produce their pharmacologic effects. As a result of
their site of action, their rapid clearance from plasma, and their long
half-lives in tissues, it is difficult to relate pharmacologic activity
to plasma concentrations of oligonucleotide. Although antisense
oligonucleotides represent a novel class of therapeutic agent, their
activity may still be described in conventional pharmacologic terms.
For antisense drugs, the receptor is mRNA, the binding motif is Watson
and Crick base-pairing, and the ligand-receptor interactions result in
destruction of the receptor (mRNA). Studies in this laboratory and
others have characterized the pharmacology of phosphorothioate
oligodeoxynucleotides with MOE modifications, which represent
second-generation chemistries. Structure activity relationships (i.e.,
sequence and mismatched sequence data) have been defined and
dose-response relationships have been characterized. One significant
factor missing was the correlation between oligonucleotide tissue
concentrations and target organ pharmacologic effect.
In this study, we have demonstrated that there is a concordance between
concentrations of oligonucleotide in liver after single and multiple
doses and reductions in Fas mRNA. As concentrations of oligonucleotide
in liver increase, there is a concurrent reduction in Fas mRNA, and as
liver is cleared of oligonucleotide over time, there is an increase in
mRNA. The Fas mRNA levels correlated with the physiological responses
of mice to treatment with an antibody to Fas that binds and activates
Fas to induce apoptosis and ultimately, mortality. These reductions in
mRNA inversely correlated with mortality in antibody challenged mice
with a conventional sigmoidal dose-response relationship. The studies
presented herein went beyond relating organ concentrations to
pharmacologic effect to determine target cell concentrations. A higher
concordance was apparent between the concentrations in hepatocytes,
Fas-expressing cells in liver, and reductions in mRNA. This difference
was due to the fact that hepatocytes cleared oligonucleotide at a
faster rate than other cells in liver (Fig. 4). Thus, the whole liver concentration did not accurately represent hepatocyte concentrations. For example, duration of the reduction in mRNA following single-dose administration was approximately 7 days. Between 7 and 15 days, liver
concentration of oligonucleotide remained high, however, Fas mRNA
levels returned back to normal.
There are several possibilities that can explain this phenomena: 1)
tolerance developed following oligonucleotide treatment, or 2) whole
liver was not the target, and 3) suborgan or cell-specific kinetics
were critical. There does not appear to be tolerance to the drug
treatment based on the responses observed in the repeat administration
experiments. Each administration of ISIS 22023 induced a similar
reduction in mRNA levels, suggesting that there was equivalent response
with each re-administration. The appearance of the reductions in mRNA
with repeated administration also suggests that there is no rebound
phenomenon. The discordance between liver concentrations and response
is most likely the result of differences in suborgan kinetics and drug
binding to noncellular compartments. Examining hepatocyte
concentrations revealed that ISIS 22023 concentration in hepatocytes
dropped significantly between 7 and 15 days. These data support the
concept that there is a concentration differential between whole liver
and the more rapidly clearing hepatocytes, and help explain why mRNA
levels begin to return to normal levels before liver concentrations
have diminished appreciably and suggest that the site of ISIS 22023 action is within the hepatocyte, which is predicted based on expression.
Although, we realize that the data collected here are limited, and may
overparameterize the model, this is the first study to demonstrate the
pharmacokinetic and pharmacodynamic relationships of an antisense drug
in vivo. Based on the estimated parameters for both the kinetic model
and dynamic model, a 24 mg/kg dose given once a week should provide
sufficient protection of the mice subjected to agonistic Fas antibody
challenge. In fact, a recent study showed that using the above
simulation-based regimen, 100% protection was achieved for up to 5 months in the mice receiving ISIS 22023 treatment and subjected to
agonistic Fas antibody challenge (data not shown). These studies
demonstrate that the model has a good predictive power. In conclusion,
this study demonstrated that: 1) concentrations in the target cells are
critical to activity; 2) antisense activity follows a classic
concentration-response relationship; and 3) dose regimens for this
class of drugs like other classes should be designed based on
concentrations at the active site or cell.
Special acknowledgment is due to Patricia DeLeon and Jesse Cook
for excellent technical assistance. We are especially grateful for the
technical review of this manuscript by Stanley Crooke and Frank
Bennett. We also wish to express our gratitude for the help provided by
Karen Keyer in the preparation of this manuscript.
![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
Top
Abstract
Introduction
cancer and inflammation, in
Methods in Molecular Medicine: Antisense Therapeutics (Agrawal S ed) pp 13-46,
Humana Press Inc., Totowa, NJ.
